InteractiveFly: GeneBrief

Glu-RIIA and Glu-RIIB: Biological Overview | Regulation | Developmental Biology | Effects of Mutation | Evolutionary Homologs | References

Gene name - Glutamate receptor IIA
and Glutamate receptor IIB

Synonyms -

Cytological map position - 25F1--2

Function - glutamate-receptor - Ca2+ channel

Keywords - muscle, neuromuscular junction

Symbol - GluRIIA and GluRIIB

FlyBase IDs: FBgn0004620 and FBgn0020429

Genetic map position - 2-[17]

Classification - transmembrane domain protein

Cellular location - surface

NCBI links: GluRIIA - Entrez Gene
NCBI links: GluRIIB - Entrez Gene

GluRIIA orthologs: Biolitmine
GluRIIB orthologs: Biolitmine

Recent literature
Ziegler, A. B., et al. (2016). The amino acid transporter JhI-21 coevolves with glutamate receptors, impacts NMJ physiology, and influences locomotor activity in Drosophila larvae. Sci Rep 6: 19692. PubMed ID: 26805723
Changes in synaptic physiology underlie neuronal network plasticity and behavioral phenomena, which are adjusted during development. The Drosophila larval glutamatergic neuromuscular junction represents a powerful synaptic model to investigate factors impacting these processes. Amino acids such as glutamate have been shown to regulate Drosophila NMJ physiology by modulating the clustering of postsynaptic glutamate receptors and thereby regulating the strength of signal transmission from the motor neuron to the muscle cell. This study used Evolutionary Rate Covariation (ERC), a recently developed bioinformatic tool, to identify amino acid transporters impacting glutmatergic signal transmission. This screen identified ten proteins co-evolving with NMJ glutamate receptors. One candidate transporter, the SLC7 (Solute Carrier) transporter family member JhI-21 (Juvenile hormone Inducible-21), which is expressed in Drosophila larval motor neurons, was selected for further study. JhI-21 was shown to suppress postsynaptic muscle glutamate receptor abundance, and JhI-21 expression in motor neurons regulates larval crawling behavior in a developmental stage-specific manner.
Wang, T., Jones, R. T., Whippen, J. M. and Davis, G. W. (2016). α2δ-3 is required for rapid transsynaptic homeostatic signaling. Cell Rep 16: 2875-2888. PubMed ID: 27626659
The homeostatic modulation of neurotransmitter release, termed presynaptic homeostatic potentiation (PHP), is a fundamental type of neuromodulation, conserved from Drosophila to humans, that stabilizes information transfer at synaptic connections throughout the nervous system. This study demonstrates that α2δ-3 (straitjacket), an auxiliary subunit of the presynaptic calcium channel, is required for PHP. The α2δ gene family has been linked to chronic pain, epilepsy, autism, and the action of two psychiatric drugs: gabapentin and pregabalin. Loss of α2δ-3 blocks both the rapid induction and sustained expression of PHP due to a failure to potentiate presynaptic calcium influx and the RIM-dependent readily releasable vesicle pool. These deficits are independent of α2δ-3-mediated regulation of baseline calcium influx and presynaptic action potential waveform. α2δ proteins reside at the extracellular face of presynaptic release sites throughout the nervous system, a site ideal for mediating rapid, transsynaptic homeostatic signaling in health and disease.
Wang, M., Chen, P. Y., Wang, C. H., Lai, T. T., Tsai, P. I., Cheng, Y. J., Kao, H. H. and Chien, C. T. (2016). Dbo/Henji modulates synaptic dPAK to gate glutamate receptor abundance and postsynaptic response. PLoS Genet 12: e1006362. PubMed ID: 27736876
In response to environmental and physiological changes, the synapse manifests plasticity while simultaneously maintains homeostasis. This study analyzed mutant synapses of henji, also known as dbo, at the Drosophila neuromuscular junction (NMJ). In henji mutants, NMJ growth is defective with appearance of satellite boutons. Transmission electron microscopy analysis indicates that the synaptic membrane region is expanded. The postsynaptic density (PSD) houses glutamate receptors GluRIIA and GluRIIB, which have distinct transmission properties. In henji mutants, GluRIIA abundance is upregulated but that of GluRIIB is not. Electrophysiological results also support a GluR compositional shift towards a higher IIA/IIB ratio at henji NMJs. Strikingly, dPAK, a positive regulator for GluRIIA synaptic localization, accumulates at the henji PSD. Reducing the dpak gene dosage suppresses satellite boutons and GluRIIA accumulation at henji NMJs. In addition, dPAK associated with Henji through the Kelch repeats which is the domain essential for Henji localization and function at postsynapses. It is proposed that Henji acts at postsynapses to restrict both presynaptic bouton growth and postsynaptic GluRIIA abundance by modulating dPAK.
Goel, P., Li, X. and Dickman, D. (2017). Disparate postsynaptic induction mechanisms ultimately converge to drive the retrograde enhancement of presynaptic efficacy. Cell Rep 21(9): 2339-2347. PubMed ID: 29186673
Retrograde signaling systems are fundamental modes of communication synapses utilize to dynamically and adaptively modulate activity. However, the inductive mechanisms that gate retrograde communication in the postsynaptic compartment remain enigmatic. This study investigated retrograde signaling at the Drosophila neuromuscular junction, where three seemingly disparate perturbations to the postsynaptic cell trigger a similar enhancement in presynaptic neurotransmitter release. This study shows that the same presynaptic genetic machinery and enhancements in active zone structure are utilized by each inductive pathway. However, all three induction mechanisms differ in temporal, translational, and CamKII activity requirements to initiate retrograde signaling in the postsynaptic cell. Intriguingly, pharmacological blockade of postsynaptic glutamate receptors, and not calcium influx through these receptors, is necessary and sufficient to induce rapid retrograde homeostatic signaling through CamKII. Thus, three distinct induction mechanisms converge on the same retrograde signaling system to drive the homeostatic strengthening of presynaptic neurotransmitter release.
Patron, L. A., Nagatomo, K., Eves, D. T., Imad, M., Young, K., Torvund, M., Guo, X., Rogers, G. C. and Zinsmaier, K. E. (2019). Cul4 ubiquitin ligase cofactor DCAF12 promotes neurotransmitter release and homeostatic plasticity. J Cell Biol. PubMed ID: 30670470
This study genetically characterized the synaptic role of the Drosophila homologue of human DCAF12, a putative cofactor of Cullin4 (Cul4) ubiquitin ligase complexes. Deletion of Drosophila DCAF12 impairs larval locomotion and arrests development. At larval neuromuscular junctions (NMJs), DCAF12 is expressed presynaptically in synaptic boutons, axons, and nuclei of motor neurons. Postsynaptically, DCAF12 is expressed in muscle nuclei and facilitates Cul4-dependent ubiquitination. Genetic experiments identified several mechanistically independent functions of DCAF12 at larval NMJs. First, presynaptic DCAF12 promotes evoked neurotransmitter release. Second, postsynaptic DCAF12 negatively controls the synaptic levels of the glutamate receptor subunits GluRIIA, GluRIIC, and GluRIID. The down-regulation of synaptic GluRIIA subunits by nuclear DCAF12 requires Cul4. Third, presynaptic DCAF12 is required for the expression of synaptic homeostatic potentiation. It is suggested that DCAF12 and Cul4 are critical for normal synaptic function and plasticity at larval NMJs.
Lecompte, M., Cattaert, D., Vincent, A., Birman, S. and Cherif-Zahar, B. (2019). The Drosophila ammonium transporter Rh50 is required for integrity of larval muscles and neuromuscular system. J Comp Neurol. PubMed ID: 31273786
Rhesus glycoproteins (Rh50) have been shown to be ammonia transporters in many species from bacteria to human. They are involved in various physiological processes including acid excretion and pH regulation. Rh50 proteins can also provide a structural link between the cytoskeleton and plasma membranes that maintains cellular integrity. Although ammonia plays essential roles in the nervous system, in particular at glutamatergic synapses, a potential role for Rh50 proteins at synapses has not yet been investigated. To better understand the function of these proteins in vivo, the unique Rh50 gene of Drosophila melanogaster which encodes two isoforms, Rh50A and Rh50BC, was studied. Drosophila Rh50A is expressed in larval muscles and enriched in the postsynaptic regions of the glutamatergic neuromuscular junctions (NMJs). Rh50 inactivation by RNA interference (RNAi) selectively in muscle cells caused muscular atrophy in larval stages and pupal lethality. Interestingly, Rh50-deficiency in muscles specifically increased glutamate receptor subunit IIA (GluRIIA) level and the frequency of spontaneous excitatory postsynaptic potentials (EPSPs). This work therefore highlights a new role for Rh50 proteins in the maintenance of Drosophila muscle architecture and synaptic physiology which could be conserved in other species.
Goel, P., Nishimura, S., Chetlapalli, K., Li, X., Chen, C. and Dickman, D. (2020). Distinct Target-Specific Mechanisms Homeostatically Stabilize Transmission at Pre- and Post-synaptic Compartments. Front Cell Neurosci 14: 196. PubMed ID: 32676010
Neurons must establish and stabilize connections made with diverse targets, each with distinct demands and functional characteristics. At Drosophila neuromuscular junctions (NMJs), synaptic strength remains stable in a manipulation that simultaneously induces hypo-innervation on one target and hyper-innervation on the other. However, the expression mechanisms that achieve this exquisite target-specific homeostatic control remain enigmatic. This study identified the distinct target-specific homeostatic expression mechanisms. On the hypo-innervated target, an increase in postsynaptic glutamate receptor (GluR) abundance is sufficient to compensate for reduced innervation, without any apparent presynaptic adaptations. In contrast, a target-specific reduction in presynaptic neurotransmitter release probability is reflected by a decrease in active zone components restricted to terminals of hyper-innervated targets. Finally, loss of postsynaptic GluRs on one target induces a compartmentalized, homeostatic enhancement of presynaptic neurotransmitter release called presynaptic homeostatic potentiation (PHP) that can be precisely balanced with the adaptations required for both hypo- and hyper-innervation to maintain stable synaptic strength. Thus, distinct anterograde and retrograde signaling systems operate at pre- and post-synaptic compartments to enable target-specific, homeostatic control of neurotransmission.
Basargekar, A., Yogi, S., Mushtaq, Z., Deivasigamani, S., Kumar, V., Ratnaparkhi, G. S. and Ratnaparkhi, A. (2020). Drosophila Mon1 and Rab7 interact to regulate glutamate receptor levels at the neuromuscular junction. Int J Dev Biol 64(4-5-6): 299-307. PubMed ID: 32658990
Regulation of post-synaptic receptors plays an important role in determining synaptic strength and plasticity. The Drosophila larval neuromuscular junction (nmj) has been used extensively as a model to understand some of these processes. In this context, this study explored in the role of Drosophila Monensin sensitivity protein 1 (DMon1) in regulating glutamate receptor (GluRIIA) levels at the nmj. DMon1 is an evolutionarily conserved protein which, in complex with calcium caffeine zinc sensitivity1 (CCZ1), regulates the conversion of early endosomes to late endosomes through recruitment of Rab7. C-terminal deletion mutants of Dmon1 (Dmon1(Δ181)) exhibit lethality. The escapers have a short life span and exhibit severe motor defects. At the nmj, these mutants show defects in synaptic morphology and a strong increase in GluRIIA levels. The mechanism by which Dmon1 regulates GluRIIA is unclear. This study has characterized an EMS mutant referred to as pog(1) and demonstrates it to be an allele of Dmon1. Further, this study has examined the role of rab7 in regulating GluRIIA. Similar to Dmon1, knock-down of rab7 using RNAi in neurons, but not muscles, leads to an increase in GluRIIA. Loss of one copy each of Dmon1 and rab7 leads to a synergistic increase in receptor expression. Further, overexpression of an activated Rab7 can rescue the GluRIIA phenotype observed in Dmon1 (Δ181) mutants. Together, these results highlight a neuronal role for Rab7 in GluRIIA regulation and underscore the importance of the endo-lysosomal pathway in this process.
Yu, Y., Niccoli, T., Ren, Z., Woodling, N. S., Aleyakpo, B., Szabadkai, G. and Partridge, L. (2020). PICALM rescues glutamatergic neurotransmission, behavioural function, and survival in a Drosophila model of Aβ42 toxicity. Hum Mol Genet. PubMed ID: 32592479
Alzheimer's disease (AD) is the most common form of dementia and the most prevalent neurodegenerative disease. Genome Wide Association Studies have linked PICALM to AD risk. PICALM has been implicated in Aβ42 production and turn-over, but whether it plays a direct role in modulating Aβ42 toxicity remains unclear. This study found that increased expression of the Drosophila PICALM orthologue like-AP180 (lap) could rescue Aβ42 toxicity in an adult-onset model of AD, without affecting Aβ42 level. Imbalances in the glutamatergic system, leading to excessive, toxic stimulation have been associated with AD. This study found that Aβ42 caused accumulation of pre-synaptic vesicular glutamate transporter (VGlut) and increased spontaneous glutamate release. Increased lap expression reversed these phenotypes back to control levels, suggesting that lap may modulate glutamatergic transmission. This study also found that lap modulated the localisation of Amph, the homologue of another AD risk factor BIN1, and that Amph itself modulated post-synaptic glutamate receptor (GluRII) localisation. A model is proposed where PICALM modulates glutamatergic transmission, together with BIN1, to ameliorate synaptic dysfunction and disease progression.
Metwally, E., Zhao, G., Wang, Q. and Zhang, Y. Q. (2020). Ttm50 facilitates calpain activation by anchoring it to calcium stores and increasing its sensitivity to calcium. Cell Res. PubMed ID: 32848200
Calcium-dependent proteolytic calpains are implicated in a variety of physiological processes, as well as pathologies associated with calcium overload. However, the mechanism by which calpain is activated remains elusive since intracellular calcium levels under physiological conditions do not reach the high concentration range required to trigger calpain activation. From a candidate screening using the abundance of the calpain target glutamate receptor GluRIIA at the Drosophila neuromuscular junction as a readout, this study uncovered that calpain activity was inhibited upon knockdown of Ttm50, a subunit of the Tim23 complex known to be involved in the import of proteins across the mitochondrial inner membrane. Unexpectedly, Ttm50 and calpain are co-localized at calcium stores Golgi and endoplasmic reticulum (ER), and Ttm50 interacts with calpain via its C-terminal domain. This interaction is required for calpain localization at Golgi/ER, and increases calcium sensitivity of calpain by roughly an order of magnitude. These findings reveal the regulation of calpain activation by Ttm50, and shed new light on calpain-associated pathologies.
Dannhauser, S., Mrestani, A., Gundelach, F., Pauli, M., Komma, F., Kollmannsberger, P., Sauer, M., Heckmann, M. and Paul, M. M. (2022). Endogenous tagging of Unc-13 reveals nanoscale reorganization at active zones during presynaptic homeostatic potentiation. Front Cell Neurosci 16: 1074304. PubMed ID: 36589286
Neurotransmitter release at presynaptic active zones (AZs) requires concerted protein interactions within a dense 3D nano-hemisphere. Among the complex protein meshwork the (M)unc-13 family member Unc-13 of Drosophila melanogaster is essential for docking of synaptic vesicles and transmitter release. This study employed minos-mediated integration cassette (MiMIC)-based gene editing using GFSTF (EGFP-FlAsH-StrepII-TEV-3xFlag) to endogenously tag all annotated Drosophila Unc-13 isoforms enabling visualization of endogenous Unc-13 expression within the central and peripheral nervous system. Electrophysiological characterization using two-electrode voltage clamp (TEVC) reveals that evoked and spontaneous synaptic transmission remain unaffected in unc-13 (GFSTF) 3rd instar larvae and acute presynaptic homeostatic potentiation (PHP) can be induced at control levels. Furthermore, multi-color structured-illumination shows precise co-localization of Unc-13(GFSTF), Bruchpilot, and GluRIIA-receptor subunits within the synaptic mesoscale. Localization microscopy in combination with HDBSCAN algorithms detect Unc-13(GFSTF) subclusters that move toward the AZ center during PHP with unaltered Unc-13(GFSTF) protein levels.
Mallik, B., Brusich, D. J., Heyrman, G. and Frank, C. A. (2023). Precise mapping of one classic and three novel GluRIIA mutants in Drosophila melanogaster. MicroPubl Biol 2023. PubMed ID: 37334199
Mutation of the Drosophila melanogaster GluRIIA gene or pharmacological agents targeting it are commonly used to assess homeostatic synaptic function at the larval neuromuscular junction (NMJ). The commonly used mutation, GluRIIA (SP16), is a null allele created by a large and imprecise excision of a P-element which affects GluRIIA and multiple upstream genes. This study mapped the exact bounds of the GluRIIA (SP16) allele, refined a multiplex PCR strategy for positive identification of GluRIIA (SP16) in homozygous or heterozygous backgrounds, and sequenced and characterized three new CRISPR-generated GluRIIA mutants. The three new GluRIIA alleles are apparent nulls that lack GluRIIA immunofluorescence signal at the 3 (rd) instar larval NMJ and are predicted to cause premature truncations at the genetic level. Further, these new mutants have similar electrophysiological outcomes as GluRIIA (SP16), including reduced miniature excitatory postsynaptic potential (mEPSP) amplitude and frequency compared to controls, and they express robust homeostatic compensation as evidenced by normal excitatory postsynaptic potential (EPSP) amplitude and elevated quantal content. These findings and new tools extend the capacity of the D. melanogaster NMJ for assessment of synaptic function.

Four ionotropic glutamate receptors had been identified in Drosophila: DGluRI, a kainate-type receptor expressed in the CNS (Ultsch, 1992); DNMDAR, an NMDA-like receptor expressed in brain (Ultsch, 1993), and DGluRIIA and DGluRIIB, the subjects of this overview. DGluRIIA and DGluRIIB are closely linked genes encoding muscle-specific glutamate receptors (Schuster, 1991 and Petersen, 1997). Receptors are classified by whether or not they serve as ion channels, and by the type of chemical that serves to stimulate receptor activity. The two DGluRII receptors are ionotropic receptors, referring to the fact that they serve as ion channels. Although they are classified as non-NMDA types, they cannot be classified as either an AMPA or kainate type. These two genes are adjacent in the genome, have similar genomic organization, and are more closely related to each other than to any other glutamate receptors. Nonetheless, DGluRIIA and DGluRIIB share only 44% amino acid identity. The sequence of DGluRIIA is identical to vertebrate channels in the putative pore region, an area that is critical for Ca2+ permeability, while the sequence of DGluRIIB is divergent. These genes are expressed in all somatic muscles and are excluded from the nervous system (Currie, 1995 and Petersen, 1997). Such a configuration is known as postsynaptic, referring to the location of the glutamate channel on muscular synaptic boutons (Saitoe, 1997 and Petersen, 1997) targeted by neurons. By analogy with other channel proteins, the ligand-gated glutamate channels are likely to be multimeric assemblies of individual subunits (Peterson, 1997).

The identification of the existence of an unknown retrograde signal at the Drosophila NMJ leaves a number of open questions. (1) What is being sensed by the muscle that initiates the generation of this signal? The muscle could respond to synaptic depolarization, or it could be sensitive to a second messenger that is regulated by glutamate receptor function, such as calcium influx. (2) What is the presynaptic target of the postsynaptic signal? No sprouting of synaptic boutons or change in the calcium dependence of transmitter release is observed, suggesting that the increase in transmitter release is not secondary to gross structural plasticity or to a change in the function of the calcium sensor. The increase in presynaptic release may be due to an increase in calcium influx into the presynaptic terminal or to changes in the function of the release machinery. (3) What is the nature of the retrograde signal initiated by activity in the muscle? In Drosophila, unlike in vertebrates, each muscle is not regulated by a sensory neuron-to-motor neuron circuit, so the mechanism is unlikely to be cellular. Precedent exists for diffusible signals such as nitric oxide and arachidonic acid to function as retrograde signals for synaptic plasticity. Since the presynaptic and postsynaptic cell are in tight apposition throughout development, the signal could also involve membrane-bound molecules. These questions will be the subject of future genetic and physiological analyses (Peterson, 1997).

A retrograde signal from muscle to motoneuron could influence presynaptic release by increasing presynaptic structure. Bouton number was quantified for the junction at muscles 6 and 7 in abdominal segment A3. There was no change in bouton number when comparing wild type with larvae expressing increased PKA in muscle. Similarly, there is no change in presynaptic bouton number in larvae in which PKA is inhibited in muscle. Thus, a retrograde signal most likely regulates either the number of presynaptic active zones present in each presynaptic bouton or regulates some aspect of the presynaptic release mechanism (Davis, 1998).

Drosophila iGluRs Are Not AMPA or Kainate Receptors

AMPA, kainate, and NMDA, the canonical ligands used to classify vertebrate iGluRs, were applied to Xenopus oocytes to test cell responses to those evoked by glutamate, aspartate, and quisqualate. All agonists were applied at a concentration of 3 mM, close to the EC50 for glutamate, with substitution of extracellular Ca2+ by 0.8 mM Ba2+ used to prevent activation of TMEM16A channels. For GluRIIA/C/D/E responses to glutamate and quisqualate, responses were of similar amplitude, whereas for AMPA, kainate, aspartate, and NMDA there was no detectable response. A similar profile was obtained for GluRIIB/C/D/E, with glutamate and quisqualate responses, and there was no detectable response for AMPA, kainate, aspartate, and NMDA (Han, 2005).

To investigate the structural basis for this unique profile ligand binding domain (LBD) S1S2 constructs of the five Drosophila NMJ iGluRs were screened for expression as soluble proteins in Escherichia coli, and GluRIIB was identified as a promising candidate for crystallization. X-ray diffraction data for the GluRIIB S1S2 complex with glutamate, at a resolution of 2 Å, revealed the classic back-to-back LBD dimer assembly, as first reported for the GluA2 AMPA receptor. In both subunits glutamate was bound in a cavity of volume 208 Å3 together with three trapped water molecules. The glutamate ligand α-carboxyl and α-amino groups make ion pair and hydrogen bond contacts with conserved arginine and glutamate residues, identical to the binding mechanism for AMPA and kainate receptors, with the γ-carboxyl group forming a series of solvent mediated interactions with main-chain and side-chain groups in domain 2. The cavity volume for GluRIIB is similar to that for GluA2 (218 Å3), which binds both AMPA and kainate, as well as quisqualate, but smaller than that for GluK1, GluK2, and GluK3 volumea, suggesting that structural features unique to Drosophila NMJ iGluRs control ligand selectivity (Han, 2005).

To gain insight into why quisqualate but not AMPA or kainate can activate Drosophila NMJ iGluRs, crystal structures were were superimposed for vertebrate GluA2 and GluK2 LBD complexes with these ligands on the GluRIIB LBD crystal structure. This process revealed that, as is the case for GluA2 and GluK2, the quisqualate ligand is easily accommodated in the GluRIIB binding site by displacement of water molecule W1. Within domain 1 of the GluRIIB LBD structure, in the loop between β-strand 7 and α-helix D, the side chain of Asp509 forms a hydrogen bond with the hydroxyl group of Tyr481, a conserved aromatic residue that caps the entrance to the ligand binding cavity, sealing it from extracellular solvent. Stacked above Tyr481, the side chain of Arg429 forms a cation π interaction with the aromatic ring, further stabilizing the conformation of Tyr481. Amino acid sequence alignments reveal that Asp509 is conserved in all Drosophila NMJ iGluRs, whereas in all vertebrate AMPA and kainate receptor subunits there is a proline at this position; similarly, cation π stacking by Arg429 is unique to GluRIIA, GluRIIB, and GluRIIC, because vertebrate AMPA and kainate receptor subunits have an Ile residue at this position. As a result, because of the different conformation of the isoxazazole group, AMPA is unable to bind to GluRIIB because the ligand 5-methyl group makes steric clashes with Asp509 and Asn736. Similarly, although the ligand α-carboxyl, α-amino, and γ-carboxyl groups of kainate are isosteric with those of glutamate, the isopropenyl group makes steric clashes with the Asp509 and Tyr481 side chains (Han, 2015).

A novel, noncanonical BMP pathway modulates synapse maturation at the Drosophila neuromuscular junction

At the Drosophila NMJ, BMP signaling is critical for synapse growth and homeostasis. Signaling by the BMP7 homolog, Gbb, in motor neurons triggers a canonical pathway-which modulates transcription of BMP target genes, and a noncanonical pathway-which connects local BMP/BMP receptor complexes with the cytoskeleton. This study describes a novel noncanonical BMP pathway characterized by the accumulation of the pathway effector, the phosphorylated Smad (pMad), at synaptic sites. Using genetic epistasis, histology, super resolution microscopy, and electrophysiology approaches, it was demonstrated that this novel pathway is genetically distinguishable from all other known BMP signaling cascades. This novel pathway does not require Gbb, but depends on presynaptic BMP receptors and specific postsynaptic glutamate receptor subtypes, the type-A receptors. Synaptic pMad is coordinated to BMP's role in the transcriptional control of target genes by shared pathway components, but it has no role in the regulation of NMJ growth. Instead, selective disruption of presynaptic pMad accumulation reduces the postsynaptic levels of type-A receptors, revealing a positive feedback loop which appears to function to stabilize active type-A receptors at synaptic sites. Thus, BMP pathway may monitor synapse activity then function to adjust synapse growth and maturation during development (Sulkowski, 2016).

BMPs fulfill multiple functions during NMJ development via canonical and noncanonical pathways. In motor neurons, signaling by Gbb triggers a canonical BMP signaling that regulates transcription of BMP target genes and a noncanonical BMP pathway that connects Wit with LIMK1 and the cytoskeleton. This study describes a novel non-canonical BMP pathway, which induces selective accumulation of pMad at presynaptic sites. This pathway does not require Gbb, but depends on presynaptic BMP receptors Wit and Sax and postsynaptic GluRIIA. This novel pathway does not contribute to the NMJ growth and instead appears to set up a positive feedback loop that modulates the postsynaptic distribution of type-A and type-B receptors as a function of synapse activity (Sulkowski, 2016).

At the Drosophila NMJ, BMP signaling controls NMJ growth and promotes synapse homeostasis. BMP fulfills all these functions via canonical and noncanonical pathways. Canonical BMP signaling activates presynaptic transcriptional programs with distinct roles in the structural and functional development of the NMJ. For example, the BMP pathway effector Trio can rescue NMJ growth in BMP pathway mutants, but does not influence synapse physiology, whereas Target of Wit (Twit) can partially restore the mini frequency but has no effect on NMJ growth. It has been shown that both muscle and neuron derived Gbb are required for the structural and functional integrity of NMJ, and multiple mechanisms that regulate Gbb expression, secretion and extracellular availability have been described. Binding of Gbb to its receptors also triggers a noncanonical, Mad-independent pathway that requires the C-terminal domain of Wit. This domain is conserved among Drosophila Wit and vertebrate BMPRII and functions to recruit and activate cytoskeletal regulators such as LIMK1. In flies, Wit-mediated activation of LIMK1 mediates synapse stability and enables rapid, activity-dependent synaptic growth (Eaton, 2005; Piccioli, 2014; Sulkowski, 2016 and references therein).

This study uncovered a novel, noncanonical BMP pathway that triggers accumulation of presynaptic pMad in response to postsynaptic GluRIIA receptors. This pathway requires Wit and Sax, suggesting that various BMP pathways compete for shared components. Super resolution imaging mapped the pMad domains at active zones, in close proximity to the presynaptic membrane. These domains concentrate the pMad immunoreactivities into thin discs that reside mostly within individual synapse boundaries. The size and shape of pMad domains suggest that pMad could associate with membrane-anchored complexes at the active zone. Since BMP signals are generally short lived, these pMad domains likely represent pMad that, upon phosphorylation, remains associated with the BMP/BMPR kinase complexes at synaptic sites. Alternatively, pMad may accumulate in synaptic aggregates that protect it from dephosphorylation. While the second possibility cannot be excluded, several lines of evidence support the first scenario. First, excess Mad cannot increase the levels of synaptic pMad. Second, neuronal expression of activated Tkv/Sax but not Mad can restore the synaptic pMad at Importin impβ11 mutant NMJs. Finally, during neural tube closure, junctional pSmad1/5/8 and its association with PAR complexes depend on BMPs. Previous studies indicate a reduction of synaptic pMad signals in response to muscle-specific Mad RNAi. This study too has observed such a reduction. In addition, this study found a significant decrease of postsynaptic IIA/IIB ratio in Mad-depleted muscles: GluRIIA and GluRIIB synaptic levels were reduced to 49% and respectively 78% of control. Since GluRIIA is key to the synaptic pMad accumulation it is suspected that the muscle Mad RNAi phenotype is due to perturbation in synaptic GluRIIA levels, perhaps by interference with the Activin signaling pathway (Sulkowski, 2016).

How are the BMP/BMPR complexes stabilized at synaptic sites? Studies on single receptors demonstrate that the confined mobility of BMPRI on the plasma membrane is key to stabilize BMP/BMPR complexes and differentially stimulate canonical versus noncanonical signaling. Direct interactions between phosphorylated Smad5 and the Par3-Par6-aPKC polarity complex occur at the apical junctions. Similarly, synaptic pMad, which remains associated with BMP/BMPR complexes, may engage in interactions that restrict the mobility of BMP/BMPR complexes on the presynaptic membrane. Nemo-mediated phosphorylation of Mad-S25 could disrupt the pMad/BMPR association and expose the BMP/BMPR complexes, so they could dissociate and/or be internalized. The heteromeric BMPR complexes are transient; ligand binding greatly increases their lifespan and stability. Albeit Gbb is not essential for synaptic pMad, it may act redundantly with other ligands to stabilize BMP/BMPR local complexes. Several ligands secreted in the synaptic cleft have been shown to bind and signal via BMPRII; they include glia secreted Maverik, Myoglianin, which could be secreted from muscle and/or glia, and Activins. However, these ligands also appear to signal via a canonical Activin pathway, which regulates the postsynaptic GluRIIA/GluRIIB abundance at the Drosophila NMJ. Alterations in the Activin signaling pathway drastically alter the synaptic recruitment of both iGluR subtypes, in particular the GluRIIA, which controls synaptic pMad, making it difficult to identify the nature and the directionality of the signaling molecule(s) involved in the synaptic pMad accumulation. Interestingly, all of these ligands are substrates for BMP-1/Tolloid-type enzymes, which control their activity and distribution. Treatments that induce long-term stimulation up-regulate a BMP-1/Tolloid homolog in Aplysia neurons (Sulkowski, 2016).

An intriguing aspect of this novel BMP pathway is the dependence on active postsynaptic GluRIIA, which is both required and sufficient for pMad accumulation at active zones. Since pMad and BMP/BMPR complexes cluster at synaptic sites, it is speculated that trans-synaptic complexes may couple postsynaptic type-A glutamate receptors with presynaptic BMP/BMPRs. The synaptic cleft is 200 Å; the iGluR tetramer expands 135 Å in the synaptic cleft, and the BMP/BMPR complexes ~55 Å. The iGluRs auxiliary subunit Neto has extracellular CUB and LDLa domains predicted to expand 120-130 Å in the synaptic cleft, based on related structures. CUB domains are BMP binding motifs that may localize BMP activities and/or facilitate ligand binding to BMPRs. In this model, Neto provides the link between postsynaptic GluRIIA and presynaptic BMP/BMPR complexes. During receptors gating cycle, the iGluRs undergo corkscrew motions that shorten the channels as revealed by cryo-electron microscopy. Such movements may influence the stability of trans-synaptic complexes and allow synaptic pMad to function as a sensor for GluRIIA activity (Sulkowski, 2016).

While more components of this novel pathway remain to be determined, it is clear that this pathway does not contribute to NMJ growth and instead has a critical role in synapse maturation. Unlike canonical BMP signaling, loss of local pMad induces minor reductions in bouton number and does not rescue the NMJ overgrowth of endocytosis mutants. Local pMad accumulates independently of Wit-mediated LIMK1 activation and does not appear to influence synapse stabilization; in fact, nrx mutants have synapse adhesion defects but show increased synaptic pMad levels. The striking correlation between synaptic pMad levels and GluRIIA activity, together with previous findings that GluRIIA activity and gating behavior directly impacts receptor mobility and synaptic stabilization suggest a positive feedback mechanism in which active GluRIIA receptors induce stabilization of BMP/BMPR complexes at synaptic sites which, in turn, promote stabilization of type-A receptors at PSDs. In this scenario, presynaptic pMad marks active BMP/BMPR complexes and acts to maintain the local BMP/BMPR complexes in large clusters that evade endocytosis. Selective disruption of local pMad via a neuronal dominant-negative MadS25D presumably destabilizes the large presynaptic BMP/BMPR clusters and causes a significant reduction in the IIA/IIB ratio and quantal size (Sulkowski, 2016).

This positive feedback couples synaptic activity with synapse development and is controlled by (1) active GluRIIA receptors, (2) presynaptic BMP receptors, Wit, Sax, and likely Tkv, (3) mechanisms regulating BMPR heteromers assembly, endocytosis and turnover, and (4) the ability of pMad to remain associated with its own kinase upon phosphorylation. Perturbations of any of these components trigger variations in local pMad levels accompanied by changes in the IIA/IIB ratio and/or quantal size. For example, nemo mutants have increased synaptic pMad levels and increased mEJCs, while imp mutants have decreased synaptic pMad levels and decreased mEJPs. The assembly and function of these putative trans-synaptic complexes, in particular ligand availability, should be influenced by the composition and organization of the synaptic cleft. Indeed, local pMad and quantal size are increased in mutants lacking heparan sulfate 6-O-endosulfatase (sulf1), or 6-O-sulfotransferase (hs6st). Since this Mad-dependent, noncanonical pathway shares components with the other BMP signaling pathways, the balance among different BMP pathways may coordinate the NMJ development and function (Sulkowski, 2016).

The complexity of BMP signaling at the Drosophila NMJ is reminiscent of the neurotrophin-regulated signaling in vertebrate systems. Neurotrophins were first identified as neuronal survival factors. Like BMPs, they are secreted as pro-proteins that must be processed to form mature ligands. The active dimers bind to transmembrane kinase receptors and induce their activation through trans-phosphorylation. Neurotrophin/receptor complexes are internalized and transported along axons to the cell soma; signaling in the cell soma controls gene expression and promotes neuronal differentiation and growth. In addition, local neurotrophin signaling regulates growth cone motility, enhances the presynaptic release of neurotransmitter and mediates activity-dependent synapse formation and maturation. At the Drosophila NMJ, several neurotrophins have been implicated in neuron survival, axon guidance and synapse growth. It will be interesting to test for the crosstalk between neurotrophin and BMP signaling at these synapses (Sulkowski, 2016).

The novel noncanonical BMP pathway reported in this study is the first example of a BMP pathway triggered by selective neurotransmitter receptors and influencing receptor distribution at PSDs. It is expected that some of these functions will apply to mammalian glutamatergic synapses: First, as indicated in the Allen Brain Atlas, glutamate receptors and Neto proteins are widely expressed in mammalian brain structures where BMPs, BMPRs and Smads are expressed. Second, BMPs have been shown to rapidly potentiate glutamate-mediated currents in human retina neurons, presumably via a noncanonical pathway. Finally, mice lacking Chordin, a BMP antagonist, have enhanced paired-pulse facilitation and LTP and show improved learning in a water maze test. Such changes could not be explained by Smad-dependent transcriptional responses and were not accompanied by structural alterations in synapse morphology. Instead, presynaptic noncanonical BMP pathway may influence the activity of postsynaptic glutamate receptors by modulating their synaptic distribution and stability (Sulkowski, 2016).

Disparate postsynaptic induction mechanisms ultimately converge to drive the retrograde enhancement of presynaptic efficacy

Retrograde signaling systems are fundamental modes of communication synapses utilize to dynamically and adaptively modulate activity. However, the inductive mechanisms that gate retrograde communication in the postsynaptic compartment remain enigmatic. This study investigated retrograde signaling at the Drosophila neuromuscular junction, where three seemingly disparate perturbations to the postsynaptic cell trigger a similar enhancement in presynaptic neurotransmitter release. This study shows that the same presynaptic genetic machinery and enhancements in active zone structure are utilized by each inductive pathway. However, all three induction mechanisms differ in temporal, translational, and CamKII activity requirements to initiate retrograde signaling in the postsynaptic cell. Intriguingly, pharmacological blockade of postsynaptic glutamate receptors, and not calcium influx through these receptors, is necessary and sufficient to induce rapid retrograde homeostatic signaling through CamKII. Thus, three distinct induction mechanisms converge on the same retrograde signaling system to drive the homeostatic strengthening of presynaptic neurotransmitter release (Goel, 2017).

The Drosophila neuromuscular junction (NMJ) is an established system to study retrograde synaptic signaling. At this model glutamatergic synapse, genetic or pharmacological perturbations to postsynaptic receptor functionality trigger retrograde signaling that instructs the neuron to precisely increase presynaptic neurotransmitter release, maintaining stable levels of synaptic strength. This process is termed presynaptic homeostatic potentiation (PHP) and can be induced through two distinct disruptions to postsynaptic glutamate receptor functionality. First, acute pharmacological blockade of receptors by application of philanthotoxin-433 (PhTx) reduces miniature excitatory postsynaptic potential (mEPSP) amplitude, initiating rapid expression of PHP (increase in quantal content) within 10 min. Second, genetic loss of the postsynaptic glutamate receptor subunit GluRIIA leads to a similar reduction in mEPSP amplitudes over chronic timescales (days) and a similar expression of PHP. Although these perturbations each disrupt receptors and lead to adaptive increases in presynaptic neurotransmitter release, PhTx- and GluRIIA-mediated PHP signaling exhibit important differences. First, some genes have been identified that are only necessary for GluRIIA-dependent PHP expression, whereas PHP is robustly expressed following acute PhTx application in larvae with mutations in these genes (Frank, 2009, Kauwe, 2016, Penney, 2016, Spring, 2016, Tsurudome, 2010). In addition, PhTx-induced PHP expression is translation-independent (Frank, 2006), whereas GluRIIA-induced PHP is blocked by inhibitions to postsynaptic translation through loss of the translational regulator target of rapamycin (Tor) (Kauwe, 2016, Penney, 2012). Although several genes and mechanisms necessary for the expression of PHP in the presynaptic neuron have been identified, far less is known about the mechanistic differences in postsynaptic transduction between PhTx- and GluRIIA-induced PHP signaling (Goel, 2017).

Recently, a novel manipulation to the postsynaptic muscle that does not affect glutamate receptors was demonstrated to induce retrograde PHP signaling at the Drosophila NMJ. This was accomplished by postsynaptic overexpression of the non-specific translational regulator Tor (Tor-OE) (Penney, 2012), which leads to a chronic and global increase in muscle protein synthesis (Chen, 2017). Although Tor-OE does not functionally affect glutamate receptors, somehow the increased muscle protein synthesis is converted into an instructive retrograde signal that appears to induce an enhancement in presynaptic glutamate release of a magnitude comparable with that observed in PhTx- and GluRIIA-mediated PHP (Penney, 2012). Although PhTx application, loss of GluRIIA, and Tor-OE each induce a similar enhancement in presynaptic release, to what extent they utilize separate or shared postsynaptic induction pathways, retrograde signaling systems, and modulations to presynaptic function is not known (Goel, 2017).

This study has characterized PHP signaling and expression when induced through PhTx application, loss of GluRIIA, and Tor-OE. This analysis has revealed that a common retrograde signaling system drives similar homeostatic adaptations in the presynaptic terminal but that separate inductive pathways differentially respond to glutamate receptor perturbation, Ca2+/calmodulin-dependent protein kinase II (CamKII) activity, and protein synthesis (Goel, 2017).

There appears to be a core set of genes necessary for both acute and chronic PHP expression, including ones involved in the homeostatic modulation of synaptic vesicle trafficking, presynaptic excitability, calcium channel activity, and active zone remodeling. However, other genes appear to be dispensable for this core program and may rather be involved in secondary functions, such as maintaining PHP expression over chronic timescales or supporting other aspects of homeostatic adaptation. Interestingly, the existence of multiple retrograde signaling pathways may be one reason for the failure of forward genetic approaches to identify any individual genes required in the muscle for the core process of PHP induction, suggesting some level of redundancy. This convergence of diverse induction mechanisms in the postsynaptic cell enables multiple pathways to detect and respond to homeostatic challenges by feeding into a unitary retrograde signaling system that potentiates presynaptic neurotransmitter release to stabilize synaptic strength (Goel, 2017).

CamKII activity plays a crucial role in gating diverse forms of synaptic plasticity. At the Drosophila NMJ, transgenic manipulations that affect postsynaptic CamKII activity have been reported to modulate the expression of PHP in GluRIIA mutants (Haghighi, 2003, Newman, 2017). The current results indicate that pCamKII levels are reduced to similar levels in GluRIIA mutants or following acute pharmacological receptor blockade, consistent with CamKII activity being capable of modulation in seconds at postsynaptic compartments. An attractive model would be that a reduction in calcium influx, either over 10 min or during chronic timescales, triggers diminished pCamKII levels and activates PHP signaling. However, this study found that pharmacological blockade of receptors is necessary and sufficient to reduce pCamKII levels at postsynaptic densities, independent of extracellular calcium, and that incubation in calcium-free saline alone is not sufficient to acutely induce PHP expression. Although there are several indications that reduced calcium influx in the postsynaptic muscle over chronic timescales likely contributes to PHP signaling, perhaps necessitating translation-dependent pathways, a calcium-independent system drives the acute expression of PHP following PhTx application, implying a distinct mechanism (Goel, 2017).

Two possibilities are considered to explain how PhTx application to glutamate receptors is transduced into PHP retrograde signaling without requiring calcium signaling through extracellular sources. First, PhTx binding to receptors may induce a conformational perturbation, distinct from ion influx through the receptor, to initiate PHP signaling. Such a mechanism could operate through a metabotropic mechanism, which would be unanticipated but not unprecedented. For example, at mammalian central synapses, the induction of N-methyl-D-aspartate (NMDA) receptor-dependent long term depression (LTD) does not require calcium influx through NMDA, but, rather, pharmacological perturbation to the receptor is sufficient. A metabotropic pathway has been proposed. Further, mammalian kainate receptors, to which the Drosophila glutamate receptors are homologous, are also capable of signaling through metabotropic mechanisms. Thus, pharmacological perturbation to GluRIIA-containing receptors could, in principle, initiate PHP signaling through an undefined metabotropic mechanism. However, at present, there is no evidence for such a mechanism in Drosophila (Goel, 2017).

Alternatively, pharmacological disruption of glutamate receptors may lead to local signaling at the NMJ through interactions with scaffolds such as Discs large (DLG)/PSD-95 and Dalcium/calmodulin dependent serine protein kinase (CASK). These scaffolds are known to be in complexes with CamKII and capable of modulating CamKII activity and phosphorylation at the subsynaptic reticulum (SSR). Intriguingly, defects in the elaboration of the SSR have recently been reported to disrupt retrograde homeostatic plasticity (Koles, 2015). CamKII signaling during PHP appears to be restricted to postsynaptic densities of type 1b boutons (Newman, 2017), suggesting that compartmentalized signaling at the SSR orchestrates local PHP signal transduction. In contrast, Tor-OE is capable of initiating PHP signaling independent of pCamKII reduction, where it promotes translation throughout the cell. This implies that protein synthesis modulates retrograde signaling downstream of or in parallel to CamKII signal transduction but ultimately feeds back into local post-translational signaling pathways. Future experiments probing the interactions between glutamate receptors, postsynaptic scaffolds, translation, and CamKII activity will clarify the signaling at this compartmentalized synapse (Goel, 2017).

The finding that PHP can be acutely induced by pharmacological perturbation of glutamate receptors and not through reductions in calcium influx over rapid timescales may help to explain perplexing observations regarding the phenomenology of PhTx-mediated PHP. For example, it was noted that PHP can be induced and expressed by a 10-min incubation of PhTx with only mEPSP events occurring. Although a reduction in calcium during these mEPSP events was discussed as a possible induction mechanism, estimates are that, at most, six mEPSP events occur per active zone during this induction time, a very low level and frequency of activity to reliably and robustly produce PHP expression. Indeed, a recent study demonstrated that mEPSP events account for a very small fraction (<1%) of the total postsynaptic calcium signal at individual NMJs (Newman, 2017), making a reduction in calcium even more implausible to explain acute PHP induction. Hence, pharmacological perturbation of postsynaptic glutamate receptors, rather than a reduction in calcium through these receptors, is an attractive mechanism to explain the characteristics of the acute induction of PHP by PhTx and raises interesting questions for future studies about how pharmacological receptor perturbation is transduced into PHP induction (Goel, 2017).

Why might a single retrograde signaling system exist to homeostatically stabilize synaptic strength at the Drosophila NMJ? In central neurons, diverse forms of synaptic plasticity, including Hebbian and homeostatic, dynamically operate over multiple timescales to bi-directionally adjust synaptic strength. Further, translation-dependent and independent processes also contribute to retrograde homeostatic signaling in the hippocampus following AMPA receptor blockade. In contrast, the NMJ is built for stable excitation and is acutely sensitive to reductions in receptor function. However, when neurotransmitter sensitivity in muscle is enhanced by increased receptor expression, no retrograde signaling system exists to homeostatically downregulate presynaptic efficacy. Thus, the muscle is endowed with multiple signaling systems to respond to perturbations but appears limited to signal retrograde increases in neurotransmitter release. Hence, a single retrograde signaling system might provide an efficient means to ensure non-additive potentiation in synaptic strength and prevent hyper-excitation when conflicting signals and multiple inductive mechanisms are simultaneously activated (Goel, 2017).

The maintenance of synaptic homeostasis at the Drosophila neuromuscular junction is reversible and sensitive to high temperature

Homeostasis is a vital mode of biological self-regulation. The hallmarks of homeostasis for any biological system are a baseline set point of physiological activity, detection of unacceptable deviations from the set point, and effective corrective measures to counteract deviations. Homeostatic synaptic plasticity (HSP) is a form of neuroplasticity in which neurons and circuits resist environmental perturbations and stabilize levels of activity. One assumption is that if a perturbation triggers homeostatic corrective changes in neuronal properties, those corrective measures should be reversed upon removal of the perturbation. This study tests the reversibility and limits of HSP at the well-studied Drosophila melanogaster neuromuscular junction (NMJ). At the Drosophila NMJ, impairment of glutamate receptors causes a decrease in quantal size, which is offset by a corrective, homeostatic increase in the number of vesicles released per evoked presynaptic stimulus, or quantal content. This process has been termed presynaptic homeostatic potentiation (PHP). Taking advantage of the GAL4/GAL80TS/UAS expression system, PHP was triggered by expressing a dominant-negative glutamate receptor subunit at the NMJ. PHP was then reversed by halting expression of the dominant-negative receptor. The data show that PHP is fully reversible over a time course of 48-72 h after the dominant-negative glutamate receptor stops being genetically expressed. As an extension of these experiments, it was found that when glutamate receptors are impaired, neither PHP nor NMJ growth is reliably sustained at high culturing temperatures (30-32°C). These data suggest that a limitation of homeostatic signaling at high temperatures could stem from the synapse facing a combination of challenges simultaneously (Yeates, 2017).

Homeostasis is a strong form of biological regulation. It permits individual cells or entire systems of cells to maintain core physiologic properties that are compatible with life. In the nervous system, decades of study have shown that while synapses and circuits are inherently plastic, they also possess robust homeostatic regulatory systems to maintain physiologic stability. Homeostatic plasticity in the nervous system is a non-Hebbian strategy to counteract challenges to neuronal function that may threaten to disrupt essential neuronal and circuit activities (Turrigiano, 2017). Depending on the synaptic preparation examined and the environmental challenge presented to the synapse, homeostatic responses may be executed via compensatory adjustments to presynaptic neurotransmitter release (Cull-Candy, 1980; Petersen, 1997; Murthy, 2001; Thiagarajan, 2005; Frank, 2006; Davis, 2015), postsynaptic neurotransmitter receptor composition (O'Brien, 1998; Turrigiano, 1998; Rongo , 1999; Turrigiano, 2008), neuronal excitability (Marder, 2002; Marder, 2006; Marder and Bucher, 2007; Bergquistet, 2010; Parrish, 2014), or even developmentally, via changes in synaptic contact formation and maintenance (Davis and Goodman, 1998; Burrone, 2002; Wefelmeyer, 2016; Yeates, 2017 and references therein).

Bidirectionality has been documented in several homeostatic systems, perhaps most prominently in the case of synaptic scaling of neurotransmitter receptors. For vertebrate neuronal culture preparations -- such as cortical neurons or spinal neurons—global silencing of network firing can induce increases in excitatory properties, such as increased AMPA-type glutamate receptor accumulation; by contrast, global enhancement of activity can induce the opposite type of response (O'Brien, 1998; Turrigiano, 1998; Wierenga, 2005; Turrigiano, 2008). Bidirectionality is also a key feature underlying homeostatic alterations of neurotransmitter release at peripheral synapses such as the neuromuscular junction (NMJ). At the NMJs of Drosophila melanogaster and mammals, impairing neurotransmitter receptor function postsynaptically results in diminished sensitivity to single vesicles of transmitter. Electrophysiologically, this manifests as decreased quantal size. NMJs respond to this challenge by enhancing neurotransmitter vesicle release (Cull-Candy, 1980; Plomp, 1992, 1995; Petersen, 1997; Davis, 1998; Frank, 2009). By contrast, perturbations that enhance quantal size (for example, overexpression of a vesicular neurotransmitter transporter in Drosophila) can result in decreased quantal content (Daniels, 2004; Gavino, 2015; Yeates, 2017 and references therein).

Synapses and circuits possess myriad solutions to assume appropriate functional outputs in the face of perturbations (Marder, 2002; Marder, 2006). Therefore, a corollary to bidirectional regulation is that homeostatic forms of regulation should also be reversible. There are experimental difficulties of presenting and removing a synaptic challenge in the context of a living synapse, so homeostatic reversibility has not been rigorously studied in an in vivo system or over extended periods of developmental time. Understanding how homeostatic regulatory systems are reversibly turned on and off could have profound implications for elucidating fundamental properties of circuit regulation (Yeates, 2017).

This study exploits the Drosophila NMJ as a living synapse to test homeostatic reversibility. At the Drosophila NMJ, a canonical way to challenge synapse function is through glutamate receptor impairment (Frank, 2014), either genetically (Petersen, 1997) or pharmacologically (Frank, 2006). Impairments of muscle glutamate receptor function decrease quantal size. Decreased quantal size spurs muscle-to-nerve signaling that ultimately results in a homeostatic increase in presynaptic vesicle release, a process that has been termed presynaptic homeostatic potentiation (PHP). The most widely used experimental homeostatic challenges to Drosophila NMJ function are not easily reversed. These challenges include genetic deletion of the glutamate receptor subunit GluRIIA (Petersen, 1997) and the mostly irreversible pharmacological inhibition of glutamate receptors with Philanthotoxin-433 (PhTox; Frank, 2006; Yeates, 2017 and references therein).

For this study, a way was engineered to challenge NMJ function in vivo for significant periods of time, verify the effectiveness of the challenge at a defined developmental time point, remove the challenge, and then assess the homeostatic capacity of the NMJ at a later developmental time point. By using the temporal and regional gene expression targeting (TARGET) GAL4/GAL80TS/UAS expression system (McGuire, 2003) to temporally control the expression of a dominant-negative GluRIIA receptor subunit (DiAntonio, 1999), this study found that homeostatic potentiation of neurotransmitter release is fully reversible. In the course of conducting these studies, a high temperature limitation of homeostatic potentiation at the NMJ was uncovered (Yeates, 2017).

Thus study presents evidence that PHP at the Drosophila neuromuscular synapse is a reversible process. In doing so, prior findings were confirmed showing that there is a tight inverse relationship between quantal amplitude and QC at the NMJ. Those findings were complemented with the results of temperature shift experiments. PHP is measurable at an early stage of larval development and can be erased over a matter of days. Interestingly, at high temperatures, PHP induced by impairing glutamate receptor function either fails or falls short of full compensation. This failure appears to correlate with impaired NMJ growth in the same animals. Why is reversibility slow after dominant-negative GluRIIAM/R removal (Yeates, 2017)?

There was a robust expression of PHP for NMJs of MHC-Gal4 >> UAS-GluRIIAM/R larvae with the TubP-Gal80TS transgene raised at 29°C for 48 h after egg-laying. Once the expression of the dominant-negative UAS-GluRIIAM/R transgene was halted, this expression of PHP was erased over a slow 48- to 72-h period (Yeates, 2017).

If PHP is a readily reversible homeostatic process, why is there a days-long time lag to reverse it? The data likely reflect a constraint of the dominant-negative GluRIIAM/R experimental perturbation, rather than the NMJ's capacity to respond quickly to the changed environment. In a prior study, researchers expressed functional, tagged GluRIIA transgenic subunits at the NMJ and performed fluorescence recovery after photobleaching (FRAP) experiments. Those experiments demonstrated that receptor turnover rates at the Drosophila NMJ are extremely slow: it appears that once postsynaptic densities (PSDs) reach a critical size, GluRIIA subunits are stably incorporated. For the current study, this likely means that the temperature downshifts in the reversibility experiments represented an opportunity for the NMJ to incorporate endogenous WT GluRIIA into a significant number of new PSDs while it continued to grow. Given sufficient growth, the endogenously expressed GluRIIA would gradually overcome the previously incorporated dominant-negative GluRIIAM/R subunits. This would restore electrophysiological parameters to normal levels, which is consistent with the data. Reversibility of rapid and sustained forms of homeostatic plasticity (Yeates, 2017).

The majority of recent studies about synaptic homeostasis at the Drosophila NMJ have emphasized that presynaptic adjustments to neurotransmitter release properties must occur within minutes of drug-induced (PhTox) postsynaptic receptor inhibition to induce a rapid and offsetting response to PhTox challenge. Some important parameters uncovered through these studies include regulation of presynaptic Ca2+ influx; regulation of the size of the readily releasable pool (RRP) of presynaptic vesicles; control of presynaptic SNARE-mediated fusion events; control of neuronal excitability; and recently, implication of endoplasmic reticulum calcium-sensing activities; presynaptic glutamate receptor activity ; and finally, identification of a retrograde, trans-synaptic signaling system governed by the Semaphorin 2b ligand and the Plexin B receptor (Yeates, 2017).

For almost all of the cases in which a mutation or an experimental condition blocks the short-term induction of homeostatic signaling, the same perturbation has also been shown to block its long-term maintenance. Interestingly, however, the converse is not true. Additional studies have shown that the long-term consolidation (or expression) of homeostatic signaling at the NMJ can be genetically uncoupled from its induction. Select molecules seem to be dedicated to a long-term maintenance program that involves protein translation and signaling processes in both the neuron and the muscle. Recent data suggest that such long-term processes may take 6 h or more to take full effect (Yeates, 2017).

As more molecular details about HSP are elucidated, it will be interesting to test whether the rapid induction and sustained consolidation of PHP can be reversed by similar or separate mechanisms, and what the time courses of those reversal mechanisms are. At the mouse NMJ, reversibility was recently demonstrated pharmacologically. d-Tubocurarine was applied at a subblocking concentration to impair postsynaptic acetylcholine receptors. Within seconds of drug application, QC increased—and then within seconds of drug washout, it decreased again (Wang, 2016b). Follow-up experiments suggested that those rapid, dynamic changes in PHP dynamics at the mouse NMJ were mediated by a calcium-dependent increase in the size of the RRP of presynaptic vesicles (Wang, 2016b). Because there seem to be several similarities between the mouse NMJ and the Drosophila NMJ, it is possible that PHP at the insect NMJ can also be rapidly reversed (Yeates, 2017).

It is instructive to consider mammalian synaptic preparations and study how homeostatic forms of synaptic plasticity are turned on and off. Groundbreaking work on cultured excitatory vertebrate synapses showed that in response to activity deprivation (or promotion), synapses employ scaling mechanisms by adding (or subtracting) AMPA-type glutamate receptors to counteract the perturbation. Bidirectional scaling suggested that reversible mechanisms likely dictate homeostatic scaling processes. Complementary studies testing scaling reversibility have borne out this prediction. Additionally, evidence for reversible forms of homeostatic scaling have been found in rodent sensory systems, such as auditory synapses after hearing deprivation (and restoration to reverse) and in the visual cortex after light deprivation. Collectively, the vertebrate and invertebrate studies support the notion that reversible fine-tuning is an efficient strategy used to stabilize activities in metazoan nervous systems. One advantage offered by the Drosophila system is a genetic toolkit to uncover possible reversibility factors. Partial or failed homeostatic signaling at high temperatures (Yeates, 2017).

Are there environmental limitations for homeostatic potentiation at the Drosophila NMJ? Clearly there are. The data suggest that a combination of high temperature (30°C-32°C) plus impaired glutamate receptor function can severely limit the NMJ's ability to compensate for reduced neurotransmitter sensitivity. High temperature alone does not seem to be a severe enough restriction to impair PHP, because Gal4 driver controls at 30°C have somewhat reduced quantal size, but a fully (or nearly fully) offsetting increase in QC. Likewise, reduced glutamate receptor function alone does not appear to be a sufficient barrier to impair PHP. For example, quantal size is severely diminished when the dominant negative UAS-GluRIIAM/R transgene is expressed at 25°C, but PHP is nevertheless intact (Yeates, 2017).

It is not clear what the molecular or anatomic basis of this limit on PHP is. It is known that it is not an issue of NMJ excitation at high temperatures, because evoked neurotransmission for WT (or driver control) NMJs remains remarkably robust over a range of temperatures, including 30°C. Nor does it seem to be an elimination of PHP in general, because PHP was still present in the case of GluRIIASP16 animals raised at 30°C, revealing a distinction between the null GluRIIA condition and the dominant-negative GluRIIA condition. The limitation seems to be on homeostatic signaling that supports PHP at high temperatures in the face of the dominant-negative transgene expression (Yeates, 2017).

Temperature effects on neurophysiology are well documented. Recent work in crustaceans demonstrates that robust and reliable circuits such as the neurons driving the rhythmicity stomatogastric nervous system can fail under extreme temperature challenges. For the Drosophila NMJ, prior studies of larval development documented a significant enhancement of synaptic arborization when larvae were raised at higher temperatures. Additional studies have shown that NMJ growth plasticity can be additionally affected by mutations that affect neuronal excitability. Given the backdrop of these data, it is not unreasonable to hypothesize that the tolerable limits of synaptic activity challenge could be different at different temperatures (Yeates, 2017).

For the current experiments, 29°C-32°C represents a potential failure zone for homeostatic potentiation at the Drosophila NMJ. It must be noted that the data suggest that the coping capacity of the NMJ is dependent on genotype. WT NMJs cope at all temperatures. By contrast, for dominant-negative GluRIIA-expressing NMJs, 29°C is a point at which PHP becomes partial, and 30°C is a point at which it fails. For GluRIIASP16 subunit deletion NMJs, there is robust, but partial PHP at 30°C, not unlike the compensation seen for the dominant-negatives at 29°C. GluRIIASP16 NMJs fail to execute PHP only when temperature is pushed to an extreme range, such as 32°C. Why do these differences persist? It is not clear. The answer could relate to the well-documented temperature-induced alterations in NMJ growth, or alternatively, a limited availability of synaptic factors that are needed to cope with a double challenge of high temperature and particular impairment of glutamate receptor function. Future molecular and physiologic work will be needed to unravel those possibilities in the contexts of different genetic backgrounds and culturing conditions (Yeates, 2017).

Distinct homeostatic modulations stabilize reduced postsynaptic receptivity in response to presynaptic DLK signaling

Synapses are constructed with the stability to last a lifetime, yet sufficiently flexible to adapt during injury. Although fundamental pathways that mediate intrinsic responses to neuronal injury have been defined, less is known about how synaptic partners adapt. This study investigated responses in the postsynaptic cell to presynaptic activation of the injury-related Dual Leucine Zipper Kinase pathway at the Drosophila neuromuscular junction. The postsynaptic compartment reduces neurotransmitter receptor levels, thus depressing synaptic strength. Interestingly, this diminished state is stabilized through distinct modulations to two postsynaptic homeostatic signaling systems. First, a retrograde response normally triggered by reduced receptor levels is silenced, preventing a compensatory enhancement in presynaptic neurotransmitter release. However, when global presynaptic release is attenuated, a postsynaptic receptor scaling mechanism persists to adaptively stabilize this diminished neurotransmission state. Thus, the homeostatic set point of synaptic strength is recalibrated to a reduced state as synapses acclimate to injury (Goel, 2018).

Neurons are endowed with robust surveillance systems that detect injury and initiate latent plasticity programs involving regenerative and degenerative responses. A fundamental signaling system induced after neuronal injury is mediated by an evolutionarily conserved mitogen-activated protein kinase called Dual Leucine Zipper Kinase (DLK). DLK signaling initiates translational changes in axons and transcriptional responses in the nucleus that ultimately promotes degeneration at the distal axon and regeneration proximal to the site of injury. Members of the Phr1/Highwire/Rpm-1 (PHR) protein family control DLK signaling to gate neuronal injury-related signaling programs. At the Drosophila neuromuscular junction, highwire (hiw) encodes an E3 ubiquitin ligase that constitutively degrades the DLK homolog Wallenda (Wnd). However, after axonal injury, Wnd is no longer degraded by Hiw, leading to increased Wnd protein levels and activation of regenerative and degenerative signaling programs. Genetic loss of hiw in neurons constitutively activates Wnd signaling, while neuronal overexpression of wnd can overcome Hiw-mediated degradation to activate this same program. This relationship between PHR proteins, DLK activity, and injury-related signaling is conserved in other invertebrate and mammalian systems. Thus, loss of hiw or overexpression of wndf in neurons activates an intrinsic signaling system that transforms the cell into a state of persistent degenerative and regenerative adaptations related to a programmed response to injury (Goel, 2018).

Although, Wnd/DLK signaling and other injury-related responses occur cell autonomously and are intrinsic to the specific neuron, there is emerging evidence that synaptically connected cells may sense this perturbation and adapt in response. For example, foundational studies at the mouse NMJ have demonstrated that motor neuron injury, denervation, or synapse elimination can provoke disassembly and remodeling of the postsynaptic specialization, including a diminution of neurotransmitter receptor levels, which can precede obvious changes in the overlying nerve terminal. There is also evidence for neuronal and synaptic remodeling in the spinal cord and other areas in the central nervous system following injury. At the glutamatergic Drosophila NMJ, an apparent reduction in the synaptic response to glutamate (quantal size) was shown in hiw mutants as well as following a 'nerve crush' injury to otherwise wild-type NMJs. Neuronal expression of hiw restores normal synaptic strength and quantal size in hiw mutants, indicating that while the postsynaptic muscle does not itself experience hiw-related signaling, a reduction in either the amount of glutamate released and/or the postsynaptic sensitivity to neurotransmitter occurs in response. However, it remains unclear how synaptic function and plasticity change in response to presynaptic Wnd/DLK signaling, a state in which injury-related signaling is active but before complete degeneration or loss of the presynaptic terminal has occurred (Goel, 2018).

At the Drosophila NMJ, two forms of homeostatic plasticity have been described that stabilize synaptic strength in response to perturbations that would otherwise disrupt functionality. First, pharmacological or genetic manipulations that diminish postsynaptic glutamate receptor (GluR) function initiate a retrograde signaling system. Specifically, a retrograde signal emitted from the muscle instructs the neuron to compensate by increasing presynaptic glutamate release to restore set point levels of synaptic strength. This process is conserved in rodents and humans, and is termed presynaptic homeostatic potentiation (PHP) because the expression mechanism of this form of plasticity is a presynaptic increase in neurotransmitter release. Second, hypo-innervation of the NMJ induces a distinct form of adaptive plasticity. In this situation, a reduction in presynaptic neurotransmitter release is observed proportional to the reduction in synapse number. In response, a homeostatic increase in quantal size is induced that stabilizes synaptic strength. However, to what extent these homeostatic mechanisms operate at synapses that have adapted to injury-related signaling is unknown (Goel, 2018).

This study has characterized synaptic structure, function, and plasticity at NMJs with active Wnd signaling in presynaptic neurons, with a particular focus on how the postsynaptic muscle responds. This analysis has revealed that the postsynaptic muscle responds by diminishing GluR abundance and by silencing the retrograde homeostatic signaling system that would normally enhance presynaptic release following reduced GluR levels. However, this subdued state of synaptic strength is homeostatically maintained through adaptive modulations to postsynaptic GluR levels following hypo-innervation. Together, this illuminates the distinct signaling systems targeted for modulation in the postsynaptic cell that stabilize a muted synaptic state in response to presynaptic Wnd signaling (Goel, 2018).

The conserved hiw/PHR gene family regulates Wnd/DLK signaling and has diverse roles in synaptic development, degeneration and regeneration, and neurotransmission. Although there are clearly myriad changes induced in neurons following Wnd activity, at the Drosophila NMJ, no evidence was found for hiw or wnd having any functions in the postsynaptic muscle. Although Wnd has no apparent function in regulating presynaptic neurotransmitter release, its negative regulator, Hiw, does have an independent function in promoting neurotransmitter release. The putative substrate that mediates this role for Hiw is unknown, but must be separate from Wnd and the downstream pathways orchestrated through Wnd signaling. Therefore, following injury, inhibition of Hiw exerts two separate influences that mutes functionality at both synaptic compartments: First, Wnd is activated to provoke a diminished state of responsiveness in the muscle, and second, presynaptic neurotransmitter release is reduced. The substrate that mediates the Wnd-independent role of Hiw to reduce presynaptic efficacy is unknown, but will be an interesting target to define in future studies (Goel, 2018).

It is remarkable that although Wnd signaling is driven solely in the presynaptic compartment at the NMJ, this process is sensed and transforms the postsynaptic muscle into a novel state characterized by subdued responsiveness to presynaptic activity. Neuronal Wnd signaling must therefore impart anterograde information to the muscle. Importantly, this information is communicated independently of altered synaptic development and structure, as well as evoked neurotransmission. An attractive possibility is that trans-synaptic adhesion proteins may mediate the activity-independent anterograde signaling following presynaptic Wnd activation. Indeed, multiple extracellular cues and trans-synaptic signals interpose a rich dialog between presynaptic terminals, postsynaptic compartments, and the extracellular matrix at synapses. Trans-synaptic 'nanocolumns' have recently emerged as inter-cellular signaling complexes that orchestrate synaptic remodeling and plasticity in addition to development, maturation, and structural alignment at synapses. It is tempting to speculate that direct interactions through these nanocolumns may communicate injury-related signaling from presynaptic terminals to postsynaptic partners to remodel receptor fields and plasticity pathways. Intriguingly, there is evidence that many new proteins are expressed in neurons following activation of Wnd/DLK signaling, providing possible candidates to test for roles in this process (Goel, 2018).

One major outcome of the response in muscle to presynaptic Wnd signaling is a diminution of postsynaptic neurotransmitter receptors. This parallels observations at the mouse NMJ following denervation or synapse elimination, in which a reduction in neurotransmitter receptor levels and protein synthesis in muscle has been demonstrated. This selective loss-of-acetylcholine receptors at synaptic sites is a result of removal of receptors from these areas coupled with a lack of insertion of new receptors. It is likely that similar mechanisms work to reduce GluR levels at the Drosophila NMJ following neuronal Wnd signaling. In addition, the decreased synaptic protein levels observed in the muscle following neuronal Wnd signaling may be result from modulation of Tor activity, as postsynaptic overexpression of Tor globally elevates protein synthesis and partially restores receptor levels. Finally, an intriguing study in the rodent visual system revealed that ablation of presynaptic photoreceptors leads to remodeling of the postsynaptic apparatus, including the rapid and localized disappearance of GluRs. Together, these studies demonstrate that postsynaptic targets adapt to injury, disease, and loss-of-presynaptic inputs by remodeling postsynaptic neurotransmitter receptor complexes to reduce sensitivity, suggesting a conserved response (Goel, 2018).

While beneficial effects of fasting on organismal function and health are well appreciated, little is known about the molecular details of how fasting influences synaptic function and plasticity. Genetic and electrophysiological experiments demonstrate that acute fasting blocks retrograde synaptic enhancement that is normally triggered as a result of reduction in postsynaptic receptor function at the Drosophila larval neuromuscular junction (NMJ). This negative regulation critically depends on transcriptional enhancement of eukaryotic initiation factor 4E binding protein (4E-BP) under the control of the transcription factor Forkhead box O (Foxo). Furthermore, the findings indicate that postsynaptic 4E-BP exerts a constitutive negative input, which is counteracted by a positive regulatory input from the Target of Rapamycin (TOR). This combinatorial retrograde signaling plays a key role in regulating synaptic strength. These results provide a mechanistic insight into how cellular stress and nutritional scarcity could acutely influence synaptic homeostasis and functional stability in neural circuits (Goel, 2018).

One mechanism that stabilizes the reduced state of synaptic strength following neuronal Wnd signaling is a selective occlusion of postsynaptic PHP transduction. Neurons experiencing Wnd signaling are competent to homeostatically modulate presynaptic neurotransmitter release, but apparently do not receive the retrograde information necessary from the muscle, even following additional perturbations to or restorations of postsynaptic GluRs. Interestingly, perturbations to Cap-dependent protein synthesis in the postsynaptic muscle have been shown to disrupt PHP retrograde signaling, and metabolic changes in the muscle can also impinge on this pathway to modulate PHP signaling (Kauwe, 2016). This indicates that one mechanism utilized by the muscle to respond to neuronal Wnd signaling, distinct from a general reduction in protein synthesis, may be a selective inhibition of Tor-dependent protein synthesis which, in turn, contributes to the occlusion of PHP signaling. It is interesting to note that while PHP signaling is highly compartmentalized, it can be still be expressed in neurons despite perturbations to a variety of processes that disrupt synaptic structure and function independently of injury-related signaling. Thus, an instructive cue mediated by Wnd signaling selectively impairs retrograde PHP communication in the muscle to stabilize a reduced level of synaptic strength (Goel, 2018).

Although PHP signaling appears to be inhibited, a second mechanism stabilizes the reduced set point of synaptic strength following neuronal Wnd signaling, GluR scaling. This postsynaptic form of homeostatic plasticity parallels the postsynaptic scaling of GluRs observed following silencing of neuronal activity in mammalian central synapses. At the Drosophila NMJ, the induction of this form of homeostatic plasticity was known to require hypo-innervation, and this study has shown that a selective increase in postsynaptic GluRIIA-containing receptors compensates for reduced neurotransmitter release and maintains stable synaptic strength. Although the signal transduction system that mediates this form of plasticity is enigmatic, it is clearly distinct from the retrograde signaling system that drives PHP. The homeostatic set point of synaptic strength has been demonstrated to be plastic, and can be stabilized at levels distinct from baseline values in mutations that disrupt synaptic function and during aging of the synapse. The current results define injury-related signaling as an additional process that has the capacity to adjust the homeostatic set point of synaptic strength (Goel, 2018).

Why might synaptic strength be adjusted to a reduced level following injury? The coordinated reduction in Wnd-mediated postsynaptic responsiveness and Hiw-mediated presynaptic efficacy may promote sufficient time for the process of repair vs. degeneration to be adjudicated within an injured neuron, while still maintaining synaptic communication at a reduced level. Loss or further reductions to neurotransmission may destabilize synaptic integrity, impairing the series of subsequent steps necessary to restore normal synaptic strength and structure should the injury be successfully overcome. In addition, a further reduction in postsynaptic receptivity may deprive the neuron of necessary trophic support and promote neuronal degeneration. Indeed, a lack of trophic support from the muscle due to weakened synaptic activity contributes to neuromuscular disease pathogenesis. In the central nervous system, there is evidence that synaptically coupled cells sense and respond to injury. Hence, injury to an individual neuron can propagate and destabilize an entire neural circuit without adaptive counter measures. The current findings illustrate the acclimations that occur in postsynaptic targets to neurons experiencing injury-related signaling and highlight the adaptations to synaptic plasticity that maintain stable functionality around a reduced set point of synaptic strength (Goel, 2018).

The SMAD2/3 interactome reveals that TGFβ controls m6A mRNA methylation in pluripotency

The TGFβ pathway has essential roles in embryonic development, organ homeostasis, tissue repair and disease. These diverse effects are mediated through the intracellular effectors SMAD2 and SMAD3 (hereafter SMAD2/3), whose canonical function is to control the activity of target genes by interacting with transcriptional regulators. Therefore, a complete description of the factors that interact with SMAD2/3 in a given cell type would have broad implications for many areas of cell biology. This study describes the interactome of SMAD2/3 in human pluripotent stem cells. This analysis reveals that SMAD2/3 is involved in multiple molecular processes in addition to its role in transcription. In particular, a functional interaction was identified with the METTL3-METTL14-WTAP complex, which mediates the conversion of adenosine to N6-methyladenosine (m6A) on RNA4. SMAD2/3 promotes binding of the m6A methyltransferase complex to a subset of transcripts involved in early cell fate decisions. This mechanism destabilizes specific SMAD2/3 transcriptional targets, including the pluripotency factor gene NANOG, priming them for rapid downregulation upon differentiation to enable timely exit from pluripotency. Collectively, these findings reveal the mechanism by which extracellular signalling can induce rapid cellular responses through regulation of the epitranscriptome. These aspects of TGFβ signalling could have far-reaching implications in many other cell types and in diseases such as cancer (Bertero, 2018)

Calcium-activated Calpain specifically cleaves Glutamate Receptor IIA but not IIB at the Drosophila neuromuscular junction

Calpains are calcium-dependent, cytosolic proteinases active at neutral pH. They do not degrade but cleave substrates at limited sites. Calpains are implicated in various pathologies, such as ischemia, injuries, muscular dystrophy, and neurodegeneration. Despite so, the physiological function of calpains remains to be clearly defined. Using the neuromuscular junction of Drosophila of both sexes as a model, RNAi screening was performed, and calpains were found to negatively regulated protein levels of the glutamate receptor GluRIIA but not GluRIIB. Calpains enrich at the postsynaptic area, and the calcium-dependent activation of calpains induced cleavage of GluRIIA at Q788 of its C terminus. Further genetic and biochemical experiments revealed that different calpains genetically and physically interact to form a protein complex. The protein complex was required for the proteinase activation to downregulate GluRIIA. These data provide a novel insight into the mechanisms by which different calpains act together as a complex to specifically control GluRIIA levels and consequently synaptic function (Metwally, 2019).

Calpains are a family of calcium-activated cytoplasmic cysteine proteases, which are ubiquitously expressed in various mammalian tissues and are functionally active at neutral pH. Members of the calpain family act in pathological processes associated with calcium overload, such as ischemia and Alzheimer's disease. For example, calpains cleave multiple synaptic proteins, including both inotropic and metabotropic glutamate receptors in excitotoxic conditions, as well as in synaptic plasticity (Chan, 1999; Xu, 2007; Doshi, 2009; Baudry, 2016). However, little is known of the precise physiological function of calpain-dependent cleavage of target proteins at synapses during normal development (Metwally, 2019).

Calpains exist in organisms ranging from bacteria to humans. To date, 15 different calpains have been identified in mammals and four in Drosophila. Unlike most proteases, calpains do not destroy but cleave their substrates at limited sites to modulate their function. Calpain 1 (μ-calpain) and calpain 2 (m-calpain) are the most ubiquitous and well-studied calpain family members. They are activated in vitro by micromolar and millimolar concentrations of calcium, respectively. Conventional calpains 1 and 2 are heterodimers composed of a large catalytic subunit with four domains (dI-dIV) and a small regulatory subunit with two domains (dV and dVI). Heterodimerization of the large and small subunits occurs through a unique interaction between their C-terminal domains. There are two calcium-binding sites within the crystal structure of the enzymatically active domain II of μ-calpain. These sites, together with domains IV and VI, interact with calcium and are required for full enzymatic activity. Calpastatin is an endogenous calpain inhibitor that binds and inhibits calpains via its calpain-inhibitor domains when the proteases are activated by calcium (Hanna, 2008). Unlike mammalian calpains, Drosophila calpains comprise a large subunit only and calpastatin has not been identified in Drosophila (Metwally, 2019).

The efficiency of neurotransmission is determined by the level of neurotransmitter receptors at the postsynaptic densities (PSDs). However, the molecular mechanism by which the level of neurotransmitter receptors at synapses is regulated is not well understood. At Drosophila neuromuscular junction (NMJ) synapses, there are two subtypes of glutamate receptors (GluR), GluRIIA and GluRIIB, which are developmentally, biophysically, and pharmacologically distinct. For example, the postsynaptic sensitivity to glutamate is much higher at synapses enriched for GluRIIA than at synapses enriched for GluRIIB. Previous studies show that GluRIIA and GluRIIB receptors are differentially regulated via independent pathways, although the molecular mechanisms that control different subtypes of GluRs remain elusive. To uncover new pathways that control GluR expression at NMJ synapses, a genetic screen was carried out identified a critical role was identified for calpains in specifically regulating the abundance of GluRIIA, but not GluRIIB, at the Drosophila NMJs. Valpains enrich at the postsynaptic area and cleave GluRIIA, but not GluRIIB, at its C-terminal in a calcium-dependent manner. As calcium is a highly versatile intracellular signal that regulates many different cellular functions, the finding of calpain proteinases activated by calcium under physiological conditions at NMJ synapses offers novel insights into the myriad of calcium-mediated processes at other cellular contexts (Metwally, 2019).

Neurotransmission efficiency is determined by the level of neurotransmitter receptors at the PSDs. Therefore, the regulation of glutamate receptors at synapses has been under intensive studies. However, the molecular mechanism responsible for regulating the expression of neurotransmitter receptors at synapses is still poorly understood. Using genetic screening, together with optogenetic manipulation and calcium treatment, this study has revealed that GluRIIA protein levels were specifically and negatively regulated by calcium-dependent calpains during development (Metwally, 2019).

Extensive studies have demonstrated robust homeostatic regulation at the Drosophila NMJ. Genetic or pharmacological manipulations that impair the protein level or activity of postsynaptic GluRs cause a compensatory increase in presynaptic action potential-evoked neurotransmitter release; the increase in evoked release precisely compensates for the decrease in GluR sensitivity and, thereby, maintains a normal level of transmission. This study has shown that calpains negatively regulated mEJP amplitude and frequency at NMJs, consistent with previous studies demonstrating that quantal size and mEJP frequency are increased when GluRIIA is overexpressed. The increase in mEJP amplitude and frequency in calpain mutants can be explained by the elevated levels of GluRIIA and Brp protein at synapses. GluRIIA mutant exhibits a large decrease in mEJP amplitude with no change in EJP amplitude, probably due to homeostatic regulation, whereas overexpression of GluRIIA leads to an increase in mEJP amplitude with an increase of EJP amplitude, indicating that there is no compensatory downregulation of quantal content. In the latter case, different effects on EJP amplitude were observed in a more quantitative manner by using a different gene dosage of GluRIIA. Specifically, increased EJP amplitude is observed at synapse overexpressing a low level of GluRIIA, whereas overexpression of GluRIIA at a high level leads to normal EJP amplitude. This latter case is consistent with the finding that a high threefold increase of GluRIIA level in calpains mutant did not lead to upregulation of EJP amplitude. It is possible that the increase in mEJP amplitude is compensated by a downregulation of quantal content to maintain normal EJP amplitude in calpain mutants. The finding of normal EJP amplitudes regardless of upregulation or downregulation of GluRIIA abundance upon calpain manipulations further support homeostatic regulation at Drosophila NMJ (Metwally, 2019).

Previous studies showed that GluRIIA and GluRIIB exhibit competing effects and interdependence as a result of the selective absence of GluRIIA or GluRIIB type at NMJ. If the primary effect of calpains was specifically on GluRIIA, we would expect to see an opposite change in GluRIIB level. However, this is not what we observed. Instead, we observed that GluRIIB was normally localized and its levels were unchanged when GluRIIA was upregulated, resulting from calpain mutation or knockdown, suggesting a distinct regulation of subunit composition by calpains at the NMJ (Metwally, 2019).

At the Drosophila NMJ, the postsynaptic calcium has been speculated to regulate the sensitivity and synaptic localization of GluRIIA. However, how calcium regulates GluRIIA at NMJ is yet to be investigated. The findings in the present study showed a critical role of calcium in regulating postsynaptic GluRIIA levels via calpain. In Drosophila muscles and S2 cells, calcium-dependent activation of calpain resulted in the C terminus cleavage of GluRIIA but not GluRIIB. It is noted that the percentage of cleaved GluRIIA by calpains is low by Western analysis. This low cleavage efficiency is consistent with previous findings. However, immunostaining showed apparent changes in GluRIIA level at NMJ upon manipulating calpain expressions. Multiple possibilities could explain the discrepancy between low cleavage efficiency by Western analysis and apparent changes in GluRIIA level by immunostaining. First, synaptic GluRIIA is part of the total GluRIIA and calpain may preferentially cleave a subset of synaptic GluRIIA, which is physically accessible and sensitive to calpain cleavage due to specific post-translational modifications. The enzymatic activity of calpains could also be differentially regulated at different cellular locations. Alternatively, calpain may indirectly regulate GluRIIA level by other mechanism, such as translation regulation (Metwally, 2019).

There are a few reports that calpain is involved in the regulation of protein synthesis. For example, ischemia-induced calpain activation targets translation initiation factor 4G1 for cleavage, leading to reduced translation. On the other hand, calpain stimulates protein synthesis by truncating B56a, the regulatory subunit of PP2A which inhibits the Akt/mTOR/S6 signaling during mGluR-dependent LTD. Although multiple independent lines of evidence support that Drosophila calpains specifically target GluRIIA for cleavage, there is still a possibility that calpain negatively regulates GluRIIA level through other unknown mechanisms, including translational regulation (Metwally, 2019).

Where does GluRIIA get cleaved by calpains within the cell? As calpains are enriched in the postsynaptic area where calcium transients are produced by spontaneous and evoked transmitter release, it is possible that calpain cleaves GluRIIA at the postsynaptic area. Alternatively, calpain might cleave GluRIIA at the endoplasmic reticulum when it is synthesized. As an obvious decrease in GluRIIA intensity was observed upon calcium treatment for a short time of 30 min, the first scenario is favored, in which calpain cleaves GluRIIA mostly at the postsynaptic site (Metwally, 2019).

Calpain interacts with and cleaves dozens of target proteins. The present studyfound that calpains specifically target GluRIIA but not other synaptic proteins at the Drosophila NMJ terminals. A physical interaction between calpain and GluRIIA in a calcium-dependent manner. This is the first demonstration of calcium-facilitated interaction of calpain with its target protein in cultured cells. It will be of interest to test whether other substrates of calpains identified so far also interact with calpains in a calcium-dependent manner (Metwally, 2019).

Calpain activation is triggered by abnormally high calcium levels associated with pathologies, such as ischemic insults and neurodegeneration, but the in vivo role of calpain under physiological conditions is largely unknown. This genetic and optogenetic results indicate that calcium-induced activation of calpains is likely to occur under certain physiological conditions at NMJ synapses during development. As calcium signaling is involved in a variety of molecular and cellular processes in development and disease, the current findings offer a novel insight into the role of calcium signaling pathways. How and where calpains are activated by calcium within a cell remain to be elucidated (Metwally, 2019).

Multiple independent lines of evidence showed that different calpains, at least calpains A, B, and D, form a complex in which each is essential for the normal proteinase activity to control GluRIIA level. Drosophila calpain B was originally demonstrated to work as monomers based on in vitro studies of Escherichia coli-produced recombinant protein (Park, 2008), rather than as heterodimers, such as mammalian calpains 1 and 2. However, the current results showed that different calpains acted together in vivo to downregulate GluRIIA. Specifically, this study showed that different calpains colocalized and interacted physically in vivo and in cultured cells. Drosophila calpains do not form homodimers, at least for calpain A (this study) and calpain B (Park, 2008), but instead formed multimers consisting of different calpains, although the domain mediating calpain-calpain interaction remains to be defined. The finding of a calpain complex is supported by a previous study demonstrating that different calpains in Drosophila function synergistically downstream of calcium transients to trigger dendritic pruning of sensory neurons during metamorphosis (Kanamori, 2013). Similarly, mammalian calpains 8 and 9, which are specifically expressed in the gastrointestinal tract, interact physically to form a protein complex called 'gastric calpain' involving in gastric mucosal defense. Thus, in addition to the conventional calpain heterodimers, such as calpains 1 and 2, consisting of a large and a small subunit, heterodimers or heteromultimers of different calpains without a small subunit are emerging as a new mode of calpain activity regulation (Metwally, 2019).

Excess glutamate release triggers subunit-specific homeostatic receptor scaling

Ionotropic glutamate receptors (GluRs) are targets for modulation in Hebbian and homeostatic synaptic plasticity and are remodeled by development, experience, and disease. This study has probed the impact of synaptic glutamate levels on the two postsynaptic GluR subtypes at the Drosophila neuromuscular junction, GluRIIA and GluRIIB. It was demonstrated that GluRIIA and GluRIIB compete to establish postsynaptic receptive fields, and that proper GluR abundance and composition can be orchestrated in the absence of any synaptic glutamate release. However, excess glutamate adaptively tunes postsynaptic GluR abundance, echoing GluR scaling observed in mammalian systems. Furthermore, when GluRIIA vs. GluRIIB competition is eliminated, GluRIIB becomes insensitive to glutamate modulation. In contrast, GluRIIA is now homeostatically regulated by excess glutamate to maintain stable miniature activity, where Ca(2+) permeability through GluRIIA receptors is required. Thus, excess glutamate, GluR competition, and Ca(2+) signaling collaborate to selectively target GluR subtypes for homeostatic regulation at postsynaptic compartments (Han, 2023).

By generating null mutations in GluRIIA and GluRIIB subunits, this study has shown a competition exists between GluR subtypes that establishes stable postsynaptic fields. While synaptically released glutamate is not required to organize this process, excess glutamate triggers an adaptive downscaling of both GluR subtypes in the postsynaptic compartment. However, when this GluR subtype competition is eliminated, a clear and distinctive relationship is revealed between excess glutamate and GluR plasticity: GluRIIB receptors become completely insensitive to excess glutamate, with stable GluRIIB levels maintained. In contrast, GluRIIA receptors constitute the 'plastic' receptor subtype, homeostatically tuned to excess synaptic glutamate release to maintain stable miniature activity. Further, Ca2+ influx through GluRIIA receptors is a key transducer of this signaling system, rendering GluRIIA non-plastic when Ca2+ permeability is lost. Together, these results highlight the interplay between GluR subtype competition, synaptic glutamate, and Ca2+ signaling at postsynaptic compartments and reveal the existence of homeostatic receptor scaling at the Drosophila NMJ (Han, 2023).

Several layers of regulation operate at postsynaptic compartments to establish GluR fields at the fly NMJ. First, the relative level of GluR transcription and translation between subtypes can ultimately set GluRs at synapses. This is demonstrated by overexpression of either GluRIIA or GluRIIB subunits, which can saturate the entire GluR field at postsynaptic compartments and lead to the concomitant loss of the other GluR subtype. Second, post-translational processes, mediated by such factors as enzymatic cleavage, phosphorylation, and degradation modulate GluR activity and abundance. For example, Ca2+-dependent protein cleavage by calpain, phosphorylation control by p21-activated kinase, and proteosomal degradation by the E3 ubiquitin ligase adapter Diablo have all been shown to modulate GluRs at the fly NMJ, which parallel findings in vertebrates. Third, while glutamate released from synaptic vesicles is not necessary to establish or maintain GluR fields in Drosophila or in mammals there is evidence that ambient glutamate modulation from non-vesicular glutamate release and glial transporters might regulate GluR clustering and receptor field size, while excess vesicular glutamate release triggers adaptive reductions in GluRs. There is also evidence that correlated or diminished activity selectively regulates GluRIIA abundance (Han, 2023).

In C. elegans, parallel processes regulate GluR trafficking and plasticity. It will be of interest to determine how these many layers of control intersect and are coordinated to establish GluR fields during development and remodeling in plasticity. There appears to be a hierarchy of regulatory steps controlling GluR plasticity in response to excess glutamate. While excess glutamate downregulates both GluRIIA and GluRIIB abundance when both receptor subtypes are present, this plasticity is adaptive but not homeostatic-miniature amplitude is still enhanced. However, when GluRIIA vs. GluRIIB competition is eliminated, a complete distinction in GluR behavior is revealed: GluRIIB is not responsive to excess glutamate, while GluRIIA receptors are sensitively tuned to glutamate to now enable the homeostatic control of miniature activity. Ca2+ influx through GluRIIA is crucial to this plasticity, where loss of this secondary messenger converts GluRIIA to behave like static GluRIIB receptors. It is possible that excess glutamate drives Ca2+-related signaling through GluRIIA in postsynaptic compartments that ultimately acts on both GluRIIA and GluRIIB when both receptor subtypes are present, at least at MN-Ib NMJs, and this signaling may be lost in GluRIIB-only NMJs. Although the downstream effectors that respond to excess glutamate and Ca2+ to modulate postsynaptic GluRs are not known, an attractive candidate is the auxiliary KAR subunit Neto, which regulates GluR abundance at the fly NMJ46 and functionally modulates AMPARs in worms (Han, 2023).

Interestingly, KARs in mammals are also under homeostatic control, where the auxiliary subunit Neto controls key properties of these receptors. What purpose might two GluR subtypes, differing in their current amplitudes, biophysics, and plasticity, subserve? One idea is that GluRIIB receptors provide a basal signal at postsynaptic compartments to maintain synaptic dialogue, while GluRIIA receptors are the potent subtype that sets synaptic strength, drives muscle contraction, and is targeted for plasticity. These differential functions may be reflected in their distinct subsynaptic localizations, with GluRIIA enriched opposite active zone centers where glutamate is released, while GluRIIB is enriched in the outside periphery of these areas. Another possibility, not mutually exclusive, is that GluRIIBs serve as 'back-up' receptors to maintain NMJ transmission and locomotion when GluRIIAs are blocked, a phenomenon that occurs naturally at larval NMJs due to toxins injected by parasitoid wasps and other organisms. Indeed, presynaptic homeostatic potentiation, a conserved form of retrograde plasticity modeled at the fly NMJ, is induced when GluRIIA is lost or pharmacologically inhibited to maintain stable NMJ excitation (Han, 2023).

Hence, stable GluRIIB receptors provide robustness to buffer NMJ function from perturbations while also allowing flexibility for GluRIIA receptors to dynamically change with plasticity (Han, 2023).

Although the Drosophila NMJ has long been used as a model to study presynaptic forms of adaptive plasticity such as homeostatic potentiation and depression,recent work has found parallel modes of adaptive plasticity that target postsynaptic GluR abundance at this model glutamatergic synapse. In addition to excess glutamate targeting GluRs, at least three additional examples of adaptive GluR plasticity have been observed at the fly NMJ. First, in synaptic undergrowth mutants, where presynaptic innervation is reduced, overall synaptic strength can be maintained at least in some cases through an adaptive enhancement of postsynaptic GluR abundance. Second, when innervation by a single motor neuron is biased at adjacent muscles, stable synaptic strength is maintained through both pre- and postsynaptic mechanisms, with a homeostatic increase in postsynaptic GluR abundance necessary at hypo-innervated NMJs. Third, activation of injury-related signaling in motor neurons induces a downregulation in postsynaptic GluR abundance to adaptively reduce the set point of synaptic strength (Han, 2023).

These examples suggest an underappreciated level of postsynaptic plasticity exists at the Drosophila NMJ, which when combined with the sophisticated genetic and functional tools available, highlights the great potential for this system to illuminate how GluR subtype competition and presynaptic function target GluRs for adaptive modulation (Han, 2023).

There are several limitations regarding understanding of the timing and dynamics of the GluR plasticity described in this study. Because GluR composition, abundance, and function were assessednthrough imaging and electrophysiology at fixed times in later stages of development (third-instar larvae), the time course of the changes in GluR levels (minutes, hours, days) is uncertain. It is also not clear whether the reduction in GluRIIA levels, in response to excess glutamate release, is due to reduced transcription, translation, trafficking, and/or enhanced degradation. Finally, it is not clear to what extent action-potential-driven patterns of activity are altered during development in vGlut-OE NMJs and whether these potential differences contribute to GluR plasticity in addition to the enhanced glutamate emitted from individual synaptic vesicles. Future studies, including intravital live imaging of GluRs, will help to address these limitations (Han, 2023).

The Drosophila postsynaptic DEG/ENaC channel ppk29 contributes to excitatory neurotransmission

The protein family of Degenerin/Epithelial Sodium Channels (DEG/ENaC) is comprised of diverse animal-specific, non-voltage-gated ion channels that play important roles in regulating cationic gradients across epithelial barriers. However, the specific neurophysiological functions of most DEG/ENaC-encoding genes remain poorly understood. This study demonstrates that ppk29 contributes specifically to the postsynaptic modulation of excitatory synaptic transmission at the larval neuromuscular junction (NMJ). Electrophysiological data indicate that the function of pppk29 in muscle is necessary for normal postsynaptic responsivity to neurotransmitter release, and for normal coordinated larval movement. The ppk29 mutation does not affect gross synaptic morphology and ultrastructure, which indicates the observed phenotypes are likely due to defects in glutamate receptor function. Together, these data indicate that DEG/ENaC ion channels play a fundamental role in the postsynaptic regulation of excitatory neurotransmission (Hill, 2017).

The identities and functions of genes that regulate neuronal synaptic functions in health and disease remains a major goal of neuroscience research. Although the principle molecular mechanisms that regulate synapse formation and function are relatively well understood, mechanisms for synaptic plasticity, especially at the physiological timescale, are still mostly unknown. This study describes a novel role for pickpocket (ppk) 29, a member of the degenerin/epithelial sodium channel (DEG/ENaC) family of non-voltage-gated sodium channels, in the postsynaptic modulation of baseline excitatory neurotransmission at the Drosophila larval neuromuscular junction (NMJ) (Hill, 2017).

Members of the DEG/ENaC family are exclusively found in animal genomes. They function as trimeric cation channels, which are expressed in both neuronal and non-neuronal tissues. DEG/ENaC channels can be gated by diverse extracellular stimuli, including extracellular ligands and mechanical forces (Ben-Shahar, 2011; Eastwood, 2012; Hill, 2017 and references therein).

Several studies in invertebrate and mammalian species suggest that some DEG/ENaC family members also directly contribute to synaptic functions (Younger, 2013; Urbano, 2014; Ievglevskyi, 2016; Miller-Fleming, 2016), which may explain their reported contributions to long-term potentiation, learning and memory (Wemmie, 2002), and addiction (Kreple, 2014). In addition, mutations in DEG/ENaC-encoding genes have been implicated in neuropathologies such as multiple sclerosis and epilepsy (Wemmie, 2013). However, whether the observed neuronal and behavioral phenotypes of mutations in DEG/ENaC-encoding genes are due to presynaptic or postsynaptic processes is not well understood (Hill, 2017).

In contrast to mammalian genomes, which typically harbor eight to nine independent DEG/ENaC-encoding genes, the genome of the fruit fly Drosophila melanogaster encodes >30 independent family members, named ppk genes. Analyses of mutations in several ppk genes indicates that Drosophila DEG/ENaC channels contribute to diverse sensory functions such as salt taste, water sensing, and the detection of mating pheromone. In addition, some ppk genes have been implicated in the maintenance of synaptic homeostasis (Younger, 2013). Nevertheless, the specific molecular mechanisms by which these channels exert their synaptic functions remain elusive (Hill, 2017).

This study reports that ppk29, which has been reported previously as a neuronally enriched Drosophila DEG/ENaC subunit implicated in pheromone-sensing functions (Thistle, 2012; Mast, 2014; Vijayan, 2014; Yuan, 2014), is also required for normal neurotransmission at the larval NMJ, a model glutamatergic synapse, via postsynaptic processes, possibly via modulation of postsynaptic glutamate receptors (Hill, 2017).

Despite their emerging importance in neurological and cognitive pathologies, the precise neurophysiological functions of DEG/ENaC channels remain elusive. This study demonstrates that a Drosophila DEG/ENaC-encoding gene, ppk29, is required for normal synaptic functions. However, in contrast to the ppk11/ppk16 complex (Younger, 2013), ppk29 action is restricted to the postsynaptic site and is associated with baseline spontaneous neurotransmission but not PhTX-dependent synaptic homeostasis. Therefore, it is proposed that individual DEG/ENaC-like channels may play independent roles in regulating synaptic functions, which may explain some of the contradicting reports about their functions in the mammalian synapse (Hill, 2017).

Previous studies have shown that some DEG/ENaC-encoding genes are expressed in human skeletal muscles, with their function remaining unknown. In Caenorhabditis elegans, the DEG/ENaC-encoding gene unc-105 is also expressed in muscles, where it is important for proper muscle organization, growth, and contraction. The current data indicate that, in addition to the contributions of DEG/ENaC proteins to muscle development and physiology, they also contribute to excitatory neurotransmission via postsynaptic processes in muscle. Nonetheless, because the fly NMJ is glutamatergic, the findings presented in this study could also provide important insights about postsynaptic functions of DEG/ENaC signaling in glutamatergic central synapses of vertebrates (Hill, 2017).

The postsynaptic impact of ppk29 on excitatory signaling may be mediated directly by PPK29 or indirectly via interaction with other proteins. Since currently available tools did not allow localization of PPK29 to specific subcellular compartments, it is too early to conclude whether the observed phenotypes in the ppk29 mutants are the consequence of a direct synaptic function or possibly of an indirect function via its action in other subcellular domains. It is also noted the that PPK29 may play more than one role in muscles and, therefore, that the electrophysiological and behavioral data may not be mediated through the same mechanism (Hill, 2017).

Nevertheless, one intriguing way that PPK29 might directly contribute to synaptic transmission is by acting as a direct postsynaptic receptor for glutamate or other molecules that are coreleased during spontaneous excitatory neurotransmission. For example, in mouse brain slices, extracellular protons increase with the stimulation of glutamatergic inputs, which can activate acid-sensitive ion channels, which are also members of the DEG/ENaC family (Du, 2014). Therefore, although it is not known whether PPK29 is an acid-activated channel, the corelease of protons with glutamate during spontaneous neurotransmission at the Drosophila NMJ may directly activate PPK29 channels (Hill, 2017).

The ppk29 gene may also affect neurotransmission indirectly through the modulation of expression or function of other synaptic proteins. Postsynaptic glutamate receptors are promising candidates to mediate an indirect impact of ppk29 on synaptic physiology since these ionotropic receptors are the main mediators of excitatory neurotransmission at the larval NMJ. This study found that ppk29 mutant animals exhibit decreased spontaneous amplitude and current flow, suggesting altered function of the postsynaptic ionotropic GluRs, which may be mediated by direct interaction between PPK29 and the GluR complex, or by indirect interaction via other postsynaptic signaling mechanisms. In line with this hypothesis, direct physical interactions between other DEG/ENaC proteins and potassium channels and sodium/chloride cotransporters have been reported in mammalian systems. It is further hypothesized that the observed differences in GluR expression levels are due to compensatory transcriptional changes. It is not known whether the ppk29 mutation independently impacts both GluRIIA and GluRIIB. Yet, previous studies have shown that genetic manipulation of expression levels of either GluRIIA or GluRIIB affect expression levels of the other (Marrus, 2004); therefore, the ppk29 mutation may directly or indirectly impact one or both subunit types (Hill, 2017).

To date, studies of spontaneous neurotransmitter release at the Drosophila larval NMJ have suggested that its main function is to regulate the development and maintenance of excitatory synaptic transmission by regulating presynaptic morphology and the postsynaptic clustering of glutamate receptors. However, spontaneous neurotransmitter release at central synapses has also been shown to impact local protein synthesis in dendrites as well as dendritic summation of EPSPs at much shorter timescales. This study has identified an important function for DEG/ENaC channels at the physiological timescale, which has an impact on both neurophysiological and behavioral phenotypes. Although it is not known how the molecular action of PPK29 might affect synapses and behavior, it is argued that these findings about the contribution of DEG/ENaC-encoding genes to spontaneous excitatory neurotransmission at the Drosophila larval NMJ may serve as an excellent model for understanding the function of spontaneous baseline excitatory neurotransmission in regulating synaptic organization. Better understanding of these processes at the physiological timescale is essential for understanding behavioral and neural plasticity in health and disease (Hill, 2017).

Synapse-specific and compartmentalized expression of presynaptic homeostatic potentiation

Postsynaptic compartments can be specifically modulated during various forms of synaptic plasticity, but it is unclear whether this precision is shared at presynaptic terminals. Presynaptic Homeostatic Plasticity (PHP) stabilizes neurotransmission at the Drosophila neuromuscular junction, where a retrograde enhancement of presynaptic neurotransmitter release compensates for diminished postsynaptic receptor functionality. To test the specificity of PHP induction and expression, this study has developed a genetic manipulation to reduce postsynaptic receptor expression at one of the two muscles innervated by a single motor neuron. PHP can be induced and expressed at a subset of synapses, over both acute and chronic time scales, without influencing transmission at adjacent release sites. Further, homeostatic modulations to CaMKII, vesicle pools, and functional release sites are compartmentalized and do not spread to neighboring pre- or post-synaptic structures. Thus, both PHP induction and expression mechanisms are locally transmitted and restricted to specific synaptic compartments (Li, 2018).

Although the genes and mechanisms that mediate retrograde homeostatic potentiation have been intensively investigated, whether this process can be expressed and restricted to a subset of synapses within a single neuron has not been determined. This study has developed a manipulation that enables the loss of GluRs on only one of the two postsynaptic targets innervated by a Type Ib motor neuron at the Drosophila NMJ. The analysis of synaptic structure and function in this condition has revealed the spectacular degree of compartmentalization in postsynaptic signaling and presynaptic expression that ultimately orchestrate the synapse- specific modulation of presynaptic efficacy (Li, 2018).

Compartmentalization of postsynaptic PHP signaling GluRs are dynamically trafficked in postsynaptic compartments where they mediate the synapse-specific expression of Hebbian plasticity such as LTP and homeostatic plasticity, including receptor scaling. In contrast, homeostatic plasticity at the human, mouse, and fly NMJ is expressed through a presynaptic enhancement in neurotransmitter release, but is induced through a diminishment of postsynaptic neurotransmitter receptor functionality. Using biased expression of Gal4 to reduce GluR levels on only one of the two muscle targets innervated by a single motor neuron, this study demonstrates that the inductive signaling underlying PHP is compartmentalized at the postsynaptic density, and does not influence activity at synapses innervating the adjacent muscle (Li, 2018).

Postsynaptic changes in CaMKII function and activity have been associated with PHP retrograde signaling. Consistent with this compartmentalized inductive signaling, this study observed pCaMKII levels to be specifically reduced at postsynaptic densities of Ib boutons in which GluR expression is perturbed, while pCaMKII was unchanged at postsynaptic compartments opposite to Is boutons and at NMJs in the adjacent muscle with normal GluR expression. Further, postsynaptic overexpression of the constitutively active CaMKII occludes the expression of PHP. Similar synapse-specific control of postsynaptic CaMKII phosphorylation, modulated by activity, has been previously observed. As noted in other studies, this localized reduction in pCaMKII provides a plausible mechanism for the inductive PHP signaling restricted to and compartmentalized at Ib synapses (Li, 2018).

How does a perturbation to GluR function lead to a reduction in CaMKII activity that is restricted to postsynaptic densities opposing Type Ib boutons? Recent evidence suggests that distinct mechanisms regulate pCaMKII levels during retrograde PHP signaling depending on pharmacologic or genetic perturbation to glutamate receptors and the role of protein synthesis. Scaffolds at postsynaptic densities are associated in complexes with GluRs and CaMKII. Intriguingly, the scaffold dCASK is capable of modulating CaMKII activity at specific densities in an activity-dependent fashion. Further, CaMKII activity can regulate plasticity with specificity at subsets of synapses in Drosophila and other systems. Although intra-cellular 'cross talk' between Is and Ib boutons cannot be ruled out, as GluRIIA is reduced at postsynaptic sites of both neuronal subtypes, it is striking that reductions in pCaMKII are restricted to Ib postsynaptic compartments. An attractive model, therefore, is that the postsynaptic density isolates calcium signaling over chronic time scales to compartmentalize PHP induction. The membranous complexity and geometry of the SSR at the Drosophila NMJ may be the key to restricting calcium signaling at these sites, as this structure can have major impacts on ionic signaling during synaptic transmission. These properties, in turn, may lead to local modulation of CaMKII function. Interestingly, Drosophila mutants with defective SSR elaboration and complexity have been associated with defects in PHP expression. In the mammalian central nervous system, it is well established that dendritic spines function as biochemical compartments that isolate calcium signaling while enabling propagation of voltage changes, and it is tempting to speculate that the SSR may subserve similar functions at the Drosophila NMJ to enable synapse-specific retrograde signaling (Li, 2018).

The homeostatic modulation of presynaptic neurotransmitter release is compartmentalized at the terminals of Type Ib motor neurons. It was previously known that PHP can be acutely induced and expressed without any information from the cell body of motor neurons. The current data suggests that the signaling necessary for PHP expression is even further restricted to specific postsynaptic densities and presynaptic boutons, demonstrated through several lines of evidence. First, quantal content is specifically enhanced at boutons innervating muscle 6 in M6>GluRIIARNAi without measurably impacting transmission on the neighboring boutons innervating muscle 7. In addition, PHP can be acutely induced at synapses innervating muscle 7 despite PHP having been chronically expressed at muscle 6. Finally, the homeostatic modulation of the RRP and enhancement of the functional number of release sites is fully expressed regardless of whether PHP is induced at all Type Ib boutons or only a subset. Thus, PHP signaling is orchestrated at specific boutons according to the state of GluR functionality of their synaptic partners and does not influence neighboring boutons within the same motor neuron. Although the compartmentalized expression of PHP was not unexpected, there was precedent to suspect inter-bouton crosstalk during homeostatic signaling. In the dynamic propagation of action potentials along the axon, the waveform could, in principle, change following PHP expression to globally modulate neurotransmission at all release sites in the same neuron. However, voltage imaging did not identify any change in the action potential waveform at individual boutons following PHP signaling, and this study did not observe any impact on neighboring boutons despite PHP being induced at a subset of synapses in the same motor neuron. Further, mobilization of an enhanced readily releasable synaptic vesicle pool is necessary for the expression of PHP, and synaptic vesicles and pools are highly mobile within and between presynaptic compartments. Hence, it was conceivable that a mobilized RRP, induced at some presynaptic compartments, may be promiscuously shared between other boutons. However, while a large enhancement was observed in the RRP at synapses innervating muscle 6 in M6>GluRIIARNAi, this adaptation had no impact on the RRP at adjacent presynaptic compartments innervating muscle 7. Thus, PHP signaling is constrained to boutons innervating one of two postsynaptic targets and does not 'spread' to synapses innervating the adjacent target despite sharing common cytosol, voltage, and synaptic vesicles (Li, 2018).

What molecular mechanisms mediate the remarkable specificity of PHP expression at presynaptic compartments? One attractive possibility is that active zones themselves are fundamental units and act as substrates for the homeostatic modulation of presynaptic function. The active zone scaffold BRP remodels during both acute and chronic PHP expression (Weyhersmuller, 2011), and other active zone proteins are likely to participate in this remodeling. Indeed, many genes encoding active zone components are required for PHP expression, including the calcium channel cac and auxiliary subunit α2-δ, the piccolo homolog fife, the scaffolds RIM (Rab3-interacting Molecule) and RIM-binding protein (RBP), and the kainite receptor DKaiR1D. If individual active zones can undergo the adaptations necessary and sufficient for PHP expression, this would imply that PHP can be induced and expressed with specificity at individual active zones. Indeed, the BRP cytomatrix stabilizes calcium channel levels at the active zone, and also controls the size of the RRP, two key presynaptic expression mechanisms that drive PHP. Further, the recruitment of new functional release sites have been observed following both chronic and acute PHP expression, suggesting that previously silent active zones become 'awakened' and utilized to potentiate presynaptic neurotransmitter release (Li, 2018).

Interestingly, presynaptic GluRs, localized near active zones, are necessary for PHP expression and have the capacity to modulate release with specificity at individual active zones. Thus, active zones have the capacity to remodel with both the specificity and precision necessary and sufficient for compartmentalized PHP expression. If each active zone operates as an independent homeostat to adjust release efficacy in response to target-specific changes, how is information transfer at individual sites integrated to ensure stable and stereotypic 'global' levels of neurotransmission? One speculative possibility is that active zones at terminals of each neuron are endowed with a total abundance of material that is tightly controlled and sets stable global levels of presynaptic neurotransmitter release. Such active zone material may be 'sculpted' with considerable heterogeneity within presynaptic terminals, varying in number, size, and density. Consistent with such a possibility, mutations in the synaptic vesicle component Rab3 exhibit extreme changes in active zone size, number, and density, but stable global levels of neurotransmission. Within this global context, plasticity mechanisms may operate at individual active zones, superimposed as independent homeostats to adaptively modulate synaptic strength. In addition, there is intriguing evidence for the existence of 'nanocolumns' between presynaptic active zones and postsynaptic GluRs that form structural and functional signaling complexes (Biederer, 2017; Tang, 2016). One particularly appealing possibility, therefore, is that a dialogue traversing synaptic nanocolumns functions to convey the retrograde signaling and active zone remodeling necessary for PHP expression at individual release sites. Studies in mammalian neurons have revealed parallel links between the functional plasticity of active zones, including their structure and size, and the homeostatic modulation of neurotransmitter release. Such intercellular signaling systems are likely to modify synaptic structure and function to not only establish precise pre- and post-synaptic apposition during development, but also to maintain the plasticity necessary for synapses to persist with the flexibility and stability to last a lifetime (Li, 2018).

The HSPG glypican regulates experience-dependent synaptic and behavioral plasticity by modulating the non-canonical BMP pathway

Under food deprivation conditions, Drosophila larvae exhibit increases in locomotor speed and synaptic bouton numbers at neuromuscular junctions (NMJs). Octopamine, the invertebrate counterpart of noradrenaline, plays critical roles in this process; however, the underlying mechanisms remain unclear. This study shows that a glypican (Dlp) negatively regulates type I synaptic bouton formation, postsynaptic expression of GluRIIA, and larval locomotor speed. Starvation-induced octopaminergic signaling decreases Dlp expression, leading to increases in synapse formation and locomotion. Dlp is expressed by postsynaptic muscle cells and suppresses the non-canonical BMP pathway, which is composed of the presynaptic BMP receptor Wit and postsynaptic GluRIIA-containing ionotropic glutamate receptor. During starvation, decreases in Dlp increase non-canonical BMP signaling, leading to increases in GluRIIA expression, type I bouton number, and locomotor speed. These results demonstrate that octopamine controls starvation-induced neural plasticity by regulating Dlp and provides insights into how proteoglycans can influence behavioral and synaptic plasticity (Kamimura, 2019).

Food deprivation induces behavioral and synaptic plasticity in Drosophila larvae through octopamine signaling. This study demonstrates that octopamine signaling dynamically regulates Dlp expression at the NMJ during starvation. Dlp was found to inhibits postsynaptic GluRIIA expression, and starvation and octopamine function were found to reduce postsynaptic Dlp expression. The inhibition of Dlp results in increased GluRIIA expression, which this study proposes is maintained for several hours during starvation by the activation of the non-canonical BMP pathway. Increased GluRIIA expression causes increased type I bouton growth at the NMJ and increased starvation-dependent larval locomotion. It has previously been shown that during the non-starvation or normal development of NMJs, postsynaptic GluRIIA expression and type I bouton growth are regulated by distinct biochemical pathways (Sulkowski, 2016). GluRIIA/GluRIIB postsynaptic composition is regulated by non-canonical BMP signaling, while type I bouton growth is regulated independently by canonical BMP signaling. However, this study finds that during starvation-dependent growth, GluRIIA expression and type I bouton growth become interrelated. Starvation increases GluRIIA expression in a Dlp-dependent manner, and type I bouton growth requires GluRIIA expression. It is not certain why this change in regulation occurs, but it is noted that independent regulatory mechanisms allow more flexibility for NMJ formation during normal development, while starvation-dependent growth may simply require a quick and transient increase in synaptic activity that may not need to be as flexible. This study found that Dlp regulates starvation-dependent changes at least in part through the regulation of non-canonical BMP signaling. However, non-canonical BMP signaling is likely only one component of how Dlp regulates NMJ plasticity and larval locomotion. This study finds that GluRIIA amounts increase upon the suppression of Dlp even in wit null mutants, indicating that BMP signaling functions to maintain increases in GluRIIA, but not necessarily to induce it. Thus, Dlp must inhibit GluRIIA expression through a pathway that is independent of non-canonical BMP signaling. Furthermore, although starvation-dependent increases in type I boutons depend on Wit and GluRIIA, there is also a component to this increase that is Wit and GluRIIA independent. This suggests that Dlp may also regulate type I boutons through a GluRIIA-independent mechanism (Kamimura, 2019).

Octopamine is required to maintain the basal expression levels of postsynaptic Dlp, and thus it positively regulates Dlp expression under normal rearing conditions. However, after food deprivation, octopamine signaling turns into a negative regulator of Dlp expression during the early phase of starvation. In the late phase of starvation, octopamine signaling reverts back to being a positive regulator. Koon (2012) and Koon (2011) revealed that type II bouton formation was not only positively regulated by Octβ2R autoreceptors but also negatively regulated by Octβ1R autoreceptors. To explain the acute growth of type II bouton growth after starvation, they postulated that Octβ1R exhibited a higher affinity for octopamine than did Octβ2R. Under normal rearing conditions, high-affinity Octβ1R may be dominantly activated by low concentrations of octopamine. Under this condition, low-affinity Octβ2R may be weakly activated. Thus, basal levels of type II bouton growth are determined by the balanced activities of these antagonistic octopamine receptors. After food deprivation, high concentrations of octopamine released from type II endings may strongly activate positive Octβ2R receptors, leading to the acute growth of type II boutons. A similar mechanism may be involved in the regulation of Dlp expression. One possibility is that high-affinity octopamine receptors (HOctRs) and low-affinity octopamine receptors (LOctRs) at type I synaptic regions positively and negatively, respectively, regulate the postsynaptic localization of Dlp. If this is the case, then the basal postsynaptic levels of Dlp may be affected by the dominant activation of positive HOctRs. In the early phase of starvation, during which high levels of octopamine are present, postsynaptic Dlp levels may be decreased by the activation of negative LOctRs. In the late phase of starvation, octopamine levels may be decreased by the exhaustion of octopaminergic neurons, leading to the dominant activation of HOctRs again. This may result in increases in postsynaptic Dlp expression. Thus, the flexible roles of octopamine signaling may be explained by differences in the octopamine receptors mainly activated in each phase. Future studies are needed to identify the octopamine receptors that regulate Dlp expression (Kamimura, 2019).

It remains unclear how octopamine signaling at type I synaptic regions regulates postsynaptic Dlp levels. Broadie and collaborators revealed that ECM metalloproteinases (Mmp1 and Mmp2) and the tissue inhibitor of metalloproteinases (Timp) play critical roles in synapse formation and activity-dependent plasticity in the Drosophila NMJ (Dear, 2016; Dear, 2017; Shilts, 2017). Mmp2 from the presynaptic and postsynaptic compartments has been shown to negatively regulate synaptic Dlp levels at the NMJ, possibly by the proteolytic cleavage of Dlp core proteins (Wang, 2014). However, Mmp1 is known to associate with Dlp via heparan sulfate (HS) chains and positively regulate Dlp expression levels at type I endings (Dear, 2016). Octopamine signaling after food deprivation may activate the proteolytic activity of Mmp2 at type I synaptic regions, leading to a decrease in postsynaptic Dlp levels. Mmp 1 may be activated in the later phase of starvation, leading to an increase in postsynaptic Dlp levels. A previous study revealed that Timp regulated the activities of these MMPs at the NMJ to restrict trans-synaptic Gbb signaling (Shilts, 2017). Thus, complex cooperation among Mmps, Timp, and Dlp may be involved in the experience-dependent plasticity of the NMJ. The local translation of Dlp may also be involved, particularly in the later phase (Kamimura, 2019).

This study has revealed that Dlp suppressed the GluRIIA-mediated non-canonical BMP signaling pathway, which forms a positive feedback loop between presynaptic pMad accumulation and postsynaptic type A iGluR expression (Sulkowski, 2016). Thus, decreases in Dlp expression after food deprivation may acutely promote this signaling pathway, leading to rapid increases in type A iGluR and pMad accumulation at the NMJ. However, Dlp expression was increased during the late phase of starvation, which may have suppressed this BMP pathway. A positive feedback loop may lead to a runaway condition with no brakes. In this sense, Dlp appears to critically regulate the levels of this GluRIIA-mediated BMP signaling as a brake during starvation. How, then, does Dlp regulate non-canonical BMP signaling? Since Dlp exclusively localizes at postsynaptic regions, it may directly associate with GluRIIA and destabilize type A iGluRs at synapses. After food deprivation, the postsynaptic expression levels of Dlp decrease, and thus type A iGluRs may be stabilized, leading to the promotion of presynaptic local Wit signaling and pMad accumulation. It is also conceivable that Dlp inhibits the trans-synaptic association between presynaptic BMP receptor complexes and postsynaptic type A iGluRs proposed by Sulkowski (2016). Many of the functions of HSPGs are mediated by HS moieties, which bind with various signaling molecules in a sulfation pattern-dependent manner. The development and function of the NMJ are known to be regulated in a complex manner by HS co-polymerase (encoded by ttv), N-deacetylase/N-sulfotransferase (encoded by sfl), HS 6-O-sulfotransferase, and extracellular sulfatase (Sulf1) (Kamimura, 2019).

These findings suggest that various signaling pathways at the NMJ are intricately regulated by structural changes in HS chains. The structure of HS chains on Dlp may be altered under specific conditions such as starvation, and Dlp may selectively regulate a particular signaling pathway among multiple target pathways in an HS sulfation pattern-dependent manner. LAR is known to associate with Dlp and Sdc through HS chains at the developing NMJ, in which Dlp showed higher affinity for LAR than Sdc, possibly because of differences in HS structures. Presynaptic LAR is activated by postsynaptic Sdc, leading to type I bouton growth, whereas Dlp competitively inhibits this signaling. Therefore, the decrease observed in Dlp expression after food deprivation may disinhibit LAR signaling and promote type I bouton growth. In addition, Dlp has been shown to regulate Wg signaling through the HS moieties during NMJ development. Wg is secreted from presynaptic terminal and glia and participates in synaptic plasticity. Therefore, it is also possible that Wg and Dlp regulate the starvation-induced synaptic and behavioral plasticities. Furthermore, a recent study revealed that neurexin synaptic organizing proteins are expressed as HSPGs. HS modifications in presynaptic neurexins are critical for high-affinity interactions with the postsynaptic ligands neuroligins and the leucine-rich repeat transmembrane neuronal proteins (LRRTMs), and are essential for synapse development at the mouse hippocampus. Drosophila Neurexin (Dnrx) is also an HSPG and positively regulates larval locomotion speed and type I bouton growth at NMJs in an HS-dependent manner. Furthermore, it has been suggested that Dnrx, Neuroligin (Dnlg1), and Wit form signaling complexes at the NMJ. Therefore, presynaptic Dnrx and postsynaptic Dlp may competitively regulate type I bouton formation and larval locomotor activity as antagonistic HSPGs, in which the HS chains of Dlp may inhibit the Dnrx-Dnlg1 interaction in a sulfation pattern-dependent manner. Further studies are needed to examine whether these pathways and HS modifications regulate starvation-induced synaptic and behavioral plasticities (Kamimura, 2019).

Only three type II octopaminergic neurons bilaterally innervate most of the body wall muscles in each abdominal segment, suggesting that type II octopaminergic neurons globally regulate the properties of type I terminals, coordinating the activities of the body wall musculature. This is similar to the mammalian brain noradrenergic system. Noradrenaline-producing neurons mainly cluster in a small nucleus called the locus coeruleus, which is located in the pons with global projections throughout the brain. The locus coeruleus noradrenergic system is activated by many stressors. For example, noradrenaline is released in many brain regions during emotional arousal and enhances long-term memory, an extreme example of which is post-traumatic stress disorder (PTSD). It has been shown that that noradrenaline facilitated the synaptic delivery of GluR1-containing AMPA receptors in the hippocampus, which lowered the threshold for long-term potentiation (LTP), thereby enhancing learning and memory. This is similar to the starvation-induced synaptic localization of type A iGluRs in the Drosophila NMJ, which is regulated by Dlp in an octopamine-dependent manner. Thus, further studies are needed to investigate whether the emotional enhancement of memory is regulated by glypican in the hippocampus. A previous study reported that astrocyte-secreted glypican 4 induced the release of the AMPA receptor clustering factor neuronal pentraxin 1 from presynaptic terminals through the signaling of RPTPδ, a LAR family receptor protein tyrosine phosphatase. This induced the formation of active synapses by recruiting AMPA receptors to postsynaptic sites. In addition, noradrenaline has been shown to activate astrocyte networks in the cortex and enhance their responsiveness to local changes in neuronal activity. Therefore, noradrenaline may stimulate the secretion of glypican 4 from astrocytes in a neuronal activity-dependent manner, thereby enhancing active synapse formation. In this case, the direction of glypican 4 activities appears to be opposite that of Dlp. However, glypicans often display opposing activities in a context-dependent manner. Thus, the interplay among Dlp, type A iGluR, LAR, and octopamine at the Drosophila NMJ may become a simple model to study the molecular mechanisms underlying emotion-enhanced memory formation and thus PTSD (Kamimura, 2019).

In the adult mammalian brain, the synapses of a subset of neurons are enwrapped with a specialized chondroitin sulfate proteoglycan-rich ECM called the perineuronal net (PNN). PNNs are mainly composed of tenascin-R, hyaluronan, and lectican family proteoglycans such as aggrecan, versican, and neurocan. The ECM enwrapping type I synaptic boutons at the Drosophila larval NMJ is an HSPG-based ECM containing Dlp, syndecan, perlecan, laminin, and the secreted lectin Mind-the-Gap. The orthologs of hyaluronan synthase and lectican family proteoglycans are not found in Drosophila, and therefore an HSPG-based ECM may be used at the synapses of Drosophila. In spite of their different compositions, both synaptic ECMs play critical roles in experience-induced neural plasticity, as shown by the present study and previous studies on mammalian ocular dominance plasticity. This study has revealed that Dlp regulated the localization of type A iGluRs during starvation. PNNs also contribute to the synaptic localization of AMPA receptors. Detailed comparative studies are needed to reveal functional and structural similarities between the PNNs and the ECM of the Drosophila NMJ (Kamimura, 2019).

Abnormalities in the expression of glypicans have been linked to various neurological disorders, such as autism spectrum disorders (ASDs), schizophrenia, attention-deficit/hyperactivity disorder (ADHD), and neuroticism. Dysfunctions in the noradrenergic system have also been reported in ASDs, schizophrenia, and ADHD. Thus, studies on the octopamine-mediated control of Dlp expression may contribute to a deeper understanding of the molecular mechanisms responsible for the pathogeneses of these disorders (Kamimura, 2019).

Rapid homeostatic modulation of transsynaptic nanocolumn rings

Robust neural information transfer relies on a delicate molecular nano-architecture of chemical synapses. Neurotransmitter release is controlled by a specific arrangement of proteins within presynaptic active zones. How the specific presynaptic molecular architecture relates to postsynaptic organization and how synaptic nano-architecture is transsynaptically regulated to enable stable synaptic transmission remain enigmatic. Using time-gated stimulated emission-depletion microscopy at the Drosophila neuromuscular junction, it was found that presynaptic nanorings formed by the active-zone scaffold Bruchpilot (Brp) align with postsynaptic glutamate receptor (GluR) rings. Individual rings harbor approximately four transsynaptically aligned Brp-GluR nanocolumns. Similar nanocolumn rings are formed by the presynaptic protein Unc13A and GluRs. Intriguingly, acute GluR impairment triggers transsynaptic nanocolumn formation on the minute timescale during homeostatic plasticity. Distinct phases of structural transsynaptic homeostatic plasticity were revealed, with postsynaptic GluR reorganization preceding presynaptic Brp modulation. Finally, homeostatic control of transsynaptic nano-architecture and neurotransmitter release requires the auxiliary GluR subunit Neto. Thus, transsynaptic nanocolumn rings provide a substrate for rapid homeostatic stabilization of synaptic efficacy (Muttathukunnel, 2022).

This study has identified a stereotypic arrangement of transsynaptically aligned molecular nanocolumns that is regulated in a modular and sequential fashion during homeostatic plasticity at the Drosophila NMJ. Moreover, a GluR subtype-specific nano-organization was revealed and it was discovered that the auxiliary GluR subunit Neto is required for rapid homeostatic modulation of transsynaptic nanocolumn number and neurotransmitter release (Muttathukunnel, 2022).

Previous work demonstrated that a cluster of voltage-gated Ca2+ channels localizes to the Brp ring center at the Drosophila NMJ. Furthermore, Unc13A, a molecule suggested as a molecular correlate of presynaptic release sites, forms ring-like arrays in close proximity to Brp C termini and GluRs. In light of these findings, the results are consistent with a model in which Ca2+ influx at the Brp/AZ center induces neurotransmitter release in the nanocolumn rings. Given that the neurotransmitter content released by a single synaptic vesicle does not activate all GluRs of a given PSD at the Drosophila NMJ and that Drosophila GluRs have a low glutamate affinity, neurotransmitter release may predominantly activate GluRs that are aligned to presynaptic release sites. Some evidence suggests that synaptic transmission predominantly occurs within transsynaptic nanocolumns. Hence, the transsynaptic nanocolumn rings discovered in this study may reflect subsynaptic transmission modules that are activated by a common Ca2+-channel cluster. Future work is needed to relate the molecular nanocolumn topography to synaptic physiology, for example, by assessing how many GluRs are activated by neurotransmitter release from a single synaptic vesicle. In this regard, the slight offset between Unc13A and GluR rings may indicate that a given release site may not only activate a single aligned GluR cluster but also neighboring GluR clusters, consistent with physiology data (Muttathukunnel, 2022).

GluR subunit composition and GluR location with regard to release sites are important factors determining synaptic efficacy. At the Drosophila NMJ, the ratio of slowly and rapidly desensitizing GluRIIA- and GluRIIB-containing receptors is a key regulator of quantal size. This study revealed that transsynaptic nanocolumns harbor a mix of GluRIIA- and GluRIIB-containing receptors, and that ambient receptors, which represent almost half of the GluRs within a PSD, mainly incorporate the GluRIIB subunit. The persistence of transsynaptic nanocolumn rings in GluRIIA and GluRIIB mutants implies that neither of these subunits alone is sufficient for ring formation or transsynaptic alignment. Previous work revealed no defects in spontaneous or AP-evoked synaptic transmission upon GluRIIA overexpression or after GluRIIB loss . Thus, two genetic manipulations that mainly decrease ambient receptor abundance, but not receptors inside the nanocolumn ring, do not induce a corresponding decrease in synaptic transmission. This indicates that synaptic transmission is largely confined to transsynaptic nanocolumn rings and/or that synaptic transmission outside the rings is dominated by rapidly desensitizing GluRIIB-containing receptors. Moreover, the observation of increased mEPSP amplitudes in GluRIIBSP5 mutants suggests that GluRIIB-containing receptors surrounding the nanocolumns have the potential to negatively regulate synaptic transmission by replacing GluRIIA-containing receptors within the nanocolumns (Muttathukunnel, 2022).

A variety of auxiliary subunits control GluR assembly, trafficking, and function. The auxiliary GluR subunit Neto has been implicated in GluR clustering at the Drosophila NMJ (Kim, 2021). This uncovered modular ring arrays of Neto-β that transsynaptically align with Brp C termini, suggesting that this auxiliary GluR subunit is a postsynaptic element of transsynaptic nanocolumn rings. The persistence of transsynaptic nanocolumn rings in hypomorphic neto109 mutants suggests that neto is not crucial for ring formation or transsynaptic alignment, or that the remaining Neto was sufficient for transsynaptic nanocolumn ring formation. In contrast to neto109 mutants, in which both Neto-α and Neto-β levels are reduced, loss of Neto-α does not decrease GluR levels or mEPSP amplitude, suggesting that this Neto isoform either does not stabilize GluRs at the Drosophila NMJ or that there is a compensation by Neto-β. While reduced levels of ambient receptors do not impair synaptic transmission in case of GluRIIA overexpression or in GluRIIBSP5 mutants, the decreased GluR abundance within the rings of neto109 mutants correlates with a decrease in spontaneous and AP-evoked synaptic transmission, again implying that synaptic transmission predominantly occurs within the rings (Muttathukunnel, 2022).

GluR impairment at the Drosophila NMJ induces a homeostatic increase in release, and there is evidence for the modulation of presynaptic nano-architecture during this form of homeostatic plasticity. A previous study reported increased GluR levels upon sustained pharmacological GluR inhibition for several days. This study demonstrates GluR modulation within 5 min after pharmacological GluR impairment that precedes the modulation of Brp, as well as Neto-β. Although it cannot be excluded that other molecules are modulated prior to GluRs, or that small changes in Brp or Neto-&beta could not be resolved; after PhTX treatment for 5 min, the data imply that GluR modulation precedes Neto-β and presynaptic regulation during homeostatic plasticity. Furthermore, GluR and Brp fluorescence intensity changes detected with confocal microscopy preceded the increase in GluR and Brp cluster numbers at STED resolution. This could either indicate that small nanostructural changes could not be detected with STED microscopy or that the modulation of transsynaptic nano-architecture lags behind the regulation of GluR and Brp levels or distribution. Similar to the data obtained with confocal microscopy, the increase in GluR cluster number preceded Brp cluster regulation upon GluR perturbation, again indicative of a temporal sequence of transsynaptic changes during PHP. Interestingly, while GluR, but not Brp cluster number increased 15 min after PhTX treatment, a larger fraction of transsynaptically aligned Brp clusters was noted. This suggests that transsynaptic nanocolumn formation likely precedes Brp cluster formation. The temporal sequence of GluR and Brp regulation may also explain the existence of GluR clusters within the ring that are not opposed by Brp. Together, these findings are consistent with a model of coordinated, transsynaptic, and modular structural plasticity during PHP that results in the addition of transsynaptic nanocolumns to the ring (Muttathukunnel, 2022).

Apparent changes in GluR fluorescence intensity, GluR cluster number, or homeostatic potentiation of release upon pharmacological GluR perturbation in hypomorphic neto109 mutants were not observed. This shows that wild-type Neto levels are required for homeostatic control of GluRs and presynaptic release. GluR inhibition also led to a slight but significant increase in Brp fluorescence intensity in neto109 mutants, which was less pronounced than in wild type. The defect in PHP seen in neto109 mutants could thus arise from impaired GluR and/or Brp regulation. Although the genetic data establish a causal relationship between the homeostatic regulation of transsynaptic nanocolumns and presynaptic release, future work is required to scrutinize the relationship between transsynaptic nano-architecture and synaptic transmission, and to dissect the molecular mechanisms controlling transsynaptic nano-architecture and its homeostatic regulation. In this regard, it will be exciting to explore which molecules are involved in transsynaptic alignment and ring formation. Synaptic cell-adhesion molecules, such as neurexins and neuroligins, represent obvious candidates (Muttathukunnel, 2022).

Postsynaptic cAMP signalling regulates the antagonistic balance of Drosophila glutamate receptor subtypes

The balance among different subtypes of glutamate receptors (GluRs) is crucial for synaptic function and plasticity at excitatory synapses. This study shows that the two subtypes of GluRs (A and B) expressed at Drosophila neuromuscular junction synapses mutually antagonize each other in terms of their relative synaptic levels and affect subsynaptic localization of each other. Upon temperature shift-induced neuromuscular junction plasticity, GluR subtype A increased but subtypeB decreased with a timecourse of hours. Inhibition of the activity of GluR subtype A led to imbalance of GluR subtypes towards more GluRIIA. To gain a better understanding of the signalling pathways underlying the balance of GluR subtypes, an RNA interference screen of candidate genes was performed and postsynaptic-specific knockdown of dunce, which encodes cAMP phosphodiesterase, was found to increase levels of GluR subtype A but decreased subtype B. Furthermore, bidirectional alterations of postsynaptic cAMP signalling resulted in the same antagonistic regulation of the two GluR subtypes. These findings thus identify a direct role of postsynaptic cAMP signalling in control of the plasticity-related balance of GluRs (Zhao, 2020).

A negative correlation between subtype A and B receptors has been reported previously at Drosophila NMJs. However, the mechanism by which the antagonistic balance of different subtypes of GluRs is regulated remains unclear. The present study revealed that bidirectional alterations of cAMP levels in the postsynaptic muscle cells alter the balance of GluR subtypes in a cell-autonomous manner. This study thus provides new insights into the mechanism underlying synaptic plasticity by altering the balance of GluR subtypes (Zhao, 2020).

Most previous conventional microscopy studies have reported substantial colocalization or differential localization of GluRIIA and GluRIIB. This study reports an apparently non-overlapping localization of GluRIIA and GluRIIB at the postsynaptic densities (PSDs) of NMJ synapses (see The subtype B forms a doughnut-shaped ring, with a smaller subtype A ring in the centre). Although clear interpretations are lacking for the distinct localization of GluRIIA and GluRIIB at PSDs, there could be two possibilities: either different classes of receptors might be associated with specific interacting proteins that could mediate, directly or indirectly, the concentric localization of GluR subtypes at PSDs or concentric rings of GluR subtypes A and B in wild-type larvae might associate with their specific biophysical properties. Desensitization is the process by which receptors are inactivated in the prolonged presence of an agonist; it occurs faster in response to a lower concentration of agonist. On the postsynaptic side, GluR subtype A exhibits slower desensitization kinetics than GluR subtype B. It is therefore speculated that the slower desensitization of subtype A receptors might be caused, in part, by a higher concentration of glutamate released on the presynaptic side, because subtype A rings are more closely juxtaposed to presynaptic Cacophony calcium channels than subtype B rings (Zhao, 2020).

This study showed that subtype A rings become enlarged (both the inner and outer ring diameters are increased) when the synaptic levels of GluRIIA are increased, whereas subtype B rings are enlarged in a specific manner, i.e., the inner diameter decreases, but the outer diameter of the ring increases when the level of GluRIIB is increased. A simple explanation for the enlarged GluR rings might therefore be increased synaptic levels of GluRIIA or GluRIIB. Given that GluR-enriched PSDs are confined to specific spatial domains by cell adhesion molecules and the spectrin-actin network, it is possible that an increase in the level of one subtype of GluR might take up the space left by a reduced level of the other (Zhao, 2020).

Synaptic plasticity is the ability of synapses to strengthen or weaken over time, in response to increases or decreases in their activity. It is well established that GluRs are involved in synaptic plasticity at excitatory synapses. However, it is not entirely known how different types or subtypes of GluRs are involved in synaptic plasticity. Drosophila glutamatergic NMJs with two subtypes of GluRs, rather than mammalian NMJs with multiple subtypes of GluRs, are an effective model for studying synaptic plasticity. Hyperexcitable double mutants of eag sh show persistent strengthening of larval NMJs, which represents long-term plasticity. This study found that increased presynaptic release by warm-activated TrpA1 led to increased GluRIIA but normal GluRIIB, which was consistent with increased GluRIIA in eag sh double mutants. In addition, increased GluRIIA but decreased GluRIIB were observed in high temperature-induced synaptic plasticity of long-term strengthening of neurotransmission. Thus, it is speculated that increased presynaptic release might result in long-term plasticity by enhancing postsynaptic responses through increased GluRIIA (Zhao, 2020).

Increased GluRIIA is consistently observed in different models of synaptic plasticity. However, this study observed normal and reduced GluRIIB in TrpA1- and high temperature-induced synaptic plasticity, respectively. The discrepant changes in GluRIIB levels in different models of synaptic plasticity could be caused by different timescales, such as a limited time (8 h) for elevating presynaptic release through activating TRPA1 versus 4 days for raising larvae at 27°C, or the change in the level of GluRIIB might be too low to be detected by overexpressing TRPA1, or both (Zhao, 2020).

Whether the antagonistic balance of GluRs is actively (as a functional requirement) or passively (as a physical competition) regulated depends on specific conditions. It appeared that GluRIIA and GluRIIB competed with each other for the essential subunits when the expression levels of either GluRIIA or GluRIIB were changed, consistent with previous reports. These results support a passive competition between GluRIIA and GluRIIB. However, an actively regulated antagonistic balance of GluRs also occurs. When the essential subunit GluRIIC, GluRIID or GluRIIE was limited, both GluRIIA and GluRIIB decreased. If only passive regulation of GluRIIA and GluRIIB occurs, GluR subtype A and B decreased at similar levels would be expected. However, the ratio of GluRIIA to GluRIIB increased, indicating that the GluRIIA subtype is maintained preferentially when the total GluRs are limited and supporting an active regulation of the balance between GluRIIA and GluRIIB. Given that GluRIIA is mainly responsible for the postsynaptic responses, the relative increase in GluRIIA when an essential subunit of GluRs was knocked down might be a functional compensation for the decrease of synaptic strength (Zhao, 2020).

In addition to the antagonism of GluRIIA and GluRIIB reported in this study, there are a few reports on the regulation of synaptic levels of single GluR subunits. For example, GluRIIA but not GluRIIB receptors are anchored at the PSD by the actin-associated Coracle (Chen, 2005a) and are regulated by a signalling pathway involving the Rho-type GEF (Pix) and its effector, Pak kinase (Albin, 2004). Recent studies also showed specific upregulation of GluRIIA but not GluRIIB when the calcium-dependent proteinase calpains were mutated (Zhao, 2020).

A previous study showed that the numbers of terminal varicosities and branches were increased in dnc but not rut mutants. Given that elevated cAMP levels induced an antagonistic balance of GluRs at the postsynaptic side, it was important to test whether the antagonistic balance of GluRs was associated with NMJ overgrowth. The results showed that the number of varicosities remained normal when dnc or rut was knocked down by RNAi in the postsynaptic muscles, suggesting that an alteration in the cAMP pathway at the postsynaptic side did not affect NMJ development (Zhao, 2020).

The importance of the GluR subtype balance in synaptic plasticity has been documented in mammals. The major forms of AMPA receptors in the hippocampus include GluA1/2 and GluA2/3 heteromers, in addition to GluA1 homomers. The relative abundance of GluA1- and GluA2-containing receptors is a well-established determinant of synaptic plasticity in diverse brain circuits; GluA1-containing receptors are recruited to synapses after long-term potentiation, whereas GluA2-containing receptors are required for long-term depression. Together with the mammalian findings, the results support the notion that the GluR subtype balance contributes to synaptic plasticity at excitatory synapses (Zhao, 2020).

It is widely known that cAMP signalling plays an important role in regulating synaptic plasticity by increasing presynaptic neurotransmitter release. However, it is not known whether the cAMP pathway acts postsynaptically in regulating the ratio of GluRs, which plays a crucial role in synaptic plasticity. In the present study, we showed, for the first time, that the cAMP pathway regulates the balance of different GluR subtypes on the postsynaptic side; either increased or reduced cAMP leads to an altered ratio of GluR subtypes at Drosophila NMJ synapses. Thus, an optimal level of cAMP in postsynaptic muscles might be required for the normal ratio of synaptic GluR subtypes (Zhao, 2020).

When cAMP levels are elevated, cAMP binds to the regulatory subunits of PKA and liberates catalytic subunits that then become active. Active PKA in muscles decreases the activity of GluRIIA in Drosophila (Davis, 1998). Thus, an increase in the level of synaptic GluRIIA might compensate for the reduced activity of GluRIIA caused by overexpression of wild-type or constitutively active PKA. Conversely, inhibition of PKA activity in muscles causes a significant increase in the average amplitude of miniature excitatory junctional currents, consistent with the notion that PKA negatively regulates the activity of GluRIIA. Surprisingly, an increase was observed in synaptic GluRIIA when the cAMP level was downregulated in rut mutants or when PKA was knocked down by RNAi. It appears that the negative regulation of GluRIIA activity by PKA is not sufficient to account for the increase of GluRIIA at NMJ synapses (Zhao, 2020).

Analysis of western blots showed that the protein level of GluRIIA increased significantly, regardless of whether postsynaptic cAMP pathway was up- (dnc RNAi and PKAOE) or downregulated (rut RNAi and PKA RNAi), suggesting that the similar antagonistic balance of GluR subtypes induced by both up- and downregulation of cAMP might be caused by an elevated protein level of GluRIIA the cAMP pathway regulates the antagonism between GluRIIA and GluRIIB at two distinct steps, GluRIIA activity and protein level. Exactly how bidirectional changes of cAMP lead to a similar alteration of GluR subtypes remains to be investigated (Zhao, 2020).

Although Dnc and Rut regulate cAMP levels in opposite directions, physiological studies in Drosophila have shown that activity-dependent short-term plasticity is altered in a similar manner at larval NMJs in both dnc and rut mutants. Specifically, synaptic facilitation and post-tetanic potentiation are both weakened, indicating that the bidirectional change of cAMP signalling might result in similar abnormalities in synapse plasticityn. The mechanisms underlying synaptic facilitation and post-tetanic potentiation are exclusively presynaptic. Synaptic facilitation and post-tetanic potentiation both result from increased presynaptic calcium concentrations, leading to an enhanced release of neurotransmitters. A bell-shaped model was proposed to explain this mode of regulation, i.e. mutations in dnc and rut, which regulate cAMP levels in opposite directions, result in a similar plasticity phenotype. It is proposed that the bell-shaped model might also explain a similar increase in GluRIIA at NMJ synapses caused by bidirectional changes in cAMP levels in postsynaptic muscles (Zhao, 2020).

The antagonistic balance of GluRIIA and GluRIIB is induced by the postsynaptic cAMP/PKA pathway. However, whether the antagonism between GluRIIA and GluRIIB requires the cAMP/PKA pathway is unclear. GluRIIA or GluRIIB nulls were recombined with postsynaptic RNAi knockdown of PKA (i.e. inhibition of the cAMP pathway). It is noted that PKA null mutants are lethal at the first larval stage and thus cannot be used for the genetic interaction assay. Compared with simple null mutants of GluRIIA (or GluRIIB), PKA RNAi in the mutant background of GluRIIA (or GluRIIB) did not change the synaptic levels of GluRIIB (or GluRIIA), suggesting that the antagonistic balance of GluRIIA and GluRIIB does not require the cAMP pathway at the postsynaptic side. Thus, an altered cAMP pathway leads to the antagonistic balance of GluRIIA and GluRIIB, but the antagonistic balance of GluRIIA and GluRIIB appears not to be dependent on the cAMP pathway, at least for the antagonism induced by null mutations of GluRIIA or GluRIIB, or the remaining PKA upon RNAi knockdown is sufficient to support the antagonistic balance of GluRIIA and GluRIIB (Zhao, 2020).

It will be of great interest to determine how the cAMP-PKA-GluR signalling pathway acts on the postsynaptic side to contribute to synaptic plasticity and whether this pathway is also effective and conserved in mammalian central synapses (Zhao, 2020).

Distinct molecular pathways govern presynaptic homeostatic plasticity

Presynaptic homeostatic plasticity (PHP) stabilizes synaptic transmission by counteracting impaired neurotransmitter receptor function through neurotransmitter release potentiation. PHP is thought to be triggered by impaired receptor function and to involve a stereotypic signaling pathway. However, this study demonstrates that different receptor perturbations that similarly reduce synaptic transmission result in different responses at the Drosophila neuromuscular junction. While receptor inhibition by the glutamate receptor (GluR) antagonist γ-D-glutamylglycine (γDGG) is not compensated by PHP, the GluR inhibitors Philanthotoxin-433 (PhTx) and Gyki-53655 (Gyki) induce compensatory PHP. Intriguingly, PHP triggered by PhTx and Gyki involve separable signaling pathways, including inhibition of distinct GluR subtypes, differential modulation of the active-zone scaffold Bruchpilot, and short-term plasticity. Moreover, while PHP upon Gyki treatment does not require genes promoting PhTx-induced PHP, it involves presynaptic protein kinase D. Thus, synapses not only respond differentially to similar activity impairments, but achieve homeostatic compensation via distinct mechanisms, highlighting the diversity of homeostatic signaling (Nair, 2021).

This study demonstrates that neurotransmitter receptor impairment by different GluR antagonists (PhTx and Gyki) induces PHP via distinct mechanisms, including differential inhibition of GluR subtypes, differential modulation of an active zone scaffold, and short-term plasticity during PHP. Importantly, genetic evidence is provided that separable molecular mechanisms promote PHP in response to Gyki and PhTx treatment. On the other hand, the GluR antagonist γDGG did not produce PHP, despite a robust inhibition of postsynaptic receptors. Together, the data suggest that distinct molecular mechanisms mediate PHP in response to GluR inhibition by different antagonists and that GluR inhibition per se is not sufficient for PHP expression (Nair, 2021).

The results contrast with prevalent models of homeostatic synaptic plasticity, which rest on the assumption that synaptic activity changes induce homeostatic signaling. In the case of PHP, the magnitude of the homeostatic increase in presynaptic release scales with the decrease in the amplitude of postsynaptic miniature events. Based on these data, it was proposed that reduced ion flux through the receptors triggers a homeostatic signaling cascade in the postsynaptic cell, which is relayed to the presynaptic compartment where it adjusts release. A prediction that directly follows is that any receptor perturbation with similar effects on the amplitude of synaptic miniature events produces a homeostatic response via similar underlying mechanisms. This study observed that two GluR antagonists, PhTx and Gyki, induced PHP, while one antagonist, γDGG, did not. All three perturbations similarly decreased quantal size (q), indicating a similar reduction of ion flux through the receptors. Although it cannot be excluded that γDGG may have directly blocked PHP expression, the results suggest that ion flux is unlikely the sole signal responsible for PHP induction at the Drosophila NMJ. This agrees with recent observations at the Drosophila and mouse NMJ indicating that Ca2+ flux through the receptor is likely dispensable for PHP induction (Goel, 2017; Wang, 2018). How could pharmacological receptor perturbation induce PHP independent of ion flux through the receptor? One intriguing possibility is that perturbation-specific conformational changes of the receptor may be involved in PHP signaling. Conformational changes of kainate-, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPA)-, and N-Methyl-d-aspartate (NMDA)-type GluRs at mammalian synapses are known to signal independent of ion flux in the context of synaptic plasticity, possibly through metabotropic signaling. It is also known that different antagonists stabilize different conformational states of AMPA receptors. Thus, it is conceivable that distinct antagonists may trigger different PHP signaling pathways, depending on the different signaling partners that could be recruited by different conformational states. This hypothesis requires further investigation in relation to molecular signaling underlying PHP induction (Nair, 2021).

Although PhTx and Gyki robustly induced PHP through an increase in RRP size, several differences were observed between PhTx- and Gyki-induced PHP. First, PhTx and Gyki differentially affected GluR subtypes. Whereas PhTx mainly inhibits GluRIIA-containing receptors, Gyki reduced mEPSP amplitude at either GluRIIA or GluRIIB mutant NMJs. Since mEPSPs at the Drosophila NMJs are mediated by GluRIIA- and GluRIIB-containing receptors, these results indicate that Gyki may act on both types of receptor complexes. An intriguing possibility to be tested in the future is that PhTx-dependent receptor inhibition may trigger PHP by affecting a signaling module associated with GluRIIA-containing receptor complexes, whereas Gyki-induced PHP may be triggered by mechanisms beyond GluRIIA-dependent signaling. Second, while Gyki similarly accelerated the decay of miniature excitatory postsynaptic currents (mEPSCs) and EPSCs, the EPSC decay was slower than that of mEPSCs upon PhTx application. The mismatch between the mEPSC and EPSC decay kinetics after PhTx treatment may either result from presynaptic changes, such as the recruitment of synaptic vesicles with a low pr during the EPSC decay phase, or different effects on GluR desensitization, saturation, or diffusion during evoked release. Regardless, as this mismatch was not observed after Gyki treatment, it either reflects a difference in antagonist action or in Gyki- and PhTx-induced PHP. Third, Gyki, but not PhTx, resulted in altered short-term plasticity, indicative of a decrease in pr or increased GluR desensitization and/or saturation. Fourth, while PhTx increased Brp-fluorescence intensity, indicating elevated Brp abundance, Gyki did not. Although it cannot be excluded that Brp modulation at a different time point with regard to Gyki application, this observation suggests differential regulation of this core active zone protein during PhTx- and Gyki-induced PHP. Fifth, this study revealed that Gyki application induced PHP in six mutants that were previously shown to block PHP upon PhTx treatment. Sixth, this study identified a gene (PKD) that is required for Gyki-induced PHP, but not PhTx-induced, PHP. Thus, although the possibility cannot be excluded of partially overlapping molecular mechanisms, these data demonstrate separable molecular pathways between Gyki- and PhTx-induced PHP (Nair, 2021).

Diverging homeostatic signaling has been observed in the context of acute versus chronic receptor perturbations. At the Drosophila NMJ, mutants with intact acute PHP in response to PhTx application displayed impaired chronic PHP upon GluRIICRNAi expression. However, it is difficult to separate whether these differences emerge because of differences in the nature or the timescale of receptor perturbation. Since the curremt experiments focus on acute perturbations, the results support the idea that distinct homeostatic signaling pathways can be triggered depending on the specific way receptors have been perturbed. PhTx, γDGG, and Gyki inhibit AMPA receptors via different modes of antagonism. While PhTx is an activity-dependent pore blocker, γDGG acts as a low-affinity competitive antagonist and Gyki is an allosteric inhibitor. PhTx and γDGG inhibit Drosophila GluRs with characteristics similar to AMPA receptor inhibition. Similarly, the conserved Gyki-interacting residues between rat and Drosophila GluRs highlight the possibility that Gyki could also act similarly on Drosophila and mammalian GluRs. Thus, PhTx, γDGG, and Gyki may inhibit Drosophila GluRs via different modes of antagonism, which may be involved in triggering distinct homeostatic signaling pathways. Differential homeostatic signaling has also been observed in the context of firing rate homeostasis, where molecular responses depend on whether the protein or the conductance of a sodium channel protein is eliminated. Thus, distinct molecular signaling mechanisms that are specific to the nature, rather than the functional effects of a perturbation, may be a general theme in the context of neuronal homeostasis (Nair, 2021).

This study also implicated presynaptic PKD signaling in synaptic homeostasis. Although PKD has been linked to various intracellular processes, such as vesicle sorting, endocytosis, and the regulation of the actin cytoskeleton, its synaptic function is less explored. Recently, PKD was associated with synaptic plasticity (Oueslati Morales, 2020). The current data clearly demonstrate that presynaptic PKD is required for Gyki-induced PHP, but not for PhTx-induced PHP and establish that separable molecular pathways govern Gyki- and PhTx-induced PHP. However, PKD was required for PHP only at low extracellular Ca2+ concentration, similar to a number of genes supporting PhTx-induced PHP. This indicates that PKD signaling may contribute to the robustness of PHP under conditions of low pr. It is worth highlighting that the PHP defect observed upon presynaptic PKD knockdown and in PKD hypomorphs may not be fully penetrant, likely because of the genetic nature of the perturbation, and/or genetic compensation by other members of the Protein Kinase C/Calmodulin-Dependent Kinase family. This genetic compensation is particularly pronounced in PKD null mutants, which precluded the analysis after complete loss of PKD function (Nair, 2021).

Unlike PhTx, Gyki induced PHP in the fully dissected NMJ preparation, which is more amenable to electrophysiology and imaging approaches, and thus allowed probing the dynamics of PHP induction. It is currently unclear why PhTx-induced PHP cannot be observed in the full preparation. One hypothesis is that the muscles/synapses may be stretched in the full preparation, which could lead to a disruption of signaling domains relevant for PhTx-induced PHP. A second possibility could be the potential disruption of neuro-glial signaling that is important for PhTx-induced PHP in the full preparation (Wang, 2020). The fact that Gyki results in PHP after full dissection demonstrates that PHP induction is not limited to the 'semi-intact' preparation per se. It will be interesting to investigate whether different molecular pathways promoting PHP in response to Gyki and PhTx treatment may be differentially affected by the type of the preparation. In this regard, Gkyi induced PHP in Multiplexin (dmp) mutants, which were shown to disrupt PhTx-dependent PHP and neuro-glia signaling (Wang, 2020; Nair, 2021 and regerences therein).

Gyki application to the fully dissected preparation revealed very rapid, low-latency PHP induction kinetics, which maintained EPSP amplitudes constant despite Gyki-induced mEPSP amplitude reduction. However, due to the sampling interval of 35 s, shorter latencies cannot be ruled out. In addition, the exact latency of PHP could not be estimated because of the relatively slow kinetics of mEPSP reduction in these experiments. The reversible GluR block by Gyki also revealed rapid, low-latency PHP reversal after Gyki washout, consistent with previous observations at mouse cerebellar mossy fiber boutons or at mouse NMJs. The results imply continuous, bidirectional PHP signaling that compensates for receptor impairment within ∼35 s, similar to observations at the mouse NMJ. It will be interesting to explore the molecular mechanisms underlying the induction of distinct PHP pathways and whether distinct molecular pathways also rapidly stabilize synaptic efficacy at other synapses (Nair, 2021).

Although the results indicate that PHP signaling depends on specific receptor perturbations, the mechanisms of antagonism of the used antagonists have not been studied for Drosophila GluRs. Moreover, the lack of PHP upon γDGG treatment may result from off-target inhibition of PHP signaling cannot be excluded, nor that Gyki acts independent of GluRs to induce PHP. However, given Gyki's known mechanism of allosteric antagonism of mammalian AMPARs, and its potent reduction of mEPSP amplitudes at the Drosophila NMJ, this is considered unlikely. Furthermore, given the evolutionary conservation of Gyki-interacting residues in Drosophila GluR subunits, it is likely that Gyki acts on all Drosophila GluR subunits. However, the effect of Gyki on individual receptor subunits cannot be tested, because GluRIIC-E subunits are essential for receptor formation. Homozygous Drosophila lines harboring mutations in genes required for PhTx-induced PHP were used to test whether PhTx and Gyki activate different downstream signaling pathways. This analysis did not systematically investigate the effects of potential off-site mutations that may have accumulated over time in these fly lines. However, most strains were kept over a balancer chromosome, and Gyki-induced PHP proceeded normally in six mutants that were previously shown to disrupt PhTx-induced PHP, suggesting that a major contribution of off-target mutations is unlikely (Nair, 2021).


Postsynaptic translation affects the efficacy and morphology of neuromuscular junctions

Long-term synaptic plasticity may be associated with structural rearrangements within the neuronal circuitry. Although the molecular mechanisms governing such activity-controlled morphological alterations are mostly elusive, polysomal accumulations at the base of developing dendritic spines and the activity-induced synthesis of synaptic components suggest that localized translation is involved during synaptic plasticity. This study shows that large aggregates of translational components as well as messenger RNA of the postsynaptic glutamate receptor subunit DGluR-IIA are localized within subsynaptic compartments of larval neuromuscular junctions of Drosophila. Genetic models of junctional plasticity and genetic manipulations using the translation initiation factors eIF4E and poly(A)-binding protein showed an increased occurrence of subsynaptic translation aggregates. This was associated with a significant increase in the postsynaptic DGluR-IIA protein levels and a reduction in the junctional expression of the cell-adhesion molecule Fasciclin II. In addition, the efficacy of junctional neurotransmission and the size of larval neuromuscular junctions were significantly increased. These results therefore provide evidence for a postsynaptic translational control of long-term junctional plasticity (Sigrist, 2000).

Translational control is primarily exerted by regulation of the initiation step of translation, which appears to be controlled by the rate-limiting initiation factor eIF4E. In addition, the interaction of the 5' cap bound eIF4E with the 3' end of mRNAs through a complex of other initiation factors and the poly(A)-binding protein (PABP) has been shown to synergistically facilitate translation initiation. To assess the potential role of regulated translation during the development of the larval neuromuscular junctions (NMJs) in Drosophila, the subcellular expression pattern of eIF4E and PABP were analyzed in filet preparations of third instar larvae. Both antigens showed a weak and ubiquitous expression in the cytoplasm of all larval cells, and they colocalized in strongly immunopositive aggregates up to 2microm in length close to NMJs. The specific localization of eIF4E/PABP aggregates close to and partially overlapping with junctional profiles revealed that eIF4E/PABP aggregates are positioned subsynaptically within or adjacent to the subsynaptic reticulum (SSR). No evidence was found for presynaptic or axonal localization of such aggregates. Therefore, the almost exclusive subsynaptic distribution of the eIF4E/PABP aggregates within larval muscles indicates that there may be a functional relationship between NMJs and the appearance of nearby eIF4E/PABP aggregates (Sigrist, 2000).

Ultrastructural examinations of larval NMJs revealed polysomal accumulations within and close to the SSR. According to their variable size, subsynaptic location and frequency of detection, the larger of these polysomal clusters are likely to represent the eIF4E/PABP aggregates detected by light microscopy. In addition, smaller polysomal aggregates were widely distributed in discrete membranous compartments throughout the SSR, whereas presynaptic and axonal profiles were free of polysomes. It is therefore concluded that mRNAs are translated within subsynaptic compartments of larval NMJs and that local centres of concentrated, subsynaptic translation are identified by large junctional eIF4E/PABP aggregates (Sigrist, 2000).

To assess whether junctional translation is subject to regulation, the number was quantified of synaptic specializations (boutons) per NMJ that were labelled by one or more translation aggregates. Animals that overexpressed PABP in larval muscles and larvae that were mutant in pabp showed a significantly increased occurrence of subsynaptic eIF4E/PABP aggregates and an unaltered level of muscular PABP staining. In addition, the total PABP levels in crude larval protein extracts were unaltered in all analysed genotypes, even when PABP mRNA levels were significantly increased or reduced under genetic gain-of-function or loss-of-function conditions, respectively. Such a homeostasis of total PABP levels is a well described phenomenon for PABP, and in crude protein extracts it might have masked the significant local increase in the number of PABP aggregates observable within subsynaptic compartments of NMJs. Although the exact reason for this increase in the occurrence of eIF4E/PABP aggregates is unknown, a local perturbation of PABP levels owing to a previously described overshooting compensation of the PABP-homeostasis mechanism might facilitate formation of subsynaptic translation aggregates (Sigrist, 2000).

A similar increase in the frequency of postsynaptic translation aggregates was also observed in two mutants representing well established genetic models of long-term synaptic plasticity in Drosophila, the hyperactive K+-channel mutant eag, Sh and the cAMP-phosphodiesterase mutant dunce. Thus, increased neuronal activity levels (in eag, Sh) as well as elevated cellular cAMP levels (in dunce) are capable of inducing subsynaptic translation aggregate formation. These findings are consistent with the hypothesis that synaptic activity can control synaptic translation (Sigrist, 2000).

To identify potential substrates and targets of subsynaptic translation at larval NMJs, quantitative immunostainings were performed of several junctionally expressed proteins, including the synaptic vesicle protein synaptotagmin, the junctional anti-horseradish peroxidase (HRP) epitope, the cell-adhesion molecule Fasciclin II (FasII), the postsynaptic glutamate receptor subunit DGluR-IIA and the conventional myosin as a nonsynaptic protein. No obvious differences were detected in the expression levels of myosin, synaptotagmin and the junctional anti-HRP immunoreactivity in all analysed genotypes; however, animals that showed elevated numbers of subsynaptic translation aggregates consistently displayed increased junctional levels of DGluR-IIA and an altered junctional distribution of FasII, which was associated with a significant reduction of synaptic FasII levels as compared with control animals. A similar FasII phenotype has been described in the plasticity models eag, Sh and dunce, and it has been shown that presynaptic FasII downregulation is essential for increased junctional outgrowth. Intriguingly, in Aplysia the FasII homologue apCAM is also presynaptically downregulated after treatments that increase synaptic efficacy and growth of new synaptic connections. This synaptic apCAM regulation is thought to be achieved by a protein-synthesis-dependent activation of an endocytic apCAM internalization. Given that FasII has been detected in membranes of a subset of presynaptic vesicles, it seems possible that subsynaptic protein synthesis affects junctional FasII levels through similar mechanisms to those in Aplysia (Sigrist, 2000).

The postsynaptic DGluR-IIA immunoreactivities were significantly stronger in translationally sensitized animals. This strong increase of synaptic DGluR-IIA expression was not due to transcriptional upregulation of dglur-IIA; the total amounts of DGluR-IIA mRNAs were unaltered or even reduced in the analysed genotypes as compared with controls. In situ hybridization experiments revealed that DGluR-IIA mRNA surrounds individual type-I boutons, with prominent staining of terminal and branch-point boutons and weak or absent staining within the SSR of interbouton connectives. Thus, the subsynaptically localized DGluR-IIA mRNA represents a direct substrate for the junctional translation machinery. These results can not exclude an extrajunctional contribution to the observed synaptic DGluR-IIA increase, but they suggest that this phenotype is due to an increased subsynaptic synthesis of DGluR-IIA in genotypes with a higher occurrence of junctional eIF4E/PABP aggregates (Sigrist, 2000).

To analyse the functional consequences of genetically modified subsynaptic translation, the strength of neurotransmission at NMJs was assessed on muscle 6 of third instar larvae. The average amplitudes of miniature excitatory junctional currents (mEJCs) and thus the quantal sizes were indistinguishable in all analysed genotypes. This finding indicates either that the additional receptor subunits that are synaptically localized may be functionally silent (for example, through physiological silencing or intracellular localization or that the amount of glutamate released from an individual quantum is not sufficient to saturate the postsynaptic receptors. In contrast, postsynaptic responses evoked by stimulation of motor nerve axons were substantially larger in all mutants exhibiting increased levels of subsynaptic translation. Thus, the derived quantal content was significantly increased above control values, suggesting that the observed larger amplitudes of evoked junctional responses arise from an increased number of released presynaptic vesicles per action potential (Sigrist, 2000).

To investigate whether the increase in junctional efficacy was due to a change in the number of synaptic specializations, the number of junctional boutons per NMJ was quantified. Genotypes that displayed an increased occurrence of subsynaptic translation aggregates had significantly larger NMJs and reduced junctional FasII levels. In addition, the junctional sizes of the analysed animals correlated in a highly significant manner with their estimated quantal contents, suggesting that junctional efficacy and the morphological elaboration of NMJs are tightly coupled. On the basis of light microscopic examinations of DGluR-IIA labelled NMJs, the density of synapses within NMJs of all mutant animals appeared similar to that of controls or even higher, indicating that the total number of synapses increased proportionally with the junctional size. This finding indicates that the increased quantal content in animals with facilitated subsynaptic translation may be because of an increase in the number of vesicle release sites per given stimulus (Sigrist, 2000).

In summary, this study has shown that translational machinery and mRNAs are associated with the subsynaptic reticulum of NMJs and that genetic manipulations that affect the occurrence of subsynaptic translation aggregates are accompanied by changes in the levels of synaptic proteins, such as DGluR-IIA and FasII. These same manipulations also affected the function and morphology of NMJs, suggesting that subsynaptic translation can instruct junctional growth and synaptic reorganization and thereby long-term functional changes. These results further suggest that subsynaptic translation can be regulated by altered levels of neuronal activity, indicating that the regulation of postsynaptic translation participates in activity-dependent junctional plasticity. Thus, the inducible recruitment of postsynaptic protein synthesis appears to render individual synapses competent to instruct long-term changes in their functions and morphological organization. Given that localized protein synthesis has been shown to act in a synapse specific stabilization of long-term facilitation in central neurons of Aplysia, it emerges that synaptic translation might represent a common principle of long-term alterations of neuronal function and connectivity (Sigrist, 2000).

Nonvesicular release of glutamate by glial xCT transporters suppresses glutamate receptor clustering in vivo

It was hypothesized that cystine/glutamate transporters (xCTs) might be critical regulators of ambient extracellular glutamate levels in the nervous system and that misregulation of this glutamate pool might have important neurophysiological and/or behavioral consequences. To test this idea, a novel Drosophila xCT gene was identified and functionally characterized, that has been named 'genderblind' (gb). Genderblind is expressed in a previously overlooked subset of peripheral and central glia. Genetic elimination of gb causes a 50% reduction in extracellular glutamate concentration, demonstrating that xCT transporters are important regulators of extracellular glutamate. Consistent with previous studies showing that extracellular glutamate regulates postsynaptic glutamate receptor clustering, gb mutants show a large (200%-300%) increase in the number of postsynaptic glutamate receptors. This increase in postsynaptic receptor abundance is not accompanied by other obvious synaptic changes and is completely rescued when synapses are cultured in wild-type levels of glutamate. Additional in situ pharmacology suggests that glutamate-mediated suppression of glutamate receptor clustering depends on receptor desensitization. Together, these results suggest that (1) xCT transporters are critical for regulation of ambient extracellular glutamate in vivo; (2) ambient extracellular glutamate maintains some receptors constitutively desensitized in vivo; and (3) constitutive desensitization of ionotropic glutamate receptors suppresses their ability to cluster at synapses (Augustin, 2007).

The primary physiological role of xCT transporters remains controversial. Although xCT transporters mediate 1:1 exchange between extracellular cystine and intracellular glutamate, glutamate excretion is generally ignored, and xCT transporters are often assumed to function primarily as a cystine-uptake mechanism for glutathione synthesis and protection from oxidative stress. However, this bias ignores several important facts: (1) xCT transporters also export glutamate. (2) Mammalian brain xCT appears most abundant in 'border areas between the brain proper and periphery', specifically 'several regions facing the CSF,' including ventricle walls and meninges, consistent with the idea that xCT transporters are important for regulation of free glutamate content of CSF but not for cystine uptake in all brain cells. (3) Mammalian xCT transporters appear to be dispensable for cystine uptake and glutathione synthesis. Instead, glutathione synthesis in neurons and glia may be regulated by excitatory amino acid transport (EAAT) family proteins. EAATs are best known as sodium-dependent transporters for glutamate uptake, but EAATs also efficiently import cysteine, the reduced form of cystine used in glutathione synthesis. In agreement, overexpression of Drosophila gb (Tub-Gal4;UAS-gb) causes shortened lifespan and neurodegeneration, consistent with increased glutamate secretion but the exact opposite phenotype that one would expect if the role of GB were cystine uptake for neuroprotection. (4) Microdialysis of rat brains with inhibitors of xCT function leads to a decrease in nonvesicular glutamate secretion (Augustin, 2007).

Accordingly, it is argued that glutamate export by xCT transporters is at least as important as cystine import, particularly in the nervous system. Full acceptance of this idea, however, requires one to accept the idea that xCT transporters maintain ambient extracellular glutamate in the nervous system for good reasons and that extracellular glutamate in the brain is not merely a potentially pathological byproduct of glutamatergic transmission. The data suggest that ambient extracellular glutamate regulates constitutive receptor desensitization for control of synaptic glutamate receptor abundance (Augustin, 2007).

A link between glutamate receptor desensitization and clustering has not previously been demonstrated. It is well known that desensitization functionally eliminates glutamate receptors on a short time scale (tens to hundreds of milliseconds). The data suggest that constitutive desensitization is, on a longer time scale (hours), also associated with removal of receptors from the synapse. The EC50 for activation of Drosophila larval muscle glutamate receptors is ~2 mM, and significant numbers of receptors can be desensitized at considerably lower concentrations. Because 2 mM is near the concentration of glutamate bathing NMJ receptors in vivo, it must be concluded that one-half or more of Drosophila larval muscle glutamate receptors are constitutively desensitized, and therefore delocalized, in vivo. This conclusion is consistent with the 200%-300% increase in postsynaptic glutamate receptor abundance that was observe after switching NMJs to culture media containing 0 mM glutamate (Augustin, 2007).

At first, the idea that many glutamate receptors should be desensitized (and subsequently delocalized) in vivo seems surprising. However, constitutive desensitization (and subsequent delocalization) of ligand-gated ion channels by ambient ligand is analogous to constitutive inactivation of voltage-gated ion channels by resting membrane potential. Constitutive inactivation of voltage-gated channels is a common and important regulator of membrane excitability. For example, at a typical rat skeletal muscle resting potential of -90 mV, approximately two-thirds of rat skm-1 skeletal muscle sodium channels are inactivated. As a result, only one-third of channels in the membrane are normally available for generation of action potentials. However, if resting membrane potential is modified or the voltage dependence of sodium channel inactivation is slightly shifted by (for example) channel phosphorylation, then the number of functionally available sodium channels in the membrane can change quickly and dramatically, with consequent large effects on cell excitability. In the case of glutamate receptors, the number of functionally available receptors at a synapse, and therefore synaptic strength, could similarly be quickly and effectively altered by relatively minor changes in ambient glutamate levels (perhaps because of regulation of xCT-mediated transport) or changes in the concentration dependence of receptor desensitization as a result of (for example) receptor phosphorylation. These possibilities have not been explored (Augustin, 2007).

A physiological role for ambient extracellular glutamate also has medical implications. Abnormal levels of CSF glutamate have been linked to a variety of human neurodevelopmental and neurodegenerative disorders, including anxiety/stress-related disorders, Rett syndrome, autism, and all forms (both familial and sporadic) of amyotrophic lateral sclerosis. Furthermore, xCT and 4F2hc have specifically been implicated in development, behavior, and disease. For example, lysinuric protein intolerance, a recessive disorder characterized by severe mental retardation, is caused by mutations in the human xCT gene SLC7A7 [solute carrier family 7 (cationic amino acid transporter, y+ system), member 7]. Similarly, 4F2hc is required for tumor transformation in human cancers. Finally, human xCT protein was recently identified as the fusion-entry receptor for Kaposi's sarcoma-associated herpes virus. Not surprisingly, therefore, extracellular glutamate and xCT transporters are beginning to be targeted for pharmacological inhibition. These results suggest that pharmacological inhibition of xCT transport could considerably ameliorate neuropathologies exacerbated by extracellular glutamate but raise the caveat that tampering with extracellular glutamate could have unexpected developmental and/or psychotropic effects (Augustin, 2007 and references therein).

Glial wingless/Wnt regulates glutamate receptor clustering and synaptic physiology at the Drosophila neuromuscular junction

Glial cells are emerging as important regulators of synapse formation, maturation, and plasticity through the release of secreted signaling molecules. This study used chromatin immunoprecipitation along with Drosophila genomic tiling arrays to define potential targets of the glial transcription factor Reversed polarity (Repo). Unexpectedly, wingless (wg), encoding a secreted morphogen that regulates synaptic growth at the Drosophila larval neuromuscular junction (NMJ), was identified as a potential Repo target gene. Repo regulates wg expression in vivo, and local glial cells secrete Wg at the NMJ to regulate glutamate receptor clustering and synaptic function. This work identifies Wg as a novel in vivo glial-secreted factor that specifically modulates assembly of the postsynaptic signaling machinery at the Drosophila NMJ (Kerr, 2014).

The diversity of genes directly regulated by Repo-a critical transcriptional regulator of glial cell development in Drosophila-has not been thoroughly explored. ChIP studies from Drosophila S2 cells identified several potential Repo targets that have been shown to govern fundamental aspects of glial development or function. For example, known targets were identified that actively promote glial cell fate specification (e.g., pointed, distalless;, blood-brain barrier formation (e.g., gliotactin, loco, coracle, Nrv1, engulfment activity (e.g., dCed-6), neurotransmitter metabolism (e.g., EAAT1, Gs2), ionic homeostasis (e.g., fray), and neuron-glia signaling during nervous system morphogenesis (e.g., Pvr). For at least two potential targets, gs2 and Cp1, this study demonstrated a key requirement for Repo in their transcriptional activation during development (Kerr, 2014).

Given the broad roles of this collection of genes in glial cell biology, this work supports the hypothesis that Repo transcriptionally regulates a diverse class of genes that modulate many aspects of glial cell development. For instance, Pointed, which is now a predicted Repo target, is a key glial factor that activates glial fate at very early developmental stages. Likewise, Repo appears to regulate Gliotactin, Coracle, and Nrv1, which are molecules essential for formation of the pleated septate junction-based blood-brain barrier at mid to late embryogenesis in Drosophila. At the same time, EAAT1 and GS2 are activated late in the embryonic glial program, with expression being retained even in fully mature glia, and these transporters are critical for synaptic neurotransmitter recycling. Since EAAT1 and GS2 are both activated by Repo, and primarily expressed in CNS glia, these data argue that Repo is directly upstream of multiple key glial factors required for glutamate clearance from CNS synapses (Kerr, 2014).

Mammalian excitatory glutamatergic synapse formation is modulated by multiple soluble glia-derived factors including TSPs, Hevin/Sparc, and glypicans 4 and 6. These factors, along with other secreted glial factors that remain to be identified, are essential for initial synapse formation and (with the exception of TSPs) can promote postsynaptic differentiation through membrane insertion and clustering of AMPA receptors. This study identified Wg as a novel glia-derived factor essential for postsynaptic structure and function in vivo at the Drosophila glutamatergic NMJ. Combined with previous findings that NMJ glia can also release a TGF-β family member to regulate presynaptic growth in a retrograde manner (Fuentes-Medel, 2012), these studies provide compelling evidence that Drosophila glia function as a major integrator of synaptic signals during developmen (Kerr, 2014).

Previous work has demonstrated that Wg/Wnt signaling potently modulates the coordinated assembly of both presynaptic and postsynaptic structures at the Drosophila NMJ (Speese, 2007). Loss of Wg, or its receptor DFz2, leads to a dramatic decrease in synaptic boutons and disrupted clustering of postsynaptic glutamate receptors (Packard, 2002). Although previous studies supported evidence implicating motor neurons in Wg release, the presence of alternative cellular sources remained an open and important question. The surprising discovery of Wg as a candidate Repo target gene by ChIP led to an exploration of the possibility that NMJ glia could act as an additional in vivo source of NMJ Wg. Consistent with this idea, this study found that peripheral glia expressed Wg, SPGs were able to deliver Wg::GFP to the NMJ, and knockdown of SPG Wg significantly reduced NMJ Wg levels and led to a partial phenocopy of wg mutant phenotypes (Kerr, 2014).

Interestingly, it was found that loss of glia-derived Wg could account for some, but not all, wg loss-of-function phenotypes. For example, whereas depletion of glia-derived Wg disrupted clustering of postsynaptic glutamate receptors, it had no effect on the formation of synaptic boutons. In contrast, depletion of neuronal Wg led to defects in both glutamate receptor clustering as well as bouton formation. Although only neuronal Wg regulated bouton growth, these data argue that both glial and neuronal Wg are capable of modulating the assembly of glutamate receptor complexes. Thus, this study has identified two in vivo sources of Wg at the NMJ: the presynaptic neuron and local glial cells (Kerr, 2014).

Regarding the modulation of neurotransmission, both glial and neuronal Wg was found to have important roles, which, as in the case of the development of synaptic structure, were only partially overlapping. Loss of glial or neuronal Wg resulted in postsynaptic defects in neurotransmission, including increased mEJP amplitude (a postsynaptic property), decreased nerve-evoked EJPs, and decreased quantal content. Consistent with Repo regulating glial Wg expression, these phenotypes were mimicked by loss of repo function. The most notable difference in functional requirements for glial versus neuronal Wg is in mEJP frequency (a presynaptic function): depletion of glial Wg resulted in a dramatic increase in mEJP frequency, whereas manipulating neuronal Wg had no effect. Thus both glial and neuronal Wg are critical regulators of synaptic physiology in vivo, likely modulating NMJ neurotransmission in a combinatorial fashion, although glial Wg has the unique ability to modulate presynaptic function (Kerr, 2014).

The increase in mEJP amplitude is consistent with findings that GluR cluster size was increased upon loss of glia- or neuron-derived Wg, and that in general this was accompanied by minor changes in GluRIIA signal intensity. A potential explanation is that neuron- and glia-derived Wg regulate the levels of GluRIIA subunits. Previously, it was demonstrated that downregulation of the postsynaptic Frizzled Nuclear Import (FNI) pathway also increased GluRs at the NMJ (Speese, 2012). This suggests that glia- and neuron-derived Wg may act in concert via the FNI pathway to stabilize the synapse by regulating GluR expression (Kerr, 2014).

An important property of the larval NMJ is the ability to maintain constant synaptic function throughout development via structural and functional modifications. The combined functions of glial and neuronal Wg likely contribute to this mechanism, as together they positively regulate synaptic growth and function as well as organize postsynaptic machinery. However, a previous study suggested that the transcription factor Gooseberry (Gsb), in its role as positive regulator of synaptic homeostasis in neurons, may be antagonized by Wg function (Marie, 2010). Mutations in gsb block the increase in neurotransmitter release observed when postsynaptic GluRs are downregulated. Furthermore, Marie (2010) showed that the gsb mutant defect can be rescued by a heterozygous wg mutant allele. However, the specific role of Gsb in this process is unclear, as rapid synaptic homeostasis was normal in the mutant, and defects appeared restricted to a long-term decrease in GluR function. It will be important to define the specific role of Gsb in synaptic homeostasis and to manipulate Wg function in alternative ways before a clear relationship between Wg and Gsb can be established (Kerr, 2014).

How could neuronal versus glial Wg differ in regulating NMJ development and physiology? One possibility is that the level or site of Wg delivery by each cell type is different. For example, since SPGs invade the NMJ only intermittently (Fuentes-Medel, 2009), it is possible that they release most of their Wg outside of the NMJ, whereas the presynaptic neuron, which is embedded in the muscle cell, delivers it more efficiently and directly to the postsynaptic muscle cell. Alternatively, the Wg morphogen released by glia versus that released by neurons could be qualitatively different through alternative post-translational modifications such as glycosylation. Either mechanism would allow for glia to modulate specific aspects of NMJ physiology independently from neuronal Wg, perhaps in an activity-dependent manner (Kerr, 2014).

Although glia-derived Wg does not modulate NMJ growth, Drosophila glia can indeed regulate synaptic growth at the NMJ in vivo. It has been demonstrated previously that Drosophila glia release the TGF-β ligand Maverick to modulate TGF-β/BMP retrograde signaling at the NMJ and thereby the addition of new synaptic boutons (Fuentes-Medel, 2012). The discovery that glia-derived Wg can exert significant control over the physiological properties of NMJ synapses expands the mechanisms by which Drosophila glia can control NMJ synapse development and function. In the future it will be important to understand how glial Wg and TGF-β signaling integrate to promote normal NMJ growth, physiology, and plasticity (Kerr, 2014).

Anterograde Activin signaling regulates postsynaptic membrane potential and GluRIIA/B abundance at the Drosophila neuromuscular junction

Members of the TGF-beta superfamily play numerous roles in nervous system development and function. In Drosophila, retrograde BMP signaling at the neuromuscular junction (NMJ) is required presynaptically for proper synapse growth and neurotransmitter release. This study analyzed whether the Activin branch of the TGF-beta superfamily also contributes to NMJ development and function. Elimination of the Activin/TGF-beta type I receptor babo, or its downstream signal transducer smox, does not affect presynaptic NMJ growth or evoked excitatory junctional potentials (EJPs), but instead results in a number of postsynaptic defects including depolarized membrane potential, small size and frequency of miniature excitatory junction potentials (mEJPs), and decreased synaptic densities of the glutamate receptors GluRIIA and B. The majority of the defective smox synaptic phenotypes were rescued by muscle-specific expression of a smox transgene. Furthermore, a mutation in actβ, an Activin-like ligand that is strongly expressed in motor neurons, phenocopies babo and smox loss-of-function alleles. These results demonstrate that anterograde Activin/TGF-beta signaling at the Drosophila NMJ is crucial for achieving normal abundance and localization of several important postsynaptic signaling molecules and for regulating postsynaptic membrane physiology. Together with the well-established presynaptic role of the retrograde BMP signaling via Glass bottom boat and Wishful thinking, these findings indicate that the two branches of the TGF-beta superfamily are differentially deployed on each side of the Drosophila NMJ synapse to regulate distinct aspects of its development and function (Kim, 2014).

Numerous reports have now implicated the Activin/TGF-β and BMP branches of the TGF-β superfamily in regulating neuronal development, synaptic plasticity and cognitive behavior. Accordingly, members from both subfamilies are widely expressed in the nervous system and are co-expressed in multiple regions of vertebrate and invertebrate brains. It is therefore quite likely that ligands of both subfamilies co-exist within the extracellular space and in some cases, act on the same neurons. Lending support to this idea, pyramidal neurons in the CA3 region of the rat hippocampus are known to accumulate both phosphorylated Smad2 and Smad1/5/8, transcriptional transducers of the canonical Activin/TGF-β and BMP-type signaling, respectively. The activation of these two closely-related signaling pathways in common sets of neurons, or different cells of a common neuronal circuit raises the intriguing question of whether the two pathways play different or redundant roles during neuronal development and function (Kim, 2014).

This study utilized the Drosophila neuromuscular junction to address this issue since ligands of both Activin/TGF-β and BMP families are expressed in both muscle and motor neurons. The data, together with previous studies on the role of BMP signaling at the NMJ, clearly demonstrate that the two pathways influence NMJ synaptogenesis in different ways. The Activin/TGF-β pathway is necessary for achieving the proper densities of GluRIIA, GluRIIB and Dlg in postsynaptic muscle membrane, while the BMP pathway has a smaller effect on the distribution of these postsynaptic proteins. In addition, the Activin/TGF-β pathway was dispensable for maintaining overall synaptic growth and homeostasis, both of which are strongly affected by mutations in the BMP pathway. In addition, tissue-specific rescue experiments indicate that the postsynaptic reception of Activin/TGF-β signaling is important in regulating synaptic GluR abundance, whereas BMP signal reception is known to act in the presynaptic motor neurons to promote synaptic growth. These observations suggest that each pathway influences NMJ synapse development and function by acting mainly in either the pre- or postsynaptic cell (Kim, 2014).

Interestingly, the BMP and Activin/TGF-β pathways have also been recently found to control different aspects of the Drosophila innate immune response (Clark, 2011). In this case BMP signaling suppresses the expression of multiple antimicrobial peptide genes following wounding, whereas the Activin/TGF-β pathway limits melanization after bacterial infection in adult flies. Therefore, it appears that the division of labor between these subpathways is not limited to just the nervous system, rather it may be the norm when these related signaling pathways act in concert to regulate a common biological process (Kim, 2014).

The fact that the pathways actually differ in how they affect a complex biological process is not surprising given that the different R-Smads are likely to have different selectivity in gene activation. Within motor neurons, BMP signaling promotes microtubule formation in axons and directly regulates expression of trio, a Rac GEF, that acts as a major regulator of actin cytoskeleton in many types of cells. Thus, it is likely that BMP signaling modulates synaptic growth, in part, by changing the structure and dynamics of the actin and microtubule cytoskeleton within motor neurons. BMP signaling also regulates the transcription of twit, a gene encoding a L-6 neurotoxin-like molecule that controls the frequency of presynaptic spontaneous vesicle release (Kim, 2012; Kim, 2014 and references therein).

Targets of Drosophila Activin/TGF-β signaling in any tissue are less well characterized. Within the central brain, glial-derived Myo signals through Smox to control expression of the Ecdysone B1 receptors in remodeling mushroom body neurons. However, it is not clear if EcR-B1 is a direct or indirect target of smox transcriptional regulation. It is also unclear if Ecdysone signaling plays a role in regulating synaptogenesis at the NMJ, although it may play a role during metamorphic remodeling of the NMJ as it does for the mushroom body neurons. The only other known targets of Smox are InR, Pi3K and Akt, all of which are Insulin signaling components and are reduced in the Drosophila prothoracic gland in the absence of Activin/TGF-β signaling. Once again the effect may be indirect, but this finding is interesting since Insulin signaling components have been shown to control synaptic clustering of GluRs (Kim, 2014).

The clustering of GluRs and Dlg at the NMJ have been shown to be regulated by both transcriptional and post-transcriptional mechanisms. For example, a recent genetic screen identified longitudinals lacking (lola), a BTN-Zn finger transcription factor, as an essential regulator of GluR and dPak expression in muscles. In contrast, the current studies on Activin/TGF-β signaling suggest, at least for GluRIIA, that this pathway functions at the post-transcriptional level since this study found that overexpression of glurIIA-gfp using an exogenous promotor and transcriptional activator does not lead to an enrichment of GluRIIAGFP at synaptic sites of Activin/TGF-β pathway mutants. This phenotype is reminiscent of that found for certain mutants in the NF-κB signaling system. Loss of Dorsal (an NF-κB homolog), Cactus (an IκB related factor), or Pelle (an IRAK kinase) leads to a substantial reduction of GluRIIA and a slight reduction of Dlg postsynaptic localization at the NMJ and a concomitant reduction in mEJP size. In addition, as was found for loss of Activin/TGF-β signaling, exogenously-expressed GluRIIA-myc did not reach the synaptic surface in NF-κB signaling mutants consistent with a possible role of Activin/TGF-β signaling in regulating NF-κB signaling. However, even if future studies show that the relationship is true, the Activin/TGF-β pathway likely regulates additional factors since its loss also affects GluRIIB levels and muscle resting potential, neither of which is altered in NF-κB pathway mutants. Interestingly, the regulation of GluRIIB levels by Activin/TGF-β signaling does appear to be at the level of transcription, indicating that this signaling pathway likely affects GluR clustering at the NMJ via both transcriptional and post-transcriptional mechanisms (Kim, 2014).

Analysis of Activin/TGF-β signaling at the NMJ, coupled with previous studies on BMP signaling and the novel ligand Maverick, indicates that TGF-β ligands are produced in, and act upon, all three cell types that contribute to NMJ function, specifically the motor neuron, wrapping glia, and muscle (see Model of controlling NMJ development and function by Activin/TGF-β and BMP pathways). This leads to the important issue of how directionality of TGF-β signaling at the NMJ is regulated. One possibility is that ligands are sequestered, either inside the secreting cells or on their surfaces, so that they have limited access to receptors on the opposing pre or postsynaptic membrane. For example, Gbb is produced both in muscle and motor neurons, leading to the issue of how directional signaling from muscle to motor neurons is achieved. On the postsynaptic muscle, Gbb release is potentiated by dRich (Rho GTPase activating protein at 92B), a Cdc42 selective Gap while in the presynaptic neuron Crimpy, a Drosophila homolog of the vertebrate Crim1 gene, has been shown to bind to a precursor form of Gbb. The Gbb/Crimpy complex is thought to either interfere with secretion or activation of motor neuron-derived Gbb thus ensuring that only muscle-derived Gbb activates the retrograde BMP signal at the NMJ. Since there are a large number of characterized TGF-β superfamily binding proteins, Drosophila homologs of some of these factors such as the BMP binding proteins Cv-2, Sog, Tsg and Dally, or the Activin-binding protein Follistatin, may sequester and regulate levels of active ligands within the NMJ. Sequestering mechanisms may also provide direction control by facilitating autocrine as opposed to juxtacrine signaling. If ligand-binding proteins are associated with the membrane surface of the ligand-producing cell, they may facilitate delivery of the ligand to receptors on the producing cell, thus enhancing autocrine signaling. It is interesting in this regard that in the developing Drosophila retina, Actβ appears to signal in an autocrine fashion to control photoreceptor connectivity in the brain (Kim, 2014).

Activin-type ligands are secreted from glia, motor neuron and muscle. The Activin-type ligands induce Babo-mediated phosphorylation of Smox that facilitates association with Med. In the muscle, the phospho-Smox/Med complexes activate the transcription of glurIIB and an unknown factor controlling post-transcriptional process or stability of glurIIA mRNA. In the motor neuron, the phospo-Smox/Med complex controls spontaneous release of synaptic vesicles via unknown mechanism(s). On the other hand, glia-secreted Mav stimulates Mad phosphorylation in the muscle resulting in increased gbb transcription. Gbb protein is released from the muscle and binds Tkv/Sax and Wit complex on the motor neuron leading to an accumulation of phospho-Mad in the nuclei by an unknown mechanism. The resultant phospho-Mad/Med complex activates the transcription of trio whose product promotes synaptic bouton formation (Kim, 2014).

Another important mechanism to control signal direction is likely to be tissue-specific receptor expression. For example, Wit is highly enriched in motor neurons compared to muscle, and this may help ensure that Gbb released from the postsynaptic muscle signals to the presynaptic motor neuron. Type I receptor diversity may be even more important in controlling directionality since at least 2 isoforms of Tkv and three isoforms of Babo have been identified. In the case of Babo, Activin-like ligands have a clear preference for signaling through different receptor isoforms, and these isoforms show differential tissue expression (Kim, 2014).

An additional factor to be considered in understanding TGF-β superfamily signal integration within different NMJ cell types is the possibility of canonical versus non-canonical and/or cross-pathway signaling. For example, in mushroom body neurons Babo can signal in a non-Smad dependent manner through Rho1, Rac and LIM kinase1 (LIMK1) to regulate axon growth and target recognition. Whether this mechanism, or another non-canonical pathway is operative at the NMJ is unclear. Cross-pathway signaling has also recently been identified in Drosophila. In this example, loss of Smox protein in the wing disc has been shown to up-regulate Mad activity in a Babo-dependent manner. Double mutants of babo and smox suppress the cross-pathway signal. As is described in this study, smox protein null mutations lead to significantly more severe GluR and mEJP defects than strong babo mutations alone, and this phenotype is suppressed in double mutants. Thus, as in wing discs, loss of Smox protein likely leads to ectopic Mad activity in muscles that further decrease GluR expression and/or localization at the NMJ. Consistent with this view, this study found that loss of Mad actually increases GluRIIB localization, suggesting that Mad acts negatively to regulate GluRIIB in muscle. One possible model to explain the Smox/Mad data is that normally the Babo/Smox signal inhibits Mad signaling which is itself a repressive signal for GluR accumulation. Thus, in babo mutants, total GluR levels decrease due to the loss of smox and therefore an increase in the repressive Mad signal. In the smox protein null mutant even more repressive Mad signal is generated by Babo further hyperactivating Mad activity leading to even lower levels of GluR accumulation. In medea mutants the activity of both pathways is reduced thereby returning the level of GluR levels close to normal. Additional experiments employing various single and double mutants, together with tissue-specific expression of various ligands, receptor isoforms and ligand-binding proteins will be needed to fully elucidate how vectorial TGF-β signaling is accomplished at the NMJ. Likewise, the identifcation of directly responding target genes and how they are influenced by both Smox and Mad signals is needed to fully appreciate how these two TGF-β signaling branches regulates NMJ functional activity (Kim, 2014).

RNA-binding FMRP and Staufen sequentially regulate the coracle scaffold to control synaptic glutamate receptor and bouton development

Both mRNA-binding Fragile X Mental Retardation Protein (FMRP) and mRNA-binding Staufen regulate synaptic bouton formation and glutamate receptor (GluR) levels at the Drosophila neuromuscular junction (NMJ) glutamatergic synapse. This study tested whether these RNA-binding proteins (RBPs) act jointly in a common mechanism. Both dfmr1 and staufen mutants, and trans-heterozygous double mutants, were shown to display increased synaptic bouton formation and GluRIIA accumulation. With cell-targeted RNAi, a downstream Staufen role within postsynaptic muscle. With immunoprecipitation, this study showed that FMRP binds staufen mRNA to stabilize postsynaptic transcripts. Staufen is known to target actin-binding, GluRIIA anchor Coracle, and this study confirmed that Staufen binds to coracle mRNA. FMRP and Staufen were shown to act sequentially to co-regulate postsynaptic Coracle expression, and show Coracle, in turn, controls GluRIIA levels and synaptic bouton development. Consistently, this study found dfmr1, staufen and coracle mutants elevate neurotransmission strength. FMRP, Staufen and Coracle all suppress pMad activation, providing a trans-synaptic signaling linkage between postsynaptic GluRIIA levels and presynaptic bouton development. This work supports an FMRP-Staufen-Coracle-GluRIIA-pMad pathway regulating structural and functional synapse development (Song, 2022).

This study reveals the mechanism of the established FMRP negative regulation of postsynaptic GluRIIA receptors and presynaptic bouton formation in the Drosophila FXS disease model. Specifically, the mRNA-binding FMRP-positive translational regulator binds to staufen mRNA as predicted, within the postsynaptic cell. Consequently, both dfmr1 and staufen mutants share the elevated GluRIIA level and bouton number phenotypes based on a common postsynaptic pathway function, and genetically interact as trans-heterozygotes to reproduce these phenotypes. Staufen acts as a dsRBP to bind coracle mRNA as predicted; both dfmr1 and staufen mutants exhibit elevated postsynaptic Coracle levels, and genetically interact as trans-heterozygotes to reproduce this phenotype. Coracle acts as a GluRIIA-binding anchoring scaffold within the postsynaptic domain to regulate local receptor accumulation (Chen, 2005). Consequently, dfmr1, staufen and coracle mutants all increase NMJ synaptic functional differentiation to elevate neurotransmission strength. Finally, the elevated postsynaptic GluRIIA levels mediate retrograde BMP receptor trans-synaptic signaling that induces pMad to drive new presynaptic bouton development. dfmr1, staufen and coracle mutants all exhibit elevated presynaptic pMad levels, thereby linking the postsynaptic GluRIIA accumulation and presynaptic supernumerary bouton formation defects shared by all of these mutants (Song, 2022).

The staufen mutant increased synaptic Coracle levels, GluRIIA levels and bouton number are all internally consistent. In a previous study, opposite phenotypes were measured in staufenHL/Df(2R)Pcl7B, which reduces another 14 genes in heterozygous deficiency, including loci involved in neuronal development (e.g. grh, nopo). Importantly, this study similarly found reduced synaptic protein levels and bouton number in staufenHL/Df(2R)Pcl7B, suggesting that heterozygosity of one or more of the neighboring genes impairs synaptic development. However, this study showed that a staufen RNAi that reduces transcript levels by ~90% replicates the staufen mutant NMJ phenotypes of increased GluRIIA levels and synaptic bouton numbers. This was also replicated with a second, independent staufen RNAi line. Moreover, this study showed that the effect is entirely restricted to postsynaptic muscle RNAi, with no effect from presynaptic neuron RNAi, consistent with restricted postsynaptic Staufen function. In addition, postsynaptic staufen rescue of the staufen mutant restored normal synaptic bouton formation, with OE reducing GluRIIA levels in staufen mutants and rescuing GluRIIA levels in dfmr1 mutants. Both staufen mutants and postsynaptic staufen RNAi also share the arrested supernumerary satellite bouton development characterizing dfmr1 null mutants. These many independent lines of evidence confirm the results, and are consistent with the known parallel FMRP role in restricting GluRIIA levels and synaptic bouton formation (Song, 2022).

To regulate Staufen, FMRP binds staufen mRNA and protects targeted staufen transcripts from degradation. FMRP contains at least three distinct RNA-binding domains (RBDs), and Staufen has five RBDs. Staufen reportedly binds a specific RNA hairpin structure formed by long 3' UTRs, but RIP shows that Staufen also binds mRNAs that are not predicted to generate this secondary structure. Although the decreased staufen mRNA levels in both dfmr1 mutants and muscle-targeted dfmr1 RNAi are predicted to be due to the lack of FMRP binding, it is also possible that other unregulated interactors cause the downregulated staufen mRNA expression (Shah et al., 2020). Localized labeling with an anti-Staufen antibody has been reported in the postsynaptic NMJ, which can be confirmed, but it was not possible to reduce labeling in staufen hypomorphic mutants. Therefore Staufen labeling was not shown in the current study. Moreover, western blots have been reported with the same anti-Staufen antibody; however, attempts were unsuccessful. Therefore qPCR was used to measure staufen mRNA levels. Staufen binds to coracle mRNA, but does so in a non-selective manner. This result is consistent with Staufen acting as a very broad spectrum dsRBP, and suggests that Staufen likely acts with a translational regulator partner to generate specificity. FMRP is very well established to partner with other RBPs to mediate the translational regulation of its target transcripts (Song, 2022). ------

The postsynaptic Coracle scaffold acts in a GluRIIA local anchoring mechanism, presumably to link the receptors to the underlying actin cytoskeleton (Chen, 2005). The jointly elevated Coracle and GluRIIA levels in both dfmr1 and staufen mutants are consistent with this scaffold function. Because the dfmr1/+; staufen/+ trans-heterozygotes share this correlated Coracle and GluRIIA upregulation in the postsynaptic domain, a single common signaling pathway is indicated. Coracle also restricts terminal branching development in peripheral sensory neurons. Both coracle mutants and sensory neuron-targeted coracle RNAi also display increased dendritic branch and termini numbers. These phenotypes are similar to the expanded NMJ terminals and increased synaptic bouton development reported in this study. Importantly, both coracle loss of function (mutants and muscle-targeted RNAi) and gain of function (muscle-targeted OE) increase postsynaptic GluRIIA levels and generate supernumerary boutons. Likewise, the knockdown and OE of many other similar scaffolds are known to cause phenocopying defects. Some examples include the muscle chaperone UNC-45, the tight junction scaffold zonula occludens-1 (ZO-1) and synaptic UNC-13. Indeed, both coracle loss and OE similarly cause increased dendritic crossing in Drosophila sensory neurons, similar to the phenocopy of developmental defects reported in this study. Combining the roles of postsynaptic FMRP-Staufen-Coracle in GluRIIA clustering, it was reasoned that this pathway must be a regulatory determinant of synaptic functional development (Song, 2022).

Removing FMRP, Staufen and Coracle strongly enhances functional synaptic differentiation and NMJ neurotransmission strength. This is consistent with expectations from the postsynaptic GluRIIA accumulation in all of these mutants. Elevated GluRIIA levels are well known to be associated with increased evoked functional responses and prolonged channel open times. A GluRIIA pore sequence (MQQ) critically required for the Drosophila channel Ca2+ permeability is conserved in mammalian receptors. This selectivity allows Ca2+-dependent participation in spontaneous (mEJC) and evoked (EJC) neurotransmission. Although enhanced evoked EJC amplitudes are typically accompanied by mEJC alterations, this study found that mEJC amplitude and frequency are unchanged in both the staufen and coracle mutants, and show only minimal changes in the dfmr1 mutants. Classically, both evoked and spontaneous neurotransmission were thought to be mediated by the same vesicles; however, more recent evidence has indicated that spontaneous and evoked neurotransmission have distinct machinery and vesicle pools. Postsynaptic receptors can be segregated into different compartments that are activated by either spontaneous or evoked release. This work supports this growing body of evidence for differential regulation. Importantly, GluRIIA has unique functions, modulating both presynaptic glutamate release and presynaptic bouton development (Song, 2022).

The dfmr1, staufen and coracle mutants all showed upregulated presynaptic pMad correlated with postsynaptic activated GluRIIA accumulation. GluRIIA activation triggers presynaptic pMad signaling via BMP receptors surrounding active zones, which, in turn, stabilizes GluRIIA receptors in the postsynaptic domains. This trans-synaptic signaling mechanism induces new presynaptic bouton development. The targeted postsynaptic RNAi for all three genes confirms this intercellular link. Synaptic BMP signaling involves both the type I serine/threonine kinase receptors and the type II receptor Wit. Although BMP ligand Glass bottom boat (Gbb) signaling via Wit presynaptic receptors is well established at the NMJ to modulate synaptogenesis, the mechanism of presynaptic bouton formation induced by activated GluRIIA signaling does not involve canonical BMP signaling via Gbb. In the dfmr1 mutants, it is suggested that postsynaptic GluRIIA accumulation induces presynaptic bouton development via non-canonical GluRIIA-Wit trans-synaptic retrograde signaling. Similarly, the muscle postsynaptic glypican Dally-like protein (Dlp) negatively regulates NMJ synaptic development by inhibiting this same non-canonical BMP pathway through decreased activated GluRIIA expression. Postsynaptic GluRIIA clustering can thus trigger presynaptic bouton formation, although supernumerary boutons do not always induce reciprocal GluRIIA changes. It is concluded that an FMRP-Staufen-Coracle-GluRIIA-pMad pathway regulates intertwined structural and functional glutamatergic synapse development (Song, 2022).

Drosophila Nesprin-1 controls glutamate receptor density at neuromuscular junctions

Nesprin-1 is a core component of a protein complex connecting nuclei to cytoskeleton termed LINC (linker of nucleoskeleton and cytoskeleton). Nesprin-1 is anchored to the nuclear envelope by its C-terminal KASH domain, the disruption of which has been associated with neuronal and neuromuscular pathologies, including autosomal recessive cerebellar ataxia and Emery-Dreifuss muscular dystrophy. This study describes a new and unexpected role of Drosophila Nesprin-1, Msp-300, in neuromuscular junction. Larvae carrying a deletion of Msp-300 KASH domain (Msp-300ΔKASH) present a locomotion defect suggestive of a myasthenia, and demonstrate the importance of muscle Msp-300 for this phenotype, using tissue-specific RNAi knock-down. Msp-300ΔKASH mutants display abnormal neurotransmission at the larval neuromuscular junction, as well as an imbalance in postsynaptic glutamate receptor composition with a decreased percentage of GluRIIA-containing receptors. Msp-300ΔKASH locomotion phenotypes could be rescued by GluRIIA overexpression, suggesting that the locomotion impairment associated with the KASH domain deletion is due to a reduction in junctional GluRIIA. In summary, this study found that Msp-300 controls GluRIIA density at the neuromuscular junction. Theses results suggest that Drosophila is a valuable model for further deciphering how Nesprin-1 and LINC disruption may lead to neuronal and neuromuscular pathologies (Morel, 2014).

This work describes a new and unexpected role of Drosophila Nesprin-1, Msp-300, in neuromuscular junction function. It was first shown that larvae carrying a deletion of Msp-300 KASH domain present a locomotion defect suggestive of a myasthenia, and the importance of muscle Msp-300 for this phenotype was demonstrated using tissue-specific RNAi knock-down. It was then shown that Msp-300 ΔKASH mutants display abnormal neurotransmission at the larval neuromuscular junction, as well as an imbalance in postsynaptic glutamate receptor composition with a decreased percentage of GluRIIA-containing receptors. Finally, Msp-300 ΔKASH locomotion phenotypes could be rescued by GluRIIA overexpression, suggesting that the locomotion defects associated with the KASH domain deletion are partly due to a reduction in junctional GluRIIA (Morel, 2014).

Biological evidence is presented supporting previous bioinformatics prediction that Msp-300 is a Nesprin-1, thus validating the use of Drosophila to study LINC complex and Nesprin-1-related diseases. Msp-300 forms filaments with a 'beads on a string' pattern, which seems to assemble as sheets at the level of Z-discs and form a web closely apposed to nuclei. This perinuclear localization requires the presence of the KASH domain, showing that the predicted KASH is functional. A strong nuclear clustering associated with the KASH domain deletion was documented, in agreement with nuclei anchoring defects recently reported by Elhanany-Tamir (2012) in larvae carrying a genomic deletion removing the 3' half of Msp-300 gene. Nesprin-1 was shown to be an important player of nuclear positioning in mouse muscles, where the KASH domain deletion causes nuclei mislocalization with the occurrence of extrasynaptic nuclei clustering and in C. elegans hypoderm syncitia, where loss of the Nesprin-1 homolog ANC-1 causes nuclei clustering. These observations therefore constitute biological evidence that Msp-300 is a bona fide Nesprin-1 (Morel, 2014).

Nuclei clustering is a striking feature of KASH domain deletion in proteins such as Msp-300/Nesprin1, ANC-1 and Klarsicht. Correlation between nuclei clustering and locomotion impairment in Syne-1 KO mice, klarsicht, and Msp-300 mutants together with the occurrence of nuclei abnormal localization in pathologies such as centronuclear myopathies raise the question of the possible contribution of nuclei clustering to the locomotion phenotype. Since there were no Msp-300 mutants presenting a locomotion/junctional phenotype without nuclei clustering, therefore establishing the contribution of Msp-300 alone, mutant conditions perturbing nuclear localization and presenting locomotion impairment were examined, and it was asked if these phenotypes occur independently of Msp-300. klarsicht mutations result in nuclei clustering and locomotion impairment but also Msp-300 mislocalization. They thus could not be used to discriminate between a role of nuclei clustering or an independent contribution of Msp-300 to the locomotion phenotype. On the other hand, ens swo mutants present altered locomotion, Msp-300 mislocalization, together with irregularly spaced nuclei, but no nuclei clusters. Based on these results, a contribution of fine nuclei position to the locomotion phenotype cannot be excluded. However, the locomotion defects together with the Msp-300 subcellular mislocalization observed in both ens swo and Msp-300 ΔKASH larvae independently of the presence or absence of nuclei clusters suggest that nuclei clustering itself is not responsible for the locomotion impairment and rather point toward a direct contribution of Msp-300 localization (Morel, 2014).

Interestingly, Z-disc localization is not affected by the KASH domain deletion. The antibody used in this work was generated using a partial cDNA of Msp-300. Blast analysis reveals that this cDNA 3' end aligns with all Msp-300 isoforms, the 5' end being shared by fewer isoforms. Msp-300 localizations observed in this study are thus likely to result from the superposition of several discrete localization patterns corresponding to different isoforms. It is proposed that KASH-containing isoforms are responsible for the perinuclear Msp-300 staining while the Z-disc staining corresponds to a different subset of Msp-300 isoforms, which could perform different tasks in the cell (Morel, 2014).

Msp-300 ΔKASH larvae present a clear locomotion defect, which is fully recapitulated in larvae with muscle-specific knock-down of Msp-300 KASH-containing isoforms. This locomotion defect can be explained by a decreased percentage of GluRIIA-containing receptors at the NMJ. Indeed, deletion of one copy of the gluRIIA gene results in identical locomotion defects while overexpression of GluRIIA in Msp-300 ΔKASH rescues the locomotion phenotype. Interestingly, Z-disc localization is not affected by the KASH domain deletion. The antibody used in this work was generated using a partial cDNA of Msp-300. Blast analysis reveals that this cDNA 3' end aligns with all Msp-300 isoforms, the 5' end being shared by fewer isoforms. Msp-300 localizations observed in this study are thus likely to result from the superposition of several discrete localization patterns corresponding to different isoforms. It is proposed that KASH-containing isoforms are responsible for the perinuclear Msp-300 staining while the Z-disc staining corresponds to a different subset of Msp-300 isoforms, which could perform different tasks in the cell (Morel, 2014).

When performing electrophysiological analysis, a decrease was detected in eEJC's amplitude and quantal content in Msp-300 ΔKASH larvae but no alteration of the mEJCs. This result was at first surprising knowing that these mutant larvae have a decreased GluRIIA density and that GluRIIA density somehow controls mEJPs amplitude. Indeed, increasing GluRIIA density by twofold in GluRIIB null background leads to an increase in mEJP amplitude. According to immunostainings, GluRIIA density was only decreased by 35% in Msp-300 ΔKASH larvae when compared to WT conditions. Thus, it is proposed that in Msp-300 ΔKASH larvae, the density of postsynaptic GluRIIA-containing receptors is sufficient to give a normal response to the spontaneous exocytosis of one neurotransmitter containing vesicle, hence the lack of modifications in the mEJCs. It is also proposed that the amount of GluRIIA-containing receptors would be limiting upon stimulation and release of several quanta of neurotransmitter, resulting in decreased eEJCs. This hypothesis is further supported by the observation that although larvae carrying a single copy of the gluRIIA gene present a clear locomotion defect (similar to Msp-300 ΔKASH larvae), larvae with a single copy of both gluRIIA and gluRIIB genes only present a small decrease in mEJPs. Both the GluR density measures and the electrophysiological analysis are thus in agreement with Msp-300 KASH domain deletion resulting in a decreased postsynaptic sensitivity to neurotransmitter release (Morel, 2014).

Investigations on the mechanisms underlying GluR synaptic localization have revealed that A- and B-type receptor localization are governed by different processes. Indeed, Discs-Large, a prototypical MAGUK protein localized in the subsynaptic reticulum (SSR), positively regulates B-type receptor NMJ targeting, without affecting A-type receptors while A-type, but not B-type, receptor targeting is under the control of Dorsal (NF-κB) and Cactus (IκB). This study has shown that neither Dorsal nor Cactus SSR localization are affected by Msp-300 KASH domain deletion. It can therefore be concluded that Msp-300 contributes to A-type GluR NMJ localization independently of either Dorsal or Cactus (Morel, 2014).

Several non-mutually exclusive hypotheses can be proposed to explain how Msp-300 KASH domain deletion alters GluR composition at the NMJ. Disconnecting Msp-300 from the nuclei could directly impact transcription of GluR subunits. Indeed, an increasing number of results suggest that the LINC complex controls gene expression: the LINC complex has been involved in chromatin organization in mammalian cells and S. cerevisiae and disconnecting the LINC complex from the actin cytoskeleton leads to altered cellular response to mechanical stress and abnormal gene expression. Msp-300 could also control GluR subunit proteins levels by controlling mRNAs access to the translation machinery or post-translational modification of GluRIIA. In agreement with this, human glutamate receptors undergo important post-translational modifications impacting their activity, trafficking, or localization. Similar modifications could occur on GluRIIA and control the assembly of functional postsynaptic glutamate receptors, their activity or localization at the synapse (Morel, 2014).

Is Msp-300 an organizer of the perinuclear region? Msp-300 size (13,000 aa for the CH and KASH domain-containing isoforms), the presence of numerous spectrin repeats (up to 52), the localization pattern observed at Z-band and in the cytoplasm surrounding nuclei, suggest that Msp-300 could be a scaffold organizing the perinuclear region. Several lines of evidence support that hypothesis, which could explain the phenotypes described in this work for Msp-300 ΔKASH mutants (Morel, 2014).

In higher eukaryotes, ER is seen as a highly dynamic continuum consisting of three different subcompartments, the rough ER, the smooth ER, and the nuclear envelope. ER dynamics is thought to play an important role in both the morphology and the functions of ER, and relies mostly on microtubules in mammals and Drosophila. Elhanany-Tamir (2012) documented ER and microtubule organization in WT Drosophila larval muscle and showed that astral microtubules are attached to the nuclear envelope from which they radiate, while ER localizes around myonuclei and at Z-bands. This organization is lost in Msp-300 mutants. In Msp-300 ΔKASH mutants, microtubules detach from the nucleus and form a loose perinuclear ring overlapping with Msp-300 ring. Elhanany-Tamir further reports an important disorganization of ER staining in Msp-300-3' deletion mutants (Morel, 2014).

Considering microtubules' important role in ER dynamics and shaping and their disorganization in Msp-300 ΔKASH, it is tempting to speculate that ER dynamics or fine subcellular organization might be altered upon Msp-300 KASH domain deletion (Morel, 2014).

Syne-1, mammalian Msp-300, was isolated in a screen for Golgi-specific spectrin repeats containing proteins. Observation of the subcellular localization of Syne-1 and Golgi in myoblasts and myotubes, together with comparison of physical distance between nuclei and Golgi apparatus and Syne-1 size, led to the idea that Syne-1 could physically couple Golgi, ER, and nuclei in muscle cells (Morel, 2014 and references therein).

It is therefore speculated that altering Msp-300 anchorage to nuclei could directly impact both ER and Golgi organization and localization with respect to myonuclei or dynamics, thus resulting in altered translation or post-translational modification of proteins including glutamate receptors. Modification of organelle subcellular organization upon Msp300 mutation could thus in turn impact NMJ function (Morel, 2014).

Finally, Nesprin-1 was originally isolated in a yeast two-hybrid screen for MuSK interactors and called Syne-1 (for synaptic nuclear envelope-1) based on its enrichment at the nuclear envelope of synaptic nuclei. MuSK is a receptor tyrosine kinase involved in acetylcholine receptors clustering at the NMJ in mammals. Although acetylcholine receptor density or molecular architecture of the NMJ are not altered by the expression of a dominant negative form of Syne-1 in transgenic mice, effects of Syne-1 KASH domain deletion on NMJ organization were not described. Further investigation is thus necessary to exclude a potential contribution of Syne-1 to NMJ organization. The parallel between the potential involvement of Syne-1 together with MuSK in the clustering of the acetylcholine receptors and the role of Msp-300 in type-A glutamate receptor density at NMJ should nevertheless be kept in mind when investigating the molecular mechanisms of Msp-300/Nesprin-1 contribution to synapse function (Morel, 2014). Mutations in Nesprin-1 have been associated with autosomal recessive cerebellar ataxia (ARCA1), EDMD, and autosomal recessive arthrogryposis diseases. ARCA1 is a neural disorder associated with cerebellar atrophy and impaired walking. Seven mutations were identified in ARCA1 patients, either in introns or exons, leading to a premature stop and resulting in Nesprin-1 C-terminus deletion. These were interestingly associated with mislocalized subsynaptic nuclei at the NMJ. Autosomal recessive arthrogryposis is a rare disease associated with congenital contractures. Analysis of two generations of a congenital family led to the identification of a mutation in nesprin-1 gene also resulting in a premature stop of the protein and deletion of its KASH domain. Finally, EDMD has been associated with mutations in LMNA, EMD , and nesprin-1 and 2 genes, all proposed to affect LINC complex organization. These three pathologies are associated with impaired muscle function, attributed either to neuromuscular or neuronal defects. In all cases, the nesprin-1 mutations identified result in Nesprin-1 disconnection from the LINC complex, often due to the KASH domain loss, explaining the increasing interest for the contribution of the LINC complex in muscle and neural functions (Morel, 2014).

This study has shown that Msp-300 ΔKASH larvae display obvious signs of locomotion defects that are not due to a lack of muscle contractility but rather to a defective synaptic function. Indeed, the results establish that Msp-300 is involved in the control of glutamate receptor density at the NMJ in a KASH-dependent manner. Considering the role of Msp-300 in controlling postsynaptic homeostasis, it is tempting to speculate that EDMD, ARCA1 and autosomal recessive arthrogryposis could all result from alterations of the postsynaptic fields associated with Nesprin-1 mutations (Morel, 2014).

Since Drosophila NMJ, being glutamatergic, is widely used as a model for central glutamatergic synapses, it is proposed that Drosophila is a new relevant model to study the function of Nesprin-1 in the accumulation of postsynaptic glutamate receptors and more generally to decipher the mechanisms by which Nesprin-1 impacts synapse physiology and understand its implications in neuromuscular and neuronal pathologies (Morel, 2014).

The extracellular-regulated kinase effector Lk6 is required for Glutamate receptor localization at the Drosophila neuromuscular junction

The proper localization and synthesis of postsynaptic glutamate receptors are essential for synaptic plasticity. Synaptic translation initiation is thought to occur via the target of rapamycin (TOR) and mitogen-activated protein kinase signal-integrating kinase (Mnk) signaling pathways, which is downstream of extracellular-regulated kinase (ERK). This study used the model glutamatergic synapse, the Drosophila neuromuscular junction, to better understand the roles of the Mnk and TOR signaling pathways in synapse development. These synapses contain non-NMDA receptors that are most similar to AMPA receptors. The data show that Lk6, the Drosophila homolog of Mnk1 and Mnk2, is required in either presynaptic neurons or postsynaptic muscle for the proper localization of the GluRIIA glutamate receptor subunit. Lk6 may signal through eukaryotic initiation factor (eIF) 4E to regulate the synaptic levels of GluRIIA as either interfering with eIF4E binding to eIF4G or expression of a nonphosphorylatable isoform of eIF4E resulted in a significant reduction in GluRIIA at the synapse. It was also found that Lk6 and TOR may independently regulate synaptic levels of GluRIIA. (Hussein, 2016).

This study is the first to provide information on the properties and regulation of the Drosophila protein kinase LK6. Its catalytic domain is strikingly similar to those of mammalian Mnks; similar to them, in mammalian cells LK6 can bind to ERK, can be activated by ERK signalling and can phosphorylate eIF4E. This occurs at the physiological site, Ser209. The MAPK-binding motif of LK6 is of the type previously shown to bind ERK but not p38 MAPK. Consistent with this, when expressed in mammalian cells, LK6 is not activated by stimuli that turn on p38 MAPK (Hussein, 2016).

It is more challenging to perform similar experiments in Drosophila cells owing to the difficulty in transfecting, e.g. S2 cells with high efficiency. However, importantly, this study shows that LK6 also interacts with the ERK homologue Rolled, but not with the Drosophila p38 homologue. The results, furthermore, show that LK6 is activated by Phorbol myristate acetate (PMA), but not by arsenite, which activates p38 MAPK. The regulatory properties of LK6 thus appear to be similar in mammalian and Drosophila cells, indicating that the specificity of the MAPK-interaction motifs is probably similar in both mammals and Diptera. Similar to Mnk1 and Mnk2a, LK6 is primarily, if not exclusively, cytoplasmic. It does contain a basic region of the type that, in Mnk1 and Mnk2, can bind to the nuclear shuttling protein importin-α. It therefore seems probable that either (1) it contains an NES, which ensures its efficient re-export from the nucleus, or (2) the basic region is not accessible to importin-α. The lack of effect of LMB on the localization of LK6 rules out the operation of a CRM1-type NES of the kind found in Mnk1, although the very long C-terminal extension of LK6 might contain an LMB-insensitive NES (Hussein, 2016).

By analogy with the Mnks, it is probable that the N-terminal polybasic region of LK6 mediates its binding to eIF4G and could also interact with importin-α. Given that full-length LK6 shows less efficient binding to eIF4G when compared with Mnk1, it also seems possible that it binds importin-α less efficiently, which may contribute to the finding that LK6 is cytoplasmic. It has been shown previously that even the much shorter C-terminus of Mnk2a impedes access to the N-terminal basic region in that protein, so it is entirely possible that the much larger C-terminal part of LK6 has a similar effect. This could explain why the fragment of LK6 that lacks the C-terminus bound better to eIF4G than did the full-length protein. It may also be that the low degree of binding reflects the fact that the association of LK6 with the heterologous human protein was being studied, rather than with Drosophila eIF4G. Repeated attempts have been made to use the available antisera to examine the association of LK6 with eIF4G in S2 cells, but without success. Comparison of the polybasic region of LK6 with those of Mnk1 and Mnk2a (which do bind eIF4G and importin-α), and recent results for mutants with alterations in these features, do not reveal any difference that might obviously explain the decreased ability of LK6 to bind mammalian eIF4G. As argued above, the C-terminus of LK6 may also impair its activation by ERK, based on the observation that the catalytic domain is more effectively activated than a mutant of the full-length protein that also lacks the ERK-binding motif (Hussein, 2016).

The results support the idea that LK6 is a Drosophila eIF4E kinase. LK6 can phosphorylate eIF4E in vitro and its overexpression in cells leads to increased phosphorylation of endogenous eIF4E. Furthermore, the activation of LK6 by ERK signalling but not by p38 MAPK signalling correlates well with the observed behaviour of the phosphorylation of eIF4E in PMA- or arsenite-treated Drosophila cells, and the fact that LK6 is activated by stimuli that stimulate ERK but is not activated by stimuli that activate p38 MAPK, in HEK-293 cells. The ability of LK6 to bind eIF4G also supports the contention that it can act as an eIF4E kinase in vivo (Hussein, 2016).

The observation that phosphorylation of the endogenous eIF4E in S2 cells is increased by PMA but not by arsenite is consistent with the regulatory properties of LK6 and with the notion that LK6 may phosphorylate eIF4E in these cells. The fact that it is the only close homologue of the Mnks in the fruitfly genome is also consistent with this notion. Phosphorylation of eIF4E has previously been shown to play an important role in growth in this organism and in its normal development. The current data show that LK6 can phosphorylate Drosophila eIF4E in vitro, consistent with the idea that LK6 acts as an eIF4E kinase in this organism. The dsRNAi data that was obtained, which show that two different interfering dsRNAs directed against LK6 each markedly decrease eIF4E phosphorylation in S2 cells, offer strong support to the conclusion that LK6 acts as an eIF4E kinase in Drosophila. Unfortunately, the poor quality of the available anti-LK6 antisera prevented assessing whether the incomplete nature of the loss of phosphorylation of eIF4E reflects incomplete elimination of LK6 expression (Hussein, 2016).

Previous genetic studies have linked LK6 to Ras signalling in Drosophila. This agrees very well with the finding that LK6 is activated by ERK signalling, since ERK lies downstream of Ras. LK6 was first identified as interacting with microtubules and centrosomes. Overexpression of LK6 led to defects in microtubule organization, indicative of their increased stability. The connections between the phosphorylations of eIF4E and microtubules are not immediately obvious. However, it is entirely possible that LK6 has additional substrates that interact with microtubules or are components of centrosomes and their phosphorylation may be important in the regulation of, for example, mitosis. Numerous microtubule-associated proteins are indeed phosphorylated. Microtubules undergo massive reorganization during mitosis and this involves an array of phosphorylation events and protein kinases. It may therefore be relevant that LK6 is activated by mitogenic signalling (i.e. through ERK and thus Ras) (Hussein, 2016).

A pre-synaptic regulatory system acts trans-synaptically via Mon1 to regulate Glutamate receptor levels in Drosophila

Mon1 is an evolutionarily conserved protein involved in the conversion of Rab5 positive early endosomes to late endosomes through the recruitment of Rab7. This study has identified a role for Drosophila Mon1 in regulating glutamate receptor levels at the larval neuromuscular junction. Mutants were generated in Dmon1 through P-element excision. These mutants are short-lived with strong motor defects. At the synapse, the mutants show altered bouton morphology with several small supernumerary or satellite boutons surrounding a mature bouton; a significant increase in expression of GluRIIA and reduced expression of Bruchpilot. Neuronal knockdown of Dmon1 is sufficient to increase GluRIIA levels suggesting its involvement in a pre-synaptic mechanism that regulates post-synaptic receptor levels. Ultrastructural analysis of mutant synapses reveals significantly smaller synaptic vesicles. Overexpression of vglut suppresses the defects in synaptic morphology and also downregulates GluRIIA levels in Dmon1 mutants suggesting that homeostatic mechanisms are not affected in these mutants. It is proposed that DMon1 is part of a pre-synaptically regulated trans-synaptic mechanism that regulates GluRIIA levels at the larval neuromuscular junction (Deivasigamani, 2015).

Neurotransmitter release at the synapse is modulated by factors that control synaptic growth, synaptic vesicle recycling, and receptor turnover at postsynaptic sites. Endolysosomal trafficking modulates the function of these factors and therefore plays an important role in regulating synaptic development and function. Intracellular trafficking is regulated by Rabs, which are small GTPases. These proteins control specific steps in the trafficking process. A clear understanding of the role of Rabs at the synapse is still nascent. Drosophila has 31 Rabs, and most of these are expressed in the nervous system. Rab5 and Rab7, present on early and late endosomes, respectively, are critical regulators of endolysosomal trafficking and loss of this regulation affects neuronal viability underscored by the fact that mutations in Rab7 are associated with neurodegeneration (Verhoeven, 2003). Rab5 along with Rab3 is present on synaptic vesicles, and both play a role in regulating neurotransmitter release. In Drosophila, Rab3 is involved in the assembly of active zones by controlling the level of both Bruchpilot-a core active zone protein-and the calcium channels surrounding the active zone (Graf, 2009). In hippocampal and cortex neurons, Rab5 facilitates LTD through removal of AMPA receptors from the synapse. In Drosophila, Rab5 regulates neurotransmission; it also functions to maintain synaptic vesicle size by preventing homotypic fusion. Compared to Rab5 or Rab3, less is known about the roles of Rab7 at the synapse. In spinal cord motor neurons, Rab7 mediates sorting and retrograde transport of neurotrophin-carrying vesicles. In Drosophila, tbc1D17-a known GAP for Rab7-affects GluRIIA levels (J. Lee, 2013); the effect of this on neurotransmission has not been evaluated. Excessive trafficking via the endolysosomal pathway also affects neurotransmission. This has been observed in mutants for tbc1D24-a GAP for Rab35. A high rate of turnover of synaptic vesicle proteins in these mutants is seen to increase neurotransmitter release (Uytterhoeven, 2011; Fernandes, 2014; Deivasigamani, 2015 and references therein).

This study has examined the synaptic role of DMon1-a key regulator of endosomal maturation. Multiple synaptic phenotypes are found associated with Dmon1 loss of function, and one of these is altered synaptic morphology. Boutons in Dmon1 mutants are larger with more satellite or supernumerary boutons-a phenotype strongly associated with endocytic mutants (Dickman, 2006). Formation of satellite boutons is thought to occur due to loss of bouton maturation, with the initial step of bouton budding being controlled postsynaptically and the maturation step being regulated presynaptically (J. Lee, 2010). Supporting this, a recent study shows that miniature neurotransmission is required for bouton maturation. The presence of excess satellite boutons in Dmon1 mutants suggests that the number of 'miniature' events is likely to be affected in these mutants. The fact that this phenotype can be rescued upon expression of vGlut supports this possibility. However, this does not fit with the observed decrease in size and intensity of Brp positive puncta in these mutants. Active zones with low or nonfunctional Brp are known to be more strongly associated with increased spontaneous neurotransmission. Considering the involvement of postsynaptic signaling in initiating satellite bouton formation, it is thought that altered neurotransmission possibly together with impaired postsynaptic or retrograde signaling, contributes to the altered synaptic morphology in Dmon1 mutants. This may also explain why no satellite boutons are observed in neuronal RNAi animals (Deivasigamani, 2015).

A striking phenotype associated with loss of Dmon1 is the increase in GluRIIA levels. This phenotype seems presynaptic in origin since neuronal loss of Dmon1 is sufficient to increase GluRIIA levels. Is the increase in GluRIIA due to trafficking defects in the neuron? This seems unlikely for the following reasons: First, it has been shown that although neuronal overexpression of wild-type and dominant negative Rab5 alters evoked response in a reciprocal manner, there is no change in synaptic morphology, glutamate receptor localization and density, or change in synaptic vesicle size. The role of Rab7 at the synapse is less clear. In a recent study, loss of tbc1D15-17, which functions as a GAP for Rab7, was shown to increase GluRIIA levels at the synapse. Selective knockdown of the gene in muscles, and not neurons, was seen to increase GluRIIA levels, indicating that the function of the gene is primarily postsynaptic (J. Lee, 2013). These data are not consistent with the current results from neuronal knockdown of Dmon1, suggesting that the presynaptic role of Dmon1 in regulating GluRIIA levels is likely to be independent of Rab5 and Rab7 and therefore novel (Deivasigamani, 2015).

The current experiments to evaluate the postsynaptic role of Dmon1 have been less clear. Although a modest increase in GluRIIA levels are seen upon knockdown in muscles, the increase is not always significant when compared to controls. However, the fact that muscle expression of Dmon1 can rescue the GluRIIA phenotype in the mutant suggests that it is likely to be one of the players in regulating GluRIIA postsynaptically. Further, it is to be noted, that while overexpression of vGlut leads to down-regulation of the receptor at the synapse, the receptors do not seem to get trapped in the muscle, suggesting that multiple pathways are likely to be involved in regulating receptor turnover in the muscle, and the DMon1-Rab7-mediated pathway may be just one of them (Deivasigamani, 2015).

How might neuronal Dmon1 regulate receptor expression? One possibility is that the increase in receptor levels is a postsynaptic homeostatic response to defects in neurotransmission, given that Dmon1Δ181 mutants have smaller synaptic vesicles. However, in dvglut mutants, presence of smaller synaptic vesicles does not lead to any change in GluRIIA levels, given that receptors at the synapse are generally expressed at saturating levels. Therefore, it seems unlikely that the increase in GluRIIA is part of a homeostatic response, although one cannot rule this out completely. The other possibility is that DMon1 is part of a transsynaptic signaling mechanism that regulates GluRIIA levels in a post-transcriptional manner. The observation that presynaptically expressed DMon1 localizes to postsynaptic regions and the results from neuronal RNAi and rescue experiments support this possibility. The involvement of transsynaptic signaling in regulating synaptic growth and function has been demonstrated in the case of signaling molecules such as Ephrins, Wingless, and Syt4. In Drosophila, both Wingless and Syt4 are released by the presynaptic terminal via exosomes to mediate their effects in the postsynaptic compartment. It was hypothesized that DMon1 released from the boutons either directly regulates GluRIIA levels or facilitates the release of an unknown factor required to maintain receptor levels. The function of DMon1 in the muscle is likely to be more consistent with its role in cellular trafficking and may mediate one of the pathways regulating GluRIIA turnover. These possibilities will need to be tested to gain a mechanistic understanding of receptor regulation by Dmon1 (Deivasigamani, 2015).

Dbo/Henji modulates synaptic dPAK to gate glutamate receptor abundance and postsynaptic response

In response to environmental and physiological changes, the synapse manifests plasticity while simultaneously maintains homeostasis. This study analyzed mutant synapses of henji, also known as diablo (dbo), at the Drosophila neuromuscular junction (NMJ). In henji mutants, NMJ growth is defective with appearance of satellite boutons. Transmission electron microscopy analysis indicates that the synaptic membrane region is expanded. The postsynaptic density (PSD) houses glutamate receptors GluRIIA and GluRIIB, which have distinct transmission properties. In henji mutants, GluRIIA abundance is upregulated but of GluRIIB is not. Electrophysiological results also support a GluR compositional shift towards a higher IIA/IIB ratio at henji NMJs. Strikingly, dPAK, a positive regulator for GluRIIA synaptic localization, accumulates at the henji PSD. Reducing the dpak gene dosage suppresses satellite boutons and GluRIIA accumulation at henji NMJs. In addition, dPAK associated with Henji through the Kelch repeats which is the domain essential for Henji localization and function at postsynapses. It is proposed that Henji acts at postsynapses to restrict both presynaptic bouton growth and postsynaptic GluRIIA abundance by modulating dPAK (Wang, 2016a).

Coordinated action and communication between pre- and postsynapses are essential in maintaining synaptic strength and plasticity. Presynaptic strength or release probability of synaptic vesicles involves layers of regulation including vesicle docking, fusion, and recycling, as well as endocytosis and exocytosis. Also, how postsynapses interpret the signal strength from presynapses depends largely on the abundance of neurotransmitter receptors at the synaptic membrane. During long-term potentiation, lateral diffusion of extrasynaptic AMPA receptor to synaptic sites is accelerated and the exocytosis of AMPAR is enhanced near the postsynaptic density (PSD), causing an accumulation of synaptic receptors. In contrast, under the long-term depression condition, synaptic AMPAR is reduced by hastened endocytosis. While molecular mechanisms are proposed to play roles in regulating and fine-tuning postsynaptic glutamate receptor (GluR) abundance in plasticity models, the developmental regulation of GluR abundance at the synaptic surface still needs to be elucidated (Wang, 2016a).

Synapses at the Drosophila neuromuscular junction (NMJ) use glutamate as the neurotransmitter, and have properties reminiscent of mammalian central excitatory synapses. Homologous to vertebrate AMPAR and kainate receptors, Drosophila GluR subunits assemble as tetramers to gate ion influx. Each functional receptor contains essential subunits (GluRIIC, GluRIID and GluRIIE) and either GluRIIA or GluRIIB; therefore, synaptic GluRs can be classified according to their subunit compositions as either A- or B-type receptors. These two types of receptors exhibit distinct developmental and functional properties. Newly-formed PSDs tend to accumulate more GluRIIA channels, while the IIA/IIB ratio becomes more balanced when PSDs mature. In addition, GluRIIB channels have much faster desensitization kinetics, which results in smaller quantal size than GluRIIA channels. Therefore, the synaptic composition of these two types of GluRs greatly influences the postsynaptic interpretation of neuronal activities. The Drosophila homolog of p21-activated kinase (dPAK) regulates GluRIIA abundance at the PSD; GluRIIA receptor clusters at the postsynaptic membrane are strongly reduced in dpak mutants. However, overexpression of dPAK in postsynapses is not sufficient to increase GluRIIA cluster size, suggesting that dPAK activity in regulating GluRIIA abundance is tightly controlled (Wang, 2016a).

Ubiquitination and deubiquitination play critical roles in regulating synaptic functions. In loss-of-function mutants for highwire, a gene encoding a conserved E3 ubiquitin ligase, NMJs overgrow, producing supernumerary synaptic boutons. This phenotype is duplicated by overexpression of the deubiquitinating enzyme Fat facets (Faf) in presynapses. These studies underline the importance of balanced ubiquitination in synapse formation and function. Cullin-RING ubiquitin ligases (CRLs) are large protein complexes that confer substrate ubiquitination. Importantly, CRLs promote ubiquitination through substrate receptors that provide specific recognition of substrates for ubiquitination. The BTB-Kelch proteins are suggested to be the substrate receptors for Cul3-scaffolded CRLs. This study identified a BTB-Kelch-containing protein, Henji, also known as Dbo, which regulates NMJ growth and synaptic activity by restricting the clustering of GluRIIA. Synaptic size of henji mutants was significantly expanded, as viewed under transmission electron microscopy (TEM). Immunostaining for dPAK and GluRIIA also suggests larger areas of PSDs in the absence of Henji, and the intensity of each fluorescent punctum becomes stronger, indicating abnormal accumulation of these PSD proteins. By genetically reducing one gene dosage of dpak in henji mutants, GluRIIA accumulation and abnormal bouton morphology was suppressed. In contrast, reducing the gluriia gene dosage in henji mutants restored bouton morphology but failed to suppress dPAK accumulation. Thus, Henji regulates bouton morphology and GluRIIA clustering levels likely through a control of dPAK. Interestingly, while overexpression of dPAK, either constitutively active or dominantly negative, had no effects on GluRIIA clustering, overexpression of these dPAK forms in henji mutants modulated GluRIIA levels, indicating that Henji limits the action of dPAK to regulate GluRIIA synaptic abundance. Henji localized to the subsynaptic reticulum (SSR) surrounding synaptic sites, consistent with the idea that Henji functions as a gatekeeper for synaptic GluRIIA abundance (Wang, 2016a).

This study shows that Henji functions at the postsynapse to regulate synaptic development and function at the NMJ. The PSD area is expanded and GluRIIA clusters abnormally accumulate at the PSD. Genetic evidences are provided to support that the elevation of GluRIIA synaptic abundance is at least partially caused by a corresponding accumulation of dPAK in henji mutants. Henji is sufficient to downregulate dPAK and GluRIIA levels and the Kelch repeats of Henji play the most critical role in this process. Henji tightly gates dPAK in regulating GluRIIA abundance, as dPAK enhances GluRIIA cluster abundance only when Henji is absent. Therefore, this study has identified a specific negative regulation of dPAK at the postsynaptic sites that contributes to the PSD formation and GluR cluster formation at the NMJ (Wang, 2016a).

PAK proteins transduce various signaling activities to impinge on cytoskeleton dynamics. Through kinase activity-dependent and -independent mechanisms, PAK regulates not only actin- and microtubule-based cytoskeletal rearrangement but also the activity of motors acting on these cytoskeletal tracks. In mammalian systems, PAKs participate in many synaptic events including dendrite morphogenesis, neurotransmitter receptor trafficking, synaptic strength modulation, and activity-dependent plasticity. Pathologically, PAK dysregulation also contributes to serious neurodegenerative diseases, Huntington's disease and X-linked mental retardation (Wang, 2016a).

At Drosophila NMJs, dPAK has divergent functions; loss of dpak causes a dramatic reduction in both Dlg and GluRIIA synaptic abundance, but the underlying molecular mechanisms have not been revealed. The current data show that Henji functions to restrict GluRIIA clustering but has no effect on Dlg levels, suggesting that Henji regulates one aspect of dPAK activities, probably via the SH2/SH3 adaptor protein Dock. Alternatively, Henji may function to limit dPAK protein levels locally near the postsynaptic region, rendering its influence on GluRIIA clustering, while dPAK that regulates Dlg may localize outside of the Henji-enriched region. Supporting this idea, Henji is specifically enriched around the SSR region instead of dispersed throughout the muscle cytosol. Moreover, ectopic Myc-dPAK localized at the postsynapse only when henji was mutated, indicating that Henji regulates dPAK postsynaptic localization (Wang, 2016a).

The interaction with Rac, Cdc42, or both triggers autophosphorylation and subsequent conformational changes of PAK, resulting in kinase activation. The myristoylated dPAK that has been shown to be active in growth cones failed to enhance GluRIIA abundance at the NMJ. This result shows that dPAK is necessary to regulate GluRIIA synaptic abundance, but is itself tightly regulated at the synaptic protein level or the kinase activity. Indeed, evidence is provided to show specific negative regulation of dPAK by Henji; overexpression of dPAK CA that could not enhance GluRIIA abundance in WT larvae further increased the already enhanced GluRIIA levels in the henji mutant. Similar to the CA form, the DN form also showed no effect on GluRIIA when simply overexpressed in the WT background, but exhibited strong suppression of GluRIIA in the henji mutant background. Thus, regardless of the possible conformational differences between the CA and DN forms, Henji appears to confer a constitutive negative regulation of dPAK at postsynapses, suggesting a tight control that could be at subcellular localization. In contrast to CA and DN forms, activation of dPAK requires binding to Rac1 and Cdc42, and subsequent protein phosphorylation. This additional layer of regulation may serve as a limiting factor rendering dPAK WT from recruiting GluRIIA to PSDs regardless in WT or henji mutant background (Wang, 2016a).

The structural feature suggests that Henji could function as a conventional substrate receptor of the Cul3-based E3 ligase complex. At Drosophila wing discs, Dbo functions as a Cul3-based E3 ligase to promote Dishevelled (Dsh) downregulation. Similar to the henji alleles, it was confirmed that the dbo [Δ25.1] allele and dbo RNAi were competent to induce dPAK and GluRIIA accumulation at the postsynapse. An immunoprecipitation experiment detected Henji and dPAK in the same complex, and dPAK also forms a complex with the C-terminal substrate-binding Kelch-repeats region. However, no notable or consistent increase was detected in Henji-dependent dPAK poly-ubiquitination in both S2 cells and larval extracts. Also, the Cul3-binding BTB domain of Henji seems dispensable in the suppression of dPAK levels in henji mutants. Importantly, Cul3 knockdown in muscle cells failed to cause any accumulation of GluRIIA and dPAK at the NMJ. Sensitive genetic interaction between henji and Cul3 failed to induced dPAK and GluRIIA accumulation. Dbo functions together with another BTB-Kelch protein Kelch (Kel) to downregulate Dsh. However, Kel negatively regulates GluRIIA levels without affecting dPAK localization at the postsynaptic site. This data argues that Kel functions in a distinct pathway to Henji in postsynaptic regulation of GluRIIA. Taken together, no direct evidence was found to support that dPAK is downregulated by Henji through ubiquitination-dependent degradation. Alternately, Henji could bind dPAK near the postsynaptic region and this interaction may block the recruitment or localization of dPAK onto postsynaptic sites. Under this model, dPAK is less restricted and has a higher propensity to localize at postsynaptic sites in the absence of Henji, resulting in synaptic accumulation of dPAK and GluRIIA expansions (Wang, 2016a).

As many synaptic events require rapid responses, local regulation of protein levels becomes crucial in synapses. To achieve accurate modulation, certain synaptic proteins should be selectively controlled under different developmental or environmental contexts. Indeed, emerging evidence shows that various aspects of synapse formation and function are under the control of the ubiquitin proteasome system (UPS), including synapse formation, morphogenesis, synaptic pruning and elimination, neurotransmission, and activity-dependent plasticity. In particular, the membrane abundance of postsynaptic GluR that modulates synaptic function can be regulated by components of the UPS. When Apc2, the gene encoding Drosophila APC/C E3 ligase, is mutated, GluRIIA shows excess accumulation but the molecular mechanism was not elucidated. Similarly, loss of the substrate adaptor BTB-Kelch protein KEL-8 in C. elegans also results in the stabilization of GLR-1-ubiquitin conjugates. However, no evidence shows direct ubiquitination and degradation of GLR-1 by KEL-8. Also, absence of the LIN-23-APC/C complex in C. elegans affects GLR-1 abundance at postsynaptic sites without altering the level of ubiquitinated GLR-1. Therefore, GLR-1 receptor endocytosis and recycling or ubiquitination and degradation of GLR-1-associated scaffold proteins are proposed to be the underlying mechanism for E3 ligase regulation. In mammals, endocytosis of AMPAR can be influenced by poly-ubiquitination and degradation of the prominent postsynaptic scaffold protein PSD-95 (Wang, 2016a).

This study describes a novel regulation by the BTB-Kelch protein Henji on synaptic GluRIIA levels. By limiting GluRIIA synaptic levels, Henji modulates the postsynaptic output in response to presynaptic glutamate release. In the absence of Henji, quantal size is elevated, coinciding with an increase in the postsynaptic GluRIIA/GluRIIB ratio. In a previous study, increases in the GluRIIA/GluRIIB ratio by overexpressing a GluRIIA transgene in the muscle or by reducing the gene copy of gluriib promote NMJ growth, but co-expression of both GluRIIA and GluRIIB did not alter the bouton number. Combined with the current findings, those data provide a link between an increased GluRIIA-mediated postsynaptic response and bouton addition at NMJs. However, satellite boutons were not detected following GluRIIA overexpression. One possibility is that satellite boutons are considered as immature boutons and their appearance may indicate the tendency for NMJ expansion, as in the case of excess BMP signaling. Failure to become mature boutons may be caused by the lack of cooperation with other factors such as components of the presynaptic endocytic pathway, actin cytoskeleton rearrangement or neuronal activity. No significant alterations in endocytosis and the BMP pathway in the henji mutant. Nevertheless, it cannot be ruled out that Henji may modulate other presynaptic events that are defective in henji mutants to interfere with bouton maturation (Wang, 2016a).

Filamin, a synaptic organizer in Drosophila, determines glutamate receptor composition and membrane growth

Filamin is a scaffolding protein that functions in many cells as an actin-crosslinker. FLN90, an isoform of the Drosophila ortholog Filamin/cheerio that lacks the actin-binding domain, is shown in this study to govern the growth of postsynaptic membrane folds and the composition of glutamate receptor clusters at the larval neuromuscular junction. Genetic and biochemical analyses reveal that FLN90 is present surrounding synaptic boutons. FLN90 is required in the muscle for localization of the kinase dPak and, downstream of dPak, for localization of the GTPase Ral and the exocyst complex to this region. Consequently, Filamin is needed for growth of the subsynaptic reticulum. In addition, in the absence of filamin, type-A glutamate receptor subunits are lacking at the postsynapse, while type-B subunits cluster correctly. Receptor composition is dependent on dPak, but independent of the Ral pathway. Thus two major aspects of synapse formation, morphological plasticity and subtype-specific receptor clustering, require postsynaptic Filamin (Lee, 2016).

Proper postsynaptic function depends on appropriate localization of receptors and signaling molecules. Scaffolds such as PSD-95/SAP90 and members of the Shank family are critical to achieving this organization. While usually without intrinsic enzymatic activity, scaffolds recruit, assemble, and stabilize receptors and protein networks through multiple protein-protein interactions: they can bind to receptors, postsynaptic signaling complexes, and the cytoskeleton at the postsynaptic density. Mutations in these proteins are associated with neuropsychiatric disorders. While understanding synapse assembly has begun, much remains to be investigated (Lee, 2016).

The Drosophila larval neuromuscular junction (NMJ) is a well-studied and genetically accessible glutamatergic synapse. Transmission is mediated by AMPA-type receptors, and several postsynaptic proteins important for its development and function have related proteins at mammalian synapses, including the PDZ-containing protein Discs-Large (DLG) and the kinase Pak. In addition, the postsynaptic membrane forms deep invaginations and folds called the subsynaptic reticulum (SSR), which are hypothesized to create subsynaptic compartments comparable to dendritic spines. Recently, we found that the SSR is a plastic structure whose growth is regulated by synaptic activity. This phenomenon may be akin to the use-dependent morphological changes, such as growth of dendritic spines, that occur postsynaptically in mammalian brain. The addition of membrane and growth of the SSR requires the exocyst complex to be recruited to the synapse by the small GTPase Ral; the SSR fails to form in ral mutant larvae. Moreover, the localization of Ral to a region surrounding synaptic boutons is likely to direct the selective addition of membrane to this domain. Ral thus provided a tractable entry point for better understanding postsynaptic assembly. The mechanism for localizing the Ral pathway, however, was unknown. The present study determined that Ral localization is dependent on cheerio, a gene encoding filamin, which is critical for proper development of the postsynapse (Lee, 2016).

Filamin is a family of highly conserved protein scaffolds with a long rod-like structure of Ig-like repeats. With over 90 identified binding partners, some of which are present also at the synapse, mammalian filamin A (FLNA) is the most abundant and commonly studied filamin. Filamin can bind actin and has received the most attention in the context of actin cytoskeletal organization. Drosophila filamin, encoded by the gene cheerio (cher), shares its domain organization and 46% identity in amino acid sequence with human FLNA. Drosophila filamin has a well-studied role in ring canal formation during oocyte development, where it recruits and organizes actin filaments. This study now shows that filamin has an essential postsynaptic role at the fly NMJ. An isoform of this scaffold protein that lacks the actin-binding domain acts via dPak to localize GluRIIA receptors and Ral; filamin thereby orchestrates both receptor composition and membrane growth at the synapse (Lee, 2016).

Filamin is a highly conserved protein whose loss of function is associated with neurodevelopmental disorders. In humans, mutations in the X-linked FLNA cause periventricular heterotopia, a disorder of cortical malformation with a wide range of clinical manifestations such as epilepsy and neuropsychiatric disturbances. Studies in rodent models have shown that abnormal filamin expression causes dendritic arborization defects in a TSC mouse and that filamin influences neuronal proliferation. Filamin is present in acetylcholine receptor clusters at the mammalian NMJ, but its function there is unknown. In lysates of the mammalian brain, filamin associates with known synaptic proteins such as Shank3, Neuroligin 3, and Kv4.2. A recent report indicated that filamin degradation promotes a transition from immature filopodia to mature dendritic spines, a phenomenon that is likely to be related to the actin-bundling properties of the long isoform of filamin. Data in the present study have uncovered a novel pathway that does not require the actin-binding domain of filamin. In this pathway, postsynaptically localized filamin, via Pak, directs two distinct effector modules governing synapse development and plasticity: (1) the Ral-exocyst pathway for activity-dependent membrane addition and (2) the composition of glutamate receptor clusters. These pathways determine key structural and physiological properties of the postsynapse (Lee, 2016).

Although loss of filamin had diverse effects on synapse assembly, they were selective. Muscle-specific knockdown or the cherQ1415sd allele disrupted type-A but not type-B GluR localization at the postsynaptic density. Likewise, the phenotypes for muscle filamin were confined to the postsynaptic side: the presynaptic active zone protein Brp and overall architecture of the nerve endings were not altered by muscle-specific knockdown. The specificity of its effect on particular synaptic proteins, and the absence of the actin-binding region in FLN90, suggests that filamin's major mode of action here is not overall cytoskeletal organization, but rather to serve as a scaffold for particular protein-protein interactions (Lee, 2016).

Analysis of the distribution of the SSR marker Syndapin and direct examination of the subsynaptic membrane by electron microscopy revealed that formation of the SSR required filamin. Genetic analysis uncovered a sequential pathway for SSR formation from filamin to the Pak/Pix/Rac signaling complex, to Ral, to the exocyst complex and consequent membrane addition. The SSR is formed during the second half of larval life and may be an adaptation for the low input resistance of third-instar muscles. Like dendritic spines, the infoldings of the SSR create biochemically isolated compartments in the vicinity of postsynaptic receptors and may shape physiological responses, although first-order properties of the synapse, such as mini- or EPSP amplitude are little altered in mutants that lack an SSR. The formation of the SSR requires transcriptional changes driven by Wnt signaling and nuclear import, proteins that induce membrane curvature (such as Syndapin, Amphiphysin, and Past1), and Ral-driven, exocyst-dependent membrane addition. The activation of Ral by Ca2+ influx during synaptic transmission allows the SSR to grow in an activity-dependent fashion. The localization of Ral to the region surrounding the bouton appears crucial to determining the site of membrane addition because Ral localization precedes SSR formation and exocyst recruitment and because exocyst recruitment occurred selectively surrounding boutons even when Ca2+-influx occurred globally in response to calcimycin. This study has now shown that Ral localization, and consequently exocyst recruitment, membrane growth, and the presence of the SSR marker Syndapin, are all dependent on a local action of filamin at the synapse. FLN90, the filamin short isoform, localized to sites of synaptic contact and indeed surrounded the boutons just as does Ral. When this postsynaptic filamin was removed by muscle-specific filamin knockdown or the cherQ1415sd allele, the downstream elements of the pathway, Pak, Rac, Ral, the exocyst, and Syndapin, were no longer synaptically targeted. The mislocalization is not a secondary effect of loss of the SSR but likely a direct consequence of filamin loss, as Pak and Ral can localize subsynaptically even in the absence of the SSR. Unlike the likely mode of action of nuclear signaling by Wnt, the delocalization of Ral was not a consequence of altered protein production; its expression levels did not change. Thus filamin may be viewed as orchestrating the formation of the SSR and directing it to the region surrounding synaptic boutons (Lee, 2016).

The second major feature of the filamin phenotype was the large reduction in the levels of the GluRIIA receptor subunit from the postsynaptic membranes. GluRIIA and GluRIIB differ in their electrophysiological properties and subsynaptic distribution. Because type B GluRs, which contain the IIB subunit, desensitize more rapidly than type A, the relative abundance of type A and type B GluRs is a key determinant of postsynaptic responses and changes with synapse maturation. The selective decrease in GluRIIA at filamin-null NMJs is likely a consequence of dPak mislocalization: filamin-null NMJs lack synaptic dPak, and dPak null NMJs lack synaptic GluRIIA. Moreover, the first-order electrophysiological properties at NMJs lacking filamin resembled those reported at NMJs missing dPak. In the current study, though, only the change in mEPSP frequency was statistically significant. At filamin-null NMJs, the decrease in GluRIIA is accompanied by an increase in GluRIIB, suggestive of a partial compensation that could account for the relatively normal synaptic transmission. Because the IIA and IIB subunits differ in desensitization kinetics and regulation by second messengers, functional consequences of filamin loss may become more apparent with more extensive physiological characterizations at longer time scales (Lee, 2016).

While both SSR growth and receptor composition required the kinase activity of dPak, receptor composition was independent of Ral and thus represents a distinct branch of the pathway downstream of dPak. As with Ral, the loss of GluRIIA from the synapse was due to delocalization and not a change in expression of the protein, consistent with unaltered GluRIIA transcripts in dPak null animals. Thus filamin, via dPak, alters proteins with functional significance for the synapse as well as its structural maturation (Lee, 2016).

Mammalian filamin, via its many Ig-like repeats, has known scaffold functions in submembrane cellular compartments and filamin is therefore likely also to serve as a scaffold at the fly NMJ. These epistasis data indicate that filamin recruits Ral through recruitment of a signaling complex already known to function at the fly NMJ: dPak and its partners dPix and Rac. Mammalian filamin is reported to directly interact with Ral during filopodia formation, however the details of their interaction at the fly NMJ are less clear. Because Ral localization requires filamin to recruit dPak and dPix and specifically requires the kinase activity of dPak, it is possible that either Ral or filamin need to be phosphorylated by dPak to bind one another. Mammalian filamin interacts with some components of the Pak signaling complex and is a substrate of Pak. This study has now shown that Drosophila filamin and PAK interact when coexpressed in HEK cells, and thus a direct scaffolding role for FLN90 in the recruitment of Pak and the organization of the postsynapse is likely (Lee, 2016).

The overlapping but different distributions of filamin and its downstream targets indicate that its scaffolding functions must undergo regulation by additional factors. The proteins discussed in this study take on either of two patterns at the synapse. Some, like Ral, Syndapin, and filamin itself, are what can be termed subsynaptic and, like the SSR, envelope the entire synaptic bouton. Others, like dPak and its partners and the GluRIIA proteins, are concentrated in much smaller regions, immediately opposite presynaptic active zones, where the postsynaptic density (PSD) is located. It is hypothesized that filamin interacts with additional proteins, including potentially transsynaptic adhesion proteins, that localize filamin to the subsynaptic region and also govern to which of the downstream effectors it will bind. Indeed, it appears paradoxical that dPak, though predominantly at the postsynaptic density is nonetheless required for Ral localization throughout the subsynaptic region. If dPak is needed to phosphorylate either filamin or Ral to permit Ral localization, the phosphorylations outside the PSD may be due to low levels of the dPak complex in that region; synaptic dPak was previously shown to be a relatively mobile component of the PSD (Lee, 2016).

Filamin was the first nonmuscle actin-crosslinking protein to be discovered. With an actin-binding domain at the N terminus, the long isoform of filamin and its capacity to integrate cellular signals with cytoskeletal dynamics have subsequently been the focus of the majority of the filamin literature. At the NMJ, however, this was not the case. Several lines of evidence indicated that the short FLN90 isoform of filamin, which lacks the actin-binding domain, plays an essential role in postsynaptic assembly. First, the short FLN90 isoform was the predominant and perhaps the only isoform of filamin found expressed in the muscles. Second, both endogenous and overexpressed FLN90 localized subsynaptically. Third, loss of the short isoform disrupted localization of postsynaptic components while lack of just the long isoform had little or no effect. Lastly, exogenous expression of just the short isoform in filamin null background sufficiently rescued the defect in SSR growth. The modest postsynaptic phenotypes of the cher1allele, which predominantly disrupts the long isoform, may be due to small effects of the allele on expression of the short isoform or may be an indirect consequence of the presence of the long isoform in the nerve terminals (Lee, 2016).

The existence of the short isoform has been reported in both flies and mammals and may be produced either by transcriptional regulation or calpain-mediated cleavage. The short isoform can be a transcriptional co-activator, but its functional significance and mechanisms of action have been largely elusive. The short isoform has little or no affinity for actin, but most of the known sites for other protein-protein interactions are shared by both isoforms. Thus the structure of FLN90, with nine predicted Ig repeats and likely protein-protein interactions, is consistent with a scaffolding function to localize key synaptic molecules independent of interactions with the actin cytoskeleton (Lee, 2016).

This study has introduced filamin as a major contributor to synapse development and organization. The severity of the phenotypes indicates filamin has a crucial role that is not redundant with other scaffolding proteins. The effects of filamin encompass several much-studied aspects of the Drosophila NMJ: the clustering and subunit subtype of glutamate receptors and the plastic assembly of specialized postsynaptic membrane structures. The pathways that govern these two phenomena diverge downstream of Pak kinase activity and are dependent on filamin for the proper localization of key signaling modules in the pathways. By likely acting as a scaffold protein, the short isoform of filamin may function as a link between cell surface proteins, as yet unidentified, and postsynaptic proteins with essential localizations to and functions at the synapse. Because many of the components of these pathways at the fly NMJ are also present at mammalian synapses and can interact with mammalian filamin, a parallel set of functions in CNS dendrites merits investigation (Lee, 2016).

The maintenance of synaptic homeostasis at the Drosophila neuromuscular junction is reversible and sensitive to high temperature

Homeostasis is a vital mode of biological self-regulation. The hallmarks of homeostasis for any biological system are a baseline set point of physiological activity, detection of unacceptable deviations from the set point, and effective corrective measures to counteract deviations. Homeostatic synaptic plasticity (HSP) is a form of neuroplasticity in which neurons and circuits resist environmental perturbations and stabilize levels of activity. One assumption is that if a perturbation triggers homeostatic corrective changes in neuronal properties, those corrective measures should be reversed upon removal of the perturbation. This study tested the reversibility and limits of HSP at the well-studied Drosophila melanogaster neuromuscular junction (NMJ). At the Drosophila NMJ, impairment of glutamate receptors causes a decrease in quantal size, which is offset by a corrective, homeostatic increase in the number of vesicles released per evoked presynaptic stimulus, or quantal content. This process has been termed presynaptic homeostatic potentiation (PHP). Taking advantage of the GAL4/GAL80(TS)/UAS expression system, PHP was triggered by expressing a dominant-negative glutamate receptor subunit GluRIIA at the NMJ. PHP was then reversed by halting expression of the dominant-negative receptor. These data show that PHP is fully reversible over a time course of 48-72 h after the dominant-negative glutamate receptor stops being genetically expressed. As an extension of these experiments, it was found that when glutamate receptors are impaired, neither PHP nor NMJ growth is reliably sustained at high culturing temperatures (30-32 ° C). These data suggest that a limitation of homeostatic signaling at high temperatures could stem from the synapse facing a combination of challenges simultaneously (Yeates, 2017).

For the Drosophila NMJ, homeostatic potentiation is a robust and sensitive process. One assumption supported by all available data is that the larval NMJ is capable of modulating its vesicle release at any time point during development, in accordance with the presence or absence of a homeostatic challenge to synapse function. Rapid, acute induction of homeostatic signaling has been demonstrated at the Drosophila NMJ by application of PhTox to impair the glutamate receptors, but the reversibility of this particular modulation (or any other modulation at the Drosophila NMJ) had not been studied. More generally, it is not clear what happens in metazoan nervous systems when strong perturbations that induce homeostatic signaling are introduced for long periods of developmental time and then later removed (Yeates, 2017).

This study presents evidence that PHP at the Drosophila neuromuscular synapse is a reversible process. In doing so, prior findings were confirmed showing that there is a tight inverse relationship between quantal amplitude and QC at the NMJ. This study complements those findings with the results of temperature shift experiments. It was found that PHP is measurable at an early stage of larval development and can be erased over a matter of days. Interestingly, at high temperatures, PHP induced by impairing glutamate receptor function either fails or falls short of full compensation. This failure appears to correlate with impaired NMJ growth in the same animals (Yeates, 2017).

There was a robust expression of PHP for NMJs of MHC-Gal4 >> UAS-GluRIIAM/R larvae with the TubP-Gal80TS transgene raised at 29°C for 48 h after egg-laying. Once the expression of the dominant-negative UAS-GluRIIAM/R transgene was halted, this expression of PHP was erased over a slow 48- to 72-h period (Yeates, 2017).

If PHP is a readily reversible homeostatic process, why is there a days-long time lag to reverse it? The data likely reflect a constraint of the dominant-negative GluRIIAM/R experimental perturbation, rather than the NMJ's capacity to respond quickly to the changed environment. In a prior study, researchers expressed functional, tagged GluRIIA transgenic subunits at the NMJ and performed fluorescence recovery after photobleaching (FRAP) experiments. Those experiments demonstrated that receptor turnover rates at the Drosophila NMJ are extremely slow: it appears that once postsynaptic densities (PSDs) reach a critical size, GluRIIA subunits are stably incorporated. For the current study, this likely means that the temperature downshifts in the reversibility experiments represented an opportunity for the NMJ to incorporate endogenous WT GluRIIA into a significant number of new PSDs while it continued to grow. Given sufficient growth, the endogenously expressed GluRIIA would gradually overcome the previously incorporated dominant-negative GluRIIAM/R subunits. This would restore electrophysiological parameters to normal levels, which is consistent with the current data (Yeates, 2017).

The majority of recent studies about synaptic homeostasis at the Drosophila NMJ have emphasized that presynaptic adjustments to neurotransmitter release properties must occur within minutes of drug-induced (PhTox) postsynaptic receptor inhibition to induce a rapid and offsetting response to PhTox challenge. Some important parameters uncovered through these studies include regulation of presynaptic Ca2+ influx; regulation of the size of the readily releasable pool (RRP) of presynaptic vesicles; control of presynaptic SNARE-mediated fusion events; control of neuronal excitability; and recently, implication of endoplasmic reticulum calcium-sensing activities; presynaptic glutamate receptor activity; and finally, identification of a retrograde, trans-synaptic signaling system governed by the Semaphorin 2b ligand and the Plexin B receptor (Yeates, 2017).

For almost all of the cases in which a mutation or an experimental condition blocks the short-term induction of homeostatic signaling, the same perturbation has also been shown to block its long-term maintenance. Interestingly, however, the converse is not true. Additional studies have shown that the long-term consolidation (or expression) of homeostatic signaling at the NMJ can be genetically uncoupled from its induction. Select molecules seem to be dedicated to a long-term maintenance program that involves protein translation and signaling processes in both the neuron and the muscle. Recent data suggest that such long-term processes may take 6 h or more to take full effect (Yeates, 2017).

As more molecular details about HSP are elucidated, it will be interesting to test whether the rapid induction and sustained consolidation of PHP can be reversed by similar or separate mechanisms, and what the time courses of those reversal mechanisms are. At the mouse NMJ, reversibility was recently demonstrated pharmacologically. d-Tubocurarine was applied at a subblocking concentration to impair postsynaptic acetylcholine receptors. Within seconds of drug application, QC increased—and then within seconds of drug washout, it decreased again (Wang, 2016b). Follow-up experiments suggested that those rapid, dynamic changes in PHP dynamics at the mouse NMJ were mediated by a calcium-dependent increase in the size of the RRP of presynaptic vesicles (Wang, 2016b). Because there seem to be several similarities between the mouse NMJ and the Drosophila NMJ, it is possible that PHP at the insect NMJ can also be rapidly reversed (Yeates, 2017).

It is instructive to consider mammalian synaptic preparations and study how homeostatic forms of synaptic plasticity are turned on and off. Groundbreaking work on cultured excitatory vertebrate synapses showed that in response to activity deprivation (or promotion), synapses employ scaling mechanisms by adding (or subtracting) AMPA-type glutamate receptors to counteract the perturbation. Bidirectional scaling suggested that reversible mechanisms likely dictate homeostatic scaling processes. Complementary studies testing scaling reversibility have borne out this prediction. Additionally, evidence for reversible forms of homeostatic scaling have been found in rodent sensory systems, such as auditory synapses after hearing deprivation (and restoration to reverse) and in the visual cortex after light deprivation (and restoration to reverse). Collectively, the vertebrate and invertebrate studies support the notion that reversible fine-tuning is an efficient strategy used to stabilize activities in metazoan nervous systems. One advantage offered by the Drosophila system is a genetic toolkit to uncover possible reversibility factors (Yeates, 2017).

Are there environmental limitations for homeostatic potentiation at the Drosophila NMJ? Clearly there are. The current data suggest that a combination of high temperature (30°C-32°C) plus impaired glutamate receptor function can severely limit the NMJ's ability to compensate for reduced neurotransmitter sensitivity. High temperature alone does not seem to be a severe enough restriction to impair PHP, because Gal4 driver controls at 30°C have somewhat reduced quantal size, but a fully (or nearly fully) offsetting increase in QC. Likewise, reduced glutamate receptor function alone does not appear to be a sufficient barrier to impair PHP. For example, quantal size is severely diminished when the dominant negative UAS-GluRIIAM/R transgene is expressed at 25°C, but PHP is nevertheless intact (Yeates, 2017).

It is not clear what the molecular or anatomic basis of this limit on PHP is. It is known that it is not an issue of NMJ excitation at high temperatures, because evoked neurotransmission for WT (or driver control) NMJs remains remarkably robust over a range of temperatures, including 30°C. Nor does it seem to be an elimination of PHP in general, because PHP was still present in the case of GluRIIASP16 animals raised at 30°C, revealing a distinction between the null GluRIIA condition and the dominant-negative GluRIIA condition. The limitation seems to be on homeostatic signaling that supports PHP at high temperatures in the face of the dominant-negative transgene expression (Yeates, 2017).

Temperature effects on neurophysiology are well documented. Recent work in crustaceans demonstrates that robust and reliable circuits such as the neurons driving the rhythmicity stomatogastric nervous system can fail under extreme temperature challenges. For the Drosophila NMJ, prior studies of larval development documented a significant enhancement of synaptic arborization when larvae were raised at higher temperatures. Additional studies have shown that NMJ growth plasticity can be additionally affected by mutations that affect neuronal excitability. Given the backdrop of these data, it is not unreasonable to hypothesize that the tolerable limits of synaptic activity challenge could be different at different temperatures (Yeates, 2017).

For the current experiments, 29°C-32°C represents a potential failure zone for homeostatic potentiation at the Drosophila NMJ. It is noted that the data suggest that the coping capacity of the NMJ is dependent on genotype. WT NMJs cope at all temperatures. By contrast, for dominant-negative GluRIIA-expressing NMJs, 29°C is a point at which PHP becomes partial, and 30°C is a point at which it fails. For GluRIIASP16 subunit deletion NMJs, there is robust, but partial PHP at 30°C, not unlike the compensation seen for the dominant-negatives at 29°C. GluRIIASP16 NMJs fail to execute PHP only when temperature is pushed to an extreme range, such as 32°C. Why do these differences persist? It is not clear. The answer could relate to the well-documented temperature-induced alterations in NMJ growth, or alternatively, a limited availability of synaptic factors that are needed to cope with a double challenge of high temperature and particular impairment of glutamate receptor function. Future molecular and physiologic work will be needed to unravel those possibilities in the contexts of different genetic backgrounds and culturing conditions (Yeates, 2017). -----

Protein interactions and retrograde signals

Two distinct mechanisms regulate synaptic efficacy at the Drosophila neuromuscular junction: a PKA-dependent modulation of quantal size and a retrograde regulation of presynaptic release. Postsynaptic expression of a constitutively active PKA catalytic subunit decreases quantal size, whereas overexpression of a mutant PKA regulatory subunit (inhibiting PKA activity) increases quantal size. Increased PKA activity also decreases the response to direct iontophoresis of glutamate onto postsynaptic receptors. The PKA-dependent modulation of quantal size (or response of the muscle to the spontaneous release of a single synaptic vesicle) requires the presence of the muscle-specific glutamate receptor DGluRIIA, since PKA-dependent modulation of quantal size is lost in viable homozygous DGluRIIA- mutants. The DGluRIIA sequence contains an optimal PKA consensus phosphorylation site on the C-terminal tail (RRXS), believed to be located in the cytoplasmic portion of the protein. Elevated postsynaptic PKA reduces the quantal amplitude and the time constant of miniature excitatory junctional potential (mEJP) decay to values that are nearly identical to those observed in DGluRIIA mutants. PKA modulation of quantal size is sensitive to the copy number of DGluRIIA. Larvae heterozygous for a deletion of DGluRIIA show significantly less modulation by PKA than wild-type controls. This suggests that PKA-dependent modulation of receptor function may be influenced by the subunit composition of postsynaptic receptors. PKA activity appears to constitutively regulate synaptic function at the wild-type synapse. The demonstration that inhibition of PKA leads to a large increase in quantal size suggests that there is a high basal phosphorylation of DGluRIIA at the wild-type synapse. The PKA-dependent reduction in quantal size is accompanied developmentally by an increase in presynaptic quantal content, indicating the presence of a retrograde signal that regulates presynaptic release. A retrograde regulation of presynaptic transmitter release may serve to maintain postsynaptic excitation during the developmental growth of this synapse (Davis, 1998).

During the development of the Drosophila NMJ, a large increase in muscle volume is tightly coupled to increases in both presynaptic structure and presynaptic function. This coupling assures that the presynaptic motoneuron is able to appropriately excite postsynaptic muscle. It has been proposed that this correlation between presynaptic and postsynaptic growth is maintained by a signal from muscle to motoneuron (Schuster, 1996a). This hypothesis was tested by manipulating postsynaptic excitation by increasing PKA activity and then assaying whether there is a presynaptic compensation for changes in postsynaptic excitation (Davis, 1998).

The hypothesis that an activity-dependent retrograde signal regulates presynaptic release at this synapse is supported by the data. There is a significant increase in quantal content (>50% increase compared to wild-type and genetic controls) that accompanies the PKA-dependent decrease in quantal size. Quantal content is a measure of the number of vesicles released by the nerve. This increase in quantal content is observed in each of the three experimental genotypes in which the PKA activator decreases quantal size. Furthermore, this increase in quantal content is observed in both voltage-clamp and current-clamp experiments. Since presynaptic release is increased in response to a decrease in the postsynaptic sensitivity to transmitter, it is proposed that there exists a retrograde signal capable of regulating presynaptic transmitter release at this synapse. These results agree with those of Petersen (1997), who observed an increased presynaptic release in glutamate receptor mutants that decrease quantal size (Davis, 1998).

The PKA-dependent increase in quantal size due to PKA inhibition is not compensated for by a change in presynaptic release, however. There is no change in presynaptic quantal content despite a substantial (>40%) increase in quantal size when PKA is inhibited in muscle. As a result, there is a significant increase in the average compound extrajunctional potential (EJC), when compared to wild type. PKA activity was inhibited in muscle by expression of a mutant PKA regulatory subunit under UAS control. These regulatory subunits carry a mutated cAMP binding site and as a result have reduced sensitivity to cAMP resulting in decreased PKA catalytic activity. Again, these results are in agreement with those of Petersen (1997), who showed that an overexpression of DGluRIIA increases quantal size without a compensatory change in presynaptic release. Thus, under these conditions, there does not appear to be a presynaptic compensation for increased postsynaptic excitation (Davis, 1998).

A retrograde signal from muscle to motoneuron could influence presynaptic release by increasing presynaptic structure. Bouton number was quantified for the junction at muscles 6 and 7 in abdominal segment A3. There was no change in bouton number comparing wild type and with larvae expressing increased PKA in muscle. Similarly, there is no change in presynaptic bouton number in larvae in which PKA is inhibited in muscle. Thus, a retrograde signal most likely regulates either the number of presynaptic active zones present in each presynaptic bouton or regulates some aspect of the presynaptic release mechanism (Davis, 1998).

The rate of spontaneous release events is altered by postsynaptic PKA. In both current clamp and voltage clamp experiments, a reduction in quantal size (due to increased PKA in the muscle) correlates with a significant reduction in the frequency of spontaneous mEJP events. This reduction in mEJP frequency may originate postsynaptically, since presynaptic release is actually increased, not decreased, in these mutants. It is possible that a PKA-dependent decrease in quantal size reduces the amplitude of many release events below the noise level. Alternately, these results may indicate that postsynaptic receptors are functionally silenced by increased postsynaptic PKA activity. There was no significant change in miniature EJC frequency when PKA activity was inhibited in muscle, indicating that silent synapses, or very small events, are not revealed by increasing quantal size (Davis, 1998).

Glutamate channel expression and properties

In Drosophila, mutations in specific ion channel genes can increase or decrease the level of neural/synaptic activity. These genetic tools have been used, in combination with classical pharmacological agents, to modulate neural activity during embryogenesis and examine effects on the differentiation of an identified neuromuscular junction. Electrical activity is found to be required for the neural induction of transmitter receptor expression during synaptogenesis. Likewise, neural electrical activity is required to localize transmitter receptors to the synaptic site. In muscles with activity-blocked synapses, a low level of receptors is expressed homogeneously in the muscle membrane as in muscles developing without innervation. Thus, presynaptic electrical activity is required to mediate the neural induction of the transmitter receptor field in the postsynaptic membrane (Broadie, 1993).

Pronounced alterations in synaptic activity have a relatively small impact on functional glutamate receptor expression during neurogenesis. Null mutations in paralytic (para) (which codes for a sodium channel), significantly reduce synaptic activity but does not significantly alter GluR expression by the end of embryogenesis. Likewise, single mutations in Shaker or ether a-go-go, a structural gene encoding a fast voltage-gated potassium channel and another potassium channel respectively, do not significantly alter synaptic GluR expression by the time of hatching. However, more dramatic changes in synaptic activity have a small but significant effect on GluR expression during synaptogenesis. For example, a deficiency in the para locus reduces GluR expression at the neuromuscular junction. Suppression of presynaptic electrical activity, whether by using mutations or injection tetrodotoxin, which binds neuronal sodium channels and so blocks the propagation of action potentials, reduces the expression of GluRs during synaptogenesis. Different alleles of para block synaptic transmission at different temperatures. Embryos raised at these restrictive temperatures show a dramatic reduction in GluR expression at the NMJ at the end of embryogenesis. Evidence is also provided that presynaptic electrical activity is required to localize GluRs to the synaptic site during synaptogenesis. Synaptic morphogenesis is independent of synaptic activity during embryogenesis, and it is concluded that postsynaptic activity, mediated through GluRs, similarly plays no role in synaptic morphogenesis in the embryo. Unlike synaptic structure, the size of the synaptic zone is significantly altered in line with alterations in presynaptic electrical activity. Hyperactive synapses expand to occupy a large area of the muscle surface relative to wild type. Likewise, inactive synapses occupy an even smaller synaptic domain relative to hyperactive synapses, although inactive synapses are not significantly smaller than wild type. Thus, activity modulates the size of the synaptic domain during synaptogenesis. The decreased synaptic density of hyperactive synapses explains the decrease in GluR expression in the postsynaptic membrane observed in eag Sh double mutants; GluR expression is not suppressed by increased activity but, rather, GluRs are localized over a larger surface area and so decrease in density at any given synaptic site (Broadie, 1993)

Glutamate receptor channels in Drosophila embryos and larvae were examined with the patch-clamp technique in various configurations. In the cell-attached mode, only one type of channel is observed in the extrajunctional region at any stage. The burst duration histogram was fit with three exponentials. The burst duration of long component lengthens with increasing glutamate concentration. In excised outside-out patches the unitary channel current is 7.1 pA at -60 mV and the direction of current reversed at zero membrane potential. In contrast, junctional receptor channels have different properties. In the whole-cell configuration, spontaneous synaptic currents with steps on the falling phase are observed. The step amplitudes have two discrete values of 9.4 and 18.5 pA at -60 mV, due to the opening of junctional glutamate receptor channels. Synaptic currents change amplitudes linearly with the membrane potential in the negative potential range but nonlinearly above zero. With 1 mM glutamate in the bath, synaptic currents are no longer observed. Instead, there are single channel events with the current amplitude varying between 8 and 12 pA at -60 mV. Their long burst duration depends on glutamate concentration, indicating that they are glutamate receptor channel events. The extrapolated reversal potential of these channel currents is around +12 mV. These junctional receptor channels are strictly localized at the junction. These findings suggest that the channel conversion mechanism in Drosophila is different from that observed in vertebrates. Further close examination of other intermediate steps during neuromuscular junction formation is needed (Nishikawa, 1995).

Outside-out patches of membrane were excised from muscle fibers 6 and 7 of third-instar wild-type Drosophila larvae. Channels were observed to open in response to short pulses of L-glutamate. At a holding potential of -60 mV, the channels open to one main conductance level of about 120 pS. The current-voltage plot for the channels is linear and reverses around 0 mV holding potential. The channels are also activated by quisqualate but not by aspartate, N-methyl-D-aspartate (NMDA), kainate, glycine, gamma-aminobutyric acid (GABA) and acetylcholine. At high glutamate concentrations (3 or 10 mM), channel activation reached a peak within 0.3 ms. The channels open in 'bursts', flickering between open and closed states. The channels opened only for a few milliseconds after switching on the glutamate and channel activity declines to zero after the initial surge, with time constants between 5 and 20 ms. During applications of low glutamate concentrations (0.2-0.5 mM) to the same patch the channels open much less frequently and during most applications no openings are observed. The openings observed are short and 'bursts' of openings are rare. Two exponential components were identified in the open-time distribution obtained with pulsed applications of glutamate (0.5-10 mM) with time constants of about 0.1 and 2.0 ms. The kinetics of the channels seem to be similar to the kinetics of certain glutamate gated channels found on the muscles of crayfish and locust (Heckmann, 1995).

Outside-out patches from wild-type Drosophila larval muscle were exposed briefly to L-Glutamate (Glu) using a piezo-driven application system. Glu in concentrations of 0.1 to 30 mM was applied and the responses to repeated applications of a given concentration were averaged. The peak current, i, and the current rise time, tr, from 0.1 i to 0.9 i were determined from the averages. Half-maximum activation of the channels is reached with approximately 2 mM Glu. i increases proportional to the power n = 3.5 to n = 5.8 (average of four experiments, n = 4.4) for Glu concentrations between 0.3 and 0.5 mM. tr increases from approximately 0.2 ms at 10 mM Glu to a value of approximately 3.5 ms at 0.2 mM Glu. A linear reaction scheme with five binding steps preceding the channel-opening conformational change is proposed as the kinetic mechanism of channel activation: this scheme was investigated in computer simulations. A set of rate constants assuming the same affinity for each binding site is found to describe the data better than one assuming positive cooperativity. The results are very similar to those for Glu-gated channels of crayfish and locust muscle, which is evidence of a common kinetic mechanism for these channels (Heckmann, 1996).

Glutamate receptor channels are ubiquitous agonist-gated channels. Pharmacologically they are classified into several subtypes but may have common fundamental channel properties. To build a foundation for future molecular biological and genetic studies, kinetics of the glutamate receptor channel were studied in embryonic Drosophila myotubes in culture using the patch clamp technique. There are many channel events of brief duration, together with prolonged ones. Brief events are frequently observed in low concentrations whereas the frequency of prolonged events increases with agonist concentrations. Long openings (> 5 ms) were often interrupted by brief closures, most of which lasted less than 100 mu s, thus showing a bursting behavior. At all agonist concentrations, the burst duration was fitted with three exponential components (brief, intermediate and long). The mean duration of the long component increases linearly with the glutamate concentration. The mean closed time and number of brief closures per ms within long bursts are independent of agonist concentration. The mean burst durations of the brief (30-250 mu s) and intermediate component (300-1050 mu s) do not change significantly with agonist concentration. The closing episodes within bursts are rare in the brief and intermediate burst components. The ratio of the fractional areas of the brief or intermediate and long burst components increases linearly with agonist concentration in the log-log plot with a slope of one. These findings suggest that the brief and intermediate components are due to singly-liganded openings and the long component is the result of doubly-liganded openings (Chang, 1996).

Focal extracellular excitatory postsynaptic currents were recorded to investigate short-term depression at glutamatergic Drosophila neuromuscular synapses. The amplitudes of quantal excitatory postsynaptic currents (qEPSCs) elicited before and after depolarizations eliciting large release were compared. Depression reduces the amplitude of the qEPSCs to 0.65 +/- 0.14 of control. Recovery from depression and of the receptor channels from desensitization follow a similar time course. Thus receptor desensitization seems to be involved in short-term depression at Drosophila neuromuscular junctions (Adelsberger, 1997).

Outside-out patches from wild-type Drosophila larval muscle were exposed to L-glutamate (glu) using a piezo-driven application system. Glu receptor channels open and desensitize in response to rapid applications of 10 mM glu. Desensitization was fitted with an exponential function with a mean time constant of desensitization (tau d) of 15 ms in response to 10 mM glu. The tau d is concentration dependent and decreases to 6 ms (on average) with 0.7 mM glu and increases again to 12 ms (on average) in response to 0.5 mM glu. Desensitization in response to longer applications of glu is almost complete, but surprisingly, even a 1-ms pulse of 3 mM glu produces about 30% desensitization. In the presence of low glu concentrations, the response to a pulse is reduced and is about halved by preequilibration with 30 microM glu. Recovery from desensitization is not concentration dependent and was fitted with an exponential function with a mean time constant of 150 ms. During recovery the channels rarely open. Kinetic schemes were fitted to these results, and a circular reaction scheme was found to fit the data best. An important feature of the scheme is desensitization from a lower ligated closed state. This allows substantial desensitization of synaptic receptor channels in response to quantal release of transmitter, in part without opening of the channels. Desensitization reduces the probability of the channels opening in response to a subsequent release for a period of time determined by the rate of recovery from desensitization and might serve as a form of molecular short-term memory (Heckmann, 1997).

Evoked excitatory postsynaptic currents (EPSC) were recorded with an extracellular macropatch electrode from glutamatergic neuromuscular junctions of Drosophila larvae. At 20 degrees C quantal current amplitude is about -400 pA and the 10%-90% rise time is slightly below 0.2 ms for the fastest rising events and on average 0.3+/-0.1 ms in the best recordings. The quantal currents often have 'shoulders' but decay approximately monoexponentially from half amplitude. The time constant of the exponential fit varies with mean values ranging from 2.5 ms to 7.7 ms in 13 experiments and an average value of 4.4+/-1.6 ms. Comparison of these results with data obtained earlier with outside-out patches of larval muscle membrane leads to the conclusion that glutamate has to reach a saturating peak concentration of at least 10 mM in the synaptic cleft to allow the observed short quantal current rise times. To account for the time course of the quantal current decay one has to assume that the glutamate concentration in the synaptic cleft remains in the millimolar range for more than a millisecond and that the time course of the decay of the quantal currents is in part due to desensitization of the postsynaptic receptor channels (Heckmann, 1998).

Mutations in rho-type guanine exchange factor (rt/GEF), also called dpix, were recovered from a large-scale screen in Drosophila for genes that control synaptic structure. dPix/rtGEF is homologous to mammalian Pix. dPix plays a major role in regulating postsynaptic structure and protein localization at the Drosophila glutamatergic neuromuscular junction. dpix mutations lead to decreased synaptic levels of the PDZ protein Discs large, the cell adhesion molecule Fas II, and the glutamate receptor subunit GluRIIA, and to a complete reduction of the serine/threonine kinase Pak and the subsynaptic reticulum. The electrophysiology of these mutant synapses is nearly normal. Many, but not all, dpix defects are mediated through dPak, a member of the family of Cdc42/Rac1-activated kinases. Direct interaction of mammalian Pix with Pak has been detected. Thus, a Rho-type GEF (Pix) and Rho-type effector kinase (Pak) regulate postsynaptic structure (Parnas, 2001).

In mammals, the Pix family contains two members: alphaPix (Cool-2) and ßPix (Cool-1). Pix has an SH3 domain, a DBL-homology GEF domain, and a pleckstrin homology domain. The Cool (for cloned-out of library)/Pix (for PAK-interactive exchange factor) proteins directly bind to members of the PAK family of serine/threonine kinases and regulate their activity. In Drosophila, dPix is localized to the PSD: dpix mutations lead to the loss of synaptic Pak kinase. Paks are a family of Cdc42/Rac1-activated serine/threonine kinases important in regulating actin-containing structures. In the fly NMJ, Pak kinase is localized to the PSD. In mammals, Pak is recruited to focal complexes in a Cdc42-, Rac1-, and Pix- dependent manner (Parnas, 2001).

In addition to the absence of Pak kinase at the synapse, dpix mutations lead to the decrease in synaptic levels of the PDZ protein Discs-large (Dlg), the cell adhesion molecule Fasciclin II (Fas II), the glutamate receptor (GluR) subunit GluRIIA, and to the elimination of the subsynaptic reticulum (SSR). In Drosophila, the PSD-95 homolog Dlg has been shown to be directly responsible for the clustering of the Shaker potassium channel and to partially control the clustering of the cell adhesion molecule Fas II to the NMJ. Many, but not all, dpix defects are mediated through Pak kinase. Thus, the data suggest a pathway for synaptic clustering from dPix to Pak kinase to Dlg to Shaker and to Fas II (Parnas, 2001).

The dpix phenotype is consistent with at least two functions at the postsynaptic compartment: targeting and stabilization of postsynaptic components. In dpix mutants, Pak kinase is completely missing from the synapse. Since Pix is known to directly interact with Pak in mammals and target it to focal complexes, the data best fit with the model in which dPix targets Pak kinase to the synapse via a direct interaction. Furthermore, overexpressing either Pak kinase or a membrane-tethered gain-of-function form of Pak kinase does not result in any accumulation of Pak kinase at the synapse. Still, it is possible that Pak kinase is targeted to the synapse via a different mechanism and fails to stabilize in dpix mutants (Parnas, 2001).

In contrast to Pak kinase, Dlg and GluRIIA are not completely eliminated from the synapse in dpix mutants, but rather, their levels are reduced. In the case of Dlg, its localization pattern is also disrupted, indicating that dPix controls the postsynaptic targeting of Dlg at least to some extent, as well as its stabilization at the synapse. The localization pattern of GluRIIA (in subsynaptic domains opposite active zones) is intact. Thus, dPix is not necessary for the synaptic targeting of GluRIIA per se, but rather, it is important for maintenance of its levels and/or stabilization (Parnas, 2001).

Mutation and activation of Galphas similarly alters pre- and postsynaptic mechanisms modulating neurotransmission: DGluRIIA expression is under the control of Galphas

Constitutive activation of Galphas in the Drosophila brain abolishes associative learning, a behavioral disruption far worse than that observed in any single cAMP metabolic mutant, suggesting that Galphas is essential for synaptic plasticity. The intent of this study was to examine the role of Galphas in regulating synaptic function by targeting constitutively active Galphas to either pre- or postsynaptic cells and by examining loss-of-function Galphas mutants (dgs) at the glutamatergic neuromuscular junction (NMJ) model synapse. Surprisingly, both loss of Galphas and activation of Galphas in either pre- or postsynaptic compartment similarly increases basal neurotransmission, decreases short-term plasticity (facilitation and augmentation), and abolished posttetanic potentiation. Elevated synaptic function is specific to an evoked neurotransmission pathway because both spontaneous synaptic vesicle fusion frequency and amplitude are unaltered in all mutants. In the postsynaptic cell, the glutamate receptor field is regulated by Galphas activity; based on immunocytochemical studies, GluRIIA receptor subunits are dramatically downregulated (>75% decrease) in both loss and constitutive active Galphas genotypes. In the presynaptic cell, the synaptic vesicle cycle is regulated by Galphas activity; based on FM1-43 dye imaging studies, evoked vesicle fusion rate is increased in both loss and constitutively active Galphas genotypes. An important conclusion of this study is that both increased and decreased Galphas activity very similarly alters pre- and post-synaptic mechanisms. A second important conclusion is that Galphas activity induces transynaptic signaling; targeted Galphas activation in the presynapse downregulates postsynaptic GluRIIA receptors, whereas targeted Galphas activation in the postsynapse enhances presynaptic vesicle cycling (Renden, 2003).

This study provides no evidence that postsynaptic function is regulated by the level of Galphas activity or that alterations in the postsynaptic glutamate receptor field play any role in short-term plasticity at the Drosophila NMJ. In both gain- and loss-of-function Galphas mutants, there is no substantial change in glutamate receptor conductance, density, or distribution based on mEJC amplitude analyses and direct assay of glutamate-gated currents in the muscle. This finding is extremely surprising in light of the dramatic alteration of the molecular character of the postsynaptic glutamate receptor field in both loss and gain of function Galphas mutants. Two different antibodies were used to assay the GluR fields: a polyclonal antibody against all GluRII subunits showed a significant reduction of signal in all Galphas mutants and a monoclonal antibody specific to GluRIIA showed a nearly complete loss of signal in all Galphas mutants. Immunoreactivity against DGluRIIA in the embryo appeared normal in dgsR60 homozygous mutants, indicating a postembryonic modification of DGluRIIA expression under the control of dgs. At a minimum, these analyses reveal a striking molecular alteration of the GluR field downstream of Galphas, possibly to the extent of nearly eliminating GluRIIA subunits (Renden, 2003).

Complete loss of GluRIIA has been shown to cause significantly decreased mEJC amplitudes, whereas a nearly complete loss of GluRIIA immunoreactivity, using two antibodies, is reported here without a similar change in mEJC amplitudes. One way to rationalize this apparent contradiction is to postulate that the reduced presence of GluRIIA after Galphas manipulation is not sufficient to alter significantly mEJC kinetics or amplitudes. The present report shows a 75% reduction of receptor abundance, whereas GluRIIA genetic nulls were examined previously. More recently, the effect of graded expression levels of GluRIIA was examined, revealing that low levels of GluRIIA, in the absence of GluRIIB, results in an overcompensation of presynaptic transmitter release, doubling the amplitude of glutamatergic transmission at the NMJ. At higher levels of GluRIIA expression, this phenotype was eliminated. If the levels of DGluRIIB were also downregulated (or eliminated) by altered Galphas signaling, these findings would be in agreement with those of the previous study. A second possibility is that the loss of GluRIIA immunoreactivity caused by Galphas manipulation may represent epitope masking rather than loss of GluRIIA subunits. Extracellular binding of an auxiliary protein to glutamate receptors has recently been reported in C. elegans, and an essential auxiliary subunit of mammalian AMPA receptors (stargazin) has recently been found. Stargazin is essential for proper insertion and localization of receptors with the postsynaptic density and is modulated by PKA phosphorylation, thereby controlling receptor number. Interaction with such proteins, or other changes in the accessibility/confirmation of GluRIIA in the postsynaptic compartment, might alter its recognition by antibodies. A final possibility is that there may be compensatory increases in the levels of the other GluRII subunits present at the NMJ. Such a compensatory mechanism might permit loss of GluRIIA subunits without an appreciable change in mEJC amplitudes. The loss of GluRIIA immunoreactivity demonstrates conclusively that the postsynaptic GluR field is strikingly controlled by the level of Galphas activity, but the functional significance of this regulation remains elusive and awaits further investigation (Renden, 2003).

Numerous lines of evidence have demonstrated the existence of both anterograde and retrograde transynaptic signals at the Drosophila NMJ. Such signals are involved in induction of postsynaptic receptor fields, pruning of postsynaptic receptor fields, and compensatory regulation of presynaptic quantal content. The present study shows that increasing Galphas function either pre- or postsynaptically results in nearly identical phenotypes, and independent assays of presynaptic and postsynaptic function indicate similar mechanisms. Specifically, presynaptic Galphas activation modifies the postsynaptic GluRIIA receptor field, and postsynaptic Galphas activation heightens presynaptic vesicle cycling. Moreover, global loss-of-function Galphas mutants also modify the postsynaptic GluRIIA field. Are these paired pre- and post-synaptic alterations a form of compensation or are they independent, Galphas-dependent mechanisms? What signals are used to communicate the level of Galphas activity in both directions across the synaptic cleft (Renden, 2003)?

The identity of the messenger(s) is still unclear, but there are a few likely suspects. Glutamate itself has been shown to act as an anterograde regulatory message at the Drosophila NMJ; presynaptic glutamatergic tone inversely controls the levels of DGluRIIA postsynaptically. Thus it is possible that the elevated glutamatergic transmission in Galphas mutants directly causes the downregulation of GluRIIA expression. At the Drosophila NMJ and in mammalian systems, integrin function has been shown to be required for functional synaptic plasticity. Integrins are known to signal between cells within a short period of time through activation of associated intracellular cascades. At the Drosophila NMJ, the hypertonicity response is mediated in part by integrins dependent on intracellular cAMP levels, and in Xenopus cultured neurons, PKA-dependent transmission is inhibited by disintegrin. These studies suggest that integrins may function as anterograde and/or retrograde messengers mediating physical transynaptic signaling. Another possible retrograde messenger is nitric oxide, produced by phosphorylation of nitric oxide synthase (NOS). There is evidence that NOS is present in Drosophila and is localized to epithelial and neuronal tissues . Application of nitric oxide to the NMJ induces presynaptic vesicle fusion, making it a formal candidate as a retrograde messenger (Renden, 2003).

In conclusion, tissue-specific expression of constitutively active Galphas on either side of the Drosophila NMJ synaptic cleft greatly enhances basal neurotransmission to disrupt expression of short-term synaptic plasticity, specifically in low [Ca2+]bath conditions. This Galphas-dependent alteration does not affect the probability of spontaneous vesicle fusion or the basal function of the postsynaptic receptor field and so is specific to evoked release of neurotransmitter. Increases in Galphas activity on either side of the synapse greatly increases evoked amplitude in low Ca2+, primarily due to a cAMP-dependent increased synaptic vesicle mobility, but also dramatically reduce GluRIIA receptor levels. When Galphas activity is decreased, neurotransmission is similarly enhanced, GluRIIA receptor levels are similarly downregulated, but synaptic vesicle mobility is not detectably altered. It is clear that there is a bidirectional transynaptic communication network at the Drosophila NMJ that responds to altered Galphas activity to modify both pre- and post-synaptic compartments. However, the functional significance of some of these changes remains unclear, and the messengers mediating transynaptic signaling remain to be identified (Renden, 2003).

Differential localization of glutamate receptor subunits at the Drosophila neuromuscular junction

The subunit composition of postsynaptic neurotransmitter receptors is a key determinant of synaptic physiology. Two glutamate receptor subunits, Drosophila glutamate receptor IIA (DGluRIIA) and DGluRIIB, are expressed at the Drosophila neuromuscular junction and are redundant for viability, yet differ in their physiological properties. A third glutamate receptor subunit at the Drosophila neuromuscular junction, DGluRIII, has been identified that is essential for viability. DGluRIII is required for the synaptic localization of DGluRIIA and DGluRIIB and for synaptic transmission. Either DGluRIIA or DGluRIIB, but not both, is required for the synaptic localization of DGluRIII. DGluRIIA and DGluRIIB compete with each other for access to DGluRIII and subsequent localization to the synapse. These results are consistent with a model of a multimeric receptor in which DGluRIII is an essential component. At single postsynaptic cells that receive innervation from multiple motoneurons, DGluRIII is abundant at all synapses. However, DGluRIIA and DGluRIIB are differentially localized at the postsynaptic density opposite distinct motoneurons. Hence, innervating motoneurons may regulate the subunit composition of their receptor fields within a shared postsynaptic cell. The capacity of presynaptic inputs to shape the subunit composition of postsynaptic receptors could be an important mechanism for synapse-specific regulation of synaptic function and plasticity (Marrus, 2004a).

The localization of glutamate receptors is essential for the formation and plasticity of excitatory synapses. These receptors cluster opposite neurotransmitter release sites of glutamatergic neurons, but these release sites have heterogeneous structural and functional properties. At the Drosophila neuromuscular junction, receptors expressed in a single postsynaptic cell are confronted with an array of hundreds of apposed active zones. Hence, this is an ideal preparation for the investigation of whether receptor clustering is sensitive to the morphological and physiological properties of the apposed active zones. To investigate the relationship between the localization of glutamate receptors and the properties of the apposed active zones, receptor localization was investigated in mutants in which receptors are limited. It was found that receptors are not uniformly distributed opposite the full array of active zones but that some active zones have a disproportionately large share of receptors as assayed by receptor levels and response to transmitter. The active zones at which receptors preferentially cluster are larger and have a higher neurotransmitter release probability than the average active zone. A similar relationship is found between glutamate receptor clusters and active-zone size at wild-type synapses. It is concluded that when confronted with an array of active zones, glutamate receptors preferentially cluster opposite the largest and most physiologically active sites. These results suggest an activity-dependent matching of pre- and post-synaptic function at the level of a single active zone (Marrus, 2004b).

A novel ionotropic glutamate receptor, DGluRIII, has been identified and characterized at the Drosophila NMJ (Marrus, 2005a). DGluRIII is an essential subunit that is required for the synaptic localization of the two previously described receptors, DGluRIIA and DGluRIIB. A strong hypomorphic mutant was generated that expresses low levels of wild-type DGluRIII protein. In this mutant, where receptor levels are limited, glutamate receptors are found to localize opposite the largest and most physiologically active release sites. These results demonstrate a matching of pre- and post-synaptic function at individual active zones. Such a mechanism could ensure the alignment of receptors opposite functional active zones during development and maximize synaptic strength at this high-fidelity synapse (Marrus, 2004b).

A strong hypomorphic mutation of DGluRIII was generated (Marrus, 2004a) by rescuing a genetic null for DGluRIII with low-level expression of wild type DGluRIII protein (referred to as the DGluRIII mutant). Analysis of glutamate receptor (GluR) expression in this DGluRIII mutant revealed faint puncta of staining within the synaptic region (Marrus, 2004a). To investigate whether these puncta could represent functional receptors, it was first asked whether they localize opposite active zones. NMJs were double stained for an active-zone marker (Wucherpfennig, 2003) and DGluRIII in wild-type and DGluRIII mutant larvae. DGluRIII staining is dramatically reduced in the mutant, but those puncta that are visible colocalize with active zones. Because DGluRIII is thought to function as a component of a heteromultimeric glutamate receptor (Marrus, 2004a), double staining was performed for DGluRIII and a second subunit, DGluRIIA. Each punctum of DGluRIII colocalizes with DGluRIIA. Because the DGluRIII receptor puncta in the DGluRIII mutant localize opposite presynaptic release sites and appear to coassemble with other glutamate receptor subunits, they likely represent functional glutamate receptor complexes (Marrus, 2004b).

To quantify the reduction in DGluRIII levels in the DGluRIII mutant, the intensity of DGluRIII immunoreactivity opposite each active zone was measured. There is an 18-fold reduction in DGluRIII levels in the mutant. Although the intensity of all receptor puncta is down, it appears that some active zones have relatively high levels of apposed receptors, whereas many others have no detectable receptor. It was enquired whether residual receptors are uniformly localized opposite active zones or, alternatively, whether receptors preferentially cluster opposite certain active zones. To investigate this question, the distribution of glutamate receptor intensities was compared opposite each active zone from wild-type and mutant synapses. If receptors are allocated opposite active zones in the same manner in the mutant as in the wild-type, then scaling the wild-type distribution by 18 (the difference in mean intensities) should mimic the mutant distribution. However, these distributions are significantly different. The mutant has many more active zones with relatively brighter apposed GluR puncta, and many more with no detectable GluR puncta, than would be expected from a uniform 18-fold scaling. It is concluded that receptors preferentially localize opposite certain active zones in the mutant (Marrus, 2004b).

To investigate the function of these receptors, electrophysiology was performed at the NMJ. To assess the function of the entire postsynaptic receptor field, ionophoretic glutamate was applied to the synapses. The ionophoretic response in the DGluRIII mutant is approximately 13-fold less than that of wild-type larvae. Although the precise amplitude of ionophoretic responses is somewhat variable, there is good agreement between the reduction in receptor function (13-fold) and receptor staining (18-fold). This observation suggests that there are no unknown receptors that are capable of mediating a substantial glutamate response in the DGluRIII mutant (Marrus, 2004b).

It is concluded that when receptors are limited, receptor patches are most likely to cluster opposite the largest and most active release sites. Even when receptors are in excess, there is a correlation between the density of postsynaptic receptors and the size of the apposed active zone. These findings suggest a model of activity-dependent matching of the functional properties of the pre- and postsynaptic specializations at individual release sites (Marrus, 2004b).

A number of findings support this model. (1) In the DGluRIII mutant the remaining receptors are not evenly distributed opposite each active zone. Instead, certain receptor puncta contain a disproportionate number of receptors as assayed by immunocytochemistry and electrophysiology. (2) Sites with more receptors are located opposite larger active zones. The range of sizes in active zones in the mutant is very similar to that in the wild-type, but receptors are preferentially located opposite the larger active zones. For hippocampal neurons, the largest active zones have been found to have the highest probability of release. (3) In the DGluRIII mutant, the EJC saturates at a lower calcium level than does the wild-type EJC, even though presynaptic release is not saturated. This change in the calcium dependence of the EJC cannot be explained by postsynaptic saturation of a limited pool of receptors that are evenly distributed across the array of active zones. Release of a single vesicle may saturate the receptors in the apposed puncta (explaining the decrease in mEJP amplitude). However, the postsynaptic response to each release site is independent, so this type of saturation would affect the EJC in low and high calcium equally. Instead, the disparity in release rates and EJC amplitude at different calcium concentrations strongly suggests that there is preferential localization of receptors opposite sites of relatively higher release rates. If most receptors are utilized at lower calcium levels, then the progressive recruitment of additional release sites at higher calcium levels will have a reduced impact on EJC amplitude as a result of the relative paucity of receptors at these sites. (4) At the wild-type synapse there is a correlation between the brightest receptor puncta and the largest active zones. It has been demonstrated that the intensity of glutamate receptor staining correlates with both the amount and function of glutamate receptors (Marrus, 2004a). Based on the results from the DGluRIII mutant, it is suggested that these larger active zones at the wild-type synapse are also more active. The physiological dissociation of postsynaptic response and presynaptic release is not seen at the wild-type synapse because there are a sufficient number of receptors to localize opposite each active zone. Nonetheless, the correlation between receptor levels and active-zone size suggests that this matching mechanism functions during normal synaptic development and is not exclusively a homeostatic compensation used by the DGluRIII mutant (Marrus, 2004b).

Three possible mechanisms are considered for the matching of presynaptic release with postsynaptic receptor levels. (1) Presynaptic activity, potentially mediated by neurotransmitter release, could promote the local clustering or stabilization of apposed receptors. More active release sites would be more likely to activate postsynaptic receptors that would, in turn, promote the retention or growth of that receptor patch. During initial synapse formation, there is evidence that presynaptic activity does regulate the global localization of glutamate receptors to the synaptic region. The role of neurotransmitter release in regulating this global localization of receptors in the embryo is controversial, but its role during synaptic growth has not been investigated. (2) The second model postulates a retrograde specification of presynaptic release properties. In this view, larger receptor patches may initially localize opposite active zones in a random manner but then induce an increase in presynaptic function during development. There is good evidence for retrograde regulation of presynaptic properties at this synapse, but there is no evidence that such mechanisms act locally at a single active zone. (3) It is possible that the matching of pre- and postsynaptic functional properties is not, in fact, activity dependent. A transynaptic signal could coordinately regulate pre- and post-synaptic development and thus simultaneously control the function of the pre- and post-synaptic specializations. Although this third model is a formal possibility, the more parsimonious explanation, that the matching of activity levels is an activity-dependent process, is preferred (Marrus, 2004b).

What might be the purpose of an activity-dependent matching of pre- and post-synaptic function at individual release sites? During development, it would be an effective fail-safe mechanism for ensuring that receptors only localize opposite properly developed, i.e., functional, release sites. In addition, for synapses such as the NMJ, which demands high-fidelity synaptic transmission, it would increase synaptic strength by placing the most receptors at the sites of highest release (Marrus, 2004b).

GluRIIE, an essential Drosophila glutamate receptor subunit that functions in both central neuropil and neuromuscular junction

A Drosophila forward genetic screen for mutants with defective synaptic development has been identified and termed bad reception (brec). Homozygous brec mutants are embryonic lethal, paralyzed, and show no detectable synaptic transmission at the glutamatergic neuromuscular junction (NMJ). Genetic mapping, complementation tests, and genomic sequencing show that brec mutations disrupt a previously uncharacterized ionotropic glutamate receptor subunit, named here 'GluRIID.' GluRIID is expressed in the postsynaptic domain of the NMJ, as well as widely throughout the synaptic neuropil of the CNS. In the NMJ of null brec mutants, all known glutamate receptor subunits are undetectable by immunocytochemistry, and all functional glutamate receptors are eliminated. Thus, it is concluded that GluRIID is essential for the assembly and/or stabilization of glutamate receptors in the NMJ. In null brec mutant embryos, the frequency of periodic excitatory currents in motor neurons is significantly reduced, demonstrating that CNS motor pattern activity is regulated by GluRIID. Although synaptic development and molecular differentiation appear otherwise unperturbed in null mutants, viable hypomorphic brec mutants display dramatically undergrown NMJs by the end of larval development, suggesting that GluRIID-dependent central pattern activity regulates peripheral synaptic growth. These studies reveal GluRIID as a newly identified glutamate receptor subunit that is essential for glutamate receptor assembly/stabilization in the peripheral NMJ and required for properly patterned motor output in the CNS (Featherstone, 2005 ).

The Drosophila genome encodes 30 putative ionotropic glutamate receptor subunits (Littleton, 2000), but only 21 genes contain amino acid sequences thought to be required for pore formation (Sprengel, 2001). Three genes, called 'GluRIIA,' 'GluRIIB,' and 'GluRIII' (also known as 'GluRIIC') (Sprengel, 2001), have been shown to encode functional ionotropic glutamate receptor subunits localized to the NMJ. GluRIIC null mutants are embryonic lethal, and strong hypomorphs have many fewer GluRs at the larval NMJ (Marrus, 2004a; Marrus, 2004b). GluRIIA null mutants are viable but display reduced receptor channel open time, smaller miniature excitatory junction potentials, and reduced sensitivity to the antagonist argiotoxin 636. GluRIIB null mutants are also viable but show no significant change in receptor function (DiAntonio, 1999), suggesting that GluRIIB is less important for channel function or that most native receptors lack GluRIIB. Simultaneous deletion of both GluRIIA and GluRIIB causes embryonic lethality (Petersen, 1997; DiAntonio, 1999) and a presumed complete loss of functional glutamate receptors. Antibody staining suggests that GluRIIA and GluRIIB occupy adjacent partially overlapping domains (Marrus, 2004b; Chen, 2005a), indicating that at least some receptors contain either GluRIIA or GluRIIB but not both. Thus, it has been proposed that glutamate receptors at the Drosophila NMJ are composed of GluRIIC plus either GluRIIA or GluRIIB (Featherstone, 2005).

Another study (G. Qin, 2005) supports the findings reported here and also introduces a fifth NMJ subunit, GluRIIE. Thus, the Drosophila NMJ contains, at least, five different ionotropic glutamate receptor subunits, each encoded by a different gene: GluRIIA, GluRIIB, GluRIIC, GluRIID, and GluRIIE. Null mutations in GluRIIC, GluRIID, and GluRIIE each cause embryonic lethality, loss of functional NMJ glutamate receptors, and decreased immunoreactivity for other subunits. This suggests that GluRIIC, GluRIID, and GluRIIE are essential subunits contained by each glutamate receptor at the NMJ. In contrast, GluRIIA and GluRIIB are each individually dispensable, although at least one of these subunits is required for a functional receptor because deletion of both GluRIIA and GluRIIB is lethal. The subunit stoichiometry of mammalian non-NMDA glutamate receptors has never been definitively solved, but recent evidence from partial crystal structures strongly suggests that each ionotropic glutamate receptor is a 'dimer of dimers', e.g., composed of four subunits (Sun, 2002; Gouaux, 2004; Mayer, 2004). If Drosophila NMJ glutamate receptors are similarly tetrameric, then all existing data suggest that they are heterotetramers composed of one GluRIIC subunit, one GluRIID subunit, and one GluRIIE subunit, plus either one subunit of GluRIIA or one subunit of GluRIIB (but not both GluRIIA and GluRIIB). In other words, the Drosophila NMJ contains two subclasses of ionotropic glutamate receptor: (1) GluRIIA-containing receptors and (2) GluRIIB-containing receptors. This model is consistent with immunocytochemical and genetic results: (1) immunoreactivity for GluRIIA and GluRIIB is segregated such that clusters appear to contain either GluRIIA or GluRIIB but not both (Marrus, 2004b; Chen, 2005); (2) only some GluRIID clusters are immunoreactive for GluRIIA (Chen, 2005a), and (3) GluRIIA and GluRIIB are differentially trafficked and stabilized (Chen, 2005a). If GluRIIA and GluRIIB can be found in the same receptor (which presumably also contains the required subunits GluRIIC, GluRIID, and GluRIIE), then it must be concluded that Drosophila NMJ glutamate receptors are likely pentameric (Featherstone, 2005).

Because at least four different GluR subunits are required in the Drosophila NMJ in vivo, it suggests that there are at least four distinct subunit-dependent requirements for receptor assembly, trafficking, and/or stabilization. It is not clear how the different subunits play these roles. Amino acid sequence alignment shows that GluRIIA, GluRIIB, GluRIIC, GluRIID, and GluRIIE subunits differ most from each other near their N termini, a region that is known to be involved in ligand binding and possibly receptor assembly (Gouaux, 2004; Mayer, 2004). However, GluRIIA and GluRIIB show no obvious similarity in this region to explain why they might to be able to substitute for each other. GluRIIC has a class II C-terminal consensus PDZ (PSD-95/DLG/zona occludens-1)-binding domain (Marrus, 2004b), suggesting that GluRIIC might have a unique anchoring role. Neither GluRIIA nor GluRIIB contains recognizable PDZ-binding motifs, although stabilization of GluRIIB-containing receptors requires (apparently indirectly) the presence of the PDZ domain protein DLG (Chen, 2005a). Thus, it remains unclear how individual Drosophila GluR subunits contribute to receptor assembly and function. For mammalian receptors, answers to this question have typically been sought using heterologously expressed receptor subunits. However, this study suggests that there may be important differences in the mechanisms of receptor assembly and function in vivo. GluRIIA forms functional homomeric receptors when expressed in Xenopus oocytes, but in vivo overexpression of GluRIIA in muscle is essentially unable to overcome the requirement for GluRIID or form functional GluRs. It is conceivable that Xenopus oocytes contain endogenous proteins similar to kainite receptor subunits, and these proteins are sufficient for GluRIIA assembly and/or stabilization. Indeed, Xenopus oocytes are known to contain an endogenous protein (XenU1) that can substitute for the mammalian NMDA receptor subunit NR2. Alternatively, homomeric GluRIIA receptors may be only very inefficiently formed in oocytes, similar to the in vivo situation, but this inefficiency is not apparent outside of a synaptic context. Heterologous expression of other Drosophila NMJ GluR subunits has not been reported. Thus, these results suggest caution when interpreting some results using expressed subunits and highlight the importance of in vivo studies (Featherstone, 2005).

This study also suggests a functional role for glutamate receptors in the Drosophila CNS. Uniquely among known fly GluR subunits, GluRIID is expressed both in the NMJ and at central synapses. Excitatory transmission in the Drosophila CNS is thought to be predominantly cholinergic, although in situ data for several putative ionotropic glutamate receptor subunits shows that many subunits are expressed in the CNS. GluRIID is expressed at high levels throughout the synaptic neuropil of the ventral nerve cord, indicating that glutamatergic synapses in Drosophila might be much more widespread and pervasive than has been speculated previously. Consistent with this, it is shown that, in the absence of GluRIID, there is severe disruption of endogenous central motor pattern output. Interestingly, the only glutamategated responses that have been demonstrated in Drosophila neurons are inhibitory; glutamate-gated currents in voltage-clamped larval CNS neurons are prolonged (2-5 s), reverse at -55mV, and are blocked by picrotoxin. Nevertheless, GluRIID in the CNS could be part of an excitatory receptor that remains functionally unidentified. More intriguing is the possibility that GluRIID is an essential component of a kainate receptor-like glutamate-gated cation channel in Drosophila muscle but part of a glutamate-gated anion channel in the CNS. Glutamate-gated currents in embryos are, unfortunately, so far undetectable in embryos; thus, it has not been possible to determine whether GluRIID is required for CNS glutamate-gated anion currents. Although the nature of the GluRIID-containing receptor is unknown, loss of its function clearly causes dramatic changes in endogenous patterned activity within motor neurons. In mutants, many motor neurons lack detectable patterned motor output, and all cells show a striking reduction in the frequency of patterned motor output. This result minimally demonstrates that GluRIID-dependent glutamatergic transmission plays a vital modulatory role in controlling motor output from the CNS (Featherstone, 2005).

Although embryonic synaptogenesis appears normal in the absence of GluRIID, partial loss of GluRIID in viable brec mutants dramatically reduces postembryonic synaptic growth and differentiation. The loss of glutamate receptors in either the CNS or NMJ could cause NMJ morphology defects in two ways: (1) loss of muscle depolarization could disrupt a retrograde signal that induces presynaptic growth, or (2) disruption of endogenous central motor pattern activity could alter electrical activity-dependent presynaptic growth. In Drosophila, there is not good support for the former mechanism, because inhibition of muscle depolarization does not detectably alter NMJ arborization. In contrast, it is well established that neuronal electrical activity is positively correlated with the growth of the Drosophila presynaptic motor terminal. Thus the second model is the most parsimonious explanation. This conclusion raises the exciting prospect that the endogenous pattern of central electrical activity plays a critical role in sculpting postembryonic NMJ development (Featherstone, 2005).

Retrograde signaling by Syt 4 induces presynaptic release and synapse-specific growth

The molecular pathways involved in retrograde signal transduction at synapses and the function of retrograde communication are poorly understood. This study demonstrates that postsynaptic calcium 2+ ion (Ca2+) influx through glutamate receptors and subsequent postsynaptic vesicle fusion trigger a robust induction of presynaptic miniature release after high-frequency stimulation at Drosophila neuromuscular junctions. An isoform of the synaptotagmin family, synaptotagmin 4 (Syt 4), serves as a postsynaptic Ca2+ sensor to release retrograde signals that stimulate enhanced presynaptic function through activation of the cyclic adenosine monophosphate (cAMP)-cAMP-dependent protein kinase pathway. Postsynaptic Ca2+ influx also stimulates local synaptic differentiation and growth through Syt 4-mediated retrograde signals in a synapse-specific manner (Yoshihara, 2005).

Neuronal development requires coordinated signaling to orchestrate pre- and postsynaptic maturation of synaptic connections. Synapse-specific enhancement of synaptic strength as occurs during long-term potentiation, as well as compensatory homeostatic synaptic changes, have been suggested to require retrograde signals for their induction. Although retrograde signaling has been implicated widely in synaptic plasticity, the molecular mechanisms that transduce postsynaptic Ca2+ signals during enhanced synaptic activity to alterations in presynaptic function are poorly characterized. Because postsynaptic Ca2+ is essential for synapse-specific potentiation, it is important to characterize how Ca2+ can regulate retrograde communication at synapses (Yoshihara, 2005).

To dissect the mechanisms underlying activity-dependent synaptic plasticity, test were performed to see whether newly formed Drosophila glutamatergic neuromuscular junctions (NMJs), which have ∼30 active zones, show physiological changes after 100-Hz stimulation (5-1552+ chelator EGTA from the patch pipette caused a modest suppression of HFMR, whereas the fast Ca2+ chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA) induced strong suppression by 2.5 min of perfusion. Longer perfusion with BAPTA for 5 min before stimulation abolished HFMR, indicating HFMR is induced after postsynaptic Ca2+ influx (Yoshihara, 2005).

Ca2+-induced vesicle fusion in presynaptic terminals provides a temporally controlled and spatially restricted signal essential for synaptic communication. Postsynaptic vesicles within dendrites have been visualized by transmission electron microscopy, and dendritic release of several neuromodulators has been reported. To test whether postsynaptic vesicle fusion might underlie the Ca2+-dependent release of retrograde signals, postsynaptic vesicle recycling was blocked by using the dominant negative shibirets1 mutation, which disrupts endocytosis at elevated temperatures. shibirets1 was expressed specifically in postsynaptic muscles by driving a UAS-shibirets1 transgene with muscle-specific myosin heavy chain (Mhc)-Gal4, keeping presynaptic activity intact. At the permissive temperature (23°C), high-frequency stimulation induced normal HFMR. However, raising the temperature to 31°C suppressed HFMR in the presence of postsynaptic shibirets1, whereas wild-type animals displayed normal HFMR at 31°C. Basic synaptic properties in Mhc-Gal4, UAS-shibirets1 animals were not affected at either the permissive or the restrictive temperature. The suppression of HFMR is not due to irreversible damage induced by postsynaptic UAS-shibirets1 expression, because a second high-frequency stimulation after recovery to the permissive temperature induced normal HFMR (Yoshihara, 2005).

The synaptic vesicle protein synaptotagmin 1 (Syt 1) is the major Ca2+ sensor for vesicle fusion at presynaptic terminals but is not localized postsynaptically. It has recently been shown that another isoform of the synaptotagmin family, synaptotagmin 4 (Syt 4), is present in the postsynaptic compartment (Adolfsen, 2004), suggesting Syt 4 might function as a postsynaptic Ca2+ sensor. Syt 4 immunoreactivity is observed in a punctate pattern surrounding presynaptic terminals, suggesting Syt 4 is present on postsynaptic vesicles. Postsynaptic vesicle recycling was blocked by using the UAS-shibirets1 transgene driven with Mhc-Gal4. Without a temperature shift, Syt 4-containing vesicles showed their normal postsynaptic distribution surrounding presynaptic terminals. When the temperature was shifted to 37oC for 10 min in the presence of high-K+ saline containing 1.5 mM Ca2+ to drive synaptic activity, Syt 4-containing vesicles translocated to the plasma membrane. After recovery at 18oC for 20 min, postsynaptic vesicles returned to their normal position. Removing extracellular Ca2+ during the high-K+ stimulation resulted in vesicles that did not translocate to the postsynaptic membrane (Yoshihara, 2005).

To further test whether the Syt 4 vesicle population undergoes fusion with the postsynaptic membrane as opposed to mediating fusion between intracellular compartments, transgenic animals were constructed expressing a pH-sensitive green fluorescent protein (GFP) variant (ecliptic pHluorin) fused at the intravesicular N terminus of Syt 4. Ecliptic pHluorin increases its fluorescence 20-fold when exposed to the extracellular space from the acidic lumen of intracellular vesicles during fusion. Expression of Syt 4-pHluorin in postsynaptic muscles resulted in intense fluorescence at specific subdomains in the postsynaptic membrane, defining regions where Syt 4 vesicles undergo exocytosis. The fluorescence was not diffusely present over the postsynaptic membrane but directed to restricted compartments. Mhc-Gal4, UAS-Syt 4-pHluorin larvae were costained with antibodies against the postsynaptic density protein, DPAK, and nc82, a monoclonal antibody against a presynaptic active zone protein. Syt 4-pHluorin colocalized with DPAK and localized adjacent to nc82, demonstrating that Syt 4-pHluorin translocates from postsynaptic vesicles to the plasma membrane at postsynaptic densities opposite presynaptic active zones (Yoshihara, 2005).

To examine the function of Syt 4-dependent postsynaptic vesicle fusion, the phenotypes of a syt 4 null mutant (syt 4BA1) and a syt 4 deficiency (rn16) were tested. Mutants lacking Syt 4 hatch from the egg case 21 hours after egg laying at 25oC, similar to wild type, and grow to fully mature larvae that pupate and eclose with a normal time course. To determine whether postsynaptic vesicle fusion triggered by Ca2+ influx is required for HFMR, the effects of high-frequency stimulation in syt 4 mutants were analyzed. In contrast to controls, the increase of miniature release was eliminated in syt 4 mutants. Postsynaptic expression of a UAS-syt 4 transgene completely restored HFMR in the null mutant, demonstrating that postsynaptic Syt 4 is required for triggering enhanced presynaptic function. Presynaptic expression of a UAS-syt 4 transgene did not restore HFMR. In addition, postsynaptic expression of a mutant Syt 4 with neutralized Ca2+-binding sites in both C2A and C2B domains did not rescue HFMR, indicating that retrograde signaling by Syt 4 requires Ca2+ binding (Yoshihara, 2005).

The large increase in miniature frequency observed during HFMR is similar to the enhancement of presynaptic release after activation of cyclic adenosine monophosphate (cAMP)-dependent protein kinase (PKA) described in Aplysia and Drosophila. Bath application of forskolin, an activator of adenylyl cyclase, results in a robust enhancement of miniature frequency at Drosophila NMJs similar to that observed during HFMR, suggesting retrograde signals may function to increase presynaptic cAMP. To test the role of the cAMP-PKA pathway in HFMR, DC0 mutants were assayed for the presence of HFMR. DC0 encodes the major catalytic subunit of PKA in Drosophila and has been implicated in olfactory learning. Similar to the lack of forskolin-induced miniature induction, DC0 null mutants lacked HFMR. Bath application of forskolin in syt 4 mutants resulted in enhanced miniature frequency, suggesting activation of the cAMP pathway can bypass the requirement for Syt 4 in synaptic potentiation (Yoshihara, 2005).

To further explore the role of retrograde signaling at Drosophila synapses, the role of activity was tested in synapse differentiation and growth. During Drosophila embryonic development, presynaptic terminals undergo a stereotypical structural change from a flat path-finding growth cone into varicose synaptic terminals through dynamic reconstruction. Such developmental changes in synaptic structure may share common molecular mechanisms with morphological changes induced during activity-dependent plasticity. Synaptic transmission was eliminated by using a deletion mutation that removes the postsynaptic glutamate receptors, DGluRIIA and DGluRIIB (referred to as GluRs). Postsynaptic currents normally induced by nerve stimulation were completely absent in the mutants (gluR). Miniatures were also eliminated, even at elevated extracellular Ca2+ concentrations of 4 mM. In the absence of GluRs, the presynaptic morphology of motor terminals is abnormal, even though GluRs are only expressed in postsynaptic muscles. GluR-deficient terminals maintain a flattened growth cone-like structure and fail to constrict into normal synaptic varicosities. Synaptic development was also assayed in a null mutant of the presynaptic plasma membrane t-SNARE [SNAP (soluble N-ethylmaleimide-sensitive factor attachment protein) receptor], syntaxin (syx), which eliminates neurotransmitter release, providing an inactive synapse similar to that in the gluR mutant. syx null mutants also have abnormal growth cone-like presynaptic terminals with less varicose structure (Yoshihara, 2005).

Because activity is required for synapse development, whether Syt 4-dependent vesicle fusion may be required, similar to its role in acute retrograde signaling during HFMR, was tested. Physiological analysis revealed that the amplitude of evoked currents in mutants lacking Syt 4 was moderately reduced compared with wild type, suggesting weaker synaptic function or development. Similar to the morphological phenotype of the gluR mutant, syt 4 null mutant embryos showed defective presynaptic differentiation. Nerve terminals lacking Syt 4 displayed reduced varicose structure, whereas wild-type terminals have already formed individual varicosities at this stage of development. Postsynaptic expression with a UAS-syt 4 transgene rescued the physiological and morphological phenotypes. Syt 4 Ca2+-binding deficient mutant transgenes did not rescue either the morphological immaturity or the reduced amplitude of evoked currents, even though Syt 4 immunoreactivity at the postsynaptic compartment was restored by muscle-specific expression of the mutant syt 4 transgene, similar to the wild-type syt 4 transgene and endogenous Syt 4 immunoreactivity (Yoshihara, 2005).

Mammalian syt 4 was originally identified as an immediate-early gene that is transcriptionally up-regulated by nerve activity in certain brain regions. Therefore, this study analyzed gain-of-function phenotypes caused by postsynaptic Syt 4 overexpression specifically in muscle cells to increase the probability of postsynaptic vesicle fusion. Syt 4 overexpression induced overgrowth of presynaptic terminals in mature third instar larvae, in contrast to overexpression of Syt 1, which does not traffic to Syt 4-containing postsynaptic vesicles. In addition to synaptic overgrowth, Syt 4 overexpression occasionally induced the formation of abnormally large varicosities. Postsynaptic overexpression of the Syt 4 Ca2+-binding mutant did not induce synaptic overgrowth, indicating that retrograde signaling by Syt 4 also requires Ca2+ binding to promote synaptic growth (Yoshihara, 2005).

To determine whether the cAMP-PKA pathway is important in activity-dependent synaptic growth, the effects of PKA on synaptic morphology were assayed. Expression of constitutively active PKA presynaptically using a motor neuron-specific Gal4 driver induced not only synaptic overgrowth but also larger individual varicosities in mature third instar larvae, similar to those induced by postsynaptic overexpression of Syt 4. These observations are consistent with the presynaptic overgrowth observed in the learning mutant, dunce, which disrupts the enzyme that degrades cAMP, and with studies in Aplysia implicating PKA in synaptic varicosity formation. The loss-of-function phenotype of PKA mutants (DC0B3) were characterized at the embryonic NMJ to compare with gluR and syt 4 mutants. Presynaptic terminals in the DC0 mutant were morphologically aberrant, with abnormal growth cone-like features and less varicose structure. Postsynaptic expression of a constitutively active PKA transgene in the DC0 or syt 4 mutant backgrounds rescued the immature morphology, suggesting activation of PKA is downstream of Syt 4-dependent release of retrograde signals (Yoshihara, 2005).

Similar to the role of Syt 1-dependent synaptic vesicle fusion in triggering synaptic transmission at individual synapses, Syt 4-dependent vesicle fusion might trigger synapse-specific plasticity and growth. To test synapse specificity, advantage was taken of the specific properties of the Drosophila NMJ at muscle fibers 6 and 7, where two motorneurons innervate both muscle fibers 6 and 7 during development. Syt 4 was expressed specifically in embryonic muscle fiber 6 but not muscle fiber 7 by using the H94-Gal4 driver. If Syt 4-dependent retrograde signals induce general growth of the motorneuron, one would expect to see a proliferation of synapses on both muscle fibers. Alternatively, if Syt 4 promoted local synaptic growth, one would expect specific activation of synapse proliferation only on target muscle 6, releasing the Syt 4-dependent signal. UAS-syt 4 driven by H94-Gal4 increased innervation on muscle fiber 6 compared with that on muscle fiber 7 in third instar larvae. Control experiments with Syt 4 Ca2+-binding deficient mutant transgenes, or a transgene encoding Syt 1, did not result in proliferation. Thus, synaptic growth can be preferentially directed to specific postsynaptic targets where Syt 4-dependent retrograde signals predominate, allowing differential strengthening of active synapses via local rewiring (Yoshihara, 2005).

On the basis of the results described in this study, a local feedback model is proposed for activity-dependent synaptic plasticity and growth at Drosophila NMJs. Synapse-specific Ca2+ influx triggers postsynaptic vesicle fusion through Syt 4. Fusion of Syt 4-containing vesicles with the postsynaptic membrane releases locally acting retrograde signals that activate the presynaptic terminal, likely through the cAMP pathway. Active PKA then triggers cytoskeletal changes by unknown effectors to induce presynaptic growth and differentiation. Moreover, PKA is well known to facilitate neurotransmitter release directly, triggering a local synaptic enhancement of presynaptic release as shown in HFMR. Therefore, postsynaptic vesicular fusion might initiate a positive feedback loop, providing a localized activated synaptic state that can be maintained beyond the initial trigger (Yoshihara, 2005).

As a general mechanism for memory storage, Hebb postulated that potentiated synapses maintain an activated state until structural changes occur to consolidate alterations in synaptic strength. The current results demonstrate that acute plasticity and synapse-specific growth require Syt 4-dependent retrograde signaling at Drosophila NMJs. The feedback mechanism described in this study could be a molecular basis for both input-specific postsynaptic tagging and an output-specific presynaptic mark or tag for long-lasting potentiation. The regenerative nature of a positive feedback signal allows individual synapses to be tagged in a discrete all-or-none manner until synaptic rewiring is completed. The synaptic tag is maintained as a large increase in miniature frequency at Drosophila NMJs, suggesting a previously unknown role for miniature release in neuronal function. The spatial resolution for input and output specificity would result from the accuracy insured by Ca2+-dependent vesicle fusion and subsequent diffusion, similar to the precision of presynaptic neurotransmitter release (Yoshihara, 2005).

NF-kappaB, IkappaB, and IRAK control glutamate receptor density at the Drosophila NMJ

NF-κB signaling has been implicated in neurodegenerative disease, epilepsy, and neuronal plasticity. However, the cellular and molecular activity of NF-κB signaling within the nervous system remains to be clearly defined. This study shows that the NF-κB and IκB homologs Dorsal and Cactus surround postsynaptic glutamate receptor (GluR) clusters at the Drosophila NMJ. Mutations in dorsal, cactus, and IRAK/pelle kinase specifically impair GluR levels, assayed immunohistochemically and electrophysiologically, without affecting NMJ growth, the size of the postsynaptic density, or homeostatic plasticity. Additional genetic experiments support the conclusion that cactus functions in concert with, rather than in opposition to, dorsal and pelle in this process. Finally, evidence is provided that Dorsal and Cactus act posttranscriptionally, outside the nucleus, to control GluR density. Based upon these data it is speculated that Dorsal, Cactus, and Pelle function together, locally at the postsynaptic density, to specify GluR levels (Heckscher, 2007).

NF-κB signaling has been implicated in the mechanisms of neural plasticity, learning, epilepsy, neurodegeneration, and the adaptive response to neuronal injury. The data presented in this study advance the understanding of neuronal NF-κB signaling in two ways. First, multiple lines of evidence are presented that NF-κB/Dorsal signaling is required for the control of GluR density at the NMJ. These data provide a synaptic function for NF-κB signaling that may be directly relevant to the diverse activities ascribed to NF-κB in the nervous system. Second, molecular and genetic evidence is provided that Dorsal, Cactus, and Pelle may function posttranscriptionally, at the postsynaptic side of the NMJ, to specify GluR density during postembryonic development (Heckscher, 2007).

Several independent lines of experimentation suggest that Cactus, Dorsal, and Pelle function together at the PSD to specify GluR density. Evidence is provided that Cactus and Dorsal localize to a similar postsynaptic domain. In addition, overexpression of a GFP-tagged Pelle protein that is sufficient to rescue a pelle mutation, can traffic to the PSD where Cactus and Dorsal reside. Next, genetic evidence is presented that cactus, dorsal, and pelle function together, in the same genetic pathway, to control GluR density. It is particularly surprising that mutations in cactus behave similarly to dorsal and pelle. In other systems (embryonic patterning and immunity), Cactus inhibits Dorsal-mediated transcription by binding and sequestering cytoplasmic Dorsal protein. As a result, in these other systems, cactus mutations cause phenotypes that are opposite to those observed in dorsal mutations. This study used the same cactus and dorsal mutations that previously have been observed to generate the predicted opposing phenotypes during embryonic patterning, and yet it was observed that cactus phenocopies the dorsal mutations. In addition, genetic epistasis experiments indicate that these genes function together to facilitate GluR density. Thus, at the NMJ, Cactus functions in concert with, rather than in opposition to, Dorsal (Heckscher, 2007).

One explanation for this observation could be that Dorsal does not function as a nuclear transcription factor during the control of GluR levels. In support of this idea it has been demonstrated that (1) Dorsal protein is not detected in the nucleus, (2) reporters of Dorsal-dependent transcription fail to show activity in muscle nuclei, and (3) mutation of the Dorsal transactivation domain, dlU5 does not impair GluR abundance even though this same mutation has been shown to impair transcription-dependent patterning during embryogenesis. An alternative explanation for the observation that dorsal and cactus have similar phenotypes at the NMJ could be that Cactus and Dorsal act synergistically to control the transcription of GluRs at the NMJ. Indeed, there is evidence in other systems that IκB can shuttle with NF-κB to the nucleus. A previous study shows Cactus accumulation in Drosophila larval muscle nuclei in a dorsal mutant background (Cantera, 1999). However, this result could not be repeated despite examination of Cactus localization in five allelic combinations of dorsal. Furthermore, the data from vertebrate systems suggest that IκB should shuttle into the nucleus with NF-κB, not in its absence. Thus, a model is favored in which Dorsal and Cactus function together at the postsynaptic membrane to facilitate GluR abundance during development (Heckscher, 2007).

If this model is correct, then it is predicted that NF-κB does not control GluR density through transcriptional regulation. This prediction is supported by two experimental observations: (1) GluR transcript levels (assessed by QT PCR) are not statistically different from wild-type in dorsal and cactus mutations that cause an ~50% decrease in GluR abundance; (2) it was demonstrated that overexpression of a myc-tagged GluRIIA cDNA using a heterologous, muscle-specific promoter is not able to restore synaptic GluRIIA levels in either the cactus or dorsal mutant backgrounds. These data are consistent with Dorsal and Cactus acting posttranscriptionally to control GluR density at the NMJ. There are two general mechanisms by which GluR levels could be controlled posttranscriptionally: (1) altered receptor delivery to the NMJ or (2) altered receptor internalization/degradation. If receptor internalization/degradation were enhanced in the cactus, dorsal, or pelle mutant backgrounds, one might expect GluRIIA-myc overexpression to overcome this change and restore normal receptor levels. In addition, less myc-tagged protein might be seen in the mutants in comparison to wild-type. This is not what was observed. Therefore, the hypothesis is favored that Cactus, Dorsal, and Pelle function together to promote the delivery of glutamate receptors to the NMJ during development (Heckscher, 2007).

The possibility that Cactus, Dorsal, and Pelle act posttranscriptionally to control GluR density raises many questions. For example, do Dorsal and Cactus exist as a protein complex at the PSD? If so, is this complex regulated and how might such a complex influence GluR density? Since pelle kinase-dead mutants impair GluR density, it is possible that Dorsal and Cactus recruit Pelle to the PSD. If so, what are the targets of Pelle kinase that are relevant to establishing or maintaining the proper density of glutamate receptors at the PSD? Finally, the demonstration that cytoplasmic NF-κB/Dorsal can influence GluR density does not rule out the possibility that NF-κB/Dorsal may also translocate to the muscle nucleus at the Drosophila NMJ under certain stimulus conditions. Indeed, in both the vertebrate central and peripheral nervous systems NF-κB is found within neuronal and muscle nuclei, and nuclear translocation can be stimulated by neuronal activity, glutamate, injury, and disease. For nuclear entry of Dorsal, two events must occur: (1) Cactus must be degraded and (2) Dorsal must be phosphorylated. It remains possible that one or both of these criteria are not met during the normal development of the Drosophila NMJ but could be met under as-yet-to-be-identified stimulus conditions. The possibility that NF-κB acts both locally at the synapse and globally via the nucleus is not unique to this signaling pathway. A similar organization has been documented for wingless/wnt signaling where noncanonical cytoplasmic signaling can impact cytoskeletal organization while canonical signaling involves the nuclear translocation of downstream beta-catenin and TCF-dependent gene transcription (Heckscher, 2007).

It remains unknown how NF-κB signaling is activated at the Drosophila NMJ. In Drosophila embryonic patterning and innate immunity, NF-κB signaling is initiated through activation of Toll or Toll-like receptors. There are nine Toll and Toll-like receptors encoded in the Drosophila genome. However, none of these receptors appear to be present in Drosophila larval muscle. The Toll receptor is expressed in a subset of embryonic muscle fibers, but is absent from larval muscle. None of the Toll-like receptors are expressed in Drosophila embryonic muscle and none appear to be expressed in larval muscle. An alternative possibility is that TNF-α receptors activate NF-κB in Drosophila muscle as has been observed in vertebrate skeletal muscle. Indeed, a TNF-α receptor homolog (Wengen) has been identified, and it is expressed in Drosophila skeletal muscle. The possibility that TNF-α signaling is mediated via NF-κB is intriguing given the recent demonstration that TNF-α regulates GluR abundance in the vertebrate central nervous system. In both cultured neurons and hippocampal slices glial-derived TNF-α signaling is required for the increase in postsynaptic AMPA receptor abundance observed following chronic activity blockade. Thus, the current data in combination with work in the vertebrate CNS raise the possibility that a conserved TNFα/NF-κB signaling system controls GluR abundance at both neuromuscular and central synapses during development and in response to chronic activity blockade (Heckscher, 2007).

Presynaptic establishment of the synaptic cleft extracellular matrix is required for post-synaptic differentiation

Formation and regulation of excitatory glutamatergic synapses is essential for shaping neural circuits throughout development. A genetic screen for synaptogenesis mutants in Drosophila identified mind the gap (mtg), which encodes a secreted, extracellular N-glycosaminoglycan-binding protein. Mtg is expressed neuronally and detected in the synaptic cleft, and is required to form the specialized transsynaptic matrix that links the presynaptic active zone with the post-synaptic glutamate receptor (GluR) domain. Null mtg embryonic mutant synapses exhibit greatly reduced GluR function, and a corresponding loss of localized GluR domains. All known post-synaptic signaling/scaffold proteins functioning upstream of GluR localization are also grossly reduced or mislocalized in mtg mutants, including the dPix-dPak-Dock cascade and the Dlg/PSD-95 scaffold. Ubiquitous or neuronally targeted mtg RNA interference (RNAi) similarly reduce post-synaptic assembly, whereas post-synaptically targeted RNAi has no effect, indicating that presynaptic MTG induces and maintains the post-synaptic pathways driving GluR domain formation. These findings suggest that MTG is secreted from the presynaptic terminal to shape the extracellular synaptic cleft domain, and that the cleft domain functions to mediate transsynaptic signals required for post-synaptic development (Rohrbough, 2007).

Glutamatergic synapse formation and maturation is critical for sculpting neural circuits. Synaptogenesis defects cause crippling neurological disabilities ranging from motor ataxias to profound mental retardation, and subsequent modulation of glutamatergic synapses is a lifelong dynamic process underlying the ability to learn and remember. A critical hypothesis, developed largely from the vertebrate neuromuscular junction (NMJ) model, is that presynaptic signals trigger post-synaptic assembly and modulation. Proposed secreted signals include Agrin, WNTs, FGFs and Narp. Transmembrane synaptic signaling proteins include SynCAM, EphrinB-EphB, and β-neurexin/neuroligin. Both classes of signals are proposed to interact intimately with a specialized synaptic cleft extracellular matrix (ECM), which is molecularly distinct from nonsynaptic ECM (Rohrbough, 2007).

Synaptic cleft ECM components include heterotrimeric (α/β/γ) laminin glycoproteins and heparan sulfate proteoglycans (HSPGs), which bind extracellular glucuronic acid and N-acetyl glucosamine (GlcNAc) polysaccharides. The secreted HSPGs Agrin and Perlecan are established regulators of NMJ synaptogenesis. The HSPG Syndecan (Sdc)-2 is similarly implicated in hippocampal synapse formation and plasticity. ECM signaling at the mammalian NMJ also acts via the α/β-dystroglycan-glycoprotein complex (DGC). Numerous integrin receptors localize to mammalian NMJs and central glutamatergic synapses, and integrin-ECM interactions regulate aspects of synaptic development and modulation, including glutamatergic transmission and plasticity (Rohrbough, 2007).

The Drosophila glutamatergic NMJ contains an ultrastructurally distinctive synaptic cleft ECM present only between pre- and post-synaptic densities, but little is known of its molecular composition. HSPGs, including laminin-binding Sdc and GPI-anchored Dallylike (Dlp), and the secreted Hikaru Genki (HIG) are localized to the cleft ECM and regulate synaptic differentiation. A post-synaptic Dystrophin scaffold has recently been shown to regulate NMJ maturation. Drosophila WNT Wingless (Wg) is a secreted anterograde synaptic maturation signal that acts via its pre/post-synaptic receptor Frizzled (Dfz2). Integrin ECM receptors and laminin are localized to NMJ synapses and regulate synapse formation, and both structural and functional development (Rohrbough, 2007).

The Drosophila NMJ post-synaptic domain contains two subclasses of tetrameric AMPA/kainate-like glutamate receptors (GluRs), composed of either IIA (A-class) or IIB (B-class) subunits, in combination with common required IIC, IID, and IIE subunits. The PDZ-domain scaffold Discs Large (Dlg), a PSD-95/SAP70 homolog, plays a key role in localizing post-synaptic proteins including B-class GluRs. The Drosophila p21-activated kinase (dPak), a serine threonine kinase activated by GTPases Rac and Cdc42, and its localizing Rho-type GEF, dPix, also play essential roles in the post-synaptic domain. dPAK interacts directly via its kinase domain with Dlg, and is required for Dlg synaptic expression. In a second pathway, dPak binding to the adaptor Dreadlocks (Dock), a Src homology (SH3 and SH2)-containing Nck homolo, regulates A-class GluR abundance. The dPix-dPak-Dock and dPak-Dlg pathways therefore converge to regulate localization of both GluR classes in the post-synaptic domain (Rohrbough, 2007).

Extensive work in this model system has established that the presynaptic neuron induces post-synaptic differentiation, inducing development and modulation of GluR domains. GluR domain formation involves lateral membrane receptor diffusion, receptor sequestration/anchoring mechanisms, regulated GluR subunit transcription, and local post-synaptic translation. To define the molecular mechanisms regulating functional post-synaptic differentiation, a systematic mutagenesis screen was undertaken to isolate mutants with defective post-synaptic assembly. This approach has revealed mind the gap (mtg), which encodes a secreted protein required for the formation of the synaptic cleft matrix, as well as a localization of the signaling pathways regulating GluR domains. Presynaptic mtg knockdown inhibits post-synaptic differentiation, indicating that presynaptically secreted MTG organizes the extracellular cleft domain and is critically required for transsynaptic signaling that induces post-synaptic differentiation (Rohrbough, 2007).

The mind the gap (mtg) gene was isolated in an unbiased forward Drosophila genetic screen for novel mutants blocking functional differentiation of the glutamatergic NMJ synapse. This screen has now revealed numerous novel genes and mechanisms regulating pre- or post-synaptic development, including GluR subunits and genes regulating functional receptor expression. Loss of mtg results in a severe, ~80% loss of GluR function at the NMJ. Notably, this phenotype represents the most severe mutant GluR impairment ever reported, with the exception of genetic mutants for the requisite GluR subunits (IIC-IIE) themselves . MTG is expressed neuronally and localized synaptically, contains a well-conserved secretory signal sequence, and is secreted and binds GlcNAc in vitro. These findings support the working hypothesis that MTG is secreted into the synaptic cleft to bind ECM glycosaminoglycans (GAG) or proteoglycans (PG) during glutamatergic synaptogenesis (Rohrbough, 2007).

The mtg developmental expression profile closely parallels the timing of NMJ synapse formation and functional differentiation. Expression increases sharply with initial nerve-muscle contact (12-13 h after fertilization), and peaks at 16-17 h, correlating with post-synaptic GluR domain assembly (15-17 h). MTG protein is concentrated in embryonic neurons, and becomes increasingly localized with development to NMJ synaptic domains, and to other cell adhesion sites such as muscle attachment sites, where many synaptic signaling (Pix, Pak, integrins) and scaffold proteins (Dlg) are colocalized. MTG is found in the presynaptic terminal and post-synaptic SSR, clearly detected within the extracellular synaptic cleft domain, consistent with presynaptic secretion as well as a transsynaptic regulatory role. In the absence of MTG, post-synaptic GluR puncta apposed to the presynaptic terminal are lost, and GluRs are dispersed in the nonsynaptic membrane, resulting in profound functional transmission loss. The mislocalized GluRs appear nonfunctional based on the muscle response to exogenously applied glutamate, which demonstrates a dramatic overall loss of functional cell surface GluRs. All PSD proteins known to act in the upstream GluR regulatory pathways (dPix, dPak, Dlg, Dock) are severely reduced or mislocalized in mtg mutants, indicating that MTG acts at an upstream organizing step required to establish this cascade of post-synaptic interactions. The loss of synaptic cleft matrix material at mtg mutant synapses indicates first, that MTG is required for this signaling domain to be established, and second, suggests that this domain has an important inductive, instructive role in post-synaptic assembly (Rohrbough, 2007).

Genetically rescuing the mtg mutant has not been accomplished with a wild-type copy of the mtg gene. Since mtg is present at low overall levels during much of development, it is likely that mtg expression timing and/or level must be precisely regulated for normal protein function and animal viability. Transgenic rescue with tissue-targeted Gal4-driven mtg expression may therefore result in a deleterious overexpression condition. Transgenically expressed mtg RNAi, however, phenocopies mtg mutant phenotypes. Ubiquitous mtg RNAi causes early lethality and defective post-synaptic assembly, with reduced synaptic localization of GluRs and other upstream regulatory proteins. These phenotypes are more severe in the mtg1 mutant, which exhibits lethality at the hatching stage, and also nonsynaptic mislocalization of post-synaptic proteins. Therefore, the findings are consistent with the expected effects of RNAi as a partial loss-of-function condition. Most conclusively, targeted RNAi knockdown of MTG in the presynaptic neuron impairs post-synaptic differentiation and the assembly of GluR domains with a similar severity to the ubiquitous RNAi condition. In contrast, muscle-targeted mtg RNAi has no detectable effect on movement or animal viability, and does not cause any detectable synaptic impairment or defects in post-synaptic assembly. Taken together, these results support the identity of the mtg gene and suggest that presynaptically secreted MTG protein is required for post-synaptic development. It is concluded that MTG is a critical element in the presynaptic inductive mechanism (Rohrbough, 2007).

MTG has a cysteine-rich carbohydrate-binding module (CBM) with homology with ChtBDs found in peritrophic matrix proteins, lectins, and other known ECM proteins. This domain contains six conserved cysteines predicted to form three disulfide bridges within a β-folded carbohydrate-binding structure that binds GlcNAc moieties. Lectins that specifically bind GlcNAc (WGA) and GalNAc (DBA, VVA) are commonly used synaptic cleft or post-synaptic markers at the vertebrate NMJ. This study confirms that WGA-binding targets are localized to Drosophila NMJ boutons, and that extracted MTG protein binds GlcNAc in vitro, suggesting that MTG recognizes GlcNAc-containing target(s) in the synaptic cleft. Drosophila S2 cells transformed with MTG-GFP secrete the protein, which accumulates both on the outer surface of the cells and in the medium, supporting its extracellular localization. However, it has not yet been possible to replicate the GlcNAc binding with purified MTG protein, the ideal experiment to confirm and further probe the binding specificity. GlcNAc-containing carbohydrate and GAG scaffolds (e.g., chitin, hyaluronic acid [HA], heparin) and PG (e.g., heparin sulfate, chondroitin sulfate) are components of neuronal and muscle ECM, and concentrated within the specialized synaptic cleft matrix at the NMJ and other synapses in multiple species. It was recently shown that loss of GlcNAc transferase alters In Drosophila NMJ synaptic structure, function, and locomotory behavior, independently demonstrating GlcNAc-mediated interactions have roles in synaptic maturation (Rohrbough, 2007).

MTG does not share significant overall whole-sequence homology with an identified vertebrate protein. It is increasingly recognized that many Drosophila proteins are conserved at a structural level to serve identical functions with mammalian functional homologs; among many examples are neurotrophin-like proteins and their receptors and olfactory receptors. Several vertebrate protein families contain cysteine-rich domains with predicted structural homology with MTG, including the TGF-β, GPH, PDGF, and NGF growth factor families. CDM/ChtBD-related domains are common in other secreted protein families, including knottins, mucins and lectins. These domains contain six to 10 cysteine residues predicted to form disulfide bridges, which mediate homo- or heterodimer formation, carbohydrate binding, and extracellular ligand-target/receptor interactions. Laminin integrin ligands contain the structurally related type-1 EGF domain, which is a site of receptor recognition and ECM binding. The consistent extracellular function of these related disulfide-forming protein domains is to mediate ECM protein interactions and ligand-receptor intercellular signaling. A similar function is proposed for MTG in organizing the synaptic cleft matrix and mediating transsynaptic signaling critical for synaptogenesis (Rohrbough, 2007).

Null mtg mutants show a reduction or complete absence of electron-dense synaptic matrix, suggesting a loss or gross disorganization of multiple synaptic ECM components and binding proteins. This is the first report of such a cleft phenotype reported for a functional synaptic mutant. Vertebrate neuronal and perisynaptic ECM consists of a GAG scaffold matrix (e.g., heparin sulfate, chondroitin sulfate, HA), numerous bound proteins, laminins, and transmembrane molecules/receptors interacting with the matrix, including NCAM family proteins and integrins. Much less is known, however, about the mechanisms linking the synaptic cleft matrix structural and signaling environment to post-synaptic assembly. The regulation of acetylcholine receptor (AChR) expression/localization at vertebrate NMJs by the agrin-agrin receptor (MusK)-rapsyn pathway provides an obvious framework for comparison, although the role of agrin in post-synaptic receptor maintenance versus domain assembly has recently been redefined. The vertebrate cholinergic NMJ is concentrated in GlcNAc- and GalNAc-containing GAG and PG, in particular secreted (agrin, perlecan) and transmembrane (syndecan) heparin sulfate proteoglycans (HSPG); laminins, which act as integrin ligands and interact directly with other membrane proteins including Ca2+ channels; and ECM transmembrane receptors, including receptor tyrosine kinases (RTK) and integrins, activated by binding of ECM ligands. Integrins represent an appealing ECM receptor-mediated link to Pix-Pak pathway activation in vertebrates, and hippocampal synapse formation and plasticity in the hippocampus (Rohrbough, 2007).

Three integrin receptor subtypes (αPS1/βPS, αPS2/βPS, αPS3/βPS) localize to the Drosophila NMJ, and regulate synaptic structural development and functional transmission properties, including activity-dependent plasticity. Integrins, RGD domain-containing laminins, and the secreted synaptic cleft protein Hikaru Genki regulate synapse formation and ultrastructure. More recently, two synaptic HSPGs, Syndecan (Sdc) and Dallylike (Dlp), were shown to localize to the NMJ and regulate presynaptic terminal growth and AZ formation, respectively. The receptor tyrosine phosphatase dLAR is a receptor for both Sdc and Dlp and interacts with these ligands via their GAG chains. Drosophila dystrophin has also recently been shown to localize post-synaptically, and form a synaptic glycoprotein complex with extracellular dystroglycan. Surprisingly, Drosophila dystrophin regulates presynaptic properties, but not post-synaptic GluR expression. These studies do not suggest that MTG is a major structural component of the synaptic cleft matrix, but rather that MTG has a necessary role in organizing the broader structure and transsynaptic signaling capabilities of the synaptic cleft ECM. If this hypothesis is correct, the severity of the mtg mutant phenotypes may be due to a disruption of multiple transsynaptic signaling pathways (Rohrbough, 2007).

The findings of this study suggest that interactions between MTG and its GAG/PG-binding partner(s) in the synaptic cleft matrix are linked to the activation and localization of PSD signaling pathways. Protein localization/binding studies and three-dimensional structural models suggest that the cleft domain is a dense GAG scaffold extensively linked by secreted matrix PG, including chondroitin sulfate PG (tenascins, lecticans, phosphacans), and HSPGs (perlecan, β-glycans, agrins), as well as by transmembrane HSPGs (syndecans, systroglycans), GPI-linked membrane-bound proteins (glypicans), NCAMs, and integrins. This matrix could act in part to sterically trap or limit lateral movement of transmembrane proteins to maintain them in the synapse. Such a mechanism could preferentially or selectively serve to localize a key upstream signaling molecule, such as dPix, thus localizing the downstream dPak-Dlg-Dock cascade necessary for GluR aggregation. Similarly, the synaptic matrix may directly inhibit GluR lateral membrane diffusion, effectively ensnaring GluRs at post-synaptic sites. In the absence of the matrix, GluR dispersal could prevent accumulation of functional puncta. Alternatively, MTG may be more directly involved in forming a synaptic cleft signaling environment that allows signal molecules and/or receptors to be properly presented or anchored. It is also possible that MTG may function directly as an inductive signal by binding to an unidentified receptor. For example, synaptic localization of Dock requires its interaction with an unidentified SH2 domain-containing RTK, but the identity of this RTK or its effectors is unknown (Rohrbough, 2007).

MTG is required for the post-synaptic localization/activation of the dPix-dPak-Dock-Dlg pathways. dPix (Rho-type guanine exchange factor) binds to and is required to localize dPak; dpix and dpak mutants equally reduce synaptic Dlg level, and essentially eliminate formation of the post-synaptic SSR domain where these proteins reside. A dPak-Dock interaction is required to regulate synaptic levels of A-class GluRs. The GluR phenotypes of dpix, dpak, and dock mutants are all similar, reducing A-class GluRs by ~50%. These mutants are nevertheless viable through larval development in the near absence of dPix, dPak, or Dock, and NMJ synaptic transmission strength in basal evoked recordings is normal, due in large part to compensatory mechanisms leading to increased transmitter release. In studies in mature larvae, dpix, dpak, and dock mutations cause decreased expression level of post-synaptic proteins, without the dramatic mislocalization phenotypes characteristic of mtg functional null mutants at the embryonic NMJ. However, this study shows that strong MTG knockdown by RNAi in embryos and larvae causes a loss of post-synaptic protein levels that appears comparable with the phenotype in dpix and dpak mutant larvae. These findings suggest that the greater severity of the mtg1 mutant phenotype is likely due to a more severe loss of MTG function in the null condition, rather than simply due to a developmental disruption during the embryonic synaptogenesis period. Alternatively, since loss of MTG affects synaptic localization of both dPak and Dock individually, the full loss-of-function mtg phenotype severity may result from an additive block of several branches of the intertwined post-synaptic differentiation pathways (Rohrbough, 2007).

It is hypothesized that presynaptically secreted MTG establishes the synaptic cleft matrix signaling environment required for transsynaptic ligand-receptor pathways inducing post-synaptic differentiation. Several questions must next be addressed to test this hypothesis. One goal is to thoroughly test the GlcNAc binding specificity of MTG using purified protein and competitive binding assays. The future task will be to identify GlcNAc-containing GAG- or PG-binding target(s) of MTG resident in the synaptic cleft. Synaptic labeling with carbohydrate-specific lectins and matrix-specific antibodies, and genetically/pharmacologically altering synaptic protein glycosylation, will identify predicted glycosylated matrix components and potential targets. Site-directed mutational analysis of the key disulfide-forming cysteines in the GlcNAc-binding domain will allow testing this domain's role in MTG function. Another major goal will be to test other known transsynaptic signal/receptor pathways include Wg/Frz, Gbb/Wit, Syndecan/dLAR, DG/Dystrophin and integrin ligands (Hig, laminin)/integrins. Future studies will test whether these pathway components are lost/mislocalized in mtg mutants, as predicted by the model. Genetic interactions should exist between established signaling pathways and MTG, which can be tested in double mutant combinations to determine whether mtg phenotypes result from the additive disruption of multiple signaling pathways, as predicted. Finally, MTG may itself be an anterograde transsynaptic signaling molecule acting through a post-synaptic receptor. The potential receptor identities and assay possible signal-receptor interaction mechanisms during synaptogenesis will be the subject of future studies (Rohrbough, 2007).

Activity-dependent site-specific changes of glutamate receptor composition in vivo

The subunit composition of postsynaptic non-NMDA-type glutamate receptors (GluRs) determines the function and trafficking of the receptor. Changes in GluR composition have been implicated in the homeostasis of neuronal excitability and synaptic plasticity underlying learning. This study imaged GluRs in vivo during the formation of new postsynaptic densities (PSDs) at Drosophila neuromuscular junctions coexpressing GluRIIA and GluRIIB subunits. GluR composition was independently regulated at directly neighboring PSDs on a submicron scale. Immature PSDs typically had large amounts of GluRIIA and small amounts of GluRIIB. During subsequent PSD maturation, however, the GluRIIA/GluRIIB composition changed and became more balanced. Reducing presynaptic glutamate release increased GluRIIA, but decreased GluRIIB incorporation. Moreover, the maturation of GluR composition correlated in a site-specific manner with the level of Bruchpilot, an active zone protein that is essential for mature glutamate release. Thus, this study shows that an activity-dependent, site-specific control of GluR composition can contribute to match pre- and postsynaptic assembly (Schmid, 2008).

Fluorescence recovery after photobleaching (FRAP) experiments indicated that there is a principal difference between GluRIIA and GluRIIB incorporation at individual PSDs, showed that a genetically evoked suppression of presynaptic glutamate release results in an increase in GluRIIA and a decrease in GluRIIB incorporation, and suggested that the amount of the presynaptic active zone protein BRP correlates with a shift toward GluRIIB incorporation. Thus, when PSDs begin to mediate mature current amplitudes, they shift their composition toward the rapidly desensitizing GluRIIB complex. Put differently, a synaptic site will stop GluRIIA incorporation only if both relevant parameters, presynaptic release of glutamate (sufficient amount of BRP) and sufficient levels of efficiently conducting glutamate receptors (GluRIIA), coincide (Schmid, 2008).

Newly formed PSDs possess a rather balanced composition and then become rich in GluRIIA, before finally balancing the GluR composition again. Potentially, a positive feedback initially promotes GluRIIA incorporation, whereas a negative feedback dominates with further maturation. It has been suggested that postsynaptic Ca2+ influx through GluRIIA controls retrograde signaling. Similarly, GluR composition could be controlled during PSD maturation by such a signal through GluRIIA. Metabotropic glutamate receptors or nonvesicular glutamate release may also be involved. Moreover, the role of Ca2+-activated kinase/phosphatase signaling in GluR composition dynamics should be studied. Notably, postsynaptic stimulation of aPKC activity has been shown to reduce the PSD levels of GluRIIA. It is tempting to speculate that differences in presynaptic glutamate release might steer postsynaptic GluR composition through spatio-temporal modulation of Ca2+ signals (Schmid, 2008).

In vivo mobility of synaptic acetylcholine receptors fluorescently labeled with bungarotoxin has been elegantly studied at the mouse NMJ. There, activity-evoked changes in Ca2+ influx through these receptors seem to control receptor lifetime in the postsynaptic compartment. Mechanistic similarities between these two systems are now open for analysis (Schmid, 2008).

The cytoplasmic C termini of mammalian GluRs are considered to be important in the control of activity-dependent GluR targeting. Notably, this study found that subtype-specific differences were not fully switched after genetically switching the CTDs and subtype-specific behaviors were not fully eliminated after deprivation of presynaptic glutamate release. Thus, differences in the affinities of the GluRIIA and GluRIIB ectodomains and differences in their ionic transmission and signals conveyed by the CTDs seem to affect PSD assembly synergistically. In fact, it was recently shown that PSD assembly is greatly disrupted in the absence of all GluRs and is severely compromised without GluRIIA. Thus, potential protein interactions, which specifically allow GluRIIA to incorporate into immature PSDs, might be particularly important for efficient PSD assembly (Schmid, 2008).

Positive feedback mechanisms clearly operate during long-term potentiation (LTP). Indeed, it was demonstrated that LTP increased the amount of glutamate receptors containing the GluR1 subunit at glutamatergic synapses in the vertebrate CNS. On the single-molecule level, local activity restricts the mobility of the GluR1 subunit on a submicron scale. In addition, mechanisms must exist that stabilize the total synaptic strength of a neuron during exposure to synaptic plasticity. Recently, it was shown that adaptation to prolonged AMPAR blockade in cultivated hippocampal neurons increases the amount of GluR1 per PSD and elevates the presynaptic efficacy, which is consistent with the current findings. According to the current model, a blockade of receptors would lead to an omission of the negative feedback loop on GluRIIA incorporation and BRP accumulation. Notably, large synapses would be affected most, but diffusion of receptors in the spine and the spine neck would also have to be considered. It will be interesting to investigate how positive and negative feedback during development are related to positive feedback during LTP and negative feedback during homeostatic control, and how these mechanisms interact in terms of metaplasticity (Schmid, 2008).

Recently, pharmacological blockade of glutamate receptors at the Drosophila NMJ was shown to trigger a fast negative-feedback mechanism involving presynaptic CaV2.1 calcium channels. This negative feedback, however, was independent of presynaptic action potentials (Frank, 2006). This study used presynaptic TNT expression to suppress synaptic activity, resulting in a complete block of evoked glutamate release and a reduced frequency of mEPSCs25. However, although mEPSC frequency is unaltered in BRP mutants, clear changes occurred in GluR composition. Thus, evoked glutamate release, rather than mEJCs, might be mainly responsible for determining GluR composition at larval NMJs over extended periods. Therefore, these observations may be reconciled by differences in the duration of activity suppression and/or the site of intervention. Similarly, in cultured hippocampal neurons, AMPA receptor blockade or action potential suppression lead to different synaptic responses to inactivity (Schmid, 2008).

Taken together, this analysis indicates that the incorporation rate of GluRIIA is high at immature PSDs, but increasing levels of GluRIIA establish a negative feedback, reducing PSD incorporation of this complex and further growth of the presynaptic site. In addition, it is hypothesized that high levels of glutamate release lower the consumption of GluRIIA at maturing PSDs, and thereby support the growth of additional nascent synapses. In this manner, synapse-specific control of GluR composition can also execute cell-wide control over activity-dependent synapse formation in a long-term fashion. Indeed, experience-dependent increases in the number of NMJ synapses depend on GluRIIA, but not GluRIIB (Schmid, 2008).

Maturation of active zone assembly by Drosophila Bruchpilot

Synaptic vesicles fuse at active zone (AZ) membranes where Ca2+ channels are clustered and that are typically decorated by electron-dense projections. Recently, mutants of the Drosophila ERC/CAST family protein Bruchpilot (BRP) were shown to lack dense projections (T-bars) and to suffer from Ca2+ channel-clustering defects. This study used high resolution light microscopy, electron microscopy, and intravital imaging to analyze the function of BRP in AZ assembly. Consistent with truncated BRP variants forming shortened T-bars, BRP was identified as a direct T-bar component at the AZ center with its N terminus closer to the AZ membrane than its C terminus. In contrast, Drosophila Liprin-α, another AZ-organizing protein, precedes BRP during the assembly of newly forming AZs by several hours and surrounds the AZ center in few discrete punctae. BRP seems responsible for effectively clustering Ca2+ channels beneath the T-bar density late in a protracted AZ formation process, potentially through a direct molecular interaction with intracellular Ca2+ channel domains (Fouquet, 2009).

This study addressed whether BRP signals T-bar formation (without being a direct component of the T-bar) or whether the protein itself is an essential building block of this electron-dense structure. Evidence is provided that BRP is a direct T-bar component. Immuno-EM identifies the N terminus of BRP throughout the whole cross section of the T-bar, and genetic approaches show that a truncated BRP, lacking the C-terminal 30% of the protein's sequence, forms truncated T-bars. Immuno-EM and light microscopy consistently demonstrate that N- and C-terminal epitopes of BRP are segregated along an axis vertical to the AZ membrane and suggest that BRP is an elongated protein, which directly shapes the T-bar structure (Fouquet, 2009).

In brp5.45 (predicted as aa 1-866), T-bars were not detected, whereas brp1.3 (aa 1-1,389) formed T-bar-like structures, although fewer and smaller than normal. Moreover, the BRPD1-3GFP construct (1-1,226) did not rescue T-bar assembly. Thus, domains between aa 1,226 and 1,390 of BRP may also be important for the formation of T-bars. Clearly, however, the assembly scheme for T-bars is expected to be controlled at several levels (e.g., by phosphorylation) and might involve further protein components. Nonetheless, it is highly likely that the C-terminal half of BRP plays a crucial role (Fouquet, 2009).

Since BRP represents an essential component of the electron-dense T-bar subcompartment at the AZ center, it might link Ca2+ channel-dependent release sites to the synaptic vesicle cycle. Interestingly, light and electron microscopic analysis has located CAST at mammalian synapses both with and without ribbons. Overall, this study is one of the first to genetically identify a component of an electron-dense synaptic specialization and thus paves the way for further genetic analyses of this subcellular structure (Fouquet, 2009).

The N terminus of BRP is found significantly closer to the AZ membrane than the C terminus, where it covers a confined area very similar to the area defined by the CacGFP epitope. Electron tomography of frog NMJs suggested that the cytoplasmic domains of Ca2+ channels, reminiscent of pegs, are concentrated directly beneath a component of an electron-dense AZ matrix resembling ribs. In addition, freeze-fracture EM identified membrane-associated particles at flesh fly AZs, which, as judged by their dimensions, might well be Ca2+ channels. Peg-like structures were observed beneath the T-bar pedestal. Similar to fly T-bars, the frog AZ matrix extends up to 75 nm into the presynaptic cytoplasm. Based on the amount of cytoplasmic Ca2+ channel protein it has been concluded that Ca2+ channels are likely to extend into parts of the ribs. Thus, physical interactions between cytoplasmic domains of Ca2+ channels and components of ribs/T-bars might well control the formation of Ca2+ channel clusters at the AZ membrane. However, a short N-terminal fragment of BRP (aa 1-320) expressed in the brp-null background was unable to localize to AZs efficiently and consistently failed to restore Cac clustering (unpublished data) (Fouquet, 2009).

The mean Ca2+ channel density at AZs is reduced in brp-null alleles. In vitro assays indicate that the N-terminal 20% of BRP can physically interact with the intracellular C terminus of Cacaphony (Cac). Notably, it was found that the GFP epitope at the very C terminus of CacGFP was closer to the AZ membrane than the N-terminal epitope of BRP. It is conceivable that parts of the Cac C terminus extend into the pedestal region of the T-bar cytomatrix to locally interact with the BRP N terminus. This interaction might play a role in clustering Ca2+ channels beneath the T-bar pedestal (Fouquet, 2009).

Clearly, additional work will be needed to identify the contributions of discrete protein interactions in the potentially complex AZ protein interaction scheme. This study should pave the way for a genetic analysis of spatial relationships and structural linkages within the AZ organization. Moreover, the current findings should integrate in the framework of mechanisms for Ca2+ channel trafficking, clustering, and functional modulation (Fouquet, 2009).

The imaging assays allowed a temporally resolved analysis of AZ assembly in vivo. BRP is a late player in AZ assembly, arriving hours after DLiprin-α and also clearly after the postsynaptic accumulation of DGluRIIA. Accumulation of Cac was late as well, although it slightly preceded the arrival of BRP, and impaired Cac clustering at AZs lacking BRP became apparent only from a certain synapse size onwards, form at sites distant from preexisting ones and grow to reach a mature, fixed size. Thus, the late, BRP-dependent formation of the T-bar seems to be required for maintaining high Ca2+ channel levels at maturing AZs but not for initializing Ca2+ channel clustering at newly forming sites. As the dominant fraction of neuromuscular AZs is mature at a given time point, the overall impression is that of a general clustering defect in brp mutants. In reverse, it will be of interest to further differentiate the molecular mechanisms governing early Ca2+ channel clustering. Pre- to postsynaptic communication via neurexin-neuroligin interactions might well contribute to this process. A further candidate involved in early Ca2+ channel clustering is the Fuseless protein, which was recently shown to be crucial for proper Cac localization at AZs (Fouquet, 2009).

In summary, during the developmental formation of Drosophila NMJ synapses, the emergence of a presynaptic dense body, which is involved in accumulating Ca2+ channels, appears to be a central aspect of synapse maturation. This is likely to confer mature release probability to individual AZs and contribute to matching pre- and postsynaptic assembly by regulating glutamate receptor composition. Whether similar mechanisms operate during synapse formation and maturation in mammals remains an open question (Fouquet, 2009).

This study concentrated on developmental synapse formation and maturation. The question arises whether similar mechanisms to those relevant for AZ maturation might control activity-dependent plasticity as well and whether maturation-dependent changes might be reversible at the level of individual synapses. Notably, experience-dependent, bidirectional changes in the size and number of T-bars (occurring within minutes) were implied at Drosophila photoreceptor synapses by ultrastructural means. Moreover, at the crayfish NMJ, multiple complex AZs with double-dense body architecture were produced after stimulation and were associated with higher release probability. In fact, a recent study has correlated the ribbon size of inner hair cell synapses with Ca2+ microdomain amplitudes. Thus, a detailed understanding of the AZ architecture might permit a prediction of functional properties of individual AZs (Fouquet, 2009).

PP2A and GSK-3β act antagonistically to regulate active zone development
The synapse is composed of an active zone apposed to a postsynaptic cluster of neurotransmitter receptors. Each Drosophila neuromuscular junction comprises hundreds of such individual release sites apposed to clusters of glutamate receptors. This study shows that protein phosphatase 2A (PP2A) is required for the development of structurally normal active zones opposite glutamate receptors. When PP2A is inhibited presynaptically, many glutamate receptor clusters are unapposed to Bruchpilot (Brp), an active zone protein required for normal transmitter release. These unapposed receptors are not due to presynaptic retraction of synaptic boutons, since other presynaptic components are still apposed to the entire postsynaptic specialization. Instead, these data suggest that Brp localization is regulated at the level of individual release sites. Live imaging of glutamate receptors demonstrates that this disruption to active zone development is accompanied by abnormal postsynaptic development, with decreased formation of glutamate receptor clusters. Remarkably, inhibition of the serine-threonine kinase GSK-3beta completely suppresses the active zone defect, as well as other synaptic morphology phenotypes associated with inhibition of PP2A. These data suggest that PP2A and GSK-3beta function antagonistically to control active zone development, providing a potential mechanism for regulating synaptic efficacy at a single release site (Viquez, 2009).

This study demonstrates that the serine-threonine phosphatase PP2A is required in the presynaptic neuron for normal development and maturation of presynaptic release sites. This action of PP2A is opposed by the serine-threonine kinase GSK-3β, suggesting that this phosphatase/kinase pair co-regulate the phosphorylation state and activity of proteins that are required for proper synaptic development (Viquez, 2009).

At the Drosophila NMJ, the synaptic terminal of a motoneuron is a branched chain of synaptic boutons whose gross structure is strongly influenced by the cytoskeleton. Within each synaptic terminal, there are hundreds of individual synapses, neurotransmitter release sites with an active zone directly apposed to a cluster of postsynaptic glutamate receptors. Most studies in Drosophila have focused on genes controlling synaptic terminal development. However with the recent development of antibodies to the active zone component Bruchpilot and the essential glutamate receptor DGluRIII, a genetic analysis of active zone and postsynaptic density development is now feasible. Previous studies have demonstrated that PP2A acts in the motoneuron to control synaptic terminal morphology likely via regulation of microtubules. This study demonstrate that PP2A is also essential for the proper development of the individual synaptic unit, the active zone and glutamate receptor dyad (Viquez, 2009).

Presynaptic inhibition of PP2A impairs synaptic transmission, leading to a large decrease in quantal content. While investigating potential morphological explanations for defective transmitter release, it was observed that many glutamate receptor clusters are unapposed to the active zone protein Bruchpilot. This is not due to retraction of the presynaptic terminal, since apposed and unapposed GluR clusters are intermingled throughout the terminal in a salt and pepper pattern, and presynaptic structures such as synaptic vesicles are still apposed to the entire extent of the postsynaptic specialization. Instead, there is a defect at the level of the individual synapse. These GluR clusters may be unapposed to active zones, or may be apposed to abnormal active zones lacking Bruchpilot. Two lines of evidence suggest that these GluR clusters may be unapposed to active zones. First, Bruchpilot is required for the localization of T-bars to the active zone, so if many active zones are missing Bruchpilot, then there should be a decrease in the proportion of active zones with T-bars. However when PP2A is inhibited no change was seen in the proportion of active zones with T-bars. Second, with PP2A inhibition the number of Brp puncta is down, as is the density of active zones as defined by ultrastructural analysis. This suggests that there is not a large pool of active zones without Brp. Both of these findings suggest that there are fewer active zones, and that those active zones that form do contain Brp. If this is so, then why are GluR clusters present that are unapposed to active zones? This could be due either to a problem with synapse formation/maturation or maintenance. While it is not known which is the case, the model that there is a defect in the formation or maturation is preferred for the following reasons. First, unapposed receptors are more prevalent in the distal regions of the NMJ where new synapses tend to be added. Second, the unapposed receptors form quite small clusters, while newly forming GluR clusters in wild type are also quite small. Finally, live imaging reveals that fewer GluR clusters form late in larval development, demonstrating a defect in synapse formation (Viquez, 2009).

A model is proposed in which PP2A activity is required for the maturation phase of synapse development. In this view, at a wild type synapse a signal would initiate synapse formation, leading to postsynaptic clustering of glutamate receptors as well as transsynaptic interactions that form the tightly apposed pre- and postsynaptic membranes as seen in electron micrographs. Later, additional active zone components such as Brp would be recruited to the active zone, a process known to occur after GluR clustering. With PP2A inhibition, this unknown signal would still initiate synapse formation and induce GluR clusters. However, at some fraction of nascent synapses the maturation process would fail. The GluR clusters could be trapped in their small, immature state or lost, while the transsynaptic process leading to the tight apposition of pre- and postsynaptic membranes would also fail and Brp would not be recruited. The alternate model that synaptic maintenance is disrupted, and that unapposed GluR clusters are the remains of synapses at which the presynaptic terminal has been lost, cannot, however, be ruled out. Regardless of the precise mechanism, these data demonstrate that PP2A is required to ensure the correct apposition of structurally normal active zones and glutamate receptors at the synapse (Viquez, 2009).

PP2A is one of the major serine/threonine phosphatases in the cell, so inhibiting its function likely leads to hyperphosphorylation of many proteins. Hence, phenotypes could be due to the pleiotropic effects of misregulating numerous pathways. The data, however, argue for a good deal of specificity in the function of PP2A for the synaptic morphology phenotypes assayed. Inhibiting PP2A in the neuron leads to misapposed GluR clusters, a disrupted synaptic cytoskeleton, and an altered bouton morphology. Each of these phenotypes is suppressed when GSK-3β is inhibited. This suggests that these synaptic phenotypes are due to the misregulation of a pathway that is antagonistically regulated by PP2A and GSK-3β. Opposite phenotypes are not seen, however, when PP2A is overexpressed, suggesting that hyperphosphorylation affects this pathway more than hypophosphorylation. While genetic studies cannot prove that this phosphatase/kinase pair act directly on the same substrate, the simplest interpretation of the data is that PP2A and GSK-3β co-regulate the phosphorylation state and activity of a protein or proteins that are required for the proper development of active zones and the synaptic cytoskeleton. While these PP2A phenotypes are all suppressed by inhibition of GSK-3β, there is no suppression of the accumulation of synaptic material in the axon, a phenotype consistent with defects in axonal transport. Decreased transport of active zone material such as Brp is a plausible mechanism for the active zone defects in this mutant. However, the failure of GSK-3β inhibition to suppress the axonal transport phenotype demonstrates that the active zone maturation and axon transport phenotypes are genetically separable. Hence, the accumulation of Brp in the axon cannot be responsible for the defects in synaptic maturation (Viquez, 2009).

The identity of the pathway regulated by PP2A and GSK-3β is not known. One candidate substrate is APC2, which binds to and stabilizes the plus end of microtubules and which is a characterized substrate of both PP2A and GSK-3β. In hippocampal cells phosphorylation of APC by GSK-3β inhibits APC function and so disrupts microtubule stability and axon outgrowth. It was shown that loss of APC2 dominantly enhances the PP2A phenotype, which is consistent with the model from hippocampal cells that phosphorylating APC decreases its function. However, if APC2 were the key substrate, then it would be predicted that homozygous APC2 mutants, where all APC2 function is lost, should replicate the PP2A phenotype. However a synaptic apposition phenotype is not seen in recessive mutants for APC2 or in APC1/APC2 double mutants. Instead, the enhancement of the PP2A phenotype by the loss of APC2 suggests that APC2 promotes PP2A function, possibly in its role as a scaffolding molecule. Wnt signaling is candidate pathway for mediating these synaptic phenotypes because wnt signaling is required for normal Drosophila NMJ development and because GSK-3β and PP2A regulate the phosphorylation state of β-catenin in canonical wnt signaling. Inhibition of PP2A would be predicted to lead to hyperphosphorylation and destruction of β-catenin, thereby blocking wnt signaling. However it is unlikely that the PP2A synaptic phenotype is due to loss of canonical wnt signaling. First, this study found that expression of a constitutively active β-catenin does not suppress the PP2A synaptic phenotype but instead has a slight tendency to enhance the cytoskeletal defect. Second, APC functions as part of the destruction complex that leads to degradation of β-Catenin and block of wnt signaling, however APC mutants enhance rather than suppress the PP2A phenotype. These results are inconsistent with the model that the phenotype is due to decreased canonical wnt signaling through β-Catenin. However, the data are consistent with a role for β-catenin-independent wnt signaling. A third candidate substrate is Futsch, since it can be phosphorylated by GSK-3β and the effect of reduction of PP2A activity on Futsch structure is suppressed by reduction in GSK-3β levels. Continued genetic analysis may lead to the identification of the relevant substrate(s) that are antagonistically regulated by PP2A and GSK-3β to control synaptic development (Viquez, 2009).

There are interesting parallels between the function of PP2A and GSK-3β in the developing Drosophila neuromuscular system and in the pathogenesis of neurodegenerative diseases such as Alzheimer's. In Drosophila, PP2A antagonizes GSK-3β function to stabilize the synaptic cytoskeleton and promote synapse formation. In models of Alzheimer's Disease, PP2A and GSK-3β also act antagonistically, for example in regulating the phosphorylation state of tau. In addition, disruptions to the axonal cytoskeleton and synapse loss are early events in Alzheimer's pathogenesis. Characterizing the function of PP2A/GSK-3β in regulating cytoskeletal and synaptic integrity during development may provide insights into their role in regulating cytoskeletal and synaptic integrity during disease (Viquez, 2009).

Structure-function analysis of endogenous lectin Mind-the-gap in synaptogenesis

Mind-the-Gap (MTG) is required for neuronal induction of Drosophila neuromuscular junction (NMJ) postsynaptic domains, including glutamate receptor (GluR) localization. It has previously been hypothesized that MTG is secreted from the presynaptic terminal to reside in the synaptic cleft, where it binds glycans to organize the heavily glycosylated, extracellular synaptomatrix required for transsynaptic signaling between neuron and muscle. This study tests this hypothesis with MTG structure-function analyses of predicted signal peptide (SP) and carbohydrate-binding domain (CBD), by introducing deletion and point-mutant transgenic constructs into mtg null mutants. The SP is shown to be required for MTG secretion and localization to synapses in vivo. It is further shown that the CBD is required to restrict MTG diffusion in the extracellular synaptomatrix and for postembryonic viability. However, CBD mutation results in elevation of postsynaptic GluR localization during synaptogenesis, not the mtg null mutant phenotype of reduced GluRs as predicted by the hypothesis, suggesting that proper synaptic localization of MTG limits GluR recruitment. In further testing CBD requirements, it was shown that MTG binds N-acetylglucosamine (GlcNAc) in a Ca(2+)-dependent manner, and thereby binds HRP-epitope glycans, but that these carbohydrate interactions do not require the CBD. It is concluded that the MTG lectin has both positive and negative binding interactions with glycans in the extracellular synaptic domain, which both facilitate and limit GluR localization during NMJ embryonic synaptogenesis (Rushton, 2012).

The core hypothesis for the role of MTG in synaptogenesis has proven to be wrong. It was postulated that MTG is secreted from the presynaptic neuron to bind extracellular glycans via a well-conserved CBD, to build an extracellular synaptomatrix required for the anterograde transsynaptic signaling inducing postsynaptic GluR domains (Rohrbough, 2007; Rushton, 2009; Rohrbough, 2010; Rushton, 2012 and references therein).

In this study a systematic test of this hypothesis was undertaken with structure-function analyses of SP and CBD requirements in vivo. It was confirmed that the SP is required for secretion, and MTG was discovered to be strongly down regulated when introduction into the secretory pathway is prevented. However, contrary to our hypothesis, the CBD is not required for binding glycans (as shown in vitro on GlcNAc-conjugated beads), glycoproteins (as shown by IP of HRP-epitope proteins), or binding to the extracellular membrane and/or pericellular matrix (as shown by in vivo imaging). The CBD does play an important regulatory role in the localization/anchoring of MTG in the synaptic ECM, since δCBD MTG migrates much further from the secretory synaptic boutons. This function may well be dependent on glycan binding, as indicated by the alteration in glycoprotein binding in δCBD MTG IP experiments. Moreover, contrary to the hypothesis, the CBD is not required for the MTG role in functional GluR postsynaptic localization during embryonic synaptogenesis. Rather, the CBD plays an unexpected role in limiting GluR recruitment to the developing NMJ synapse, and is essential for postembryonic viability (Rushton, 2012).

Although the CBD requirement is clearly fundamentally different from the previous hypothesis, this structure-function analysis reveals intriguing CBD functions. The MTG CBD has a strong homology to CBD14 (Rohrbough, 2007), yet it is not required for binding to GlcNAc, GlcNAc polymer, or other glycans, showing that another cryptic CBD must be present that is sufficient to mediate this carbohydrate binding. Candidate carbohydrate-binding regions include the glutamine-rich region near the N-terminus, and the coiled-coil domain near the C terminus. The glutamine-rich region is a particularly intriguing candidate. This domain bears a striking resemblance to the prion-like domains of Aplysia CPEB and Drosophila CPEB (Orb2), as well as Drosophila fragile X mental retardation protein (FMRP), all of which are necessary for synaptic mechanisms underlying learning and memory. Interestingly, the prion-like domain of Aplysia CPEB causes mouse CPEB (which lacks this domain) to aggregate into puncta, very similar to MTG puncta. Similarly, the FMRP prion-like domain drives aggregation and puncta formation. It would therefore be of great interest to extend the structure-function analysis to the MTG glutamine-rich region, specifically to investigate whether this domain is involved in carbohydrate-binding and/or extracellular puncta formation (Rushton, 2012).

Endogenous animal lectins are extremely diverse, and are organized into many families, including C-type, R-type, siglecs, galectins, and chitinase-like lectins, among others. Each lectin family has characteristic carbohydrate- binding-fold consensus sequences, which differ greatly from one family to another. Within several families, examples exist of lectins that have the characteristic disulfide-bonding fold domain, but do not bind predicted carbohydrate substrates. For example, the C-lectin family includes several members with canonical C-type lectin domains (CTLD) that do not bind carbohydrates, nor is calcium always required for their carbohydrate binding. Specifically, collectin- like tetranectin binds calcium and the protein kringle 4 via its CTLD, yet binds the carbohydrate heparin in a separate, non-CTLD domain. Likewise, while lecticans bind tenascin- R in a Ca2+-dependent manner, this binding does not require tenascin-linked carbohydrates, but rather appears to be a protein-protein interaction. Thus, the relationship between lectins and their carbohydrate-binding partners is not simple or easily defined, and it is obvious that CBDs acquire new functions and new binding properties during evolution. This certainly appears to be the case for the MTG CBD, which is not required for carbohydrate to binding, yet regulates MTG mobility/anchoring at the NMJ, and regulates the recruitment of postsynaptic GluRs (Rushton, 2012).

It is particularly intriguing that MTG carbohydrate binding is Ca2+-dependent, and it is tempting to place MTG in the C-type lectin family based on this characteristic. However, MTG lacks a canonical CTLD, and does not appear to resemble any of the very diverse but well-characterized C-type lectin families. Nor does it resemble the pentraxin domain, another Ca2+-dependent lectin domain. Rather, MTG has the CBD14 domain of the peritrophin lectin family. Invertebrate lectins of this family have not been reported to require calcium for carbohydrate binding, although the related Clostridium endo beta-1.3-glucanase Lic16A binds GlcNAc polymer far more strongly in the presence of calcium and the related mammalian FIBCD1 also has a Ca2+-dependent binding requirement. It is speculated that the calcium requirement for MTG binding to synaptic carbohydrates may be physiologically important, since extracellular calcium concentration in the synaptic cleft and surrounding synaptomatrix is modulated by synaptic activity, involving presynaptic and postsynaptic calcium influx and calcium exchange (Rushton, 2012).

Although the MTG CBD has homology to chitin-binding domains, there is no indication that MTG binds chitin in vivo. MTG does not localize in chitin-rich tissues and when exogenously expressed in these tissues, it is not retained to any detectable degree. When MTG:GFP is expressed ubiquitously, the protein does not localize in the trachea, nor in other chitinous structures such as the external cuticle. In contrast, Drosophila Serpentine and Vermiform chitin deacetylases with canonical chitin-binding domains strongly colocalize with chitin in the trachea, and a Serp N-terminus construct with the SP and CBD fused to GFP likewise strongly co-localizes with chitin in the trachea. Indeed, it is very striking that MTG is strongly downregulated outside nervous tissue when driven ubiquitously, and does not detectably accumulate in epidermis or muscle, except immediately surrounding the NMJ terminal. Thus, there must be a mechanism to specifically retain and preserve MTG in neurons, particularly in synaptic domains. Likewise, within the embryonic CNS, expression of MTG is virtually identical whether it is driven by a neural-specific or a ubiquitous GAL4 driver, indicating that MTG:GFP is accumulated at synapses in a very specific manner and down-regulated elsewhere. The one notable exception is the salivary gland, a tissue specialized for secretion, but it appears even the salivary gland cannot maintain MTG without the SP required for secretion into the extracellular lumenal domain. These data suggest that MTG likely binds GlcNAc inside the chains of N-glycans, O-glycans, and/or GAGs within the pericellular matrix and particularly within the specialized extracellular synaptomatrix (Rushton, 2012).

How might the MTG CBD affect synaptic functional development? It is well established that ECM glycoproteins and proteoglycans are essential for the organization of synaptic components. Lectins that bind selectively to the carbohydrate component of these molecules can regulate, modulate, stabilize, or sequester their activities. At the vertebrate NMJ, the Agrin lectin is required for AChR cluster maintenance in a fashion similar to the MTG requirement for GluR localization at the Drosophila NMJ. In the vertebrate CNS, lecticans brevican and neurocan bind to tenascin and hyaluronic acid to stabilize the extracellular synaptomatrix lattice, and have been implicated in affecting synaptic development and plasticity. Digestion of this lattice by hyaluronidase causes increased lateral diffusion of AMPA GluRs, suggesting the matrix acts as an important diffusion/ mobility barrier. Similarly, the NP lectin family has been implicated in AMPA GluR trafficking: NP-1 and NP-2 (also known as Narp) form a complex with the NP receptor to colocalize with and trigger clustering of AMPA GluRs at postsynaptic sites. Specifically, this lectin mechanism mediates postsynaptic recruitment of the AMPA GluRs with GluR1 and GluR4 subunits. The results presented in this study show that the CBD of MTG is similarly important for regulating GluR trafficking and postsynaptic maintenance at the Drosophila developing embryonic NMJ (Rushton, 2012).

In conclusion, this study has shown that MTG is a secreted, Ca2+-dependent carbohydrate-binding protein resident in the extracellular matrix surrounding synapses. The predicted SP is required for the secretion of MTG, but the CBD is not demonstrably required for glycan interaction, indicating that a cryptic CBD must also be present within MTG. The CBD appears to regulate binding affinity of MTG to the ECM, and is clearly required to anchor properly MTG close to the synaptic interface. In the absence of the CBD, excess GluRs are recruited to the embryonic NMJ postsynaptic domain. This is the opposite consequence to the loss of postsynaptic GluRs occurring with complete removal of MTG. It is concluded that the MTG lectin has both positive and negative roles regulating GluR recruitment during synaptogenesis (Rushton, 2012).

N-glycosylation requirements in neuromuscular synaptogenesis

Neural development requires N-glycosylation regulation of intercellular signaling, but the requirements in synaptogenesis have not been well tested. All complex and hybrid N-glycosylation requires MGAT1 (UDP-GlcNAc:alpha-3-D-mannoside-beta1,2-N-acetylglucosaminyl-transferase I) function, and Mgat1 nulls are the most compromised N-glycosylation condition that survive long enough to permit synaptogenesis studies. At the Drosophila neuromuscular junction (NMJ), Mgat1 mutants display selective loss of lectin-defined carbohydrates in the extracellular synaptomatrix, and an accompanying accumulation of the secreted endogenous Mind the gap (MTG) lectin, a key synaptogenesis regulator. Null Mgat1 mutants exhibit strongly overelaborated synaptic structural development, consistent with inhibitory roles for complex/hybrid N-glycans in morphological synaptogenesis, and strengthened functional synapse differentiation, consistent with synaptogenic MTG functions. Synapse molecular composition is surprisingly selectively altered, with decreases in presynaptic active zone Bruchpilot (BRP) and postsynaptic Glutamate receptor subtype B (GLURIIB), but no detectable change in a wide range of other synaptic components. Synaptogenesis is driven by bidirectional trans-synaptic signals that traverse the glycan-rich synaptomatrix, and Mgat1 mutation disrupts both anterograde and retrograde signals, consistent with MTG regulation of trans-synaptic signaling. Downstream of intercellular signaling, pre- and postsynaptic scaffolds are recruited to drive synaptogenesis, and Mgat1 mutants exhibit loss of both classic Discs large 1 (DLG1) and newly defined Lethal (2) giant larvae [L(2)gl] scaffolds. It is concluded that MGAT1-dependent N-glycosylation shapes the synaptomatrix carbohydrate environment and endogenous lectin localization within this domain, to modulate retention of trans-synaptic signaling ligands driving synaptic scaffold recruitment during synaptogenesis (Parkinson, 2013).

This study began with the hypothesis that disruption of synaptomatrix N-glycosylation would alter trans-synaptic signaling underlying NMJ synaptogenesis (Dani, 2012). MGAT1 loss transforms the synaptomatrix glycan environment. Complete absence of the HRP epitope, α1-3-fucosylated N-glycans, is expected to require MGAT1 activity: key HRP epitope synaptic proteins include fasciclins, Neurotactin and Neuroglian, among others. This study shows that HRP epitope modification of the key synaptogenic regulator Fasciclin 2 is not required for stabilization or localization, suggesting a role in protein function. However, complete loss of Vicia villosa (VVA) lectin reactivity synaptomatrix labeling is surprising because the epitope is a terminal β-GalNAc. This result suggests that the N-glycan LacdiNAc is enriched at the NMJ, and that the terminal GalNAc expected on O-glycans/glycosphingolipids may be present on N-glycans in this synaptic context. Importantly, VVA labels Dystroglycan and loss of Dystroglycan glycosylation blocks extracellular ligand binding and complex formation in Drosophila, and causes muscular dystrophies in humans. This study shows that VVA-recognized Dystroglycan glycosylation is not required for protein stabilization or synaptic localization, but did not test functionality or complex formation, which probably requires MGAT1-dependent modification. Conversely, the secreted endogenous lectin MTG is highly elevated in Mgat1 null synaptomatrix, probably owing to attempted compensation for complex and hybrid N-glycan losses that serve as MTG binding sites. MTG binds GlcNAc in a calcium-dependent manner and pulls down a number of HRP-epitope proteins by immunoprecipitation (Rushton, 2012), although the specific proteins have not been identified. It will be of interest to perform immunoprecipitation on Mgat1 samples to identify changes in HRP bands. Importantly, MTG is crucial for synaptomatrix glycan patterning and functional synaptic development. MTG regulates VVA synaptomatrix labeling, suggesting a mechanistic link between the VVA and MTG changes in Mgat1 mutants. The MTG elevation observed in Mgat1 nulls provides a plausible causative mechanism for strengthened functional differentiation (Parkinson, 2013).

Consistent with recent glycosylation gene screen findings (Dani, 2012), Mgat1 nulls exhibit increased synaptic growth and structural overelaboration. Therefore, complex and hybrid N-glycans overall provide a brake on synaptic morphogenesis, although individual N-glycans may provide positive regulation. Likely players include MGAT1-dependent HRP-epitope proteins (e.g., fasciclins, Neurotactin, Neuroglian), and position-specific (PS) integrin receptors and their ligands, all of which are heavily glycosylated and have well-characterized roles regulating synaptic architecture. An alternative hypothesis is that Mgat1 phenotypes may result from the presence of high-mannose glycans on sites normally carrying complex/hybrid structures, suggesting possible gain of function rather than loss of function of specific N-glycan classes. NMJ branch and bouton number play roles in determining functional strength, although active zones and GluRs are also regulated independently. Thus, the increased functional strength could be caused by increased structure at Mgat1 null NMJs. However, muscle-targeted UAS-Mgat1 rescues otherwise Mgat1 null function, but has no effect on structural defects, demonstrating that these two roles are separable. Presynaptic Mgat1 RNAi also causes strong functional defects, showing there is additionally a presynaptic requirement in functional differentiation. Neuron-targeted Mgat1 causes lethality, indicating that MGAT1 levels must be tightly regulated, but preventing independent assessment of Mgat1 presynaptic rescue of synaptogenesis defects (Parkinson, 2013).

Presynaptic glutamate release and postsynaptic glutamate receptor responses drive synapse function. Using lipophilic dye to visualize SV cycling, this study found Mgat1 null mutants endogenously cycle less than controls, but have greater cycling capacity upon depolarizing stimulation. The endogenous cycling defect is consistent with the sluggish locomotion of Mgat1 mutants, whereas the elevated stimulation-evoked cycling is consistent with electrophysiological measures of neurotransmission. Similarly, mutation of dPOMT1, which glycosylates VVA-labeled Dystroglycan, decreases SV release probability (Wairkar, 2008), although dPOMT1 adds mannose not GalNAc. Null Mgat1 mutants display no change in SV cycle components (e.g. Synaptobrevin, Synaptotagmin, Synaptogyrin, etc.), but exhibit reduced expression of the key active zone component Bruchpilot. Other examples of presynaptic glycosylation requirements include the Drosophila Fuseless (FUSL) glycan transporter, which is critical for Cacophony (CAC) voltage-gated calcium channel recruitment to active zones, and the mammalian GalNAc transferase (GALGT2), whose overexpression causes decreased active zone assembly. Postsynaptically, Mgat1 nulls show specific loss of GLURIIB-containing receptors. Similarly, dPOMT1 mutants exhibit specific GLURIIB loss (Wairkar, 2008), although dystroglycan nulls display GLURIIA loss. Selective GLURIIB loss in Mgat1 nulls may drive increased neurotransmission owing to channel kinetics differences in GLURIIA versus GLURIIB receptors (Parkinson, 2013).

Bidirectional trans-synaptic signaling regulates NMJ structure, function and pre/postsynaptic composition. This intercellular signaling requires ligand passage through, and containment within, the heavily glycosylated synaptomatrix, which is strongly compromised in Mgat1 mutants. In testing three well-characterized signaling pathways, this study found that Wingless (Wg) accumulates, whereas both GBB and JEB are reduced in the Mgat1 null synaptomatrix. WG has two N-glycosylation sites, but these do not regulate ligand expression, suggesting WG build-up occurs owing to lost synaptomatrix N-glycosylation. Importantly, WG overexpression increases NMJ bouton formation similarly to the phenotype of Mgat1 nulls, suggesting a possible causal mechanism. GBB is predicted to be N-glycosylated at four sites, but putative glycosylation roles have not yet been tested. Importantly, GBB loss impairs presynaptic active zone development similarly to Mgat1 nulls, suggesting a separable causal mechanism. JEB is not predicted to be N-glycosylated, indicating that JEB loss is caused by lost synaptomatrix N-glycosylation. Importantly, it has been shown that loss of JEB signaling increases functional synaptic differentiation similarly to Mgat1 nulls (Rohrbough, 2013). In addition, jeb mutants exhibit strongly suppressed NMJ endogenous activity, similarly to the reduced endogenous SV cycling in Mgat1 nulls. Moreover, the MTG lectin negatively regulates JEB accumulation in NMJ synaptomatrix, consistent with elevated MTG causing JEB downregulation in Mgat1 nulls (Parkinson, 2013).

Trans-synaptic signaling drives recruitment of scaffolds that, in turn, recruit pre- and postsynaptic molecular components. Specifically, DLG1 and L(2)GL scaffolds regulate the distribution and density of both active zone components (e.g. BRP) and postsynaptic GluRs, and both of these scaffolds are reduced at Mgat1 null NMJs. Importantly, dlg1 mutants display selective loss of GLURIIB, with GLURIIA unchanged, similar to Mgat1 nulls, suggesting a causal mechanism. Moreover, l(2)gl mutants display both a selective GLURIIB impairment as well as reduction of BRP aggregation in active zones, similarly to Mgat1 nulls, suggesting a separable involvement for this synaptic scaffold. DLG1 and L(2)GL are known to interact in other developmental contexts, indicating a likely interaction at the developing synapse. Although synaptic ultrastructure has not been examined in l(2)gl mutants, dlg1 mutants exhibit impaired NMJ development, including a deformed SSR. These synaptogenesis requirements predict similar ultrastructural defects in Mgat1 mutants, albeit presumably due to the combined loss of both DLG1 and L(2)GL scaffolds. Future work will focus on electron microscopy analyses to probe N-glycosylation mechanisms of synaptic development (Parkinson, 2013).

The 4.1 protein coracle mediates subunit-selective anchoring of Drosophila glutamate receptors to the postsynaptic actin cytoskeleton

Glutamatergic Drosophila neuromuscular junctions contain two spatially, biophysically, and pharmacologically distinct subtypes of postsynaptic glutamate receptor (GluR). These receptor subtypes appear to be molecularly identical except that A receptors contain the subunit GluRIIA (but not GluRIIB), and B receptors contain the subunit GluRIIB (but not GluRIIA). A- and B-type receptors are coexpressed in the same cells, in which they form homotypic clusters. During development, A- and B-type receptors can be differentially regulated. The mechanisms that allow differential segregation and regulation of A- and B-type receptors are unknown. Presumably, A- and B-type receptors are differentially anchored to the membrane cytoskeleton, but essentially nothing is known about how Drosophila glutamate receptors are localized or anchored. This study identified Coracle, a homolog of mammalian brain 4.1 proteins, in yeast two-hybrid and genetic screens for proteins that interact with and localize Drosophila glutamate receptors. Coracle interacts with the C terminus of GluRIIA but not GluRIIB. To test whether coracle is required for glutamate receptor localization, receptors were immunocytochemically and electrophysiologically examined in coracle mutants. In coracle mutants, synaptic A-type receptors are lost, but there is no detectable change in B-type receptor function or localization. Pharmacological disruption of postsynaptic actin phenocopies the coracle mutants, suggesting that A-type receptors are anchored to the actin cytoskeleton via Coracle, whereas B-type receptors are anchored at the synapse by another (yet unknown) mechanism (Chen, 2005b; full text of article).

Drosophila glutamate receptor mRNA expression and mRNP particles

The processes controlling glutamate receptor expression early in synaptogenesis are poorly understood. This study examined glutamate receptor (GluR) subunit mRNA expression and localization in Drosophila embryonic/larval neuromuscular junctions (NMJs). It was shown that postsynaptic GluR subunit gene expression is triggered by contact from the presynaptic nerve, approximately halfway through embryogenesis. After contact, GluRIIA and GluRIIB mRNA abundance rises quickly approximately 20-fold, then falls within a few hours back to very low levels. Protein abundance, however, gradually increases throughout development. At the same time that mRNA levels decrease following their initial spike, GluRIIA, GluRIIB, and GluRIIC subunit mRNA aggregates become visible in the cytoplasm of postsynaptic muscle cells. These mRNA aggregates do not colocalize with eIF4E, but nevertheless presumably represent mRNP particles of unknown function. Multiplex FISH shows that different GluR subunit mRNAs are found in different mRNPs. GluRIIC mRNPs are most common, followed by GluRIIA and then GluRIIB mRNPs. GluR mRNP density is not increased near NMJs, for any subunit; if anything, GluR mRNP density is highest away from NMJs and near nuclei. These results reveal some of the earliest events in postsynaptic development and provide a foundation for future studies of GluR mRNA biology (Genesan, 2011).

This study has shown that GluR subunit gene expression depends on contact between pre and postsynaptic cells. In response to cell-cell contact, GluR subunit mRNA abundance increases rapidly, but then drops off again to very low levels within a few hours and continues to fall throughout larval development. At the same time that overall GluR mRNA levels decrease during embryogenesis, GluR mRNA aggregates appear throughout the postsynaptic muscle cell cytoplasm. Different GluR mRNAs are not colocalized, but seem to form separate aggregates whose density is proportional to amount of protein required by the cell. eIF4E protein does not appear to be a component of the GluR mRNA aggregates (Genesan, 2011).

It is reasonable that contact between the pre and postsynaptic cells turns on glutamate receptor subunit gene expression, given the huge increase in postsynaptic receptor protein that contact triggers. However, this was not a forgone conclusion. It was equally reasonable for GluR subunit gene expression to be a normal part of muscle development following myoblast fusion, or for contact to trigger translation of pre-existing mRNAs. The signaling cascade that mediates contact-dependent postsynaptic glutamate receptor expression remains unknown. It is now known that the pathway ends in muscle nuclei and triggers a large but transient burst of GluR mRNA production. This will facilitate genetic screens for pathway components (Genesan, 2011).

It is interesting that GluR mRNA aggregate density correlates with the amount of protein required. All glutamate receptors in the Drosophila NMJ contain a GluRIIC subunit, plus either GluRIIA or GluRIIB, with GluRIIA being dominant. The number of GluRIIC mRNA aggregates is approximately equal to the number of GluRIIA and GluRIIB mRNA aggregates combined, and there are more GluRIIA aggregates than GluRIIB aggregates. This suggests that the GluR mRNA aggregates described in this study may be involved in translation. However, eIF4E does not appear to be a component of GluR mRNA aggregates (at least for GluRIIA). Since mRNA is only associated with eIF4E while being actively translated, the GluR mRNA aggregates are probably not translating, but rather might serve to preserve or traffick GluR mRNA in preparation for translation (Genesan, 2011).

The data shows that GluR mRNA aggregates are not preferentially localized near NMJs. These results are consistent with a previous study that used In Situ Hybridization with DIG labeled RNA probes against GluRIIA followed by colorimetric detection and light microscopy. Although the techniques of the previous study could not reveal the small GluR mRNA aggregates that were observed, they definitely showed that GluRIIA was distributed throughout muscles, consistent with the current results. Another study (Sigrist, 2002) used similar techniques and focused on GluRIIA mRNA near NMJs, but did not explicitly claim that GluRIIA mRNA was found exclusively at NMJs. It cannot determine whether or not GluR mRNA may be preferentially translated near NMJs (Genesan, 2011).

If GluR mRNA or any preferentially translated subset of GluR mRNA were localized near NMJs, it would have to be trafficked there. The fact that GluR mRNA aggregates are distributed throughout muscle does not mean that they are not trafficked in interesting ways. Indeed, GluRIIA overexpression led to accumulation of GluRIIA mRNA aggregates near nuclei, possibly due to overload of unknown transport systems. Unfortunately FISH uses fixed tissue and therefore it was not possible to observe GluR mRNA aggregate movement. Attempts were made to tag and visualize GluRIIA transcripts using the 'MS2/MCP-GFP system', which allows live motion tracking of mRNA down to single mRNA level. However, unlike FISH, the MS2/MCP-GFP system does not allow visualization native mRNA. Rather, one or more MS2 'stem-loop' sequences are inserted into a transgenic mRNA of interest. This confers a specific 3D structure that can be bound by GFP-tagged 'MS2 coat binding proteins' (MCPs) expressed in the same cells. A 9X repeat of MS2 sequence was synthesized with multiple cloning sites and was inserted tinto a GluRIIA transgene capable of rescuing loss of GluRIIA in vivo. After verifying that this tagged mRNA was indeed being produced and not being degraded, flies carrying the GluRIIA-MS2 transgene were crossed to flies expressing MCP-GFP. The resulting GluRIIA-MS2;MCP-GFP embryos and larvae were then examined. Unfortunately, most of the MCP-GFP was nuclear, consistent with the fact that MCP-GFP contains a default nuclear localization signal and the idea that MCP-GFP was unable to efficiently associate with GluRIIA-MS2. A small amount of GFP was visible in the cytoplasm, but only in rare instances were punctae seeb mimicking the distribution of native GluRIIA mRNA. Since FISH represents the 'gold standard' for visualization of native mRNA, it is concluded that GluRIIA mRNA aggregation is probably disrupted by incorporation of MS2 or association with MCP-GFP, and results involving indirect tagging of transgenic GluRIIA should be viewed with caution. Although it might be possible to optimize relative expression of GluRIIA-MS2 and MCP-GFP such that tagged GFP behaves similar to native mRNA, it should be noted that the MS2/MCP system has been previously used only for highly expressed mRNAs like nanos, gurken and bicoid, whereas GluRIIA is expressed at relatively low levels (Genesan, 2011).

The best way to determine the function of GluR mRNA aggregates is to disrupt them and see what happens to mRNA stability and GluR protein production. Toward this end, attempts have begun to biochemically isolate and proteomically identify Drosophila GluR mRNA-associated proteins. The results confirm that specific proteins are associated with GluR mRNAs. These proteins include highly conserved but previously unnamed proteins representing novel protein families. Disruption of these proteins causes dramatic loss of GluR protein, consistent with the idea that the GluR mRNA aggregates described describe in this study represent GluR messenger ribonucleoprotein (mRNP) particles. GluR mRNPs have not been previously described (Genesan, 2011).

Consistent with multiplex FISH results, the proteomic screen did not identify eIF4E. This is not surprising. eIF4E associates with mRNA only during active translation. Given the stability of GluR protein (>24 h) and relatively low demand for new protein after initial synaptogenesis, it is likely that GluR mRNAs are not being actively translated most of the time. Instead, they appear to be sequestered in aggregates for utilization as needed (Genesan, 2011).

In summary, this study has presented the first description of GluR mRNA aggregates in relation to each other and the glutamatergic synapse they support. Despite intense interest in synapse formation, receptor trafficking and receptor localization, relatively little interest has been paid to various aspects of glutamate receptor subunit gene expression. Gene expression encompasses many processes, including nuclear transcription and transcript processing (capping, splicing, editing, polyadenylation, etc., and finally nuclear export), through cytoplasmic mRNA trafficking, sequestration, translation and eventual degradation. All of these processes are mediated and/or regulated by mRNPs, which are visible as aggregates, or 'granules' in the nucleus or cytoplasm. It is proposed that the GluR mRNA aggregates described in this study are bona fide GluR mRNP particles, probably required for GluR mRNA stability and/or translation, and that studying their composition and function will lead to important insights concerning glutamate receptor gene expression and nervous system development (Genesan, 2011).

Hebbian plasticity guides maturation of glutamate receptor fields in vivo

Synaptic plasticity shapes the development of functional neural circuits and provides a basis for cellular models of learning and memory. Hebbian plasticity describes an activity-dependent change in synaptic strength that is input-specific and depends on correlated pre- and postsynaptic activity. Although it is recognized that synaptic activity and synapse development are intimately linked, a mechanistic understanding of the coupling is far from complete. Using Channelrhodopsin-2 to evoke activity in vivo, this study investigated synaptic plasticity at the glutamatergic Drosophila neuromuscular junction. Remarkably, correlated pre- and postsynaptic stimulation increased postsynaptic sensitivity by promoting synapse-specific recruitment of GluR-IIA-type glutamate receptor subunits into postsynaptic receptor fields. Conversely, GluR-IIA was rapidly removed from synapses whose activity failed to evoke substantial postsynaptic depolarization. Uniting these results with developmental GluR-IIA dynamics provides a comprehensive physiological concept of how Hebbian plasticity guides synaptic maturation and sparse transmitter release controls the stabilization of the molecular composition of individual synapses (Ljaschenko, 2013).

Repeated light-triggered neurotransmitter release from presynaptic active zones provoked synaptic depression via a decrease in quantal content. Interestingly, muscle depolarization itself also led to a drop in quantal content despite bypassing synapses (Post animals). The latter observation is highly reminiscent of homeostatic communication whereby a retrograde pathway of inverted polarity operates to increase quantal content in response to reduced muscle excitability. Future studies can now test whether molecular components involved in the homeostatic upregulation of quantal content also contribute to its downregulation following postsynaptic ChR2 stimulation. (Ljaschenko, 2013).

Pairing pre- and postsynaptic depolarizations repetitively (pre and post) triggered a synapse-specific increase in postsynaptic GluR-IIA-type GluRs. Hence, correlated activity initiated a Hebbian form of synaptic plasticity at the Drosophila NMJ. A comparison of the cluster size distributions following brief and long pulses suggests two phases of plasticity. The first phase of activity-induced plasticity (15 ms pulses) promotes an evenly distributed increase in the number of clusters. Therefore, despite an increase in total number, the average size of GluR-IIA clusters is not significantly altered. The next phase (2 s pulses) then leads to an increase mainly in the number of large clusters, and hence the average GluR-IIA cluster size increases. Neurotransmitter pr varies across active zones at the NMJ. Because the size of GluR clusters is largest opposite high-pr active zones, it is to be expected that functional recordings of synaptic currents preferentially sample large receptor fields. For this reason, GluR-IIA incorporation likely remained below the detection threshold in electrophysiological recordings following brief light pulses (Ljaschenko, 2013).

In view of the unchanged total number of receptor fields (anti-GluR-IID staining) and active zones, paired stimulation did not appear to give rise to the formation of new synapses. Instead, GluR-IIA was likely incorporated into receptor fields with previously undetectable IIA levels (Ljaschenko, 2013).

In vivo imaging suggests that positive feedback initially promotes GluR-IIA incorporation during synapse growth and that GluR-IIA entry is specifically restrained with further maturation, whereas the rate of GluR-IIB recruitment remains constant. The physiological signals that guide these synapse-specific molecular dynamics are unknown. It is argued that the Hebbian mechanism identified in the present study represents the signal that promotes GluR-IIA entry during synapse development. This is consistent with the observed increase in small clusters following short pulses. Furthermore, in this framework, paired pre- and postsynaptic stimulation would be able to override the inhibition of GluR-IIA incorporation at relatively mature receptor fields and thereby restore the 'juvenile behavior' of the PSDs (Ljaschenko, 2013).

At the developing Drosophila NMJ, receptor field growth is accompanied by BRP-dependent, active zone maturation. Correspondingly, large receptor fields are found opposite high-pr active zones that are rich in BRP. Therefore, small, growing receptor fields opposite immature, low-pr active zones will tend to be exposed to glutamate only when pr is elevated, e.g., during trains of action potentials. Because a large number of other synapses will also be active at these time points, transmitter release will coincide with strong postsynaptic depolarization, leading to Hebbian GluR-IIA incorporation (Ljaschenko, 2013).

A comprehensive model conversely demands a signal to remove GluR-IIA from mature receptor fields in order to describe their diminished rate of IIA incorporation in vivo and to limit receptor-field growth. It was reasoned that such a physiological cue could be provided by sparse (i.e., unsynchronized) transmitter release that preferentially occurs at high-pr, mature synapses and does not trigger substantial muscle depolarization. This hypothesis is experimentally supported by GluR-IIA removal from synapses when muscle depolarization is prevented during neurotransmission (Ljaschenko, 2013).

This study introduces a physiological model in which GluR-IIA is increased at simultaneously active synapses via Hebbian plasticity and is decreased at solitarily active synapses. Such solitary activity may be provided by spontaneous transmitter release (i.e., minis). The physiological function of minis has been controversially discussed. The results suggest that they contribute to 'taming the beast'; in other words, restraining the extent of Hebbian plasticity. The model can account for developmental, synapse-specific receptor subunit dynamics, and explains why GluR-IIA levels are higher opposite low-pr Ib motorneurons than opposite high-pr Is motorneurons. This conceptual framework can account for an increase in GluR-IIA following chronic activity elevation and is consistent with low synaptic IIA levels in the presence of ambient extracellular glutamate, although, intriguingly, sustained glutamate exposure also affects GluR-IIB (Ljaschenko, 2013).

Trains of action potentials are likely the physiological equivalent of paired pre- and postsynaptic depolarization, which simply triggers the Hebbian change more efficiently than solely presynaptic ChR2 stimulation. Notably, rapid GluR-IIA exit can be acutely provoked. This observation is compatible with fast GluR dynamics in mammals, which can operate on a timescale of minutes and well below. Hence, rapid receptor trafficking also occurs in Drosophila, though this probably remains concealed when receptor exit is not explicitly provoked during time-lapse imaging of synapse development in vivo (Ljaschenko, 2013).

Perhaps most conspicuously, activity-dependent bidirectional GluR-IIA mobility is reminiscent of subunit-specific AMPA receptor trafficking at mammalian central synapses, which mediates manifold forms of synaptic plasticity. Local activity has been shown to drive synapse-specific accumulation of GluR1 AMPA receptors. Whereas high-frequency stimulation triggers LTP and synaptic GluR1 incorporation, low-frequency stimulation triggers LTD and GluR1 removal. Collectively, these considerations support the notion that fundamental mechanisms of synaptic plasticity have been strongly conserved during evolution (Ljaschenko, 2013).

Drosophila Neto is essential for clustering glutamate receptors at the neuromuscular junction

Neurotransmitter receptor recruitment at postsynaptic specializations is key in synaptogenesis, since this step confers functionality to the nascent synapse. The Drosophila neuromuscular junction (NMJ) is a glutamatergic synapse, similar in composition and function to mammalian central synapses. Various mechanisms regulating the extent of postsynaptic ionotropic glutamate receptor (iGluR) clustering have been described, but none are known to be essential for the initial localization and clustering of iGluRs at postsynaptic densities (PSDs). This study identified and characterized the Drosophila neto (neuropilin and tolloid-like) as an essential gene required for clustering of iGluRs (GluRIIA, GluRIIB, and GluRIIC) at the NMJ. Neto colocalizes with the iGluRs at the PSDs in puncta juxtaposing the active zones. neto loss-of-function phenotypes parallel the loss-of-function defects described for iGluRs. The defects in neto mutants are effectively rescued by muscle-specific expression of neto transgenes. Neto clustering at the Drosophila NMJ coincides with and is dependent on iGluRs. These studies reveal that Drosophila Neto is a novel, essential component of the iGluR complexes and is required for iGluR clustering, organization of PSDs, and synapse functionality (Kim, 2012).

Once neurons reach their correct postsynaptic targets, a cascade of events marks the beginning of synaptogenesis. The pre- and postsynaptic compartments are kept in register by adhesion molecules, while active zone precursor vesicles and synaptic vesicles arrive at the presynaptic specialization. The assembly of the presynaptic active zones appears to involve the delivery of prefabricated transport packets, although sequential arrival of components has been observed at specialized synapses. The postsynaptic assembly, however, seems to largely depend on gradual de novo clustering of component proteins. The formation of the postsynaptic densities (PSDs) culminates with the recruitment of neurotransmitter receptors. Neuronal activity triggers further synthesis and aggregation of receptor complexes and synapse maturation, stabilization, and growth (Kim, 2012 and references therein).

In contrast to the rich understanding of nicotinic acetylcholine receptor (nAChR) clustering at the mammalian neuromuscular junction (NMJ), clustering of the ionotropic glutamate receptors (iGluRs) that form the majority of central synapses remains less understood. Considerable advances have been made toward identifying proteins that interact with the C-terminal tails of iGluRs and regulate their membrane trafficking, anchoring at the synapses, and involvement in intracellular signaling cascades. In the postsynaptic compartment, proteins that contribute to glutamate receptor clustering at the synapses include PDZ domain-containing proteins, cytoskeleton-binding and scaffolding components, and proteins that control endosomal trafficking. Receptor trafficking and assembly signals have also been found in the N-terminal domains of the iGluRs. Moreover, recent studies using reconstituted synapses have identified a number of presynaptic adhesion molecules and secreted factors that participate in receptor clustering through trans-synaptic protein interactions. For example, Narp (neuronal activity-regulated pentraxin) or other pentraxins secreted from the presynaptic neurons (NP1 and NRP) bind to the N-terminal domain of GluA4 and are critical trans-synaptic factors for GluA4 recruitment at the synapses. The direct coupling of the N-terminal domain of GluA2 to N-cadherin promotes enrichment of AMPA receptors at synapses and maturation of spines, although this interaction could occur in cis or in trans, since N-cadherin is present on both pre- and postsynaptic membranes. These trans-synaptic clustering strategies apply to subsets of iGluR subunits, and it is not clear whether they have a central role in the organization of postsynaptic domains in vivo or rather provide modulatory functions (Kim, 2012).

The Drosophila NMJ is a glutamatergic synapse similar in composition and function to the mammalian central AMPA/Kainate synapses. The fly NMJ iGluRs are heterotetrameric complexes composed of three essential subunits-IIC, IID, IIE-and either IIA or IIB. Type A and type B receptor complexes differ in their single-channel properties, synaptic responses and localization, and regulation by second messengers. Previous studies have shown that the nascent synapses are predominantly type A complexes and change their subunit compositions toward more B-type complexes upon maturation that relies at least in part on CaMKII activity (Kim, 2012 and references therein).

How do iGluR complexes traffic to and cluster at the NMJ? In flies, none of the NMJ iGluR subunits have PDZ- binding motifs. Live-imaging studies on growing synapses have shown that iGluRs from diffuse extrasynaptic pools stably integrate into immature PSDs, but Discs large (Dlg), the fly PSD-95 ortholog, and other postsynaptic proteins remain highly mobile. Dlg does not colocalize with the iGluR receptors at the PSDs and instead is adjacent to the PSDs. Moreover, iGluRs are localized and clustered normally at the NMJ of dlg mutants, although the type B receptor is reduced in levels. The only protein shown to bind directly to iGluR subunits is Coracle, a homolog of mammalian brain 4.1 proteins. Coracle appears to stabilize type A but not type B receptors by anchoring them to the postsynaptic spectrin-actin cytoskeleton. Several more postsynaptic proteins have been identified that regulate the subunit compositions and the extent of iGluR synaptic localization, but no molecules other than the receptors themselves were shown to be absolutely required for clustering of the receptor complexes (Kim, 2012).

One possible link in understanding the trafficking and clustering of iGluRs at the fly NMJ could be provided by the emerging families of auxiliary subunits. Auxiliary subunits are transmembrane proteins that avidly and selectively bind to mature iGluRs and form stable complexes at the cell surface. They can modulate the functional characteristics of iGluRs and may also mediate surface trafficking and/or targeting to specific subcellular compartments. Auxiliary proteins described so far include stargazin and its relatives, cornichon homolog-2 and homolog-3, Cysteine-knot AMPAR-modulating protein, SynDIG1, neuropillin and tolloid-like proteins Neto1 and Neto2, and Caenorhabditis elegans SOL-1. Studies in tissue culture and heterologous systems suggested that some of the auxiliary subunits have the potential to contribute to clustering of iGluRs, since they promote the accumulation of receptors at the cell surface . However, no auxiliary protein has been implicated in the clustering of iGluRs in vivo. In fact, it is unclear whether surface iGluRs must be associated with auxiliary subunits to be functional. For C. elegans, auxiliary subunits are essential for functional receptors, but for vertebrate and Drosophila iGluRs, this remains an open question (Kim, 2012).

Drosophila has several genes reported to encode for auxiliary subunits, including a stargazin-type molecule (Stg1), two cornichon proteins (cni and cnir), the SOL-1-related protein CG34402, and one Neto-like protein. Among them, neto mRNA was found to be expressed in the Drosophila striated muscle. Similar to vertebrate Neto1 and Neto2, this study found that Drosophila Neto is a multidomain, transmembrane protein with two extracellular CUB (for complement C1r/C1s, UEGF, BMP-1) domains followed by an LDLa (low-density lipoprotein receptor domain class A) motif. Unlike vertebrate Netos, Drosophila neto was found to be an essential locus: neto-null embryos are completely paralyzed and cannot hatch into the larval stages. Flies with suboptimal Neto levels, such as in neto hypomorphs, do not fly and have defective NMJ structure and function. Neto was found to be essential in the striated muscle for the synaptic trafficking and clustering of the iGluRs at the PSDs. Moreover, Neto and iGluR synaptic clustering depend on each other. It is proposed that Neto functions as an essential nonchannel component of the iGluR complexes at the Drosophila NMJ (Kim, 2012).

Drosophila neto is an essential locus that encodes for a protein dynamically expressed throughout development. The neto transcript is maternally loaded, and the protein could be detected by Western analysis at all stages of embryogenesis. In spite of a significant maternal pool, the absence of zygotic neto expression produces 100% embryonic paralysis and lethality, suggesting a crucial role for Neto in the later stages of embryogenesis. Fully penetrant embryonic paralysis has been described only for two types of mutants: with defects in epithelial integrity or with nonfunctional NMJ. In the first class, disruption of the blood-brain barrier allows for the potassium-rich hemolymph to flood the CNS, causing hyperactivity and action potential failure. The second class includes mutants that impair the NMJ function. Muscle expression of Neto rescued the lethality and defects of neto- null mutants, indicating an essential role for Neto at the NMJ. These findings fit best with a model in which Neto and iGluRs are engaged in targeting each other to PSDs via direct interaction. In this model, Neto functions as a nonchannel, essential subunit of the iGluR complexes (Kim, 2012).

Indeed, neto loss-of-function phenotypes parallel the loss-of-function defects described for iGluR complexes. First, neto-null mutant embryos lack any body wall peristalsis and hatching movements and have no detectable iGluR clusters at the NMJ. Second, the animals with suboptimal Neto levels have a dramatically reduced number of synaptic iGluR clusters and reduced frequency and amplitude of miniature synaptic potentials. The sparse iGluR clusters in neto109 always colocalize with Neto clusters, indicating that the complexes must contain Neto and iGluRs in order to be incorporated at the PSDs. Finally, Neto-deprived animals exhibit a deficit in the maintenance of mature PSDs. A similar deficit was reported for NMJ synapses developing in the near absence of iGluRs. During synapse formation, iGluR incorporation into the postsynaptic membrane is critical to enlarge PSDs. By clustering in concert to iGluRs, Neto is essential for functional iGluR complexes and directly controls synapse formation at the Drosophila NMJ. An important difference between neto109 and glutamate receptor hypomorpic mutants is that quantal content remains unchanged in neto109 and there is no presynaptic compensation, as seen in receptor mutants. The cause for this difference is not understood, but it is speculated that the lack of presynaptic compensation in neto mutants may reflect a role for Neto in PSD development and maturation and/or in retrograde signaling (Kim, 2012).

Similar to other postsynaptic components, Neto is distributed between junctional and extrajunctional locations on the muscle, as assessed by antibody staining. Outside the NMJ, Neto appears tightly associated with the muscle membrane in a pattern reminiscent of the T tubules. This distribution suggests that Neto could traffic on the muscle surface and perhaps could be mobilized to the junctions as needed. Fully functional iGluR complexes were also detected on the muscle surface at extrajunctional locations (Kim, 2012).

One way in which Neto could control the iGluR clustering is by engaging the receptor complexes on the muscle membrane followed by trafficking to the synaptic junction. This model would be consistent with the Neto/iGluR codependence for clustering at the synapse; i.e., only components engaged in a productive complex could traffic and be stabilized at the NMJ. This model also predicts that, at suboptimal Neto levels, iGluRs will accumulate on the muscle surface at extrajunctional locations. Indeed, this seems to be the case, since in neto hypomorphs, GluRIIA was detected on the muscle surface, accessible by antibodies in the absence of membrane-permeable detergents (Kim, 2012).

In addition, Neto may have a regulatory role in the synaptic targeting of the iGluRs and control the extent of iGluR clustering at the NMJ. Neto may receive and integrate signals about the cellular status and transduce that information into targeting a certain amount of receptors to the synapses. The intracellular domain of Neto is rich in putative phosphorylation sites that may be used to modulate Neto engagement of iGluRs or to connect the complexes with motors and scaffold proteins. Several kinases have been described to control the extent of the iGluR accumulation at the NMJ. Their substrates may include Neto as part of signaling networks that couple cell status to growth of postsynaptic structures (Kim, 2012).

Live-imaging studies have shown that iGluRs from diffuse extrasynaptic pools stably integrate into immature PSDs, while other postsynaptic proteins remain highly mobile. Neto may mediate stable incorporation and stabilization of iGluRs to newly formed PSDs. For example, Neto could promote iGluR aggregation via CUB-mediated self-association and/or extracellular interactions. CUB-containing proteins have been implicated in the formation of acetylcholine receptor aggregates in C. elegans (Gally, 2004). In flies and vertebrates, synaptic aggregation of the neurotransmitter receptors at the NMJ does not occur in the absence of innervating neurons. In vertebrates, neuronally secreted agrin participates in extracellular interactions that enable receptor clustering and synapse stabilization. In Drosophila, the molecular mechanisms that underlie the requirement for innervation to initiate synaptogenesis at the NMJ are not known. A forward genetic screen identified Mind the gap (MTG), a presynaptically secreted protein that appears to organize the extracellular millieu, but it is unclear how MTG could induce postsynaptic differentiation (Rohrbough, 2007). Neto may provide an entry point in understanding these requirements. Our data indicate that by controlling the iGluRs clustering, Neto plays a significant role in the organization and maintenance of the PSDs. Although Neto does not have a PDZ-binding motif, it may participate in both intracellular and extracellular interactions that help stabilize the PSDs (Kim, 2012).

Vertebrate Netos bind to and have a profound impact on the properties of selective kainate receptors: They modulate the agonist-binding affinities and the off kinetics, thus determining the characteristically slow rise time and decay kinetics of synaptic kainate receptors. A role for vertebrate Netos in surface expression of kainate receptors or their redistribution between synaptic and extrasynaptic locations is less clear at this time, as it appears to depend on specific kainate receptor subunits, the neurons and tissues analyzed, and/or the genetic background of the knockout mice tested. Nevertheless, it is possible that Drosophila Neto also modulates the ligand-gated channel properties for iGluRs and shapes the function of synapses at the NMJ (Kim, 2012).

Recent work from vertebrates changed this view on iGluRs: They are not companionless complexes at the PSDs, but rather dynamic supramolecular signaling complexes that include components that regulate the trafficking, scaffolding, stability, signaling, and turnover of the receptors. The discovery of Neto reveals that Drosophila iGluRs also form multisubunit complexes modulated by auxiliary proteins at the fly NMJ. Neto is the first auxiliary iGluR subunit described in Drosophila. In vertebrates, Neto and other auxiliary subunits impart diversity and richness to iGluR function, but no auxiliary protein was shown to be essential for in vivo clustering of the receptors. Auxiliary subunits in C. elegans are essential for functional receptors but not for clustering. The fly Neto is the first example of an auxiliary subunit required for iGluR clustering (Kim, 2012).

An intriguing question is why the requirements for Neto are so different in various species. Neto1/Neto2 double knockout mice have defects in long-term potentiation, learning, and memory but are viable (Tang, 2011). More importantly, Neto1 and Neto2 are not essential for iGluR clustering. In contrast, Drosophila neto-null mutants are embryonic-lethal, and Neto is absolutely required for iGluR clustering. This difference could be due to variations in the properties of individual domains of Netos, or it could reflect the diversity among synapse types and the nature and composition of multiprotein complexes where various Netos function. Indeed, there are primary sequence differences among Netos that could translate into functional differences. For example, the LDLa motif in Neto2 binds Ca2+; the fly Neto lacks the conserved residues predicted to chelate Ca2+ ions. The fly Neto has a long insert between the signal peptide and the first CUB motif. In all Neto proteins, the intracellular domain is rich in potential phosphorylations sites, but in flies, this domain is very acidic (pI 3.86), unlike Neto1 (pI 8.28) and Neto2 (pI 6.62). Secreted isoforms have been reported/predicted for vertebrate Netos but not for Drosophila. Instead, a new transmembrane Neto isoform has been recently entered in the fly database (cDNA reference RE42119). This isoform is predicted to share the exons encoding for extracellular and transmembrane parts, but has alternative exons to encode for a basic (pI 9.17) intracellular domain, with no similarity with vertebrate proteins. While the validated fly Neto isoform is sufficient to provide the essential Neto activity at the NMJ, it will be interesting to investigate whether flies use multiple Neto isoforms at the NMJ or alternate them for tissue- or synapse-specific functions (Kim, 2012).

In flies, Neto is also expressed in subsets of neurons in the CNS; thus, Neto may have additional functions at glutamatergic central synapses. As in vertebrates, neuronal Neto is not essential; only the NMJ function of Neto is required for viability. While a role for Neto at central synapses remains to be determined, it is tempting to speculate that Drosophila Netos might have attained tissue- or context-specific roles in modulation of iGluRs. Thus, Netos constitute a family of conserved proteins that influence the function of glutamatergic synapses and have acquired species- and tissue-specific roles during evolution (Kim, 2012).

Postsynaptic glutamate receptors regulate local BMP signaling at the Drosophila neuromuscular junction

Effective communication between pre- and post-synaptic compartments is required for proper synapse development and function. At the Drosophila neuromuscular junction (NMJ), a retrograde BMP signal functions to promote synapse growth, stability and homeostasis and coordinates the growth of synaptic structures. Retrograde BMP signaling triggers accumulation of the pathway effector pMad in motoneuron nuclei and at synaptic termini. Nuclear pMad, in conjunction with transcription factors, modulates the expression of target genes and instructs synaptic growth; a role for synaptic pMad remains to be determined. This study reports that pMad signals are selectively lost at NMJ synapses with reduced postsynaptic sensitivities. Despite this loss of synaptic pMad, nuclear pMad persisted in motoneuron nuclei, and expression of BMP target genes was unaffected, indicating a specific impairment in pMad production/maintenance at synaptic termini. During development, synaptic pMad accumulation followed the arrival and clustering of ionotropic glutamate receptors (iGluRs) at NMJ synapses. Synaptic pMad was lost at NMJ synapses developing at suboptimal levels of iGluRs and Neto, an auxiliary subunit required for functional iGluRs. Genetic manipulations of non-essential iGluR subunits revealed that synaptic pMad signals specifically correlate with the postsynaptic type-A glutamate receptors. Altering type-A receptor activities via protein kinase A (PKA) revealed that synaptic pMad depends on the activity and not the net levels of postsynaptic type-A receptors. Thus, synaptic pMad functions as a local sensor for NMJ synapse activity and has the potential to coordinate synaptic activity with a BMP retrograde signal required for synapse growth and homeostasis (Sulkowski, 2013).

Previous work has described Neto as the first nonchannel subunit required for the clustering of iGluRs and formation of functional synapses at the Drosophila NMJ. Neto and iGluR complexes associate in the striated muscle and depend on each other for targeting and clustering at postsynaptic specializations. This study shows that Neto/iGluR synaptic complexes induce accumulation of pMad at synaptic termini in an activity-dependent manner. The effect of Neto/iGluR clusters on BMP signaling is selective, and limited to synaptic pMad; nuclear accumulation of pMad appears largely independent of postsynaptic glutamate receptors. This study demonstrates that synaptic pMad mirrors the activity of postsynaptic type-A receptors. As such, synaptic pMad may function as an acute sensor for postsynaptic sensitivity. Local fluctuations in synaptic pMad may provide a versatile means to relay changes in synapse activity to presynaptic neurons and coordinate synapse activity status with synapse growth and homeostasis (Sulkowski, 2013).

Drosophila NMJs maintain their evoked potentials remarkably constant during development, from late embryo to the third instar larval stages. This coordination between motoneuron and muscle properties requires active trans-synaptic signaling, including a retrograde BMP signal, which promotes synaptic growth and confers synaptic homeostasis. Nuclear pMad accumulates in motoneurons during late embryogenesis. However, embryos mutant for BMP pathway components hatch into the larval stages, indicating that BMP signaling is not required for the initial assembly of NMJ synapses and instead modulates NMJ growth and development. This study demonstrates that synaptic accumulation of pMad follows GluRIIA arrival at nascent NMJs and depends on optimal levels of synaptic Neto and iGluRs. As type-A receptors have been associated with nascent synapses, and type-B receptors mark mature NMJs, accumulation of synaptic pMad appears to correlate with a growing phase at NMJ synapses. Furthermore, synaptic pMad correlates with the activity and not the net levels of postsynaptic type-A receptors. In fact, expression of a GluRIIA variant with a mutation in the putative ion conduction pore triggered reduction of synaptic pMad levels. Thus, synaptic pMad functions as a molecular sensor for synapse activity and may constitute an important element in synapse plasticity (Sulkowski, 2013).

The synaptic pMad pool has been localized primarily to the presynaptic compartment. However, a contribution for postsynaptic pMad to the pool of synaptic pMad is also possible. Postsynaptic pMad accumulates in response to glia-secreted Mav, which regulates gbb expression and indirectly modulates the Gbb-mediated retrograde signaling (Fuentes-Medel, 2012). RNAi experiments revealed that knockdown of mad in muscle induces a decrease in synaptic pMad, albeit much reduced in amplitude compared with knockdown of mad in motoneurons (Fuentes-Medel, 2012). Also, knockdown of wit in motoneurons, but not in muscle, and knockdown of put in muscle, but not in motoneurons, triggers reduction of synaptic pMad (Fuentes-Medel, 2012). Intriguingly, the synaptic pMad is practically abolished in GluRIIA and neto109 mutants and cannot be further reduced by additional decrease in Mad levels. Whereas loss of postsynaptic pMad could be due to a Mav-dependent feedback mechanism that controls Gbb secretion from the muscle, the absence of presynaptic pMad demonstrates a role for GluRIIA and Neto in modulation of BMP retrograde signaling (Sulkowski, 2013).

As BMP signals are generally short lived, synaptic pMad probably reflects accumulation of active BMP/receptor complexes at synaptic termini. Recent evidence suggests that BMP receptors traffic along the motoneuron axons, with Gbb/receptors complexes moving preferentially in a retrograde direction. By contrast, Mad does not appear to traffic. Thus, Mad is likely to be phosphorylated and maintained locally by a pool of active Gbb/BMP receptor complexes that remain at synaptic termini for the time postsynaptic type-A receptors are active (Sulkowski, 2013).

The activity of type-A glutamate receptors may control synaptic pMad accumulation (1) indirectly via activity-dependent changes that are relayed to both pre- and postsynaptic cells, or (2) directly by influencing the production and signaling of varied Gbb ligand forms or by localizing Gbb activities. For example, inhibition of postsynaptic receptor activity induces trans-synaptic modulation of presynaptic Ca2+ influx. Such Ca2+ influx changes may trigger events that induce a local change in synaptic pMad accumulation. One possibility is that changes in Ca2+ influx may recruit Importin-β11 at presynaptic termini, which in turn mediate synaptic pMad accumulation (Sulkowski, 2013).

At the Drosophila NMJ, Gbb is secreted in the synaptic cleft from both pre- and postsynaptic compartments. The secretion of Gbb is regulated at multiple levels, transcriptionally and post-translationally. Furthermore, the Gbb prodomain could be processed at several cleavage sites to generate Gbb ligands with varying activities. The longer, more active Gbb ligand retains a portion of the prodomain that could influence the formation of Gbb/BMP receptor complexes. Synaptic pMad may result from signaling by selective forms of Gbb. Or type-A receptors could modulate secretion and processing of Gbb in an activity-dependent manner. Understanding the function of different pools and active forms of Gbb within the synaptic cleft will help explain the multiple roles for Gbb at Drosophila NMJs (Sulkowski, 2013).

Alternatively, active postsynaptic type-A receptor complexes may directly engage and stabilize presynaptic Gbb/BMP receptor signaling complexes via trans-synaptic interactions. CUB domains can directly bind BMPs; thus Neto may utilize its extracellular CUB domains to engage Gbb and/or presynaptic BMP receptors. As synaptic pMad mirrors active type-A receptors, such trans-synaptic complexes will depend on Neto in complexes with active type-A receptors. No capture has yet been shown of a direct interaction between Gbb and Neto CUB domains in co-immunoprecipitation experiments. Nonetheless, a trans-synaptic complex that depends on the activity of type-A receptors could offer a versatile means for relaying synapse activity status to the presynaptic neuron via fast assembly and disassembly (Sulkowski, 2013).

Irrespective of the strategy that correlates synaptic pMad pool with the active type-A receptor/Neto complexes, further mechanisms must act to maintain the Gbb/BMP receptor complexes at synapses and protect them from endocytosis and retrograde transport. Such mechanisms must be specific, as general modulators of BMP receptors endocytosis impact both synaptic and nuclear pMad. A candidate for differential control of BMP/receptor complexes is Importin-β11. Loss of synaptic pMad in importin-β11 is rescued by neuronal expression of activated BMP receptors, by blocking retrograde transport, but not by neuronal expression of Mad. As Mad does not appear to traffic, presynaptic Importin-β11 must act upstream of the BMP receptors, perhaps to stabilize active Gbb/BMP receptor complexes at the neuron membrane. By contrast, local pMad cannot be restored at Neto-deprived NMJs by overactivation of presynaptic BMP receptors or by blocking retrograde transport. As neto and gbb interact genetically, it is tempting to speculate that postsynaptic Neto/type-A complexes localize Gbb activities and stabilize Gbb/BMP receptor complexes from the extracellular side. Additional extracellular factors, for example heparan proteoglycans, or intracellular modulators, such as Nemo kinase, may control the distribution of sticky Gbb molecules within the synaptic cleft and their binding to BMP receptors, or may stabilize Gbb/BMP receptor complexes at synaptic termini (Sulkowski, 2013).

Synaptic pMad may act locally and/or in coordination with the transcriptional control of BMP target genes to ensure proper growth and development of the synaptic structures. A presynaptic pool of pMad maintained by Importin-β11 neuronal activities ensures normal NMJ structure and function. Like importin-β11, GluRIIA and Neto-deprived synapses show a significantly reduced number of boutons. Intriguingly, the absence of GluRIIA induces up to 20% reduction in bouton numbers, whereas knockdown of GluRIIB does not appear to affect NMJ growth. Although the amplitude of the growth phenotypes observed in normal culturing conditions (25°C) was modest, this phenomenon may explain the requirement for GluRIIA reported for activity-dependent NMJ development (at 29°C). Furthermore, knockdown of Neto or any iGluR essential subunit affect synaptic pMad and NMJ growth in a dose-dependent manner. Not significant changes were found in nuclear pMad or expression of BMP target genes in GluRIIA or Neto-deprived animals, but the restoration of synaptic pMad by presynaptic constitutively active BMP receptors rescues the morphology and physiology of importin-β11 mutant NMJs. The smaller NMJs observed in the absence of local pMad may reflect a direct contribution of synaptic pMad to retrograde BMP signaling, a pathway that provides an instructive signal for NMJ growth. Thus, BMP signaling may integrate synapse activity status with the control of synapse growth (Sulkowski, 2013).

Synaptic pMad may also contribute to synapse stability. Mutants in BMP signaling pathway have an increased number of 'synaptic footprints': regions of the NMJ where the terminal nerve once resided and has retracted. It has been proposed that Gbb binding to its receptors activates the Williams Syndrome-associated Kinase LIMK1 to stabilize the NMJ. Synaptic pMad may further contribute to the stabilization of synapse contacts by engaging in interactions that anchor the Gbb/BMP receptor complexes at synaptic termini. During neural tube closure, local pSmad1/5/8 mediates stabilization of BMP signaling complexes at tight junction via binding to apical polarity complexes. Flies may utilize a similar anchor mechanism that relies on pMad-mediated interactions for stabilizing BMP signaling complexes and other components at synaptic junctions. Local active BMP signaling complexes are thought to function in this manner in the maintenance of stemness and in epithelial-to-mesenchymal transition (Sulkowski, 2013).

Separate from its role in synapse growth and stability, BMP signaling is required presynaptically to maintain the competence of motoneurons to express homeostatic plasticity. The requirements for BMP signaling components for the rapid induction of presynaptic response may include a role for synaptic pMad in relaying acute perturbations of postsynaptic receptor function to the presynaptic compartment. At the very least, attenuation of local pMad signals, when postsynaptic type-A receptors are lost or inactive, may release local Gbb/BMP receptor complexes and allow them to traffic to neuron soma and increase the BMP transcriptional response, promoting expression of presynaptic components and neurotransmitter release. In addition, synaptic pMad-dependent complexes may influence the composition and/or activity of postsynaptic glutamate receptors. Although future experiments will be needed to address the nature and function of local pMad-containing complexes, the current findings clearly demonstrate that synaptic pMad constitutes an exquisite monitor of synapse activity status, which has the potential to relay information about synapse activity to both pre- and postsynaptic compartments and contribute to synaptic plasticity. As BMP signaling plays a crucial role in synaptic growth and homeostasis at the Drosophila NMJ, the use of synaptic pMad as a sensor for synapse activity may enable the BMP signaling pathway to monitor synapse activity then function to adjust synaptic growth and stability during development and homeostasis (Sulkowski, 2013).

Neto-mediated intracellular interactions shape postsynaptic composition at the Drosophila neuromuscular junction

The molecular mechanisms controlling the subunit composition of glutamate receptors are crucial for the formation of neural circuits and for the long-term plasticity underlying learning and memory. This study use the Drosophila neuromuscular junction (NMJ) to examine how specific receptor subtypes are recruited and stabilized at synaptic locations. In flies, clustering of ionotropic glutamate receptors (iGluRs) requires Neto (Neuropillin and Tolloid-like), a highly conserved auxiliary subunit that is essential for NMJ assembly and development. Drosophila neto encodes two isoforms, Neto-α and Neto-β, with common extracellular parts and distinct cytoplasmic domains. Mutations that specifically eliminate Netoβ or its intracellular domain were generated. When Neto-β is missing or is truncated, the larval NMJs show profound changes in the subtype composition of iGluRs due to reduced synaptic accumulation of the GluRIIA subunit. Furthermore, neto-β mutant NMJs fail to accumulate p21-activated kinase (PAK), a critical postsynaptic component implicated in the synaptic stabilization of GluRIIA. Muscle expression of either Neto-α or Neto-β rescued the synaptic transmission at neto null NMJs, indicating that Neto conserved domains mediate iGluRs clustering. However, only Neto-β restored PAK synaptic accumulation at neto null NMJs. Thus, Neto engages in intracellular interactions that regulate the iGluR subtype composition by preferentially recruiting and/or stabilizing selective receptor subtypes (Ramos, 2015).

At the Drosophila NMJ, Neto enables iGluRs clustering at synaptic sites and promotes postsynaptic differentiation. This study shows that Neto-β, the major Neto isoform at the fly NMJ, plays a crucial role in controlling the distribution of specific iGluR subtypes at individual synapses. Similar to other glutamatergic synapses, the subunit composition determines the activity and plasticity of the fly NMJ. The data are consistent with a model whereby Neto-β, via its conserved domains, fulfills a significant part of Neto-dependent iGluRs clustering activities during synapse assembly. At the same time, Neto-β engages in intracellular interactions that regulate iGluR subtypes distribution by preferentially recruiting and/or stabilizing type-A receptors. In this model, Neto-β could directly associate with the GluRIIA-containing complexes and/or regulate the synaptic abundance of type-A receptors indirectly, by recruiting PSD components such as PAK. Thus, Neto-β employs multiple strategies to control which flavor of iGluR will be at the synapses and to modulate PSD composition and postsynaptic organization (Ramos, 2015).

Neto proteins have been initially characterized as auxiliary subunits that modulate the function of kainate (KA) and NMDA receptors. In vertebrates, Neto1 and Neto2 directly interact with KAR subunits and increase channel function by modulating gating properties. Since loss of KAR currents in mice lacking Neto1 and/or Neto2 exceed a reduction that could be attributed to alterations of channel gating, an additional role for Neto proteins in synaptic targeting of receptors has been proposed. The role for vertebrate Neto proteins in KAR membrane and/or synaptic targeting remains controversial and appears to be cell type-, receptor subunit-, and Neto isoform-dependent. Furthermore, the C. elegans Neto has a very small intracellular domain (24 amino acids beyond the conserved domains). This implies that 1) Neto without an intracellular domain constitutes the minimal conserved functional moiety, and 2) the divergent intracellular domains of Neto proteins may fulfill tissue and/or synapse specific modulatory functions. Indeed, Neto2 bears a class II PDZ binding motif that binds to the scaffold protein GRIP and appears to mediate KARs stabilization at selective synapses (Ramos, 2015).

In flies, Neto is an essential protein that plays active roles in synapse assembly and in the formation and maintenance of postsynaptic structures at the NMJ. The Drosophila Neto isoforms do not have PDZ binding motifs, but they use at least two different mechanisms to regulate the synaptic accumulation and subunit composition of iGluRs. First, Neto participates in extracellular interactions that enable formation of iGluR/Neto synaptic complexes; formation of stable aggregates is presumably prevented by the inhibitory prodomain of Neto. Second, the two Neto isoforms appear to facilitate the selective recruitment and/or stabilization of specific iGluR subtypes. It is speculated that Neto-β may selectively associate with and recruit type-A receptors, perhaps by engaging the C-terminal domain of GluRIIA, which is critical for the synaptic stabilization of these receptors. Aside from a possible role in the selective recruitment of iGluR subtypes, Neto-β participates in intracellular interactions that facilitate the recruitment of PAK at PSDs; in turn, PAK signals through two independent, genetically separable pathways (a) to modulate the GluRIIA synaptic abundance and (b) to facilitate formation of SSR (Ramos, 2015).

Whether Neto-β recruits PAK directly or via a larger protein complex remains to be determined. Neto-β contains an SH3 domain that may bind to the proline-rich SH3 binding domain of PAK. However, in tissue culture experiments, attempts to detect a direct interaction between PAK and Neto-β (full-length or intracellular domain) failed. PAK synaptic accumulation is completely abolished at NMJ with mutations in dPix, although not all dpix defects are mediated through PAK. Conversely, PAK together with Dreadlocks (Dock) controls the synaptic abundance of GluRIIA, while PAK and dPix regulate the Dlg distribution. The reduction of GluRIIA and Dlg synaptic abundance observed at neto-β mutant NMJs suggests that Neto-β may interact with both dPix and Dock and enable both PAK activities. In addition, Neto-β may stabilize postsynaptic type-A receptors by enhancing their binding to Coracle, which anchors GluRIIA to the postsynaptic actin cytoskeleton (Ramos, 2015).

Importantly, this study connects the complex regulatory networks that modulate the PSD composition to the Neto/iGluR clusters themselves. The Neto-β cytoplasmic domain is rich in putative protein interaction motifs, and may function as a scaffold platform to mediate multiple protein interactions that act synergistically during synapse development and homeostasis. Loss of Neto-β-mediated intracellular interactions at netoβshort NMJs reduced the GluRIIA synaptic abundance, but did not affect the GluRIIB synaptic signals. It is unlikely that the remaining cytoplasmic part of Neto-β facilitates the GluRIIB synaptic accumulation at these NMJs at the expense of GluRIIA and PAK. Instead, a model is favored whereby the synaptic stabilization of GluRIIA requires a Neto-β-dependent intracellular network. Disruption of this network diminishes GluRIIA and increases GluRIIB synaptic abundance, pending the availability of limiting subunits, GluRIIC-E and Neto. Conversely, the presence of this network ensures that at least some type-A receptors are stabilized at synaptic sites, even at Neto-deprived synapses, such as in netohypo larvae [12]. Assembly of this network does not require GluRIIA since both Neto-β and PAK accumulated normally at GluRIIA mutant NMJs. Furthermore, in the absence of Neto-β the synaptic abundance of GluRIIA can be partly restored by excess Neto-α or a δ-intracellular Neto variant, suggesting that excess iGluRs 'clustering capacity' overrides the cellular signals that shape PSD composition. What intracellular domain(s) of Neto bind to and how they are modified by post-translational modifications will be critical questions to understand how postsynaptic composition is regulated during development and homeostasis (Ramos, 2015).

The discovery of Drosophila Neto isoforms with alternative cytoplasmic domains and isoform specific activities expands the repertoire of Neto-mediated functions at glutamatergic synapses. All Neto proteins share the highly conserved CUB1, -2, LDLa and transmembrane domains that have been implicated in engaging and modulating the receptors, the central function of Neto proteins. In flies this conserved part is both required and sufficient for iGluRs clustering and NMJ development. In C. elegans the entire Neto appears to be reduced to this minimal functional unit. The only exception is a retina-specific CUB1-only Neto1 isoform with unknown function. In contrast to shared domains, the cytoplasmic domains are highly divergent among Neto proteins. This diversity might have evolved to facilitate intracellular, context specific function for Neto proteins, such as the need to couple the iGluR complexes to neuron or muscle specific scaffolds in various phyla. By engaging in different intracellular interactions, via distinct cytoplasmic domains, different Neto isoforms may undergo differential targeting and/or retention at the synapses and thus acquire isoform-specific distributions and functions within the same cell (Ramos, 2015).

Phylogenetic analyses indicate that the intracellular domains of Neto are rapidly evolving in insects. Blocks of high conservations could be clearly found in the genome of short band insect Tribolium castaneum (Coleoptera) or in Apis mellifera (Hymenoptera). However, most insects outside Diptera appear to have only one Neto isoform, more related to Neto-β. In fact, the only Neto-α isoform outside Drosophila was found in Musca domestica (unplaced genomic scaffold NCBI Reference Sequence: XM_005187241.1). Other neto loci, from Hydra to vertebrates, appear to encode Neto proteins with unique and highly divergent intracellular domains. An extreme example is the C. elegans Neto/Sol-2, with a very short cytoplasmic tail, which requires additional auxiliary subunits, Sol-1 and Stargazin, to control the function of glutamate receptors. Neto proteins appear to utilize their intracellular domains to connect to the signaling networks that regulate the distribution and subunit composition for iGluRs. Such cellular signals converge onto and are integrated by the intracellular domains of the receptors and/or by various auxiliary subunits associated with the iGluR complexes (Ramos, 2015).

Neto proteins modulate the gating behavior of KAR but also play crucial roles in the synaptic recruitment of glutamate receptors in vivo. At the fly NMJ, Neto enables iGluRs synaptic clustering and initiates synapse assembly. In addition, the intracellular domain of Neto-β recruits PSD components and triggers a cascade of events that organize postsynaptic structures and shape the composition of postsynaptic fields. The cytoplasmic domains of Neto proteins emerge as versatile signaling hubs and organizing platforms that directly control the iGluRs subunit composition and augment the previously known Neto functions in modulation of glutamatergic synapses (Ramos, 2015).

Presynaptic DLG regulates synaptic function through the localization of voltage-activated Ca(2+) channels

The DLG-MAGUK subfamily of proteins plays a role on the recycling and clustering of glutamate receptors (GLUR) at the postsynaptic density. discs-large1 (dlg) is the only DLG-MAGUK gene in Drosophila and originates two main products, DLGA and DLGS97 which differ by the presence of an L27 domain. Combining electrophysiology, immunostaining and genetic manipulation at the pre and postsynaptic compartments, this study examined the DLG contribution to the basal synaptic-function at the Drosophila larval neuromuscular junction. The results reveal a specific function of DLGS97 in the regulation of the size of GLUR fields and their subunit composition. Strikingly the absence of any of DLG proteins at the presynaptic terminal disrupts the clustering and localization of the calcium channel DmCa1A subunit (Cacophony), decreases the action potential-evoked release probability and alters short-term plasticity. These results show for the first time a crucial role of DLG proteins in the presynaptic function in vivo (Astorga, 2016).

dlg1 is the only gene of the DLG-MAGUK subfamily in Drosophila. Similar to vertebrate genes, two forms of the gene are expressed as the result of two transcription start sites. DLGA (α form) and DLGS97 (β form) are distinguished by the inclusion of an L27 domain in beta forms located in the amino terminus of the protein. In vertebrates DLG4/PSD95 is predominantly expressed as α form while DLG1/SAP97 is mainly expressed as β form. DLGA is expressed in epithelial tissues, somatic muscle and neurons; in turn, DLGS97 is not expressed in the epithelial tissue. In the larval neuromuscular junction (NMJ), a glutamatergic synapse, both dlg products are expressed pre and postsynaptically. Hypomorphic dlg alleles display underdeveloped subsynaptic reticulum, bigger glutamate receptors fields and an increased size of synaptic boutons, active zones and vesicles. Additionally to these morphological defects, altered synaptic responses such as increased excitatory junction currents (EJC) and increased amplitude of miniature excitatory junction potentials have been observed. The strong morphological defects make difficult to distinguish developmental defects from the role of DLGs in the basal function of the mature synapse. Previously studies have reported form-specific null mutant strains for DLGA (dlgA40.2) and DLGS97, (dlgS975). These mutants do not show the gross morphological defects observed in hypomorphic mutants, although still show functional synaptic defects, supporting a role of DLG proteins in the mature synaptic function (Astorga, 2016).

Combining genetic, electrophysiology and immunostaining techniques this study dissected the role of DLG proteins at the pre and postsynaptic compartments. The results show the specific requirement of postsynaptic DLGS97 for normal glutamate receptor (GLUR) distribution. In turn, both DLG proteins increase the release probability by promoting voltage-dependent Ca2+ channel localization. The results demonstrate for the first time a relevant role to DLG proteins in the presynaptic function contributing to Ca2+ mediated short-term plasticity and probability of release (Astorga, 2016).

Flies carrying the severe hypomorph dlg1 mutant allele, dlgXI-2 and the isoform specific dlgS97 null mutant displayed increased amplitude of the spontaneous excitatory postsynaptic (junctional) potential (mEJP) without changes in frequency. In addition all mutants displayed a decreased quantal content as measured by evoked post-synaptic potentials. The specific defects behind these results were explored. To characterize the synaptic transmission in WT and dlg mutants, post synaptic currents were recorded in HL3.1 solution on muscles 6 or 7 of third instar male larvae of the various genotypes. Recordings of spontaneous excitatory postsynaptic currents (mEJC) were obtained after blocking the voltage activated sodium channels. Thereafter, histogram distributions were constructed with the amplitudes of the miniature events and the quantal size was estimated by the peak value obtained adjusting a Log-Normal distribution in each genotype. It is worth to emphasize that finding a phenotype on dlgA or dlgS97 mutants means that DLGA or DLGS97 proteins by themselves cannot replace DLG function (Astorga, 2016).

The average amplitude of spontaneous postsynaptic potentials were compared and, supporting previous results, it was found that the average amplitude of the mEJC of the mutants dlgXI-2 (0.99 ± 0.05 nA) and dlgS97 (0.98 ± 0.03 nA) were significantly larger compared to the average amplitudes of the mEJC of Canton-S strain used as WT control (0.81 ± 0.04 nA) and of dlgA (0.78 ± 0.02 nA) specific mutant. The same result was obtained comparing the quantal size. None of the mutants showed a significant change compared to the WT in the frequency of the mEJC. As an additional control, all mutants were recorded over a deficiency covering the dlg gene, finding similar results. These findings are in accordance with the idea that DLGS97 protein and not DLGA is necessary for the quantal size determination (Astorga, 2016).

Bigger quantal size could be of pre or postsynaptic origin as the result of increased neurotransmitter (NT) content in vesicles or increased glutamate receptor field's size respectively. First, to determine the pre or post-synaptic origin of this phenotype, a UAS-dsRNA construct that targets all dlg products, was expressed under the control of the motoneuron promoter OK6-GAL4 or the muscle promoter C57-GAL4. As expected for a post-synaptic defect, the increased quantal size observed in dlgS97 mutants was mimicked only by the decrease of DLG in the muscle. The specific role of DLGS97 in the muscle is supported by the rescue of the dlgS97 mutant phenotype only by the expression of DLGS97 in the muscle and not in the motor neuron. The effect of GAL4 expression was examined in the mutant background in all experiments; neither of the GAL4 lines without the specific UAS constructs changed the phenotype of the mutants. Again, none of the genotypes studied displayed differences with the WT in the frequency of the minis (Astorga, 2016).

Changes in quantal size of postsynaptic origin could be due to higher number of post-synaptic receptors and/or a different composition of the postsynaptic receptors. An increase in the size of glutamate receptors fields has been described in dlg hypomorphic alleles including dlgXI-2 mutants. Therefore, the size of the glutamate receptor fields was compared among the mutants and with WT, and also the active zones were measured using antibodies for the active zone protein Bruchpilot. Consistently with previous results bigger glutamate receptors fields were found compared to WT only in dlgXI-2 and dlgS97 mutants but not in dlgA mutants. Surprisingly, an increased number of active zones per bouton was also found in all mutants, a phenotype usually associated with an increase in the frequency of minis that were not observe. In addition, an increased active zone area was found in dlgA and dlgXI-2 mutants (Astorga, 2016).

As expected for a postsynaptic defect, the bigger size of the glutamate fields in dlgS97 mutants was rescued by the expression of DLGS97 in the muscle but not by its expression in the motor neuron. These results confirm that DLGS97, but not DLGA is responsible for the regulation of the size of the receptors fields in the muscle (Astorga, 2016).

The strict requirement of DLGS97 in the regulation of the size of GLUR fields supports results that have involved other DLGS97 interacting proteins in the regulation of the size of the glutamate receptors fields. METRO, an MPP-like MAGUK protein, has been shown to form a complex with DLGS97 and LIN-7 through the L27 domains present in each of the three proteins. metro mutants display decreased DLGS97 at the synapse and larger GLUR receptors fields than WT, even bigger than dlgS97 mutants. METRO and DLGS97 depend on each other for their stability on the synapse, thus, in dlgS97 mutants, METRO and dLIN-7 are highly reduced at the synapse. The similar post-synaptic phenotype of metro and dlgS97 and the reported interaction between these two proteins suggests the proposal that the increase size of GLUR fields is consequence of the loss of METRO due to the loss of DLGS97 protein (Astorga, 2016).

As stated before, changes in quantal size of postsynaptic origin can also reflect a different composition of the receptors. Drosophila NMJ GLUR receptors are tetramers composed by obligatory subunits and two alternative subunits, GLURIIA and GLURIIB. Receptors composed by one of these two subunits differ in their kinetics; GLURIIB receptors desensitizes faster than GLURIIA receptors. Thus, the kinetic of the spontaneous currents (mEJCs), is associated to the relative abundance of these two types of receptors in the GLUR fields. It has been shown that the abundance of GLURIIB but not of GLURIIA in the synapse is associated with the expression of dlg. To analyze if dlg mutants display a change in the composition of the subunits abundance relative to the control, the kinetics of the mEJCs were studied. Kinetics analyses of the mEJCs revealed that only dlgS97 and the double mutant display a slower kinetic in the off response, which is compatible with a different composition of the glutamate receptors fields regarding the proportion between GLURIIA and B receptors. The value of tau also increased in larvae expressing dsRNA-dlg in the muscle, but not by its expression in the motor neuron. Finally tau-off values recovered the WT value only with the expression of DLGS97 in the muscle. As slower mEJCs were observed, the results suggest an increase in the ratio of GLURIIA/GLURIIB. It is known that receptors containing the GLURIIA subunit display bigger conductance and slower inactivation kinetics than receptors containing the GLURIIB subunit. Thus, synapses with post-synaptic receptors fields containing proportionally less GLURIIB subunits would display bigger and slower mEJCs similar to the phenotype observed in dlgS97 mutants. To confirm this hypothesis, the abundance of GLURIIA and GLURIIB receptors was evaluated by immunofluorescence in the NMJ of WT and dlg mutant larvae. The immunofluorescence that allowed the detection and quantification of GLURIII and GLURIIB fields was performed with paraformaldehyde (PFA) fixative. However, the immunofluorescence to detect GLURIIA receptors only works fixating the tissue with Bouin reagent. Thus, in order to be able to compare between these two fixations, the size of the GLUR fields was normalized by the HRP staining that labels the whole presynaptic bouton. First, as a control, GLURIIA and GLURIII were double stained in the same larvae. The results show that using PFA fixative, GLURIII fields display bigger size only in dlgS97 mutants and not in dlgA mutants. Even more, as predicted from the kinetic data, only dlgS97 and not dlgA mutants display bigger GLURIIA fields while there are not difference in the size of GLURIIB fields between WT and the mutants. Additionally the results show no difference in the number of GLURIIA or GLURIIB clusters between WT and dlg mutants. Immunohistochemical results confirm the prediction from the electrophysiological data revealing that in dlgS97 mutants, GLURIIA subunits are proportionally more abundant in GLUR fields than in control larvae. In conclusion, the results show that dlgS97 mutants display larger quanta and mEJCs with slower kinetic establishing its participation in the regulation of the size of GLUR fields where the increased size is obtained mainly through the recruitment of receptors containing GLURIIA subunits. As a similar result was obtained in another study that observed that the loss of GLURIIB receptors in the NMJ of dlgXI-2 mutant embryos, these observation suggest that either of the two DLG proteins are necessary for the localization of GLURIIB in the synapse but only DLGS97 is actively limiting the size of the clusters by regulating the number of GLURIIA receptors (Astorga, 2016).

Taking into account previous reports that show the regulation of the synaptic localization of DLG by CAMKII, the regulation of the subunit composition by CAMKII and these results, a mechanism is proposed by which, after a strong activation of CAMKII, the phosphorylation of DLGS97 would detach it from the synapse allowing the increase of the size of the GLUR fields by the recruitment of GLURIIA over GLURIIB. These changes should increase the synaptic response by two different mechanisms (Astorga, 2016).

To determine if DLG proteins modulate the presynaptic release probability, excitatory junction currents (EJC) were recorded in the muscle by stimulating the nerve at 0.5 Hz in low extracellular Ca2+ (0.2 mM), both conditions to avoid synaptic depression. For all mutant genotypes the average peak amplitude and quantal content (EJC amplitude/quantal size) of the evoked responses were significantly smaller than WT. In congruence with previous results, the lower amplitude of the current response is accompanied by a decrease in the quantal content. Taking into account the results on the size of the GLUR fields in the mutant's muscles, these results are compatible with a reduction of the neurotransmitter release in dlg mutants. A decreased neurotransmitter release could be associated with a decreased number of release sites in the boutons. However, the number of active zones per bouton is increased in all dlg mutants with bigger active zones in dlgXI-2 and dlgA mutants (Astorga, 2016).

The decrease in the evoked response could be a consequence of the absence of the specific form of DLG in the postsynaptic side, transmitted by unknown mechanisms or, alternatively, it could be the result of an effect of DLG on the probability of release. In order to explore where this phenotype originates (pre or post-synaptically) DLG levels were downregulated by expressing dsRNA against all forms of dlg. Compatible with a presynaptic defect, the expression of UAS-dsRNA-dlg presynaptically decreases the amplitude of the evoked response while the same construct expressed postsynaptically using C57 promoter did not changed the amplitude of the EJCs. The presynaptic expression of the dsRNA-dlg also mimics the reduction in quantal content of the mutants, displaying a severe reduction in this parameter. On the other hand, the postsynaptic expression of the dsRNA-dlg associates to a moderate but significant decrease on the quantal content, as expected from the effect already reported of the postsynaptic dsRNA-dlg on the quantal size and the lack of effect on the amplitude of the EJCs. The presynaptic effect of DLG is supported further by the rescue experiments. Thus, the amplitude of the evoked response and the quantal content in dlgA40.2 mutant is completely rescued by the selective expression of DLGA in the presynaptic compartment but not by its expression in the postsynaptic compartment. The pre-synaptic expression of DLGS97 improves the synaptic function increasing the average size of the EJCs such that the difference between the WT and the presynaptic-rescue is not significant, suggesting a complete rescue. However, the average EJC in the presynaptic rescue is not different either from the control mutant animal, which is interpreted as the rescue not being complete and thus the term partial rescue is used. DLGS97 does not, however rescued at al the quantal content. This is explained because although the amplitude of the current increased, the quantal size remains unchanged by the presynaptic expression of DLGS97. In consequence the quantal content does not increase as much as the current. On the other hand, the postsynaptic expression did not increase the amplitude of the evoked current. However, since it does rescue the quantal size the quantal content augmented enough to be different from the mutant control. Notably, DLGA expressed presynaptically in dlgA40.2mutants not only rescued the EJC amplitude but also the number of active zones per bouton and the size of the active zones. On the other hand DLGS97 expression only partially rescued the increased number of active zones in dlgS975 mutants. These results support a role of DLG proteins in the presynaptic function where DLGA seems to regulate more aspects than DLGS97. Despite the fact that both forms of DLG share most of their protein domains, neither of the two-forms is able to fully rescue the absence of the other, suggesting that both of them participate in a complex. The binding between the SH3 and GUK domains of MAGUK proteins has been described; this interaction (at least in vitro) is able to form intra or intermolecular associations and offers a mechanism by which DLGA and DLGS97 proteins could be associated to recruit proteins to a complex (Astorga, 2016).

Changes in the overall quantal content at these synapses may reveal presynaptic defects. However, genetic background and other independent modification could alter apparent release. To independently scrutinize alteration in the presynaptic release probability two presynaptic properties were examined, the short-term plasticity and the calcium dependency of quantal release (Astorga, 2016).

To explore the EJC phenotype observed in dlg mutants, stimulation paradigms were carried out that allow characterization of aspects of the short term plasticity that are known to depend on presynaptic functionality and give clues about the mechanisms involved in the observed defects. First, the response were studied of the mutants to high frequency stimulation, 150 stimuli at 20 Hz. WT responses at high frequency stimulation show a fast increase in the amplitude of the response that then slows down. The fast initial increase is called facilitation and the second phase with smaller slope is called augmentation. The time constant of the facilitation is believed to reflex the calcium dynamics in the terminal and its slope to be the product of the accumulation of calcium and the consequent calcium dependent increase in the probability of release. The fractional increment in the mutants' responses showed an increased facilitation in all mutants, while an increased augmentation was only significant in dlgA mutants compared to WT. Additionally all mutants showed a trend toward steeper slopes than WT, but only the augmentation slope in dlgA mutants reached statistical significance. These results support that the mutants display a lower probability of release than WT, which could reflect defects in the calcium dynamics or in the response to calcium (Astorga, 2016).

Previous work in dlg mutants did not report defects in short-term plasticity. These works differ from the current one in methodological aspects, mainly that they were carried out in a media with high concentration of magnesium (20 mM) and calcium (1.5 mM). This work was carried out in a media containing low magnesium (4 mM) and calcium (0.2 mM) concentration. It is known that magnesium reduces neurotransmitter release, probably due to partial blockade of VGCC. Additionally, magnesium permeates more than sodium and potassium through GLURs (Astorga, 2016).

To better evaluate the calcium dynamics in the terminal pair pulse (PP) experiments, a well-known paradigm to evaluate presynaptic calcium dynamics, were carried out. In PP, a second depolarization shortly after the first one carried out in low extracellular calcium concentration elicits an increased release of neurotransmitter thought to reflect the increased calcium concentration in the terminal reached after the first stimulus. According to this, and posing as the working hypothesis that DLG affects presynaptic calcium dynamics, a second pulse would be expected to increase the release in a bigger proportion, since the first stimulus did not release much of the ready releasable pool. Conversely, a second stimulus given at high calcium concentration produces a decrease in the release of neurotransmitter, which is considered to originate in the partial depletion of the ready releasable pool at the release sites. Thus, a second pulse at high calcium concentration should elicit a smaller decrease of the release since an inferior entrance of calcium should produce less depletion of the ready releasable pool of vesicles (Astorga, 2016).

Consistently with a decreased calcium entrance, all mutants displayed increased pair pulse facilitation at low calcium concentration and decreased pair pulse depression at high calcium concentration. These results support a defect in the calcium entrance to the terminal as the underlying defect in dlg mutants causing the evoked stimuli defects. To characterize the calcium dependency of the release in the mutants the evoked responses were measured at different calcium concentrations. It can be observed that for all the mutants and at most calcium concentrations, the quantal content of the evoked response is lower than the control. The only exception is seen at 2 mM calcium where the quantal content of the dlgA mutants and the control are not different to each other. However, even at this calcium concentration the quantal content of dlgS97 and the double mutant dlgXI-2 are significantly lower than the control. To get insight about the release process the responses were fit to a Hill equation. This type of fitting better estimate the maximum responses and the EC50, which is masked in the overall release of different backgrounds. This is observed in the graph with the normalized responses by the maximal quantal predicted. The adjusted curves show that mutants reach the theoretical maximal quantal content at higher calcium concentration than the WT and that the EC50 for the mutants is diminished respect to the WT. To confirm the presynaptic origin of the defect in the calcium dependency, the calcium dependency was carried out in the mutant genotypes expressing DLGA or DLGS97 pre or postsynaptically. The quantal content analysis shows that only the presynaptic expression of DLGA in dlgA mutants completely rescued the calcium dependency, in line with previous results that show the importance of DLGA in the presynaptic compartment. On the other hand the presynaptic expression of DLGS97, although it rescued the calcium dependency, failed to rescue the maximal quantal content. Observing the graph with the normalized responses, DLGA as well as DLGS97 both are able to restore the WT calcium dependency. The inability to rescue the maximal quantal content could be explained by the existence of synaptic compensatory mechanisms that allow to counterweigh the bigger quantal size in dlgS97 mutants, which were shown before not to be rescued by the presynaptic expression of DLGS97 (Astorga, 2016).

Facilitation is thought to depend on the resultant of the calcium entrance, calcium release from intracellular stores and the clearance of cytosolic calcium. So, the defects in facilitation observed in the mutants could be due to a decreased calcium entrance but also they could be due to a defect on the clearance of calcium. In a preliminary experiment, the relative changes were measured of the total intracellular calcium concentration in the bouton using the genetically encoded calcium indicator GCamp6f. GCamp6f expressed in control flies (OK6-GAL4/UAS-GCamp6f) respond with a fast and transitory change in the cytoplasmic calcium of the boutons when they are exposed to a local pulse of potassium. The same experimental approach in dlgS97 mutant larvae reveals that the rise of the calcium response is significantly slower than the control; additionally the recovery of the response is also significantly slower. These preliminary experiments suggest a defect in calcium entrance in the mutants but they also support a defect in the extrusion that hint to additional defects. Further experiments are needed to clarify the calcium kinetics involved since these experiments were measuring the bulk of calcium change and in doing this approximation, the nanodomain changes that are known to be the ones that regulate the neurotransmitter release are being lost (Astorga, 2016).

Since the results described above including the calcium dependency of the release as well as the parameters of the short-term plasticity suggest that the calcium entrance to the terminal is impaired, a view that is supported by the preliminary data measuring the cytosolic calcium, the next experiments focused on the calcium entrance. The main calcium entry to the terminal is the voltage gated calcium channel (VGCC) encoded by the Drosophila gene cacophony. Advantage was taken of a UAS-cacophony1-EGFP transgenic fly (CAC-GFP) to study the distribution of the channel in WT and mutant genotypes. CAC-GFP overexpressed in WT background localizes in the synapse in a strictly plasma membrane-associated manner in big clusters closely associated with release sites. However, CAC-GFP overexpressed in dlgS97 or dlgA mutant background displays a significant decrease in the expression accompanied by a more disperse localization with significantly smaller clusters, suggesting that the Cacophony protein might not be properly delivered or anchored to the plasma membrane in dlg mutants (Astorga, 2016).

It was reasoned that if dlg mutants had a defect on calcium entrance, the over expression of calcium channels should rescue at least partially the phenotype. Advantage was taken of the fact that CAC-GFP construct encodes a functional channel, and recordings were taken from control and dlg mutants overexpressing CAC. As expected and supporting a decreased calcium entrance in the mutants, dlgS97 and dlgA mutants that overexpress CAC-GFP display significantly bigger evoked EJCs compared to dlg mutants, without a change of phenotype in the spontaneous currents. Additionally, the over expression of CAC-GFP partially rescued the pair pulse facilitation and the pair pulse depression as well as the calcium curve (Astorga, 2016).

The disrupted localization of CAC could result from the disturbance of a normal direct association to DLG or it could be affected indirectly. To test an immunoprecipitation assay was carried out using flies that express CAC-EGFP in all neurons. Antibodies against GFP were able to precipitate DLG together with Cacophony-GFP, supporting that Cacophony channel is part of the DLG complex in the boutons (Astorga, 2016).

A possible interaction between DLG and voltage-gated calcium channels (VGCCs is) the VGCC auxiliary subunits. The α2δ auxiliary subunit (Straightjacket in Drosophila) increases calcium channel activity and plasma membrane expression of CaV2 α1 subunits and Cacophony. The β auxiliary subunit increases plasma membrane expression of several mammalian VGCC classes. Intriguing β subunits are also MAGUK proteins and they are able to release the VGCC α subunit from the endoplasmic reticulum retention. It may be speculated that DLG through their SH3-GUK domain might be playing the role of the β subunit (Astorga, 2016).

On the other hand, in mammalian cultured neurons it has been proposed that a complex formed by the scaffold proteins LIN-2/CASK, LIN-10/MINT and LIN-7 is involved in the localization of VGCCs at the synapse and that SAP97 forms a complex with CASK. The association of DLGS97 with LIN7 has been reported in the postsynaptic compartment in the Drosophila NMJ. Furthermore, an association between the L27 domain of DLGS97 and the L27 domain of Drosophila CASK has been shown in vitro, however there are no reports of this type of association with DLGA. Another protein involved in the localization of calcium channels in the active zone is RIM. Drosophila rim has been involved in synaptic homeostasis and the modulation of vesicle pools. Surprisingly rim mutants, display low probability of release and altered responses to different calcium concentrations. Recently it was shown that spinophilin mutants display a phenotype with bigger quantal size and GLUR fields size with a higher proportion of GLURIIA subtype of receptors as well as decreased EJCs and decreased pair pulse facilitation. This is a phenotype very similar to the one described here for dlg mutants. The authors in this report did not explore the calcium channels abundance or distribution and the current study did not explore the link of DLG to Neuroligins, Neurexins and Syd. It would be interesting to determine if there is a link between Spinophilin and DLG (Astorga, 2016).

Taken together these results show that dlgS97 is the main isoform responsible for the postsynaptic defects in the dlgXI-2 mutants; which comprise the increase in the size of the receptors fields and the change in the ratio of GLURIIA/GLURIIB. The results as well support a model in which DLG forms a presynaptic complex that includes Cacophony where the absence of either form of DLG leads to defects in the localization of the voltage dependent calcium channel and to a decrease in the entrance of calcium to the bouton; which in turn affect the probability of release and the short-term plasticity in the mutants. The results described in this study highlight the specificity of the function of DLGS97 and DLGA proteins and describe for the first time an in vivo presynaptic role of DLG proteins (Astorga, 2016).

Acute fasting regulates retrograde synaptic enhancement through a 4E-BP-dependent mechanism

While beneficial effects of fasting on organismal function and health are well appreciated, little is known about the molecular details of how fasting influences synaptic function and plasticity. Genetic and electrophysiological experiments demonstrate that acute fasting blocks retrograde synaptic enhancement that is normally triggered as a result of reduction in postsynaptic receptor function at the Drosophila larval neuromuscular junction (NMJ). This negative regulation critically depends on transcriptional enhancement of eukaryotic initiation factor 4E binding protein (4E-BP) under the control of the transcription factor Forkhead box O (Foxo). Furthermore, these findings indicate that postsynaptic 4E-BP exerts a constitutive negative input, which is counteracted by a positive regulatory input from the Target of Rapamycin (TOR). This combinatorial retrograde signaling plays a key role in regulating synaptic strength. These results provide a mechanistic insight into how cellular stress and nutritional scarcity could acutely influence synaptic homeostasis and functional stability in neural circuits (Kauwe, 2016).

Many forms of dietary restriction can reduce cellular stress, improve organismal health, and in many instances extend lifespan in a number of model organisms. A major cellular function that is highly sensitive to nutrient intake from yeast to mammals is cap-dependent translation under the regulation of the target of rapamycin (TOR). TOR promotes cap-dependent translation primarily through phosphorylation of 4E-BPs (eukaryotic initiation factor 4E binding proteins) and p70 S6Ks (S6 ribosomal protein kinases). Phosphorylation of 4E-BP suppresses its ability to bind and inhibit the interaction between eIF4E (eukaryotic initiation factor 4E) and the initiation factor eIF4G, a critical step for translation initiation. In addition to the regulation by TOR, 4E-BP undergoes upregulation in response to dietary restriction and starvation. Together these two responses result in a strong inhibition of protein synthesis and act as a metabolic brake. Multiple lines of evidence suggest that fasting-induced increase in ketone bodies influences neuronal excitability and aspects of neurotransmitter release; however, little is known about how different forms of dietary restriction, by influencing protein translation, can exert an effect on the regulation of synaptic function and plasticity (Kauwe, 2016).

At the Drosophila larval neuromuscular junction (NMJ), the genetic removal of GluRIIA, one of five glutamate receptor subunits, reduces the postsynaptic response to unitary release of neurotransmitter. As a result of this reduced response to neurotransmitter, a retrograde signal is triggered in the postsynaptic muscle that ultimately leads to a compensatory enhancement in presynaptic release from the motor neuron, a process that is conserved at the vertebrate NMJs. The maintenance of this homeostatic synaptic compensation or retrograde synaptic enhancement is highly sensitive to postsynaptic cap-dependent translation in Drosophila; mutations in either Target of Rapamycin (TOR) or eIF4E can dominantly suppress the synaptic compensation in GluRIIA mutants (Penney, 2012). Interestingly, postsynaptic overexpression of TOR or S6K, in an otherwise wild-type muscle, is also sufficient to trigger a retrograde enhancement in presynaptic neurotransmitter release, suggesting that normal synaptic strength may be affected by a postsynaptic signal from the muscle (Penney, 2012; Kauwe, 2016 and references therein).

Previous findings have demonstrated that postsynaptic translation plays a critical role in the regulation of retrograde synaptic enhancement at the NMJ. Therefore, in light of the effect of dietary restriction on TOR-dependent translation, this study set out to investigate the consequence of nutrient restriction on retrograde synaptic compensation in GluRIIA mutants. Electrophysiological analysis indicates that acute fasting, but not amino acid restriction, blocks this retrograde synaptic compensation. This block is not merely due to reduced TOR activity, but rather a result of transcriptional upregulation of postsynaptic 4E-BP under the control of the transcription factor Foxo. These results indicate that the retrograde regulation of synaptic strength at the NMJ depends on the balance between 4E-BP and TOR (Kauwe, 2016).

A few hours of fasting can have a strong impact on retrograde synaptic enhancement at the Drosophila larval NMJ. Removal of food source acutely activates 4E-BP transcription in postsynaptic muscles in a Foxo-dependent manner, thereby leading to the inhibition of retrograde synaptic enhancement at the NMJ. The results indicate that Foxo and 4E-BP act cell autonomously in postsynaptic muscles to exert a retrograde negative regulation on presynaptic neurotransmitter release. Future studies are needed to test whether fasting-induced alterations in insulin signaling underlie the transcriptional upregulation of 4E-BP via its effect on Foxo in postsynaptic muscles. While 4E-BP-mediated suppression of synaptic enhancement as a result of fasting could be considered undesirable during development, it can be beneficial under conditions of abnormally high synaptic activity. As such, 4E-BP-mediated inhibition of retrograde synaptic enhancement and the subsequent dampening of circuit activity might provide an explanation for the beneficial effects of fasting in reducing seizures in some cases. Similarly, in cases where dysregulation of TOR activity is thought to underlie abnormal circuit activity, such as in TSC models, intermittent fasting could potentially dampen the increase in synaptic release through a 4E-BP-dependent inhibition, thereby stabilizing neuronal circuits (Kauwe, 2016).

In addition to its role as a molecular responder to stress, 4E-BP exerts a constitutive negative regulation on presynaptic neurotransmitter release at the NMJ. Electrophysiological analysis of loss-of-function mutant larvae indicates that 4E-BP functions in postsynaptic muscles to constitutively provide a retrograde negative influence on synaptic strength. In light of these findings, a two-pronged scheme is proposed for the retrograde regulation of synaptic strength at the NMJ. On the one hand, a positive input from TOR is mediated through S6K/eIF4A and eIF4E to enhance postsynaptic translation. Synaptic compensation in GluRIIA mutant larva appears to rely mostly on this axis as evidenced by strong sensitivity to S6K heterozygosity and no change in the proportion of phosphorylated 4E-BP versus non-phosphorylated 4E-BP levels. Opposing this positive input, 4E-BP inhibits translation by sequestering eIF4E and adjusting the degree of retrograde compensation. Indeed, loss of 4E-BP leads to a strong increase in quantal content that is highly sensitive to eIF4E heterozygosity but not sensitive to S6K heterozygosity. The balance between these two forces reveals itself also when 4E-BP loss-of-function mutants are rescued by a non-phosphorylatable 4E-BP transgene. In this combination TOR can no longer inhibit 4E-BP, and this study finds that the presynaptic neurotransmitter release is lower than wild-type, similarly to what is observed in TOR hypomorphic mutants (Penney, 2012). A working model is proposed in which the negative force of 4E-BP is under constant check via phosphorylation by TOR, and the positive input from TOR/S6K is constitutively countered by 4E-BP’s ability to sequester eIF4E, a dynamic duel that ensures a tight regulation of synaptic strength (Kauwe, 2016).


Embryonic and Larval

The RNA expression pattern of DGluRIIB was established by means of embryonic whole-mount in situ hybridization. DGluRIIB RNA is observed exclusively in muscle. It first appears at late stage 12 and reaches its highest levels at stage 14. In stages 15-17, DGluRIIB expression is lower but is still present in somatic musculature. Low levels are observed in the gut-associated muscle. This expression pattern is similar but not identical to that of DGluRIIA. DGluRIIA is also first observed at stage 12 and is expressed exclusively in muscle, but in contrast to DGluRIIB, it increases gradually until it reaches its highest levels in stages 16 and 17 (Currie, 1995 and Peterson, 1997).

Glutamate is the excitatory transmitter at neuromuscular junction synapses in Drosophila, and electrophysiological studies indicate that the receptors for glutamate are concentrated in muscle fibers at synaptic sites. Acetylcholine is the excitatory transmitter at vertebrate neuromuscular synapses, and previous studies have shown that accumulation of acetylcholine receptors (AChRs) at synaptic sites is controlled both by transcriptional and post-translational mechanisms. The transcriptional pathway culminates in selective expression of AChR subunit genes in nuclei near the synaptic site, causing AChR mRNA to accumulate in the synaptic region of the muscle fiber. A cDNA encoding a subunit of the Drosophila muscle glutamate receptor (DGluR-IIA) was used to determine the temporal and spatial expression pattern of the DGluR-IIA gene during embryogenesis and in larval muscle. DGluR-IIA mRNA is first expressed at stage 12 of embryogenesis and that expression is detected in developing dorsal, lateral, and ventral somatic muscles within the next 2 hr. By stage 16 DGluR-IIA mRNA is expressed in all somatic muscles and in pharyngeal muscles. In third instar larvae DGluR-IIA mRNA is expressed in all body-wall muscle fibers. DGluR-IIA mRNA, however, is expressed throughout the larval muscle fibers and is not concentrated within muscle fibers at neuromuscular synapses. These results indicate that although the DGluR-IIA gene is expressed in somatic muscle cells it is not selectively expressed in nuclei near the synaptic site (Currie, 1995).

To investigate the subcellular localization of DGluRIIA and DGluRIIB, each receptor was tagged with the myc-epitope. The epitope was incorporated immediately following a heterologous signal sequence that was used to replace the endogenous signal sequences of each gene. As such, the myc-epitope is present at the extracellular N terminus of each receptor. Transgenic flies were generated that express the tagged receptors under the control of the muscle-specific myosin heavy-chain promoter. Both DGluRIIB and DGluRIIA are localized to the neuromuscular junctions (NMJs) of body-wall muscles in third instar larvae (Peterson, 1997).

The NMJs of third instar larvae have been subdivided by morphological and physiological criteria. Type I synapses have larger boutons and contain small, clear, glutamate-filled vesicles, while Type II synapses have small boutons and are primarily peptidergic. While many muscles possess only Type I synapses, a number of muscles are innervated by both Type I and Type II synapses. To assess whether glutamate receptors are differentially localized to a particular class of synapse within a single cell, double staining was performed for Synaptotagmin, a marker of all presynaptic terminals and for glutamate receptor. Confocal microscopy reveals that DGluRIIB is localized to the postsynaptic specialization surrounding presynaptic terminals of Type I boutons, but no staining for receptors is observed at Type II boutons. Staining for DGluRIIA gives the same result. This indicates that at the Drosophila NMJ, as in vertebrate central neurons, glutamate receptors are differentially localized to particular synapses within a single cell (Peterson, 1997).

A hallmark of Type I boutons is the presence of an elaborate postsynaptic specialization, the subsynaptic reticulum (SSR), which consists of numerous layers of invaginated membrane surrounding the presynaptic terminal. Molecules localized to the SSR such as the PDZ-containing protein Discs-Large (Dlg) and the cell adhesion molecule Fasciclin II (FasII) appear to form a halo surrounding the entire presynaptic terminal when analyzed by confocal microscopy. In contrast, both DGluRIIA and DGluRIIB appear as bright spots adjacent to the presynaptic terminal. These hot spots of receptor localization are of the appropriate size and pattern to represent postsynaptic receptor clusters opposite presynaptic release sites. To investigate this possibility immunoelectron microscopy was performed. The EM analysis confirms the patchy distribution of receptors surrounding the bouton. Receptors are localized to particular regions of the synaptic cleft and are nearly undetectable in the underlying invaginations of the SSR. These patches of receptors around synaptic boutons are always observed opposite a presynaptic terminal containing accumulations of synaptic vesicles and tightly apposed, parallel presynaptic and postsynaptic membranes that are characteristic of active zones. Hence, the clusters of receptors visible by confocal microscopy appear to be postsynaptic markers of vesicle release sites. Shaker (Sh) potassium channels and FasII require Dlg for clustering at synapses. It was of interest to see whether DGluRIIA or DGluRIIB also require Dlg for their localization. To address this question, the myc-tagged proteins were stained in a dlg mutant in which Shaker fails to cluster to the NMJ. No change was seen in glutamate receptor localization. Hence, other proteins are likely to function in the localization of these glutamate receptors (Peterson, 1997).

Changes in the distribution and density of transmitter receptors in the postsynaptic cell are required steps for functional synapse formation. Antibodies were raised against Drosophila glutamate receptors (DGluR-IIA) and the distribution of receptors were visualized during neuromuscular junction formation in embryos. In wild-type embryos, embryonic development is complete within 22 hr after egg lying (AEL) and neuromuscular junction (NMJ) formation begins at 13 hr AEL. At the time of initial synapse formation, DGluR-IIAs appears as clusters closely associated with some muscle nuclei. Subsequently, these nonjunctional clusters disperse while DGluR-IIAs accumulate at the junctional region. In a paralytic temperature-sensitive mutant, para(ts1), neural activity decreases drastically at restrictive temperatures. When neural activity is blocked throughout synaptogenesis by rearing embryos at a restrictive temperature prior to the beginning of synaptogenesis, 12 hr AEL, the dispersal of extrajunctional clusters is significantly suppressed and no accumulation of receptors at the junction is observed at 22 hr AEL. However, when neural activity is blocked later, by rearing embryos at a restrictive temperature from 13 hr AEL, DGluR-IIAs does not accumulate at the NMJ, although extrajunctional clusters disperse normally. These findings suggest that the neural activity differentially regulates dissipation of receptor clusters in the nonjunctional region and accumulation of receptors at the junctional region (Saitoe, 1997).

During the formation of neuromuscular junctions in Drosophila embryos, glutamate receptors undergo a drastic change in distribution. To study the underlying mechanism of this developmental process, it is desirable to map the distribution of functional receptors with accurate spatial resolution. Since glutamate receptors desensitize within several milliseconds, the agonist must be applied rapidly. To fulfil these requirements laser stimulation of a caged compound was used to release L-glutamate at a focal spot. Since the glutamate receptor channel is permeable to Ca2+, the change in internal Ca2+ concentration was examined using a Ca2+ indicator, fluo-3. Using this approach, the distribution of functional glutamate receptors were mapped in cultured embryonic Drosophila myotubes and myoblasts. Consistent with previous immunofluorescence studies using an antibody against a glutamate receptor subunit, a large increase of internal Ca2+ concentration is observed when laser stimulation is located close to some nuclei in the myotube. No change is detected when the laser stimulus is applied over any regions of the myoblasts. No increase of the internal Ca2+ concentration in myotubes is observed when the external solution contains either glutamate at a desensitizing concentration (1 mM) or a glutamate receptor channel blocker, argiotoxin (1 microg/ml). These results indicate that a rise in intracellular Ca2+ concentration can be used to show the distribution of the functional receptor on the muscle surface membrane (Saitoe, 1998).

Little is known about the functional significance of spontaneous miniature synaptic potentials, which are the result of vesicular exocytosis at nerve terminals. By using Drosophila mutants with specific defects in presynaptic function it has been found that glutamate receptors cluster normally at neuromuscular junctions of mutants that retain spontaneous transmitter secretion but have lost the ability to release transmitter in response to action potentials. In contrast, receptor clustering is defective in mutants in which both spontaneous and evoked vesicle exocytosis are absent. Thus, spontaneous vesicle exocytosis appears to be tightly linked to the clustering of glutamate receptors during development (Saitoe, 2001).

The existence of miniature end-plate potentials provides a basis for the theory of quantal synaptic transmission. A single miniature end-plate potential arises when a synaptic vesicle fuses spontaneously with the presynaptic membrane and releases a quantum of transmitter (spontaneous vesicle exocytosis). However, little is known about the functional importance of this process. Presynaptic and postsynaptic neurotoxins that allow spontaneous vesicle exocytosis to persist have little effect on synaptic development, including postsynaptic accumulation of receptors. During the development of Drosophila neuromuscular junctions (NMJs), glutamate receptors (GluRs) cluster in the postsynaptic membrane in a manner that depends on nerve-muscle contact. To investigate the role of spontaneous secretory events in receptor clustering, Drosophila mutants with distinctive secretory defects were used. Mutations of neuronal-synaptobrevin (n-syb) or cysteine string protein (csp) selectively prevent nerve-evoked exocytosis whereas spontaneous vesicle exocytosis persists. In contrast, syntaxin-1A (syx) or shibire (shi) mutations eliminate both spontaneous and evoked exocytosis, thereby allowing one to distinguish the role of spontaneous secretory events (Saitoe, 2001).

Neuromuscular transmission in wild-type and mutant Drosophila embryos or larvae were characterized. A typical burst of excitatory synaptic currents (ESCs) often exceeded 600 pA in amplitude in a newly hatched wild-type larva (control). In the presence of tetrodotoxin, the bursting of ESCs is suppressed, and ESCs seldom exceeded 400 pA. A similar suppression of bursting and reduction in the amplitude of ESCs is observed when the ventral nerve cord is removed. Thus, propagated activity in the nervous system triggers multiple vesicle exocytosis and contributes to the ESCs. Concomitantly, the residual events [miniature ESCs (mESCs)] in these wild-type larvae are due to spontaneous vesicle exocytosis (Saitoe, 2001).

An n-syb null mutant was investigated in which nerve-evoked ESCs but not mESCs are lost. Consistent with these findings, ESCs were detected in n-syb embryos but virtually no large-amplitude ESCs characteristic of nerve-evoked ESCs. This apparent absence of evoked responses (but persistence of mESCs) was confirmed by the fact that TTX had no effect on the frequency or amplitude of ESCs and that no evoked ESCs were elicited by nerve stimulation (Saitoe, 2001).

In syx mutants, both nerve-evoked and mESCs are undetectable. In agreement with this phenotype, neither nerve-evoked nor mESCs were detected during observations exceeding 15 min each in seven cells. Although the possibility of missing very infrequent occurrences could not be completely eliminated, it is clear that the frequency of mESCs in syx embryos is far lower than the frequency of mESCs in n-syb embryos. Given the distinct phenotypes of the n-syb and syx mutants, the distribution of postsynaptic GluRs was examined in these embryos (Saitoe, 2001).

Preparations were also stained with antibody against horseradish peroxidase (anti-HRP), which binds to a neuronal surface antigen and reveals the presynaptic terminals. Immunoreactive GluRs formed prominent junctional clusters that closely mirrored the presynaptic elements in wild-type and n-syb mutants. Although this finding apparently contrasts with the observation in para (Na+ channel) mutants, that neural electrical activity is essential for the clustering of receptors, it should be noted that the n-syb mutation is a more subtle perturbation of this system. Unlike n-syb, however, syx mutants rarely have discernible junctional GluR clusters, although they invariably form neuromuscular contacts (Saitoe, 2001).

These findings raised the possibility that the absence of detectable mESCs in these mutants may have arisen from a deficit of postsynaptic GluRs. Moreover, the lack of receptor clusters could either be a developmental consequence of a lack of vesicle fusions in the nerve terminal, or could be due to a requirement for syntaxin in the trafficking of GluRs to the postsynaptic membrane. Indeed, syx is required for cell viability in Drosophila, and the development of both the neuron and muscle in syx embryos is likely to be due to small amounts of maternal Syx. If this residual Syx is not adequate for the maintenance of normal vesicular traffic to the cell surface, GluRs may not be inserted appropriately in the sarcolemma. To address these issues, it was determined whether syx mutants responded to applied glutamate and also whether junctional GluR clusters are restored in syx mutants by selectively inducing the presynaptic or postsynaptic expression of a syx transgene (Saitoe, 2001).

The application of L-glutamate by pressure ejection onto the junctional region of syx muscles indicates that some sensitivity is lost concomitantly with the loss of clusters. The pressure ejection of L-glutamate at the junctional region yielded robust inward currents in wild-type and n-syb mutants, but glutamate-evoked currents are much smaller in syx mutants. What is important here is that even with this diminished glutamate sensitivity, mESCs would have been detected in syx mutants if their terminals were spontaneously releasing transmitter. Thus, the complete absence of detectable mESCs could not be attributed to a lack of postsynaptic sensitivity (Saitoe, 2001).

The low sensitivity to glutamate in syx mutants could reflect the diminished trafficking of GluRs to the surface. Thus, a test was performed to see whether the clustering defect is due to a pre- or post-synaptic action of syx by the targeted expression of the syx transgene with either the neuron-specific elav-GAL4 driver or the muscle-specific G14-GAL4 driver. Neuron-specific expression of the syx transgene restores GluR clusters but muscle-specific expression does not. Evoked and mESCs are readily observed in transgenic embryos in which neuron-specific expression of the syx transgene is restored but not in transgenic embryos in which syx transgene is expressed in muscles. Thus, a comparison of the phenotypes of n-syb and syx has led to a hypothesis that spontaneous secretory events at the NMJ are critical to the formation of GluR clusters (Saitoe, 2001).

Two temperature-sensitive (ts) paralytic mutants were used to examine further the correlation between spontaneous vesicle exocytosis and GluR clustering. At elevated temperatures, a defect in dynamin in shits blocks synaptic vesicle recycling and thereby depletes the terminals of synaptic vesicles. In contrast, cspts mutations appear to interfere with excitation-secretion coupling in the terminal. Synapses in shits mutants become completely silent at a nonpermissive temperature (32°C), whereas cspts mutants lose evoked responses while retaining mESCs. Because of these differences in release properties at the nonpermissive temperature, the distribution of GluRs in these lines was compared. At a permissive temperature (25°C), when release properties are similar among these embryos, GluR clustering is comparable for wild-type, cspts, and shits mutant embryos. However, this situation changes when embryos are moved to the nonpermissive temperature at 13 hours after fertilization, which is when nerve-muscle contacts first form. The development of GluR clusters is not perceptibly altered in wild-type and cspts mutants at 32°C. However, no detectable GluR clusters are observed in shits mutants, as is the case in syx mutants. These results again suggest a tight link between spontaneous vesicle exocytosis and GluR clustering (Saitoe, 2001).

Further insight into the nature of interaction of presynaptic and postsynaptic elements has come from the injection of argiotoxin into wild-type embryos at concentrations that block all muscle contractile activity. In these embryos, GluRs still cluster postsynaptically. Thus, it was not the activation of postsynaptic GluRs that directs GluR clustering. Similar findings have been reported in vertebrates, where alpha-bungarotoxin does not impede the clustering of acetylcholine receptor (AChR). As in vertebrates, secretion of molecules, such as agrin for AChRs, ephrins for N-methyl-D-aspartate (NMDA)-type GluRs, and neuronal activity-regulated pentraxin for AMPA-type GluRs, may drive receptor clustering by being released with, or in parallel to, the neurotransmitter at Drosophila NMJs (Saitoe, 2001).

A positive correlation has been documented between ongoing spontaneous vesicle exocytosis and the embryonic development of GluR clusters at Drosophila NMJs. Nerve-evoked vesicle exocytosis is not necessary for this process, because although neither n-syb nor cspts mutants show any demonstrable nerve-evoked ESCs, GluRs still cluster. mECSs persist in both mutants. However, when spontaneous secretory events are absent (as in shits mutants at the nonpermissive temperature or in syx), junctional GluR clusters are exceedingly infrequent. If GluR clustering is solely contingent on the nerve-muscle contact, GluRs should cluster at the contacts in shits mutants raised at the nonpermissive temperature and in syx mutants. In a recent study of mice lacking an isoform of munc 18-1, there was no demonstrable change in AChR clustering, although both spontaneous and evoked neurotransmitter release are absent. Together with the observations made in this paper, these data suggest that munc 18-1 is not involved in the secretion of the agent that induces clustering of neurotransmitter receptors, whereas syntaxin is essential for this process (Saitoe, 2001).

The absence of clusters in Drosophila in syx and shi mutants implies that spontaneous secretory events are related to GluR clustering and probably to cluster stabilization as well. Moreover, it is the clustering of these receptors, rather than their surface expression, that depends on spontaneous secretion: Functional GluRs are detected in syx mutants although they rarely form detectable clusters at the synapse. The link between spontaneous vesicle exocytosis and receptor clustering must now be clarified (Saitoe, 2001).

Experience-dependent strengthening of Drosophila neuromuscular junctions

The genetic analysis of larval neuromuscular junctions (NMJs) of Drosophila has provided detailed insights into molecular mechanisms that control the morphological and physiological development of these glutamatergic synapses. However, because of the chronic defects caused by mutations, a time-resolved analysis of these mechanisms and their functional relationships has been difficult so far. This study provides a first temporal map of some of the molecular and cellular key processes that are triggered in wild-type animals by natural larval locomotor activity and then mediate experience-dependent strengthening of larval NMJs. Larval locomotor activity was increased either by chronically rearing a larval culture at 29° C instead of 18°C or 25°C or by acutely transferring larvae from a culture vial onto agar plates. Within 2 hr of enhanced locomotor activity, NMJs showed a significant potentiation of signal transmission that was rapidly reversed by an induced paralysis of the temperature-sensitive mutant parats1. Enhanced locomotor activity was also associated with a significant increase in the number of large subsynaptic translation aggregates. After 4 hr, postsynaptic DGluR-IIA glutamate receptor subunits started to transiently accumulate in ring-shaped areas around synapses, and they condensed later on, after chronic locomotor stimulation at 29°C, into typical postsynaptic patches. These NMJs showed a reduced perisynaptic expression of the cell adhesion molecule Fasciclin II, an increased number of junctional boutons, and significantly more active zones. Such temporal mapping of experience-dependent adaptations at developing wild-type and mutant NMJs will provide detailed insights into the dynamic control of glutamatergic signal transmission (Sigrist, 2003).

One of the prerequisites for a time-resolved analysis of experience-dependent adaptations at Drosophila NMJs was the tight control of larval locomotor activity. Acute enhancement of larval locomotor activity has been achieved by transferring larvae from food vials onto agar plates, a procedure that has been used extensively before as a locomotor reference in the genetic analysis of larval foraging behavior. In addition, larval locomotor activity is persistently modified at different temperatures (18°C, 25°C, and 29°C) that are well within the natural temperature range of Drosophila development. Both paradigms enable the control of larval locomotor activity and therefore allow a time-resolved analysis of experience-dependent adaptations at developing NMJs of Drosophila (Sigrist, 2003).

Three independent lines of evidence suggest that the morphological and physiological changes at NMJs described in this study are triggered by enhanced larval locomotor activity and not caused by temperature treatment or plate transfer itself: (1) the considerable bouton outgrowth seen in wild-type larvae reared at 29°C was significantly suppressed in 29°C reared dglurIIA-ko mutants, which show defective postsynaptic signal transmission, rapid depression of spike train-evoked junctional signal transmission, reduced locomotor activity, and reduced subsynaptic protein synthesis; (2) exposing wild-type and parats1 larvae to permissive 22°C agar plates for 2-3 hr resulted in a significant and similar strengthening of junctional signal transmission in both genotypes -- strengthening was rapidly reversed by induced paralysis in parats1 animals; (3) wild-type larvae reared at 25°C and exposed to 25°C agar plates for up to 18 hr showed significantly enhanced locomotor activity, and they developed more boutons than comparable animals that remained in the food slurry. These experiments show that whenever synaptic signal transmission and larval locomotor activity was compromised, such as in dglurIIA or paralyzed parats1 mutants, the junctional phenotypes were strongly suppressed. It is therefore concluded that the acute and chronic exposure of Drosophila larvae to elevated temperatures or agar plates leads to an enhanced larval locomotor activity, which results initially in reversible physiological changes and later on in molecular and cellular adaptations that ensure persistently enhanced junctional signal transmission and efficient muscle contraction (Sigrist, 2003).

One of the first obvious consequences of enhanced locomotor activity was the fast enhancement of evoked junctional signal transmission, which was already maximal after 2-4 hr of locomotor stimulation on agar plates and was rapidly reversed by paralysis. The observation that the quantal sizes remained unaltered at the indicated time points of locomotor stimulation, whereas evoked junctional responses increased significantly, strongly suggests that the phases of experience-dependent strengthening of Drosophila NMJs are based primarily on an enhanced release of presynaptic vesicles per NMJ (Sigrist, 2003).

Mechanisms that can result in a fast increase in the number of released vesicles include an enhanced presynaptic Ca2+ influx, alterations in the Ca2+ sensitivity of the presynaptic release process, activation of presynaptic metabotropic glutamate receptors, or signaling mediated by the presynaptic G-protein-coupled receptor Methuselah. These mechanisms are typically involved in transient short-term enhancements of synaptic signal transmission. It appears likely that these or similar mechanisms are active during early phases of the experience-dependent junctional strengthening described in this study; however, their exact involvement as well as their temporal regulation remain to be investigated (Sigrist, 2003).

The number of released presynaptic vesicles can also increase at NMJs with a larger number of active release sites. Ultrastructural and morphological analysis of NMJs has revealed that animals that experienced persistently enhanced locomotor activity (rearing at 29°C) develop larger NMJs with an increased total number of T-bar-harboring active zones and an unaltered density of active zones per bouton. Because active zones represent synapses with a high probability of vesicle release, this mechanism could account for the observed strengthening of junctional signal transmission at larger NMJs. In fact, such a typical relationship between the number of active zones and the number of junctional boutons is readily apparent in the consistently observed correlation between junctional transmission strength and NMJ size. These data suggest that the fast-developing NMJs of Drosophila larvae consolidate induced functional changes by recruiting active zones and controlling their density by growing additional boutons. Recent experiments have shown that such NMJs not only transmit single stimuli more efficiently than control NMJs, they also show an enhanced faithfulness in the transmission of high-frequency stimuli (Sigrist, 2003).

It is intriguing to note that the scored electrophysiological parameters were indistinguishable among most locomotor-stimulated animals. This included larvae, which experienced 2-6 hr of locomotor stimulation. NMJs of these larvae showed no detectable bouton outgrowth compared with their controls, suggesting that this early phase of junctional strengthening does not rely on large-scale morphological alterations. It has been shown that dglurIIA-ko mutants are unable to greatly enlarge their NMJs by bouton addition. This mutant mediates enhanced presynaptic vesicle release by increasing the number of active zones; however, these additional active zones are squeezed into a smaller number of preexisting boutons compared with wild type. It is therefore tempting to speculate that within 2 hr of locomotor stimulation NMJs start to increase the number of active zones by de novo synaptogenesis and by rapidly recruiting presynaptic T-bars (dense bodies) onto a large reservoir of preexisting and T-bar-free synapses. In fact, such a fast recruitment of presynaptic dense bodies to synapses has been proposed for synapses in the fly visual system. It is therefore possible that experience-dependent strengthening of junctional signal transmission is mediated primarily by the functional recruitment of additional active zones, which are later distributed in newly grown boutons at their typical density. These processes would leave the efficacy of junctional signal transmission unchanged even during the morphological expansion of NMJs. Unfortunately, because of the current lack of probes that could specifically recognize T-bars or active zones in vivo or in light-microscopic preparations, it was not possible to address these potentially highly dynamic processes at larval NMJs (Sigrist, 2003).

The results show that enhanced locomotor activity results within 2 hr in a rapid and reversible enhancement of evoked junctional signal transmission and in a similarly fast stimulation of local subsynaptic protein synthesis. Although it remains to be investigated whether localized subsynaptic protein synthesis could play an instructive role during these early physiological events, it has been found that the mRNA encoding the glutamate receptor subunit DGluR-IIA is stored within the subsynaptic compartment of NMJs. It therefore represents a likely substrate of localized subsynaptic protein synthesis. It was found that DGluR-IIA-specific immunoreactivity increases visibly after 4 hr of locomotor stimulation, first in the form of ring-shaped accumulations at the edge of preexisting synapses and after chronic stimulation within typical postsynaptic patches. Given that several ionotropic neurotransmitter receptors perform lateral diffusion movements into and out of postsynaptic complexes, it appears likely that the ring-shaped DGluR-IIA accumulations described in this study similarly reflect a transient step in the maturation of postsynapses. Thus, experience-induced subsynaptic protein synthesis seems to instruct the DGluR-IIA-mediated functional maturation of postsynapses, which together with added presynaptic dense bodies mediates the observed increase in the total number of active zones. Finally, NMJs grow more boutons to reestablish the typical active zone density to consolidate the earlier induced physiological alterations. On the basis of this first temporal map of processes contributing to experience-dependent plasticity at Drosophila NMJs, future experiments will incorporate the behavioral assays introduced here to uncover the dynamic control of glutamatergic signal transmission (Sigrist, 2003).

Glutamate receptor dynamics organizing synapse formation in vivo

Insight into how glutamatergic synapses form in vivo is important for understanding developmental and experience-triggered changes of excitatory circuits. Postsynaptic densities (PSDs) expressing a functional, GFP-tagged glutamate receptor subunit (GluR-IIAGFP) were imaged at neuromuscular junctions of Drosophila melanogaster larvae for several days in vivo. New PSDs, associated with functional and structural presynaptic markers, form independently of existing synapses and grow continuously until reaching a stable size within hours. Both in vivo photoactivation and photobleaching experiments show that extrasynaptic receptors derived from diffuse, cell-wide pools preferentially enter growing PSDs. After entering PSDs, receptors are largely immobilized. In comparison, other postsynaptic proteins tested (PSD-95, NCAM and PAK homologs) exchange faster and with no apparent preference for growing synapses. New glutamatergic synapses form de novo and not by partitioning processes from existing synapses, suggesting that the site-specific entry of particular glutamate receptor complexes directly controls the assembly of individual PSDs (Rasse, 2005).

Glutamate receptors localized within the PSD region transmit the excitatory responses in the brain and in many other neuronal systems. Thus, a detailed molecular and cell-biological insight into the formation of glutamatergic synapses is important for understanding of the development of excitatory neuronal circuits and for long-term information storage in the CNS. So far, formation of glutamatergic synapses has been studied mainly in cultivated brain neurons. These studies have been helpful, for example, in delineating a temporal sequence of pre- and postsynaptic assembly and characterizing mechanisms of glutamate receptor trafficking during synapse formation. How glutamatergic synapses assemble in vivo remains to be addressed. It is conceivable that synapse formation is more tightly controlled temporally and spatially in vivo than in vitro, particularly when synapses are added to strengthen already functional circuits. It is thus important to follow 'the entire history' of identified synapses over time in the intact organism while monitoring their molecular dynamics and functional features. Analysis of synapse formation in vivo might profit from the use of synaptic models that are optically and genetically highly accessible. The Drosophila neuromuscular junction (NMJ) is a well established glutamatergic model, widely used for functional genetic descriptions of principle glutamatergic transmission. A mature NMJ comprises a few hundred individual synapses, which are ultrastructurally similar to central mammalian synapses and express glutamate receptor subunits (GluR-IIA, GluR-IIB, GluR-IIC, GluR-IID, GluR-IIE) related to mammalian non-NMDA type glutamate receptors. NMJs comprised of motor neurons and somatic muscles form during late embryonic development in Drosophila. The number of individual synaptic sites per NMJ increases throughout subsequent larval development. Thus, the NMJ should be a suitable model for studying the new formation of glutamatergic synapses within a functional circuit (Rasse, 2005).

Functional, GFP-labeled glutamate receptors (GluR-IIAGFP) have been imaged during the formation of new NMJ synapses in vivo. New small receptor fields form de novo and not by discrete partitioning events. These new small receptor fields correspond to the PSD region of functional synapses, as they are tightly associated with both independent PSD markers and functional (styryl dye labeling) and molecular (active zone components, calcium channels) markers of presynaptic active sites. Small PSDs grow for many hours before finally stabilizing at a mature size. Fluorescence recovery after photobleaching (FRAP) and photoactivation experiments indicate that the incorporation of GluR-IIA-containing glutamate receptor complexes from extrasynaptic pools is directly instructive for PSD formation and growth and, as a result, synapse formation (Rasse, 2005).

This study investigated in fully native settings how glutamatergic synapses form and how glutamate receptor dynamics are organized during this process. The transparent nature of Drosophila larvae make them an ideal subject in which to examine the glutamatergic synapses forming at the developing larval NMJ. To label glutamate receptors for in vivo imaging, enhanced GFP (EGFP) was inserted into the middle of the intracellular C terminus of GluR-IIA. GluR-IIAGFP was then expressed from a genomic transgene. In Western blots probed with antibodies to GFP, GluR-IIAGFP was detected at the predicted 140 kDa in extracts of GluR-IIAGFP transgenic embryos. Next, the subcellular distribution of GluR-IIAGFP was evaluated by immunofluorescent staining of Drosophila NMJs. In such stainings, GluR-IIA is known to label individual PSDs. Likewise, GluR-IIAGFP expression was confined to individual PSDs. GluR-IIAGFP strictly colocalizes with p21/rac1-activated kinase (PAK), an established PSD marker, and with endogenous GluR-IIA. Furthermore, GluR-IIAGFP patches are surrounded by the typical perisynaptic expression of the neural cell adhesion molecule (NCAM) homolog fasciclin II (FasII) In the GluRIIA and GluRIIB double-mutant background, the transgenic expression of GluR-IIA mediated by either GluR-IIAGFP or GluR-IIA is indistinguishable, and it is similar to the level of GluR-IIA expression found at wild-type NMJs. Also, GluR-IIA and GluR-IIAGFP colocalize with GluR-IIC. The GluR-IIC subunit is essential for NMJ neurotransmission, probably by acting as an obligate binding partner of GluR-IIA for forming functional channels. In short, GluR-IIAGFP is expressed at physiological levels, and individual receptor fields correspond to individual PSDs. From this point on, receptor fields identified by means of GluR-IIAGFP will be referred to as PSDs (Rasse, 2005).

Several key findings were made. (1) New small glutamatergic PSDs form separately from existing synapses and then grow to a mature size. Mature PSDs are discrete entities that seem to be stable in size and shape over periods of days. (2) Growing receptor fields invariably associate with a presynaptic active zone during further outgrowth. The coordination between pre- and postsynaptic assembly seems to be very tight. In fact, 99.5% of all GluR-IIA spots older than 10 h were associated with active zone markers, and, vice versa, no active zone accumulations were observed without GluR-IIA accumulation. In contrast, not all younger GluR-IIA spots are associated with a corresponding active zone. It was somewhat surprising, given the studies on cultured mammalian neurons, to find that receptor field/PSD assembly might precede certain aspects of presynaptic active zone assembly and maturation in this system. It will be interesting to address whether glutamatergic synapse types differ in this regard, or whether initial synaptogenesis in culture differs intriniscally from the mode chosen for adding synapses to an already functional circuitry (Rasse, 2005).

(3) New synapses form de novo, whereas split-like redistributions of glutamate receptors from existing synapses into new synapses do not seem to contribute to the formation of new synapses. This is important, because splitting events at the glutamatergic PSDs of mammalian CNS synapses have been postulated but are controversial. (4) Two complementary strategies for in vivo photolabeling of glutamate receptors (FRAP and photoactivation) allow a comparison of glutamate receptor dynamics at growing and stable PSDs. Both approaches show that GluR-IIA preferentially enters growing PSDs from diffuse extrasynaptic pools. This entry is directly correlated with PSD growth. Once glutamatergic PSDs have reached a certain size, they stabilize, and their glutamate receptor population becomes largely immobilized. In contrast, other postsynaptic transmembrane proteins (for example, FasII) and scaffolding proteins such as the SAP-97/PSD-95 homolog Dlg show equally high dynamics over all synapses. Previously, genetic experiments showed that more synapses form per NMJ when the level of GluR-IIA expression is increased. Vice versa, the reduction of the GluR-IIA level by one gene dose inhibits the NMJ from producing additional synapses when genetically or behaviorally challenged. Thus, the entry of glutamate receptors into PSDs (but not the entry of the other postsynaptic proteins investigated) might directly control the growth of the postsynaptic specialization and thereby the growth of synapses. Notably, local translation of GluR-IIA has been suggested to promote activity-dependent synapse formation in this system. It will be interesting to further test the role of GluR-IIA as a potential organizer of postsynaptic assembly by combining in vivo imaging and functional genetics at the Drosophila NMJ. However, it is well established in the mammalian brain that entry of specific AMPA-type glutamate receptor complexes into preformed synapses can control synapse efficacy over shorter time periods. It will be interesting to investigate whether glutamate receptor entry can control the formation of mammalian glutamatergic synapses under certain circumstances (Rasse, 2005).

In addressing the origin of receptors integrated into growing synapses, no evidence was obtained for internal stores of glutamate receptors. It will thus be interesting to determine the local cues and signals at PSDs that specifically control GluR-IIA entry and immobilization during PSD growth. It is suspected that GluR-IIA populations residing in the extrasynaptic muscle plasma membrane support PSD growth. Electrophysiology has, in fact, demonstrated the existence of glutamate receptors in the extrasynaptic membrane of Drosophila muscles. Such receptors might diffuse laterally in the membrane partitioning in and out of synapses, where they likely have a low residence time. This pool of unbound receptors might be fundamentally different from a second pool of receptors, which is 'trapped' in PSDs, as demonstrated recently by tracking individual glutamate receptor complexes in cultured mammalian neurons (Rasse, 2005).

During the few days of larval development, the synaptic current per Drosophila NMJ increases by nearly two orders of magnitude to keep pace with the growing postsynaptic muscle cell. The data imply that GluRIIA-containing glutamate receptors seem to become particularly 'invested' the formation and subsequent outgrowth of new synapses, but less so in the enlargement of preexisting synapses. At average NMJ synapses, it is likely that glutamate receptors are not saturated for glutamate during vesicle exocytosis. It is speculated that at small, newly forming synapses, a higher proportion of glutamate receptors might be activated during glutamate release because of the smaller physical distance from the exact position of the presynaptic release site. To achieve an increase in the overall NMJ current, it therefore might be more efficient to insert new receptors into these small synapses than to insert them into the large receptor fields of mature size synapses, which are unlikely to be saturated upon presynaptic stimulation. Thereby, the 'drive' of the Drosophila NMJ toward increasing overall transmission strength might be the deeper reason behind strongly stabilizing receptors once they are integrated (Rasse, 2005).

Examining the dynamic molecular composition of glutamatergic synapses in more detail by further applying both in vivo imaging and the powerful genetics of Drosophila should help clarify principal rules for assembly and remodeling of glutamatergic synapses. Notably, at the Drosophila NMJ, activity-dependent formation of additional synapses has been described on the basis of experience-dependent and genetically evoked processes (Rasse, 2005).

Non-NMDA-type glutamate receptors are essential for maturation but not for initial assembly of synapses at Drosophila neuromuscular junctions

The assembly of glutamatergic postsynaptic densities (PSDs) seems to involve the gradual recruitment of molecular components from diffuse cellular pools. Whether the glutamate receptors themselves are needed to instruct the structural and molecular assembly of the PSD has hardly been addressed. This study engineered Drosophila neuromuscular junctions (NMJs) to express none or only drastically reduced amounts of their postsynaptic non-NMDA-type glutamate receptors. At such NMJs, principal synapse formation proceeded and presynaptic active zones showed normal composition and ultrastructure as well as proper glutamate release. At the postsynaptic site, initial steps of molecular and structural assembly took place as well. However, growth of the nascent PSDs to mature size was inhibited, and proteins normally excluded from PSD membranes remained at these apparently immature sites. Intriguingly, synaptic transmission as well as glutamate binding to glutamate receptors appeared dispensable for synapse maturation. Thus, these data suggest that incorporation of non-NMDA-type glutamate receptors and likely their protein-protein interactions with additional PSD components triggers a conversion from an initial to a mature stage of PSD assembly (Schmid, 2006).

A detailed molecular and cell biological insight into the formation of glutamatergic synapses is important for understanding the development of excitatory neuronal circuits and also the process of long-term information storage in the CNS. So far, studies on cultivated brain neurons analyzed mechanisms of glutamate receptor trafficking during synapse formation and have suggested a temporal sequence of presynaptic and postsynaptic assembly. However, whether in turn the process of incorporating glutamate receptors is needed for the establishment of synaptic structures has hardly been addressed (Schmid, 2006).

The relationship between neurotransmitter receptor incorporation and synapse assembly was addressed by genetically reducing or eliminating the expression of all neurotransmitter receptors at a certain synapse type. Consequences of eliminating all postsynaptic glutamate receptors expressed at a specific glutamatergic synapse has so far not been described. This study showed that a lack of glutamate receptors provoked a specific block in the molecular and ultrastructural maturation of PSDs (Schmid, 2006).

Notably, loss of transmission concomitant with losing glutamate receptor complexes seemed not involved, based on the fact that neither blocking synaptic transmission nor affecting glutamate binding by site-directed mutagenesis did provoke similar defects. Thus, consistent with studies in other synaptic systems, ionic transmission through the postsynaptic neurotransmitter receptors does not appear essential for principal synapse assembly. Instead, the data clearly imply that a critical level of glutamate receptor protein is needed to allow synapse maturation (Schmid, 2006).

A model for the maturation of individual NMJ synapses in either the presence or absence of postsynaptic glutamate receptors is presented in this study. At glutamate receptor-deprived synapses, synaptic vesicles appeared normally distributed, and their activity-mediated release appeared increased, likely as part of a compensation for reduced postsynaptic sensitivity. Moreover, functional active zones with presynaptic dense bodies still formed. Thus, active zones still assemble when the mature organization of synaptic membranes ('tight planar apposition') is not established. Consistently, previous work had shown that the formation of presynaptic dense bodies persisted even after genetic elimination of postsynaptic muscle cells. In contrast, active zone formation is severely affected in bruchpilot mutants, whereas the presynaptic and postsynaptic membranes remain tightly apposed (Schmid, 2006).

At developing NMJs, newly forming 'nascent' PSDs are characterized by small GluRIIA accumulations strictly colocalized with PAK kinase. Even in the complete absence of glutamate receptors (gluRIIC single or gluRIIA&IIB double mutant), postsynaptic PAK patches, as typical for small nascent synapses, still formed, indicating that principal cues for the definition of postsynaptic membrane patches persisted in this situation. However, these PAK patches consistently failed to reach mature size. PAK, which mediates effects of Rho-GEF dPIX has been implicated in postsynaptic maturation, with PAK mutants showing a partial depletion of GluRIIA, and reduced SSR formation. However, neither pak nor dpix mutants have so far been reported to show defects in synaptic membrane apposition. Thus, postsynaptic differentiation is not completely blocked in the absence of glutamate receptors. Instead, two postsynaptic 'assembly modules' (PAK/dPIX signaling and glutamate receptor localization) appear only partly dependent on each other, with glutamate receptor localization being essential for PSD maturation but not for initial PSD assembly (Schmid, 2006).

At the cholinergic mouse NMJ, genetic deletion of the adult acetylcholine receptor subunit leads to severely reduced AChR density. Notably, a profound reorganization of AChR-associated components of the postsynaptic membrane and cytoskeleton has been observed in this situation (Schmid, 2006).

Synaptic membranes are electron dense and apposed to each other, leaving a cleft of consistent width, likely essential for robust timing and efficacy of neurotransmission. In contrast, perisynaptic membranes are less electron dense and tend to undulate. At NMJs lacking glutamate receptors, FasII/Dlg complexes ectopically remained at synaptic sites and membranes now appeared undulated, indicating perisynaptic type of membrane adhesion. Thus, glutamate receptors seem essential to establish the type of membrane adhesion found at the synapse, whereas usually perisynaptic adhesion molecules as FasII mediate a qualitatively different type of membrane adhesion. Notably, undulation of perisynaptic membranes was impaired at NMJs lacking glutamate receptors leading to a less developed SSR. Moreover, boutons often appeared atypically round, further indicating that membrane–membrane adhesion is fundamentally affected at NMJ terminals lacking glutamate receptors (Schmid, 2006).

Several classes of bona fide cell adhesion molecules (CAMs) have been implicated in mediating membrane adhesion at synapses, particularly transsynaptic neurexin–neuroligin pairs and cadherins. The specific contributions of these synaptic CAMs during initial synapse assembly and maturation are under intense investigation. The data are consistent with the idea that the C-terminal, intracellular domains of glutamate receptors might engage in interactions with other PSD components, which in turn cluster postsynaptic CAM-type membrane proteins. These would then mediate interactions to cluster presynaptic CAMs or bind components of the extracellular matrix to allow synaptic membrane apposition. Alternatively, direct interactions of glutamate receptors with other membrane protein complexes, as recently demonstrated for Stargazins/TARPs (transmembrane AMPA receptor regulatory proteins), might be involved (Schmid, 2006).

No CAM single mutant has so far been reported to provoke a defect in synaptic membrane apposition as severe as the one observed in this study for glutamate receptor mutant situations. Thus, multivalent interactions of the heterotetrameric glutamate receptor complexes as well as the redundant involvement of several CAM species might occur (Schmid, 2006).

PAK labeling suggested that initial steps in defining postsynaptic membranes persisted even in the total absence of glutamate receptors, whereas these initial assemblies could not mature on the ultrastructural level when glutamate receptors were lacking. in vivo imaging has been achieved of photolabeled GluRIIA at the developing NMJ, and it was found that newly forming PSDs in fact grow by a continuous incorporation of glutamate receptors, whereby the accumulation of presynaptic active zone material (BRP) appeared slightly delayed. Thereby, the entry of GluRIIA, likely derived from cell-wide plasma membrane pools via lateral diffusion, directly correlated with PSD growth. Once glutamatergic PSDs reached a certain size, they stabilized and GluRIIA was essentially immobilized. In comparison, other postsynaptic proteins showed high turnover equally over all synapses. This slow turnover of glutamate receptors is consistent with the view that multiple interactions of glutamate receptor set the core of a transsynaptic interaction matrix. Several lines of genetic and experience-dependent manipulations point toward a rate-limiting role of GluRIIA levels in NMJ synapse formation. In summary, the available data suggest that incorporation of glutamate receptors might be a key event to allow additional expansion of initial postsynaptic assemblies, finally leading to mature PSDs. Thereby, the overall level of glutamate receptors available in the muscle membrane might control the total number of synapses forming per NMJ (Schmid, 2006).

Understanding the plasticity processes taking place at glutamatergic synapses has been a focus of attention within cellular neuroscience. Hereby, rapid changes in synaptic receptor number were reported to mediate plastic changes of synaptic transmission, often on the timescale of tens of minutes in mammalian preparations. Notably, however, a recent study indicated that the cycling of synaptic glutamate receptors needed 16 h or more. Similar timing was observed for nicotinic acetylcholine and GABA receptors. Thus, parts of the synaptic glutamate receptor population might be needed to reside stably within the PSD to maintain synapse stability. In fact, only severe receptor deprivation interfered with proper postsynaptic assembly at the NMJ, suggesting that the glutamate receptor level should not fall below a certain critical threshold (Schmid, 2006).

Notably, the extracellular domain of the mammalian AMPA receptor subunit GluR2 has been shown to increase the size and density of spines in hippocampal neurons, and to induce spine formation in GABAergic interneurons normally lacking spines. It will be interesting to see whether these structural roles of glutamate receptors have a common mechanistic denominator (Schmid, 2006).

Different types of synapses differ strongly in the ultrastructural detail of their postsynaptic specializations. Thus, a typical brain neuron, acting as a postsynaptic partner for different types of presynaptic inputs, has to establish and maintain different postsynaptic architectures, suggesting the existence of 'identity molecules' allowing the self-assembly of such architectures, and potentially a match with membrane cues of the presynaptic partner cell. Obvious candidates for such molecules are the postsynaptic neurotransmitter receptors themselves. This study is consistent with such a view (Schmid, 2006).

Synaptic homeostasis is consolidated by the cell fate gene gooseberry, a Drosophila pax3/7 homolog

In a large-scale screening effort, the gene gooseberry (gsb) was identified as being necessary for synaptic homeostasis at the Drosophila neuromuscular junction. The gsb gene encodes a pair-rule transcription factor that participates in embryonic neuronal cell fate specification. This study defines a new postembryonic role for gooseberry. gsb becomes widely expressed in the postembryonic CNS, including within mature motoneurons. Loss of gsb does not alter neuromuscular growth, morphology, or the distribution of essential synaptic proteins. However, gsb function is required postembryonically for the sustained expression of synaptic homeostasis. In GluRIIA mutant animals, miniature EPSP (mEPSP) amplitudes are significantly decreased, and there is a compensatory homeostatic increase in presynaptic release that restores normal muscle excitation. Loss of gsb significantly impairs the homeostatic increase in presynaptic release in the GluRIIA mutant. Interestingly, gsb is not required for the rapid induction of synaptic homeostasis. Furthermore, gsb seems to be specifically involved in the mechanisms responsible for a homeostatic increase in presynaptic release, since it is not required for the homeostatic decrease in presynaptic release observed following an increase in mEPSP amplitude. Finally, Gsb has been shown to antagonize Wingless signaling during embryonic fate specification, and initial evidence is presented that this activity is conserved during synaptic homeostasis. Thus, gsb was identified as a gene that distinguishes between rapid induction versus sustained expression of synaptic homeostasis and distinguishes between the mechanisms responsible for homeostatic increase versus decrease in synaptic vesicle release (Marie, 2010).

This study has advanced understanding of synaptic homeostasis in several important ways. First, gsb was identified as required in postmitotic, postembryonic neurons for synaptic homeostasis at the Drosophila NMJ. Drosophila gsb is the homolog of vertebrate pax3/pax7. Thus, these data identify a new function for a conserved gene family that has been traditionally studied in the context of neuronal fate specification. Second, it was demonstrated that loss of gsb selectively disrupts the expression of synaptic homeostasis without impairing the rapid induction of synaptic homeostasis. These data suggest that genetically separable mechanisms exist for the induction versus the expression of synaptic homeostasis at the Drosophila NMJ. Third, it was demonstrated that loss of Gsb selectively disrupts the mechanisms responsible for a homeostatic increase in presynaptic release without impairing the mechanisms responsible for homeostatic decrease in presynaptic release. Therefore, these two forms of homeostatic modulation, both expressed at the Drosophila NMJ, appear to involve genetically separable mechanisms. Fourth, because gsb is a transcription factor, these data highlight the possibility that the persistent expression of synaptic homeostasis in the GluRIIA mutant is consolidated through transcription-/translation-dependent mechanisms, while the rapid induction of homeostatic signaling is independent of new protein synthesis. Thus, there may exist genetically separable phases of homeostatic signaling at the Drosophila NMJ analogous to the induction versus expression of long-term synaptic plasticity in other systems. Finally, evidence is presented to support the hypothesis that Gsb functions similarly during cell fate specification and synaptic homeostasis. According to the emerging model, Gsb may antagonize Wingless signaling in motoneurons to facilitate the consolidation of synaptic homeostasis. While there remains considerable work to prove this model, the data, in combination with prior work during embryonic cell fate specification, provide the basis for a compelling model that can be examined in greater detail in future studies (Marie, 2010).

In this study, motoneurons with decreased levels of Gsb are unable to express synaptic homeostasis in the background of a GluRIIA mutation. By contrast, the rapid, protein synthesis-independent induction of synaptic homeostasis following application of the glutamate receptor antagonist PhTx is normal. One possible explanation for this difference is that PhTx and the GluRIIA mutant cause different postsynaptic perturbations and initiate separate homeostatic signaling systems, only one of which is affected by loss of gsb. This seems unlikely, however, because previously published data indicate that PhTx primarily acts upon postsynaptic glutamate receptors including those that contain the GluRIIA receptor subunit. Furthermore, several mutations have been shown to block synaptic homeostasis both in the GluRIIA mutant and following PhTx application, demonstrating that these two perturbations share, at some level, a common molecular mechanism of homeostatic signaling. It is hypothesized, therefore, that loss of Gsb impairs a molecular process that is selectively involved in the sustained expression of synaptic homeostasis. This would be consistent with the sustained expression of synaptic homeostasis requiring new protein synthesis (Marie, 2010).

The possibility that Gsb participates specifically in the sustained expression or consolidation of synaptic homeostasis has interesting implications. In one model of homeostatic signaling, the GluRIIA mutation represents a persistent stress that induces a continuous, rapidly induced form of homeostatic compensation. In this model, the homeostatic modulation of presynaptic release is continually updated and never consolidated. An alternative model is that, once induced, the homeostatic modulation of presynaptic release is consolidated and maintained for prolonged periods of time. If this is the case, it should be possible to selectively disrupt the consolidation of synaptic homeostasis independently of the mechanisms of induction. Loss of Gsb appears to do just this. It is not possible to persistently inhibit protein synthesis during larval development. However, the demonstration that decreased Gsb disrupts synaptic homeostasis in the GluRIIA mutant suggests that transcription and translation may be involved in the mechanisms that consolidate a homeostatic change in presynaptic release (Marie, 2010).

In Drosophila, like in vertebrates, combinations of transcriptional regulators determine the fate of neurons. Indeed, transcription factors control all stages of early neuronal development and neuronal circuit formation, from the direction in which the axon initially extends from the neuronal cell body, the location of the terminal zone of the axonal arborization, and the specificity of synaptic targeting to the choice of neurotransmitter. More recently, some evidence suggests that expression of the transcription factor evenskipped during early embryogenesis could affect a motoneuron's complement of ion channels and neuron excitability. However very little is known regarding the role of transcriptional regulators in mature neurons. Recently, it was demonstrated that mild perturbation of the engrailed gene lead to mice with an adult phenotype that resembles key pathological features of Parkinson's disease. In this study, the expression of Gsb-RNAi using a postmitotic neuronal GAL4 driver leads to the conclusion that Gsb has a postmitotic activity that is essential to the maintenance of synaptic homeostasis. These data provide evidence that the transcription factors involved in embryonic development may have potent postembryonic functions that are necessary for the maintenance of stable neural function (Marie, 2010).

How do embryonic transcriptional regulators influence the expression of synaptic homeostasis? This study presents data to support a model in which Gsb function is conserved during embryonic patterning and synaptic homeostasis. Specifically, decreased wg levels rescue synaptic homeostasis in GluRIIA; gsb/+ double mutant animals. According to this model, Wg antagonizes the expression or consolidation of synaptic homeostasis, providing new insight into the activity of this potent intercellular, synaptic signaling molecule. It could be important, for example, to suppress homeostatic signaling during anatomical synaptic plasticity, a process in which Wg has been implicated. These data are strengthened by two observations. First, these data are supported by the well established embryonic activity of Gsb. Second, since partial loss of wg rescues the homeostatic defect in GluRIIA; gsb/+ double mutant animals, it is unlikely that this represents a nonspecific genetic interaction. Many additional experiments will be necessary to prove the function of wg as an antagonist of synaptic homeostasis. The data, however, take this model beyond the stage of pure speculation and suggest that this will be an important avenue of future experimental investigation (Marie, 2010).

Distinct presynaptic and postsynaptic dismantling processes of Drosophila neuromuscular junctions during metamorphosis

Synapse remodeling is a widespread and fundamental process that underlies the formation of neuronal circuitry during development and in adaptation to physiological and/or environmental changes. However, the mechanisms of synapse remodeling are poorly understood. Synapses at the neuromuscular junction (NMJ) in Drosophila larvae undergo dramatic and extensive remodeling during metamorphosis to generate adult-specific synapses. To explore the molecular and cellular processes of synapse elimination, confocal microscopy, live imaging, and electron microscopy (EM) of NMJ synapses were performed during the early stages of metamorphosis in Drosophila in which the expressions of selected genes were genetically altered. It is reported that the localization of the postsynaptic scaffold protein Disc large (Dlg) becomes diffuse first and then undetectable, as larval muscles undergo histolysis, whereas presynaptic vesicles aggregate and are retrogradely transported along axons in synchrony with the formation of filopodia-like structures along NMJ elaborations and retraction of the presynaptic plasma membrane. EM revealed that the postsynaptic subsynaptic reticulum vacuolizes in the early stages of synapse dismantling concomitant with diffuse localization of Dlg. Ecdysone is the major hormone that drives metamorphosis. Blockade of the ecdysone signaling specifically in presynaptic neurons by expression of a dominant-negative form of ecdysone receptors delayed presynaptic but not postsynaptic dismantling. However, inhibition of ecdysone signaling, as well as ubiquitination pathway or apoptosis specifically in postsynaptic muscles, arrested both presynaptic and postsynaptic dismantling. These results demonstrate that presynaptic and postsynaptic dismantling takes place through different mechanisms and that the postsynaptic side plays an instructive role in synapse dismantling (Liu, 2010).

The Drosophila NMJ is an attractive model system for studying synaptogenesis but has only rarely been exploited to study synapse elimination. The present study used the Drosophila NMJ to study synapse elimination in the early stages of metamorphosis during which extensive synapse elimination occurs. This study unveiled distinct presynaptic and postsynaptic dismantling processes. Presynaptic elimination is characterized by the formation of prominent filopodial structures. The presynaptic membrane then retracts toward the nerve-muscle contact site with decreased bouton number and enlarged bouton size, accompanied by SV aggregation and retrograde axonal transport of SVs. It is worth pointing out that the precise timing of synapse dismantling revealed by immunostaining and live imaging is different, and this is probably attributable to the fact that the samples were analyzed under different conditions. For example, animals were kept at 25°C for immunostaining but were maintained at 20°C during live imaging. It is well known that filopodia are present in growth cones and play an important role in neurite outgrowth. The filopodia-like structures observed during synapse elimination presumably sense and explore the environment. The data also demonstrate that retrograde axonal transport plays an important role in presynaptic elimination. It is expected that the retrogradely transported synaptic constituents are reused to form adult-specific synaptic connections, although the final fate of the motor neuron MN4a innervating muscle 4 has not been determined. During the metamorphic period from 4 to 11 h APF we examined for the complete synapse elimination, neither disrupted synaptic microtubules were observed as reported for the local synapse disassembly in Drosophila larvae nor was an axonal 'retraction bulb' as seen during mammalian NMJ synapse elimination; a retraction bulb appears when a presynaptic terminal is detached mostly or completely from the postsynaptic specialization (Liu, 2010).

The first signs of postsynaptic elimination were the blurred and diffuse localization of postsynaptic markers Dlg and CD8-GFP-Shaker at 4 h APF, followed by a more expanded distribution of GluRs and vacuolization of SSR in 6 h APF. The postsynaptic components of Dlg and CD8-GFP-Shaker were almost completely eliminated at 9 h APF. Completion of the synapse dismantling process starting from the diffusion of postsynaptic Dlg at 4 h APF takes ~7 h. It is remarkable to note that the patterns of elimination of postsynaptic Dlg and GluRs are different; Dlg is eliminated by a diffusion-degradation process, whereas no diffusion of GluRs was observed before degradation, indicating that they are eliminated by different mechanisms. The differential elimination of Dlg and GluRs is consistent with the previous finding that the synaptic localization of GluR IIA is independent of Dlg. Interestingly, as for Drosophila GluRs, mammalian NMJ postsynaptic acetylcholine receptors are eliminated without the intermediate process of diffusion (Liu, 2010).

Synapse elimination is distinct from axonal and dendritic pruning. In Drosophila, the pruning of axonal and dendritic processes during metamorphosis closely resembles the pathological process of Wallerian degeneration, a process in which part of the axon separated from the nucleus of the neuron degenerates. In both axonal pruning of central mushroom body neurons and dendritic elimination of peripheral sensory neurons, severing of neuronal processes is preceded by microtubule depolymerization and followed by cytoplasmic blebbing and degeneration. It is conceivable that synapse elimination and pruning of neuronal processes are closely interconnected, but the time course of the two discrete processes has yet to be determined. It has been shown that specific E2/E3 ubiquinating enzymes and caspases (i.e., UbcD1-Diap1-Dronc) are involved in dendritic pruning. But it is unknown whether those molecules also participate in NMJ synapse elimination. However, this study has demonstrated that disruption of ubiquitination and apoptosis pathways on the postsynaptic side arrests synaptic elimination. It will be of great interest to identify the specific ubiquinating enzymes and caspases that participate in synapse dismantling during metamorphosis. It is worth noting that glial cells play an important role in axonal pruning of mushroom body neurons and olfactory receptor neurons. This study found no evidence to suggest that glial cells play a role in NMJ synapse elimination (Liu, 2010).

Synapse disassembly or instability in Drosophila NMJ terminals have been described. However, this process is fundamentally different from that reported in this paper in several aspects. First, local synapse disassembly, the disassembly of distal synaptic boutons or a branch of the whole synaptic terminals of a motor neuron, a process that occurs during synapse growth in larval development was examined in previous papers, whereas this paper studied the elimination of complete NMJ 4 synapses in synchrony with muscle histolysis during metamorphosis. Also, the synapse elimination reported in this study is different from that of mammalian NMJ synapses; the former involves muscle destruction, whereas the latter does not. Second, the processes of local versus general synapse disassembly are different. In local synapse disassembly, presynaptic dismantling precedes postsynaptic dismantling: the presynaptic microtubule cytoskeleton retracts first, followed by the elimination of synaptic release machinery (i.e., the vesicle-associated protein synapsin) and ultimately the disassembly of the postsynaptic apparatus including the postsynaptic GluRs and the scaffold Dlg. However, in the elimination of complete synaptic terminals during metamorphosis, postsynaptic dismantling starts first, followed by presynaptic dismantling. Third, disrupting the dynactin complex destabilizes local synapses, leading to more synaptic 'footprints' that are defined as the withdrawal of presynaptic components from clearly defined postsynaptic specialization containing Dlg was studied previously. It was argued that the dynactin complex functions locally within presynaptic terminals to maintain synapse stability. Although the current study not examined synaptic footprints during metamorphosis, this study reports that disrupting the dynactin complex in presynaptic neurons delays presynaptic dismantling specifically, whereas the postsynaptic components disassemble normally, indicating that dynactin-mediated retrograde axonal transport is required for presynaptic elimination. These results demonstrate that the dynactin complex functions differently in distinct cellular contexts (Liu, 2010).

Two independent lines of evidence indicate that the postsynaptic rather than presynaptic side plays an instructive role in synapse elimination during metamorphosis. First, immunostaining showed that postsynaptic dismantling precedes presynaptic elimination by ~1 h. Second, blockade of retrograde axonal transport and ecdysone signaling specifically in presynaptic neurons delayed presynaptic dismantling only but postsynaptic dismantling proceeded normally. However, inactivation of ecdysone signaling, ubiquitination, or apoptosis pathways in postsynaptic muscles arrested both presynaptic and postsynaptic dismantling. It is noted that synapse elimination is closely correlated with muscle destruction. Indeed, muscle histolysis might be the primary cause of NMJ synapse elimination during metamorphosis. These results together indicate that postsynaptic elimination is independent of presynaptic elimination, but presynaptic elimination depends on postsynaptic elimination; in other words, postsynaptic elimination triggers presynaptic elimination. This hypothesis is supported by a previous report that local ecdysone treatment of the hawkmoth, Manduca sexta, to induce local muscle degeneration results in loss of synaptic contacts in the treated region, whereas neighboring NMJ synapses remain intact. The current data are also consistent with mounting evidence from mammalian studies supporting a major role for the postsynaptic side in synapse elimination. However, the downstream targets of the ubiquitination, apoptosis, or ecdysone pathways in the postsynaptic muscles that are crucial for initiating presynaptic dismantling are currently unknown. It will be of great interest to identify these targets (Liu, 2010).

Neuroligin 2 is required for synapse development and function at the Drosophila neuromuscular junction

Neuroligins belong to a highly conserved family of cell adhesion molecules that have been implicated in synapse formation and function. However, the precise in vivo roles of Neuroligins remain unclear. This study has analyzed the function of Drosophila neuroligin 2 (dnl2) in synaptic development and function. dnl2 is strongly expressed in the embryonic and larval CNS and at the larval neuromuscular junction (NMJ). dnl2 null mutants are viable but display numerous structural defects at the NMJ, including reduced axonal branching and fewer synaptic boutons. dnl2 mutants also show an increase in the number of active zones per bouton but a decrease in the thickness of the subsynaptic reticulum and length of postsynaptic densities. dnl2 mutants also exhibit a decrease in the total glutamate receptor density and a shift in the subunit composition of glutamate receptors in favor of GluRIIA complexes. In addition to the observed defects in synaptic morphology, it was also found that dnl2 mutants show increased transmitter release and altered kinetics of stimulus-evoked transmitter release. Importantly, the defects in presynaptic structure, receptor density, and synaptic transmission can be rescued by postsynaptic expression of dnl2. Finally, this study shows that dnl2 colocalizes and binds to Drosophila Neurexin (dnrx) in vivo. However, whereas homozygous mutants for either dnl2 or dnrx are viable, double mutants are lethal and display more severe defects in synaptic morphology. Altogether, these data show that, although dnl2 is not absolutely required for synaptogenesis, it is required postsynaptically for synapse maturation and function (Sun, 2011).

Analysis of the Drosophila melanogaster genome revealed the presence of four neuroligin-like genes (CG13772, CG34127, CG34139, and CG31146). All four of the putative neuroligin genes encode proteins that share significant amino acid similarity with vertebrate Neuroligin and a similar predicted protein structure; however, based on protein sequence alignments, the four Drosophila neuroligins and mammalian neuroligins evolved from a common ancestor. As such, it is not possible to draw a direct correlation between any one Drosophila neuroligin and the mammalian neuroligins. A full-length cDNA clone has been identified from a Drosophila brain cDNA library that corresponds to the CG13772 gene, which was submitted to flybase as dneuroligin (dnl). A recent unbiased screen for genes affecting NMJ structure, however, identified CG31146 and named that homolog Drosophila neuroligin 1 (Banovic, 2010). The present study examines the role of the CG13772 homolog originally named dneuroligin, which is now term Drosophila neuroligin 2 (dnl2) to avoid confusion. The dnl2 gene is located on the left arm of the second chromosome at cytological position 27C3-4 and is composed of 13 exons and 12 introns. dnl2 encodes a 1248 amino acid long protein with a predicted molecular weight of 137 kDa. Similar to vertebrate neuroligins, Dnl2 is also predicted to comprise three distinct regions: an N-terminal extracellular acetylcholinesterase-like domain, a single transmembrane region, and a C-terminal cytoplasmic region with a conserved PDZ binding motif. The extracellular domain also contains several putative N-glycosylation sites and a serine/ threonine-rich region for potential O-glycosylation between the acetylcholinesterase-like domain and the transmembrane domain that may affect neurexin binding (Sun, 2011).

Neurexins and Neuroligins are highly conserved cell adhesion molecules that form an asymmetric, trans-synaptic complex required for synapse formation (Sudhof, 2008). The present study examined a homolog of neuroligin (dnl2) in Drosophila expressed at NMJ synapses and in the CNS. dnl2 null mutants are viable and exhibit numerous defects in synaptic morphology and function. Presynaptically, a reduction was observed in axonal branching and fewer synaptic boutons, although the number of active zones per bouton was increased. Postsynaptically, dnl2 mutants exhibit a decrease in GluR density and a shift in the ratio of GluRIIA to GluRIIB receptor complexes in favor of GluRIIA complexes. dnl2 mutants also showed a decrease in complexity of the subsynaptic reticulum. Both presynaptic and postsynaptic defects observed in dnl2 mutants could be recapitulated by knockdown of dnl2 in muscle, and, more importantly, the defects can be rescued by postsynaptic expression of a wild-type dnl2 transgene. Functionally, dnl2 mutants showed an increase in transmitter release and a decrease in paired-pulse plasticity indicative of an increase in transmitter release probability. It is also possible that changes in the active (voltage-gated) properties of the postsynaptic membrane may contribute to the increased amplitude of EJPs. Indeed, the changes in the kinetics of EJPs with little or no change in the kinetics of mEJPs may be indicative of altered membrane conductance. Together, these data indicate that, although dnl2 is not absolutely required for synaptogenesis, it does play an important role in the postsynaptic cell in synapse maturation and function (Sun, 2011).

dnl2 mutants showed several defects in postsynaptic architecture. In vertebrates, neuroligins are thought to regulate postsynaptic organization via a direct interaction with PSD-95 (Irie, 1997). In Drosophila, the homolog of PSD-95, dlg, has been shown to be required for several aspects of postsynaptic organization. Furthermore, the defects in postsynaptic architecture in dnl2 mutants are reminiscent of a defect in dlg function. Loss of dnl2 does not appear to prevent or impair clustering of dlg at postsynaptic sites because the overall dlg levels was not different in dnl2 mutants. Rather, dnl2 may be required for proper dlg signaling (Sun, 2011).

dnl2 mutants also showed several defects in presynaptic architecture mediated by trans-synaptic interactions. The most likely candidate for a trans-synaptic signaling partner is neurexin. Both dnl2 and dnrx null flies are viable but display significant reductions in the number of synaptic boutons. Strong colocalization of dnl2 and dnrx is observed in the CNS and the NMJ, and a complex was detected between the two proteins in vivo. dnl2;dnrx double mutants, however, are lethal and display more severe phenotypes than those observed in single mutants, implying that dnl2 and dnrx may interact with additional partners during synaptic development. For example, dnrx can also interact with other neuroligins in Drosophila to mediate synaptic development. Banovic (2010) found that loss of dnrx did not enhance the morphological defects in dnl1 mutants, suggesting that both genes function within a common pathway. Furthermore, a point mutation in dnl1 that is predicted to abolish binding to dnrx suppressed the phenotype associated with overexpression of dnl1 (Banovic, 2010). There are also two other predicted homologs of neuroligin in flies, but the function of these genes remains to be determined. Neuroligins and neurexins may also form additional complexes with other proteins involved in synaptogenesis. A recent study found that neuroligin was able to induce increases in synaptic density independently of neurexin binding. Similarly, two other recent studies showed that leucine-rich repeat transmembrane proteins can induce presynaptic differentiation when bound to neurexin, providing a novel trans-synaptic neurexin-dependent mechanism for development of presynaptic specializations (Sun, 2011).

If the presynaptic defects observed in dnl2 mutants are not mediated via an interaction with neurexin, the question remains, how does dnl2 affect presynaptic morphology? One possibility is that these changes occur indirectly. The level of postsynaptic GluRIIA expression is correlated with changes in the number of presynaptic T-bars. It is possible that GluRIIA expression is regulated in part by dnl2, and increased GluRIIA expression in dnl2 mutants is responsible for the increased density of T-bars. Normally, the density of T-bars in individual boutons is held constant via homeostatic changes in the expression of the cell adhesion molecule Fas II. In the present study, however, an increase was observed in the number of T-bars per bouton. Moreover, a decrease was seen in the total number of boutons rather than an increase as might be expected based on previous studies. Because the addition of synaptic boutons during NMJ growth requires the downregulation of Fas II expression, the uncoupling between T-bar numbers and bouton expansion in dnl2 mutants may suggest that dnl2 is involved in the GluRIIA-mediated downregulation of Fas II (Sun, 2011).

The Drosophila genome is predicted to have four neuroligin homologs and a single neurexin homolog. Whether all four Drosophila neuroligins are required for synapse development is presently unknown. To date, the only other neuroligin that has been studied in Drosophila is dnl1 (Banovic, 2010). Neither dnl1 nor dnl2 are required for synaptogenesis, yet both genes play a role in the development/maturation of the NMJ, suggesting that the two genes may be functionally redundant. Both proteins are expressed at the NMJ in wild-type animals, and null mutations in either gene lead to significantly reduced bouton numbers, defects in GluR organization, and alterations in the complexity of the subsynaptic reticulum (Sun, 2011).

Despite these similarities, however, there are also a number of differences between dnl1 and dnl2 null mutants. First, dnl2 is expressed in both the CNS and muscles, whereas dnl1 is only expressed in muscle. Second, dnl1 mutants showed a complete loss of GluR expression in ~10% of boutons (Banovic, 2010), whereas dnl2 showed a uniform decrease in total GluR expression in all boutons and an increased abundance of GluRIIA receptor complexes at the expense of GluRIIB complexes. Third, dnl1 mutants showed a decrease in transmitter release (Banovic, 2010), whereas dnl2 mutants showed an increase in transmitter release. Although both mutants showed a decrease in bouton number, dnl2 mutants also showed a significant increase in the number of active zones. As such, the differences in the amplitude of transmitter release observed in dnl1 and dnl2 mutants likely reflect the different presynaptic morphologies of the two mutants (Sun, 2011). dnrx). dnl1 shows very little colocalization with dnrx and none outside the NMJ (Banovic, 2010). In contrast, dnl2 showed a much stronger colocalization at the NMJ and also shows strong colocalization within the CNS. Furthermore, dnl2 forms a complex with dnrx in vivo, although it remains to be shown whether the same is true for dnl1 (Banovic, 2010). dnl2;dnrx double mutants were lethal and showed more severe defects in bouton morphology than either dnl2 or dnrx mutants alone, whereas dnl1;dnrx mutants were viable and did not show any exacerbation of the morphological defects (Banovic, 2010). Together, these results suggest that the interactions between dnl1 or dnl2 and dnrx serve different functions at the NMJ (Sun, 2011)

Because mutations in dnrx or either of the two dnl genes studied thus far all give rise to a reduction in the number of synaptic boutons, it seems likely that bouton number is regulated via an interaction between dnrx and dnl1 and/or dnl2. It is also apparent, however, that these three genes perform functions independently of each other, such that single mutants have similar yet distinct synaptic phenotypes. Banovic (2010) concluded that dnrx promotes but is not necessary for dnl1 function. The results of the present study, however, showing exaggerated synaptic phenotypes and lethality in dnl2;dnrx double mutants may suggest some redundancy between the functions of dnrx and dnl2. Consistent with this model, a recent publication showed that dnrx is expressed both presynaptically and postsynaptically in embryonic NMJs (Chen, 2010). Furthermore, postsynaptic dnrx appears to specifically promote GluRIIA receptor complexes (Chen, 2010). This raises an interesting possibility of a cis-interaction between dnl2 and dnrx in addition to trans-synaptic interactions, although additional work will be required to assess whether cis-interactions occur, and if so, what role they play (Sun, 2011).

Together, the results of the present study combined with the results of Banovic (2010) suggest that neither dnl1 nor dnl2 are absolutely required for synaptogenesis, but both genes play an essential role in synaptic development. Additional studies will be required to determine the function of other neuroligin genes in Drosophila and to determine whether these genes have functionally redundant roles in synapse development and function (Sun, 2011).

Drosophila Neuroligin3 regulates neuromuscular junction development and synaptic differentiation

Neuroligins (Nlgs) are a family of cell adhesion molecules thought to be important for synapse maturation and function. Studies in mammals have shown that different Nlgs have different roles in synaptic maturation and function. The functions of Drosophila Neuroligin1 (DNlg1), DNlg2, and DNlg4 have also been examined. This study reports the role of DNlg3 in synaptic development and function by using Drosophila neuromuscular junctions (NMJs) as a model system. DNlg3 was found to be expressed in both CNS and NMJs where it was largely restricted to the postsynaptic site. By generating and examining dnlg3 mutants, the mutants mutants were found to exhibit an increased bouton number and reduced bouton size compared to the wild-type. Consistent with alterations in bouton properties, pre- and postsynaptic differentiations were also affected including abnormal synaptic vesicle endocytosis, increased PSD length and reduced GluRIIA recruitment. Additionally, synaptic transmission was reduced. Altogether, this study shows that DNlg3 is required for NMJ development, synaptic differentiation and function (Xing, 2014a).

Activity-dependent retrograde laminin A signaling regulates synapse growth at Drosophila neuromuscular junctions

Retrograde signals induced by synaptic activities are derived from postsynaptic cells to potentiate presynaptic properties, such as cytoskeletal dynamics, gene expression, and synaptic growth. However, it is not known whether activity-dependent retrograde signals can also depotentiate synaptic properties. This study shows that laminin A (LanA) functions as a retrograde signal to suppress synapse growth at Drosophila neuromuscular junctions (NMJs). The presynaptic integrin pathway consists of the integrin subunit βν and focal adhesion kinase 56 (Fak56), both of which are required to suppress crawling activity-dependent NMJ growth. LanA protein is localized in the synaptic cleft and only muscle-derived LanA is functional in modulating NMJ growth. The LanA level at NMJs is inversely correlated with NMJ size and regulated by larval crawling activity, synapse excitability, postsynaptic response, and anterograde Wnt/Wingless signaling, all of which modulate NMJ growth through LanA and βν. These data indicate that synaptic activities down-regulate levels of the retrograde signal LanA to promote NMJ growth (Tsai, 2012).

This study proposes a plasticity mechanism by which the synapse growth (or size) can be modulated by larva crawling and synaptic activities. These activities modulate LanA-integrin signaling that functions to constrain NMJ growth. This trans-synaptic signaling functions in a retrograde manner, which requires postsynaptic muscle-derived LanA and presynaptic integrin. The model suggests various activities modulate NMJ growth by regulating the LanA level and integrin signaling (Tsai, 2012).

Regulation of LanA levels at NMJs is the major mechanism underlying this synaptic structural plasticity. The LanA levels at NMJs are tightly coupled to several synaptic activities that are involved in synaptic structural plasticity at NMJs. Wg signaling in both pre- and postsynaptic compartments are shown to modify synaptic structure at Drosophila NMJs. The channel mutations para and eag Sh alter both synaptic potential and NMJ size. Finally, manipulation of postsynaptic responses by altering the GluRIIA and GluRIIB compositions also fine tunes synapse size and pFAK levels. Activities that promote NMJ growth also down-regulated LanA levels at NMJs. In contrast, NMJ growth suppression was accompanied with LanA accumulation, establishing an inverse correlation between the LanA level and the NMJ size. Importantly, manipulation of the gene dosage of LanA (or βν) could override these synaptic activities in NMJ growth regulation. This study also showed that LanA down-regulation at NMJs preceded synaptic structural remodeling induced by larval crawling, further supporting that LanA is a major mediator of these activities to modulate NMJ growth (Tsai, 2012).

Integrin signaling activities play important roles in synapse development and plasticity. In mammalian central synapses, various integrin subunits are important to transmit postsynaptic signaling in various plasticity models may function redundantly with βν to mediate integrin signaling. This study indicates a distinct presynaptic integrin pathway that is likely composed of βν and αPS3 (encoded by Volado), as suggested by their strong genetic interaction in NMJ growth. In response to postsynapse-secreted LanA signals, activation of the presynaptic integrin is transmitted through Fak56 activation. Interestingly, the signaling activity is rather local, limited by the range of LanA distribution, and shown by muscle 6-specific rescue, although this does not exclude the involvement of signaling to the nuclei of motor neurons. The presynaptic integrin/Fak56 signaling is in turn mediated by two downstream signaling activities. The activation of NF1/cAMP signaling, which suppressed NMJ overgrowth induced by crawling activity or βν mutation. The integrin/Fak56 pathway also suppresses Ras/MAPK signaling Tsai, 2008), as shown by diphospho-ERK (dpERK) accumulation and Fas2 reduction at NMJs in high crawling condition. These pathways have been shown to regulate cell adhesion and cytoskeletal organization, leading to the stabilization of synapses. The activity-dependent depletion of the LanA laminins in the synaptic cleft would allow the remodeling of synapses and further growth of NMJs (Tsai, 2012).

The activity-dependent structural plasticity is specific to the presynaptic integrin pathway. hiw mutants that show large NMJ size still retained the structural plasticity and constant pFAK levels at NMJs. Interestingly, LanA levels were increased in hiw mutants, in contrast to other NMJ overgrown mutants. Two nonmutually exclusive mechanisms can regulate activity-dependent LanA expressions at NMJs. First, within hours of activity induction, the LanA levels can be regulated at NMJs by putative ECM regulators such as matrix metalloproteinases. Second, transcription regulation of LanA can provide long-term changes of LanA levels at NMJs. Activity-triggered presynaptic Wg secretion promotes Wg receptor DFz2 activation on both post- and presynaptic compartments. The LanA level is regulated by the anterograde Wg signaling that is transduced through nuclear entry of the DFz2 intracellular domain and its transcription activity. However, LanA is unlikely to mediate all aspects of Wg signaling activity as overexpression of LanA in postsynapses suppressed ghost bouton formation, a hallmark in disrupting Wg signaling. Postsynaptic BMP/Gbb functions as a retrograde signal to activate presynaptic BMP type II receptor Wit in response to synaptic activity. With the lack of genetic interaction between BMP/Gbb and integrin signaling components, and constant levels of phosphorylated Mothers against dpp (pMad) in different crawling activities, it is proposed that both BMP/Gbb and LanA pathways can function in parallel by retrograde mechanisms to regulate NMJ growth (Tsai, 2012).

Kismet positively regulates glutamate receptor localization and synaptic transmission at the Drosophila neuromuscular junction

The Drosophila neuromuscular junction (NMJ) is a glutamatergic synapse that is structurally and functionally similar to mammalian glutamatergic synapses. These synapses can, as a result of changes in activity, alter the strength of their connections via processes that require chromatin remodeling and changes in gene expression. The chromodomain helicase DNA binding (CHD) protein, Kismet (Kis), is expressed in both motor neuron nuclei and postsynaptic muscle nuclei of the Drosophila larvae. This study shows that Kis is important for motor neuron synaptic morphology, the localization and clustering of postsynaptic glutamate receptors, larval motor behavior, and synaptic transmission. The data suggest that Kis is part of the machinery that modulates the development and function of the NMJ. Kis is the homolog to human CHD7, which is mutated in CHARGE syndrome. Thus, the data suggest novel avenues of investigation for synaptic defects associated with CHARGE syndrome (Ghosh, 2014: PubMed).

Drosophila Neuroligin3 regulates neuromuscular junction development and synaptic differentiation

Neuroligins (Nlgs) are a family of cell adhesion molecules thought to be important for synapse maturation and function. Studies in mammals have shown that different Nlgs have different roles in synaptic maturation and function. The functions of Drosophila Neuroligin1 (DNlg1), DNlg2, and DNlg4 have also been examined. This study reports the role of DNlg3 in synaptic development and function by using Drosophila neuromuscular junctions (NMJs) as a model system. DNlg3 was found to be expressed in both CNS and NMJs where it was largely restricted to the postsynaptic site. By generating and examining dnlg3 mutants, the mutants mutants were found to exhibit an increased bouton number and reduced bouton size compared to the wild-type. Consistent with alterations in bouton properties, pre- and postsynaptic differentiations were also affected including abnormal synaptic vesicle endocytosis, increased PSD length and reduced GluRIIA recruitment. Additionally, synaptic transmission was reduced. Altogether, this study shows that DNlg3 is required for NMJ development, synaptic differentiation and function (Xing, 2014).


Postsynaptic sensitivity to glutamate was genetically manipulated at the Drosophila neuromuscular junction (NMJ) to test whether postsynaptic activity can regulate presynaptic function during development. DGluRIIB, the gene encoding a second muscle-specific glutamate receptor was cloned. This gene is closely related to the previously identified DGluRIIA and located adjacent to it in the genome. Mutations that eliminate DGluRIIA (but not DGluRIIB) or transgenic constructs that increase DGluRIIA expression were generated. When DGluRIIA is missing, the response of the muscle to a single vesicle of transmitter is substantially decreased. However, the response of the muscle to nerve stimulation is normal because quantal content is significantly increased. Thus, a decrease in postsynaptic receptors leads to an increase in presynaptic transmitter release, indicating that postsynaptic activity controls a retrograde signal that regulates presynaptic function (Petersen, 1997).

Double mutants for DGluRIIA and DGluRIIB are embryonic lethal. The homozygous mutants develop to be late embryos but are unable to hatch. When mechanically removed from the chorion and viteline membranes, the mutant embryos appear to be normal, in terms of gross anatomy, but they are unable to crawl. The head is capable of some coordinated movements, but the abdominal body wall muscles merely fibrillate and there are no coordinated peristaltic waves. Therefore, these two receptors are essential for synaptic transmission at the neuromuscular junctions of the abdominal musculature. Transgenic expression of either DGluRIIA or DGluRIIB is able to rescue viability. This demonstrates that either gene is sufficient and that neither is necessary for viability. Expression of either gene in muscle, via a muscle-specific myosin heavy chain promoter is also able to rescue lethality. Therefore, the essential function of these genes is in the somatic musculature (DiAntonio, 1999).

Although the two receptors are redundant at the level of viability, the many differences in amino acid sequence suggests that they might have physiological differences. One measure of receptor function is the quantal size, or response of the muscle to the spontaneous release of a single synaptic vesicle. Quantal size reflects the postsynaptic sensitivity to transmitter, which is determined in large part by the properties of the transmitter receptor. With the genetic tools at hand, both the receptor subunit composition and gene dosage could be varied and the effect on quantal size in vivo could be assessed (DiAntonio, 1999).

Comparison of quantal size at synapses expressing one or the other receptor reveals that DGluRIIA-expressing synapses exhibit a significantly larger response to transmitter than DGluRIIB-expressing synapses. In addition to the difference in amplitude, there is also a difference in the kinetics of the synaptic potentials. The time constant of the miniature extrajunctional potential (mEJP) decay is significantly shorter in DGluRIIB expressing larvae than in DGluRIIA-expressing larvae (21.6 ± 0.6 msec and 32.9 ± 1.2 msec) (DiAntonio, 1999).

The data above suggest that the ratio of receptor subunits at the wild-type synapse could regulate quantal size. A larger proportion of DGluRIIA would increase quantal size, whereas more DGluRIIB would decrease quantal size. When DGluRIIA is overexpressed in a wild-type background, there is a significant increase in quantal size (Petersen, 1997). However, this result is equally consistent with quantal size being regulated by receptor subunit composition or receptor density. To distinguish between these two possibilities, DGluRIIB was overexpressed in a wild-type background. A late driver, MHC Gal4, initiates expression in the first larval instar after endogenous receptor expression has begun, and an early driver, 24B Gal4, initiates expression in myoblasts. In both cases, there is a significant decrease in quantal size. Late expression of DGluRIIB leads to a 44% reduction in mEJP amplitude, whereas earlier expression produces a 68% decrease. A similar change in mEJP amplitude is seen when DGluRIIB is directly overexpressed from the myosin heavy chain promoter. Despite the likely increase in receptor density caused by overexpression, quantal size fell because of a change in the relative abundance of receptor subtype (DiAntonio, 1999).

Although receptor subunit composition is a primary determinant of quantal size, the data suggest that receptor density may also regulate postsynaptic sensitivity to single quantum. When the double mutant is rescued with increasing gene dosages of DGluRIIA, there is a significant increase in quantal size. There is an 18% increase from one to two genomic copies of A and a further 20% increase from two genomic copies to gross overexpression of the cDNA. Because no DGluRIIB is expressed in any of these genotypes, these results cannot be explained by a change in subunit composition between these two receptors, although the existence of a third receptor that may function at this synapse cannot be ruled out. Similarly, there is a 24% increase in quantal size when the gene dosage of DGluRIIB is doubled while rescuing the null mutant. However, there is no further increase in mEJP amplitude when the DGluRIIB cDNA is overexpressed. These data are consistent with a model in which receptor density is a determinant of quantal size (DiAntonio, 1999).

Since quantal size is reduced in the absence of DGluRIIA and quantal size is increased when DGluRIIA is overexpressed (Petersen, 1997) it has been suggested that receptor density is a primary determinant of postsynaptic responsiveness. However, these observations are equally consistent with a model in which the relative level of DGluRIIA regulates quantal size, with a higher proportion of DGluRIIA favoring a larger postsynaptic response. This second model is supported by the data. Regardless of the level of expression, synapses lacking DGluRIIB have a large quantal size, and synapses lacking DGluRIIA have a small quantal size. In fact, overexpression of the DGluRIIB subunit at a wild-type synapse leads to a dose-dependent decrease in quantal size. In this case, the receptor density should be increasing, but the quantal size is decreasing. This is most easily explained if the primary determinant of quantal size at this synapse is the relative abundance of each receptor subtype (DiAntonio, 1999).

Although subunit composition is the primary factor controlling postsynaptic responsiveness, the data does suggest that receptor density may also regulate quantal size. In the absence of the DGluRIIB subunit, increasing the gene dosage of DGluRIIA increases the quantal size. However, the alternate explanation, that the subunit composition is changing between DGluRIIA and an unidentified third receptor, cannot be excluded (DiAntonio, 1999).

How might the cell exploit the differences in receptor function to regulate synaptic strength? (1) The two receptors could be differentially expressed. During embryonic development, DGluRIIB is initially expressed at a high level and then declines, whereas DGluRIIA expression slowly rises throughout embryogenesis (Petersen, 1997). Such a mechanism is used at the vertebrate NMJ in the switch from a fetal to adult acetylcholine receptor subunit. (2) The two receptors could be differentially regulated by second messengers. Davis (1998) has demonstrated that activation of PKA decreases the quantal size at the Drosophila NMJ and that this modulation requires the presence of DGluRIIA. Similar subunit-specific modulation has been seen for numerous vertebrate transmitter receptors (DiAntonio, 1999).

Does the cell use these postsynaptic mechanisms to regulate synaptic strength? When a Drosophila muscle is hypoinnervated, it compensates with an increase in quantal size (Davis, 1998a). It is suggested that this increase in postsynaptic sensitivity may reflect an increase in the proportion of DGluRIIA at the synapse or a decrease in the PKA-dependent modulation of DGluRIIA (DiAntonio, 1999).

To investigate the underlying biophysical basis for the observed differences in quantal properties, a single-channel analysis of the two receptor subunits was undertaken. Outside-out patches were isolated from extrajunctional regions of muscle 6 of wild-type third instar, as well as DGluRIIA&BSP22 mutant larvae rescued with either DGluRIIA or DGluRIIB. The patches were held at minus 60 mV, and 10 mM glutamate was applied with a rapid application system. In response to glutamate, the channels open rapidly, flicker between open and closed states, and desensitize in the continued presence of glutamate. There is no significant difference in single-channel current amplitudes in the three genotypes. Their single-channel conductance is very similar to what has been observed previously for wild-type channels from larvae (Heckmann, 1995) and embryos (Broadie, 1993; Nishikawa, 1995). There is, however, a marked difference in the time course of desensitization. When fit with an exponential function, the time constant of decay is 18 msec for channels from wild-type larvae; 19 msec from larvae expressing DGluRIIA, and 2.0 msec for channels from larvae expressing DGluRIIB (DiAntonio, 1999).

Because channels from wild-type patches that desensitize as quickly as the DGluRIIB channels (Heckmann, 1997 and DiAntonio, 1999) have not been observed, DGluRIIB homomultimers must be quite rare in a wild-type cell. The channels analyzed were extrajunctional; however, there is no evidence for a difference in the time course of patch and quantal currents (Heckmann, 1998). Therefore, this difference in the time course of desensitization seen with single channels may explain some of the differences in quantal amplitude and time course seen at synapses in larvae rescued with either DGluRIIA or DGluRIIB (DiAntonio, 1999).

In DGluRIIA mutants, the amplitude of evoked synaptic events remains normal despite a large decrease in quantal size because of a compensatory increase in quantal content; that is, the number of vesicles released by the nerve (Petersen, 1997). These data have been taken as evidence for a retrograde signal linking postsynaptic activity with presynaptic transmitter release properties. Does a similar form of retrograde signaling occur at synapses mutant for DGluRIIB? At synapses lacking DGluRIIB, the quantal size is near wild-type levels. To assess the relationship between quantal size and quantal content over a wide range of values, the double mutant was rescued with a transgenic UAS DGluRIIA cDNA driven by a Gal4 line (H94) that gives quite variable levels of expression. Recordings of spontaneous miniature junctional potentials and evoked excitatory junctional potentials were made from muscle 6, segment A3 of third instar larvae, and quantal content was estimated by dividing the mean EJP amplitude by the mean mEJP amplitude. There is a significant difference in quantal content when cells were grouped by quantal size; cells with the smallest quantal size tend to have the largest quantal content. This suggests that, at synapses lacking DGluRIIB, changes in postsynaptic activity are compensated for by regulating presynaptic transmitter release. In this genotype, the amplitude of the evoked events is significantly larger than in wild type (25.1 ± 1.4 mV and 15.4 ± 2.0 mV respectively). It is argued in this paper that the compensatory mechanism involves an increase in quantal content, or the number of residues released by the nerve (Di Antonio, 1999).

To assess in a more quantitative manner the relationship between gene dosage of DGluRIIA and quantal content, the synaptic response was compared in 0.3 mM external calcium at the wild-type synapse and in the double mutant rescued with one genomic DGluRIIA transgene, two genomic DGluRIIA transgenes, or by overexpression of the DGluRIIA cDNA. As would be expected from the results above, the single genomic DGluRIIA, with the smallest mean quantal size, gave the largest quantal content. The single DGluRIIA shows a significant increase in quantal content, when compared with wild type (237%); two copies of genomic DGluRIIA have a smaller increase (180%), and overexpression of DGluRIIA has no change in quantal content (114%) (DiAntonio, 1999).

Whereas the inverse relationship between the gene dosage of DGluRIIA and quantal content was expected, the magnitude of the change in quantal content was a surprise. Although the null mutant rescued with a single genomic DGluRIIA transgene does have a slightly smaller quantal size than wild type (0.97 ± 0.06 vs 1.19 ± 0.07 mV), the increase in quantal content more than compensates for this postsynaptic deficit. As a result, the postsynaptic response to nerve stimulation is significantly increased (182%; p < 0.05). As the gene dosage of DGluRIIA (and hence the response to a single vesicle) is increased, the response to nerve stimulation decreases because quantal content is no longer upregulated. This result, in addition to the increase seen in the EJP amplitude in the H94-DGluRIIA rescued larvae described above, suggests that the mechanism monitoring postsynaptic activity and regulating presynaptic transmitter release is not directly sensitive to depolarization of the muscle. Thus when low levels of DGluRIIA are expressed postsynaptically, there is a large increase in presynaptic transmitter release that overcompensates for the decrease in postsynaptic sensitivity to transmitter (DiAntonio, 1999).

Activation of ionotropic glutamate receptors leads to the generation of two types of signals. The postsynaptic cell is depolarized by the influx of cations through the open channel, and second messenger systems can be activated through either the influx of calcium or the interaction of the receptor with other signaling molecules. The overcompensation of quantal content seen in the single genomic DGluRIIA rescue suggests that depolarization is not the determinant being sensed in the postsynaptic cell. In fact, this result could suggest that the retrograde signal is not even sensitive to the activity of the channel but instead is measuring the amount of channel present. To distinguish between the activity and amount of postsynaptic receptor, a dominant negative mutant of DGluRIIA was generated. Using site-directed mutagenesis, a single residue in the channel pore M614 to an R. The analogous mutation in homologous vertebrate channels is thought to coassemble with wild-type receptors and produce nonfunctional channels. Transgenic flies were generated carrying the M/R mutant cloned downstream of the UAS promoter. Expression of two copies of this transgene in a wild-type background driven by the strong mesodermal promoter 24B Gal4 is lethal. Driving expression of a single copy of the mutant with 24B Gal4 produces viable adults with no obvious behavioral abnormalities. Staining of the larval neuromuscular junction shows that this mutant receptor does localize to the synapse. As with overexpression of the wild-type receptor, however, much of the transgenic receptor is present extrasynaptically. Recordings of spontaneous excitatory junctional potentials reveal that expression of this mutant receptor leads to a dramatic decrease in quantal size (1.01 ± 0.05 vs 0.33 ± 0.02 mV). Hence, this pore mutant acts as a dominant negative receptor in vivo. Analysis of evoked synaptic potentials revealed no significant change, indicative of a large increase in quantal content in the mutant. These data do not support the model that a low channel density is the signal controlling the retrograde regulation of presynaptic transmitter release. Normal levels of the endogenous DGluRIIA and DGluRIIB receptors are expressed in addition to the transgenic expression of a DGluRIIA pore mutant, and yet quantal content is upregulated. Although the possibility that the mutant channel could disrupt localization of the endogenous receptors to the synapse cannot be ruled out, the model that it acts as a dominant negative by disrupting the pore of the channel is favored. Therefore, these data imply that the activity of the channel and ion flux through the pore are the initiating events for the measurement of postsynaptic activity and the regulation of presynaptic function. These data lead to the simple model of a homeostatic mechanism in which a muscle-to-motoneuron signal regulates presynaptic release to ensure appropriate depolarization of the muscle. Similar compensation may occur at the vertebrate and crayfish NMJ and at central excitatory and inhibitory synapses. Such a mechanism could function during development to match the release capacity of the nerve to the ever growing requirements of the muscle (DiAntonio, 1999).

In this study a transgenic Ca2+ imaging technique was established in Drosophila that enabled the Ca2+ sensor protein yellow Cameleon-2 to be targeted specifically to larval neurons. This noninvasive method allows the measurement of evoked Ca2+ signals in presynaptic terminals of larval neuromuscular junctions (NMJs). Transgenic Ca2+ imaging was combined with electrophysiological recordings and morphological examinations of larval NMJs to analyze the mechanisms underlying persistently enhanced evoked vesicle release in two independent mutants. Persistent strengthening of junctional vesicle release relies on the recruitment of additional active zones, the spacing of which correlates with the evoked presynaptic Ca2+ dynamics of individual presynaptic terminals. Knock-out mutants of the postsynaptic glutamate receptor (GluR) subunit DGluR-IIA, which showed a reduced quantal size, develop NMJs with a smaller number of presynaptic boutons but a strong compensatory increase in the density of active zones. This results in an increased evoked vesicle release on single action potentials and larger evoked Ca2+ signals within individual boutons; however, the transmission of higher frequency stimuli is strongly depressed. A second mutant, pabpP970/+ in a gene coding for a [poly(A)-binding protein (pabp)], shows genetically elevated subsynaptic protein synthesis, which shows unaltered quantal size but strongly increased eEJCs caused by enhanced evoked vesicle release. pabpP970/+ showed enhanced evoked vesicle release triggered by elevated subsynaptic protein synthesis, developed NMJs with an increased number of presynaptic boutons and active zones; however, the density of active zones is maintained at a value typical for wild-type animals. This resulted in wild-type evoked Ca2+ signals but persistently strengthened junctional signal transmission. These data suggest that the consolidation of strengthened signal transmission relies not only on the recruitment of active zones but also on their equal distribution in newly grown boutons (Reiff, 2002).

This study addresses the question of how NMJs of Drosophila larvae achieve the continuous enhancement of evoked vesicle release seen throughout their development and during activity-dependent strengthening. Using wild-type animals and two independent mutants that genetically represent both phases of junctional strengthening, it was found that Ca2+-dependent presynaptic mechanisms, which are known to result in fast and reversible modifications of presynaptic vesicle release, may provide only a minor or transient contribution to enhanced vesicle release during the development and long-term strengthening of junctional signal transmission. Instead, a persistent enhancement of vesicle release relies primarily on the recruitment of active zones. This conclusion is further supported by yCam2-based Ca2+ imaging results, which together with ultrastructural data reveal that evoked presynaptic Ca2+ signals correlate with the density of active zones. The data therefore suggest that enhanced vesicle release is realized by a differential regulation of active zone density in different genotypes: NMJs of dglurIIA-ko mutants compensate for their postsynaptic defect by packing more active zones into preexisting boutons. This leads to a functional compensation, which approaches homeostasis of evoked junctional signal transmission compared with wild type presumably to ensure muscle contraction and animal survival. In contrast, enhanced junctional signal transmission as seen in elav-Campabp animals is mediated by distributing added active zones into newly grown boutons. This leads to homeostasis of active zone density compared with wild-type controls and therefore may reflect the cellular basis of strengthened junctional signal transmission at Drosophila NMJs (Reiff, 2002).

Previous ultrastructural observations from other Drosophila mutants and larvae of the flesh fly Sarcophaga bullata have already suggested that the density of active zones is tightly regulated, presumably to ensure that individual synapses have sufficient access to reserve pool vesicles, vesicle recycling machinery, efficient Ca2+-buffering systems, or neurotransmitter uptake mechanisms, for example. Data from wild-type and elav-Campabp animals show a similar active zone density and evoked Ca2+ signals per bouton and thus suggest that individual boutons represent functional compartments that are likely to be maintained constant during junctional development and its strengthening. This seems to guarantee uncompromised signal transmission on a single bouton level. From these observations a model emerged that predicts that additional active zones need to be distributed in newly grown boutons. This would explain the increasing number of genotypes that show a strict relationship between bouton number and transmission strength. Intriguingly, in several other systems the recruitment of active synapses has been observed, as well as local morphological alterations of synaptic compartments; both are thought to represent long-lasting changes in the strength of synaptic communication (Reiff, 2002).

On the basis of the above considerations, it appears surprising that DGluR-IIA-ko mutants pack the additional active zones in a smaller number of presynaptic boutons. This results in an increased density of active zones, a larger stimulation-evoked Ca2+ entry per bouton, an enhanced evoked vesicle release, and a wild-type muscle depolarization on single action potentials. These phenotypes show that mutants with impaired postsynaptic glutamate receptor function are capable of efficiently triggering the recruitment of active zones to compensate for the mutationally induced postsynaptic defect. However, this recruitment fails to induce the proportional outgrowth of new boutons that can be observed at wild-type NMJs and several other genotypes. Indeed, a recent analysis of the role of DGluR-IIA subunits in junctional development revealed that the increased expression of DGluR-IIA is sufficient to induce bouton outgrowth. Although it is currently not clear why DGluR-IIA-ko mutants accumulate active zones at such an unusual density, it appears that this mechanism alone is not sufficient to ensure uncompromised repetitive signal transmission. Although the latter may be attributable to increased postsynaptic desensitization in this mutant, presynaptic factors like the depletion of the readily releasable vesicle pool also appear likely to contribute to this observation. It is therefore tempting to speculate that this mutant is trapped in a transient phase of junctional strengthening (Reiff, 2002).

According to such a model, a postsynaptic sensor would trigger signals that control the recruitment of active zones. The transiently increased density of active zones would trigger a second signal that instructs the resetting of active zone density by distributing them into newly grown boutons. Intriguingly, in the chronically hyperactive mutant eag, Sh represents a precedence for this scenario because the eag mutant shows, presumably because of the continuous hyperactivity stimulus, an increased density of T-bar-harboring active zones in an already increased number of junctional boutons. These findings provide further evidence for the suggestion that developmental processes and activity-dependent phenomena may use closely related mechanisms (Reiff, 2002).

The postsynaptic glutamate receptor subunit DGluR-IIA mediates long-term plasticity in Drosophila

The developing neuromuscular junctions (NMJs) of Drosophila larvae can undergo long-term strengthening of signal transmission, a process that has been shown recently to involve local subsynaptic protein synthesis and that is associated with an elevated synaptic accumulation of the postsynaptic glutamate receptor subunit DGluR-IIA. To analyze the role of altered postsynaptic glutamate receptor expression during this form of genetically induced junctional plasticity, the expression levels of two so far-described postsynaptic receptor subunit genes, dglur-IIA and dglur-IIB, were manipulated in wild-type animals and plasticity mutants. Elevated synaptic expression of DGluR-IIA, which was achieved by direct transgenic overexpression, by genetically increased subsynaptic protein synthesis, or by a reduced dglur-IIB gene copy number, results in an increased recruitment of active zones, a corresponding enhancement in the strength of junctional signal transmission, and a correlated addition of boutons to the NMJ. Ultrastructural evidence demonstrates that active zones appear throughout NMJs at a typical density regardless of genotype, suggesting that the space requirements of active zones are responsible for the homogeneous synapse distribution and that this regulation results in the observed growth of additional boutons at strengthened NMJs. These phenotypes were suppressed by reduced or eliminated DGluR-IIA expression, which resulted from either a reduced dglur-IIA gene copy number or transgenic overexpression of DGluR-IIB. These results demonstrate that persistent alterations of neuronal activity and subsynaptic translation result in an elevated synaptic accumulation of DGluR-IIA, which mediates the observed functional strengthening and morphological growth apparently through the recruitment of additional active zones (Sigrist, 2002).

The focus of this study was the question of how local subsynaptic protein synthesis can regulate the transmission strength and the morphological development of larval neuromuscular junctions of Drosophila. The increased synaptic accumulation of the glutamate receptor subunit DGluR-IIA alone is responsible for the functional recruitment of additional synapses within a given NMJ. This was evident in an apparent increase in the number of DGluR-IIA-expressing postsynapses on transgenic overexpression of DGluR-IIA, an increased total number of T-bar-harboring release sites per NMJ, an increased frequency of spontaneous vesicle fusion events, and significantly larger junctional responses on nerve stimulation (eEJCs) compared with control animals. These physiological changes, which have been evoked solely by manipulating the expression level of DGluR-IIA, were indistinguishable from those seen in animals with genetically increased subsynaptic translation, and they were suppressed in the latter genotypes by reducing the dglur-IIA gene doses. These observations suggest that most if not all of the physiological effects of subsynaptic protein synthesis at NMJs are mediated by increased synaptic expression of the glutamate receptor subunit DGluR-IIA. Because subsynaptically stored mRNAs encoding DGluR-IIA represent a likely substrate of synaptic translation, the local synthesis of DGluR-IIA subunits and their subsequent synaptic delivery could therefore provide the means for a site-specific functional recruitment of synapses and thus for local alterations of glutamatergic signal transmission (Sigrist, 2002).

Interestingly, it was found that the synaptic expression level of DGluR-IIA and its associated physiological phenotypes are inversely related to the expression of the glutamate receptor subunit DGluR-IIB: a reduced dglur-IIB gene copy number results in a significant increase of synaptic DGluR-IIA accumulation, whereas the transgenic overexpression of DGluR-IIB reduces synaptic DGluR-IIA levels. One possibility to explain this inverse relationship of both glutamate receptor expression levels could be a competition of both subunits in the formation of hetero-oligomeric receptor complexes. Another possibility may reside in the opposing roles of DGluR-IIA and DGluR-IIB for synaptic signal transmission: synapses expressing DGluR-IIA resemble wild-type transmission characteristics, whereas DGluR-IIB-expressing synapses exhibit very fast desensitization kinetics, resulting in strongly reduced quantal sizes. Given that NMJs with small quantal sizes are accompanied by suppressed subsynaptic protein synthesis this could result in an inefficient subsynaptic synthesis and a reduced synaptic delivery of DGluR-IIA in DGluR-IIB-overexpressing animals. In turn, synapses with reduced or no DGluR-IIB may efficiently activate the DGluR-IIA synthesis and their synaptic deposition. Although it is currently not possible to differentiate between these and other possibilities, it is important to note that NMJs are apparently equipped with two subunit-specific mechanisms, which because of their opposing effects on synaptic DGluR-IIA accumulation are well suited to tightly control the subunit composition of postsynaptic glutamate receptors (Sigrist, 2002).

On the basis of these data, it appears that a crucial factor for the implementation of persistently strengthened junctional signal transmission is the controlled upregulation of DGluR-IIA, which results in the functional recruitment of additional synapses. These added synapses show postsynaptic responses to released quanta of glutamate that are typical for wild-type NMJs, suggesting that increased DGluR-IIA expression results primarily in a larger number of normally operating postsynapses. Very strong overexpression of DGluR-IIA, which has been achieved using a cDNA-based transgene, appears to further increase the amount of DGluR-IIA at individual postsynapses and has been shown to result in a dose-dependent increase of miniature excitatory junctional potential (mEJP) amplitudes. These findings suggest that not only the number of responsive postsynapses can be changed by DGluR-IIA, but also the postsynaptic sensitivity can be changed. They further support the idea that the number of postsynaptic glutamate receptor complexes per synapse determines the amplitudes of mEJPs, and they are consistent with results from hippocampal synapses, which propose that a neurotransmitter from a single vesicle saturates all postsynaptic glutamate receptors of that synapse (Sigrist, 2002).

Strengthening of glutamatergic synapses and activation of silent postsynapses during long-term potentiation has recently gained much attention. Several lines of evidence have suggested that the targeted trafficking of specific glutamate receptor subunits and their incorporation into preexisting synapses represent a prominent route of synaptic activation and functional modification in hippocampal preparations. The results from the Drosophila NMJ indicate that these glutamatergic synapses use similar postsynaptic mechanisms to functionally recruit additional synapses, indicating that the local synthesis and the targeted trafficking of receptor subunits may represent an evolutionary conserved mode to alter glutamatergic circuits in a site-specific manner (Sigrist, 2002).

There is increasing evidence from this and several other recent studies that the strength of junctional signal transmission is correlated with the number of synapse-harboring boutons. Strikingly, the density of active zones, which represent sites of high-probability vesicle release, is approximately constant within individual boutons in all analyzed genotypes, even in NMJs with strongly enhanced signal transmission. This observation is consistent with a recent report suggesting that the spacing of active zones at NMJs and in the visual system of Drosophila and Sarcophaga is tightly regulated, presumably because each active zone requires a large enough surrounding surface area for proper function. It therefore seems likely that synapse recruitment leads to a transient increase in the density of active zones at larval NMJs of Drosophila, which are induced to grow to provide the now-required additional synaptic surface area. Interestingly, this growth does not involve a simple size increase of preexisting boutons, but it uses in a FasII-dependent manner the rather costly addition of new boutons to NMJs. This suggests that the axonal compartmentalization, which is given in form of type I boutons, generates functional units that, similarly to the spacing of active zones, need to be homeostatically preserved. It therefore appears that the consistent correlation between the strength of junctional signal transmission and the number of junctional boutons reflects the consolidation of induced functional changes, which include the functional recruitment of synapses and their distribution in newly grown boutons (Sigrist, 2002).

On the basis of the prominent role of the glutamate receptor subunit DGluR-IIA during long-term strengthening of signal transmission, the question arises whether NMJs can develop without DGluR-IIA. Surprisingly, NMJs with genetically eliminated DGluR-IIA expression develop to a size that resembles that of wild-type NMJs, despite strong defects in synaptic signal transmission. The same effect of eliminated DGluR-IIA expression was found in animals with increased subsynaptic protein synthesis, that normally develop significantly larger NMJs. Moreover, mutants in the synaptic vesicle protein synaptotagmin that have substantial defects in junctional signal transmission show similar basal development of NMJs. These observations demonstrate that neither DGluR-IIA expression itself nor intact synaptic physiology or subsynaptic translation is required to develop NMJs with a relatively simple morphology. They suggest that larval NMJs can develop according to a program that appears to be independent of neuronal activity and the expression of the glutamate receptor subunit DGluR-IIA. This 'programmed development' appears to establish a minimal innervation that would typically ensure baseline synaptic signal transmission and muscle contraction (Sigrist, 2002).

Superimposed on such programmed development, some of the mechanisms underlying the functional and structural modulation of the initially established synaptic connectivity are described in this study. Because postsynaptic DGluR-IIA expression plays a key role in this form of plasticity, which is likely regulated by neuronal activity and local subsynaptic protein synthesis, it is proposed that the 'activity-dependent' mode of junctional development helps adjust the junctional performance to the prevailing needs of the individual animal. A similar concept of activity-induced modifications of previously established neural circuits has been implicated in the development and functional tuning of various neural networks, for example, during the formation of barrels in the somatosensory cortex, and of ocular dominance columns in the primary visual cortex. On the basis of these similarities, the molecular and genetic analysis of developing NMJs of Drosophila might yield further important insights into the mechanisms underlying the activity-dependent remodeling of synaptic networks (Sigrist, 2002).

Retrograde control of synaptic transmission by postsynaptic CaMKII at the Drosophila neuromuscular junction

Retrograde signaling plays an important role in synaptic homeostasis, growth, and plasticity. A retrograde signal at the neuromuscular junction (NMJ) of Drosophila controls the homeostasis of neurotransmitter release. This retrograde signal is regulated by the postsynaptic activity of Ca2+/calmodulin-dependent protein kinase II (CaMKII). Reducing CaMKII activity in muscles enhances the signal and increases neurotransmitter release, while constitutive activation of CaMKII in muscles inhibits the signal and decreases neurotransmitter release. Postsynaptic inhibition of CaMKII increases the number of presynaptic, vesicle-associated T bars at the active zones. Consistently, it is shown that glutamate receptor mutants also have a higher number of T bars; this increase is suppressed by postsynaptic activation of CaMKII. Furthermore, presynaptic BMP receptor Wishful thinking is required for the retrograde signal to function. These results indicate that CaMKII plays a key role in the retrograde control of homeostasis of synaptic transmission at the NMJ of Drosophila (Haghighi, 2003).

Reducing the function of postsynaptic glutamate receptors at the neuromuscular junction (NMJ) of Drosophila triggers a retrograde signal from the postsynaptic muscle to the presynaptic motor neuron, leading to an increase in the amount of neurotransmitter release. This retrograde signal is regulated by the postsynaptic activity of CaMKII. Reducing postsynaptic CaMKII activity by expressing a CaMKII inhibitory peptide in somatic muscles increases quantal content, mimicking the effect of reducing postsynaptic glutamate receptor activity. Furthermore, in glutamate receptor GluRIIA-/- mutants, constitutive activation of CaMKII in muscles inhibits the retrograde signal and decreases quantal content. These changes in retrograde signaling and neurotransmitter release are not accompanied by any significant changes in the number of synaptic boutons per muscle surface area or any gross structural or ultrastructural alterations. However, upon inhibition of CaMKII postsynaptically the number of T bars per active zone in presynaptic boutons is significantly increased. Similarly, the number of T bars per active zone is doubled in GluRIIA-/- mutant larvae. This increase is suppressed by constitutive activation of CaMKII in postsynaptic muscles in GluRIIA-/- mutants. These results point to CaMKII as a key regulator of the retrograde signal controlling homeostasis of synaptic transmission at the NMJ of Drosophila (Haghighi, 2003).

Postsynaptic inhibition of CaMKII activity is sufficient to increase presynaptic neurotransmitter release in a retrograde fashion. This increase in quantal content can be potentiated by expressing additional doses of the inhibitory transgene and suppressed by expressing a constitutively active CaMKII transgene simultaneously. These results suggest a direct involvement of CaMKII in controlling the retrograde signal that maintains the homeostasis of neurotransmitter release at the Drosophila NMJ (Haghighi, 2003).

While increasing the postsynaptic activity of CaMKII in wild-type larvae has no effect on neurotransmitter release, once the retrograde signal is induced (i.e., in GluRIIA-/- mutants), activation of CaMKII can inhibit the signal. This is consistent with the observation that while removal of GluRIIA causes a decrease in quantal size and an increase in quantal content, overexpression of GluRIIA, which leads to an increase in quantal size, does not change quantal content. These results suggest that quantal content may be increased only when CaMKII activity is reduced to a critical threshold. As long as CaMKII activity remains above this critical threshold, quantal content is unchanged. This could be a mechanism through which the synapse can compensate for any reduction in muscle activity and ultimately maintain homeostasis (Haghighi, 2003).

In a recent study, Kazama (2003) has provided evidence for the involvement of postsynaptic CaMKII in the retrograde control of neurotransmission at the Drosophila NMJ. Changes in the activity of postsynaptic CaMKII have been shown to affect both neurotransmitter release and synaptic structure in early first instar larvae. Kazama also reports an apparent change in localization of postsynaptic glutamate receptors in response to postsynaptic activation of CaMKII. This is in contrast to the current observations; no changes were found in the Highaghi (2003) study in overall synaptic structure or localization of glutamate receptors in response to either inhibition or activation of postsynaptic CaMKII. The differences in these findings could be partially due to differences in the level and pattern of expression of transgenes (using different Gal4 lines) or due to the fact that the NMJ was examined at very different developmental stages. Haghighi (2003) examined late third instar larvae, while Kazama (2003) examined early first instar larvae. Interestingly, the results are in agreement in that both studies observed no changes in the amplitude or kinetics of spontaneous potentials, indicating that CaMKII does not directly modulate glutamate receptors at the Drosophila NMJ; this is not the case with vertebrates (Haghighi, 2003).

When the activity of postsynaptic glutamate receptors at the Drosophila NMJ is reduced, a retrograde signal from the muscle to the motor neuron is triggered that causes an increase in quantal content. It has been suggested that this retrograde signal could be triggered in response to changes in muscle depolarization or in response to Ca2+ conducted by glutamate receptors. The data indicate that postsynaptic activity of CaMKII plays an important role in controlling the signal. Both muscle depolarization and Ca2+ flux through glutamate receptors could be involved in changing the levels of intracellular Ca2+ and thus that of CaMKII. The idea is favored that calcium influx through glutamate receptors is at least in part responsible for activating CaMKII and triggering the retrograde signal. There are several lines of evidence that support this hypothesis (Haghighi, 2003).

One line of evidence is based on the glutamate receptor ion channel properties and how they are changed in GluRIIA-/- mutants. Compared to wild-type receptors, glutamate receptors in these mutants have a greatly reduced single-channel mean open time; this also affects the kinetics of EPSPs. Therefore, evoked currents that give rise to similar EPSP peak amplitudes in wild-type and GluRIIA-/- mutants will lead to less ion influx in the mutants (ion flux is a product of time and current). Considering the high Ca2+ permeability of glutamate receptors (PCa/PNa = 9.55), due to this reduced ion influx, Ca2+ influx will also be reduced in GluRIIA-/- mutants both during spontaneous and evoked activities. Therefore, it is conceivable that this change in Ca2+ influx, monitored by CaMKII, could act as a trigger for the retrograde signal (Haghighi, 2003).

This hypothesis is further supported by data demonstrating that the retrograde increase in quantal content in GluRIIA-/- mutants can be counteracted by overexpressing multiple copies of GluRIIA, independent of the size of EPSPs. The conclusion from these results is that the retrograde control of presynaptic release for these genotypes is not directly related to muscle depolarization. In addition, it is argued that if muscle depolarization were the sole trigger for the homeostatic retrograde signaling, then quantal content in highwire (hiw) mutants should have been compensated for. hiw mutants have 60%-70% less quantal release, while retrograde control of neurotransmission is still intact in these mutants. Therefore, it is proposed that postsynaptic CaMKII regulates presynaptic release by responding to calcium influx through glutamate receptors during evoked and spontaneous neurotransmitter release (Haghighi, 2003).

The role of postsynaptic membrane depolarization in homeostatic control of presynaptic release at the Drosophila NMJ has been investigated by Paradis (2001). This study demonstrates that the expression of an inward-rectifying potassium channel, Kir2.1, in postsynaptic muscles leads to an increase in quantal content. Kir2.1-expressing muscles show severe defects in muscle properties, including input resistance and membrane potential. More importantly, muscle excitability is affected to the point that mEPSP amplitude is reduced to less than half of wild-type levels (Paradis, 2001). The authors further show that mEPSCs are still wild-type under voltage-clamp conditions, suggesting no change in glutamate receptor function. However, considering the kinetics of membrane depolarization, it is conceivable that, under physiological conditions, GluRIIA function could be compromised during an evoked response that is severely reduced in duration. In other words, while glutamate receptor function is not affected directly, these results suggest that ion influx through glutamate receptors could be affected due to membrane defects. Therefore, the moderate increase in quantal content could be partially due to this apparent reduction in glutamate receptor activity (Haghighi, 2003).

How does the motor neuron respond to the retrograde signal? Based on the results, inhibition of postsynaptic CaMKII mimics the reduction in postsynaptic activity in glutamate receptor mutants and triggers the retrograde signal, leading to an increase in neurotransmitter release at the NMJ. This increase in neurotransmitter release does not appear to induce the NMJ to grow more synaptic boutons, since the numbers of synaptic boutons remained unchanged. Similarly, the overall ultrastructure of boutons remained indistinguishable from wild-type. In contrast, Koh (1999) has reported an overdevelopment of the subsynaptic reticulum in larvae expressing a CaMKII inhibitory peptide (Ala) (Haghighi, 2003).

In that study, Ala or CaMKIIT287D were expressed in both muscles and the nervous system simultaneously, whereas Haghighi (2003) manipulated CaMKII only in neurons or muscles exclusively. It is conceivable that the level and the pattern of expression of these transgenes could have led to differences between experiments. Furthermore, Koh analyzed boutons at the midline section only, while the Haghighi study looked at complete serial sections of boutons. These differences in the levels or pattern of Ala expression as well as differences in analyses could underlie this discrepancy (Haghighi, 2003).

The Haghighi study found a 60% increase in the number of T bars per active zone in response to inhibition of CaMKII in muscles. Often present at active zones at the Drosophila NMJ, T bars are electron-dense structures associated with clusters of synaptic vesicles. Higher numbers of active zones and T bars appear to correlate with an increase in the strength of synaptic transmission. For example, hyperexcitable eag shaker mutants contain a higher number of T bars than wild-type at NMJ synapses. The Haghighi study further demonstrates that induction of retrograde signaling in GluRIIA-/- mutants leads to a doubling of the number of T bars per active zone, similar to the effect of postsynaptic inhibition of CaMKII. Reiff (2002) has recently reported an increase in T bars in another allelic combination of glutamate receptor mutants. Finally, the Haghighi study shows that the increase in T bars could be surpressed by postsynaptic activation of CaMKII. Since postsynaptic activation of CaMKII in glutamate receptor mutants also suppresses quantal content, these results further support a direct correlation between presynaptic T bars and neurotransmitter release at the Drosophila NMJ. These findings suggest that postsynaptic reduction of CaMKII activity may boost presynaptic neurotransmitter release by upregulating T bars at active zones in presynaptic boutons, a potential mechanism for the control of synaptic transmission induced by the retrograde signal. The number of T bars per active zone could therefore be used as an index for the presence of the homeostatic retrograde signal, independent of quantal content measurements (Haghighi, 2003).

It has been demonstrated that a BMP type II receptor, wishful thinking (wit), is required for both growth and function of the NMJ in Drosophila. To further explore the mechanism by which motor neurons respond to the retrograde signal, whether the retrograde enhancement of quantal content can occur in wit mutants was examined. The results indicate that the retrograde signal cannot increase neurotransmitter release in the absence of Wit. Activation of the retrograde signal by either postsynaptic expression of GluRIIAM/R or postsynaptic inhibition of CaMKII did not lead to any increase in quantal content. These results indicate a requirement for wit presynaptically for the functioning of the retrograde mechanism that controls the homeostasis of neurotransmitter release at the NMJ of Drosophila and that postsynaptic inhibition of CaMKII requires the function of presynaptic BMP signaling to enhance quantal release (Haghighi, 2003).

Glass bottom boat (Gbb), a BMP ortholog, functions as a retrograde ligand for Wit at the Drosophila NMJ. Mutations in gbb lead to NMJ defects similar to those observed in wit mutants, and postsynaptic transgenic expression of Gbb can rescue many of these defects. In light of these findings, it is possible that there is a link between postsynaptic activity of CaMKII and the level and function of Gbb at the NMJ of Drosophila (Haghighi, 2003).

Another candidate protein for interacting with CaMKII in controlling retrograde signaling is Discs large (DLG). DLG has been shown to be phosphorylated by CaMKII (Koh, 1999) and to be involved in synaptic transmission at the NMJ of Drosophila. However, the defects in synaptic transmission in dlg mutants are rescued by presynaptic rather than postsynaptic expression of DLG. This suggests that the role of DLG in neurotransmission is primarily presynaptic. Furthermore, in the rho-type guanine nucleotide exchange factor dpix mutants quantal release is not greatly affected, while DLG levels are reduced by 80%. Therefore, it seems unlikely that the effects observed are due to changes in DLG phosphorylation levels. Additional experiments are needed to further elucidate the mechanism through which CaMKII activity controls the homeostasis of neurotransmitter release and to identify target proteins that CaMKII may interact with in the postsynaptic cell (Haghighi, 2003).

Activity-dependent synaptic plasticity at glutamatergic neuromuscular junctions: Retrograde signaling by Syt 4 induces presynaptic release and synapse-specific growth

The molecular pathways involved in retrograde signal transduction at synapses and the function of retrograde communication are poorly understood. Postsynaptic calcium 2+ ion (Ca2+) influx through glutamate receptors and subsequent postsynaptic vesicle fusion trigger a robust induction of presynaptic miniature release after high-frequency stimulation at Drosophila neuromuscular junctions. An isoform of the synaptotagmin family, Synaptotagmin 4 (Syt 4), serves as a postsynaptic Ca2+ sensor to release retrograde signals that stimulate enhanced presynaptic function through activation of the cyclic adenosine monophosphate (cAMP)-cAMP-dependent protein kinase pathway. Postsynaptic Ca2+ influx also stimulates local synaptic differentiation and growth through Syt 4-mediated retrograde signals in a synapse-specific manner (Yoshihara, 2005).

Neuronal development requires coordinated signaling to orchestrate pre- and post-synaptic maturation of synaptic connections. Synapse-specific enhancement of synaptic strength as occurs during long-term potentiation, as well as compensatory homeostatic synaptic changes, have been suggested to require retrograde signals for their induction. Although retrograde signaling has been implicated widely in synaptic plasticity, the molecular mechanisms that transduce postsynaptic Ca2+ signals during enhanced synaptic activity to alterations in presynaptic function are poorly characterized. Because postsynaptic Ca2+ is essential for synapse-specific potentiation, it is important to characterize how Ca2+ can regulate retrograde communication at synapses (Yoshihara, 2005).

To dissect the mechanisms underlying activity-dependent synaptic plasticity, tests were performed to see whether newly formed Drosophila glutamatergic neuromuscular junctions (NMJs), which have ~30 active zones, show physiological changes after 100-Hz stimulation. Within 1 min after stimulation, a gradual 100-fold increase in miniature excitatory postsynaptic current (miniature) frequency was observed, from a baseline of 0.03 Hz to often more than 5 Hz. The high-frequency-stimulation-induced miniature release (termed HFMR) continued for a few minutes to as long as 20 min before subsiding to baseline levels. Perfusion of postsynaptic muscles with the Ca2+ chelator EGTA from the patch pipette caused a modest suppression of HFMR, whereas the fast Ca2+ chelator BAPTA induced strong suppression by 2.5 min of perfusion. Longer perfusion with BAPTA for 5 min before stimulation abolished HFMR, indicating HFMR is induced after postsynaptic Ca2+ influx (Yoshihara, 2005).

Ca2+-induced vesicle fusion in presynaptic terminals provides a temporally controlled and spatially restricted signal essential for synaptic communication. Postsynaptic vesicles within dendrites have been visualized by transmission electron microscopy, and dendritic release of several neuromodulators has been reported. To test whether postsynaptic vesicle fusion might underlie the Ca2+-dependent release of retrograde signals, postsynaptic vesicle recycling was blocked by using the dominant negative shibirets1 mutation, which disrupts endocytosis at elevated temperatures. shibirets1 was expressed specifically in postsynaptic muscles by driving a UAS-shibirets1 transgene with muscle-specific myosin heavy chain (Mhc)-Gal4, keeping presynaptic activity intact. At the permissive temperature (23°C), high-frequency stimulation induced normal HFMR. However, raising the temperature to 31°C suppressed HFMR in the presence of postsynaptic shibirets1, whereas wild-type animals displayed normal HFMR at 31°C. Basic synaptic properties in Mhc-Gal4, UAS-shibirets1 animals were not affected at either the permissive or the restrictive temperature. The suppression of HFMR is not due to irreversible damage induced by postsynaptic UAS-shibirets1 expression, because a second high-frequency stimulation after recovery to the permissive temperature induced normal HFMR (Yoshihara, 2005).

The synaptic vesicle protein synaptotagmin 1 (Syt 1) is the major Ca2+ sensor for vesicle fusion at presynaptic terminals but is not localized postsynaptically. Another isoform of the synaptotagmin family, synaptotagmin 4 (Syt 4), is present in the postsynaptic compartment (Adolfsen, 2004), suggesting Syt 4 might function as a postsynaptic Ca2+ sensor. Syt 4 immunoreactivity is observed in a punctate pattern surrounding presynaptic terminals, suggesting Syt 4 is present on postsynaptic vesicles. Postsynaptic vesicle recycling was again blocked by using the UAS-shibirets1 transgene driven with Mhc-Gal4. Without a temperature shift, Syt 4-containing vesicles show their normal postsynaptic distribution surrounding presynaptic terminals. When the temperature is shifted to 37°C for 10 min in the presence of high-K+ saline containing 1.5 mM Ca2+ to drive synaptic activity, Syt 4-containing vesicles translocate to the plasma membrane. After recovery at 18°C for 20 min, postsynaptic vesicles return to their normal position. Removing extracellular Ca2+ during the high-K+ stimulation results in vesicles that do not translocate to the postsynaptic membrane (Yoshihara, 2005).

To further test whether the Syt 4 vesicle population undergoes fusion with the postsynaptic membrane as opposed to mediating fusion between intracellular compartments, transgenic animals were constructed expressing a pH-sensitive green fluorescent protein (GFP) variant (ecliptic pHluorin) fused at the intravesicular N terminus of Syt 4. Ecliptic pHluorin increases its fluorescence 20-fold when exposed to the extracellular space from the acidic lumen of intracellular vesicles during fusion. Expression of Syt 4-pHluorin in postsynaptic muscles resulted in intense fluorescence at specific subdomains in the postsynaptic membrane, defining regions where Syt 4 vesicles undergo exocytosis. The fluorescence was not diffusely present over the postsynaptic membrane but directed to restricted compartments. Mhc-Gal4, UAS-Syt 4-pHluorin larvae were co-stained with antibodies against the postsynaptic density protein, DPAK, and nc82, a monoclonal antibody against a presynaptic active zone protein. Syt 4-pHluorin colocalized with DPAK and localized adjacent to nc82, demonstrating that Syt 4-pHluorin translocates from postsynaptic vesicles to the plasma membrane at postsynaptic densities opposite presynaptic active zones (Yoshihara, 2005).

To examine the function of Syt 4-dependent postsynaptic vesicle fusion, the phenotype of a syt 4 null mutant (syt 4BA1) ( Adolfson, 2004) and a syt 4 deficiency (rn16) was characterized. Mutants lacking Syt 4 hatch from the egg case 21 hours after egg laying at 25°C, similar to wild type, and grow to fully mature larvae that pupate and eclose with a normal time course. To determine whether postsynaptic vesicle fusion triggered by Ca2+ influx is required for HFMR, the effects of high-frequency stimulation were analyzed in syt 4 mutants. In contrast to controls, the increase of miniature release was eliminated in syt 4 mutants. Postsynaptic expression of a UAS-syt 4 transgene completely restored HFMR in the null mutant, demonstrating that postsynaptic Syt 4 is required for triggering enhanced presynaptic function. Presynaptic expression of a UAS-syt 4 transgene did not restore HFMR. In addition, postsynaptic expression of a mutant Syt 4 with neutralized Ca2+-binding sites in both C2A and C2B domains did not rescue HFMR, indicating that retrograde signaling by Syt 4 requires Ca2+ binding (Yoshihara, 2005).

The large increase in miniature frequency observed during HFMR is similar to the enhancement of presynaptic release after activation of cyclic adenosine monophosphate (cAMP)-dependent protein kinase (PKA) described in Aplysia and Drosophila. Bath application of forskolin, an activator of adenylyl cyclase, results in a robust enhancement of miniature frequency at Drosophila NMJs similar to that observed during HFMR, suggesting retrograde signals may function to increase presynaptic cAMP. To test the role of the cAMP-PKA pathway in HFMR, DC0 mutants were assayed for the presence of HFMR. DC0 encodes the major catalytic subunit of PKA in Drosophila and has been implicated in olfactory learning. Similar to the lack of forskolin-induced miniature induction, DC0 null mutants lack HFMR. Bath application of forskolin in syt 4 mutants resulted in enhanced miniature frequency, suggesting activation of the cAMP pathway can bypass the requirement for Syt 4 in synaptic potentiation (Yoshihara, 2005).

To further explore the role of retrograde signaling at Drosophila synapses, the role of activity in synapse differentiation and growth was characterized. During Drosophila embryonic development, presynaptic terminals undergo a stereotypical structural change from a flat path-finding growth cone into varicose synaptic terminals through dynamic reconstruction. Such developmental changes in synaptic structure may share common molecular mechanisms with morphological changes induced during activity-dependent plasticity. Synaptic transmission was eliminated by using a deletion mutation that removes the postsynaptic glutamate receptors, DGluRIIA and DGluRIIB (hereafter referred to as GluRs). Postsynaptic currents normally induced by nerve stimulation were completely absent in the mutants (gluR). Miniatures were also eliminated, even at elevated extracellular Ca2+ concentrations of 4 mM. In the absence of GluRs, the presynaptic morphology of motor terminals is abnormal, even though GluRs are only expressed in postsynaptic muscles. GluR-deficient terminals maintain a flattened growth cone-like structure and fail to constrict into normal synaptic varicosities. Synaptic development was assayed in a null mutant of the presynaptic plasma membrane t-SNARE [SNAP (soluble N-ethylmaleimide-sensitive factor attachment protein) receptor], syntaxin (syx), which eliminates neurotransmitter release, providing an inactive synapse similar to that in the gluR mutant. syx null mutants also have abnormal growth cone-like presynaptic terminals with less varicose structure (Yoshihara, 2005).

Because activity is required for synapse development, whether Syt 4-dependent vesicle fusion may be required, similar to its role in acute retrograde signaling during HFMR, was tested. Physiological analysis revealed that the amplitude of evoked currents in mutants lacking Syt 4 was moderately reduced compared with wild type, suggesting weaker synaptic function or development. Similar to the morphological phenotype of the gluR mutant, syt 4 null mutant embryos show defective presynaptic differentiation. Nerve terminals lacking Syt 4 display reduced varicose structure, whereas wild-type terminals have already formed individual varicosities at this stage of development. Postsynaptic expression with a UAS-syt 4 transgene rescues the physiological and morphological phenotypes. Syt 4 Ca2+-binding deficient mutant transgenes did not rescue either the morphological immaturity or the reduced amplitude of evoked currents, even though Syt 4 immunoreactivity at the postsynaptic compartment was restored by muscle-specific expression of the mutant syt 4 transgene, similar to the wild-type syt 4 transgene and endogenous Syt 4 immunoreactivity (Yoshihara, 2005).

Mammalian syt 4 was originally identified as an immediate-early gene that is transcriptionally up-regulated by nerve activity in certain brain regions (Vician, 1995). Therefore, gain-of-function phenotypes caused by postsynaptic Syt 4 overexpression were examined specifically in muscle cells to increase the probability of postsynaptic vesicle fusion. Syt 4 overexpression induced overgrowth of presynaptic terminals in mature third instar larvae, in contrast to overexpression of Syt 1, which does not traffic to Syt 4-containing postsynaptic vesicles. In addition to synaptic overgrowth, Syt 4 overexpression occasionally induced the formation of abnormally large varicosities. Postsynaptic overexpression of the Syt 4 Ca2+-binding mutant did not induce synaptic overgrowth, indicating that retrograde signaling by Syt 4 also requires Ca2+ binding to promote synaptic growth (Yoshihara, 2005).

To determine whether the cAMP-PKA pathway is important in activity-dependent synaptic growth, the effects of PKA on synaptic morphology were assayed. Expression of constitutively active PKA presynaptically using a motor neuron-specific Gal4 driver induced not only synaptic overgrowth but also larger individual varicosities in mature third instar larvae, similar to those induced by postsynaptic overexpression of Syt 4. These observations are consistent with the presynaptic overgrowth observed in the learning mutant, dunce, which disrupts the enzyme that degrades cAMP, and with studies in Aplysia implicating PKA in synaptic varicosity formation. The loss-of-function phenotype of PKA mutants (DC0B3) was characterized at the embryonic NMJ, to compare with gluR and syt 4 mutants. Presynaptic terminals in the DC0 mutant were morphologically aberrant, with abnormal growth cone-like features and less varicose structure. Postsynaptic expression of a constitutively active PKA transgene in the DC0 or syt 4 mutant backgrounds rescued the immature morphology, suggesting activation of PKA is downstream of Syt 4-dependent release of retrograde signals (Yoshihara, 2005).

Similar to the role of Syt 1-dependent synaptic vesicle fusion in triggering synaptic transmission at individual synapses, Syt 4-dependent vesicle fusion might trigger synapse-specific plasticity and growth. To test synapse specificity, advantage was taken of the specific properties of the Drosophila NMJ at muscle fibers 6 and 7, where two motorneurons innervate both muscle fibers 6 and 7 during development. Syt 4 was expressed specifically in embryonic muscle fiber 6 but not muscle fiber 7 by using the H94-Gal4 driver. If Syt 4-dependent retrograde signals induce general growth of the motorneuron, one would expect to see a proliferation of synapses on both muscle fibers. Alternatively, if Syt 4 promoted local synaptic growth, one would expect specific activation of synapse proliferation only on target muscle 6, releasing the Syt 4-dependent signal. UAS-syt 4 driven by H94-Gal4 increased innervation on muscle fiber 6 compared with that on muscle fiber 7 in third instar larvae. Control experiments with Syt 4 Ca2+-binding deficient mutant transgenes, or a transgene encoding Syt 1, did not result in proliferation. Thus, synaptic growth can be preferentially directed to specific postsynaptic targets where Syt 4-dependent retrograde signals predominate, allowing differential strengthening of active synapses via local rewiring (Yoshihara, 2005).

On the basis of these results, a local feedback model is proposed for activity-dependent synaptic plasticity and growth at Drosophila NMJs. Synapse-specific Ca2+ influx triggers postsynaptic vesicle fusion through Syt 4. Fusion of Syt 4-containing vesicles with the postsynaptic membrane releases locally acting retrograde signals that activate the presynaptic terminal, likely through the cAMP pathway. Active PKA then triggers cytoskeletal changes by unknown effectors to induce presynaptic growth and differentiation. Moreover, PKA is well known to facilitate neurotransmitter release directly, triggering a local synaptic enhancement of presynaptic release as shown in HFMR. Therefore, postsynaptic vesicular fusion might initiate a positive feedback loop, providing a localized activated synaptic state that can be maintained beyond the initial trigger (Yoshihara, 2005).

As a general mechanism for memory storage, Hebb postulated that potentiated synapses maintain an activated state until structural changes occur to consolidate alterations in synaptic strength (Hebb, 1949). The current results demonstrate that acute plasticity and synapse-specific growth require Syt 4-dependent retrograde signaling at Drosophila NMJs. The feedback mechanism described here could be a molecular basis for both input-specific postsynaptic tagging and an output-specific presynaptic mark or tag for long-lasting potentiation. The regenerative nature of a positive feedback signal allows individual synapses to be tagged in a discrete all-or-none manner until synaptic rewiring is completed. The synaptic tag is maintained as a large increase in miniature frequency at Drosophila NMJs, suggesting a previously unknown role for miniature release in neuronal function. The spatial resolution for input and output specificity would result from the accuracy insured by Ca2+-dependent vesicle fusion and subsequent diffusion, similar to the precision of presynaptic neurotransmitter release (Yoshihara, 2005).

Gating characteristics control glutamate receptor distribution and trafficking in vivo

Glutamate-releasing synapses dominate excitatory release in the brain. Mechanisms governing their assembly are of major importance for circuit development and long-term plasticity underlying learning and memory. AMPA/Kainate-type glutamate receptors (GluRs) are tetrameric ligand-gated ion channels that open their ion-conducting pores in response to binding of the neurotransmitter. Changes in subunit composition of postsynaptic GluRs are highly relevant for plasticity and development of glutamatergic synapses. To date, posttranslational modifications, mostly operating via the intracellular C-terminal domains (CTDs) of GluRs, are presumed to be the major regulator of trafficking. In recent years, structural and electrophysiological analyses have improved understanding of GluR gating mechanism. However, whether conformational changes subsequent to glutamate binding may per se be able to influence GluR trafficking has remained an unaddressed question. Using a Drosophila system allowing for extended visualization of GluR trafficking in vivo, this study provides evidence that mutations changing the gating behavior alter GluR distribution and trafficking. GluR mutants associated with reduced charge transfer segregated from coexpressed wild-type GluRs on the level of individual postsynaptic densities. Segregation was lost upon blocking of evoked glutamate release. Photobleaching experiments suggested increased mobility of mutants with reduced charge transfer, which accumulated prematurely during early steps of synapse assembly, but failed to further increase their level in accordance with assembly of the presynaptic scaffold. In summary, gating characteristics seem to be a new variable for the understanding of GluR trafficking relevant to both development and plasticity (Petzoldt, 2014).


Novel Functional Properties of Drosophila CNS Glutamate Receptors

Phylogenetic analysis reveals AMPA, kainate, and NMDA receptor families in insect genomes, suggesting conserved functional properties corresponding to their vertebrate counterparts. However, heterologous expression of the Drosophila kainate receptor DKaiR1D and the AMPA receptor DGluR1A revealed novel ligand selectivity at odds with the classification used for vertebrate glutamate receptor ion channels (iGluRs). DKaiR1D forms a rapidly activating and desensitizing receptor that is inhibited by both NMDA and the NMDA receptor antagonist AP5; crystallization of the KaiR1D ligand-binding domain reveals that these ligands stabilize open cleft conformations, explaining their action as antagonists. Surprisingly, the AMPA receptor DGluR1A shows weak activation by its namesake agonist AMPA and also by quisqualate. Crystallization of the DGluR1A ligand-binding domain reveals amino acid exchanges that interfere with binding of these ligands. The unexpected ligand-binding profiles of insect iGluRs allows classical tools to be used in novel approaches for the study of synaptic regulation. See the Video Abstract.

AMPA receptors: Developmental expression

In almost all nervous systems, rapid excitatory synaptic communication is mediated by a diversity of ionotropic glutamate receptors. In C. elegans, 10 putative ionotropic glutamate receptor subunits have been identified, a surprising number for an organism with only 302 neurons. Sequence analysis of the predicted proteins has identified two NMDA and eight non-NMDA receptor subunits. The relationship between the putative C. elegans and known vertebrate glutamate receptor subunits was analyzed by generating a neighbor-joining tree of receptor subunits. Clear groupings of the subunits can be distinguished. Thus, GLR-1 and GLR-2 group together and are most similar to rat AMPA receptors. GLR-3 and GLR-4, and more weakly GLR-5 and GLR-6, also group together. GLR-3-GLR-7 are clearly non-NMDA receptor subunits but cannot be obviously classified into pharmacological subtypes. GLR-8 is the farthest outlier and is even more divergent than glutamate receptors identified in Arabidopsis. NMR-1 is grouped with rat NR1 and Drosophila Nmdar1, and NMR-2 is grouped with rat NR2 receptors. The complete distribution of these subunits in the nervous system of C. elegans is described in this study. Receptor subunits were found almost exclusively in interneurons and motor neurons, but no expression was detected in muscle cells. Interestingly, some neurons express only a single subunit, suggesting that these may form functional homomeric channels. Conversely, interneurons of the locomotory control circuit (AVA, AVB, AVD, AVE, and PVC) coexpress up to six subunits, suggesting that these subunits interact to generate a diversity of heteromeric glutamate receptor channels that regulate various aspects of worm movement. Expression of these subunits in this circuit is differentially regulated by the homeodomain protein UNC-42 (Note: there is no known Drosophila homolog) and UNC-42 is also required for axonal pathfinding of neurons in the circuit. In wild-type worms, the axons of AVA, AVD, and AVE lie in the ventral cord, whereas in unc-42 mutants, the axons are anteriorly, laterally, or dorsally displaced, and the mutant worms have sensory and locomotor defects (Brockie, 2001).

Recording of glutamate-activated currents in membrane patches was combined with RT-PCR-mediated AMPA receptor (AMPAR) subunit mRNA analysis in single identified cells of rat brain slices. Analysis of AMPARs in principal neurons and interneurons of hippocampus and neocortex and in auditory relay neurons and Bergmann glial cells indicates that the GluR-B subunit in its flip version determines formation of receptors with relatively slow gating, whereas the GluR-D subunit promotes assembly of more rapidly gated receptors. The relation between Ca2+ permeability of AMPAR channels and the relative GluR-B mRNA abundance is consistent with the dominance of this subunit in determining the Ca2+ permeability of native receptors. The results suggest that differential expression of GluR-B and GluR-D subunit genes, as well as the splicing and editing of their mRNAs, account for the differences in gating and Ca2+ permeability of native AMPAR channels (Geiger, 1995).

The hypothesis that subtypes of glutamate receptors (GluRs) are differentially expressed during corticogenesis was tested. The neocortex of fetal sheep (term = approximately 145 days) was evaluated by immunoblotting and immunohistochemistry to determine the protein expression of AMPA receptors (GluR1, GluR2/GluR3 [GluR2/3], and GluR4); kainate (KA) receptors (GluR6/GluR7 [GluR6/7]), and a metabotropic GluR (mGluR5). AMPA/KA receptors and mGluR5 are expressed in neocortex by midgestation. GluR1 and mGluR5 expression increases progressively, with expression being maximal just before birth and then decreasing postnatally. GluR2/3 and GluR6/7 levels increase progressively during corticogenesis to reach adult levels near term. GluR4 is expressed at low levels during corticogenesis and in adult neocortex. The localizations of GluRs in the developing neocortex are distinct. Each GluR has a differential localization within the marginal zone, cortical plate, and subplate. GluR subtypes are expressed in laminar patterns before major cytoarchitectonic segregation occurs based on Nissl staining, although connectional patterns are emergent by midgestation based on labeling of corticostriatal projections with DiI. The GluR localizations change during cortical plate segregation, resulting in highly differential distributions in the neocortex at term. AMPA/KA receptors are expressed transiently in proliferative zones and in developing white matter. Oligodendrocytes in fetal brain express AMPA receptors. The expression of ion channel and metabotropic GluR subtypes is dynamic during corticogenesis, with subtype- and subunit-specific regulation occurring during the laminar segregation of the cortical plate and differentiation of the neocortex (Furuta, 1999).

Early in postnatal development, glutamatergic synapses transmit primarily through NMDA receptors. As development progresses, synapses acquire AMPA receptor function. The molecular basis of these physiological observations is not known. Single excitatory synapses were examined with immunogold electron-microscopic analysis of AMPA and NMDA receptors along with electrophysiological measurements. Early in postnatal development, a significant fraction of excitatory synapses have NMDA receptors and lack AMPA receptors. As development progresses, synapses acquire AMPA receptors with little change in NMDA receptor number. Thus, synapses with NMDA receptors but no AMPA receptors can account for the electrophysiologically observed 'silent synapse' (Petralia, 1999).

It has been suggested that some glutamatergic synapses lack functional AMPA receptors. Quantitative immunogold localization was used to determine the number and variability of synaptic AMPA receptors in the rat hippocampus. Three classes of synapses show distinct patterns of AMPA receptor content. Mossy fiber synapses on CA3 pyramidal spines and synapses on GABAergic interneurons are all immunopositive, have less variability, and contain 4 times as many AMPA receptors as synapses made by Schaffer collaterals on CA1 pyramidal spines and by commissural/ associational (C/A) terminals on CA3 pyramidal spines. Up to 17% of synapses in the latter two connections are immunonegative. After calibrating the immunosignal (1 gold = 2.3 functional receptors) at mossy synapses of a 17-day-old rat, the AMPA receptor content of C/A synapses on CA3 pyramidal spines is estimated to ranges from less than 3 to 140. A similar range is found in adult Schaffer collateral and C/A synapses (Nusser, 1998).

The superficial dorsal horn is a major site of termination of nociceptive primary afferents. Fast excitatory synaptic transmission in this region is mediated mainly by release of glutamate onto postsynaptic AMPA and NMDA receptors. NMDA receptors are known to be Ca2+-permeable and to provide synaptically localized Ca2+ signals that mediate short-term and long-term changes in synaptic strength. Less well known is a subpopulation of AMPA receptors that is Ca2+-permeable and has been shown to be synaptically localized on dorsal horn neurons in culture and expressed by dorsal horn neurons in situ. Kainate-induced cobalt uptake was used as a functional marker of neurons expressing Ca2+-permeable AMPA receptors and this was combined with markers of nociceptive primary afferents in the postnatal rat dorsal horn. Cobalt-positive neurons are located in lamina I and outer lamina II, a region strongly innervated by nociceptors. These cobalt-positive neurons colocalize with afferents labeled by LD2, and with the most dorsal region of capsaicin-sensitive and IB4- and LA4-positive afferents. In contrast, inner lamina II has a sparser distribution of cobalt-positive neurons. Some lamina I neurons expressing the NK1 receptor, the receptor for substance P, are also cobalt positive. These neurons are likely to be projection neurons in the nociceptive pathway. On the basis of all of these observations, it is proposed that Ca2+-permeable AMPA receptors are localized to mediate transmission of nociceptive information (Engelman, 1999).

The GluR2 subunit controls three key features of ion flux through the AMPA subtype of glutamate receptors-calcium permeability, inward rectification, and channel block by external polyamines, but whether each of these features is equally sensitive to GluR2 abundance is unknown. The relations among these properties were compared in native AMPA receptors expressed by acutely isolated hippocampal interneurons and in recombinant receptors expressed by Xenopus oocytes. The shape of current-voltage (I-V) relations between -100 and +50 mV for either recombinant or native AMPA receptors is well described by a Woodhull block model in which the affinity for internal polyamine varies over a 1000-fold range in different cells. In oocytes injected with mixtures of GluR2:non-GluR2 mRNA, the relative abundance of GluR2 required to reduce the log of internal blocker affinity by 50% is two- to fourfold higher than that needed to half-maximally reduce divalent permeability or channel block by external polyamines. Likewise, in interneurons the affinity of externally applied argiotoxin for its blocking site is a steep function of internal blocker affinity. These results indicate that the number of GluR2 subunits in AMPA receptors is variable in both oocytes and interneurons. More GluR2 subunits in an AMPA receptor are required to maximally reduce internal blocker affinity than to abolish calcium permeability or external polyamine channel block. Accordingly, single-cell RT-PCR shows that approximately one-half of the physiologically characterized interneurons exhibiting inwardly rectifying AMPA receptors express detectable levels of edited GluR2. Thus, the physiological effects of a moderate change in GluR2 relative abundance, such as occurs after ischemia or seizures or after chronic exposure to morphine, will be dependent on the ambient GluR2 level in a cell-specific manner (Washburn, 1997).

Formation of glutamatergic synapses entails development of 'silent' immature contacts into mature functional synapses. To determine how this transformation occurs, the development of neurotransmission was investigated at single synapses in vitro. Maturation of presynaptic function, assayed with endocytotic markers, follows accumulation of synapsin I. During this period, synaptic transmission is primarily mediated by activation of NMDA receptors, suggesting that most synapses are functionally silent. However, local glutamate application to silent synapses indicates that these synapses contain functional AMPA receptors, suggesting a possible presynaptic locus for silent transmission. Interference with presynaptic vesicle fusion by exposure to tetanus toxin reverts functional to silent transmission, implicating SNARE-mediated fusion as a determinant of the ratio of NMDA:AMPA receptor activation. This work reveals that functional maturation of synaptic transmission involves transformation of presynaptic silent secretion into mature synaptic transmitter release (Renger, 2001).

It is difficult to conclude that silent or AMPA-quiet synaptic transmission events were solely due to the absence or highly dynamic nature of AMPA receptors. Therefore, alternative explanations were explored that take into account both the immature nature of presynaptic terminals and the postsynaptic inclusion of AMPA and NMDA receptors during AMPA-quiet synaptic transmission. One possible mechanism is a change in the flux of transmitter release. It was found that the activation of AMPA receptors is significantly reduced when the speed of transmitter release is slowed. However, the concentration profile of neurotransmitter does not alter the amount of NMDA receptor activation. Although there are multiple possibilities that might affect the flux of transmitter, the preferred explanation is a change in the process of vesicular fusion during synaptic transmission. If the fusion pore conductance is reduced, it would slow the release of glutamate, and for presynaptic functional marker molecules like FM1-43, which are 4-fold larger than glutamate, it could prevent passage into the vesicle. This would explain the lack of FM staining, the relatively small amount of AMPA receptor-mediated current, and the increased rise time of NMDA currents at young synapses. The perturbation of vesicle fusion through TeNTx-treatment provided a strong test of this possibility. Following toxin treatment, mature synapses generally fail to label with FM dye, have higher frequencies of miniature and evoked AMPA-quiet transmission events, higher failure rates of transmission, and slowed NMDA receptor activation, reminiscent of young synapses. To conclude that maturation of the synaptic vesicle fusion process does underlie the conversion of AMPA-quiet to functional transmission, future experiments will have to directly monitor the developmental changes in the proteins involved in the formation of the synaptic vesicle fusion complex (Renger, 2001).

AMPA receptors: Alternative splicing

AMPA receptor channels mediate the fast component of excitatory postsynaptic currents in the central nervous system. Site-selective nuclear RNA editing controls the calcium permeability of these channels, and RNA editing at a second site is shown here to affect the kinetic aspects of these channels in rat brain. In three of the four AMPA receptor subunits (GluR-B, -C, and -D), intronic elements determine a codon switch [AGA (arginine) to GGA (glycine)] in the primary transcripts in a position termed the R/G site, which immediately precedes the alternatively spliced modules 'flip' and 'flop'. The extent of editing at this site progresses with brain development in a manner specific for subunit and splice form, and edited channels possess faster recovery rates from desensitization (Lomeli, 1994).

Glutamatergic transmission converging on calcium signaling plays a key role in dendritic differentiation. In early development, AMPA receptor (AMPAR) transcripts are extensively spliced and edited to generate subunits that differ in their biophysical properties. Each GluA gene is subject to alternative splicing into the flip and flop isoforms. Flip-containing receptors are more efficiently activated and desensitize with slower kinetics. During early development, flip variants are prominently expressed, but towards adulthood they become replaced by flop-containing subunits. Whether the various AMPAR subunits have specific roles in the context of structural differentiation is unclear. This study investigated the role of nine GluA variants and revealed a correlation between the expression of flip variants and the period of major dendritic growth. In interneurons, only GluA1(Q)-flip increased dendritic length and branching. In pyramidal cells, GluA2(Q)-flop, GluA2(Q)-flip, GluA3(Q)-flip and calcium-impermeable GluA2(R)-flip promoted dendritic growth, suggesting that flip variants with slower desensitization kinetics are more important than receptors with elevated calcium permeability. Imaging revealed significantly higher calcium signals in pyramidal cells transfected with GluA2(R)-flip as compared with GluA2(R)-flop, suggesting a contribution of voltage-activated calcium channels. Indeed, dendritic growth induced by GluA2(R)-flip in pyramidal cells was prevented by blocking NMDA receptors (NMDARs) or voltage-gated calcium channels (VGCCs), suggesting that they act downstream of AMPARs. Intriguingly, the action of GluA1(Q)-flip in interneurons was also dependent on NMDARs and VGCCs. Cell class-specific effects were not observed for spine formation, as GluA2(Q)-flip and GluA2(Q)-flop increased spine density in pyramidal cells as well as in interneurons. The results suggest that AMPAR variants expressed early in development are important determinants for activity-dependent dendritic growth in a cell type-specific and cell compartment-specific manner (Hamad, 2011).

AMPA receptors: Effects of RNA editing

RNA editing by site-selective deamination of adenosine to inosine alters codons and splicing in nuclear transcripts, and therefore protein function. ADAR2 (Drosophila homolog: ADAR) is a candidate mammalian editing enzyme that is widely expressed in brain and other tissues, but its RNA substrates are unknown. ADAR2-mediated RNA editing has been studied by generating mice that are homozygous for a targeted functional null allele. Editing in ADAR2-/- mice is substantially reduced at most of 25 positions in diverse transcripts; the mutant mice become prone to seizures and die young. The impaired phenotype appears to result entirely from a single underedited position, as it reverts to normal when both alleles for the underedited transcript are substituted exonically, with alleles encoding the edited version. The critical position specifies an ion channel determinant, the Q/R site, in AMPAreceptor GluR-B pre-messenger RNA. It is concluded that this transcript is the physiologically most important substrate of ADAR2 (Higuchi, 2000).

Mammalian transcripts that are known to be edited by site-selective adenosine deamination are expressed largely in brain: most encode subunits of ionotropic glutamate receptors (GluRs) that mediate fast excitatory neurotransmission. The only position edited to nearly 100% is the Q/R site of GluR-B, for which the mRNA contains an arginine (R) codon (CIG) in place of the genomic glutamine (Q) codon (CAG). The physiological importance of this codon substitution wrought by RNA editing has been revealed by early onset epilepsy and premature death of mice heterozygous for an intron-11-modified GluR-BECS allele with Q/R site-uneditable transcripts (Higuchi, 2000).

Heterozygous ADAR2+/- mice are phenotypically normal, but ADAR2-/- mice die between P0 and P20 and become progressively seizure-prone after P12, akin to GluR-B+/delta ECS mice. Therefore, this investigation focussed on the effect of ADAR2 deficiency on Q/R site editing of GluR-B pre-mRNA, the substrate for a nuclear RNA-dependent adenosine deaminase activity. As determined from cloned polymerase chain reaction with reverse transcription (RT-PCR) products from brain RNA6, Q/R site editing in primary GluR-B transcripts is tenfold lower in ADAR2-/- than in wild-type mice (10% compared with 98%). This identifies ADAR2 as the principal RNA-editing enzyme at the Q/R site. The remaining low level of Q/R site editing in GluR-B pre-mRNA cannot be mediated by the residual, enzymatically inactive, truncated ADAR2 protein, but is mediated by another ADAR, perhaps ADAR1, for which gene expression appeared unchanged in ADAR2 -/- mice (Higuchi, 2000).

The low extent of Q/R site editing of GluR-B pre-mRNA led to nuclear accumulation of incompletely processed primary GluR-B transcripts and to a fivefold reduction in GluR-B mRNA, as assessed by RNase protection and quantitative RT-PCR. The increased level of intron 11-containing GluR-B transcripts and the decrease in GluR-B mRNA are easily visualized by in situ hybridization. Editing is thus a prerequisite for efficient splicing and processing of the pre-mRNA. The edited GluR-B transcripts are preferentially spliced, as revealed by a shift in Q/R site editing from 10% to 40% when comparing intron-11-containing transcripts with GluR-B mRNA. A defect in transcript processing caused by the interaction of the residual truncated ADAR2 protein with RNA can be excluded because GluR-B pre-mRNA accumulation is also observed in ADAR2+/+ mice expressing the Q/R site-uneditable GluR-BdeltaECS allele (Higuchi, 2000).

AMPA-type glutamate receptors (AMPARs) play a major role in excitatory synaptic transmission and plasticity. Channel properties are largely dictated by their composition of the four subunits, GluR1-4 (or A-D). AMPAR assembly and subunit stoichiometry are determined by RNA editing in the pore loop. Editing at the GluR2 Q/R site is specific for GluR2 since GluR1, -3, and -4 carry a Gln (Q) at this critical, pore-lining position. The vast majority of GluR2 (>99%) in adult brain is edited to Arg; editing at this site is very efficient and crucial for brain development and function. Editing at the GluR2 Q/R site regulates AMPAR assembly at the step of tetramerization. Specifically, edited R subunits are largely unassembled and ER retained, whereas unedited Q subunits readily tetramerize and traffic to synapses. This assembly mechanism restricts the number of the functionally critical R subunits in AMPAR tetramers. Therefore, a single amino acid residue affects channel composition and, in turn, controls ion conduction through the majority of AMPARs in the brain (Greger, 2003).

What are the functional implications of these finds for AMPAR transmission? Subunit composition is a major determinant for AMPAR conductance properties. In brain, the majority of AMPAR transmission is functionally dominated by the GluR2 subunit, which renders receptors Ca2+ impermeable, alters their voltage sensitivity, and reduces conductance. Only a subset of interneurons express low levels of GluR2, and therefore, Ca2+-permeable AMPARs. It has also been shown that the number of GluR2 subunits in AMPAR tetramers can vary (at least in interneurons), and that transmission properties are differently affected by GluR2 abundance. GluR2 subunit numbers in tetramers may be subject to regulation. Repetitive synaptic activity results in a switch from GluR2-lacking to GluR2-containing AMPARs in cerebellar stellate cells. Similarly, activity may cause increased inclusion of GluR2 subunits into tetramers. Such a mechanism would protect neurons from excessive Ca2+ influx through AMPARs. However, only a limited number of GluR2 subunits can be included into a tetramer. The assembly rules described here will impair the formation of GluR2 homomers, and, by extension, restrict GluR2 numbers in AMPAR tetramers. Occlusion of GluR2 homomers could be crucial as such channels display a very small single-channel conductance, and conduct anions. The existence of such channels would decrease the efficiency of a synapse, and barely contribute to the generation of an EPSP (excitatory postsynaptic potential) (Greger, 2003).

Calcium-permeable AMPA receptors containing Q/R-unedited GluR2 direct human neural progenitor cell differentiation to neurons

Calcium-permeable AMPA receptors have been identified on human neural progenitor cells (NPCs) and present a physiological role in neurogenesis. RNA editing of the GluR2 subunit at the Q/R site is responsible for making most AMPA receptors impermeable to calcium. Because a single-point mutation could eliminate the need for editing at the Q/R site and Q/R-unedited GluR2 exists during embryogenesis, the Q/R-unedited GluR2 subunit presumably has some important actions early in development. Using calcium imaging, this study found that NPCs contain calcium-permeable AMPA receptors, whereas NPCs differentiated to neurons and astrocytes express calcium-impermeable AMPA receptors. RTPCR and BbvI digestion were used to demonstrate that NPCs contain Q/R-unedited GluR2, and differentiated cells contain Q/R-edited GluR2 subunits. This is consistent with the observation that the nuclear enzyme responsible for Q/R-editing, adenosine deaminase (ADAR2), is increased during differentiation. Activation of calcium-permeable AMPA receptors induces NPCs to differentiate to the neuronal lineage and increases dendritic arbor formation in NPCs differentiated to neurons. AMPA-induced differentiation of NPCs to neurons is abrogated by overexpression of ADAR2 in NPCs. This elucidates the role of AMPA receptors as inductors of neurogenesis and provides a possible explanation for why the Q/R editing process exists (Whitney, 2008).

Pin1 and WWP2 regulate GluR2 Q/R site RNA editing by ADAR2 with opposing effects

ADAR2 catalyses the deamination of adenosine to inosine at the GluR2 Q/R site in the pre-mRNA encoding the critical subunit of AMPA receptors. Among ADAR2 substrates this is the vital one as editing at this position is indispensable for normal brain function. However, the regulation of ADAR2 post-translationally remains to be elucidated. This study demonstrates that the phosphorylation-dependent prolyl-isomerase Pin1 interacts with ADAR2 and is a positive regulator required for the nuclear localization and stability of ADAR2. Pin1(-/-) mouse embryonic fibroblasts show mislocalization of ADAR2 in the cytoplasm and reduced editing at the GluR2 Q/R and R/G sites. The E3 ubiquitin ligase WWP2 plays a negative role by binding to ADAR2 and catalysing its ubiquitination and subsequent degradation. Therefore, ADAR2 protein levels and catalytic activity are coordinately regulated in a positive manner by Pin1 and negatively by WWP2 and this may have downstream effects on the function of GluR2. Pin1 and WWP2 also regulate the large subunit of RNA Pol II, so these proteins may also coordinately regulate other key cellular proteins (Marcucci, 2011).

AMPA receptors: Interaction with antagonists

AMPA-type glutamate receptors are specifically inhibited by the noncompetitive antagonists GYKI-53655 and CP-465,022, which act through sites and mechanisms that are not understood. Using receptor mutagenesis, it was found that these antagonists bind at the interface between the S1 and S2 glutamate binding core and channel transmembrane domains, specifically interacting with S1-M1 and S2-M4 linkers, thereby disrupting the transduction of agonist binding into channel opening. It was also found that the antagonists' affinity is higher for agonist-unbound receptors than for activated nondesensitized receptors, further depending on the level of S1 and S2 domain closure. These results provide evidence for substantial conformational changes in the S1-M1 and S2-M4 linkers following agonist binding and channel opening, offering a conceptual frame to account for noncompetitive antagonism of AMPA receptors (Balannik, 2005).

AMPARs assemble as tetramers in various combinations of four homologous subunits, termed GluR1-4 (or GluR-A to -D). Like all iGluR subunits, the AMPAR subunits share a modular design consisting of an extracellular N-terminal oligomerization domain (NTD); an extracellular agonist-binding domain formed by two segments, S1 and S2; a channel-forming domain consisting of three transmembrane domains, M1, M3, and M4 and a reentrant loop M2; and an intracellular C-terminal trafficking and anchoring domain (CTD). Currently, structural data at atomic resolution are available only for the S1 and S2 domains, of which the first and most extensively characterized is that derived from the AMPAR GluR2. Collectively, it has been shown that S1 and S2, which in the intact receptors are separated by the membrane regions M1 to M3, fold in a special manner, creating two globular domains (D1 and D2). Glutamate first docks in D1, which then promotes the rotation of D2 and closure of the binding cleft. Full agonists like glutamate and AMPA induce a large movement resulting in full activation. Partial agonists like kainate induce an intermediate closure and partial activation, and competitive antagonists like DNQX promote only a small extent of domain closure that is insufficient to trigger ion channel gating. Several studies have provided evidence that the agonist binding domains assemble as dimers, and the mechanism of desensitization has been further defined as a rearrangement of the dimer interface. Therefore, the idea that agonist binding to S1 and S2 evokes significant conformational changes in the extracellular domains leading to channel opening is widely accepted. However, the mechanism by which these conformational changes are transduced to channel gating is still unclear. Gating is likely to involve the linker regions between the agonist-binding and channel-forming domains, namely the S1-M1, S2-M3, and S2-M4 linkers. These regions are postulated to contribute to an extended mass at the bottom of D2 and are therefore likely to be coupled to the movement of D2 upon agonist binding. So far, experimental evidence for such conformational rearrangements is limited to the M3 linker (Balannik, 2005).

There are a number of pharmacological agents that affect AMPAR function through interactions outside of the agonist-binding domain. Thus, investigating the means by which binding of these ligands modulate channel gating may provide additional insight into mechanisms of receptor function. Toward this end, the site of interaction with AMPAR of two selective noncompetitive AMPAR antagonists, GYKI-53655 (GYKI) and CP-465,022 (CP), was investigated. GYKI belongs to a family of 2,3-benzodiazepines, and it is a more potent and selective analog of GYKI-52466, the first identified AMPAR noncompetitive antagonist. CP is a derivative of piraquilone and is ~100-fold more potent than GYKI on hippocampal neurons. Radioligand-binding assays suggested that the binding sites of these two compounds overlap with one another but that this site is distinct from the agonist-binding site. These antagonists are not open-channel blockers nor do they affect channel desensitization, suggesting a mechanism of action not involving binding to the channel pore. However, there is an allosteric interaction between GYKI and the inhibitor of desensitization cyclothiazide (CTZ), suggesting that the binding site for GYKI is affected by gating, although the molecular mechanism for this interaction is not known (Balannik, 2005).

This study shows that GYKI and CP bind with different affinity to different gating states of the AMPAR. The highest affinity is for the closed state, most likely the resting rather than the desensitized state, and the lowest is for the open state, further depending on agonist efficacy to open the channel. Using AMPA/kainate chimeras, it was shown that GYKI and CP bind at the linker regions between S1 and S2 and the channel, specifically interacting with S1-M1 and S2-M4. Therefore, the change in antagonist binding affinity upon gating is indicative of substantial conformational changes in these linkers following agonist binding and channel opening. A model in which these noncompetitive inhibitors constrain the movements of these linkers provides the insight into the way agonist binding is transduced to channel gating through the linker regions. As such, this study provides a potential template for rational drug design (Balannik, 2005).

AMPA receptors: Modification by phosphorylation

The phosphorylation of the glutamate receptor subunit GluR1 has been characterized using biochemical and electrophysiological techniques. GluR1 is phosphorylated on multiple sites that are all located on the C-terminus of the protein. Cyclic AMP-dependent protein kinase specifically phosphorylates SER-845 of GluR1 in transfected HEK cells and in neurons in culture. Phosphorylation of this residue results in a 40% potentiation of the peak current through GluR1 homomeric channels. In addition, protein kinase C specifically phosphorylates Ser-831 of GluR1 in HEK-293 cells and in cultured neurons. These results are consistent with the recently proposed transmembrane topology models of glutamate receptors, in which the C-terminus is intracellular. In addition, the modulation of GluR1 by PKA phosphorylation of Ser-845 suggests that phosphorylation of this residue may underlie the PKA-induced potentiation of AMPA receptors in neurons (Roche, 1996).

Long-term potentiation (LTP), a cellular model of learning and memory, requires calcium-dependent protein kinases. Induction of LTP increases the phosphorus-32 labeling of AMPA-type glutamate receptors (AMPA-Rs), which mediate rapid excitatory synaptic transmission. This AMPA-R phosphorylation appears to be catalyzed by Ca2+- and calmodulin-dependent protein kinase II (CaM-KII): (1) it correlates with the activation and autophosphorylation of CaM-KII, (2) it is blocked by the CaM-KII inhibitor KN-62, and (3) its phosphorus-32 peptide map is the same as that of GluR1 coexpressed with activated CaM-KII in HEK-293 cells. This covalent modulation of AMPA-Rs in LTP provides a postsynaptic molecular mechanism for synaptic plasticity (Barria, 1997a).

Ca2+/CaM-dependent protein kinase II (CaM-KII) can phosphorylate and potentiate responses of AMPA-type glutamate receptors in a number of systems: recent studies implicate this mechanism in long term potentiation, a cellular model of learning and memory. In this study, this CaM-KII regulatory site has been identified using deletion and site-specific mutants of glutamate receptor 1 (GluR1). Only mutations affecting Ser831 alter the 32P peptide maps of GluR1 from HEK-293 cells co-expressing an activated CaM-KII. Likewise, when CaM-KII is infused into cells expressing GluR1, the Ser831 to Ala mutant fails to show potentiation of the GluR1 current. The Ser831 site is specific to GluR1, and CaM-KII does not phosphorylate or potentiate current in cells expressing GluR2, emphasizing the importance of the GluR1 subunit in this regulatory mechanism. Because Ser831 has previously been identified as a protein kinase C phosphorylation site, this raises the possibility of synergistic interactions between CaM-KII and protein kinase C in regulating synaptic plasticity (Barria, 1997b).

Brief bath application of N-methyl-D-aspartate (NMDA) to hippocampal slices produces long-term synaptic depression (LTD) in CA1 that is (1) sensitive to postnatal age, (2) saturable, (3) induced postsynaptically, (4) reversible, and (5) not associated with a change in paired pulse facilitation. Chemically induced LTD (chem-LTD) and homosynaptic LTD are mutually occluding, suggesting a common expression mechanism. Using phosphorylation site-specific antibodies, induction of chem-LTD is found to produce a persistent dephosphorylation of the GluR1 subunit of AMPA receptors at serine 845, a cAMP-dependent protein kinase (PKA) substrate, but not at serine 831, a substrate of protein kinase C (PKC) and calcium/calmodulin-dependent protein kinase II (CaMKII). These results suggest that dephosphorylation of AMPA receptors is an expression mechanism for LTD and indicate an unexpected role for PKA in the postsynaptic modulation of excitatory synaptic transmission (Lee, 1998).

Modulation of AMPA-type glutamate channels is important for synaptic plasticity. Physiological evidence is provided that the activity of AMPA channels is regulated by protein phosphatase 1 (PP-1) in neostriatal neurons. Two distinct molecular mechanisms of this regulation are identified. One mechanism involves control of PP-1 catalytic activity by DARPP-32, a dopamine- and cAMP-regulated phosphoprotein highly enriched in neostriatum. The other involves binding of PP-1 to spinophilin, a protein that colocalizes PP-1 with AMPA receptors in postsynaptic densities. The results suggest that regulation of anchored PP-1 is important for AMPA-receptor-mediated synaptic transmission and plasticity (Yan, 1999).

Hippocampal N-methyl-D-aspartate (NMDA) receptor-dependent long-term synaptic depression (LTD) is associated with a persistent dephosphorylation of the GluR1 subunit of AMPA receptors at a site (Ser-845) phosphorylated by cAMP-dependent protein kinase (PKA). In the present study, it is shown that dephosphorylation of a postsynaptic PKA substrate may be crucial for LTD expression. PKA activators inhibit both AMPA receptor dephosphorylation and LTD. Injection of a cAMP analog into postsynaptic neurons prevents LTD induction and reverses previously established homosynaptic LTD without affecting baseline synaptic transmission. Moreover, infusing a PKA inhibitor into postsynaptic cells produces synaptic depression that occludes homosynaptic LTD. These findings suggest that dephosphorylation of a PKA site on AMPA receptors may be one mechanism for NMDA receptor-dependent homosynaptic LTD expression (Kameyama, 1998).

Modulation of postsynaptic AMPA receptors in the brain by phosphorylation may play a role in the expression of synaptic plasticity at central excitatory synapses. It is known from biochemical studies that GluR1 AMPA receptor subunits can be phosphorylated within their C terminal by cAMP-dependent protein kinase A (PKA), which is colocalized with the phosphatase calcineurin (i.e., phosphatase 2B). The effect of PKA and calcineurin has been studied on the time course, peak open probability, and single-channel properties of glutamate evoked responses for neuronal AMPA receptors and homomeric GluR1(flip) receptors recorded in outside-out patches. Inclusion of purified catalytic subunit Calpha-PKA in the pipette solution increases neuronal AMPA receptor P(O,PEAK) compared with recordings made with calcineurin included in the pipette. Similarly, Calpha-PKA increases peak open probability for recombinant GluR1 receptors compared with patches excised from cells cotransfected with a cDNA encoding the PKA peptide inhibitor PKI or patches with calcineurin included in the pipette. Neither PKA nor calcineurin alters the amplitude of single-channel subconductance levels, weighted mean unitary current, mean channel open period, burst length, or macroscopic response waveform for recombinant GluR1 receptors. Substitution of an amino acid at the PKA phosphorylation site (S845A) on GluR1 eliminates the PKA-induced increase in peak open probability, whereas the mutation of a Ca(2+), calmodulin-dependent kinase II and PKC phosphorylation site (S831A) is without effect. These results suggest that AMPA receptor peak response open probability can be increased by PKA through phosphorylation of GluR1 Ser845 (Banke, 2000).

Bidirectional changes in the efficacy of neuronal synaptic transmission, such as hippocampal long-term potentiation (LTP) and long-term depression (LTD), are thought to be mechanisms for information storage in the brain. LTP and LTD may be mediated by the modulation of AMPA receptor phosphorylation. LTP and LTD reversibly modify the phosphorylation of the AMPA receptor GluR1 subunit. However, contrary to the hypothesis that LTP and LTD are the functional inverse of each other, LTP and LTD are associated with the phosphorylation and dephosphorylation, respectively, of distinct GluR1 phosphorylation sites. Moreover, the site modulated depends on the stimulation history of the synapse. LTD induction in naive synapses dephosphorylates the major cyclic-AMP-dependent protein kinase (PKA) site, whereas in potentiated synapses the major calcium/calmodulin-dependent protein kinase II (CaMKII) site is dephosphorylated. Conversely, LTP induction in naive synapses and depressed synapses increases phosphorylation of the CaMKII site and the PKA site, respectively. LTP is differentially sensitive to CaMKII and PKA inhibitors depending on the history of the synapse. These results indicate that AMPA receptor phosphorylation is critical for synaptic plasticity, and that identical stimulation conditions recruit different signal-transduction pathways depending on synaptic history (Lee, 2000).

A chemically induced form of LTD (chemLTD) is associated with persistent dephosphorylation of GluR1 at a PKA phosphorylation site, Ser 845. The chemLTD approach was designed to maximize the number of affected synapses, thereby increasing the probability of detecting biochemical changes. The sensitivity of the chemLTD assay could be increased to detect small changes in GluR1 phosphorylation on Ser 831 and Ser 845 in single hippocampal slices after the synaptic induction of LTP and LTD. To analyse phosphorylation of GluR1 during synaptic plasticity, two hippocampal slices were placed in the same recording chamber and extracellular field potentials were recorded simultaneously in the CA1 dendritic region of both slices upon stimulation of the Schaffer collaterals. After collecting a stable baseline, stimulation to one slice was turned off while the other slice received low-frequency stimulation (LFS; 1 Hz, 900 pulses). After the LFS, the stimulation was returned to the baseline frequency, and stimulation to the control slice was resumed. The slice that received LFS showed homosynaptic LTD, but there was no significant change in synaptic strength in the control slice. One hour after the induction of LTD the slices were collected and frozen immediately on dry ice. Phosphorylation states of GluR1 at Ser 831 and Ser 845 in the control and LTD slices were then examined by quantitative immunoblotting (Lee, 2000).

Like chemLTD, homosynaptic LTD produces specific dephosphorylation of Ser 845, whereas there is no significant change in phosphorylation of Ser 831. The dephosphorylation at Ser 845 could be detected as early as 30 min after the onset of LFS. As expected, the magnitude of the change in GluR1 phosphorylation following homosynaptic LTD is smaller than that reported for chemLTD, because only a small percentage of synapses in the slice are depressed during LFS-induced homosynaptic LTD. However, this effect is reproducible and statistically significant. In addition, there is no dephosphorylation in slices that do not exhibit LTD after LFS, either because induction of LTD fails or because the NMDA receptor antagonist AP5 was applied. Collectively, these results confirm that chemLTD and homosynaptic LTD share similar downstream expression mechanism(s) (Lee, 2000).

Homosynaptic LTD is dependent on postsynaptic protein phosphatase activity. To test whether the dephosphorylation of GluR1 following LTD is also sensitive to protein phosphatase inhibitors, the control and experimental slices were incubated in okadaic acid for 2-3 h and transferred to the recording chamber. Okadaic acid abolishes LTD. Moreover, pretreatment of slices with okadaic acid blocks the LFS-induced dephosphorylation of Ser 845. In contrast, induction of LTD in slices kept in vehicle solution still produces significant dephosphorylation at Ser 845. This result indicates that protein phosphatase 1/2A may be critical for the LFS-induced dephosphorylation of GluR1 at Ser 845 (Lee, 2000).

LTP is probably associated with an increase in phosphorylation of GluR1 by CaMKII. A test was performed to see whether the same changes could be detected using phosphorylation-site-specific antibodies. Slices that received theta burst stimulation (TBS) show LTP, and a significant increase in GluR1 phosphorylation at Ser 831 (CaMKII/PKC site) at both 30 min and 1 h after TBS. Phosphorylation of Ser 845 (PKA site) is not significantly increased 30 or 60 min after LTP induction. In slices that do not exhibit LTP, GluR1 phosphorylation does not change. These results indicate that the early phase of LTP is dependent on postsynaptic CaMKII activation and that phosphorylation of GluR1 by CaMKII increases following LTP5 (Lee, 2000).

Thus the data indicate that LTP and LTD may be associated with changes in phosphorylation of GluR1 at different sites. LTD is associated with dephosphorylation at Ser 845, a PKA site, whereas LTP is associated with increased phosphorylation of Ser 831, a CaMKII site. This indicates that although induction of LTP and LTD results in the bidirectional control of AMPA-receptor phosphorylation, LTP and LTD do not regulate phosphorylation of the same site. Since LTP and LTD are reversible processes, a test of what happens to GluR1 phosphorylation after the reversal of LTP with LFS ('depotentiation') and the reversal of LTD with TBS ('de-depression') was performed (Lee, 2000).

To examine depotentiation both slices in the recording chamber were subjected to TBS simultaneously: this results in LTP. After 30 min of recording, LFS was given to one of the slices. This stimulation depresses the synaptic strength back to the baseline level while the synaptic response stays potentiated in slices that only receive TBS. There is significant dephosphorylation of Ser 831 in depotentiated slices but no significant change in Ser 845. This result indicates that the same stimulation protocol (LFS) results in the dephosphorylation of different sites on GluR1 depending on the previous experience of the synapse (whether it had previously undergone LTP) (Lee, 2000).

To examine de-depression, LTD was induced in two slices in the recording chamber; 30 min later TBS was delivered to one of the slices. The synaptic response in slices that received TBS was potentiated back to around the baseline response whereas the synaptic strength in slices that received only LFS stayed depressed throughout the experiment. There is no significant change in phosphorylation of Ser 831 (CaMKII/PKC site) following de-depression. Interestingly, the de-depression increases phosphorylation at Ser 845 (PKA site). These results indicate that, like LTD-inducing stimuli, LTP-inducing stimuli result in the differential regulation of GluR1 phosphorylation depending on the previous experience of the synapse. TBS delivered to 'naive' synapses causes phosphorylation of a CaMKII site, but TBS delivered to previously depressed synapses causes phosphorylation of a PKA site (Lee, 2000).

From these results it has been predicted that the synaptic potentiation due to changes in AMPA receptor phosphorylation following TBS-induced de-depression or TBS-induced LTP should be differentially sensitive to CaMKII and PKA inhibitors. To test this hypothesis, two-pathway experiments were performed in which the effects of TBS could be compared on naive and previously depressed synaptic inputs. Under control conditions, TBS produces the same amount of synaptic potentiation regardless of the initial state of the synapses, and both LTP and de-depression are prevented when NMDA receptors are blocked. However, when slices are incubated in the selective CaMKII inhibitor KN-93, TBS produces significantly less de novo potentiation than de-depression (measured 50 min after TBS). Conversely, when PKA is transiently inhibited at the time of tetanic stimulation using the selective inhibitor KT5720, TBS produced significantly less de-depression than de novo LTP (measured 50 min after TBS). Thus, the relative contributions of CaMKII and PKA to TBS-induced potentiation vary depending on the initial state of the synapse, as predicted. This conclusion agrees with a previous report suggesting that there are different LTP-expression mechanisms in naive versus depressed synapses in PKC- knockout mice (Lee, 2000).

These results show that bidirectional changes in synaptic function are associated with the reversible regulation of AMPA receptor phosphorylation. However, the phosphorylation or dephosphorylation of GluR1 occurs on distinct sites, depending on the past experience of the synapse. Although the mechanism of this differential phosphorylation remains to be determined, it may involve differential activation of protein kinases and protein phosphatases in the potentiated, naive and depressed synapses. This differential bidirectional regulation of the phosphorylation of AMPA receptors has significant computational implications for synaptic plasticity, since the results indicate that there may be at least three states of AMPA receptor activity with limited and regulated transitions between the states (Lee, 2000).

Although phosphorylation of the GluR1 subunit on either Ser 831 or Ser 845 potentiates AMPA receptor function, it appears to do so through distinct biophysical mechanisms. PKA phosphorylation of Ser 845 regulates the open channel probability of AMPA receptors, whereas CaMKII phosphorylation regulates the apparent single-channel conductance of the receptor. Note that single-channel conductance increases with LTP. Regulation of phosphorylation at these two sites should contribute to the changes in the efficacy of synaptic transmission that are observed during LTP and LTD. However, the trafficking and synaptic targeting of AMPA receptors may also be important in LTP and LTD. The role of phosphorylation of the GluR1 subunit in the regulation of AMPA receptor synaptic targeting is unclear. It is possible that phosphorylation of these sites regulates synaptic trafficking of the AMPA receptor in addition to regulating channel function. Alternatively, the synaptic trafficking of AMPA receptors may be regulated through a distinct mechanism, occurring with and complementing the modification of channel properties. The findings that different phosphorylation sites and signal-transduction pathways contribute to bidirectional synaptic modifications in the hippocampus provide evidence for unexpected complexity in the control of the gain of excitatory synaptic transmission (Lee, 2000).

Cerebellar LTD requires activation of PKC and is expressed, at least in part, as postsynaptic AMPA receptor internalization. AMPA receptor internalization requires clathrin-mediated endocytosis and depends upon the carboxy-terminal region of GluR2/3. Phosphorylation of Ser-880 in this region by PKC differentially regulates the binding of the PDZ domain-containing proteins GRIP/ABP (See Drosophila Grip) and PICK1. Peptides, corresponding to the phosphorylated and dephosphorylated GluR2 carboxy-terminal PDZ binding motif, were perfused in cerebellar Purkinje cells grown in culture. Both the dephospho form (which blocks binding of GRIP/ABP and PICK1) and the phospho form (which selectively blocks PICK1) attenuate LTD induction by glutamate/depolarization pairing, as do antibodies directed against the PDZ domain of PICK1. These findings indicate that expression of cerebellar LTD requires PKC-regulated interactions between the carboxy-terminal of GluR2/3 and PDZ domain-containing proteins (Xia, 2000).

AMPA receptors: Modification by palmitoylation

Modification of AMPA receptor function is a major mechanism for the regulation of synaptic transmission and underlies several forms of synaptic plasticity. Post-translational palmitoylation is a reversible modification that regulates localization of many proteins. Palmitoylation of the AMPA receptor regulates receptor trafficking. All AMPA receptor subunits are palmitoylated on two cysteine residues in their transmembrane domain (TMD) 2 and in their C-terminal region. Palmitoylation on TMD 2 is upregulated by the palmitoyl acyl transferase GODZ and leads to an accumulation of the receptor in the Golgi and a reduction of receptor surface expression. C-terminal palmitoylation decreases interaction of the AMPA receptor with the 4.1N protein and regulates AMPA- and NMDA-induced AMPA receptor internalization. Moreover, depalmitoylation of the receptor is regulated by activation of glutamate receptors. These data suggest that regulated palmitoylation of AMPA receptor subunits modulates receptor trafficking and may be important for synaptic plasticity (Hayashi, 2005).

AMPA receptors: Subunit constitution

The precise subunit composition of synaptic ionotropic receptors in the brain is poorly understood. This information is of particular importance with regard to AMPA-type glutamate receptors, the multimeric complexes assembled from GluA1-A4 subunits; the trafficking of these receptors into and out of synapses is proposed to depend upon the subunit composition of the receptor. This study reports a molecular quantification of synaptic AMPA receptors (AMPARs) by employing a single-cell genetic approach coupled with electrophysiology in hippocampal CA1 pyramidal neurons. In contrast to prevailing views, it was found that GluA1A2 heteromers are the dominant AMPARs at CA1 cell synapses (approximately 80%). In cells lacking GluA1, -A2, and -A3, synapses are devoid of AMPARs, yet synaptic NMDA receptors (NMDARs) and dendritic morphology remain unchanged. These data demonstrate a functional dissociation of AMPARs from trafficking of NMDARs and neuronal morphogenesis. This study provides a functional quantification of the subunit composition of AMPARs in the CNS and suggests novel roles for AMPAR subunits in receptor trafficking (Lu, 2009).

The subunit composition of most ionotropic neurotransmitter receptors in the CNS has not been precisely determined. For the AMPA subtype of glutamate receptor, this is a particularly important problem. Recent evidence suggests that the subunit composition of AMPARs determines not only their biophysical properties but their activity-dependent trafficking to the synapse as well. Thus a rigorous quantitative description of the subunit composition of AMPARs is a prerequisite for understanding their roles in both the maintenance of synaptic transmission and synaptic plasticity. By using a conditional KO approach, each of the AMPAR subunits, both individually and in combination, were selectively deleted in a subset of CA1 hippocampal pyramidal cells. Simultaneous whole-cell recording from a gene-deleted cell and a neighboring control cell was used to quantify the changes induced by these genetic manipulations. The main results of this study are as follows. (1) All surface AMPARs contain GluA2 on CA1 pyramidal neurons. (2) GluA1, GluA2, and GluA3 fully account for the AMPARs on these neurons. (3) About 80% of synaptic AMPARs and >95% of extrasynaptic AMPARs are GluA1A2 heteromers, and most of the remaining receptors are GluA2A3 heteromers. (4) Aberrant homomeric GluA1, GluA2, and GluA3 receptors are capable of forming, depending on the deletion, but are unlikely to contribute significantly to normal AMPAR EPSCs on these neurons. This indicates that there is a hierarchy in the subunit assembly process, with GluA2-containing receptor complexes strongly preferred over other combinations. (5) No detectable changes in NMDAR EPSCs, spine morphology, or presynaptic properties were observed following the removal of all surface AMPARs. These findings provide new insight concerning the roles of AMPARs in neuronal physiology and morphology (Lu, 2009).

AMPA receptors: Three dimensional structure

Crystal structures of the GluR2 ligand binding core (S1S2) have been determined in the apo state and in the presence of the antagonist DNQX, the partial agonist kainate, and the full agonists AMPA and glutamate. The domains of the S1S2 ligand binding core are expanded in the apo state and contract upon ligand binding with the extent of domain separation decreasing in the order of apo>DNQX>kainate>glutamate approximately equal to AMPA. These results suggest that agonist-induced domain closure gates the transmembrane channel and the extent of receptor activation depends upon the degree of domain closure. AMPA and glutamate also promote a 180° flip of a trans peptide bond in the ligand binding site. The crystal packing of the ligand binding cores suggests modes for subunit-subunit contact in the intact receptor and mechanisms by which allosteric effectors modulate receptor activity (Armstrong, 2000).

Function of Stargazin-like protein (STG-1) and SOL-1, a transmembrane CUB-domain protein, in C. elegans

Ionotropic glutamate receptors (iGluRs) mediate most excitatory synaptic signalling between neurons. Binding of the neurotransmitter glutamate causes a conformational change in these receptors that gates open a transmembrane pore through which ions can pass. The gating of iGluRs is crucially dependent on a conserved amino acid that was first identified in the 'lurcher' ataxic mouse. Through a screen for modifiers of iGluR function in a transgenic strain of C. elegans expressing a GLR-1 subunit containing the lurcher mutation, suppressor of lurcher (sol-1) was identified. This gene encodes a transmembrane protein that is predicted to contain four extracellular beta-barrel-forming domains known as CUB domains. SOL-1 and GLR-1 are colocalized at the cell surface and can be co-immunoprecipitated. By recording from neurons expressing GLR-1, it is shown that SOL-1 is an accessory protein that is selectively required for glutamate-gated currents. It is proposed that SOL-1 participates in the gating of non-NMDA iGluRs, thereby providing a previously unknown mechanism of regulation for this important class of neurotransmitter receptor (Zheng, 2004).

AMPA receptors (AMPARs) are a major subtype of ionotropic glutamate receptors (iGluRs) that mediate rapid excitatory synaptic transmission in the vertebrate brain. Putative AMPARs are also expressed in the nervous system of invertebrates. In C. elegans, the GLR-1 receptor subunit is expressed in neural circuits that mediate avoidance behaviors and is required for glutamate-gated current in the AVA and AVD interneurons. Glutamate-gated currents can be recorded from heterologous cells that express vertebrate AMPARs; however, when C. elegans GLR-1 is expressed in heterologous cells, little or no glutamate-gated current is detected. This finding suggests that other receptor subunits or auxiliary proteins are required for function. This study identifies Ce STG-1, a C. elegans stargazin-like protein, and shows that expression of Ce STG-1 together with GLR-1 and the CUB-domain protein SOL-1 reconstitutes glutamate-gated currents in Xenopus oocytes. Ce STG-1 and homologues cloned from Drosophila (Dro STG1; CG33670) and Apis mellifera (Apis STG1) have evolutionarily conserved functions and can partially substitute for one another to reconstitute glutamate-gated currents from rat, Drosophila, and C. elegans. Furthermore, Ce STG-1 and Apis STG1 are primarily required for function independent of possible roles in promoting the surface expression of invertebrate AMPARs (Walker, 2006a).

The neurotransmitter glutamate mediates excitatory synptic transmission by activating ionotropic glutamate receptors (iGluRs). In C. elegans, the GLR-1 receptor subunit is required for glutamate-gated current in a subset of interneurons that control avoidance behaviors. Current mediated by GLR-1-containing iGluRs depends on SOL-1, a transmembrane CUB-domain protein that immunoprecipitates with GLR-1. Reconstitution of glutamate-gated current in heterologous cells depends on three proteins, STG-1 (a C. elegans stargazin-like protein), SOL-1, and GLR-1. This study used genetic and pharmacological perturbations along with rapid perfusion electrophysiological techniques to demonstrate that SOL-1 functions to slow the rate and limit the extent of receptor desensitization as well as to enhance the recovery from desensitization. A SOL-1 homologue from Drosophila has been identified (CG31218) and it is shown that Dro SOL1 has a conserved function in promoting C. elegans glutamate-gated currents. SOL-1 homologues may play critical roles in regulating glutamatergic neurotransmission in more complex nervous systems (Walker, 2006b).

In Drosophila, a number of iGluRs have been cloned, including Dro GluRIA, which has significant identity with C. elegans GLR-1 and is expressed in the nervous system. However, only small kainate-gated and virtually no glutamate-gated currents can be recorded from oocytes that express Dro GluRIA. No glutamate-gated current was detected when Dro GluRIA was expressed alone. Whether the lack of current was secondary to a requirement for a stargazin-like auxiliary protein was tested. Unlike GLR-1-mediated currents, coexpression of Dro GluRIA with vertebrate stargazin, Ce STG-1, or SOL-1 and Ce STG-1 did not significantly increase glutamate-gated current. However, small currents were recorded with coexpression of Dro GluRIA and Dro STG1, and significantly larger currents with coexpression of Dro GluRIA and Apis STG1. The onset of the current observed with coexpression of Dro STG1 was very slow, indicating that the majority of the current is likely a consequence of secondary activation of an endogenous Ca2+-activated chloride conductance. Consistent with this idea, substitution of barium for calcium in the extracellular solution dramatically reduced the amplitude of currents in response to glutamate application. These results indicate that there are important functional differences between vertebrate, C. elegans, and insect stargazin molecules in their ability to promote Dro GluRIA-mediated current. The large glutamate-gated current observed with coexpression of Apis STG1 could not be explained by surface delivery of receptors; in fact, it was found that the fractional surface expression of Dro GluRIA alone was 2- to 3-fold higher than when coexpressed with Apis STG1 (Walker, 2006a).

Most rapid excitatory synaptic signaling in the brain is mediated by postsynaptic ionotropic glutamate receptors (iGluRs) that are gated open by the neurotransmitter glutamate. In Caenorhabditis elegans, sol-1 encodes a CUB-domain transmembrane protein that is required for currents that are mediated by the GLR-1 iGluR. Mutations in sol-1 do not affect GLR-1 expression, localization, membrane insertion, or stabilization at synapses, suggesting that SOL-1 is required for iGluR function. This study provides evidence that SOL-1 is an auxiliary subunit that modulates the gating of GLR-1 receptors. Mutant variants of GLR-1 with altered gating partially restore glutamate-gated current and GLR-1-dependent behaviors in sol-1 mutants. Domain analysis of SOL-1 indicates that extracellular CUB domain 3 is required for function and that a secreted variant partially restores glutamate-gated currents and behavior. Also, it is shown that endogenous glutamatergic synaptic currents are absent in sol-1 mutants. These data suggest that GLR-1 iGluRs are not simply stand-alone molecules and require the SOL-1 auxiliary protein to promote the open state of the receptor. This analysis presents the possibility that glutamatergic signaling in other organisms may be similarly modified by SOL-1-like transmembrane proteins (Zheng, 2006).

Stargazin function in vertebrates

Stargazer, an ataxic and epileptic mutant mouse, lacks functional AMPA receptors on cerebellar granule cells. Stargazin, the mutated protein, interacts with both AMPA receptor subunits and synaptic PDZ proteins, such as PSD-95. The interaction of stargazin with AMPA receptor subunits is essential for delivering functional receptors to the surface membrane of granule cells, whereas its binding with PSD-95 and related PDZ proteins through a carboxy-terminal PDZ-binding domain is required for targeting the AMPA receptor to synapses. Expression of a mutant stargazin lacking the PDZ-binding domain in hippocampal pyramidal cells disrupts synaptic AMPA receptors, indicating that stargazin-like mechanisms for targeting AMPA receptors may be widespread in the central nervous system (Chen, 2000).

Excitatory synapses in the central nervous system release glutamate onto a number of receptor subtypes. The principal ionotropic glutamate receptors include AMPA receptors and NMDARs. AMPARs mediate moment-to-moment signalling, whereas NMDARs initiate synaptic plasticity. Recent studies emphasize remarkable differences in synaptic expression of these receptors. NMDARs are relatively fixed components of the postsynaptic density (PSD), whereas AMPARs are more loosely associated and their density at synapses is tightly controlled by neuronal activity1-8. That the number of AMPARs at the synapse is regulated independently of NMDARs raises intriguing questions concerning mechanisms involved in synaptic targeting of glutamate receptors and the role that this plasticity plays in learning and memory (Chen, 2000).

This independent regulation of synaptic AMPARs versus NMDARs is clearly manifested in the stargazer mutant mouse, which exhibits seizures and cerebellar ataxia. Among the defects identified in the cerebellum is the lack of functional AMPARs on granule cells. The defective protein stargazin was recently identified as a relative of the -1 subunit of the skeletal muscle calcium channel. More recently, a family of subunits has been identified, each with distinct expression patterns in the brain. Why stargazin is necessary for functional AMPARs in cerebellar granule cells has remained mysterious and is the focus of this study (Chen, 2000).

Whether the defect in AMPARs in the stargazer mouse results from abnormal cerebellar circuitry or alternatively is an autonomous granule cell defect was examined. AMPARs were evaluated in granule cell cultures, which lack the normal excitatory mossy fiber input, as well as Purkinje cells, the primary target for granule cells. In +/stg cultures, spontaneous rapid inward currents are routinely recorded from individual granule cells, presumably generated by synapses between granule cells. These events are largely mediated by the action potential dependent release of glutamate onto AMPARs, since their frequency is greatly reduced by tetrodotoxin (TTX) and they are abolished by the AMPAR antagonist CNQX. In striking contrast, stg/stg neurons exhibit essentially no spontaneous activity (Chen, 2000).

The stg/stg neurons clearly form excitatory synaptic connections since NMDAR currents can be recorded when Mg2+ is removed from, and CNQX and glycine are added to, the solution. As expected for NMDARs, these currents are considerably slower than those mediated by AMPAR currents and are blocked by the NMDAR antagonist AP5. There was no difference in the amplitude of NMDAR-mediated events in +/stg cells, or in the frequency of these events (Chen, 2000).

In cerebellar cultures from +/stg mouse, the AMPAR subunit GluR4 forms discrete synaptic puncta that colocalize with the presynaptic marker synaptophysin. By contrast, few synaptic GluR4 puncta are evident in stg/stg cells. Despite lacking synaptic GluR4 puncta, the cultured stg/stg granule cells are decorated with presynaptic synaptophysin puncta. To assess the synaptic localization of AMPAR subunits at mossy fibre synapses in the intact cerebellar glomeruli, post embedding immunogold electron-microscopic analysis was used. Granule cell synapses in stg/stg cerebellum are virtually devoid of GluR2/3 labelling, whereas these synapses in +/stg are labelled abundantly. Similar results were obtained with an antibody to the GluR4 subunit; most synapses in the +/stg mouse are GluR4-positive, whereas synapses in the stg/stg mouse are rarely labelled. A 'blind' quantitative analysis showed roughly a 10-fold decrease in the number of GluR2/3-reactive particles at stg/stg synapses (Chen, 2000).

In contrast to the defect in synaptic GluR 2/3 labelling, the NMDAR subunit NR1 labelling was increased in the stg/stg mutant. The cytoplasmic GluR2/3 labelling was low in both +/stg and stg/stg granule cells, making quantification difficult, but no obvious difference was noted. The presence of cytoplasmic staining indicates that AMPARs are present in the stg/stg granule cells, in agreement with Western blot analysis. Finally, there was no obvious difference in the pre- or post-synaptic morphology of the mossy fiber to granule cell synapse between the stg/stg mice and +/stg mice (Chen, 2000).

Since stargazer granule cells lack AMPAR currents, whether stargazin associates with GluR subunits was examined. When co-transfected into COS cells, stargazin and GluR4 co-immunoprecipitate. Notably, stargazin also interacts with co-transfected GluR1 or 2 subunits. These stargazin interactions with GluR subunits appear specific, since stargazin does not bind to NMDAR. In brain homogenates, stargazin is enriched in Triton X-100 insoluble postsynaptic density fractions together with GluR4, NR1 and postsynaptic density-95 (PSD-95). Although stargazin could not be co-immunoprecipitated with GluR from brain extracts, the harsh conditions required for solubilization of the PSD disrupt many protein complexes and often preclude detection of interactions (Chen, 2000).

Stargazin is predicted to contain four putative transmembrane domains with intracellular amino- and carboxy-terminal tails. Its C-terminal region contains a type I PDZ-binding site for association with certain synaptic PDZ proteins such as PSD-95. Stargazin was found to interact with PSD-95 and SAP-97, as well as with PSD-93 and SAP-102. Deleting the final four amino acids of stargazin (stargazinC) disrupts interaction with PSD-95, although stargazinC does retain binding to GluR4 (Chen, 2000).

Because PSD-95 can mediate clustering of ion channels, it was of interest to see if its interactions with stargazin might cluster GluRs in heterologous cells. Co-expression of GluR4 with PSD-95, or with stargazin results in diffuse, overlapping distributions for these proteins, which generally resemble the localization of GluR4 expressed alone. However, transfecting the three together causes a remarkable redistribution of the proteins to patch-like clusters. This clustering requires interaction of stargazin with the PDZ domains from PSD-95; clustering is not observed in co-transfections with stargazinC. The stargazin/PSD-95-induced clusters of GluR4 occur at the cell surface; they are labelled in non-permeabilized cells by an antibody to the extracellular green fluorescent protein (GFP) tag on GluR4 (Chen, 2000).

These findings suggest that stargazin has two distinct roles in controlling AMPAR function: (1) stargazin regulates delivery of AMPARs to the membrane surface and this function does not require the PDZ-binding domain; (2) stargazin mediates synaptic targeting of AMPARs and this function does require the PDZ-binding C terminus. Moreover, it is proposed that stargazin and the related proteins mediate AMPAR targeting throughout the brain. That stargazin clusters at hippocampal synapses and that stargazinC acts as a dominant-negative support a general role for stargazin-like proteins in synaptic targeting of AMPARs. A class of PDZ-domain-containing proteins, including GRIP (glutamate receptor interacting protein) 1 and 2, ABP (AMPAR-binding protein), PICK1 (protein interacting with C kinase 1) and SAP-97 (synapse-associated protein) have been implicated in synaptic targeting/clustering of AMPARs. A recent study suggests that the association of GluR2 with ABP/GRIP is not essential for synaptic targeting, but is required for maintaining the synaptic surface accumulation of AMPARs. Together, it seems that synaptic targeting/insertion and synaptic stabilization of AMPARs may be mediated by several mechanisms. Whereas the interaction between stargazin and PSD-95 or related PDZ proteins is crucial for the initial synaptic targeting of AMPARs, a different pathway involving the association of GluR2 with ABP/GRIP may be required for receptor stabilization (Chen, 2000).

Excitatory synapses in the brain exhibit a remarkable degree of functional plasticity, which largely reflects changes in the number of synaptic AMPA receptors. However, mechanisms involved in recruiting AMPA receptors to synapses are unknown. Hippocampal slice cultures and biolistic gene transfections have been used to study the targeting of AMPA receptors to synapses. AMPA receptors are localized to synapses through direct binding of the first two PDZ domains of synaptic PSD-95 to the AMPAR-associated protein, stargazin. Increasing the level of synaptic PSD-95 recruits new AMPA receptors to synapses without changing the number of surface AMPARs. At the same time, stargazin overexpression drastically increases the number of extra-synaptic AMPA receptors, but fails to alter synaptic currents if synaptic PSD-95 levels are kept constant. Finally, compensatory mutations were made to both PSD-95 and stargazin to demonstrate the central role of direct interactions between them in determining the number of synaptic AMPARs (Schnell, 2002).

Under standard conditions, cultured ventral spinal neurons cluster AMPA-type (but not NMDA-type) glutamate receptors at excitatory synapses on their dendritic shafts, in spite of abundant expression of the ubiquitous NMDA receptor subunit NR1. The NMDA receptor subunits NR2A and NR2B are not routinely expressed in cultured spinal neurons and transfection with NR2A or NR2B reconstitutes the synaptic targeting of NMDA receptors and confers on exogenous application of the immediate early gene product Narp the ability to cluster both AMPA and NMDA receptors. The use of dominant-negative mutants of GluR2 further shows that the synaptic targeting of NMDA receptors is dependent on the presence of synaptic AMPA receptors and that synaptic AMPA and NMDA receptors are linked by Stargazin and a MAGUK protein. This system of AMPA receptor-dependent synaptic NMDA receptor localization is preserved in hippocampal interneurons but reversed in hippocampal pyramidal neurons (Mi, 2004).

Accumulation of AMPA receptors at synapses is a fundamental feature of glutamatergic synaptic transmission. Stargazin, a member of the TARP family, is an AMPAR auxiliary subunit allowing interaction of the receptor with scaffold proteins of the postsynaptic density, such as PSD-95. How PSD-95 and Stargazin regulate AMPAR number in synaptic membranes remains elusive. Using single quantum dot and FRAP imaging in live hippocampal neurons, it has been shown that exchange of AMPAR by lateral diffusion between extrasynaptic and synaptic sites mostly depends on the interaction of Stargazin with PSD-95 and not upon the GluR2 AMPAR subunit C terminus. Disruption of interactions between Stargazin and PSD-95 strongly increases AMPAR surface diffusion, preventing AMPAR accumulation at postsynaptic sites. Furthermore, AMPARs and Stargazin diffuse as complexes in and out synapses. These results propose a model in which the Stargazin-PSD-95 interaction plays a key role to trap and transiently stabilize diffusing AMPARs in the postsynaptic density (Bats, 2007).

Endogenous polyamines profoundly affect the activity of various ion channels, including that of calcium-permeable AMPA-type glutamate receptors (CP-AMPARs). Stargazin, a transmembrane AMPAR regulatory protein (TARP) known to influence transport, gating and desensitization of AMPARs, greatly reduces block of CP-AMPARs by intracellular polyamines. By decreasing CP-AMPAR affinity for cytoplasmic polyamines, stargazin enhances the charge transfer following single glutamate applications and eliminates the frequency-dependent facilitation seen with repeated applications. In cerebellar stellate cells, which express both synaptic CP-AMPARs and stargazin, it was found that the rectification and unitary conductance of channels underlying excitatory postsynaptic currents were matched by those of recombinant AMPARs only when the latter were associated with stargazin. Taken together, these observations establish modulatory actions of stargazin that are specific to CP-AMPARs, and suggest that during synaptic transmission the activity of such receptors, and thus calcium influx, is fundamentally changed by TARPs (Soto, 2007).

Miscellaneous AMPA receptor protein interactions

Synaptic clustering of neurotransmitter receptors is crucial for efficient signal transduction and integration in neurons. PDZ domain-containing proteins such as PSD-95/SAP90 interact with the intracellular C termini of a variety of receptors and are thought to be important in the targeting and anchoring of receptors to specific synapses. PICK1 (protein interacting with C kinase), a PDZ domain-containing protein, interacts with the C termini of AMPA receptors in vitro and in vivo. In neurons, PICK1 specifically colocalizes with AMPA receptors at excitatory synapses. Furthermore, PICK1 induces clustering of AMPA receptors in heterologous expression systems. These results suggest that PICK1 may play an important role in the modulation of synaptic transmission by regulating the synaptic targeting of AMPA receptors (Xia, 1999).

AMPA receptor subunits interact with a PDZ domain-containing protein called PICK1, which is known to bind protein kinase C alpha (PKC alpha). PICK1 interacts with sequences within the last ten amino acid residues containing a novel PDZ binding motif (E S V/I K I) of the short C-terminal alternative splice variants of AMPA receptor subunits. No interaction occurs with the corresponding long splice variants which do not contain the E S V/I K I motif. The PDZ domain of PICK1 is required for the interaction; the mutation of a single amino acid in this region (Lys-27 to Glu) prevents interaction between PICK1 and GluR2 in the yeast two-hybrid assay. A similar mutation has been reported to prevent the binding of PICK1 to PKC alpha, indicating that the same domain of PICK1 binds both PKC alpha and GluRs. Flag-tagged PICK1 is retained by a glutathione S-transferase (GST) fusion of the C-terminal of GluR2 (GST-ct-GluR2; short splice variant) but not by GST-ct-GluR1 (long splice variant). Recombinant full length GluR2 is coimmunoprecipitated with flag-PICK1 using an anti-flag antibody and flag-PICK1 is coimmunoprecipitated with an N-terminal directed anti-GluR2 antibody. Transient expression of both proteins in COS cells reveals colocalization and an altered pattern of distribution for each protein, in comparison to the expression patterns when expressed individually. This novel interaction provides a possible regulatory mechanism to specifically modulate distinct splice variants and may be involved in targeting the phosphorylation of short form GluRs by PKC alpha (Dev, 1999).

AMPA receptor-binding protein (ABP) is a postsynaptic density (PSD) protein related to glutamate receptor-interacting protein (GRIP: see Drosophila Grip). ABP has two sets of three PDZ domains, which bind the GluR2/3 AMPA receptor subunits. ABP exhibits widespread CNS expression and is found at the postsynaptic membrane. The protein interactions of the ABP/GRIP family differ from the PSD-95 family, which bind N-methyl-D-aspartate (NMDA) receptors. ABP binds to the GluR2/3 C-terminal VKI-COOH motif via class II hydrophobic PDZ interactions, distinct from the class I PSD-95-NMDA receptor interaction. ABP and GRIP also form homo- and hetero-multimers through PDZ-PDZ interactions but do not bind PSD-95. It is suggested that the ABP/GRIP and PSD-95 families form distinct scaffolds that anchor, respectively, AMPA and NMDA receptors (Srivastava, 1998).

The molecular mechanisms underlying the targeting and localization of glutamate receptors at postsynaptic sites is poorly understood. A PDZ domain-containing protein, glutamate receptor-interacting protein 1 (GRIP1), has been identified that specifically binds to the C termini of AMPA receptor subunits and may be involved in the synaptic targeting of these receptors. The cloning of GRIP2, a homolog of GRIP1, is reported along with the characterization of the GRIP1 and GRIP2 proteins in the rat CNS. Recently, a GluR2/3 binding protein homologous to GRIP1, AMPA receptor-binding protein (ABP), has been described (Srivastava, 1998). ABP is apparently a short splice variant of GRIP2 that lacks the N terminus and the seventh PDZ domain of GRIP2. Similar to GRIP1, GRIP2 contains seven PDZ domains that are very homologous to GRIP1 within the PDZ domains (64%-93% identity) but has little sequence similarity in the linker regions between the PDZ domains. GRIP1 and GRIP2 are ~130 kDa proteins that are highly enriched in brain. GRIP1 and GRIP2 are widely expressed in brain, with the highest levels found in the cerebral cortex, hippocampus, and olfactory bulb. Biochemical studies show that GRIP1 and GRIP2 are enriched in synaptic plasma membrane and postsynaptic density fractions. GRIP1 is expressed early in embryonic development before the expression of AMPA receptors and peaks in expression at postnatal day 8-10. In contrast, GRIP2 is expressed relatively late in development and parallels the expression of AMPA receptors. Immunohistochemistry using the GRIP1 antibodies demonstrates that GRIP1 is expressed in neurons in a somatodendritic staining pattern. At the ultrastructural level GRIP1 is enriched in dendritic spines near the postsynaptic density and is expressed in dendritic shafts and in peri-Golgi regions in the neuronal soma. GRIP1 appears to be clustered at both glutamatergic and GABAergic synapses. These results suggest that GRIP1 and GRIP2 are AMPA receptor binding proteins potentially involved in the targeting of AMPA receptors to synapses. GRIP1 also may play functional roles at both excitatory and inhibitory synapses, as well as in early neuronal development (Dong, 1999).

The NMDA and AMPA classes of ionotropic glutamate receptors are concentrated at postsynaptic sites in excitatory synapses. NMDA receptors interact via their NR2 subunits with PSD-95/SAP90 family proteins, whereas AMPA receptors bind via their GluR2/3 subunits to glutamate receptor-interacting protein (GRIP), AMPA receptor-binding protein (ABP), and protein interacting with C kinase 1 (PICK1). A novel cDNA (termed ABP-L/GRIP2) is described that is virtually identical to ABP except for additional GRIP-like sequences at the N-terminal and C-terminal ends. Like GRIP (here termed GRIP1), ABP-L/GRIP2 contains a seventh PDZ domain at its C terminus. Using antibodies that recognize both these proteins, the subcellular localization of GRIP1 and ABP-L/GRIP2 (collectively termed GRIP) and their biochemical association with AMPA receptors are reported. GRIP is present at excitatory synapses and also at nonsynaptic membranes and within intracellular compartments. The association of native GRIP and AMPA receptors was confirmed biochemically by coimmunoprecipitation from rat brain extracts. A majority of detergent-extractable GluR2/3 is complexed with GRIP in the brain. However, only approximately half of GRIP is associated with AMPA receptors. Unexpectedly, immunocytochemistry of cultured hippocampal neurons and rat brain at the light microscopic level shows enrichment of GRIP in GABAergic neurons and in GABAergic nerve terminals. Thus GRIP is associated with inhibitory as well as excitatory synapses. Collectively, these findings support a role for GRIP in the synaptic anchoring of AMPA receptors but also suggest that GRIP has additional functions unrelated to the binding of AMPA receptors (Wyszynski, 1999).

The PDZ domain-containing proteins, such as PSD-95 and GRIP, have been suggested to be involved in the targeting of glutamate receptors, a process that plays a critical role in the efficiency of synaptic transmission and plasticity. To address the molecular mechanisms underlying AMPA receptor synaptic localization, several GRIP-associated proteins (GRASPs) have been identified that bind to distinct PDZ domains within GRIP. GRASP-1 is a neuronal rasGEF associated with GRIP and AMPA receptors in vivo. Overexpression of GRASP-1 in cultured neurons specifically reduces the synaptic targeting of AMPA receptors. In addition, the subcellular distribution of both AMPA receptors and GRASP-1 is rapidly regulated by the activation of NMDA receptors. These results suggest that GRASP-1 may regulate neuronal ras signaling and contribute to the regulation of AMPA receptor distribution by NMDA receptor activity (Ye, 2000).

LTP and LTD have been proposed to be mediated, in part, by changes in AMPA receptor function. Increases in AMPA receptor responses have been observed during the expression of LTP. Recently, it has been shown that a high proportion of synapses in hippocampal CA1 region contains only NMDA receptors and acquires AMPA receptors only after the induction of LTP. This emergence of AMPA receptor current seems due to the appearance of synaptic AMPA receptors. Moreover, NMDA receptor-dependent LTD in cultured neurons has recently been observed to correlate with a decrease in the levels of synaptic AMPA receptors. Previous studies have suggested that AMPA receptor-associated proteins, such as GRIP, are involved in the synaptic targeting of AMPA receptors. In this study, GRASP-1 has been added to this complex and evidence is provided that GRASP-1 may also be important in regulation of AMPA receptor function and may play a role in AMPA receptor synaptic targeting. Overexpression of GRASP-1 in neurons downregulates synaptic AMPA receptor clusters, while it has no effect on synaptic NMDA receptor synaptic targeting. Both the rasGEF catalytic domain and the C-terminal 'regulatory' domain were required for this activity. Activation of NMDA receptors dramatically induces the redistribution of both GRASP-1 and AMPA receptors from punctate membrane structures to a more diffuse pattern. Together with the GRASP-1 overexpression data, these results suggest that the overall spatial distribution of GRASP-1, as well as the absolute levels, may be important for AMPA receptor targeting. These results suggest that GRASP-1 and possibly ras signaling may play a role in the regulation of AMPA receptor synaptic targeting and its regulation by NMDA receptor activity (Ye, 2000).

Several proteins have been shown to interact specifically with the C termini of the GluR2 and GluR3 AMPA receptor subunits. These include three PDZ proteins, ABP (AMPA Receptor Binding protein), GRIP (Glutamate Receptor Interacting Protein), and PICK1 (Protein Interacting with C Kinase). The two splice forms of ABP that contain either 6 or 7 PDZ domains, and the 7 PDZ domain GRIP are members of a novel sequence-related protein family. GRIP PDZ4-5 domains and ABP PDZ5 domain show the highest affinity for the GluR2/3 C terminus. The remaining PDZ domains of GRIP and ABP are likely to mediate additional interactions, possibly anchoring the AMPA receptor to cytoskeletal proteins or coupling the receptor to intracellular enzymes. PICK1, which was cloned as a PKC-interacting protein, contains a single N-terminal PDZ domain. When coexpressed with GluR2 in heterologous cells, PICK1 induces GluR2 surface clustering and intracellular redistribution (Osten, 2000 and references therein).

The roles of GRIP, ABP, and PICK1 in GluR2 AMPA receptor trafficking have been studied. An epitope-tagged MycGluR2 subunit, when expressed in hippocampal cultured neurons, is specifically targeted to the synaptic surface. With the mutant MycGluR2delta1-10, which lacks the PDZ binding site, the overall dendritic intracellular transport and the synaptic surface targeting are not affected. However, over time, MycGluR2delta1-10 accumulates at synapses significantly less than MycGluR2. Notably, a single residue substitution, S880A, which blocks binding to ABP/GRIP but not to PICK1, reduces synaptic accumulation to the same extent as the PDZ site truncation. It is concluded that the association of GluR2 with ABP and/or GRIP but not PICK1 is essential for maintaining the synaptic surface accumulation of the receptor, possibly by limiting its endocytotic rate (Osten, 2000).

N-ethylmaleimide-sensitive fusion protein (NSF) interacts directly and selectively with the intracellular C-terminal domain of the GluR2 subunit of AMPA receptors. The interaction requires all three domains of NSF but occurs between residues Lys-844 and Gln-853 of rat GluR2, with Asn-851 playing a critical role. Loading of decapeptides corresponding to the NSF-binding domain of GluR2 into rat hippocampal CA1 pyramidal neurons results in a marked, progressive decrement of AMPA receptor-mediated synaptic transmission. This reduction in synaptic transmission is also observed when an anti-NSF monoclonal antibody (mAb) is loaded into CA1 neurons. These results demonstrate a previously unsuspected direct interaction in the postsynaptic neuron between two major proteins involved in synaptic transmission and suggest a rapid NSF-dependent modulation of AMPA receptor function (Nishimune, 1998).

Specific interaction is demonstrated between the GluR2 AMPA receptor subunit C-terminal peptide with an ATPase N-ethylmaleimide-sensitive fusion protein (NSF) and alpha- and beta-soluble NSF attachment proteins (SNAPs). These proteins are colocated in dendrites. The assembly of the GluR2-NSF-SNAP complex is ATP hydrolysis reversible and resembles the binding of NSF and SNAP with the SNAP receptor (SNARE) membrane fusion apparatus. Evidence that the molar ratio of NSF to SNAP in the GluR2-NSF-SNAP complex is similar to that of the t-SNARE syntaxin-NSF-SNAP complex. NSF is known to disassemble the SNARE protein complex in a chaperone-like interaction driven by ATP hydrolysis. A model is proposed in which NSF functions as a chaperone in the molecular processing of the AMPA receptor (Osten, 1998).

Glutamate receptors mediate the majority of rapid excitatory synaptic transmission in the central nervous system (CNS) and play important roles in synaptic plasticity and neuronal development. Recently, protein-protein interactions with the C-terminal domain of glutamate receptor subunits have been shown to be involved in the modulation of receptor function and clustering at excitatory synapses. The N-ethylmaleimide-sensitive factor (NSF), a protein involved in membrane fusion events, specifically interacts with the C terminus of the GluR2 and GluR4c subunits of AMPA receptors in vitro and in vivo. Moreover, intracellular perfusion of neurons with a synthetic peptide that competes with the interaction of NSF and AMPA receptor subunits rapidly decreases the amplitude of miniature excitatory postsynaptic currents (mEPSCs), suggesting that NSF regulates AMPA receptor function (Song, 1998).

Disruption of N-ethylmaleimide-sensitive fusion protein- (NSF-) GluR2 interaction by infusion into cultured hippocampal neurons of a blocking peptide (pep2m) causes a rapid decrease in the frequency but no change in the amplitude of AMPA receptor-mediated miniature excitatory postsynaptic currents (mEPSCs). NMDA receptor-mediated mEPSCs were not changed. Viral expression of pep2m reduces the surface expression of AMPA receptors, whereas NMDA receptor surface expression in the same living cells is unchanged. In permeabilized neurons, the total amount of GluR2 immunoreactivity is unchanged, and a punctate distribution of GluR2 is observed throughout the dendritic tree. These data suggest that the NSF-GluR2 interaction is required for the surface expression of GluR2-containing AMPA receptors and that disruption of the interaction leads to the functional elimination of AMPA receptors at synapses. Based on these findings and the known properties of NSF, a model is favored in which the interaction between NSF and GluR2 is involved in the part of the cycling process that is necessary for the insertion and/or stabilization of AMPA receptors at the postsynaptic membrane. By analogy with its known presynaptic functions, NSF could act at the AMPA receptor complex by stripping the receptors of associated proteins. Candidate proteins interacting with GluR2 include the PDZ-containing proteins GRIP, ABP, and PICK1. Removal of associated proteins could prime or "reset" the AMPA receptor complex to a naive state, thereby allowing insertion into the postsynaptic membrane. If the action of NSF is prevented, for example, by peptide block, the receptors cannot be appropriately processed, and insertion/reinsertion of the reconfigured receptors into the postsynaptic membrane cannot occur (Noel, 1999).

Narp (neuronal activity-regulated pentraxin) is a secreted immediate-early gene (IEG) regulated by synaptic activity in brain. Narp was originally identified by a novel subtractive cloning strategy from stimulated hippocampus and is a member of the newly recognized subfamily of 'long pentraxins' that includes neuronal pentraxin 1 and 2, which are found in the brain; TSG-14, a tumor necrosis factor-inducible acute phase reactant; and apexin, which is localized to the acrosome of mature sperm. These molecules are similar in structure in that they possess a C-terminal pentraxin domain and a 200 amino acid unique N terminus whose function up to this point is unknown. The pentraxin domain on Narp is similar to the mammalian protein C-reactive protein (CRP) and to mammalian serum amyloid protein (SAP), as well as highly conserved homologs from species as distant as Limulus. Pentraxins are secreted proteins that self-multimerize to form pentamers and may further dimerize to form decamers. A crystal structure of SAP shows that the pentraxin sugar-binding motif is remarkably homologous in secondary and tertiary structure to the plant lectin concanavalin A, a feature that is conserved in Narp. The physiological roles of pentraxins have remained obscure, although CRP has been postulated to play a role in nonantibody-mediated immune responses by binding and aggregating bacteria and other pathogens (O'Brien, 1999 and references).

Narp possesses several properties that make it likely to play a key role in excitatory synaptogenesis. Narp is shown to be selectively enriched at excitatory synapses on neurons from both the hippocampus and spinal cord. Overexpression of recombinant Narp increases the number of excitatory but not inhibitory synapses in cultured spinal neurons. In transfected HEK 293T cells, Narp interacts with itself, forming large surface clusters that coaggregate AMPA receptor subunits. Moreover, Narp-expressing HEK 293T cells can induce the aggregation of neuronal AMPA receptors. These studies support a model in which Narp functions as an extracellular aggregating factor for AMPA receptors (O'Brien, 1999).

The neuronal IEG Narp is selectively expressed at the majority of excitatory, axodendritic shaft synapses on aspiny spinal cord and hippocampal neurons in vitro. In addition, a small number of spine-bearing neurons express Narp at their excitatory synapses in culture. In vivo, Narp to be present at both pre- and post-synaptic sites of spiny and aspiny synapses. The prominent presynaptic localization of Narp in mossy fiber terminals is associated with synaptic vesicles. Because Narp is dramatically upregulated in neurons in response to patterned synaptic activity and is expressed at relatively high levels in developing and adult brain, these studies suggest that Narp may play a critical role in linking activity with the development and plasticity of excitatory synapses. The family of long pentraxins, of which Narp is a member, has several characteristics that might play a role in promoting excitatory synapse formation. Included among these are the ability to form side-to-side and head-to-head multimeric aggregates and the ability to bind other proteins via a lectin-like domain. The ability of Narp to cluster AMPA receptors would not have been predicted from a knowledge of the family of pentraxins, since the association of Narp with AMPA receptor subunits in the presence of tunicamycin suggests that it is not the lectin component of Narp that mediates this interaction. Indeed, the specificity of the interaction (GluR1-GluR3 but not GluR4, GluR6, or NR1) would also argue against a nonspecific interaction such as that mediated by a lectin (O'Brien, 1999).

Another notable functional property of Narp-expressing cells is their ability to cluster AMPA receptors on apposing cells, even when the contacted cell does not express Narp. In view of the documented physical interaction between Narp and AMPA receptors when these proteins are expressed in the same cells, it seems likely that the intercellular clustering activity also involves their physical interaction. The transcellular clustering activity of Narp is further enhanced when the apposing cell coexpresses Narp with AMPA receptors, suggesting that a Narp-Narp interaction may also contribute to transcellular clustering. In this regard, Narp may potentially display similarities with cadherins, which self associate and participate in synaptogenesis from both the pre- and post-synaptic surfaces. Unlike the family of cadherins, however, Narp appears to be completely extracellular, with no transmembrane domain (O'Brien, 1999).

A model is proposed in which Narp-Narp interactions between pre- and post-synaptic cells contribute to excitatory synapse formation by a secondary clustering of synaptic AMPA receptors due to Narp-AMPA receptor interactions. In support of a 'presynaptic' effect of Narp, it is noted that Narp expressed on heterologous cells induces AMPA receptor clusters on neurons. By contrast, a 'postsynaptic' effect of Narp is suggested by the observation that when Narp expression is upregulated by transfection, it results in a 2-fold increase in the number of excitatory synapses, with no change in the number of inhibitory synapses. Since only a small percentage of cells is transfected in this experiment, the increase in the number of excitatory synapses on transfected cells as compared with nontransfected cells in the same dish argues strongly for a postsynaptic action of Narp. The synaptogenic effect of upregulated Narp expression in neurons would be a manifestation of the potentiating effect of Narp on the transcellular cluster formation seen in heterologous cells when Narp is coexpressed with AMPA receptor subunits. It is also possible that excess Narp in the postsynaptic neuron promotes Narp-Narp intercellular interactions with presynaptic elements and makes it more favorable for the overexpressing cell to attract new synapses. This would favor excitatory synapses over inhibitory synapses, since inhibitory axons do not appear to express Narp (O'Brien, 1999).

Silent synapses form between some primary sensory afferents and dorsal horn neurons in the spinal cord. Molecular mechanisms for activation or conversion of silent synapses to conducting synapses are unknown. Serotonin can trigger activation of silent synapses in dorsal horn neurons by recruiting AMPA receptors. AMPA-receptor subunits GluR2 and GluR3 interact via their cytoplasmic C termini with PDZ-domain-containing proteins such as GRIP (glutamate receptor interacting protein), but the functional significance of these interactions is unclear. Protein interactions involving the GluR2/3 C terminus are important for serotonin-induced activation of silent synapses in the spinal cord. Furthermore, PKC is a necessary and sufficient trigger for this activation. These results implicate AMPA receptor-PDZ interactions in mechanisms underlying sensory synaptic potentiation and provide insights into the pathogenesis of chronic pain mediated by the sensitization of spinal dorsal horn neurons to input from primary afferent fibers (Li, 1999).

Compartmentalization of glutamate receptors with the signaling enzymes that regulate their activity supports synaptic transmission. Two classes of binding proteins organize these complexes: the MAGUK proteins that cluster glutamate receptors and AKAPs that anchor kinases and phosphatases. Glutamate receptors and PKA are recruited into a macromolecular signaling complex through direct interaction between the MAGUK proteins, PSD-95 and SAP97, and AKAP79/150. The SH3 and GK regions of the MAGUKs mediate binding to the AKAP. Cell-based studies indicate that phosphorylation of AMPA receptors is enhanced by a SAP97-AKAP79 complex that directs PKA to GluR1 via a PDZ domain interaction. Since AMPA receptor phosphorylation is implicated in regulating synaptic plasticity, these data suggest that a MAGUK-AKAP complex may be centrally involved (Colledge, 2000).

Phosphorylation of glutamate receptors is a critical regulatory event in the control of synaptic function and plasticity. Evidence is provided for the existence of a macromolecular transduction unit in which PKA is targeted to glutamate receptors through the direct interaction of two distinct sets of synaptic organizing molecules -- the MAGUK proteins and AKAP79/150. These interactions increase the complexity of signaling networks at excitatory synapses and may provide a structural framework that permits preferential targeting of kinases to glutamate receptors. Presumably, such a highly organized kinase-substrate complex ensures rapid and efficient phosphorylation of ion channels in response to local synaptic signals (Colledge, 2000).

The MAGUK proteins provide the central scaffold upon which the complex is assembled. The N-terminal PDZ domains of PSD-95 and SAP97, two members of the MAGUK protein family, bind to the tails of NMDA and AMPA receptor subunits, respectively. AKAP79/150 and its associated kinases can be recruited to these glutamate receptor complexes via interaction with the C-terminal SH3 and GK domains of the MAGUKs. The demonstration that two independent sites of contact mediate interaction between AKAP79/150 and MAGUK proteins is interesting in light of other mapping studies that have defined linear sequences of 4-6 amino acids as ligands for PDZ, SH2, and SH3 domains. Certainly, multiple sites of contact are not unprecedented and are likely to provide additional stability to a given protein complex. For example, the KA2 subunit of the kainate receptor, like AKAP79, binds to both the SH3 and GK domains of PSD-95. In addition, AKAP79/150 appears to bind to the beta2 adrenergic receptor through sites in both the third intracellular loop and the C-terminal tail. Interestingly, mutations in the SH3 and GK domains of the Drosophila MAGUK Discs large produce severe phenotypes, suggesting that these modules mediate interactions that are critical for regulating MAGUK function. Furthermore, deletion of these regions of PSD-95 in mice produces defects in synaptic plasticity that have been attributed to altered downstream signaling events. A potential explanation for these observations is that the AKAP79/150 signaling scaffold no longer can be recruited to glutamate receptors through interaction with MAGUK proteins (Colledge, 2000 and references therein).

AKAP79/150 previously has been shown to provide a scaffold for three signaling enzymes: PKA, PKC, and calcineurin. Interestingly, PSD-95 competes with calcineurin for binding to AKAP79/150 in vitro. Preliminary mapping experiments suggest that the two proteins do not share the same binding site on the AKAP, since a deletion mutant that does not bind to calcineurin still binds to PSD-95. Thus, a more likely explanation is that PSD-95 binding to AKAP79/150 sterically hinders interaction with calcineurin. These data support the notion that, when bound to MAGUKs, AKAP79/150 may preferentially target kinases but not phosphatases to certain glutamate receptors at the PSD. This could provide a mechanism to favor ion channel phosphorylation through preferential recruitment of regulatory kinases (Colledge, 2000).

While these results clearly argue for a role for anchored PKA in receptor phosphorylation, targeted phosphatases are also certain to participate in receptor dephosphorylation. However, interaction of AKAP79 with MAGUKs appears to exclude the phosphatase calcineurin from the complex. One possibility is that phosphatases may be recruited to AMPA receptors through anchoring proteins other than AKAP79/150. In fact, recent reports suggest that the phosphatase PP1 may be targeted to AMPA receptor complexes through its association with spinophilin (see Drosophila Spinophilin). Kinase-phosphatase targeting to some NMDA receptors may be more direct. Through interaction with the NR1-1A splice variant, the anchoring protein yotiao targets both PKA and active PP1 to NMDA receptor complexes, conferring bidirectional regulation of NMDA receptor activity. When considered in light of the present data, this raises the intriguing possibility that signaling enzymes may be recruited to certain NMDA receptors through simultaneous association with two anchoring proteins: yotiao and AKAP79 (Colledge, 2000 and references therein).

Phosphorylation of the cytoplasmic tail of GluR1 potentiates receptor function. CaMKII increases the unitary channel conductance via phosphorylation of Ser-831, while PKA phosphorylation of Ser-845 increases the peak open probability. Phosphorylation-dependent changes in AMPA receptor activity have been proposed to underlie some aspects of LTP and LTD. For example, CamKII phosphorylation appears to be essential for the induction of hippocampal LTP, while recent studies have implicated a role for PKA in LTD. The present results suggest that AKAP79/150 functions as an important player in the postsynaptic regulation of excitatory transmission by targeting PKA to AMPA receptors. Specifically, cAMP-dependent phosphorylation of Ser-845, a known PKA site in GluR1, is enhanced when the kinase is targeted to the channel via a SAP97-AKAP79 complex. This enhancement in phosphorylation is significantly reduced when a PKA anchoring-defective form of AKAP79 is substituted in the complex. Furthermore, a mutation in the PDZ binding site in the tail of GluR1, which uncouples the receptor from SAP97, reduces the basal level of phosphorylation of Ser-845 compared to wild-type GluR1. Together, these results suggest that phosphorylation of Ser-845 is mediated through a SAP97-AKAP79 complex that targets PKA to GluR1 via a PDZ domain interaction. This is particularly interesting in light of recent evidence implicating a GluR1-PDZ domain interaction in the delivery of AMPA receptors into synapses. These data suggest that recruitment of a SAP97-AKAP79-PKA complex may play a role in this process. Manipulation of these protein-protein interactions in animals should provide models to study the role of this synaptic signaling unit in regulating glutamate receptor function in vivo (Colledge, 2000 and references therein).

AMPA receptor (AMPAR) trafficking is crucial for synaptic plasticity, which may be important for learning and memory. NSF and PICK1 bind the AMPAR GluR2 subunit and are involved in trafficking of AMPARs. GluR2, PICK1, NSF, and alpha-/beta-SNAPs form a complex in the presence of ATPgammaS. Similar to SNARE complex disassembly, NSF ATPase activity disrupts PICK1-GluR2 interactions in this complex. Alpha- and beta-SNAP have differential effects on this reaction. SNAP overexpression in hippocampal neurons leads to corresponding changes in AMPAR trafficking by acting on GluR2-PICK1 complexes. This demonstrates that the previously reported synaptic stabilization of AMPARs by NSF involves disruption of GluR2-PICK1 interactions (Hanley, 2002).

AMPAR trafficking is thought to involve constitutive cycling of receptors by endocytosis/exocytosis, as well as regulated events as part of LTD (endocytosis) and LTP (exocytosis). AMPAR endocytosis during some forms of LTD is dependent upon GluR2 phosphorylation and regulation of accessory protein binding. The NSF-mediated disassembly of the GluR2-PICK1 complex described in this study is therefore likely to be crucial in limiting endocytosis of AMPARs to maintain constitutive cycling at a constant rate and hence maintain a constant level of receptors at the synaptic membrane. From this baseline, LTD could be induced (in conjunction with phosphorylation events) by reducing the activity of NSF, possibly by modulation of SNAP-PICK1 binding, to stabilize GluR2-PICK1 interactions, and consequently enhance receptor endocytosis. This study has identified the molecular mechanism for the activity of NSF in AMPA receptor trafficking, and has demonstrated that NSF can function as a disassembling molecular chaperone in a protein complex other than the 20S SNARE complex. As additional NSF binding partners are identified, it is possible that this ATPase, previously thought to be faithful to the SNARE complex, will show more promiscuous chaperone behavior (Hanley, 2002).

Four PDZ domain-containing proteins, syntenin, PICK1, GRIP, and PSD95, have been identified as interactors with the kainate receptor (KAR) subunits GluR52b, GluR52c, and GluR6. Of these, it is shown that both GRIP and PICK1 interactions are required to maintain KAR-mediated synaptic function at mossy fiber-CA3 synapses. In addition, PKCalpha can phosphorylate ct-GluR52b at residues S880 and S886, and PKC activity is required to maintain KAR-mediated synaptic responses. It is proposed that PICK1 targets PKCalpha to phosphorylate KARs, causing their stabilization at the synapse by an interaction with GRIP. Importantly, this mechanism is not involved in the constitutive recycling of AMPA receptors since blockade of PDZ interactions can simultaneously increase AMPAR- and decrease KAR-mediated synaptic transmission at the same population of synapses (Hirbec, 2003).

The finding that KARs and AMPARs can bind to a common pool of PDZ proteins suggests that these proteins may play important general roles in the regulation of glutamatergic synapses. Based on the present findings and previous work on AMPARs, it is possible to speculate on the molecular mechanisms that mediate the differential regulation of AMPARs and KARs by these PDZ proteins. In this scheme, AMPARs are secured in intracellular pools via association of the GluR2 subunit with GRIP and/or ABP. These 'gripped' receptors are immobile over the time course of the electrophysiology experiments. PICK1 exchanges for GRIP and targets PKCalpha, which then phosphorylates S880 of GluR2, thereby preventing the rebinding of GRIP. The S880-phosphorylated AMPARs are mobile and available for surface expression. It is proposed that KARs are also 'gripped' by GRIP, but in this case, PICK1-targetted, PKC-dependent phosphorylation stabilizes the GRIP interaction with GluR5/6 and anchors the receptors at the postsynaptic membrane. These data are entirely consistent with the observations that blockade of either GRIP or PICK1 binding, or inhibition of PKC, results in a rapid decrease in KAR-mediated synaptic currents. It is speculated that, whereas phosphorylation of S880 of GluR2 prevents GRIP binding, phosphorylation of S880 and/or S886 of GluR52b (and/or equivalent residues of GluR6) stabilizes GRIP binding and anchors the receptors at the synapse (Hirbec, 2003).

These differences in the molecular consequences of PKC-mediated phosphorylation of AMPARs and KARs can explain the differential regulation in opposite directions of the functional synaptic responses. The results showing that, at the same population of synapses, disruption of PDZ protein interactions results in an increase in EPSCA and a simultaneous decrease in EPSCK suggests that these proteins may act to regulate the relative proportions of AMPARs and KARs at synapses. Physiologically, given the distinct biophysical and functional profiles of AMPARs and KARs, the dynamic regulation of these interactions will play important roles in the modulation of basal glutamatergic synaptic transmission. Furthermore, it has been reported previously that some forms of developmental and activity-dependent synaptic plasticity involve a switch from functionally expressed KARs to AMPARs. The differential effects of PDZ-interacting proteins demonstrated here on these two receptor types provide an attractive molecular mechanism to account for these developmental and activity-dependent changes in the AMPAR and KAR complement at synapses (Hirbec, 2003).

A clone has been isolated from a rat brain cDNA library corresponding to a 2779-bp cDNA encoding a novel splice form of the glutamate receptor interacting protein-1 (GRIP1). This 696-amino acid has been termed splice form GRIP1c 4-7 to differentiate it from longer splice forms of GRIP1a/b containing seven PDZ domains. The four PDZ domains of GRIP1c 4-7 are identical to PDZ domains 4-7 of GRIP1a/b. GRIP1c 4-7 also contains 35 amino acids at the N terminus and 12 amino acids at the C terminus that are different from GRIP1a/b. In transfected HEK293 cells, a majority of GRIP1c 4-7 was associated with the plasma membrane. GRIP1c 4-7 interacts with GluR2/3 subunits of the AMPA receptor. In low density hippocampal cultures, GRIP1c 4-7 clusters colocalized with GABAergic (where GABA is gamma-aminobutyric acid) and glutamatergic synapses, although a higher percentage of GRIP1c 4-7 clusters colocalized with gamma-aminobutyric acid, type A, receptor [GABA(A)R] clusters than with AMPA receptor clusters. Transfection of hippocampal neurons with hemagglutinin-tagged GRIP1c 4-7 showed that it targets to the postsynaptic complex of GABAergic synapses colocalizing with GABA(A)R clusters. GRIP1c 4-7-specific antibodies, which did not recognize previously described splice forms of GRIP1, recognized a 75-kDa protein that is enriched in a postsynaptic density fraction isolated from rat brain. EM immunocytochemistry experiments show that in intact brain GRIP1c 4-7 concentrates at postsynaptic complexes of both type I glutamatergic and type II GABAergic synapses although it is also presynaptically localized. These results indicate that GRIP1c 4-7 plays a role not only in glutamatergic synapses but also in GABAergic synapses (Charych, 2004).

Single-particle electron microscopy (EM) combined with biochemical measurements revealed the molecular shape of SAP97 (also known as hDlg) and a monomer-dimer transition that depends on the N-terminal L27 domain. Overexpression of SAP97 drives GluR1 to synapses, potentiates AMPA receptor (AMPAR) excitatory postsynaptic currents (EPSCs), and occludes LTP. Synaptic potentiation and GluR1 delivery are dissociable by L27 domain mutants that inhibit multimerization of SAP97. Loss of potentiation is correlated with faster turnover of monomeric SAP97 mutants in dendritic spines. It is proposed that L27-mediated interactions of SAP97 with itself or other proteins regulate the synaptic delivery of AMPARs. RNAi knockdown of endogenous PSD-95 depletes surface GluR1 and impaires AMPA EPSCs. In contrast, RNAi knockdown of endogenous SAP97 reduces surface expression of both GluR1 and GluR2 and inhibits both AMPA and NMDA EPSCs. Thus SAP97 has a broader role than its close relative, PSD-95, in the maintenance of synaptic function (Nakagawa, 2004).

A novel rat gene, tanc (GenBank Accession No. AB098072), has been cloned that encodes a protein containing three tetratricopeptide repeats (TPRs), ten ankyrin repeats and a coiled-coil region, and is possibly a rat homolog of Drosophila rolling pebbles. The tanc gene is expressed widely in the adult rat brain. Subcellular distribution, immunohistochemical study of the brain and immunocytochemical studies of cultured neuronal cells indicate the postsynaptic localization of TANC protein of 200 kDa. Pull-down experiments have shown that TANC protein binds PSD-95, SAP97, and Homer via its C-terminal PDZ-binding motif, -ESNV, and fodrin via both its ankyrin repeats and the TPRs together with the coiled-coil domain. TANC also binds the alpha subunit of Ca2+/calmodulin-dependent protein kinase II. An immunoprecipitation study shows TANC association with various postsynaptic proteins, including guanylate kinase-associated protein (GKAP), alpha-internexin, and N-methyl-D-aspartate (NMDA)-type glutamate receptor 2B and AMPA-type glutamate receptor (GluR1) subunits. These results suggest that TANC protein may work as a postsynaptic scaffold component by forming a multiprotein complex with various postsynaptic density proteins (Suzuki, 2004).

At many excitatory central synapses, activity produces a lasting change in the synaptic response by modifying postsynaptic AMPA receptors (AMPARs). Although much is known about proteins involved in the trafficking of Ca2+-impermeable (GluR2-containing) AMPARs, little is known about protein partners that regulate subunit trafficking and plasticity of Ca2+-permeable (GluR2-lacking) AMPARs. At cerebellar parallel fiber-stellate cell synapses, activity triggers a novel type of plasticity: Ca2+ influx through GluR2-lacking synaptic AMPARs drives incorporation of GluR2-containing AMPARs, generating rapid, lasting changes in excitatory postsynaptic current properties. This study examined how GRIP and protein interacting with C-kinase-1 (PICK) regulate subunit trafficking and plasticity. It was found that repetitive synaptic activity triggers loss of synaptic GluR2-lacking AMPARs by selectively disrupting their interaction with GRIP and that PICK drives activity-dependent delivery of GluR2-containing receptors. This dynamic regulation of AMPARs provides a feedback mechanism for controlling Ca2+ permeability of synaptic receptors (Liu, 2005).

The targeting and surface expression of membrane proteins are critical to their functions. In neurons, synaptic targeting and surface expression of AMPA-type glutamate receptors were found to be critical for synaptic plasticity such as long-term potentiation and long-term depression (LTD). PICK1 (protein interacting with C kinase 1) is a cytosolic protein that interacts with many membrane proteins, including AMPA receptors via its PDZ (postsynaptic density-95/Discs large/zona occludens-1) domain. Its interactions with membrane proteins regulate their subcellular targeting and surface expression. However, the mechanism by which PICK1 regulates protein trafficking has not been fully elucidated. This study shows that PICK1 directly binds to lipids, mainly phosphoinositides, via its BAR (Bin/amphiphysin/Rvs) domain. Lipid binding of the PICK1 BAR domain is positively regulated by its PDZ domain and negatively regulated by its C-terminal acidic domain. Mutation of critical residues of the PICK1 BAR domain eliminates its lipid-binding capability. Lipid binding of PICK1 controls the subcellular localization of the protein, because BAR domain mutant of PICK1 has diminished synaptic targeting compared with wild-type PICK1. In addition, the BAR domain mutant of PICK1 does not cluster AMPA receptors. Moreover, wild-type PICK1 enhances synaptic targeting of AMPA receptors, whereas the BAR domain mutant of PICK1 fails to do so. The BAR domain mutant of PICK1 loses its ability to regulate surface expression of the AMPA receptors and impairs expression of LTD in hippocampal neurons. Together, these findings indicate that the lipid binding of the PICK1 BAR domain is important for its synaptic targeting, AMPA receptor trafficking, and synaptic plasticity (Jin, 2006).

Via its extracellular N-terminal domain (NTD), the AMPA receptor subunit GluR2 promotes the formation and growth of dendritic spines in cultured hippocampal neurons. The first N-terminal 92 amino acids of the extracellular domain are necessary and sufficient for GluR2's spine-promoting activity. Moreover, overexpression of this extracellular domain increases the frequency of miniature excitatory postsynaptic currents (mEPSCs). Biochemically, the NTD of GluR2 can interact directly with the cell adhesion molecule N-cadherin, in cis or in trans. N-cadherin-coated beads recruit GluR2 on the surface of hippocampal neurons, and N-cadherin immobilization decreases GluR2 lateral diffusion on the neuronal surface. RNAi knockdown of N-cadherin prevents the enhancing effect of GluR2 on spine morphogenesis and mEPSC frequency. These data indicate that in hippocampal neurons N-cadherin and GluR2 form a synaptic complex that stimulates presynaptic development and function as well as promoting dendritic spine formation (Saglietti, 2007).

AMPA-type glutamate receptors (GluRs) play major roles in excitatory synaptic transmission. Neuronal AMPA receptors comprise GluR subunits and transmembrane AMPA receptor regulatory proteins (TARPs). Previous studies have identified five mammalian TARPs, γ-2 (or stargazin), γ-3, γ-4, γ-7, and γ-8, that enhance AMPA receptor function. This study classifies γ-5 as a distinct class of TARP that modulates specific GluR2-containing AMPA receptors and displays properties entirely dissimilar from canonical TARPs. γ-5 increases peak currents and decreases the steady-state currents selectively from GluR2-containing AMPA receptors. Furthermore, γ-5 increases rates of GluR2 deactivation and desensitization and decreases glutamate potency. Remarkably, all effects of γ-5 require editing of GluR2 mRNA. Unlike other TARPs, γ-5 modulates GluR2 without promoting receptor trafficking. γ-7 regulation of GluR2 is dictated by mRNA editing. These data establish γ-5 and γ-7 as a separate family of 'type II TARPs' that impart distinct physiological features to specific AMPA receptors (Kato, 2008).

At synapses, cell adhesion molecules (CAMs) provide the molecular framework for coordinating signaling events across the synaptic cleft. Among synaptic CAMs, the integrins, receptors for extracellular matrix proteins and counterreceptors on adjacent cells, are implicated in synapse maturation and plasticity and memory formation. However, little is known about the molecular mechanisms of integrin action at central synapses. This study reports that postsynaptic β3 integrins control synaptic strength by regulating AMPA receptors (AMPARs) in a subunit-specific manner. Pharmacological perturbation targeting β3 integrins promotes endocytosis of GluR2-containing AMPARs via Rap1 signaling, and expression of β3 integrins produces robust changes in the abundance and composition of synaptic AMPARs without affecting dendritic spine structure. Importantly, homeostatic synaptic scaling induced by activity deprivation elevates surface expression of β3 integrins, and in turn, β3 integrins are required for synaptic scaling. These findings demonstrate a key role for integrins in the feedback regulation of excitatory synaptic strength (Cingolani, 2008).

Oligophrenin-1 (OPHN1) encodes a Rho-GTPase-activating protein (Rho-GAP) whose loss of function has been associated with X-linked mental retardation (MR). The pathophysiological role of OPHN1, however, remains poorly understood. This study shows that OPHN1 through its Rho-GAP activity plays a critical role in the activity-dependent maturation and plasticity of excitatory synapses by controlling their structural and functional stability. Synaptic activity through NMDA receptor activation drives OPHN1 into dendritic spines, where it forms a complex with AMPA receptors, and selectively enhances AMPA-receptor-mediated synaptic transmission and spine size by stabilizing synaptic AMPA receptors. Consequently, decreased or defective OPHN1 signaling prevents glutamatergic synapse maturation and causes loss of synaptic structure, function, and plasticity. These results imply that normal activity-driven glutamatergic synapse development is impaired by perturbation of OPHN1 function. Thus, these findings link genetic deficits in OPHN1 to glutamatergic dysfunction and suggest that defects in early circuitry development are an important contributory factor to this form of MR (Nadif Kasri, 2009).

AMPA receptors: miscellaneous upstream modification pathways

Src-family protein tyrosine kinases (PTKs) transduce signals to regulate neuronal development and synaptic plasticity. However, the nature of their activators and the molecular mechanisms underlying these neural processes are unknown. Brain-derived neurotrophic factor (BDNF) and platelet-derived growth factor enhance expression of AMPA-type glutamate receptor 1 and 2/3 proteins in rodent neocortical neurons via the Src-family PTK(s). The increase in AMPA receptor levels is blocked in cultured neocortical neurons by the addition of a Src-family-selective PTK inhibitor. Accordingly, neocortical cultures from Fyn-knockout mice fail to respond to BDNF, whereas those from wild-type mice do respond. Moreover, the neocortex of young Fyn mutants exhibits a significant in vivo reduction in these AMPA receptor proteins but not in their mRNA levels. In vitro kinase assay reveals that BDNF can indeed activate the Fyn kinase: it enhances tyrosine phosphorylation of Fyn as well as that of exogenously supplemented enolase. All of these results suggest that the Src-family kinase Fyn, activated by the growth factors, plays a crucial role in modulating AMPA receptor expression during brain development (Narisawa-Saito, 1999).

The regulation of AMPA receptor channels by serotonin signaling in pyramidal neurons of prefrontal cortex (PFC) was studied. Application of serotonin reduced the amplitude of AMPA-evoked currents, an effect mimicked by 5-HT1A receptor agonists (see Drosophila Serotonin receptor 1A) and blocked by 5-HT1A antagonists, indicating the mediation by 5-HT1A receptors. The serotonergic modulation of AMPA receptor currents was blocked by protein kinase A (PKA) activators and occluded by PKA inhibitors. Inhibiting the catalytic activity of protein phosphatase 1 (PP1) also eliminated the effect of serotonin on AMPA currents. Furthermore, the serotonergic modulation of AMPA currents was occluded by application of the Ca(2+)/calmodulin-dependent kinase II (CaMKII) inhibitors and blocked by intracellular injection of calmodulin or recombinant CaMKII. Application of serotonin or 5-HT1A agonists to PFC slices reduced CaMKII activity and the phosphorylation of AMPA receptor subunit GluR1 at the CaMKII site in a PP1-dependent manner. It is concluded that serotonin, by activating 5-HT1A receptors, suppress glutamatergic signaling through the inhibition of CaMKII, which is achieved by the inhibition of PKA and ensuing activation of PP1. This modulation demonstrates the critical role of CaMKII in serotonergic regulation of PFC neuronal activity, which may explain the neuropsychiatric behavioral phenotypes seen in CaMKII knockout mice (Cai, 2002).

Extracellular signal-regulated kinase (ERK) signaling is important for neuronal synaptic plasticity. The protein kinase ribosomal S6 kinase (RSK2; see Drosophila RSK), a downstream target of ERK, uses a C-terminal motif to bind several PDZ domain proteins in heterologous systems and in vivo. Different RSK isoforms display distinct specificities in their interactions with PDZ domain proteins. Mutation of the RSK2 PDZ ligand does not inhibit RSK2 activation in intact cells or phosphorylation of peptide substrates by RSK2 in vitro but greatly reduces RSK2 phosphorylation of PDZ domain proteins of the Shank family in heterologous cells. In primary neurons, NMDA receptor (NMDA-R) activation leads to ERK and RSK2 activation and RSK-dependent phosphorylation of transfected Shank3. RSK2-PDZ domain interactions are functionally important for synaptic transmission because neurons expressing kinase-dead RSK2 display a dramatic reduction in frequency of AMPA-type glutamate receptor-mediated miniature excitatory postsynaptic currents, an effect dependent on the PDZ ligand. These results suggest that binding of RSK2 to PDZ domain proteins and phosphorylation of these proteins or their binding partners regulates excitatory synaptic transmission (Thomas, 2005).

Neuregulin-1 (NRG1) signaling participates in numerous neurodevelopmental processes. Through linkage analysis, nrg1 has been associated with schizophrenia, although its pathophysiological role is not understood. The prevailing models of schizophrenia invoke hypofunction of the glutamatergic synapse and defects in early development of hippocampal-cortical circuitry. This study shows that the erbB4 receptor, as a postsynaptic target of NRG1, plays a key role in activity-dependent maturation and plasticity of excitatory synaptic structure and function. Synaptic activity leads to the activation and recruitment of erbB4 into the synapse. Overexpressed erbB4 selectively enhances AMPA synaptic currents and increases dendritic spine size. Preventing NRG1/erbB4 signaling destabilizes synaptic AMPA receptors and leads to loss of synaptic NMDA currents and spines. These results indicate that normal activity-driven glutamatergic synapse development is impaired by genetic deficits in NRG1/erbB4 signaling leading to glutamatergic hypofunction. These findings link proposed effectors in schizophrenia: NRG1/erbB4 signaling perturbation, neurodevelopmental deficit, and glutamatergic hypofunction (Li, 2007).

AMPA receptors: Effects of mutation

Mouse mutants were generated with targeted AMPA receptor (AMPAR) GluR-B subunit alleles, functionally expressed at different levels and deficient in Q/R-site editing. All mutant lines have increased AMPAR calcium permeabilities in pyramidal neurons, and one shows elevated macroscopic conductances of these channels. The AMPAR-mediated calcium influx induces NMDA-receptor-independent long-term potentiation (LTP) in hippocampal pyramidal cell connections. Calcium-triggered neuronal death is not observed, but mutants have mild to severe neurological dysfunctions, including epilepsy and deficits in dendritic architecture. The seizure-prone phenotype correlate with an increase in the macroscopic conductance, as independently revealed by the effect of a transgene for a Q/R-site-altered GluR-B subunit. Thus, changes in GluR-B gene expression and Q/R site editing can affect critical architectural and functional aspects of excitatory principal neurons (Feldmeyer, 1999).

Desensitization of AMPA receptors is thought to shape the synaptic response and act as a neuroprotective mechanism at central synapses, but the molecular mechanism underlying desensitization is poorly understood. Replacing the glutamate binding domain S1 of GluR3 (an AMPA receptor) with S1 of GluR6 (a kainate receptor) results in a fully active but completely nondesensitizing receptor. Smaller substitutions within S1 identify (in addition to two additional modulatory regions) a single exchange, L507Y, that is required and sufficient for the block of desensitization. This phenotype is specific for AMPA receptors and requires an aromatic residue at this position. L507 lies between two residues (T504 and R509) that form part of the glutamate binding site. The physical proximity of these residues, which are involved in binding and gating, suggests they may form part of the link between these two events (Stern-Bach, 1998).

Although GluR1o and GluR3o are homologous at the amino acid level, GluR3o desensitizes approximately threefold faster than GluR1o. By creating chimeras of GluR1o and GluR3o and point amino acid exchanges in their S2 regions, two residues have been identified as critical for GluR1o desensitization: Y716 and the R/G RNA-edited site, R757. Intronic elements determine a codon switch in the primary transcripts at the R/G site that immediately precedes the flip/flop region. With creation of the double-point mutant (Y716F, R757G)GluR1o, complete exchange of the desensitization rate of GluR1o to that of GluR3o is obtained. In addition, both the potency and affinity of the subtype-selective agonist bromohomoibotenic acid are exchanged by the Y716F mutation. A model is proposed of the AMPA receptor binding site whereby a hydrogen-bonding matrix of water molecules plays an important role in determining both ligand affinity and receptor desensitization properties. Residues Y716 in GluR1 and F728 in GluR3 differentially interact with this matrix to affect the binding affinity of some ligands, providing the possibility of developing subtype-selective compounds (Banke, 2001).

AMPA receptor (AMPAR)-mediated ionic currents that govern gene expression, synaptic strength, and plasticity also can trigger excitotoxicity. However, native AMPARs exhibit heterogeneous pharmacological, biochemical, and ionic permeability characteristics, which are governed partly by receptor subunit composition. AMPARs exhibit heterogeneous macroscopic ionic current properties and ionic permeability characteristics. Their biophysical and pharmacological properties are governed by four genes (GluR1 to GluR4 or GluR-A to GluR-D) that encode heteromeric receptors with high AMPA affinities that are permeable to Na+ and K+ ions. However, the relative expression of these genes, as well as the splicing and editing of their mRNAs, imparts a diversity of pharmacological properties, gating characteristics, and Ca2+ permeability between cells. Specifically, impermeability to Ca2+ is determined by the presence of the GluR2 subunit, which has a positively charged arginine at position 586 of transmembrane segment 2 (Q/R site) instead of a neutral glutamine. Thus permeability to Ca2+ ions is highest in AMPARs that lack GluR2. However, the GluR2 subunit governs more than just Ca2+ permeability. GluR subunits display sequence divergence within the C-terminal (CT) cytoplasmic tail, and this region has been shown to mediate subunit-specific interactions with various cytoplasmic proteins. These AMPAR CT-protein interactions may govern the pharmacological properties of the receptor, receptor turnover at synapses, clustering, synaptic transmission, efficacy, and plasticity. Thus the influence of GluR2 subunits on neuronal function and vulnerability to excitotoxicity may occur by mechanisms other than solely those attributable to the effects of GluR2 on ionic permeability profiles (Iihara, 2001 and references therein).

GluR2 is expressed widely in mammalian neurons. For example, in cultured dissociated cortical neurons, a preparation that commonly is used to study excitotoxicity, only 8%-15% of neurons express AMPA channels lacking GluR2. In vivo, GluR2 is expressed widely in hippocampal pyramidal and granule neurons and in cortical neurons that frequently are damaged by ischemia. Thus the relative abundance and, yet, heterogeneity of GluR2 expression have made it more difficult to define its role in AMPAR-mediated excitotoxicity. In this paper mutant mice lacking the AMPAR subunit GluR2 were used to study AMPAR-mediated excitotoxicity in cultured cortical neurons and in hippocampal neurons in vivo. It was hypothesized that in these mice the level of GluR2 expression governs the vulnerability of neurons to excitotoxicity; also examined were the ionic mechanisms involved. In cortical neuronal cultures AMPAR-mediated neurotoxicity parallels the magnitude of kainate-evoked AMPAR-mediated currents, which are increased in neurons lacking GluR2. Ca2+ permeability, although elevated in GluR2-deficient neurons, do not correlate with excitotoxicity. However, toxicity is reduced by removal of extracellular Na+, the main charge carrier of AMPAR-mediated currents. In vivo, the vulnerability of CA1 hippocampal neurons to stereotactic kainate injections and of CA3 neurons to intraperitoneal kainate administration is independent of GluR2 level. Neurons lacking the GluR2 subunit do not demonstrate compensatory changes in the distribution, expression, or function of AMPARs or of Ca2+-buffering proteins. Thus GluR2 level may influence excitotoxicity by effects additional to those on Ca2+ permeability, such as effects on agonist potency, ionic currents, and synaptic reorganization (Iihara, 2001).

Ionotropic glutamate receptors are tetramers, the isolated ligand binding cores that assemble as dimers. Previous work on nondesensitizing AMPA receptor mutants, which combined crystallography, ultracentrifugation, and patch-clamp recording, show that dimer formation by the ligand binding cores is required for activation of ion channel gating by agonists. To define the mechanisms responsible for stabilization of dimer assembly in native AMPA receptors, contacts between the adjacent ligand binding cores were individually targeted by amino acid substitutions, using the GluR2 crystal structure as a guide to design mutants. Disruption of a salt bridge, hydrogen bond network, and intermolecular van der Waals contacts between helices D and J in adjacent ligand binding cores greatly accelerates desensitization. Conservation of these contacts in AMPA and kainate receptors indicates that they are important determinants of dimer stability and that the dimer interface is a key structural element in the gating mechanism of these glutamate receptor families (Horning, 2004).

AMPA receptors and neuron cell death

CA1 pyramidal neurons degenerate after transient global ischemia, whereas neurons in other regions of the hippocampus remain intact. A step in this selective injury is Ca2+ and/or Zn2+ entry through Ca2+-permeable AMPA receptor channels; reducing Ca2+ permeability of AMPA receptors via expression of Ca2+-impermeable GluR2(R) channels or activation of CRE transcription in the hippocampus of adult rats in vivo using shutoff-deficient pSFV-based vectors rescues vulnerable CA1 pyramidal neurons from forebrain ischemic injury. Conversely, the induction of Ca2+ and/or Zn2+ influx through AMPA receptors by expressing functional Ca2+-permeable GluR2(Q) channels causes the postischemic degeneration of hippocampal granule neurons that otherwise are insensitive to ischemic insult. Thus, the AMPA receptor subunit GluR2 gates entry of Ca2+ and/or Zn2+ that leads to cell death following transient forebrain ischemia (Liu, 2004).

AMPA receptors: Transcriptional regulation

Genes that specify cell fate can influence multiple aspects of neuronal differentiation, including axon guidance, target selection and synapse formation. Mutations in the unc-42 gene disrupt axon guidance along the C. elegans ventral nerve cord and cause distinct functional defects in sensory-locomotory neural circuits. unc-42 encodes a novel homeodomain protein that specifies the fate of three classes of neurons in the Caenorhabditis elegans nervous system: the ASH polymodal sensory neurons; the AVA, AVD and AVE interneurons, which mediate repulsive sensory stimuli to the nematode head and anterior body, and a subset of motor neurons that innervate head and body-wall muscles. The UNC-42 sequence contains a paired type homeodomain, most closely related to C. elegans CEH-10 and UNC-4. The homeodomain of UNC-42 is 68% identical to CEH-10 and 65% identical to UNC-4. UNC-42 exhibits the same degree of similarity to a number of homeoproteins of the paired-like classes from other species, including the vertebrate proteins, Cart-1, Phox2a and Arx -- all Aristalless type homeodomain proteins. unc-42 is required for the expression of cell-surface receptors that are essential for the mature function of these neurons. In mutant animals, the ASH sensory neurons fail to express SRA-6 and SRB-6, putative chemosensory receptors. The AVA, AVD and AVE interneurons and RME and RMD motor neurons of unc-42 mutants similarly fail to express the GLR-1 glutamate receptor (The predicted GLR-1 protein is roughly 40% identical to mammalian AMPA-class glutamate receptor (GluR) subunits). These results show that unc-42 performs an essential role in defining neuron identity and contributes to the establishment of neural circuits in C. elegans by regulating the transcription of glutamate and chemosensory receptor genes (Baran, 1999).

The first three unc-42 alleles were identified based on the uncoordinated (Unc) phenotype conferred by the mutations. Although mutant worms can move forward or backward spontaneously, this movement is slow and irregular. Mutant worms tend to kink ventrally and their head movement is restricted. These defects are likely to be the result of neural defects because muscle development and morphology are normal in unc-42 mutants. In additional assays for behavioral defects, it was found that unc-42 mutants also exhibit severe defects in response to mechanical stimuli to the nematode head and anterior body. unc-42 mutants exhibit a distinctive mechanosensory (Mec) response to light touch along the body: mutant animals fail to back up when stroked along anterior body regions, but moved forward normally when touched along the posterior body. The response to touch at the tip of the nose (Not) was also severely reduced in the mutants. Only 8% of unc-42 animals responded to nose touch, compared to 90% for wild-type worms. Nose touch and body touch are mediated through separate neural circuits defined by anatomy, cell killing experiments and genetic analysis. Wild-type animals respond to light strokes to the anterior part of the body by backing up; they respond to touch along the posterior body by moving forward. The ALMR, ALML and AVM neurons sense light touch along the anterior body, and the PLMR and PLML neurons sense touch along the posterior body. Response to anterior body touch requires the function of the AVA and AVD interneurons, which innervate type-A motor neurons and drive backward movement, while response to posterior body touch requires the AVB and PVC interneurons, which innervate type-B motor neurons and drive forward movement. The response to anterior body touch is dependent primarily on the AVD interneurons, probably via gap junctions with the ALM and AVM sensory neurons. The sensory neurons ASH and FLP at the tip of the nose are the primary detectors of touch and the avoidance responses touch generates. The ASH neurons also sense high osmolarity and volatile repellents, and may be analogous to vertebrate nociceptors that detect several pain modalities. Although the ASH and FLP sensory neurons are able to synapse with many of the same interneurons as the ALM and AVM mechanosensory neurons, the response to nose touch employs a separate mode of synaptic transmission. ASH-mediated response to nose touch requires GLR-1, an AMPA-type glutamate receptor expressed by many postsynaptic targets of ASH and FLP, including the AVA and AVD interneurons. Mutations in glr-1 block the response to nose touch, but do not alter avoidance to body touch or high osmolarity, even though these responses are also mediated by the AVA and AVD interneurons. Mutations in unc-42 alter all of these behaviors: in such cases, response to nose touch and anterior body touch are severely diminished, and osmotic avoidance is also reduced (Baran, 1999).

To determine if developmental defects in the mechanosensory neurons could account for these behavioral defects, the morphology and differentiation of mutant ALM, AVM and ASH sensory neurons were examined. UNC-86, a POU homeodomain protein required to specify ALM, AVM and FLP cell fate, is expressed normally in the mutants. The axon trajectories of the unc-42 and wild-type ALM and AVM neurons are also indistinguishable from one another. The ASH neurons and several other sensory neurons with ciliated endings take up lipophilic fluorescent dyes such as DiI. When unc-42 mutants are incubated in media containing DiI, the ASH neurons take up the dye normally, indicating that their ciliated endings are exposed. Although ASH axonal and dendritic morphology appear normal in these animals, other markers for ASH differentiation are not expressed. The ASH neurons of unc-42 mutants fail to express an srb-6-gfp transgene, while the ADL and ADF sensory neurons express this transgene normally. ASH neurons of unc-42 mutants also fail to express an sra-6-gfp transgene. sra-6 and srb-6 encode seven-transmembrane receptors that may function as chemoreceptors for the detection of volatile repellents (Baran, 1999).

Failures in ASH function could account for behavioral defects mediated solely by ASH, but cannot account for the defects in response to body touch or locomotion, or the severity of the nose-touch defects of unc-42 mutants. If the ASH neurons of wild-type animals are killed by laser microsurgery, worms can still respond to nose touch 37% of the time. By contrast, unc-42 mutants respond to nose-touch stimuli only 10% of the time. These results predict that cells in addition to ASH contribute to the nose-touch response defects of the mutants. To determine if unc-42 mutations also disrupts interneuron function, a test was performed to see if unc-42 mutants express a transcriptional glr-1-gfp reporter transgene. glr-1 encodes an AMPA-type glutamate receptor expressed by all interneurons of the forward and backward locomotory circuit, as well as by a subset of head motor neurons. No glr-1-gfp expression could be detected in the AVA, AVD and AVE interneurons in living mutants or in mutant worms that were fixed and stained with an anti-GFP antiserum to enhance GFP detection. These interneurons receive synaptic input from ASH and FLP and provide output to motor neurons that control backward movement. The cell bodies of the interneurons are still present in the mutants. However, it could not be determined if the morphology of the interneuron axons along the ventral cord was normal in unc-42 mutants because individual markers for these neurons are not available. Thus, the loss of GLR-1 receptor expression by the AVA, AVD and AVE interneurons contributes to the severe nose-touch phenotype of unc-42 mutants. Mutations in unc-42 also disrupt glr-1-gfp expression in the six RMD and two RME motor neurons, all of which innervate head muscles. The RME neurons are involved in foraging behavior, and the RMDs mediate a head withdrawal response to touch along the side of the nose that is also dependent on glr-1. Because head mobility is defective in unc-42 mutants, mutant animals were not tested for defects in head withdrawal (Baran, 1999).

UNC-42 expression can be detected in several cells as early as 260 minutes into embryogenesis and by comma stage in neurons at high levels in the head. Comma stage is the time when these neurons extend axons and establish connections. Expression in head neurons continues into adulthood. Expression of a transcriptional unc-42-gfp transgene, gmEx104, is abolished in unc-42 mutants, suggesting that unc-42 regulates its own expression. UNC-42 was strongly expressed in at least 20 pairs of neurons in the head, including the AVA, AVD and AVE interneurons, ASH sensory neurons, and RMD and SMB motor neurons. Other neurons that express high levels of UNC-42 include the AIN, AVH, AVJ, AVK, RIV, SAA and SIB interneurons. Low levels of unc-42 expression are detected in hypodermis and additional neurons in the head. Transient expression of UNC-42 protein is also detected in the DD motor neurons at hatching and at low levels in postembryonic ventral cord motor neurons derived from P11. In the tail, unc-42-gfp is strongly expressed in PVT (Baran, 1999).

UNC-42 is expressed by AVA and AVD interneurons, which transmit stimuli from the ALM and AVM mechansensory neurons and the ASH and FLP sensory neurons to motor neurons. To determine whether defects in the interneurons are responsible both for the sensory and locomotory defects of unc-42 mutants, an examination was made of the behavior of animals that were mosaic for unc-42 function. The Unc and Mec phenotypes are separable in the mosaic animals, reflecting distinct sites of unc-42 function. Mosaic animals could be isolated that were Unc but not Mec, as well as animals that were Mec but not Unc. AVA and AVD, the interneurons that mediate backward motion, are derived from the AB.a blastomere. The AVB and PVC interneurons that control forward movement, the ventral cord motor neurons that innervate body wall muscles, and many of the neurons that regulate head movement are all descendants of the AB.p blastomere. 11/11 mosaic worms that failed to respond to light touch to the anterior body, but moved normally, had lost wild-type unc-42 gene activity in AB.a. Six of these losses were in AB.aala, producing mosaic animals that had mutant AVD interneurons, wild-type AVM and ALM mechanosensory neurons and wild-type motor neurons. In contrast, 12/12 mosaic worms that were Unc, but responded to light touch, had lost wild-type unc-42 function in the AB.p lineage. A severe uncoordinated phenotype is observed only when all cells derived from AB.p are mutant for unc-42. Animals with single losses of wild-type gene function in AB.p descendents are wild type, and the animals with multiple losses in AB.p descendents exhibit only weak defects in locomotion. These results are consistent with the results of laser killing experiments, which showed that most members of a single motor neuron class or multiple classes must be killed to generate a severe uncoordinated phenotype. Because motor neurons are generated at multiple points in the AB.p cell lineage, only early losses in the lineage would affect gene function in a significant number of ventral cord motor neurons. In addition, some of the motor neurons that innervate head muscles and interneurons that coordinate head movement also express UNC-42 and are derived from AB.p. Genetic mosaic analysis supports the hypothesis that unc-42 acts cell-autonomously in the AVD interneurons for the response to body touch and is likely to act cell-autonomously in motor neurons for locomotion (Baran, 1999).

Physiology of AMPA receptors

Glutamatergic transmission at a principal neuron-interneuron synapse was investigated by dual whole-cell patch-clamp recording in rat hippocampal slices combined with morphological analysis. Evoked EPSPs with rapid time course (half duration = 4 ms; 34 degrees C) were generated at multiple synaptic contacts established on the interneuron dendrites close to the soma. The underlying postsynaptic conductance change shows a submillisecond rise and decay, due to the precise timing of glutamate release and the rapid deactivation of the postsynaptic AMPA receptors. Simulations based on a compartmental model of the interneuron indicate that the rapid postsynaptic conductance change determines the shape and the somatodendritic integration of EPSPs, thus enabling interneurons to detect synchronous principal neuron activity (Geiger, 1997).

AMPK acts as a molecular trigger to coordinate glutamatergic signals and adaptive behaviours during acute starvation

The stress associated with starvation is accompanied by compensatory behaviours that enhance foraging efficiency and increase the probability of encountering food. However, the molecular details of how hunger triggers changes in the activity of neural circuits to elicit these adaptive behavioural outcomes remains to be resolved. This study shows that AMP-activated protein kinase (AMPK; see Drosophila AMPKα) regulates neuronal activity to elicit appropriate behavioural outcomes in response to acute starvation, and this effect is mediated by the coordinated modulation of glutamatergic inputs. AMPK targets both the AMPA-type glutamate receptor GLR-1 (see Drosophila Glu-RIIA) and the metabotropic glutamate receptor MGL-1 (see Drosophila mGluR) in one of the primary circuits that governs behavioural response to food availability in C. elegans. Overall, this study suggests that AMPK acts as a molecular trigger in the specific starvation-sensitive neurons to modulate glutamatergic inputs and to elicit adaptive behavioural outputs in response to acute starvation (Ahmadi, 2016).

Trafficking and mobility of AMPA receptors

Insulin is expressed in discrete regions throughout the brain. Neurons synthesize and release insulin in response to membrane depolarization, and peripheral insulin penetrates the blood-brain barrier, entering the brain. Insulin receptors are also highly expressed in CNS neurons and localized to synapses. Since glucose utilization in neurons is largely insulin independent, CNS insulin may be involved in activities other than the regulation of glucose homeostasis, and recent evidence is consistent with a wide range of neuronal functions for insulin, including neuromodulation, growth and maturation, neuronal protection, and learning and memory. However, the detailed mechanisms by which brain insulin modifies neuronal function remains to be determined. Analogous to its function in peripheral tissues, where insulin produces rapid GLUT4 translocation to the plasma membrane to increase glucose uptake in these cells, insulin rapidly recruits functional GABAA receptors to postsynaptic domains in mature CNS neurons, resulting in a long-lasting enhancement of GABAA receptor-mediated synaptic transmission. Insulin can regulate the cell surface expression and hence the function of various other ion channels and neurotransmitter receptors. Hence, insulin may function as an important CNS neuromodulator by regulating the intracellular trafficking and plasma membrane expression of ion channels and neurotransmitter receptors (Man, 2000).

Redistribution of postsynaptic AMPA subtype glutamate receptors may regulate synaptic strength at glutamatergic synapses, but the mediation of the redistribution is poorly understood. AMPA receptors undergo clathrin-dependent endocytosis, which is accelerated by insulin in a GluR2 subunit-dependent manner. Insulin-stimulated endocytosis rapidly decreases AMPA receptor numbers in the plasma membrane, resulting in long-term depression (LTD) of AMPA receptor-mediated synaptic transmission in hippocampal CA1 neurons. Moreover, insulin-induced LTD and low-frequency stimulation-induced homosynaptic CA1 LTD are mutually occlusive and are both blocked by inhibiting postsynaptic clathrin-mediated endocytosis. Thus, controlling postsynaptic receptor numbers through endocytosis may be an important mechanism underlying synaptic plasticity in the mammalian CNS (Man, 2000).

To elucidate mechanisms that control and execute activity-dependent synaptic plasticity, AMPA receptors (AMPA-Rs) with an electrophysiological tag were expressed in rat hippocampal neurons. Long-term potentiation (LTP) or increased activity of the calcium/calmodulin-dependent protein kinase II (CaMKII) induce delivery of tagged AMPA-Rs into synapses. This effect is not diminished by mutating the CaMKII phosphorylation site on the GluR1 AMPA-R subunit, but is blocked by mutating a predicted PDZ domain interaction site. These results show that LTP and CaMKII activity drive AMPA-Rs to synapses by a mechanism that requires the association between GluR1 and a PDZ domain protein (Hayashi, 2000).

Both acute and chronic changes in AMPA receptor (AMPAR) localization are critical for synaptic formation, maturation, and plasticity. AMPARs are differentially sorted between recycling and degradative pathways following endocytosis. AMPAR sorting occurs in early endosomes and is regulated by synaptic activity and activation of AMPA and NMDA receptors. AMPAR internalization triggered by NMDAR activation is Ca2+-dependent, requires protein phosphatases, and is followed by rapid membrane reinsertion. Furthermore, NMDAR-mediated AMPAR trafficking is regulated by PKA and accompanied by dephosphorylation and rephosphorylation of GluR1 subunits at a PKA site. In contrast, activation of AMPARs (without NMDAR activation) targets AMPARs to late endosomes and lysosomes, independent of Ca2+, protein phosphatases, or PKA. These results demonstrate that activity regulates AMPAR endocytic sorting, providing a potential mechanistic link between rapid and chronic changes in synaptic strength (Ehlers, 2000).

A key feature of synaptogenesis and synapse maturation is the incorporation and stabilization of postsynaptic AMPARs. Activation of NMDARs switches AMPARs from a degradative pathway to a recycling pathway. This switch in AMPAR sorting could provide a mechanism for maintaining AMPARs at synapses that have an adequate amount of NMDAR activity and for removing AMPARs from synapses with insufficient NMDAR activity. Indeed, postsynaptic NMDARs precede AMPARs at many excitatory synapses during development, and the proportion of these NMDAR-only synapses increases when NMDARs are blocked, perhaps due to selective shunting of AMPARs to lysosomal degradation. Conversely, activation of NMDARs may trigger accumulation and stabilization of AMPARs by promoting AMPAR reinsertion. Thus, by controlling the degree of AMPAR recycling and degradation, NMDAR activation may ensure the maintenance or elimination of AMPARs at appropriate synapses (Ehlers, 2000).

Changes in AMPAR phosphorylation state and synaptic localization are two principal mechanisms proposed to account for long-term changes in synaptic strength. During LTD of hippocampal neuron synapses, GluR1 subunits are dephosphorylated on serine 845, and AMPARs redistribute away from synaptic sites, perhaps due to clathrin-dependent endocytosis. Results presented here provide a possible link between GluR1 dephosphorylation, AMPAR membrane trafficking, and LTD. In particular, these findings suggest a signaling pathway whereby activation of NMDARs triggers a protein phosphatase cascade involving PP2B and PP1 that leads to selective dephosphorylation of GluR1 subunits at serine 845. Dephosphorylation of GluR1, possibly in conjunction with dephosphorylation of components of the endocytic machinery or other AMPAR subunits, then results in AMPAR endocytosis, perhaps by promoting interaction with clathrin adaptors or destabilizing interactions with proteins such as NSF or GRIP. Potentiation of synapses may, in some cases, simply be the reverse of this process. Indeed, results here, showing that membrane insertion of AMPARs occurs simultaneously with phosphorylation at serine 845 and is reduced by PKA inhibitors, are consistent with a finding of selective phosphorylation of serine 845 following potentiation of previously depressed synapses. Coordination of AMPAR phosphorylation and dephosphorylation may thus regulate synaptic strength by regulating AMPAR trafficking (Ehlers, 2000).

A number of proteins interact with the C-terminal regions of GluR2 and GluR3. Two distinct interaction domains on the C-terminal of GluR2 (ct-GluR2) have so far been identified: an NSF binding site between residues 844 and 853 and an extreme ct-PDZ binding motif. The PDZ binding motif has been shown to interact with three PDZ domain-containing proteins: GRIP and ABP (one splice form is also known as GRIP2), which are closely related, and PICK1. The role of GRIP, ABP, and PICK1 interacting with the C-terminal GluR2 was investigated by infusing a ct-GluR2 peptide ('pep2-SVKI') into CA1 pyramidal neurons in hippocampal slices using whole-cell recordings. Pep2-SVKI, but not a control or PICK1 selective peptide, causes AMPAR-mediated EPSC amplitude to increase in approximately one-third of control neurons and in most neurons following the prior induction of LTD. Pep2-SVKI also blocks LTD; however, this occurs in all neurons. A PKC inhibitor prevents these effects of pep2-SVKI on synaptic transmission and LTD. A model is proposed in which the maintenance of LTD involves the binding of AMPARs to PDZ proteins to prevent their membrane reinsertion (reinsertion normally follows rapidly after activity dependent endocytosis). Evidence is presented that PKC regulates AMPAR reinsertion during dedepression (Daw, 2000).

AMPA-type glutamate receptors (AMPA-Rs) mediate a majority of excitatory synaptic transmission in the brain. In hippocampus, most AMPA-Rs are hetero-oligomers composed of GluR1/GluR2 or GluR2/GluR3 subunits. These AMPA-R forms display different synaptic delivery mechanisms. GluR1/GluR2 receptors are added to synapses during plasticity; this requires interactions between GluR1 and group I PDZ domain proteins. In contrast, GluR2/GluR3 receptors replace existing synaptic receptors continuously; this occurs only at synapses that already have AMPA-Rs and requires interactions by GluR2 with NSF and group II PDZ domain proteins. The combination of regulated addition and continuous replacement of synaptic receptors can stabilize long-term changes in synaptic efficacy and may serve as a general model for how surface receptor number is established and maintained (Shi, 2001).

The molecular interactions with GluR2 that may be necessary for the continuous synaptic delivery of receptors were investigated. Several proteins have been identified that interact with the carboxyl terminus of GluR2, including GRIP1/2(ABP), PICK1, and rDLG6. These proteins are all PDZ (PSD-95, DLG, ZO-1) domain-containing proteins, and the carboxyl terminus of GluR2 corresponds to a group II PDZ domain binding ligand. To examine if such interactions are necessary for the synaptic delivery of GluR2, a mutant GluR2 was generated with a tyrosine (Y) added at the end of the carboxyl terminus (+863Y), a mutation that prevents interaction between GluR2 and PDZ domain-containing proteins. Whole-cell recordings from transfected HEK 293 cells indicate that the recombinant receptor made of this mutant is functional and shows inward rectification similar to GluR2(R586Q)-GFP. However, when expressed in neurons, this mutation blocks synaptic delivery of this receptor, since cells expressing GluR2(R586Q, +863Y)-GFP show no change in rectification of AMPA-R-mediated responses. This indicates that the interactions between GluR2 and group II PDZ domain proteins are necessary for its continuous synaptic delivery. In addition, the amplitude of transmission onto these cells is depressed. The suppression of AMPA-R-mediated transmission may be explained as a dominant negative effect caused by GluR2(R586Q,+863Y)-GFP. This protein may still bind to other proteins (e.g., NSF) required for receptor synaptic delivery and thus compete with endogenous receptors for interactions with the delivery machinery and thereby block the delivery arm of the cycling of endogenous synaptic AMPA-Rs. The removal process may continue, leading to a synaptic depression (Shi, 2001).

Homomeric GluR1 receptors require activity, either LTP or increased CaMKII activity, to be driven into synapses. This process requires interactions between GluR1 and PDZ domain proteins. Hetero-oligomeric receptors composed of GluR1 and GluR2, which represent the majority of endogenous GluR1 in hippocampus, also require activity for their delivery. These results are consistent with a model in which GluR1 interacts with proteins that restrict hetero-oligomeric GluR1/GluR2 receptors from synaptic delivery. CaMKII activity may relieve this restriction. Mere relief of such restriction appears not to be sufficient for synaptic delivery, however, since GluR1 receptors lacking their carboxyl terminus or GluR1 receptors with a point mutation at the PDZ interaction site are not delivered to synapses. Thus, there appear to be additional protein interactions that effect synaptic delivery (Shi, 2001).

These studies provide direct evidence for two distinct mechanisms by which AMPA-Rs can be delivered to synapses. These two mechanisms can contribute to important aspects of synaptic function. GluR1/GluR2 delivery provides additional receptors following plasticity-inducing stimuli thereby effecting synaptic enhancement. These receptors can be delivered to silent synapses, converting them to functional ones. GluR2/GluR3 receptors can continuously replace synaptic receptors. Thus, this second process can act to preserve plastic changes in the face of protein turnover. How can the number of synaptic receptors be maintained during this continuous replacement? One possibility is that several proteins, in addition to GluR1/GluR2 hetero-oligomers, are delivered in tandem to synapses during plasticity. These proteins could serve as placeholders (i.e., 'slots') that could be filled with nonsynaptic GluR2/GluR3 hetero-oligomers if synaptic GluR1/GluR2 or GluR2/GluR3 hetero-oligomers leave the synapse. GluR1/GluR2 hetero-oligomers may leave 'slots' more slowly (in days) compared to GluR2/GluR3 hetero-oligomers (in minutes), thus explaining why expression of GluR2 carboxyl terminus or infusion of G10/pep2m (a short peptide that mimics the predicted interaction site on GluR2 with NSF) depresses transmission only partially and the GluR2 carboxyl terminus does not block LTP at 1 hr. Some of these delivered proteins could also serve as 'slots' for the eventual addition of NMDA-Rs. Delivery of proteins will likely increase the physical size of synaptic contact and could possibly communicate to the presynaptic side eventually leading to the matching of pre- and postsynaptic size and function (Shi, 2001).

Recent studies show that AMPA receptor trafficking is important in synaptic plasticity. However, the signaling controlling this trafficking is poorly understood. Small GTPases have diverse neuronal functions and their perturbation is responsible for several mental disorders. The roles of small GTPases Ras and Rap in the postsynaptic signaling underlying synaptic plasticity were examined. Ras relays the NMDA receptor and CaMKII signaling that drives synaptic delivery of AMPA receptors during long-term potentiation. In contrast, Rap mediates NMDA-receptro-dependent removal of synaptic AMPA receptors that occurs during long-term depression. Ras and Rap exert their effects on AMPA receptors that contain different subunit composition. Thus, Ras and Rap, whose activity can be controlled by postsynaptic enzymes, serve as independent regulators for potentiating and depressing central synapses (Zhu, 2002).

The cytoplasmic carboxyl tails of AMPA receptor constituent subunits, which show either long or short forms, control the trafficking characteristics of AMPA receptors. AMPA receptors with long cytoplasmic tails (e.g., GluR1 or GluR4) are restricted from synapses and delivered to synapses during activity-induced synaptic enhancement. AMPA-Rs with only short cytoplasmic tails (e.g., GluR2 or GluR3) cycle continuously from nonsynaptic to synaptic sites in an activity independent manner; their number at synapses can be reduced after activity-induced synaptic depression. The results indicate that spontaneous neural activity continuously adds into the synapses AMPA receptors containing long cytoplasmic tails via Ras activity and continuously removes from synapses AMPA receptors containing only short cytoplasmic tails via Rap activity. Similarly, this study argues that LTP adds to synapses AMPA receptors containing long cytoplasmic tails while LTD removes receptors containing only short cytoplasmic tails. These results indicate the existence of a replacement mechanism at synapses that can exchange AMPA receptors containing long cytoplasmic tails with those containing only short cytoplasmic tails, which explains the observation that LTP and LTD can reverse each other. In fact, this replacement has previously been detected and may itself be under some form of regulation. For example, a more robust replacement appears to occur in dissociated neuronal preparations where LTD stimuli lead to rapid removal of AMPA receptors with long cytoplasmic tails. Thus, the rate of receptor replacement and relative number of receptors with long or short cytoplasmic tails at a synapse may control the amount of LTP or LTD available at that synapse (Zhu, 2002).

Proteins that bind to the cytoplasmic tails of AMPA receptors control receptor trafficking and thus the strength of postsynaptic responses. AP2, a clathrin adaptor complex important for endocytosis, associates with a region of GluR2 that overlaps the NSF binding site. NSF is a hexameric ATPase involved generally in membrane fusion events. Peptides used to interfere with NSF binding also antagonize GluR2-AP2 interaction. Using GluR2 mutants and peptide variants that dissociate NSF and AP2 interaction, it has been found that AP2 is involved specifically in NMDA receptor-induced (but not ligand-dependent) internalization of AMPA receptors, and is essential for hippocampal long-term depression (LTD). NSF function, in contrast, is needed to maintain synaptic AMPA receptor responses, but is not directly required for NMDA receptor-mediated internalization and LTD (Lee, 2002).

The PICK1 protein interacts in neurons with the AMPA-type glutamate receptor subunit 2 (GluR2) and with several other membrane receptors via its single PDZ domain. PICK1 also binds in neurons and in heterologous cells to protein kinase Calpha (PKCalpha) and that the interaction is highly dependent on the activation of the kinase. The formation of PICK1-PKCalpha complexes is strongly induced by TPA, and PICK1-PKCalpha complexes are cotargeted with PICK1-GluR2 complexes to spines, where GluR2 is found to be phosphorylated by PKC on serine 880. PICK1 also reduces the plasma membrane levels of the GluR2 subunit, consistent with a targeting function of PICK1 and a PKC-facilitated release of GluR2 from the synaptic anchoring proteins ABP and GRIP. This work indicates that PICK1 functions as a targeting and transport protein that directs the activated form of PKCalpha to GluR2 in spines, leading to the activity-dependent release of GluR2 from synaptic anchor proteins and the PICK1-dependent transport of GluR2 from the synaptic membrane (Perez, 2001).

Recent studies documenting a role for local protein synthesis in synaptic plasticity have lead to interest in the opposing process, protein degradation, as a potential regulator of synaptic function. The ubiquitin-conjugation system identifies, modifies, and delivers proteins to the proteasome for degradation. Both the proteasome and ubiquitin are present in the soma and dendrites of hippocampal neurons. Since the trafficking of glutamate receptors (GluRs) is thought to underlie some forms of synaptic plasticity, whether blocking proteasome activity affects the agonist-induced internalization of GluRs in cultured hippocampal neurons was examined. Treatment with the glutamate agonist AMPA induces a robust internalization of GluRs. In contrast, brief pretreatment with proteasome inhibitors completely prevents the internalization of GluRs. To distinguish between a role for the proteasome and a possible diminution of the free ubiquitin pool, a chain elongation defective ubiquitin mutant (UbK48R) was expressed that causes premature termination of polyubiquitin chains but, importantly, can serve as a substrate for mono-ubiquitin-dependent processes. Expression of K48R in neurons severely diminishes AMPA-induced internalization establishing a role for the proteasome. These data demonstrate the acute (e.g., minutes) regulation of synaptic function by the ubiquitin-proteasome pathway in mammalian neurons (Patrick, 2003).

Long-term maintenance and modification of synaptic strength involve the turnover of neurotransmitter receptors. Glutamate receptors are constitutively and acutely internalized, presumptively through clathrin-mediated receptor endocytosis. cpg2 is a brain-specific splice variant of the syne-1 gene that encodes a protein specifically localized to a postsynaptic endocytotic zone of excitatory synapses. RNAi-mediated CPG2 knockdown increases the number of postsynaptic clathrin-coated vesicles, some of which traffic NMDA receptors, disrupts the constitutive internalization of glutamate receptors, and inhibits the activity-induced internalization of synaptic AMPA receptors. Manipulating CPG2 levels also affects dendritic spine size, further supporting a function in regulating membrane transport. These results suggest that CPG2 is a key component of a specialized postsynaptic endocytic mechanism devoted to the internalization of synaptic proteins, including glutamate receptors. The activity dependence and distribution of cpg2 expression further suggest that it contributes to the capacity for postsynaptic plasticity inherent to excitatory synapses (Cottrell, 2004).

Screens for plasticity-related genes have identified multiple transcripts that encode synaptic proteins, suggesting that genes induced by activity often function in normal synaptic processes. candidate plasticity gene 2 (cpg2) was isolated in a screen for transcripts upregulated by kainic acid-induced seizures in the rat dentate gyrus, and its expression is regulated during development and by sensory experience. cpg2 is a splice variant of the syne-1 gene, a large gene that encodes a protein with an actin binding domain at the N terminus and a nuclear transmembrane domain at the C terminus, separated by a long helical region. The cpg2 transcript is derived from a portion of the separator region, encodes a protein with homologies to dystrophin, and contains motifs predicting a structural function, including several spectrin repeats and coiled coils. Proteins with these motifs often play a central role in organizing protein complexes (Cottrell, 2004).

cpg2 is expressed only in the brain and encodes a protein that localizes specifically to the postsynaptic endocytic zone of excitatory synapses. Evidence is presented that CPG2 is a critical component of the postsynaptic endocytic pathway that mediates both constitutive and activity-regulated glutamate receptor internalization. It is hypothesized that CPG2 is a key component of a specialization that is devoted to the internalization of postsynaptic proteins at synapses capable of plasticity (Cottrell, 2004).

The synapse contains densely localized and interacting proteins that enable it to adapt to changing inputs. A Ca2+-sensitive protein complex is described involved in the regulation of AMPA receptor synaptic plasticity. The complex is comprised of (1) MUPPI, a multi-PDZ domain-containing protein, (2) SynGAP, a synaptic GTPase-activating protein, and (3) the Ca2+/calmodulin-dependent kinase CaMKII. In synapses of hippocampal neurons, SynGAP and CaMKII are brought together by direct physical interaction with the PDZ domains of MUPP1, and in this complex, SynGAP is phosphorylated. Ca2+CaM binding to CaMKII dissociates it from the MUPP1 complex, and Ca2+, entering the cell via the NMDAR, drives the dephosphorylation of SynGAP. Specific peptide-induced SynGAP dissociation from the MUPP1-CaMKII complex results in SynGAP dephosphorylation accompanied by P38 MAPK inactivation, potentiation of synaptic AMPA responses, and an increase in the number of AMPAR-containing clusters in hippocampal neuron synapses. siRNA-mediated SynGAP knockdown confirms these results. These data implicate SynGAP in NMDAR- and CaMKII-dependent regulation of AMPAR trafficking (Krapivinsky, 2004).

Removal of synaptic AMPA receptors is important for synaptic depression. This study characterizes the roles of individual subunits in the inducible redistribution of AMPA receptors from the cell surface to intracellular compartments in cultured hippocampal neurons. The intracellular accumulation of GluR2 and GluR3 but not GluR1 is enhanced by AMPA, NMDA, or synaptic activity. After AMPA-induced internalization, homomeric GluR2 enters the recycling pathway, but following NMDA, GluR2 is diverted to late endosomes/lysosomes. In contrast, GluR1 remains in the recycling pathway, and GluR3 is targeted to lysosomes regardless of NMDA receptor activation. Interaction with NSF plays a role in regulated lysosomal targeting of GluR2. GluR1/GluR2 heteromeric receptors behave like GluR2 homomers, and endogenous AMPA receptors show differential activity-dependent sorting similar to homomeric GluR2. Thus, GluR2 is a key subunit that controls recycling and degradation of AMPA receptors after internalization (Lee, 2004).

Fast excitatory synaptic transmission in the mammalian brain is mediated primarily by AMPA-type glutamate receptors. Recently, the dynamic redistribution of AMPA receptors in and out of synapses has emerged as an important mechanism for certain forms of long-lasting synaptic modification. In the hippocampal CA3-CA1 synapse, net delivery of AMPA receptors to the postsynaptic membrane leads to long-term potentiation (LTP), whereas net removal of AMPA receptors by internalization from the surface seems to underlie long-term depression (LTD) (Lee, 2004 and references therein).

AMPA receptors are heterotetrameric complexes composed of various combinations of four subunits (GluR1-4). In the adult hippocampus, two major subtypes of AMPA receptors exist that contain either GluR1 and GluR2, or GluR2 and GluR3 subunits. GluR4 is mainly expressed early in development. Individual AMPA receptor subunits interact via their cytoplasmic tails with different sets of proteins. These specific protein interactions are believed to regulate the trafficking and synaptic targeting of AMPA receptors (Lee, 2004 and references therein).

Subunit-specific functions governing the synaptic delivery of AMPA receptors have been uncovered by elegant electrophysiological assays in hippocampal slice cultures and corroborated by cell biological studies in dissociated cultures. GluR1 is the key subunit that 'drives' AMPA receptors to the surface and to synapses in response to NMDA receptor stimulation and activation of CaMKII, resulting in synaptic potentiation. GluR2 on the other hand is delivered constitutively to synapses, replacing existing receptors with no change in synaptic strength. In heteromeric receptors, GluR1 acts 'dominantly' over GluR2, whereas GluR2 acts dominantly over GluR3. Thus, in the hippocampus, it is believed that GluR1/2 heteromers are delivered to synapses during activity-dependent synaptic potentiation, such as LTP, whereas GluR2/3 heteromers cycle continuously between the postsynaptic membrane and intracellular compartments (Lee, 2004).

In contrast to synaptic delivery, little is known about the roles of individual subunits in the removal of AMPA receptors from synapses. Earlier studies on endogenous AMPA receptors suggest that both GluR1- and GluR2-containing receptors can undergo inducible internalization upon stimulation with AMPA, NMDA, or insulin, but GluR1 and GluR2 are often coassembled in the same heteromeric receptor, so these studies could not distinguish their subunit-specific roles in endocytosis (Lee, 2004).

After endocytosis, AMPA receptors undergo endosomal sorting like any other internalized membrane protein -- ultimately, they can be recycled back to the surface membrane or degraded in lysosomes. However, it is not known whether the endosomal sorting of AMPA receptor depends on subunit composition or how activity might affect the intracellular fate of specific subunits (Lee, 2004).

One caveat of AMPA receptor internalization studies is that, due to technical reasons, the quantitation of 'internalization' (e.g., by surface biotinylation or antibody feeding assays) does not strictly measure endocytosis per se but rather the amount of surface receptors that have internalized and that remain in intracellular compartments. Because AMPA receptors cycle rapidly between intracellular and plasma membranes, the amount of internalized receptor is strongly affected by the rate of recycling to the surface as well as by the rate of endocytosis. Therefore, the terms 'redistribution to intracellular compartments' or 'intracellular accumulation' are preferred rather than 'internalization' to signify the amount of surface receptor that is redistributed to internal pools. Insofar as it reflects a shift from surface to intracellular compartments, the measure of intracellular accumulation of internalized receptors is still relevant to removal of AMPA receptors from the synapse (Lee, 2004).

In this report, the subunit rules have been investigated that govern the activity-dependent redistribution of surface AMPA receptors to intracellular compartments and that determine the intracellular sorting of receptors after they are internalized. In contrast to inducible synaptic delivery, where GluR1 plays the key role, the GluR2 subunit is the primary determinant of inducible intracellular accumulation of AMPA receptors. GluR2 controls the postendocytic trafficking of internalized AMPA receptors to either recycling or lysosomal degradation pathways, at least in part dependent on its interaction with NSF (Lee, 2004).

Synaptic trafficking of AMPA-Rs, controlled by small GTPase Ras signaling, plays a key role in synaptic plasticity. However, how Ras signals synaptic AMPA-R trafficking is unknown. This study shows that low levels of Ras activity stimulate extracellular signal-regulated kinase kinase (MEK)-p42/44 MAPK (extracellular signal-regulated kinase [ERK]) signaling, whereas high levels of Ras activity stimulate additional Pi3 kinase (Pi3K)-protein kinase B (PKB) signaling, each accounting for ~50% of the potentiation during long-term potentiation (LTP). Spontaneous neural activity stimulates the Ras-MEK-ERK pathway that drives GluR2L into synapses. In the presence of neuromodulator agonists, neural activity also stimulates the Ras-Pi3K-PKB pathway that drives GluR1 into synapses. Neuromodulator release increases with increases of vigilance. Correspondingly, Ras-MEK-ERK activity in sleeping animals is sufficient to deliver GluR2L into synapses, while additional increased Ras-Pi3K-PKB activity in awake animals delivers GluR1 into synapses. Thus, state-dependent Ras signaling, which specifies downstream MEK-ERK and Pi3K-PKB pathways, differentially control GluR2L- and GluR1-dependent synaptic plasticity (Y. Qin, 2005).

The results suggest that Ras signals synaptic insertion of AMPA-Rs via stimulating phosphorylation of S845 and S831 of GluR1 and S842 of GluR2L. Because Ras downstream signaling molecules ERK and PKB are unlikely to directly phosphorylate GluR1 and GluR2L, other molecules probably exist at synapses to relay the signaling. Two likely candidates are cAMP-dependent protein kinase (PKA) and calcium/calmodulin-dependent protein kinase II (CaMKII), since they can phosphorylate S845 and S831 of GluR1, respectively. Protein kinase C (PKC) is another putative candidate because it can phosphorylate S831, as well as S845, albeit to a lesser extent. However, whether ERK and PKB stimulate PKA, CaMKII, and/or PKC remains to be examined. In contrast, serine/threonine kinases Rsk (see RSK) and mTOR-S6K, which relay downstream Ras signaling in nonneuronal cells, may also serve as the relays. In particular, both Rsk and mTOR are expressed at synapses, and disruption of Rsk and mTOR signaling leads to mental retardation. Thus, determining the precise functional relationships (i.e., sequential or parallel, and downstream or upstream) of the signaling molecules involved in Ras pathways during LTP is central to answer many important questions related to the mechanisms of synaptic plasticity (Y. Qin, 2005).

Though NMDA-R-dependent forms of synaptic plasticity have been extensively examined in vitro, little is known about their properties in the intact brain. Previous studies have shown that both the occurrence and magnitude of LTP induced by electric tetanization stimuli are higher in awake than sleeping animals. However, the mechanisms of this state-dependent LTP are unclear, because the LTP-inducing stimuli do not mimic physiological activity in these states. Both GluR2L and GluR1 mediate LTP in juvenile and adult animals. This study reports that synaptic activity in sleeping animals is sufficient for driving GluR2L but not GluR1 into synapses, whereas synaptic activity in awake animals drives more GluR2L as well as GluR1 into synapses, suggesting more synaptic plasticity in awake animals. Based on these findings, it is proposed that state-dependent physiological factors, such as neuromodulators, may control the state-dependent plasticity. Indeed, neuromodulator agonists (for example, histamine, a monoamine neuromodulator) can drive more GluR2L as well as GluR1 into synapses, by stimulating Ras signaling. These results are consistent with the previous findings that neuromodulators, whose release increases in general during the awake behavioral state, stimulate ERK and Pi3K signaling and potentiate LTP. It remains to be determined whether other state-dependent factors (i.e., neuronal firing patterns, hormones, and neurotrophic factors) regulate synaptic plasticity and how these factors interact in the intact brain (Y. Qin, 2005).

Memory consolidation seems to occur during sleep and waking, while learning occurs in the conscious state. It is believed that the learning and memory processes require synaptic plasticity. This study shows that synaptic potentiation is present in both sleeping and awake states. Interestingly, synaptic plasticity in sleeping and awake states is controlled by different levels of Ras signaling and mediated by trafficking of distinct AMPA-Rs. The obvious puzzles are whether and how Ras-regulated, subunit-specific AMPA-R trafficking correlates with the different forms of memory consolidation and learning (e.g., declarative vs. procedural or explicit vs. implicit). Manipulating Ras signaling and trafficking of AMPA-Rs in intact animals (e.g., in vivo expression of Ras mutants and GluRct-GFP) during different behavioral states (e.g., slow-wave sleep, REM sleep, quiescent alert, and active exploring) and monitoring changes in learning and memory behavior promise to reveal new insights into these pivotal questions (Y. Qin, 2005).

NMDA receptors (NMDARs) control bidirectional synaptic plasticity by regulating postsynaptic AMPA receptors (AMPARs). NMDAR activation can have differential effects on AMPAR trafficking, depending on the subunit composition of NMDARs. In mature cultured neurons, NR2A-NMDARs promote, whereas NR2B-NMDARs inhibit, the surface expression of GluR1, primarily by regulating its surface insertion. In mature neurons, NR2B is coupled to inhibition rather than activation of the Ras-ERK pathway, which drives surface delivery of GluR1. Moreover, the synaptic Ras GTPase activating protein (GAP) SynGAP is selectively associated with NR2B-NMDARs in brain and is required for inhibition of NMDAR-dependent ERK activation. Preferential coupling of NR2B to SynGAP could explain the subtype-specific function of NR2B-NMDARs in inhibition of Ras-ERK, removal of synaptic AMPARs, and weakening of synaptic transmission (Kim, 2005 ).

The ERK1/2 signaling pathway is activated by calcium influx through NMDARs and plays an important role in synaptic plasticity and cell survival. NMDAR-dependent ERK activation involves the small GTPase Ras, which is stimulated by specific guanine nucleotide exchange factors (GEFs) and inhibited by GTPase activating proteins (GAPs). The RasGEF RasGRF1 is reported to bind directly to the NR2B subunit of NMDARs. SynGAP, a RasGAP highly enriched in the postsynaptic density (PSD), can associate with NMDARs through binding to PSD-95 family proteins. The exact function of these Ras regulatory proteins in synaptic plasticity has not been established, and how they are functionally coupled to NMDARs remains unclear (Kim, 2005).

Altered AMPAR trafficking has emerged as a major postsynaptic mechanism for the expression of synaptic plasticity. A prevailing model is that NMDAR-dependent LTP is mediated by the surface insertion and synaptic delivery of GluR1, that is driven by CaM kinase II and the Ras-ERK pathway. In contrast, LTD is supposed to result, at least in part, from the removal of synaptic AMPARs by the increased endocytosis and/or reduced recycling of GluR2/3 subunits (Kim, 2005).

This study investigates the links between NMDAR subtypes, Ras-ERK signaling, and AMPAR trafficking. NR2A and NR2B are found to have antagonistic actions on Ras-ERK activation and AMPAR distribution in mature neurons. NR2A-NMDARs promote, whereas NR2B-NMDARs inhibit, the surface expression of GluR1 -- primarily by regulating GluR1 surface insertion. Potentially accounting for this difference is that NR2B is coupled to the inhibition rather than the activation of the Ras-ERK pathway. This functional coupling is correlated with the specific biochemical association of SynGAP with NR2B-NMDARs (Kim, 2005).

PICK1 and ABP/GRIP bind to the AMPA receptor (AMPAR) GluR2 subunit C terminus. Transfer of the receptor from ABP/GRIP to PICK1, facilitated by GluR2 S880 phosphorylation, may initiate receptor trafficking. This study reports protein interactions that regulate these steps. The PICK1 BAR domain interacts intermolecularly with the ABP/GRIP linker II region and intramolecularly with the PICK1 PDZ domain. Binding of PKCa or GluR2 to the PICK1 PDZ domain disrupts the intramolecular interaction and facilitates the PICK1 BAR domain association with ABP/GRIP. Interference with the PICK1-ABP/GRIP interaction impairs S880 phosphorylation of GluR2 by PKC and decreases the constitutive surface expression of GluR2, the NMDA-induced endocytosis of GluR2, and recycling of internalized GluR2. It is suggested that the PICK1 interaction with ABP/GRIP is a critical step in controlling GluR2 trafficking (Lu, 2005).

NMDA receptor-dependent long-term potentiation and long-term depression (LTD) involve changes in AMPA receptor activity and postsynaptic localization that are in part controlled by glutamate receptor 1 (GluR1) subunit phosphorylation. The scaffolding molecule A-kinase anchoring protein (AKAP)79/150 targets both the cAMP-dependent protein kinase (PKA) and protein phosphatase 2B/calcineurin (PP2B/CaN) to AMPA receptors to regulate GluR1 phosphorylation. Brief NMDA receptor activation leads to persistent redistribution of AKAP79/150 and PKA-RII, but not PP2B/CaN, from postsynaptic membranes to the cytoplasm in hippocampal slices. Similar to LTD, AKAP79/150 redistribution requires PP2B/CaN activation and is accompanied by GluR1 dephosphorylation and internalization. Using fluorescence resonance energy transfer microscopy in hippocampal neurons, it has been demonstrated that PKA anchoring to AKAP79/150 is required for NMDA receptor regulation of PKA-RII localization and that movement of AKAP-PKA complexes underlies PKA redistribution. These findings suggest that LTD involves removal of AKAP79/150 and PKA from synapses in addition to activation of PP2B/CaN. Movement of AKAP79/150-PKA complexes from the synapse could further favor the actions of phosphatases in maintaining dephosphorylation of postsynaptic substrates, such as GluR1, that are important for LTD induction and expression. In addition, these observations demonstrate that AKAPs serve not solely as stationary anchors in cells but also as dynamic signaling components (Smith, 2006).

AMPA-type glutamate receptors undergo constitutive and ligand-induced internalization that requires dynamin and the clathrin adaptor complex AP-2. An atypical basic motif within the cytoplasmic tails of AMPA-type glutamate receptors directly associates with mu2-adaptin by a mechanism similar to the recognition of the presynaptic vesicle protein synaptotagmin 1 by AP-2. A synaptotagmin 1-derived AP-2 binding peptide competes the interaction of the AMPA receptor subunit GluR2 with AP-2mu and increases the number of surface active glutamate receptors in living neurons. Moreover, fusion of the GluR2-derived tail peptide with a synaptotagmin 1 truncation mutant restores clathrin/AP-2-dependent internalization of the chimeric reporter protein. These data suggest that common mechanisms regulate AP-2-dependent internalization of pre- and post-synaptic membrane proteins (Kastning, 2007).

Synaptic activity regulates the postsynaptic accumulation of AMPA receptors over timescales ranging from minutes to days. Indeed, the regulated trafficking and mobility of GluR1 AMPA receptors underlies many forms of synaptic potentiation at glutamatergic synapses throughout the brain. However, the basis for synapse-specific accumulation of GluR1 is unknown. This study reports that synaptic activity locally immobilizes GluR1 AMPA receptors at individual synapses. Using single-molecule tracking together with the silencing of individual presynaptic boutons, it was demonstrated that local synaptic activity reduces diffusional exchange of GluR1 between synaptic and extraynaptic domains, resulting in postsynaptic accumulation of GluR1. At neighboring inactive synapses, GluR1 is highly mobile with individual receptors frequently escaping the synapse. Within the synapse, spontaneous activity confines the diffusional movement of GluR1 to restricted subregions of the postsynaptic membrane. Thus, local activity restricts GluR1 mobility on a submicron scale, defining an input-specific mechanism for regulating AMPA receptor composition and abundance (Ehlers, 2007).

A primary determinant of the strength of neurotransmission is the number of AMPA-type glutamate receptors (AMPARs) at synapses. However, a mechanistic understanding of how the number of synaptic AMPARs is regulated is lacking. This study shows that UNC-116, the C. elegans homolog of vertebrate kinesin-1 heavy chain (KIF5), modifies synaptic strength by mediating the rapid delivery, removal, and redistribution of synaptic AMPARs. Furthermore, by studying the real-time transport of C. elegans AMPAR subunits in vivo, it was demonstrated that although homomeric GLR-1 AMPARs can diffuse to and accumulate at synapses in unc-116 mutants, glutamate-gated currents are diminished because heteromeric GLR-1/GLR-2 receptors do not reach synapses in the absence of UNC-116/KIF5-mediated transport. These data support a model in which ongoing motor-driven delivery and removal of AMPARs controls not only the number but also the composition of synaptic AMPARs, and thus the strength of synaptic transmission (Hoerndli, 2013).

Syntaxin-4 defines a domain for activity-dependent exocytosis in dendritic spines

Changes in postsynaptic membrane composition underlie many forms of learning-related synaptic plasticity in the brain. At excitatory glutamatergic synapses, fusion of intracellular vesicles at or near the postsynaptic plasma membrane is critical for dendritic spine morphology, retrograde synaptic signaling, and long-term synaptic plasticity. Whereas the molecular machinery for exocytosis in presynaptic terminals has been defined in detail, little is known about the location, kinetics, regulation, or molecules involved in postsynaptic exocytosis. This study shows that an exocytic domain adjacent to the postsynaptic density (PSD) enables fusion of large, AMPA receptor-containing recycling compartments during elevated synaptic activity. Exocytosis occurs at microdomains enriched in the plasma membrane t-SNARE syntaxin 4 (Stx4), and disruption of Stx4 impairs both spine exocytosis and long-term potentiation (LTP) at hippocampal synapses. Thus, Stx4 defines an exocytic zone that directs membrane fusion for postsynaptic plasticity, revealing a novel specialization for local membrane traffic in dendritic spines (Kennedy, 2010).

This study, carried out with cultured hippocampal cells and hippocampal slices, employed a novel sensor of postsynaptic membrane trafficking that allowed simultaneous visualization of the location of exocytosis and the behavior of exocytic cargo prior to, during, and following synaptic activity. Synaptic activation was shown to trigger fusion of large AMPA receptor-containing recycling compartments to the plasma membrane (PM) of dendritic spines in an all-or-none manner. Spine exocytosis occurs at sites enriched for the SNARE protein Stx4 immediately lateral to the PSD. Accordingly, disruption of Stx4 either acutely or chronically blocks membrane fusion and cargo delivery triggered by synaptic activation and acute inhibition of Stx4 abolishes LTP. Stx4 thus defines a SNARE-based exocytic zone for activity-dependent spine modification required for synaptic plasticity (Kennedy, 2010).

Although a crucial role for postsynaptic exocytosis in synaptic plasticity has long been appreciated, the site(s) of membrane insertion has been unknown and controversial. Recent studies have suggested that activity stimulates exocytosis of synaptic cargo exclusively in the soma and dendritic shafts, while others support a more local trafficking pathway within activated dendritic spines. This study provides direct evidence for activity-regulated exocytic trafficking within dendritic spines. Upon stimulation, intra-spine endosomes undergo abrupt fusion with the spine PM. Prior to fusion, spine-localized endosomes load with endogenous GluR1 and exogenous transferrin, indicating that these structures participate in ongoing recycling of plasma membrane proteins, including neurotransmitter receptors. Importantly, this study shows that, relative to endogenous GluR1, exogenously expressed SEP- GluR1 traffics less efficiently through spine recycling endosomes (REs). Specifically, less than half of spine REs contain SEP-GluR1 compared to nearly complete labeling of spine REs after endogenous GluR1 antibody feeding. This could be due to incomplete stoichiometric association with AMPA receptor regulatory subunits (e.g., TARPS), and may explain differences between the current results and studies that failed to observe spine exocytosis using SEP-GluR1 (Kennedy, 2010).

These results indicate that spine-localized recycling endosomes sense synaptic activity and rapidly alter the composition of the spine PM in dramatic, presumably single-step exocytic events at sites adjacent to the PSD. The location of exocytosis adjacent to the PSD indicates that incorporation of newly exocytosed membrane proteins into the PSD will be limited by escape and lateral diffusion from the initial site of fusion. The spine neck sequesters newly inserted material within the spine head for tens of seconds, increasing the likelihood of synaptic incorporation and perhaps shielding unstimulated neighboring synapses from newly inserted plasticity factors. Thus, while exocytic trafficking may be necessary for synaptic plasticity, short-range diffusion to the PSD within the confines of the spine head will likewise limit the rate and range of cargo incorporation into the synapse. Intriguingly, at many spines, the fraction of newly inserted SEP-GluR1 remains nearly constant for several minutes following exocytosis, even though co-exocytosed TfR-mCh quickly diffuse away. Although measuring spine exocytosis following more refined activity manipulations will be required to directly link spine exocytosis with homosynaptic LTP, this quantal mode of incorporation of AMPA receptors mirrors the single-step potentiation of AMPA receptor currents reported at individual synapses following local glutamate uncaging or weak afferent stimulation. In particular, both events are NMDA receptor-dependent, occur in a 'digital' all-or-none fashion, and are refractory to further stimulation. A possible mechanism that could set the refractory period for additional potentiation is the genesis of new spine REs or the mobilization existing REs into spines, the latter of which involves myosin Vb-mediated translocation of dendritic endosomes into spines (Kennedy, 2010).

Stx4, a PM SNARE protein expressed in brain, mediates a majority of activity-triggered membrane trafficking events from postsynaptic recycling compartments. Stx4 localizes to dendritic spines where it marks sites where spine endosomes fuse with the spine PM, thus enabling synapse-specific membrane delivery. In concert with previously described endocytic zones, the discovery of spine exocytic machinery raises the intriguing possibility of a micron-scale membrane trafficking circuit that determines the unique properties of individual synapses. Indeed, disrupting this circuit by eliminating postsynaptic endocytic zones results in loss of synaptic AMPA receptors, loss of local AMPA receptor recycling, and impaired synaptic potentiation. This study has shown that recycling cargo, including AMPA receptors can be exocytosed directly in spines, indicating that both exocytosis and endocytosis can be spatially restricted on a micron scale. In addition to AMPA receptors, many other postsynaptic membrane proteins are known to traffic through recycling endosomal compartments, including N-cadherin, L-type voltage-gated calcium channels, A-type potassium channels and metabotropic glutamate receptors. Likewise, soluble factors, including neurotrophins, may also be released via spine exocytosis. Although further experiments are needed to reveal the complete array of factors trafficked through spine endosomes and exocytosed at Stx4 exocytic domains, it is speculated that these organelles act as input-specific, activity- triggered plasticity modules that harbor and deliver a variety of membrane and secreted factors that modify diverse features of glutamatergic synapses. Further, it is interesting to note that unlike the classic presynaptic syntaxin Stx1, Stx4 associates with and is tightly regulated by the actin cytoskeleton, an adaptation particularly suited for actin-rich dendritic spines. Ultimately, spatially directed exocytosis together with retention of receptors and other exocytic cargo in the PSD, could both acutely augment and persistently confine new components of a given synapse for enduring changes in synaptic strength. More broadly, localized Stx4 exocytic domains may provide a general paradigm for stimulus-dependent regulation of local membrane composition on a micron scale (Kennedy, 2010).

A GluR1-cGKII interaction regulates AMPA receptor trafficking

Trafficking of AMPA receptors (AMPARs) is regulated by specific interactions of the subunit intracellular C-terminal domains (CTDs) with other proteins, but the mechanisms involved in this process are still unclear. This study found that the GluR1 CTD binds to cGMP-dependent protein kinase II (cGKII) adjacent to the kinase catalytic site. Binding of GluR1 is increased when cGKII is activated by cGMP. cGKII and GluR1 form a complex in the brain, and cGKII in this complex phosphorylates GluR1 at S845, a site also phosphorylated by PKA. Activation of cGKII by cGMP increases the surface expression of AMPARs at extrasynaptic sites. Inhibition of cGKII activity blocks the surface increase of GluR1 during chemLTP and reduces LTP in the hippocampal slice. This work identifies a pathway, downstream from the NMDA receptor (NMDAR) and nitric oxide (NO), which stimulates GluR1 accumulation in the plasma membrane and plays an important role in synaptic plasticity (Serulle, 2007).

NMDAR stimulation activates nNOS and production of NO, which results in cGMP production and cGKII activation. A major mechanism for expression of NMDAR-dependent LTP involves the synaptic insertion of GluR1. This study reports that, following activation by the NMDAR, cGKII binds to GluR1 and phosphorylates S845, leading to an increase of GluR1 in the plasma membrane. Notably, a cGKII dominant-negative inhibitor peptide blocked the cGMP-dependent increase of GluR1 surface expression, prevented the increase in amplitude and frequency of mEPSCs after chemLTP, and strongly reduced LTP in hippocampal slices. These results demonstrate a mechanism in which the NMDAR regulates AMPAR trafficking during LTP via NO and cGKII (Serulle, 2007).

Because NO is produced at postsynaptic sites and can diffuse through lipid membranes, initial studies of NO-dependent plasticity focused on presynaptic NO function through retrograde mechanisms. Some results were controversial, possibly because different methodologies were employed. Indeed, cGMP derivatives only facilitate LTP maximally if briefly applied when the NMDA receptor is active, and deviating protocols would lead to conflicting results. More recently, the use of new NO donors and NOS antagonists (Bon, 2003; Puzzo, 2005), both in vitro and in vivo (Feil, 2005), has demonstrated a role of the NO cascade in synaptic plasticity. Interestingly, as reported here, both the sGC inhibitor ODQ and the cGK inhibitor KT5823 were found to block LTP. Nonetheless, specific molecular mechanisms underlying the effects of NO, in particular in NO control of AMPAR trafficking in LTP, have been wanting. S-nitrosylation of NSF enhances NSF binding to GluR2 and regulates GluR2 surface expression (Huang, 2005). Also, activation of the NO-cGMP-cGKI pathway increases both GluR1 and synaptophysin puncta and the phosphorylation of VASP in hippocampal neurons (Wang, 2005). However, as yet, a specific pathway for NO control of activity-dependent GluR1 trafficking to synapses, an essential component of LTP, has not been reported. The interaction of cGKII with GluR1 reported here, and its consequent effect on GluR1 surface levels, directly link the actions of NO to LTP via GluR1 trafficking (Serulle, 2007).

A physical association of cGKII with GluR1 enables the kinase to phosphorylate GluR1 at S845. This phosphorylation is required for cGMP-dependent GluR1 surface accumulation, since block of the phosphorylation by the S845A GluR1 mutation blocked the surface increase. Phosphorylation of S845 accompanies increases in GluR1 surface levels and is necessary for GluR1 synaptic insertion during LTP. S845 is dephosphorylated during hippocampal LTD, and S845 phosphorylation on its own is sufficient for increase of GluR1 in the extrasynaptic plasma membrane. Thus far only PKA phosphorylation of S845 has been considered, perhaps because it was the initial kinase shown to phosphorylate this site. The present study demonstrates that cGKII activity also phosphorylates S845 (Serulle, 2007).

Increases of surface GluR1 following both PKA and cGKII phosphorylation are restricted to extrasynaptic sites , and AMPAR synaptic incorporation requires at least one additional step, possibly mediated by S818 phosphorylation. Interestingly, although 8-Br-cGMP on its own did not enhance hippocampal synaptic responses, when paired with a weak tetanus that by itself does not enhance responses, 8-Br-cGMP produced an immediate potentiation. This suggests that cGMP can prime the system for potentiation by a weak tetanic stimulation, possibly by increasing the extrasynaptic surface AMPAR population (Serulle, 2007).

The NMDAR and nNOS mutually interact with PSD-95, and Ca2+ fluxes through the NMDAR activate nNOS in this complex to produce NO, which induces sGC to produce cGMP, which activates cGKII. Ca2+ fluxes also stimulate Ca2+-regulated adenylate cyclases, which produce cAMP, which activates PKA, which also phosphorylates S845. PKA binds the A kinase anchoring protein, AKAP79, which in turn binds the PDZ domain scaffolding protein, SAP97, which binds the GluR1 CTD, thus targeting PKA to the GluR1 CTD and facilitating phosphorylation of S845 (Serulle, 2007).

Unlike the SAP97-AKAP-PKA pathway, the NO-cGMP-cGKII pathway does not rely on a scaffold since the kinase binds the receptor directly. Interestingly, a knockin mouse expressing GluR1 that lacks the last 7 aa of its CTD and does not bind SAP97 exhibited normal hippocampal LTP and GluR1 trafficking. This is explained if the NO-cGMP-cGKII pathway phosphorylates S845 in this mutant (Serulle, 2007).

GluR1 interacts with cGKII via auxiliary and core contact CTD sequences that flank S845. Interestingly, a CTD contact sequence resembles an AI domain sequence of cGKII, suggesting that to bind the catalytic domain, GluR1 mimics the AI domain. Also, this receptor-kinase interaction resembles the well-studied CaMKII binding to the NR2B (Serulle, 2007).

In the absence of cGMP, cGKII is inactive. Following NMDAR stimulation, binding of cGMP to cGKII induces a cGKII conformational change that causes AI domain autophosphorylation, AI domain release from the catalytic domain, and elongation of the kinase. The GluR1 CTD binds the newly exposed cGKII catalytic domain, facilitating GluR1 phosphorylation and the increase of surface GluR1. In one model for this increase, S845 phosphorylation promotes GluR1 trafficking to the plasma membrane, perhaps by releasing of GluR1 from a cytosolic retention factor. Alternatively, GluR1 may cycle into and out of the plasma membrane constitutively, and S845 phosphorylation may stabilize the receptor at the neuron surface. With either model, S845 phosphorylation would regulate the size of an extrasynaptic pool from which receptors may be inserted into the synapse during LTP. Such transport may depend on additional GluR1 phosphorylation. Because a highly selective peptide block of cGKII strongly reduces LTP, such an increase in an extrasynaptic receptor pool is likely to be a requirement for the synaptic potentiation associated with LTP. The present work demonstrates that the NMDAR can control the size of such a receptor pool, acting through nNOS, NO, and cGMP production and the activation of cGKII (Serulle, 2007).

Signaling downstream of glutamate receptors

Excitatory synaptic transmission in the central nervous system is mediated primarily by the release of glutamate from presynaptic terminals onto postsynaptic channels gated by N-methyl-D-aspartate (NMDA) and AMPA receptors. The myriad intracellular responses arising from the activation of the NMDA and AMPA receptors have previously been attributed to the flow of Ca2+ and/or Na+ through these ion channels. Binding of the agonist AMPA to its receptor can generate intracellular signals that are independent of Ca2+ and Na+ in rat cortical neurons. In the absence of intracellular Ca2+ and Na+, AMPA, but not NMDA, brings about changes in a guanine-nucleotide-binding protein (Galpha[il]) that inhibit pertussis toxin-mediated ADP-ribosylation of the protein in an in vitro assay. This effect is observed in intact neurons treated with AMPA as well as in isolated membranes exposed to AMPA, and is also found in MIN6 cells, which express functional AMPA receptors but have no metabotropic glutamate receptors. AMPA also inhibits forskolin-stimulated activity of adenylate cyclase in neurons, demonstrating that Gi proteins are activated. Moreover, both Gbetagamma blockage and co-precipitation experiments demonstrate that the modulation of the Gi protein arises from the association of Galpha(il) with the glutamate receptor-1 (GluR1) subunit. These results suggest that, as well as acting as an ion channel, the AMPA receptor can exhibit metabotropic activity (Wang, 1997).

The AMPA receptor, ubiquitous in brain, is termed 'ionotropic' because it gates an ion channel directly. An AMPA receptor can also modulate a G-protein to gate an ion channel indirectly. Glutamate applied to a retinal ganglion cell briefly suppresses the inward current through a cGMP-gated channel. AMPA and kainate also suppress the current, an effect that is blocked both by their general antagonist CNQX and also by the relatively specific AMPA receptor antagonist GYKI-52466. Neither NMDA nor agonists of metabotropic glutamate receptors are effective. The AMPA-induced suppression of the cGMP-gated current is blocked when the patch pipette includes GDP-beta-S, whereas the suppression is irreversible when the pipette contains GTP-gamma-S. This suggests a G-protein mediator, and, consistent with this, pertussis toxin blocks the current suppression. Nitric oxide (NO) donors induce the current suppressed by AMPA, and phosphodiesterase inhibitors prevent the suppression. Apparently, the AMPA receptor can exhibit a 'metabotropic' activity that allows it to antagonize excitation evoked by NO (Kawai, 1999).

Glutamate is the major excitatory neurotransmitter in the mammalian central nervous system. The ionotropic glutamate receptors are classified into two groups: NMDA (N-methyl-D-aspartate) receptors and AMPA (alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate) receptors. The AMPA receptor is a ligand-gated cation channel that mediates the fast component of excitatory postsynaptic currents in the central nervous system. AMPA receptors function not only as ion channels but also as cell-surface signal transducers by means of their interaction with the Src-family non-receptor protein tyrosine kinase Lyn. In the cerebellum, Lyn is physically associated with the AMPA receptor and is rapidly activated following stimulation of the receptor. Activation of Lyn is independent of Ca2+ and Na+ influx through AMPA receptors. As a result of Lyn activation, the mitogen-activated protein kinase (MAPK) signaling pathway is activated, and the expression of brain-derived neurotrophic factor (BDNF) messenger RNA is increased in a Lyn-kinase-dependent manner. Thus, AMPA receptors generate intracellular signals from the cell surface to the nucleus through the Lyn-MAPK pathway, which may contribute to synaptic plasticity by regulating the expression of BDNF (Hayashi, 1999).

Ca2+-permeable AMPA receptors may play a key role during developmental neuroplasticity, learning and memory, and neuronal loss in a number of neuropathologies. However, the intracellular signaling pathways used by AMPA receptors during such processes are not fully understood. The mitogen-activated protein kinase (MAPK) cascade is an attractive target because it has been shown to be involved in gene expression, synaptic plasticity, and neuronal stress. Using primary cultures of mouse striatal neurons and a phosphospecific MAPK antibody, the ability of AMPA receptors to activate the MAPK cascade was addressed. In the presence of cyclothiazide, AMPA causes a robust and direct (no involvement of NMDA receptors or L-type voltage-sensitive Ca2+ channels) Ca2+-dependent activation of MAPK through MAPK kinase (MEK). This activation is blocked by GYKI 53655, a noncompetitive selective antagonist of AMPA receptors. Probing the mechanism of this activation reveals an essential role for phosphatidylinositol 3-kinase (PI 3-kinase) and the involvement of a pertussis toxin (PTX)-sensitive G-protein, a Src family protein tyrosine kinase, and Ca2+/calmodulin-dependent kinase II. Application of AMPA to rat cerebral cortical neurons has been shown to lead to a rapid increase in Ras activity and activation of MAPK. Ras-dependent activation of MAPK is usually associated with seven transmembrane receptors that couple to heterotrimeric G-proteins. AMPA activates ERK2 (p42) by causing a Ca2+-dependent association of G-protein betagamma subunits, probably Gi, with a Ras, Raf kinase, MEK complex. This novel involvement of a heterotrimeric G-protein in ionotropic AMPA receptor signaling was examined. Striatal neurons were pretreated with pertussis toxin (PTX) or PBS vehicle for 24 hr before experiments with AMPA/cyclothiazide. PTX treatment abolishes AMPA receptor activation of MAPK, indicating a role for a Gi or Go-type G-protein in the activation of MAPK by AMPA receptors in striatal neurons. Similarly, kainate activates MAPK in a PI 3-kinase-dependent manner. AMPA receptor stimulation leads to a Ca2+-dependent phosphorylation of the nuclear transcription factor CREB, which can be prevented by inhibitors of MEK or PI 3-kinase. These results indicate that Ca2+-permeable AMPA receptors transduce signals from the cell surface to the nucleus of neurons through a PI 3-kinase-dependent activation of MAPK. This novel pathway may play a pivotal role in regulating synaptic plasticity in the striatum (Perkinton, 1999).

Thus, although the specific protein-protein interactions that lead to activation of the Ras-MAPK pathway by AMPA receptors are not currently known, it seems reasonable to propose that AMPA receptor-evoked rises in cytosolic Ca2+ may trigger activation of PI 3-kinase: then, recruitment of the lipid kinase to the MAPK cascade may, as is the case with seven-transmembrane Gi/Go-type G-protein linked receptors, be orchestrated by free G betagamma subunits. The specific exchange factors regulating Ras activity after AMPA receptor stimulation also remain to be determined. An involvement of the neuron-specific guanine nucleotide exchange factor, Ras-GRF, seems plausible because it has recently been demonstrated that Ras-GRF can be activated in response to increases in intracellular Ca2+ and/or free G-protein betagamma subunits that induce phosphorylation of Ras-GRF by as yet unknown kinases. However, Ca2+/calmodulin-dependent activation of Ras-GRF does not appear to involve PTKs, thus, the results indicating that tyrosine phosphorylation may be an important step in AMPA receptor activation of MAP kinase suggests that additional Ca2+-dependent routes to Ras may be activated. It has been shown that CaM-KII can phosphorylate AMPA receptor subunits (Mammen et al., 1997), resulting in enhanced receptor currents, and this has been implicated in the strengthening of postsynaptic responses associated with synaptic plasticity. Selective inhibition of CaM-KII activity substantially reduces AMPA/cyclothiazide-evoked activation of MAPK without altering Ca2+ influx through the receptor. These data indicate that CaM-KII can be a positive modulator of AMPA receptor signaling but that in the presence of cyclothiazide the kinase probably regulates AMPA receptor-mediated MAPK activation at a point downstream of Ca2+ entry (Perkinton, 1999 and references).

alpha and ßCaMKII are inversely regulated by activity in hippocampal neurons in culture: the alpha/ß ratio shifts toward alpha during increased activity and ß during decreased activity. The swing in ratio is ~5-fold and may help tune the CaMKII holoenzyme to changing intensities of Ca2+ signaling. The regulation of CaMKII levels uses distinguishable pathways, one responsive to NMDA receptor blockade that controls alphaCaMKII alone, the other responsive to AMPA receptor blockade and involving ßCaMKII and possibly further downstream effects of ßCaMKII on alphaCaMKII. Overexpression of alphaCaMKII or ßCaMKII results in opposing effects on unitary synaptic strength as well as mEPSC frequency that can account in part for activity-dependent effects observed with chronic blockade of AMPA receptors. Regulation of CaMKII subunit composition may be important for both activity-dependent synaptic homeostasis and plasticity (Thiagarajan, 2002).

Calcium/calmodulin-dependent protein kinase II (CaMKII) is expressed at high levels in the central nervous system, particularly in the hippocampus, where it constitutes ~2% of total protein. As a holoenzyme, neuronal CaMKII is made up of 6-12 subunits, primarily the 52 kDa alpha isoform and the 60 kDa ß isoform. The subunits of the holoenzyme are held together by association domains in their C-terminals, which form a central globular structure from which the N-terminals extend radially. The N-terminal contains the catalytic sites of the kinase as well as the autoinhibitory domains that bind to the catalytic sites in the basal state. The binding of Ca2+/calmodulin releases this autoinhibition, allowing phosphorylation to take place at a critical threonine residue, Thr286 in alpha and Thr287 in ß. This autophosphorylation allows the molecular memory of a transient Ca2+ signal to greatly outlast the duration of the Ca2+ transient itself, a property that endows CaMKII with the ability to decode Ca2+ signals in a frequency-dependent manner (Thiagarajan, 2002 and references therein).

Immunoprecipitation with subunit-specific antibodies indicates that the majority of the CaMKII holoenzymes are alpha/ß heteromers with variable subunit ratios, although some alpha homomers can also be found. Why have two isoforms? One significant distinction between the alpha and ß isoforms lies in their sharply different affinity for calmodulin. Half-maximal autophosphorylation is achieved at 130 nM calmodulin for alphaCaMKII and at 15 nM calmodulin for ßCaMKII. Due to this difference, the two isoforms have different sensitivities to Ca2+ signals under nonsaturating levels of calmodulin. alphaCaMKII is selective for higher levels of Ca2+ signals, while ßCaMKII has better sensitivity to lower levels of signal. When the two isoforms are combined in a heteromer, the response to Ca2+ signals has been found to depend on the ratio of alpha to ß subunits. Consequently, activity-dependent regulation of alpha- and ßCaMKII expression could provide a mechanism of tuning neuronal responses to different levels of activity. This is an intriguing possibility that raises several fundamental questions. Does the cell regulate the ratio of alpha to ß in an activity-dependent manner? And if so, what pathways of synaptic activity might control the regulation of alpha- and ßCaMKII? Could their regulation be coupled? What would be the consequence of such a regulation for synaptic transmission? This study uses a combination of immunotechniques and electrophysiology to address these issues. The data show that alpha- and ßCaMKII are inversely regulated by activity in a manner that may help tune CaMKII to changing levels of Ca2+ signal. Furthermore, tilting the ratio toward alpha or ß results in opposing effects on unitary synaptic strength and mEPSC frequency and has functional significance for both activity-dependent plasticity and homeostasis (Thiagarajan, 2002).

In excitatory neurons, treatment with TTX to block action potentials decreases the levels of alphaCaMKII and raises ßCaMKII, both in cell bodies and at synapses. Conversely, exposure to bicucculine to prevent inhibitory transmission and increase firing increases alphaCaMKII and decreased ßCaMKII. Thus, the changes in alphaCaMKII correlated positively with changes in electrical activity, while changes in ßCaMKII correlated negatively. This inverse regulation gives rise to ~5-fold changes in the alpha:ß ratio between the extremes of TTX or BIC treatment, while the sum total of these isoforms remained relatively unchanged, varying only 1.0- to 1.3-fold (for assumed values of the basal alpha/ß ratio ranging between 1:1 and 3:1). One may speculate that the inverse changes in isoform levels support the widest variation in isoform ratio consistent with holding fixed the total amount of enzyme (possibly important for structural reasons). That inhibitory neurons, which lack immunoreactivity for alphaCaMKII, fail to show regulation of ß with altered activity also fits with a pattern in which the overall level of CaMKII is tightly regulated (Thiagarajan, 2002).

Changes in the balance between alpha and ß isoforms predicts interesting functional consequences. ßCaMKII has been shown to exhibit a much higher affinity than alphaCaMKII for Ca2+/CaM. If the changes in the overall alpha:ß ratio between the opposite conditions of TTX and BIC are indicative of the subunit composition of the CaMKII holoenzyme, this would result in a much higher affinity of the holoenzyme for Ca2+/CaM. Assuming a ~9-fold difference in CaM affinity of ßCaMKII relative to alphaCaMKII, a rough calculation predicts a ~2-fold variation in the holoenzyme affinity for Ca2+/CaM. The increased sensitivity of the holoenzyme to CaM with decreased activity would serve as a homeostatic mechanism to confer responsiveness to weaker Ca2+ signals (Thiagarajan, 2002).

It first came as a surprise that activity-dependent changes in alpha and ß isoforms arose from largely distinct pathways, involving different glutamate receptors: levels of alphaCaMKII (but not ß) are strongly influenced by NMDAR activity; in contrast, ßCaMKII was strongly affected by AMPAR activity. These new findings make sense if put in context of previous studies on neuronal CaMKII. The strong reduction of alphaCaMKII but sparing of ß by blockade of NMDA receptors can be interpreted in light of several related observations: (1) multimeric CaMKII takes the form of alpha homomers as well as alpha/ß heteromultimers; (2) NMDAR-dependent changes in the abundance of alphaCaMKII can be detected as soon as 5 min after stimulation in hippocampal slices, consistent with a localized dendritic translation of alphaCamKII mRNA; (3) ßCaMKII mRNA is absent in dendrites, leaving alpha homomers as the only enzyme species that could be formed there. Taken together with these observations, these findings are consistent with a simple model wherein Ca2+ entry through NMDARs in the dendrites regulates alphaCaMKII homomers locally, on a fast time scale, with little or no control of ß (Thiagarajan, 2002).

The observation that levels of ßCaMKII strongly increased in response to blockade of AMPA receptors, not NMDARs, suggested that regulation of ß may be quite different than proposed for alpha alone. Under various pharmacological conditions, the pattern of changes observed in ßCaMKII was always consistent with an inverse relationship with AMPAR activity. Because ß transcripts are restricted to the cell body, and changes in ßCaMKII occur only slowly (evident only on time scales >1 hr), the regulation by AMPARs is likely to occur at the level of nuclear transcription. Thus, AMPAR-mediated depolarization could work through recruitment of voltage-gated Ca2+ channels and regulation of nuclear transcription factors. Regulation in or near the nucleus makes additional sense for the linkage between increased ßCaMKII and the downregulation of alpha that was observed in transfection studies. All considerations seem consistent with the following working hypothesis: ßCaMKII levels are regulated in the cell body, downstream of AMPA receptor activity, leading to a reciprocal regulation of alpha and thus the formation of alpha/ß heteromers of variable subunit ratio (Thiagarajan, 2002).

This scheme invokes the inter-relationship between levels of ß and alpha that was directly observed in the transfection experiments. Another possibility, not mutually exclusive, is that AMPAR block decreases postsynaptic depolarization and thereby reduces Ca2+ entry through NMDAR, leading to a fall in the alpha isoform (Thiagarajan, 2002).

Conceptual distinctions have been drawn between synaptic homeostasis, negative feedback regulation thought of as neuron wide, and forms of synaptic plasticity such as LTP, which can be self-reinforcing and synapse specific. Both kinds of regulation may be strongly impacted by inverse changes in the abundance of alpha- and ßCaMKII. For example, increased activity, and consequent elevation of the alpha/ß ratio would decrease the Ca2+/CaM sensitivity of CaMKII in a homeostatic manner. This can be viewed as 'input tuning', wherein the holoenzyme is adjusted appropriately to the ambient level of activity. The threshold for the induction of LTP, which already is high, would be further raised, thereby changing the rules governing synaptic plasticity . However, increasing the alpha/ß ratio may also change the cellular localization of CaMKII, promoting alphaCaMKII expression at specific subsynaptic sites where it could contribute to LTP. Understanding the full implications for plasticity and metaplasticity will become easier once more is known about how the subunit composition of CaMKII affects its cellular localization and degree of autophosphorylation and how alterations in the alpha/ß ratio and its downstream effects unfold over time scales ranging from minutes to days (Thiagarajan, 2002).

Altering the levels of alpha and ß causes striking changes in both mini size and frequency. Once again, the overall change in synaptic function cannot be neatly pigeonholed into strict categories of 'synaptic homeostasis' or 'synaptic plasticity' alone. Increases in the alpha/ß ratio accentuates the contribution of individual synaptic events (augmented mini area), while also tending to decrease the number of quanta received per unit time (lowered mini frequency), thus keeping the total synaptic drive within reasonable bounds. Evidently, simple biochemical changes can induce a powerful combination of self-reinforcing local changes, but negative feedback regulation over the neuron as a whole (Thiagarajan, 2002).

NMDA-type glutamate receptors play a critical role in the activity-dependent development and structural remodeling of dendritic arbors and spines. However, the molecular mechanisms that link NMDA receptor activation to changes in dendritic morphology remain unclear. The Rac1-GEF Tiam1 is present in dendrites and spines and is required for their development. Tiam1 interacts with the NMDA receptor and is phosphorylated in a calcium-dependent manner in response to NMDA receptor stimulation. Blockade of Tiam1 function with either RNAi or dominant interfering mutants of Tiam1 suggests that Tiam1 mediates effects of the NMDA receptor on dendritic development by inducing Rac1-dependent actin remodeling and protein synthesis. Taken together, these findings define a molecular mechanism by which NMDA receptor signaling controls the growth and morphology of dendritic arbors and spines (Tolias, 2005).

Glutamate receptors and adult neural stem cells

A wide variety of in vivo manipulations influence neurogenesis in the adult hippocampus. It is not known, however, if adult neural stem/progenitor cells (NPCs) can intrinsically sense excitatory neural activity and thereby implement a direct coupling between excitation and neurogenesis. Moreover, the theoretical significance of activity-dependent neurogenesis in hippocampal-type memory processing networks has not been explored. This study demonstrates that excitatory stimuli act directly on adult hippocampal NPCs to favor neuron production. The excitation is sensed via Cav1.2/1.3 (L-type) Ca2+ channels and NMDA receptors on the proliferating precursors. Excitation through this pathway acts to inhibit expression of the glial fate genes Hes1 and Id2 and increase expression of NeuroD, a positive regulator of neuronal differentiation. These activity-sensing properties of the adult NPCs, when applied as an 'excitation-neurogenesis coupling rule' within a Hebbian neural network, predict significant advantages for both the temporary storage and the clearance of memories (Deisseroth, 2004).

Using an array of approaches, the coupling of excitation to neurogenesis in proliferating adult-derived NPCs was studied both in vitro and in vivo. Adult neurogenesis is potently enhanced by excitatory stimuli and involves Cav1.2/1.3 channels and NMDA receptors. These Ca2+ influx pathways are located on the proliferating NPCs, allowing them to directly sense and process excitatory stimuli. No effect of excitation was found on the extent of differentiation in individual cells (measured by extent of MAP2ab expression in the NPC-derived neurons) nor were effects observed on proliferative rate or fraction, survival, or apoptosis. Instead, excitation increased the fraction of NPC progeny that were neurons, both in vitro and in vivo, and total neuron number was increased as well. The Ca2+ signal in NPCs leads to rapid induction of a proneural gene expression pattern involving the bHLH genes HES1, Id2, and NeuroD, and the resulting cells become fully functional neurons defined by neuronal morphology, expression of neuronal structural proteins (MAP2ab and Doublecortin), expression of neuronal TTX-sensitive voltage-gated Na+ channels, and synaptic incorporation into active neural circuits. A monotonically increasing function characterizes excitation-neurogenesis coupling, and incorporation of this relationship into a layered Hebbian neural network suggests surprising advantages for both the clearance of old memories and the storage of new memories. Taken together, these results provide a new experimental and theoretical framework for further investigation of adult excitation-neurogenesis coupling (Deisseroth, 2004).

In the hippocampal formation, neural stem cells exist either within the adjacent ventricular zone or within the subgranular zone proper at the margin between the granule cell layer and the hilus, where proliferative activity is most robust. These cells do not express neuronal markers but proliferate and produce dividing progeny that incrementally commit to differentiated fates (such as the neuronal lineage) over successive cell divisions. Native NPC populations in vivo are therefore heterogenous with regard to lineage potential, and markers are not available that distinguish between the multipotent stem cell and the subtly committed yet proliferative progenitor cell. Excitation may therefore act on either or both types of proliferating precursor, in vitro and in vivo. The functional consequences of coupling excitation to insertion of new neurons for the neural network, however, is independent of which precursor cell types respond to excitation (Deisseroth, 2004).

The enhancement of hippocampal neurogenesis by behavioral stimuli such as environmental enrichment and running may, at least in part, be implemented at the molecular level by excitation-neurogenesis coupling. Notably, running and environmental enrichment increase adult neurogenesis in the hippocampus but not in the subventricular zone. Of course, not every neurogenic region in the brain need follow the excitation-neurogenesis coupling rule outlined here. An activity rule appropriate for the unique information processing or storage function of that brain region might be expected to operate. In this context, it is interesting to note that, while subventricular zone/olfactory bulb precursor neurogenesis is not enhanced by behavioral activity, proliferation and survival in this system can be influenced by olfactory sensory stimuli. This suggests that a different form of activity rule, appropriate for that local circuit, may govern olfactory bulb neurogenesis (Deisseroth, 2004).

AMPA receptors: activity modification of neural structure

The influence of synaptically released glutamate on postsynaptic structure was investigated by comparing the effects of deafferentation, receptor antagonists and blockers of glutamate release in hippocampal slice cultures. Spine density and length of CA1 pyramidal cells decrease after transection of Schaffer collaterals and after application of AMPA receptor antagonists or botulinum toxin to unlesioned cultures. Loss of spines induced by lesion or by botulinum toxin is prevented by simultaneous AMPA application. Tetrodotoxin does not affect spine density. Synaptically released glutamate thus exerts a trophic effect on spines by acting at AMPA receptors. It is concluded that AMPA receptor activation by spontaneous vesicular glutamate release is sufficient to maintain dendritic spines (McKenney, 1999).

The classically conditioned vertebrate eye-blink response is a model in which to study neuronal mechanisms of learning and memory. In this paradigm, an eye-blink reflex in response to a tone can be evoked when the tone is repeatedly paired with an air puff to the cornea that normally elicits the blink response. A neural correlate of this response recorded in the abducens nerve can be conditioned entirely in vitro using an isolated brainstem-cerebellum preparation from the turtle by pairing trigeminal and auditory nerve stimulation. Conditioning requires that the paired stimuli occur within a narrow temporal window of <100 msec. Conditioning is blocked by a NMDA receptor antagonist. Moreover, there is a significant positive correlation between the levels of conditioning and greater immunoreactivity with the glutamate receptor 4 (GluR4) AMPA receptor subunit in the abducens motor nuclei, but not with NMDAR1 or GluR1. It is concluded that in vitro classical conditioning of an abducens nerve eye-blink response is generated by NMDA receptor-mediated mechanisms that may act to modify the AMPA receptor by increasing GluR4 subunits in auditory nerve synapses (Keifer, 2001).

Both theoretical and experimental work have suggested that central neurons compensate for changes in excitatory synaptic input in order to maintain a relatively constant output. Inhibition of excitatory synaptic transmission in cultured spinal neurons leads to an increase in mEPSC amplitudes, accompanied by an equivalent increase in the accumulation of AMPA receptors at synapses. Conversely, increasing excitatory synaptic activity leads to a decrease in synaptic AMPA receptors and a decline in mEPSC amplitude. The time course of this synaptic remodeling is slow, similar to the metabolic half-life of neuronal AMPA receptors. Moreover, inhibiting excitatory synaptic transmission significantly prolongs the half-life of the AMPA receptor subunit GluR1, suggesting that synaptic activity modulates the size of the mEPSC by regulating the turnover of postsynaptic AMPA receptors (O'Brien, 1998).

Dynamic regulation of AMPA-type glutamate receptors represents a primary mechanism for controlling synaptic strength, though mechanisms for this process are poorly understood. The palmitoylated postsynaptic density protein, PSD-95, regulates synaptic plasticity and associates with the AMPA receptor trafficking protein, stargazin. This study identifies palmitate cycling on PSD-95 at the synapse; palmitate turnover on PSD-95 is regulated by glutamate receptor activity. Acutely blocking palmitoylation disperses synaptic clusters of PSD-95 and causes a selective loss of synaptic AMPA receptors. Rapid glutamate-mediated AMPA receptor internalization requires depalmitoylation of PSD-95. In a nonneuronal model system, clustering of PSD-95, stargazin, and AMPA receptors is also regulated by ongoing palmitoylation of PSD-95 at the plasma membrane. These studies suggest that palmitate cycling on PSD-95 can regulate synaptic strength and regulates aspects of activity-dependent plasticity (El-Husseini, 2002).

AMPA receptors: long-term potentiation and depression

The ability of central glutamatergic synapses to change their strength in response to the intensity of synaptic input, which occurs, for example, in long-term potentiation (LTP), is thought to provide a cellular basis for memory formation and learning. LTP in the CA1 field of the hippocampus requires activation of Ca2+/calmodulin-kinase II (CaM-KII), which phosphorylates Ser-831 in the GluR1 subunit of the AMPA glutamate receptor (AMPA-R), and this activation/phosphorylation is thought to be a postsynaptic mechanism in LTP. In this study, a molecular mechanism has been identified by which CaM-KII potentiates AMPA-Rs. Coexpression in HEK-293 cells of activated CaM-KII with GluR1 does not affect the glutamate affinity of the receptor, the kinetics of desensitization and recovery, channel rectification, open probability, or gating. Single-channel recordings identify multiple conductance states for GluR1, and coexpression with CaM-KII or a mutation of Ser-831 to Asp increases the contribution of the higher conductance states. These results indicate that CaM-KII can mediate plasticity at glutamatergic synapses by increasing single-channel conductance of existing functional AMPA-Rs or by recruiting new high-conductance-state AMPA-Rs (Derkach, 1999).

The mechanisms responsible for enhanced transmission during long-term potentiation (LTP) at CA1 hippocampal synapses remain elusive. Several popular models for LTP expression propose an increase in 'use' of existing synaptic elements, such as increased probability of transmitter release or increased opening of postsynaptic receptors. To test these models directly, a GluR2 knockout mouse was studied in which AMPA receptor transmission was rendered sensitive to a use-dependent block by polyamine compounds. This method can detect increases during manipulations affecting transmitter release or AMPA receptor channel open time and probability, however, no such changes are seen to occur during LTP. These results indicate that the recruitment of new AMPA receptors and/or an increase in the conductance of these receptors is responsible for the expression of CA1 LTP (Mainen, 1998).

AMPA receptors (AMPARs) are not thought to be involved in the induction of long-term potentiation (LTP), but may be involved in its expression via second messenger pathways. However, one subunit of the AMPARs, GluR2, is also known to control Ca2+ influx. To test whether GluR2 plays any role in the induction of LTP, mice were generated that lack this subunit. In GluR2 mutants, LTP in the CA1 region of hippocampal slices is markedly enhanced (2-fold) and nonsaturating, whereas neuronal excitability and paired-pulse facilitation are normal. The 9-fold increase in Ca2+ permeability, in response to kainate application, suggests one possible mechanism for enhanced LTP. Mutant mice exhibit increased mortality, and those surviving show reduced exploration and impaired motor coordination. These results suggest an important role for GluR2 in regulating synaptic plasticity and behavior (Jia, 1996).

Gene-targeted mice lacking the AMPA receptor subunit GluR-A exhibit normal development, life expectancy, and fine structure of neuronal dendrites and synapses. In hippocampal CA1 pyramidal neurons, GluR-A-/- mice show a reduction in function of AMPA receptors, with the remaining receptors preferentially targeted to synapses. Immunocytochemistry in the hippocampus of GluR-A-/- mice relative to wild type reveals a redistribution of the GluR-B subunit in pyramidal and dentate granule cells with increased staining over the somata (stratum pyramidale) and decreased staining in the basal (stratum oriens) and apical (stratum radiatum) dendrites. The altered GluR-B localization upon lack of GluR-A may indicate that the edited GluR-B subunit requires a partner for assembly or dendritic targeting of GluR-B-containing AMPA receptors. A substantial amount of GluR-B remains in the stratum lacunosum-moleculare (possibly in the form of GluR-B/C heteromeric channels) in synapses at the distalmost part of the apical dendrites of CA1 and CA3 pyramidal neurons. In GluR-A-/- mice, the CA1 soma-patch currents are strongly reduced, but glutamatergic synaptic currents are unaltered; and evoked dendritic and spinous Ca2+ transients, Ca2+-dependent gene activation, and hippocampal field potentials are as in the wild type. In adult GluR-A-/- mice, associative long-term potentiation (LTP) is absent in CA3 to CA1 synapses, but spatial learning in the water maze is not impaired. The results suggest that CA1 hippocampal LTP is controlled by the number or subunit composition of AMPA receptors and show a dichotomy between LTP in CA1 and acquisition of spatial memory (Zamanillo, 1999).

This finding adds to a growing number of examples of a dichotomy between LTP and learning. One explanation could be that mice may use extrahippocampal structures to solve the Morris water maze. It is possible that LTP, although not critical for the type of reference memory test used to solve the Morris water maze, could be important in spatial tasks that involve only episodic or working memory. Alternatively, learning may be associated with LTP at a degree of synaptic involvement that is too small to be detected with conventional electrophysiological field recordings. Furthermore, the spatial task might not require LTP in Schaffer collateral-CA1 synapses. Notably, a genetically engineered CA1 NMDA receptor deficiency also generates LTP deficiency, but learning in the water maze is impaired. One explanation for the difference in learning might be that the induction phase of LTP is impaired in the NMDA receptor but not in the AMPA receptor mutant. During this phase the spinous Ca2+ transients affect numerous signaling pathways, which might be essential for memory acquisition. In this context, it might be asked how the fEPSP-LTP phenomenon is related to normal physiology in the hippocampus. The highly synchronous ensemble activity of CA3 pyramidal neurons required to induce standard fEPSP-LTP may not normally occur. In summary, adult hippocampal LTP depends on the number and subunit composition of AMPA receptors. Therefore, in adult animals, LTP appears to be essentially a postsynaptic mechanism. However, this particular form of synapse modifiability in CA1 is not required for a reference memory test (Zamanillo, 1999 and references).

To monitor changes in AMPA receptor distribution in living neurons, the AMPA receptor subunit GluR1 was tagged with green fluorescent protein (GFP). This protein (GluR1-GFP) is functional and is transiently expressed in hippocampal CA1 neurons. In dendrites visualized with two-photon laser scanning microscopy or electron microscopy, most of the GluR1-GFP is intracellular, mimicking endogenous GluR1 distribution. Tetanic synaptic stimulation induces a rapid delivery of tagged receptors into dendritic spines as well as clusters in dendrites. These postsynaptic trafficking events require synaptic N-methyl-D-aspartate (NMDA) receptor activation and may contribute to the enhanced AMPA receptor-mediated transmission observed during long-term potentiation and activity-dependent synaptic maturation (Shi, 1999).

These results build on a number of studies suggesting that the delivery of AMPA receptors to synapses contributes to activity-dependent plasticity. Inhibition of membrane fusion processes in the postsynaptic cell blocks the action of LTP. Furthermore, the COOH-termini of AMPA receptor subunits GluR2 and GluR4c bind N-ethylmaleimide-sensitive fusion protein, a protein involved in membrane fusion processes. Vesicular organelles, possibly undergoing exocytosis and endocytosis, have been detected with electron microcopy in spines. And last, dendrites can display a calcium-evoked exocytosis of trans-Golgi-derived organelles that is mediated by the calcium/calmodulin-dependent protein kinase II, an enzyme thought to mediate LTP. Other postsynaptic mechanisms, such as an increase in conductance of AMPA receptors, may also occur in parallel. These results also do not rule out a contribution by presynaptic modifications. In addition to the spine delivery of GluR1-GFP, tetanic stimulation induces the formation of clusters of the tagged receptor within dendrites. These structures may be related to the spine apparatus, membranous structures at the base of spines that appear to contain AMPA receptors. The entry of calcium through synaptic NMDA receptors may cause nucleation of AMPA receptor-containing membranes close to active synapses. Once formed, such sites may serve several functions. These sites may replenish those receptors delivered to spines during plasticity. Additionally, they may serve as a 'synaptic tag', providing a docking site for AMPA receptors synthesized at distant sites. Last, they could provide a site for local AMPA receptor synthesis. In these capacities, such clusters could represent a structural modification serving as a long-lasting memory mechanism (Shi, 1999 and references).

The second messenger pathways linking receptor activation at the membrane to changes in the nucleus are just beginning to be unraveled in neurons. The work presented here attempts to identify in striatal neurons the pathways that mediate cAMP response element-binding protein (CREB) phosphorylation and gene expression in response to NMDA receptor activation. The phosphorylation of the transcription factor CREB, the expression of the immediate early gene c-fos, and the induction of a transfected reporter gene under the transcriptional control of CREB after stimulation of ionotropic glutamate receptors were investigated. Neither AMPA/kainate receptors nor NMDA receptors are able to independently stimulate a second messenger pathway that leads to CREB phosphorylation or c-fos gene expression. Instead, a consecutive pathway from AMPA/kainate receptors to NMDA receptors and from NMDA receptors to L-type Ca2+ channels is seen. AMPA/kainate receptors are involved in relieving the Mg2+ block of NMDA receptors, and NMDA receptors trigger the opening of L-type Ca2+ channels. The second messenger pathway that activates CREB phosphorylation and c-fos gene expression is likely activated by Ca2+ entry through L-type Ca2+ channels. It is concluded that in primary striatal neurons glutamate-mediated signal transduction is dependent on functional L-type Ca2+ channels (Rajadhyaksha, 1999).

AMPA/kainate receptor channels open after interaction with glutamate and permit Na+ entry at the synapse. The resulting local depolarization removes the Mg2+ block of the NMDA receptor, which permits the NMDA receptor to respond to extracellular glutamate and glycine. Opening of the NMDA receptor channel causes Na+ and Ca2+ influx. Unlike the AMPA/kainate receptor channel that desensitizes rapidly, NMDA receptor channels have long opening times. Therefore, NMDA receptors can trigger the opening of L-type Ca2+ channels that open during strong depolarization. The activation of L-type Ca2+ channels promotes Ca2+ entry along the dendrites and at the cell body. Second messengers activated by Ca2+ translocate to the nucleus and phosphorylate CREB. The results presented in this paper suggest an important role for L-type Ca2+ channels in neuroplasticity of the striatum and confirm previous reports about the involvement of L-type Ca2+ channels in NMDA-mediated plasticity and toxicity. Under the experimental conditions described in this study, NMDA receptors initiate a signal transduction pathway but do not initiate a significant intraneuronal second messenger pathway, either alone or together with AMPA/kainate receptors. Depolarization of L-type Ca2+ channels plays a crucial role in the activation of an intraneuronal second messenger pathway (Rajadhyaksha, 1999).

Although the supportive role of AMPA/kainate receptors for NMDA receptors is in agreement with previous findings in hippocampal culture, other findings differ. In hippocampal cultures NMDA receptors and L-type Ca2+ channels seem to contribute to independent, parallel pathways rather than to the same pathway. Like in hippocampal cultures, L-type Ca2+ channels in the striatum activate the CRE and function independently of NMDA receptors. But although a direct pathway from NMDA receptors to the SRE in the striatum cannot be excluded, this pathway in itself is not enough to mediate c-fos gene expression. This difference may be attributed to intrinsic differences between both types of neurons or to the different neurotransmitters released in either culture. Hippocampal neurons are mostly glutamatergic and express very high levels of glutamate receptors. Striatal cultures are primarily GABAergic and express much lower levels of glutamate receptors. Because neurons in culture synapse onto each other, hippocampal neurons excite each other after activation, whereas GABA in striatal neurons, dependent on the level of maturity, may be excitatory or inhibitory. To avoid trans-synaptic effects in hippocampal cultures, Na+ channels are often blocked with TTX. Thus, there are fundamental differences in glutamate-mediated gene expression in neurons of both brain areas (Rajadhyaksha, 1999).

Redistribution of postsynaptic AMPA-subtype glutamate receptors may regulate synaptic strength at glutamatergic synapses, but the mediation of the redistribution is poorly understood. AMPA receptors undergo clathrin-dependent endocytosis, which is accelerated by insulin in a GluR2 subunit-dependent manner. Insulin-stimulated endocytosis rapidly decreases AMPA receptor numbers in the plasma membrane, resulting in long-term depression (LTD) of AMPA receptor-mediated synaptic transmission in hippocampal CA1 neurons. Moreover, insulin-induced LTD and low-frequency stimulation (LFS) induced homosynaptic CA1 LTD are found to be mutually occlusive and are both blocked by inhibiting postsynaptic clathrin-mediated endocytosis. Thus, controlling postsynaptic receptor numbers through endocytosis may be an important mechanism underlying synaptic plasticity in the mammalian CNS (Man, 2000).

Endocytosis of postsynaptic AMPA receptors may not be limited to homosynaptic CA1 LTD: insulin/IGF-I also produces a rapid and long-term depression of AMPA responses mediated by postsynaptic clathrin-dependent endocytosis in cultured cerebellar neurons. The insulin/IGF-I-induced depression of AMPA currents occludes cerebellar LTD, which in turn can be blocked by the inhibition of postsynaptic clathrin-dependent endocytosis. Additionally, a rapid, activity-dependent reduction of postsynaptic AMPA receptors takes place in a culture model of LTD induced by field stimulation. Taken together, these data suggest that rapid, clathrin-dependent removal of postsynaptic AMPA receptors may be a common final step in the expression of certain forms of LTD (Man, 2000 and references therein).

How is the clathrin-dependent endocytosis of AMPA receptors stimulated by LTD-inducing protocols? Since neurons contain and are able to release insulin in an activity-dependent manner, and since insulin receptors are concentrated in the postsynaptic density, one mechanism may involve the release of insulin presynaptically in response to LFS during LTD induction. Insulin may in turn activate its postsynaptic neuronal receptors to facilitate clathrin-dependent endocytosis of AMPA receptors. However, postsynaptic injection of the insulin receptor-neutralizing antibody, while blocking insulin-induced depression of AMPA EPSCs, has little effect on either hippocampal homosynaptic LTD or cerebellar LTD. These results suggest that insulin is not itself directly involved in mediating the expression of these forms of LTD but rather that insulin and LTD-inducing stimuli may converge to cause AMPA receptor endocytosis. It is likely that multiple signal transduction pathways exist for the regulation of AMPA receptor trafficking, and the elucidation of these pathways will provide further insight into the molecular mechanisms of synaptic plasticity and may ultimately provide mechanistic clues for the role of insulin in learning and memory (Man, 2000 and references therein).

Phosphorylation of the glutamate receptor is an important mechanism of synaptic plasticity. The C terminus of GluR2 of the AMPA receptor is phosphorylated by protein kinase C and serine-880 is the major phosphorylation site. This phosphorylation also occurs in human embryonic kidney (HEK) cells by addition of 12-O-tetradecanoylphorbol 13-acetate. Immunoprecipitation experiments reveal that the phosphorylation of serine-880 in GluR2 drastically reduces the affinity for glutamate receptor-interacting protein (GRIP), a synaptic PDZ domain-containing protein, in vitro and in HEK cells. This result suggests that modulation of serine-880 phosphorylation in GluR2 controls the clustering of AMPA receptors at excitatory synapses and consequently contributes to synaptic plasticity (Matsuda, 1999).

Cerebellar long-term depression (LTD) is thought to play an important role in certain types of motor learning. However, the molecular mechanisms underlying this event have not been clarified. Using cultured Purkinje cells, it has been shown that stimulations inducing cerebellar LTD cause phosphorylation of Ser880 in the intracellular C-terminal domain of the AMPA receptor subunit GluR2. This phosphorylation is accompanied by both a reduction in the affinity of GluR2 to glutamate receptor interacting protein (GRIP), a molecule known to be critical for AMPA receptor clustering, and a significant disruption of postsynaptic GluR2 clusters. Moreover, GluR2 protein released from GRIP is shown to be internalized. These results suggest that the dissociation of postsynaptic GluR2 clusters and subsequent internalization of the receptor protein, initiated by the phosphorylation of Ser880, are the mechanisms underlying the induction of cerebellar LTD (Matsuda, 2000).

Experience-dependent regulation of synaptic strength has been suggested as a physiological mechanism by which memory storage occurs in the brain. Although modifications in postsynaptic glutamate receptor levels have long been hypothesized to be a molecular basis for long-lasting regulation of synaptic strength, direct evidence obtained in the intact brain has been lacking. In the adult brain in vivo, synaptic glutamate receptor trafficking is bidirectionally, and reversibly, modified by NMDA receptor-dependent synaptic plasticity and changes in glutamate receptor protein levels accurately predict changes in synaptic strength. These findings support the idea that memories can be encoded by the precise experience-dependent assignment of glutamate receptors to synapses in the brain (Heynen, 2000).

LTP in vivo is associated with the delivery of glutamate receptor proteins to CA1 synapses, while LTD is associated with their removal. Like LTP and LTD, the changes in glutamate receptors depend on NMDAR activation during conditioning stimulation and are reversible. Although LTP results in an increase in synaptoneurosomal glutamate receptor protein levels, while LTD results in a decrease, these changes are not perfectly symmetric. LTP de novo correlates with an increase in AMPAR protein in CA1 synaptoneurosomes without a detectable change in NMDAR protein levels, while LTD de novo correlates with a decrease in both AMPAR and NMDAR protein in this biochemical fraction. These data demonstrate, for the first time, that a bidirectional redistribution of glutamate receptors accompanies bidirectional synaptic plasticity in the adult hippocampus in vivo (Heynen. 2000).

Activity-driven delivery of AMPA receptors is proposed to mediate glutamatergic synaptic plasticity, both during development and learning. In hippocampal CA1 principal neurons, such trafficking is primarily mediated by the abundant GluR-A subunit. A study of GluR-Blong, a C-terminal splice variant of the GluR-B subunit, is reported. GluR-Blong synaptic delivery is regulated by two forms of activity. Spontaneous synaptic activity-driven GluR-Blong transport maintains one-third of the steady-state AMPA receptor-mediated responses, while GluR-Blong delivery following the induction of LTP is responsible for approximately 50% of the resulting potentiation at the hippocampal CA3 to CA1 synapses at the time of GluR-Blong peak expression -- the second postnatal week. Trafficking of GluR-Blong-containing receptors thus mediates a GluR-A-independent form of glutamatergic synaptic plasticity in the juvenile hippocampus (Kolleker, 2003).

Synaptic plasticity involves protein phosphorylation cascades that alter the density of AMPA-type glutamate receptors at excitatory synapses; however, the crucial phosphorylated substrates remain uncertain. The AMPA receptor-associated protein stargazin has been shown to be quantitatively phosphorylated, and stargazin phosphorylation promotes synaptic trafficking of AMPA receptors. Synaptic NMDA receptor activity can induce both stargazin phosphorylation, via activation of CaMKII and PKC, and stargazin dephosphorylation, by activation of PP1 downstream of PP2B. At hippocampal synapses, long-term potentiation and long-term depression require stargazin phosphorylation and dephosphorylation, respectively. These results establish stargazin as a critical substrate in the bidirectional control of synaptic strength, which is thought to underlie aspects of learning and memory (Tomita, 2005).

The related small GTPases Ras and Rap1 are important for signaling synaptic AMPA receptor (-R) trafficking during long-term potentiation (LTP) and long-term depression (LTD), respectively. Rap2, which shares 60% identity to Rap1, is present at excitatory synapses, but its functional role is unknown. This study reports that Rap2 activity, stimulated by NR2A-containing NMDA-R activation, depresses AMPA-R-mediated synaptic transmission via activation of JNK rather than Erk1/2 or p38 MAPK. Moreover, Rap2 controls synaptic removal of AMPA-Rs with long cytoplasmic termini during depotentiation. Thus, Rap2-JNK pathway, which opposes the action of the NR2A-containing NMDA-R-stimulated Ras-ERK1/2 signaling and complements the NR2B-containing NMDA-R-stimulated Rap1-p38 MAPK signaling, channels the specific signaling for depotentiating central synapses (Zhu, 2005).

Activity-dependent synaptic delivery of GluR1-, GluR2L-, and GluR4-containing AMPA receptors (-Rs) and removal of GluR2-containing AMPA-Rs mediate synaptic potentiation and depression, respectively. The obvious puzzle is how synapses maintain the capacity for bidirectional plasticity if different AMPA-Rs are utilized for potentiation and depression. This study shows that synaptic AMPA-R exchange is essential for maintaining the capacity for bidirectional plasticity. The exchange process consists of activity-independent synaptic removal of GluR1-, GluR2L-, or GluR4-containing AMPA-Rs and refilling with GluR2-containing AMPA-Rs at hippocampal and cortical synapses in vitro and in intact brains. In GluR1 and GluR2 knockout mice, initiation or completion of synaptic AMPA-R exchange is compromised, respectively. The complementary AMPA-R removal and refilling events in the exchange process ultimately maintain synaptic strength unchanged, but their long rate time constants (15-18 hr) render transmission temporarily depressed in the middle of the exchange. These results suggest that the previously hypothesized 'slot' proteins, rather than AMPA-Rs, code and maintain transmission efficacy at central synapses (McCormack, 2006).

How synaptic AMPA-R exchange maintains transmission efficacy is unclear. It is possible that during exchange, GluR2-containing AMPA-Rs get into synapses and make a one-to-one replacement of synaptic GluR1-, GluR2L-, or GluR4-containing AMPA-Rs. It is also possible that synaptic delivery of GluR1-, GluR2L-, and GluR4-containing AMPA-Rs brings with them 'slot' proteins, which allow GluR2-containing AMPA-Rs to refill the empty 'slots' after GluR1-, GluR2L-, and GluR4-containing AMPA-Rs leave synapses. The temporarily depressed AMPA responses during synaptic AMPA-R exchange indicate that GluR1-, GluR2L-, and GluR4-contaning AMPA-Rs leave synapses before GluR2-containing AMPA-Rs fill in. This view is further supported by findings that synaptic refilling of GluR2-containing AMPA-Rs after removal of AMPA-Rs with long cytoplasmic termini are required for completing exchange and maintaining transmission unaltered after exchange. Because the ultimate transmission strength does not change after the exchange, synaptic efficacy must be 'memorized' by molecule(s) other than AMPA-Rs, in particular when AMPA-R-mediated transmission is temporarily depressed. The results thus provide experimental evidence supporting the 'slot' theory: 'slot' proteins, instead of AMPA-Rs, code and maintain transmission efficacy. The remaining puzzle is what are the 'slot' proteins. Proteomic analysis and functional characterization of proteins binding to cytoplasmic termini of all AMPA-R subunits promise to reveal their identity (McCormack, 2006).

Cerebellar long-term depression (LTD) is a major form of synaptic plasticity that is thought to be critical for certain types of motor learning. Phosphorylation of the AMPA receptor subunit GluR2 on serine-880 as well as interaction of GluR2 with PICK1 have been suggested to contribute to the endocytic removal of postsynaptic AMPA receptors during LTD. This study shows that targeted mutation of PICK1, the GluR2 C-terminal PDZ ligand, or the GluR2 PKC phosphorylation site eliminates cerebellar LTD in mice. LTD can be rescued in cerebellar cultures from mice lacking PICK1 by transfection of wild-type PICK1 but not by a PDZ mutant or a BAR domain mutant deficient in lipid binding, indicating the importance of these domains in PICK1 function. These results demonstrate that PICK1-GluR2 PDZ-based interactions and GluR2 phosphorylation are required for LTD expression in the cerebellum (Steinberg, 2006).

Glycogen synthase kinase-3 (GSK3) has been implicated in major neurological disorders, but its role in normal neuronal function is largely unknown. GSK3β mediates an interaction between two major forms of synaptic plasticity in the brain, NMDA receptor-dependent long-term potentiation (LTP) and NMDA receptor-dependent long-term depression (LTD). In rat hippocampal slices, GSK3β inhibitors block the induction of LTD. Furthermore, the activity of GSK3β is enhanced during LTD via activation of PP1. Conversely, following the induction of LTP, there is inhibition of GSK3β activity. This regulation of GSK3β during LTP involves activation of NMDA receptors and the PI3K-Akt pathway and disrupts the ability of synapses to undergo LTD for up to 1 hr. It is concluded that the regulation of GSK3β activity provides a powerful mechanism to preserve information encoded during LTP from erasure by subsequent LTD, perhaps thereby permitting the initial consolidation of learnt information (Peineau, 2007).

NMDA receptor-dependent LTD is due to the internalization of AMPA receptors and involves protein interactions directly associated with the AMPA receptor subunits, particularly GluR2. It was reasoned that GSK3β might form a complex with AMPA receptors, and thus attempts were made to investigate this by probing for an association of native GSK3β with AMPA receptors in the CA1 area of hippocampal slices. A specific antibody against GSK3β was able to coimmunoprecipitate the GluR1 and GluR2 AMPA receptor subunits, and conversely, immunoprecipitation of AMPA receptors produced coimmunoprecipitation of GSK3β. To determine the functional status of AMPA receptor-associated GSK3β, AMPA receptors were immunoprecipitated, using antibodies against either GluR1 or GluR2, and then assayed for kinase activity. GSK3β activity was readily detected in both GluR1 and GluR2 immunoprecipitates relative to the background IgG control, demonstrating that endogenous GSK3β associates with native AMPA receptors in the brain, and that the bound GSK3β is functionally active. This association of GSK3β with AMPA receptors suggests a compartmentalization of this enzyme for the efficient regulation of AMPA receptors during LTD (Peineau, 2007).

It was asked whether the GSK3β activity that is associated with AMPA receptors could be regulated. Previous work has shown that transient exposure of cultured neurons to a solution containing sucrose plus glycine leads to an NMDA receptor-dependent insertion of AMPA receptors into the plasma membrane. Interestingly, this effect is associated with an increase in AMPA receptor-associated PI3K activity. Since PI3K is an upstream regulator of GSK3β, whether this treatment also affected the AMPA receptor-associated GSK3β enzyme activity was investigated. Neurons were treated with sucrose (200 mM) plus glycine (100 µM) for 2 min, and this led to the insertion of AMPA receptors into the plasma membrane as determined approximately 15 min later using surface biotinylation assays. This chemically induced AMPA receptor insertion was associated with a decrease in AMPA receptor-associated GSK3β activity (Peineau, 2007).

This study identified a form of regulation of synaptic plasticity in which the transient synaptic activation of NMDA receptors, as occurs during LTP, leads to inhibition of LTD. This regulation is very powerful since LTD is fully inhibited immediately following the conditioning stimulus and the effect lasts for approximately 1 hr. Also some of the signaling pathways responsible for this potent regulation of synaptic plasticity have been identified. GSK3β activity is an absolute requirement for the induction of LTD and the conditioning stimulus inhibits its activity via activation of the PI3K-Akt pathway. Finally, there is a correlation between the phosphorylation state of GSK3β ser9 and whether NMDA receptor activation leads to the induction or inhibition of LTD (Peineau, 2007).

GSK3β is an unusual kinase that has been implicated in many diseases. However, very little is known about its normal function in the nervous system. It is important during early development and it has been shown to play a key role in cell polarity and in the growth of neuromuscular junctions. Recently, it has been shown that GSK3β is important for determining neuronal polarity during the development of hippocampal neurons. However, though GSK3β is also highly expressed in the mature brain, its function in the nervous system has, hitherto, been largely unexplored. In the nucleus of hippocampal neurons, GSK3β is involved in the regulation of gene transcription by promoting the nuclear export of the transcription factor NF-ATc4. In addition, it has been shown that overexpression of GSK3β impairs spatial learning, though the mechanism underlying this effect is unknown. This study shows that in 2-week-old rats, an age at which the expression of GSK3β is near its peak, GSK3β activity is essential for NMDA receptor-dependent LTD in the hippocampus. This form of LTD is widespread throughout the brain and has been strongly implicated in development and learning and memory. Therefore, this novel GSK3β-dependent mechanism may be of general significance in regulating the interaction between LTP and LTD throughout the brain (Peineau, 2007).

GSK3β, unlike most enzymes, possesses high basal level constitutive activity and can be bidirectionally regulated to either further increase or decrease its activity. During LTD there is additional activation of GSK3β, probably via dephosphorylation of ser9. This effect is prevented by an inhibitor of PP1/PP2A. This suggests that the activation of PP1, which is known to occur during LTD, is responsible for the activation of GSK3β, via its dephosphorylation of ser9. LTD is associated with inhibition of Akt, probably also via the activation of PP1. These data suggest that GSK3β activity is increased during LTD because the phosphatase concomitantly inhibits Akt and directly dephosphorylates ser9 of GSK3β (Peineau, 2007).

Interestingly, the alteration in the phosphorylation status of GSK3β persists beyond the delivery of low-frequency stimulation (LFS), and lithium completely blocks LTD when applied after the delivery of LFS. These data suggest that GSK3β is required for the LTD process beyond the initial induction phase. Further studies are required to determine the full time course of the involvement of GSK3β in LTD (Peineau, 2007).

GSK3β has several upstream regulators and numerous downstream targets. In the present study, two of its upstream regulators have been identified. During LTD, GSK3β is activated via an okadaic acid-sensitive protein phosphatase, which is probably PP1. During LTP, GSK3β is inhibited via the PI3K-Akt pathway. Since GSK3β is such a ubiquitous kinase, it needs mechanisms to localize its access to its substrates. This is achieved in part via direct interactions with other proteins to form complexes. For example, in the canonical Wnt pathway, GSK3β binding proteins control access of β-catenin. It seems likely that the association between GSK3β and AMPA receptors serves to localize the kinase close to substrates that are involved in the trafficking of these receptors during synaptic plasticity. Further studies are required to establish the mechanism of this interaction as well as the downstream pathways mediated by GSK3β in the regulation of LTD (Peineau, 2007).

The finding that the synaptic activation of NMDA receptors during LTP inhibits NMDA receptor-dependent LTD raises an intriguing issue: what determines whether the synaptic activation of NMDA receptors leads to the induction or inhibition of LTD? Evidence is presented that the phosphorylation state of ser9 of GSK3β is a critical determinant. Thus, during LTP, activation of the PI3K-Akt pathway results in phosphorylation of GSK3β, and hence inhibition of its activity. In contrast, during LTD, activation of PP1 results in inhibition of Akt and the dephosphorylation of GSK3β at ser9, and this leads to an increase in the enzyme's activity. The activation of PI3K-Akt and inhibition of PP1 during LTP, but inhibition of Akt during LTD as well as the selective activation of PP1 during LTD, can be explained by the differences in the magnitude and spatiotemporal properties of the Ca2+ rise associated with the synaptic activation of NMDA receptors during these two forms of synaptic plasticity (Peineau, 2007).

Previous work has described other ways in which synaptic plasticity can be powerfully influenced by the prior history of synaptic activity. However, the mechanisms involved in these forms of metaplasticity are not known. Why synapses need such regulatory mechanisms is a matter of conjecture. One intriguing possible role for the regulation described in this study is to stabilize a synaptic modification over the short term by protecting synapses from the effects of additional NMDA receptor-dependent plasticity until the information can be either consolidated or erased by NMDA receptor-independent mechanisms (Peineau, 2007).

The regulation of synaptic plasticity is further complicated by the involvement of mGluRs, which are involved in depotentiation, LTD of baseline transmission, heterosynaptic LTD, and metaplasticity. So that focus could be placed on interactions between the NMDA receptor-dependent forms of synaptic plasticity, the additional complication of mGluR-dependent synaptic plasticity were eliminated by using the broad spectrum mGluR antagonist LY341495 and by employing stimulus protocols optimized for NMDA receptor-dependent synaptic plasticity. However, given that PI3K has been implicated in a chemically induced form of mGluR-dependent LTD and heterosynaptic LTD, it will be interesting to determine whether GSK3β is also involved in these forms of synaptic plasticity. One possibility is that the PI3K-Akt-GSK3β pathway serves to inhibit NMDA receptor-dependent LTD both homosynaptically following the induction of LTP and heterosynaptically following the induction of LTD (Peineau, 2007).

The finding that in the normal brain activation of GSK3β is essential for NMDA receptor-dependent LTD, and that its activity can be regulated by LTP, may offer clues to the pathological role of this enzyme in neurological disorders. For example, the primary therapeutic action of lithium in bipolar disorders may be via inhibition of GSK3β. Indeed, specific inhibition of GSK3β has recently been shown to produce antidepressive-like activity in vivo. Overactivity of GSK3β may, therefore, lead to this mood disorder by affecting the balance and interplay between NMDA receptor-dependent LTP and LTD (Peineau, 2007).

The scaffold protein PSD-95 promotes the maturation and strengthening of excitatory synapses, functions that require proper localization of PSD-95 in the postsynaptic density (PSD). Phosphorylation of ser-295 enhances the synaptic accumulation of PSD-95 and the ability of PSD-95 to recruit surface AMPA receptors and potentiate excitatory postsynaptic currents. Evidence is presented that a Rac1-JNK1 signaling pathway mediates ser-295 phosphorylation and regulates synaptic content of PSD-95. Ser-295 phosphorylation is suppressed by chronic elevation, and increased by chronic silencing, of synaptic activity. Rapid dephosphorylation of ser-295 occurs in response to NMDA treatment that causes chemical long-term depression (LTD). Overexpression of a phosphomimicking mutant (S295D) of PSD-95 inhibits NMDA-induced AMPA receptor internalization and blocks the induction of LTD. The data suggest that synaptic strength can be regulated by phosphorylation-dephosphorylation of ser-295 of PSD-95 and that synaptic depression requires the dephosphorylation of ser-295 (Kim, 2007).

Glutamate receptors and paired-associated learning

Paired-associate learning is often used to examine episodic memory in humans. Animal models include the recall of foodcache locations by scrub jays and sequential memory. This study reports a model in which rats encode, during successive sample trials, two paired associates (flavors of food and their spatial locations) and display better-than-chance recall of one item when cued by the other. In a first study, pairings of a particular foodstuff and its location were never repeated, so ensuring unique 'what-where' attributes. Blocking NMDA receptors in the hippocampus -- crucial for the induction of certain forms of activity-dependent synaptic plasticity -- impairs memory encoding but has no effect on recall. Inactivating hippocampal neural activity by blocking AMPA receptors impairs both encoding and recall. In a second study, two paired associates were trained repeatedly over 8 weeks in new pairs, but blocking of hippocampal AMPA receptors does not affect their recall. Thus it is concluded that unique what-where paired associates depend on encoding and retrieval within a hippocampal memory space, with consolidation of the memory traces representing repeated paired associates in circuits elsewhere (Day, 2003).

AMPA receptors and direction-selective and spatial phase-selective responses in the visual cortex

Cells in the superficial layers of primary visual cortex (area 17) are distinguished by feedforward input from thalamic-recipient layers and by massive recurrent excitatory connections between neighboring cells. The connections use glutamate as transmitter, and the postsynaptic cells contain both NMDA and AMPA receptors. The possible role of these receptor types in generating emergent responses of neurons in the superficial cortical layers is unknown. NMDA and AMPA receptors are both involved in the generation of direction-selective responses in layer 2/3 cells of area 17 in cats. NMDA receptors contribute prominently to responses in the preferred direction, and their contribution to responses in the nonpreferred direction is reduced significantly by GABAergic inhibition. AMPA receptors decrease spatial phase-selective simple cell responses and generate phase-invariant complex cell responses (Rivadulla, 2001).

By combining extracellular recording and iontophoresis of receptor blockers, the following results have been demonstrated: (1) Blocking AMPA receptors removes a proportionately larger component from nonpreferred compared with preferred responses and increases direction selectivity. The remaining responses are mediated by NMDA receptors and are overwhelmingly in the preferred direction. (2) Blocking NMDA receptors removes proportional components from preferred and nonpreferred responses and preserves directional selectivity. Because the remaining responses are mediated by AMPA receptors, these receptors are sufficient for direction selectivity. (3) Blocking inhibition preferentially enhances the contribution of NMDA receptors to nonpreferred responses and reduces direction selectivity. Thus, inhibition contributes to direction selectivity by reducing NMDA responses in the nonpreferred direction. (4) Blocking AMPA receptors increases the modulation of complex cell responses by a drifting grating stimulus. Thus, AMPA receptors decrease the selectivity of complex cells for spatial phase or the spatial location of visual stimuli. (5) Blocking NMDA receptors or inhibition has little effect on the temporal modulation of simple or complex cell responses. Together, these results allow for the proposal specific roles for NMDA and AMPA receptors in direction selectivity in the superficial layers of area 17 and in the generation of phase selectivity by simple and complex cells in these layers (Rivadulla, 2001).

Direction selectivity first appears in simple cells of layer 4 in area 17, where NMDA receptors are not present in significant proportions and contribute little to visual responses. The mechanism(s) by which direction selectivity is generated and whether the mechanism is similar in various cortical layers remain unresolved. One hypothesis is that inhibition reduces the response in the no