Although Wit shows a strong overall similarity to the vertebrate BMP type II receptor, its signaling mechanism remains untested. To examine whether Wit actually mediates a BMP type signal, use was made of an antibody that specifically recognizes the phosphorylated form of Mad (P-Mad) (Tanimoto, 2000), the major transducer of BMP signals in Drosophila. In late embryos, this antibody reveals an intricate developmental pattern of BMP signaling, including P-Mad accumulation in the developing midgut and gastric cecae that closely parallels dpp expression and is thought to be indicative of cells receiving a Dpp signal. In addition to these sites of accumulation, a novel pattern of P-Mad accumulation is noted in a subset of neurons. Staining is first evident at late stage 15 and then becomes more elaborate and intense as the embryos continue to develop. Ultimately about 35-40 neurons per hemisegment show strong staining by stage 17. This expression continues into the first instar stage but is not detectable in late third instar larvae. The pattern of P-Mad accumulation in the CNS is completely abolished in wit mutant embryos whereas all other patterns of P-Mad accumulation appear normal. Conversely, with the exception of the CNS pattern, all other sites of P-Mad accumulation in late embryogenesis are absent in zygotic punt null mutants. Maternal null embryos cannot be examined since they do not develop an organized CNS. These observations are consistent with Wit function being required specifically in the CNS to transduce BMP signals whereas other sites of P-Mad accumulation result from signaling by Punt, the primary Dpp type II receptor (Marqués, 2002).
The BMP receptor Wishful thinking (Wit) is required for synapse stabilization. In the absence of BMP signaling, synapse disassembly and retraction ensue. Remarkably, downstream Smad-mediated signaling cannot fully account for the stabilizing activity of the BMP receptor. LIM Kinase1 (DLIMK1)-dependent signaling has been identified as a second, parallel pathway that confers the added synapse-stabilizing activity of the BMP receptor. DLIMK1 binds a region of the Wit receptor that is necessary for synaptic stability but is dispensable for Smad-mediated synaptic growth. A genetic analysis demonstrates that DLIMK1 is necessary, presynaptically, for synapse stabilization, but is not necessary for normal synaptic growth or function. Furthermore, presynaptic expression of DLIMK1 in a wit or mad mutant significantly rescues synaptic stability, growth, and function. DLIMK1 localizes near synaptic microtubules and functions independently of ADF/cofilin (Twinstar), highlighting a novel requirement for DLIMK1 during synapse stabilization rather than actin-dependent axon outgrowth (Eaton, 2005).
The canonical bone morphogenic protein (BMP) signaling system has been implicated in diverse cellular and developmental processes ranging from cell growth to tissue patterning. At the Drosophila NMJ, a BMP signaling system has been identified that controls synaptic growth via canonical Smad-mediated signaling to the cell body. It has been demonstrated that mutations in the BMP ligand glass bottom boat (gbb), the type I and type II BMP receptors thick veins (tkv) and wishful thinking (wit), and the Smad homologs mad and medea all significantly impair synaptic growth and function. These data define a retrograde trophic signaling system that functions through transcriptional mechanisms in the cell soma to control motoneuron synaptic growth. This study demonstrates that BMP signaling at the Drosophila NMJ is not only required for normal synaptic growth, but also for synaptic stabilization. In the absence of BMP signaling, significant increases in synapse retraction and disassembly are observed. Signaling downstream of the BMP receptors can be genetically separated into two pathways: Smad-dependent synaptic growth and LIM Kinase1-dependent synaptic stability (Eaton, 2005).
LIM Kinase1 (LIMK1) is a cytoplasmic serine/threonine kinase that was originally isolated in screens for novel kinases expressed in the nervous system. Findings in LIMK1 knockout mice reveal defects in dendritic spine morphology and activity-dependent plasticity, although neither synaptic growth nor synaptic stability has been specifically analyzed (Endo, 2003). In the Drosophila central nervous system, DLIMK1 has been implicated in the mechanisms of axonal outgrowth during metamorphosis, acting through ADF/cofilin to modulate the actin cytoskeleton, a mechanism also documented in other tissues (Arber, 1998; Ng, 2004; Ohashi, 2000; Eaton, 2005 and references therein).
Genetic analyses have defined a new function for DLIMK1 during synaptic stability in comparison with its function in axonal outgrowth. Synaptic DLIMK1 is closely associated with the synaptic microtubule cytoskeleton. In addition, genetic manipulation of ADF/cofilin activity does not affect synaptic stability at the NMJ. These data highlight differences in LIMK1 function during the rapid, dynamic process of axon outgrowth compared to the slower, more prolonged mechanisms that govern synapse stabilization at the NMJ. Together, these data define genetically separable signaling pathways downstream of the BMP receptor that could allow a single trophic signaling event to coordinately control synaptic growth and synaptic stabilization (Eaton, 2005).
An assay to quantify synaptic retraction at the Drosophila NMJ was developed and used to identify molecules involved in synaptic stability (Eaton, 2002). This assay is based on the demonstration that the formation of organized postsynaptic muscle membrane folds, termed the subsynaptic reticulum (SSR), requires the presence of the presynaptic nerve terminal. Therefore, the SSR and proteins that localize to this structure will be present only at sites where the nerve terminal resides, or where it has recently resided. Thus, observed sites of organized postsynaptic SSR that lack opposing presynaptic neuronal markers identify regions of the neuromuscular junction (NMJ) where the nerve terminal once resided and has since retracted. This interpretation has been confirmed in previous studies using light-level, ultrastructural, and electrophysiological analyses (Eaton, 2002; Pielage, 2005). Sites of synapse retraction are referred to as 'retraction events' or 'synaptic footprints' and represent a quantitative assay for synaptic stability. A wide array of pre- and post-synaptic markers have been used to clearly define synaptic retraction events. Synaptic retractions can be identified with equal efficiency using antibodies that recognize diverse presynaptic antigens, including cytoplasmic, membrane-associated, cytoskeleton, or vesicle-associated proteins (Eaton, 2002; Pielage, 2005). Several postsynaptic markers have also been used to quantify synapse retractions, including Discs-large, Shaker-GFP, and the clustered postsynaptic glutamate receptors (Eaton, 2002; Pielage, 2005; Eaton, 2005 and references therein).
The retraction assay was used to test whether BMP signaling in Drosophila motoneurons is required for synapse stabilization. First it was confirmed that mutations in the BMP type II receptor Wit and the BMP ligand Gbb cause a significant decrease in bouton number. These mutations also cause a significant increase in synaptic footprints, demonstrating that synaptic stability is significantly compromised in the absence of BMP signaling. Synaptic footprints were identified in equal numbers using multiple presynaptic markers, including anti-Synapsin and anti-nc82, which recognizes an antigen at the presynaptic active zone. Synaptic footprints can be rescued in the wit mutant background by neuronal expression of the full-length wit transgene, demonstrating that synapse destabilization is caused by the absence of the presynaptic Wit receptor. The number of synapse retractions is slightly, but statistically significantly, less in gbb compared with that in wit. This could be due to the fact that wit is a null mutation, whereas the gbb genotype that was used is not. Null mutations in gbb do not survive through larval development and, therefore, the genetic combination previously used in studies of its effects on synaptic growth was used (Eaton, 2005).
This analysis was extended to mutations that disrupt additional downstream components of the canonical BMP signaling cascade. Mutations in the BMP type I receptor thick-veins (tkv), the Smad homolog mad, and the co-Smad medea, were used. All three mutations decrease bouton numbers to levels that are statistically identical to those observed in the wit and gbb mutations. All three mutations also cause a statistically significant increase in synaptic footprints compared to wild-type, demonstrating that canonical BMP signaling is necessary for synaptic stability as well as for growth (Eaton, 2005).
The fact that synaptic footprints in wit can be rescued by neuronal expression of UAS-wit suggests that BMP signaling is necessary in the motoneuron for synaptic stability. To confirm that Smad-mediated signaling is also required within the motoneuron for synaptic stabilization, the inhibitory Smad dad was neuronally overexpressed (dad gain of function, DAD GOF). Genetic evidence suggests that this manipulation can block both Mad and DSmad2 signaling. Neuronal DAD GOF was found to decrease bouton numbers and increases synaptic footprints to levels that are near those observed in tkv, mad, and medea. Finally, in order to examine whether Smad signaling is required in the cell soma for synaptic stabilization, a genetic interaction was examined between the wit mutation and the overexpression of a dominant-negative Glued transgene (DN-Glued) that disrupts retrograde axonal transport. Previous work has demonstrated that impairment of dynactin function in motoneurons disrupts retrograde axonal transport and causes synapse retraction at the NMJ. The number of retractions in the DN-Glued; wit double mutant is not additive with respect to the wit mutation or DN-Glued overexpression. Although the overexpression of DN-Glued cannot be considered a null mutant condition, it is sufficient to block the accumulation of nuclear phospho-Mad and therefore prevents BMP signaling to the nucleus. Thus, it is concluded that Smad-mediated signaling to the motoneuron cell soma is necessary for both synaptic growth and synaptic stability. Together, these data establish an in vivo link between retrograde axonal transport, an essential trophic signaling system, and the mechanisms of synaptic stabilization (Eaton, 2005).
Additional experiments tested whether mutations that decrease the number of boutons always increase the number of footprints. In these experiments, mutations were analyzed that decrease bouton number but that have not been implicated in the BMP signaling system. Bouton numbers are significantly decreased in these three independent mutations: the cell adhesion molecule fasciclin II (fasII); the tyrosine phosphatase Dlar, and the microtubule binding protein futsch. In each of these three mutations, bouton numbers are significantly decreased whereas the number of synaptic footprints remains unchanged. It has also been shown that impaired synaptic growth in a presynaptic calcium channel mutant does not alter synaptic stability. From these data it can be concluded that impaired synaptic growth does not necessarily impair synaptic stabilization. By extension, it is concluded that BMP signaling is required for two separable processes, synaptic growth and synaptic stability (Eaton, 2005).
While all of the mutations in canonical BMP signaling molecules decrease bouton numbers to the same extent, there are significantly fewer synaptic footprints in the tkv, mad, and medea mutations (as well as in DAD GOF) when compared to the wit mutation. Thus, while canonical BMP signaling is required for both synaptic growth and stability, the Wit receptor has an additional stabilizing influence on the NMJ that cannot be accounted for by the downstream Smad signaling system. Therefore experiments were pursued to investigate the mechanism of Wit-mediated synaptic stability that appears to be independent of Smad-mediated signaling (Eaton, 2005).
It was first determined whether the signaling associated with synaptic stabilization and synaptic growth might map to different regions of the cytoplasmic tail of the Wit receptor. To do so, a transgenic rescue approach was taken that involves neuronal expression of either the full-length wit transgene or a truncated wit transgene in the wit mutant background. The truncated Wit receptor lacks a C-terminal portion that is not required for Smad signaling in mammalian systems and has been shown to restore viability to the wit mutation. First, it was demonstrated that presynaptic expression of the full-length wit transgene in the wit mutant background rescues synaptic growth and synaptic stability. However, while presynaptic expression of the truncated wit transgene (wit-dCT) completely rescues synaptic growth, it is unable to fully restore synaptic stability to wild-type levels. It was confirmed that the Wit-dCT receptor is able to activate downstream Smad signaling, by showing that neuronal expression of the wit-dCT construct rescues the presence of nuclear phospho-Mad in the wit mutant background.
These data identify a region of the Wit receptor that is necessary for synapse stabilization, but is not required for nuclear Smad signaling, synaptic growth, or synaptic function. Experiments were performed to investigate how this region of the Wit receptor influences synaptic stability (Eaton, 2005).
The C-terminal tail of mammalian BMP type II receptors has been shown to interact with and regulate the activity of LIMK1 in vitro (Foletta, 2003; Lee-Hoeflich, 2004). This study demonstrated that Drosophila LIM Kinase1 (DLIMK1) binds the C-terminal tail of the Wit receptor that is deleted in the wit-dCT transgene. Peptides corresponding to the N-terminal region of DLIMK1 containing either the tandem LIM and PDZ domains together (LIM-PDZ) or containing the tandem LIM domains alone (LIM only) interact with the C-terminal region of the Wit receptor (C-Term), but not with the kinase domain (Kinase) in a yeast two-hybrid binding assay. No interactions were detected between the PDZ domain of DLIMK1 and the Wit receptor in this analysis. In addition, an interaction was detected between the C-terminal region of the Wit receptor (Wit-CT-FLAG) and full-length DLIMK1 (DLIMK1-HA) when these proteins were coexpressed in Drosophila S2 cells. These observations suggest that the LIM domains of DLIMK1 mediate specific binding to the C-terminal tail of the Wit receptor, consistent with findings (Foletta, 2003; Lee-Hoeflich, 2004) reported for LIMK1 binding to mammalian BMP receptors (Eaton, 2005).
Next, whether DLIMK1 is required for synaptic growth and/or stabilization was tested. A P element insertion (DLIMKP1) was acquired that resides in a 5′ untranslated exon of the DLIMK1 gene as well as two deficiency chromosomes that uncover the DLIMK1 locus. Northern analysis demonstrates that DLIMKP1/DLIMKP1 results in the near absence of detectable message, indicating that this is a strong-hypomorphic or null mutation. The DLIMK1 mutations, including DLIMKP1/DLIMKP1, DLIMKP1/Y, DLIMKP1/Df(1)JA26, and DLIMKP1/Df(1)HF368 are all viable. This is consistent with the observation that LIMK1 knockout mice are homozygous viable (Meng, 2002). This study demonstrates that DLIMKP1/DLIMKP1, DLIMKP1/Y, DLIMKP1/Df(1)JA26, and DLIMKP1/Df(1)HF368 all cause a significant increase in synaptic footprints (p <0.001) without decreasing synaptic bouton number. These genetic data indicate that DLIMK1 is necessary for synaptic stability, but is not required for normal synaptic growth. Since the number of synaptic footprints is comparable when DLIMKP1 is analyzed in trans to deficiency chromosomes that uncover the DLIMK1 locus, it is concluded that this P element insertion represents a strong loss-of-function mutation (Eaton, 2005).
To support the conclusion that DLIMK1 function is necessary for synapse stabilization and to determine whether DLIMK1 functions in the nerve or the muscle to control synaptic stability, transgenic rescue experiments and overexpression experiments were pursued using a dominant-negative DLIMK1 transgene. It was found that neuronal expression of a dominant-negative, kinase-inactive DLIMK1 (DN-DLIMK1) transgene significantly increases synaptic retractions while muscle-specific expression of DN-DLIMK1 has no effect. There was a slight decrease in bouton number observed when DN-DLIMK1 was expressed neuronally that was not observed in the DLIMK1 mutations. This may represent a dominant effect of this transgene. However, this decrease in bouton number is significantly less than that observed following disruption of BMP signaling. Finally, in agreement with experiments using a dominant-negative transgene, it was found that neuronal expression of a wild-type DLIMK1 transgene (UAS-DLIMK1) restores synapse stability to the hemizygotic DLIMKP1/Y loss-of-function mutation without altering other aspects of synapse development. Together, these data support the conclusion that DLIMK1 is specifically required in the presynaptic motoneuron for synaptic stabilization (Eaton, 2005).
The observation that DLIMK1 binds the Wit receptor and is required presynaptically for synapse stabilization suggests that DLIMK1 functions downstream of the Wit receptor and may confer the added synapse-stabilizing activity of the Wit receptor. To address this possibility, genetic interactions were examined between the wit and DLIMK1 mutations. Animals harboring one mutant copy of wit and one mutant copy of DLIMK1 (DLIMKP1/+; witA12/+) show a significant increase in synaptic footprints without a change in bouton number and without a change in synaptic function compared to wild-type. In comparison, heterozygous mutations in either gene alone do not show a significant increase in synaptic footprints compared to wild-type and have normal bouton numbers. The DN-DLIMK1 was neuronally overexpressed in the wit mutant background and no additive effect was found on the number of NMJs with retractions with respect to the wit mutation or the overexpression of DN-DLIMK1 alone. The strong transheterozygous interaction between the witA12 mutation and DLIMK1P1, as well as the lack of an additive effect in the number of NMJs with footprints when DN-DLIMK1 is overexpressed in the wit mutant background, supports the conclusion that DLIMK1 functions in the same genetic pathway as wit to control synaptic stability (Eaton, 2005).
Next whether overexpression of UAS-DLIMK1 could rescue synaptic stability in the wit mutant background was investigated. To do so, the wild-type DLIMK1 transgene (UAS-DLIMK1) was neuronally expressed in the wit mutant background. When UAS-DLIMK1 was expressed in the wit mutant background using either a pan-neuronal (elav-GAL4) or motoneuron-specific (OK6-GAL4) GAL4 driver, defects were rescued in both synaptic growth and synaptic stability. Importantly, strong presynaptic expression of UAS-DLIMK1 by elav-GAL4 raised at 29°C resulted in complete rescue of synaptic footprints to wild-type levels. In all rescue experiments, including overexpression of DLIMK1 in the wit background, the presence of the wit mutation was confirmed. These data, in combination with the demonstration that DLIMK1 binds the Wit receptor and the genetic interactions observed between wit and DLIMK1, support the conclusion that DLIMK1 functions downstream of wit to control synaptic stability (Eaton, 2005).
Remarkably, rescue of the wit mutation by UAS-DLIMK1 not only restores synaptic growth and stability, but also rescues animal viability. wit mutant animals normally die during midlarval life and are never observed to mature into adult flies. Neuronal expression of DLIMK1 (with elav-GAL4) in the wit mutant background restores adult viability. Adult flies emerge in large numbers and are able to climb normally the vertical sides of a plastic vial (Eaton, 2005).
A further analysis has demonstrated that DLIMK1 expression restores synaptic function to the wit receptor mutation. Electrophysiological analysis of the wit mutant synapse demonstrates a severe impairment of synaptic transmission including a dramatic decrease in the excitatory postsynaptic potential (EPSP) amplitude, a significant decrease in the average amplitude of the spontaneous miniature release events (mEPSP), and a dramatic decrease in presynaptic release as estimated by the average EPSP/average mEPSP. Expression of UAS-DLIMK1 restores all aspects of synaptic function to near wild-type levels, including EPSP amplitude, mEPSP amplitude, and quantal content. The degree of rescue with UAS-DLIMK1 is not statistically different from that observed when wit is rescued by the full-length wit transgene. A slight but significant decrease in quantal content was observed in the DLIMKP1/DLIMKP1 homozygotes compared to wild-type, although this decrease is much less than that observed in wit or mad mutants. It is therefore concluded that DLIMK1 is able to rescue the wit electrophysiological defects even though it is not normally involved in BMP-dependent regulation of synaptic function. Thus, DLIMK1 expression can restore all of the essential functions of the Wit receptor in the wit mutant background (Eaton, 2005).
Next to be tested was whether the rescue of synaptic growth and stability by DLIMK1 expression in the wit mutant background is caused by activation of downstream Mad signaling that is independent of the Wit receptor. It was first demonstrated that mad mutations have a significant increase in synaptic footprints and a significant reduction in bouton number. It was found that neuronal expression of DLIMK1 in the mad mutant background rescues both synaptic growth and stability to wild-type levels. However, unlike expression of DLIMK1 in the wit mutant background, expression of DLIMK1 does not restore viability to the mad mutant background. This difference is likely due to the expression of mad outside of the nervous system, whereas wit expression is largely restricted to the nervous system. Thus, DLIMK1 is sufficient to promote synaptic growth and stability in the absence of Mad-mediated signaling. These data support the conclusion that DLIMK1 functions in parallel to Mad-mediated signaling in the motoneuron to control synaptic stability and growth (Eaton, 2005).
The localization of DLIMK1 was investigated in the presence and absence of the Wit receptor. The UAS-DLIMK1 transgene used in the rescue experiments harbors an HA epitope tag that was used to visualize DLIMK1 at the NMJ. Staining for HA in animals with neuronal expression of UAS-DLIMK1 revealed a filamentous DLIMK1 localization throughout the presynaptic nerve terminal. This staining pattern is identical in wild-type or wit mutant animals that neuronally express UAS-DLIMK1. Surprisingly, DLIMK1-HA was found in very close association with the synaptic microtubules identified by anti-Futsch (Map1B-like protein). However, although DLIMK1-HA is closely associated with anti-Futsch, the two markers do not colocalize. Rather, the DLIMK1-HA staining appears as an independent continuous filament that extends throughout the NMJ. Two additional experiments were pursued to test whether the DLIMK1-HA staining pattern requires the integrity of the synaptic microtubule cytoskeleton. (1) DLIMK1-HA was overexpressed in the futsch mutant background, which severely disrupts the synaptic microtubule cytoskeleton. Filamentous DLIMK1-HA staining remains in the futsch mutation, indicating that DLIMK1 filaments are independent of stable synaptic microtubules. (2) Whether DLIMK1-HA is perturbed following treatment of the synapse with nocadozole was examined at a concentration that eliminates dynamic microtubule plus ends. It was found that nocadozole treatment slightly decreases the intensity of DLIMK1-HA staining, but clear filamentous staining remains. These data suggest that DLIMK1-HA filaments are closely associated with synaptic microtubules, but can be stable in the absence of a continuous synaptic microtubule cytoskeleton. This staining pattern does not resemble the synaptic localization of actin-GFP, which concentrates into randomly distributed patches throughout the NMJ. Thus, DLIMK1 concentrates into a unique compartment within the presynaptic nerve terminal that has not been previously described. In addition, instances were found in which this compartment approached the synaptic plasma membrane, where it could associate with membrane receptors such as the Wit receptor. Finally, the possibility that low levels of DLIMK1 are present throughout the cytoplasm and in association with synaptic actin cannot be ruled out, since steady-state protein distribution was examined in fixed tissue (Eaton, 2005).
The role of LIMK1 during axon outgrowth and growth cone motility is largely due to LIMK1 phosphorylation and inactivation of ADF/cofilin, which alters actin turnover (Endo, 2003; Ng, 2004). In these studies, the phosphorylation of ADF/cofilin by LIMK1 is antagonized by the Slingshot (Ssh) family of phosphatases (Niwa, 2002). Therefore, if ADF/cofilin is the target of LIMK1, then the overexpression of the Ssh phosphatase should mimic LIMK1 loss of function in this signaling cascade. It was found that overexpression of the ssh cDNA in the motoneuron (elav UAS-SSH wt) has no effect on synaptic growth, synaptic stability, or synaptic function. It was confirmed that Ssh protein traffics to the synapse in these experiments and that Ssh has a cytoplasmic localization that does not resemble what was observed for DLIMK1-HA. Finally, overexpression of a constitutively activated cofilin transgene (elav UAS-tsr-S3A) also has no effect on synaptic growth or stabilization. In combination with the observation that the distribution of synaptic DLIMK1 does not resemble the distribution of synaptic actin, these data suggest that LIMK1 is not acting through ADF/cofilin to control synaptic stability. These data highlight differences in LIMK1 function during the rapid, dynamic process of axon outgrowth versus the slower, more prolonged mechanisms that govern synapse stabilization at the NMJ (Eaton, 2005).
This study demonstrates that BMP-receptor signaling at the Drosophila NMJ controls both synaptic growth and synaptic stabilization. The data support a model in which signaling from the BMP receptor can coordinately activate two genetically separable developmental programs: (1) cell-wide regulation of neuronal growth via nuclear Smad signaling and (2) LIMK1-dependent synaptic stabilization. This organization of synaptic signaling systems involved in synaptic growth versus synaptic stabilization could have important consequences for neural development. Since a single trophic receptor can increase both cell-wide growth and synaptic stabilization, the efficiency of trophic signaling could be increased due to the coupling of new synapse growth with synaptic stabilization and retention. Another scenario is also possible that could help to explain how synaptic growth and elimination could occur simultaneously at the terminals of a single motoneuron. If LIMK1 functions locally at the synapse, then cell-wide growth signaling could be uncoupled from the local control of synaptic stability. Thus, the loss of trophic signaling at a single muscle target could lead to target-specific synapse destabilization while the same trophic signaling system at other muscle targets could simultaneously promote synaptic growth throughout the motoneuron via cell-wide growth signaling. While there is no direct evidence that LIMK1 functions locally at the synapse, the synaptic activation of kinase signaling downstream of the BMP receptor could support such a model (Eaton, 2005).
The data support a model in which DLIMK1 functions downstream of the Wit receptor and in parallel to Smad-mediated nuclear signaling in order to achieve wild-type synaptic stability. Several lines of evidence indicate that DLIMK1 functions downstream of the Wit receptor. DLIMK1 was shown to bind to a region of the Wit receptor that is necessary for synaptic stabilization, but which is dispensable for Smad-mediated synaptic growth. DLIMK1 mutations were shown to specifically disrupt synaptic stabilization and it was found that both DLIMK1 and Wit are necessary within the presynaptic neuron for synapse stability. A strong transheterozygotic genetic interaction was demonstrated between mutations in DLIMK1 and wit, resulting in a specific loss of synaptic stability. It was further shown that neuronal overexpression of a dominant negative DLIMK1 in the wit mutant background does not result in an additive increase in the number of synapse retractions. From these data, it is concluded that DLIMK1 functions downstream of Wit to control synaptic stability. It was also shown that the stabilizing activity of DLIMK1 does not require the presence of the Smad signaling system, since expression of DLIMK1 in either the wit or the mad mutant background is able to restore synaptic stability to the NMJ. While these data are consistent with DLIMK1 functioning either downstream or in parallel to Smad-mediated signaling, the conclusion is favored that DLIMK1 functions in parallel to Smad because DLIMK1 binds a region of the Wit receptor that is dispensable for Smad-mediated synaptic growth (Eaton, 2005).
These data demonstrate that signaling via the Wit receptor stabilizes the synapse via both Smad-dependent and LIMK1-dependent signaling. Mutations in mad, medea, and expression of the inhibitory Smad dad (DAD-GOF) all cause an increase in synapse retractions as well as a decrease in bouton number. The data indicate that Smad-mediated signaling accounts for approximately 50% of the stabilizing activity of the Wit receptor, and LIM Kinase functions in parallel to mediate the other 50% of the Wit receptor's stabilizing activity. It is hypothesized that the stabilizing functions of Smad signaling and LIMK1 signaling are quite different. First, it is suspected that Smad-dependent synaptic stabilization is directly coupled to the growth-promoting function of the nuclear Smad signaling system. It is hypothesized that Smad-dependent growth regulation involves transcriptional programs that produce the necessary raw material for synaptic growth. In the absence of Smad signaling, these raw materials may become limiting, not only for new growth but also for the maintenance of the existing synapse, since synaptic proteins will need to be turned over and replaced at some rate. According to this logic, the synapse retractions caused by mutations in the Smad signaling system are related to synaptic atrophy. In contrast, DLIMK1 is necessary for synaptic stability, but is not required for normal synaptic growth. Thus, it is suspected that synapse retractions observed in DLIMK1 mutants are caused by modulation of the synaptic stabilization machinery resident at the synapse (Eaton, 2005).
Smad-independent signaling of TGF-β, including the BMP receptors, has been observed in a number of systems and includes the activation of other diverse signaling cascades such as MAPK, PI3K, Rho-like GTPases, and LIMK1. However, only LIMK1 has been associated biochemically with the region of the BMP type II receptor that was found to be specifically required for synaptic stabilization in the Wit receptor. In vitro kinase assays demonstrate that the interaction of LIMK1 with the BMPRII tail leads to changes in LIMK1 kinase activity (Foletta, 2003; Lee-Hoeflich, 2004). Although these studies reached different conclusions about the effects of receptor binding on LIMK1 activity, both studies support a model in which binding of BMP to the receptor complex leads to an increase in LIMK1 activity. It is proposed that ligand binding to the Wit receptor activates DLIMK1 to stabilize the NMJ (Eaton, 2005).
It is remarkable that the overexpression of DLIMK1 rescues all aspects of the wit receptor mutant phenotype, including synaptic growth, synaptic function, and animal viability. DLIMK1 expression is also sufficient to restore synaptic growth and stability to the mad mutation. These data contrast with the genetic analysis demonstrating that DLIMK1 is necessary for synaptic stability but is not required for normal synaptic growth and has only a minor role in functional synapse development. How can DLIMK1 expression restore synaptic growth, function, and viability in the absence of the BMP receptor? Even if DLIMK1 independently signals to the nucleus (Yang, 2004), it seems unlikely that DLIMK1 activity would be sufficient to restore the entire transcriptional program normally mediated by nuclear Smad signaling. Instead, it is proposed that the overexpression of DLIMK1 in the wit receptor mutant background hyperstabilizes the synapse, consistent with synaptic stabilization being the primary function of DLIMK1. If UAS-DLIMK1 hyperstabilizes the synapse, a second, parallel growth factor signaling system may thereby be allowed to assume the growth-promoting functions normally mediated by BMP signaling. This hypothesis invokes the existence of an unidentified second growth factor signaling cascade at the Drosophila NMJ. It seems likely that additional growth factor signaling exists since the synapse still grows to nearly 50% of its normal size in the wit mutant background. One candidate for a second synaptic growth signaling system is the activin signaling system. In the Drosophila central nervous system, ecdysone-regulated axonal remodeling during metamorphosis is regulated, in part, by signaling via the activin type I receptor Babo, the type II BMP receptor Put, and DSmad2. It should be noted, however, that the Put type II receptor lacks the C-terminal DLIMK1 binding region found in the Wit receptor. Therefore, the Put receptor probably does not normally signal via direct interactions with DLIMK1 during the remodeling of the CNS or during stabilization at the NMJ (Eaton, 2005).
In lower motor diseases, such as ALS, two events are thought to contribute significantly to disease progression: (1) the loss of nuclear trophic signaling due to impaired axonal transport, and (2) the loss of access to trophic signal due to synapse retraction. The ability of LIMK1 to stabilize the NMJ in the absence of trophic signaling might suggest a model in which activation of stability-promoting proteins, such as LIMK1, may counteract inappropriate synaptic disassembly during disease. It is interesting to note that LIMK1 has been shown to accumulate within the presynaptic nerve terminal during the late maturation of the mammalian NMJ, after the period of developmental plasticity observed during embryonic and early postnatal development (Wang, 2000). This observation supports an intriguing model in which the maturation of the mammalian NMJ from a highly plastic synapse early in development to a more stable adult synapse involves both the reduction of plasticity-associated proteins, such as CAP-23 and GAP-43, and the enrichment of stability-promoting proteins such as LIMK1. It remains to be determined whether synaptic loss in degenerative disease is directly related to the loss of stabilizing factors such as LIMK1 and, indeed, whether enhanced signaling from these factors could lead to restabilization of diseased synapses. However, the possibility that LIMK1 may function locally at the NMJ to promote synaptic stabilization, even in the absence of an essential source of trophic signaling, suggests a potent activity of LIMK1 in the nervous system that may have important therapeutic value as a future avenue for intervention in neuromuscular degenerative disease (Eaton, 2005).
wit expression pattern was studied using in situ hybridization. In embryos, wit transcription is first observed at stage 12 within a subset of cells in the developing central nervous system (CNS). There is no maternal deposition, consistent with the absence of a requirement for wit expression in the germline. wit remains restricted to the CNS, where it is expressed for the rest of embryonic development. A larger number of cells express wit in the CNS by stage 14, and the message can also be detected in the PNS and in leading edge cells during dorsal closure. After dorsal closure is complete, continued expression of wit is seen in the CNS and PNS as well as in portions of the gut. By stage 17, expression of wit is seen in the prothoracic component of the ring gland (Marqués, 2002).
The number and the positions of the phosphorylated form of Mad (P-Mad)-positive cells, per hemisegment at stage 17, indicative of cells undergoing Wit signaling, correlates well with the estimated number and positions of motoneurons at this stage. To better characterize the cells that accumulate P-Mad in the nervous system, wild-type embryos were double stained for P-Mad and several other markers, including Even-skipped and Lim-3, two transcription factors expressed in discrete subsets of embryonic motoneurons and interneurons. Staining was also performed for Fasciclin II (Fas II), a neural adhesion molecule expressed in a subset of interneurons as well as most motoneurons. There is colocalization of Eve and P-Mad in the RP2 and U/C motoneurons but not in the EL interneurons, while P-Mad and Lim-3 colocalize in a lateral cluster of cells that include motoneurons 28 and 14/30. In addition, Lim-3 and P-Mad colocalize in RPs 1, 3, 4, 5. The Fas II antibody highlights axon tracts and cell bodies of most motoneurons as well as some interneurons. Most, if not all, P-Mad cells also show colocalization of Fas II. These findings suggest that Wit is required to transduce BMP signals in motoneurons and is consistent with the ability of UAS-wit constructs to rescue wit mutants when expressed with a motoneuron-specific driver (Marqués, 2002).
Expression pattern of wit mRNA was examined by in situ hybridization. At early embryonic stages, no wit mRNA was detected, suggesting that wit is not maternally deposited. Zygotic wit expression was detected first at stage 11 in a reiterated subset of CNS neurons and in the labral sensory cluster of the peripheral nervous system. In stages 11 to 14, wit expression persists in these spatially restricted clusters. Co-staining with the monoclonal antibody BP102 that labels the axon scaffold reveals that wit is expressed in the motoneurons of the Ventral-Unpaired Median (VUM) cluster. Co-staining with anti-Even-Skipped antibodies reveals that Wit is expressed by several identifiable motoneurons including the RP2 motoneurons but not by the EL interneurons. From stage 15 to 17, the boundaries of the CNS clusters gradually disappear, as more CNS neurons appear to express wit. At stage 16, wit mRNA is also detectable in the furrows of the developing midgut (Aberle, 2002).
Wit protein is expressed in a restricted subset of CNS neurons as detected with a monoclonal antibody (23C7) generated against Wit. Wit protein is detected in a reiterated pattern in the CNS from stage 13 onward. Beginning at stage 15, Wit is also seen in the CNS axon scaffold in both the commissures and longitudinal tracts. Weak expression is also detected in the sensory neurons of the peripheral nervous system. At stage 16, Wit staining forms a pattern resembling the staining pattern of the motoneuron marker Late Bloomer (Lbl). Similar to Lbl, Wit protein is predominantly found in a stripe of cells in a region close to the posterior commissure in each segment where motoneurons are located (Aberle, 2002).
Individual neurons express only one or a few of the many identified neurotransmitters and neuropeptides, but the molecular mechanisms controlling their selection are poorly understood. In the Drosophila ventral nerve cord (VNC), the six Tv neurons express the neuropeptide gene FMRFamide (FMRFa). Each Tv neuron resides within a neuronal cell group specified by the LIM-homeodomain (LIM-HD) gene apterous (ap). The zinc-finger gene squeeze acts in Tv cells to promote their unique axon pathfinding to a peripheral target. There, the BMP ligand Glass bottom boat activates the Wishful thinking receptor, initiating a retrograde BMP signal in the Tv neuron. This signal acts together with apterous and squeeze to activate FMRFamide expression. Reconstituting this 'code,' by combined BMP activation and apterous/squeeze misexpression, triggers ectopic FMRFamide expression in peptidergic neurons. Thus, an intrinsic transcription factor code integrates with an extrinsic retrograde signal to select a specific neuropeptide identity within peptidergic cells (Allan, 2003).
FMRFa is specifically expressed in the six Tv neuroendocrine neurons located bilaterally in the three thoracic (T1-3) segments of the embryonic and larval VNC. apterous is expressed in three interneurons per VNC hemisegment, as well as in a lateral cluster of four neurons (the ap-cluster) in each of the T1-3 hemisegments. One of the four ap-cluster cells is the FMRFa-expressing Tv neuron. All ap interneurons in the VNC, except for the Tv, join a common ipsilateral axon tract termed the ap-fascicle. The Tv axon instead projects to the midline and exits the VNC dorsally to innervate the dorsal neurohemal organ (DNH). The DNH is a club-like neuroendocrine structure formed by two glial cells protruding from the midline of each thoracic segment. Anteriorly, two additional FMRFa-expressing cells are found, denoted SE2 cells. The SE2 cells do not express, nor depend upon, any regulators described in this study for their FMRFa expression. ap is important for the expression of FMRFa in the Tv neurons, but since most ap neurons do not express FMRFa, other regulators are likely needed for FMRFa regulation (Allan, 2003).
Rotund, a zinc finger protein of the C2H2 Krüppel-type belongs to a conserved subfamily of zinc finger proteins together with Drosophila CG5557, C. elegans Lin-29, and rat CIZ. Squeeze is most closely related to Rotund, with identity greater than 90% throughout the zinc finger region; Squeeze is 78% identical to LIN-29 in the conserved zinc finger region. Both rotund and CG5557 are expressed in subsets of cells in the developing CNS. CG5557 has a larval lethal phase. Mutants eclosed at a low frequency as immotile adults that died within 24 hr. Mutant larvae display a motility defect whereby the body wall musculature over-contract radially during the peristaltic wave typical of insect larval motility, apparent as a 'squeezing' of the intestine. Since this motility phenotype is fully penetrant and scored with 100% accuracy (sqzlacZ/sqzDf), CG5557 was renamed squeeze (sqz) (Allan, 2003).
What is the identity of the retrograde FMRFa-inducing signal? Recently, a Drosophila BMP type-II receptor, wishful thinking (wit), was implicated in mediating a retrograde signal from muscles to motor neurons, responsible for presynaptic maturation. Signaling by the TGF-β/BMP superfamily occurs via activation of a receptor complex, consisting of two type I and two type II receptors, leading to phosphorylation and nuclear translocation of a receptor Smad protein. In Drosophila, BMP signaling leads to the phosphorylation and nuclear translocation of the Smad protein Mothers against dpp (Mad), which can be monitored using antibodies specific to phosphorylated Mad (pMad) (Allan, 2003).
Using antibodies to pMad, BMP activation in peptidergic neurons was assayed. Nuclear pMad was detected not only in motor neurons, but also in the Tv, Va, and Vap neurons, demonstrating that peptidergic neurons projecting out of the VNC also show evidence of BMP activation. Accumulation of pMad in the Tv neurons commences during stage 17, immediately following DNH innervation. These results led to a test of whether Tv innervation of the DNH would be critical for pMad accumulation and consequently for FMRFa expression. Indeed, it was found that the absence of the DNH (in tin mutants), Tv axon pathfinding alterations (in apGAL4/UAS-robo and apGAL4/UAS-racV12) and interference with Tv axonal transport (in apGAL4/UAS-GluedDN and apGAL4/UAS-τ-myc) are all accompanied by loss of pMad staining specifically in Tv neurons. The ectopic ap-cluster FMRFa-expressing cell induced by sqz misexpression is also pMad positive. Given the role of sqz in Tv axon pathfinding, this is interpreted as resulting from sqz dominantly altering the projection of one other ap-cluster cell, forcing it to innervate the DNH. Thus, in all genotypes examined, Tv axonal projection to the DNH is critical for pMad accumulation (Allan, 2003).
Since Wit is expressed in a restricted pattern in the developing VNC, attempts were made to address whether the Tv neurons express Wit. However, single-cell resolution could not be obtained with the Wit antibody and Wit could not be definitely localized in Tv cells. However, the wit-dependent pMad accumulation in Tv neurons, the apGAL4/UAS-tkvA, UAS-saxA-mediated rescue of wit mutants, and the UAS-gbb-mediated 'rescue' of UAS-robo misexpression, provide genetic evidence supporting the expression of wit in Tv cells. Previous studies have shown that gbb is expressed in developing endoderm and visceral mesoderm, but it has not been detected in the VNC. By in situ hybridization, no apparent expression was detected in the DNH. Given that the DNH only contains two cell bodies, low-level gbb expression may be beyond detection. Moreover, since the anterior midgut is positioned in very close proximity to the DNHs, it is possible that Gbb diffuses from the visceral mesoderm to the DNH (Allan, 2003).
Amidated neuropeptides of the FMRFamide class regulate numerous physiological processes including synaptic efficacy at the Drosophila neuromuscular junction (NMJ). Mutations in wishful thinking (wit), a gene encoding a Drosophila Bmp type 2 receptor that is required for proper neurotransmitter release at the neuromuscular junction, also eliminates expression of FMRFa in that subset of neuroendocrine cells (Tv neurons) which provide the systemic supply of FMRFa peptides. Gbb, a Bmp ligand expressed in the segmentally repeated neurohemal organ associated with the ventral cord, provides a retrograde signal that helps specify the peptidergic phenotype of the Tv neurons. Supplying FMRFa in neurosecretory cells partially rescues the wit lethal phenotype without rescuing the primary morphological or electrophysiological defects of wit mutants. It is proposed that Wit and Gbb globally regulate NMJ function by controlling both the growth and transmitter release properties of the synapse as well as the expression of systemic modulators of NMJ synaptic activity (Marqués, 2003).
wishful thinking is primarily expressed in, and required for, proper nervous system function. Mutations in wit result in pharate lethality caused, in part, by defects in the growth and physiology of motoneuron synapses. Mutations in wit also affect the peptidergic phenotype of certain FMRFa-expressing cells found in the ventral cord. In particular, FMRFa expression is eliminated in the Tv neurons that contribute to the systemic supply of FMRFa peptides through release at the neurohemal organ. The regulation of FMRFa expression in Tv neurons is mediated by the Bmp ligand Gbb, since gbb null mutations also eliminate FMRFa expression in Tv neurons. Furthermore, supplying Gbb to the dorsal neurohemal cells restores FMRFa expression in Tv neurons. Since Tv neuron axons arborize onto the neurohemal cells, this strongly suggests that Gbb signals in a retrograde manner to specify the peptidergic phenotype of Tv neurons. Consistent with this view, overexpression in neuroendocrine cells of Dynamitin or a dominant-negative form of p150/Glued, both components of the Dynactin/Dynein motor complex was found to eliminate FMRFa expression in the Tv neurons. Finally, it is shown that providing FMRFa in neuroendocrine cells using the Gal4/UAS system partially rescues the lethal phenotype of wit mutants, even though they still exhibit structural and physiological synaptic defects. It is suggested that Bmp signaling provides a global cue that not only regulates the growth of the NMJ synapses locally but also controls their systemic modulation by the neuroendocrine system (Marqués, 2003).
The Drosophila genome contains seven TGFß type ligands. Three of these, Dpp, Screw and Gbb, have been shown to transduce Bmp-type signals (Mad) and to use the type I receptors Tkv and Sax. Two others, Activin and Activin-like protein, transduce signals through Smad2. The signaling pathways used by Maverick and Myoglianin remain untested. Among the three Bmp-type ligands, Gbb seemed a likely candidate for controlling expression of FMRFa, since it is broadly expressed, at least in embryos, and can signal through Wit to regulate P-Mad accumulation in motoneurons and tissue culture cells. gbb is strongly expressed in the larval brain lobes and much more weakly in the ventral ganglia. Interestingly, gbb shows enriched expression in the NHO relative to other ventral ganglia neurons. Thus, Gbb is expressed in the correct place to be a FMRFa regulating ligand (Marqués, 2003).
In Drosophila, FMRFamide peptides have been shown to enhance synaptic transmission and muscle twitch tension when perfused onto standard larval nerve-muscle preparations; however, their in vivo role(s) are not known as no mutations in the FMRFa gene have been identified. As with most neuropeptides, FMRFamide related peptides are thought to act as neuromodulators and neurohormones. The Tv-produced FMRFamide related peptides are released into the hemolymph through the neurohemal organ and hence are able to act on every tissue in the animal that is not blocked to hemalymph contact. It has been hypothesized that the lethality of wit mutants is due to the lack of proper synaptic transmission at the NMJ, resulting in the animals not being able to eclose from the pupal case. The lack of systemic FMRFamide described in this study would be expected to further decrease synaptic efficiency and the ability of wit mutants to eclose. The fact that loss of FMRFa does contribute to the lethal phenotype is supported by the partial rescue of wit mutants by overexpression of FMRFa. These results are consistent with the view that in vivo, FMRFa peptides probably enhance NMJ synaptic activity similar to their in vitro documented effects on standard larval electrophysiological preparations (Marqués, 2003).
It is important to note that although the lethal phenotype is partially reversed, the morphological and physiological synaptic defects reported for wit mutants are not rescued by overexpression of FMRFa. The simplest interpretation is that the excess of FMRFamide related peptides enhances the efficiency of wit mutant synapses in vivo without correcting the underlying developmental defects. Although one might expect a significant improvement of the electrophysiological phenotype, this is not detected, probably because the excess FMRFamide related peptides are either washed off the preparation during standard dissection prior to recording or act for only short periods (Marqués, 2003).
How Wit signaling regulates FMRFa expression is not clear. Since Smads are well known to act as transcriptional co-activators or co-repressors, the simplest explanation is that Mad directly regulates activation of FMRFa transcription, perhaps by forming a complex with Ap. However, other indirect mechanisms are also possible and this issue will only be resolved once the FMRFa promoter is fully characterized. It is also not clear whether Gbb is the only ligand that regulates FMRFa expression through Wit. In some developmental contexts, such as wing imaginal disc patterning, Gbb acts in combination with Dpp, another Bmp-type ligand. No expression of dpp has been detected in the NHO. However, it could be that one of the as yet uncharacterized ligands, Maverick or Myoglianin, could be a partner with Gbb in regulating FMRFa expression. Conversely, it seems clear that regulating the peptidergic phenotype of the six Tv neurons is not the only role of Gbb signaling. There are hundreds of neurons that receive Bmp signaling as indicated by P-Mad nuclear localization Most of them appear to be motoneurons, which require Wit/Gbb signaling to achieve proper synaptic growth but not to specify their neurotransmitter phenotype. Given that Smads act as co-transcriptional regulators, the fact that the same signal (nuclear translocation of P-Mad) results in different phenotypic outcomes in different neurons can probably be ascribed to the presence of a different set of transcription factors available in each cell type. The Tv neurons receiving the Bmp signal express apterous, a transcription factor required in those cells for FMRFa transcription, and maybe other factors that are required, in addition to the Wit signal, to activate FMRFa (Marqués, 2003).
Another important issue to resolve is whether Gbb is constitutively released from the NHO, or is synthesized and released as part of a feedback mechanism to modulate muscle contractions. It might be that efficient muscle contraction under normal conditions requires a constant level of FMRFamide related peptides that are produced in response to a constitutive Gbb signal. Alternatively, Gbb production or release might be regulated by a sensing mechanism that would activate the pathway in response to an increased demand for FMRFamide related peptides, owing to increased locomotor activity or other stimuli, such as compensating for a synaptic developmental defect. Muscle-derived Gbb acts through neuronal Wit to convey a retrograde signal essential for NMJ synapse growth and maturation. In that context, it appears that the role of Bmp signaling is to coordinate muscle growth with synapse maturation to ensure proper synaptic efficiency. Thus, the Wit/Gbb pathway acts as a two-step regulator of NMJ function. First, there is a developmental role in which Wit signaling is required for proper synaptic growth during larval development. Second, Wit signaling is required to achieve the neuromodulatory effect of circulating FMRFamide related peptides that are required for optimal synaptic transmission. Lack of either one of these inputs probably results in a substantial decrease of the EJCs. These two examples suggest that the Gbb/Wit pathway is of general importance in neural retrograde signaling and it is speculated that it may be used in the nervous system for other as yet uncharacterized developmental and physiological purposes (Marqués, 2003).
Tv neurosecretory cells form part of a cluster of four apterous-expressing neurons on each side of the three thoracic ganglia. The axons of the Tv neurons extend proximally and dorsally to join the contralateral axon, and form a median nerve that swells and arborizes onto a group of neurons and glial cells that constitute the neurohemal organ. In wit mutants, these structures develop normally, but the Tv neuron fail to activate FMRFa transcription. Using the Gal4/UAS system Wit's requirement for FMRFa expression was narrowed down to the Tv neurons. Since these neurons accumulate nuclear P-Mad, the results strongly suggest that Wit is required in the Tv neurons themselves, as opposed to forming part of an indirect signal relay mechanism. It appears likely that the source of Gbb in this signaling system is the NHO, since gbb is expressed in the NHO and replenishing Gbb in the NHO of gbb mutants rescues FMRFa expression in the Tv neurons. These experiments do not exclude the possibility that signaling might occur at the cell soma of the Tv neurons in vivo or that the source of the diffusible ligand could be a different tissue under physiological conditions. However, the dependence of nuclear P-Mad accumulation and FMRFa expression in Tv neurons on Dynein-mediated retrograde transport strongly suggests that signaling is taking place at the Tv axon terminal. This dependency on Dynein motors is not a general requirement for FMRFa expression in all neurons because subesophageal ganglion neurons are not affected by overexpression of dominant-negative Glued or Dynamitin. Nor is the consequence of disrupting this motor likely to exert its effect at the level of P-Mad translocation to the nucleus, since nuclear accumulation of P-Mad in epithelial and mesodermal cells is not effected by retrograde transport disruption. Only in the nervous system is P-Mad accumulation specifically affected, consistent with a role for a retrograde transport mechanism in moving some component of this signaling pathway from the synapse to the nucleus (Marqués, 2003).
In larvae, wit is expressed strongly in the brain lobes and more weakly in the ventral ganglia. All imaginal discs also express wit, however, with the exception of the wing disc, without a prominent spatial pattern. In the wing disc , abundant transcription is seen in the wing pouch and also a broad band of cells along the anterior-posterior margin, particularly in the presumptive notum (Marqués, 2002).
In first and second instar larvae, Wit is expressed throughout the brain. In third instar larvae, Wit expression in the ventral nerve chord ceases but is still prominent in the synaptic centers and the lamina of the developing brain lobes. Wit protein is also detected in all imaginal discs. Endogenous Wit could not be detected at the larval NMJ using the 23C7 antibody; however, when Wit was overexpressed in all neurons, it was readily detected at the NMJ with this antibody. This analysis of the expression pattern demonstrates that wit is expressed specifically in the nervous system from embryonic stages until third instar larvae, indicating that Wit signaling is continuously required during these developmental periods (Aberle, 2002).
Metamorphosis of the Drosophila brain involves pruning of many larval-specific dendrites and axons followed by outgrowth of adult-specific processes. From a genetic mosaic screen, two independent mutations were recovered that block neuronal remodeling in the mushroom bodies (MBs). These phenotypically indistinguishable mutations affect Baboon function, a Drosophila TGF-ß/activin type I receptor, and Smad on X (Smox, or dSmad2), its downstream transcriptional effector. Punt and Wit, two type II receptors, act redundantly in this process. In addition, knocking out Activin-beta (dActivin) around the mid-third instar stage interferes with remodeling. Binding of the insect steroid hormone ecdysone to distinct Ecdysone receptor isoforms induces different metamorphic responses in various larval tissues. Interestingly, expression of the Ecdysone receptor B1 isoform (EcR-B1) is reduced in activin pathway mutants, and restoring EcR-B1 expression significantly rescues remodeling defects. It is concluded that the Drosophila Activin signaling pathway mediates neuronal remodeling in part by regulating EcR-B1 expression (Zheng, 2003).
In Drosophila, recent data suggest that a BMP signaling pathway controls synaptic growth and function at the neuromuscular junction (NMJ). Whether this pathway also contributes to activity-dependent remodeling at the NMJ remains to be determined. It is interesting to note, however, that in this pathway Wit acts as a BMP receptor, and it can not be substituted for by Punt. In contrast, the activin pathway described here appears to be able to utilize either Punt or Wit for signaling. This may reflect selectivity in the binding of some ligands to one receptor, but not the other. Additional studies will be required to resolve this issue. Since many components of several different TGF-β signaling pathways show pronounced expression in different parts of the developing and postnatal rodent brain, the demonstration that TGF-β/Activin signaling cell-autonomously controls plasticity of MB neurons may provide novel insights into how neuroplasticity is dynamically regulated in higher organisms (Zheng, 2003).
To screen for structural alterations of the neuromuscular junction (NMJ), a transmembrane GFP-fusion protein (CD8-GFP-Sh) was used that specifically localizes to the NMJs of Drosophila larvae. All somatic muscles and their neuromuscular connections can be observed through the translucent cuticle of an intact animal from first to late third instar larvae using confocal microscopy. In larvae, the somatic musculature of each hemisegment is innervated by a major nerve bundle, which defasciculates into five major nerve branches. Each branch forms synapses on a subset of muscles: for example, the ISN branch innervates dorsal and dorso-lateral muscles, while the SNa branch innervates exclusively lateral muscles. The dorsal and lateral muscles are most easily visualized in CD8-GFP-Sh larvae and have therefore been the focus of this screen (Aberle, 2002).
On the third chromosome, third instar larvae of 4973 homozygous Drosophila lines carrying EMS-induced mutations for morphological alterations of the NMJ were screened. 66 lines were saved that fell into 17 complementation groups comprising 10 phenotypical classes. Among the most prominent phenotypes, mutations were found that lead to smaller and more expanded synapses. Mutations were also recovered that affect muscle innervation, muscle morphology, and localization of the CD8-GFP-Sh transgene (Aberle, 2002).
Five mutant alleles of one complementation group, named wishful thinking (wit) (Marqués, 2002), exhibit particularly small NMJs, while muscle size remains close to normal. All neuromuscular synapses are smaller in wit mutants with nearly 100% penetrance. The size of individual synaptic boutons appears normal, but boutons are often spaced further apart from one another. The specificity of the neuromuscular connections is not affected, and no other gross morphological defects were observed in wit mutant larvae. In addition, the size of sensory neurons appears normal and the dendritic tree of multidendritic neurons is not visibly reduced in size in the mutants. wit larvae pupate at the same time as their heterozygous siblings and die as pharate adults shortly before eclosion. Thus, mutations in wit seem to primarily affect the structure of the NMJ (Aberle, 2002).
Since the synapses are drastically reduced in size in wit mutants, it was of interest to determine if there were any changes in the subcellular distribution of synaptic proteins. Discs Large (Dlg), a synaptic clustering protein, and myc-tagged glutamate receptor IIB (GluRIIB), a subunit of postsynaptic GluRs, appear to localize correctly. In contrast, alterations were found in the synaptic level of the cell adhesion molecule Fasciclin II (Fas II) and in the distribution pattern of the synaptic vesicle-associated protein Synaptotagmin (Syt). In wild-type larvae, Fas II is detected on all motor axons and at higher levels on both the pre- and post-synaptic side of synaptic boutons. wit mutant synapses generally exhibit a greatly reduced synaptic staining for Fas II. Interestingly, Fas II appears to be selectively absent from the synaptic terminals but remains at normal levels on the motor axon bundles. In wild-type, the synaptic vesicle marker Syt localizes exclusively to the presynaptic terminal and is not detected normally in motor axons. In wit mutants, however, Syt staining is typically found in a punctate pattern in the nerve bundles (Aberle, 2002).
The structural growth defects in wit mutants were quantified by counting the synaptic boutons. The bouton numbers on dorsal muscle 1 are approximately 55% reduced in wit mutants. The muscle surface area is on average 12% smaller. Similar results were obtained for muscles 2, 9, and 10. Neuromuscular synapses were examined in wit mutant embryos and at each larval stage. While neuromuscular synapses in wit mutant stage 16 embryos appear similar to those of wild-type, those of first instar wit larva have visibly smaller synapses compared to wild-type, as do those of second instar wit mutants (Aberle, 2002).
The wit gene was sequenced of flies bearing five different mutant alleles. Point mutations were found in the coding region of four alleles, all of which were located in the kinase domain. Stop codons were identified in three wit alleles (K482Stop in witHA4, W491Stop in witHA3, and R524Stop in witHA5). These three alleles and witHA2 have similar phenotypes either transheterozygous over each other or hemizygous over deficiencies, suggesting they represent strong loss-of-function alleles. The witHA1 allele appears to be a hypomorph since the synaptic phenotype is slightly less severe. Interestingly, a missense mutation was found in witHA1 that changes a glutamate residue to lysine within the coiled-coil/leucine zipper domain at position 413 (E413K). These results show that synaptic undergrowth in wit mutants is caused by single point mutations in the Drosophila BMPRII homolog Wit (Aberle, 2002).
Additional evidence for a presynaptic function for Wit was provided by the examination of the effect of a dominant negative wit construct lacking the cytoplasmic domain of the receptor protein (witDeltaC). Overexpression of WitDeltaC in motoneurons using OK6-Gal4 reduces synaptic size in a wild-type background, whereas overexpression of the same protein in muscles using G14-Gal4 has no effect. Finally, whether overexpression of Wit is sufficient to promote synaptic growth in a wild-type background was tested. When Wit is overexpressed in all neurons or in all somatic muscles, no synaptic overgrowth is seen at the NMJ, even when the copy number is increased 2-fold. In summary, these results suggest that wit function is required presynaptically for viability and normal synaptic development at the NMJ (Aberle, 2002).
Since wit is required for normal synaptic development at the NMJ, it was of interest to see whether wit also plays a role in synaptic transmission. Both the spontaneous and the evoked release were examined using intracellular recordings from muscle 6 at segment A3 in wild-type and wit mutant third instar larvae. Evoked excitatory junctional potentials (EJP) are greatly reduced in wit mutants reaching only 10% of those of the wild-type. The amplitude of spontaneous miniature excitatory junctional potentials (mEJPs) is also slightly reduced in wit mutants. To further examine the effect of wit loss of function on neurotransmitter release, the frequency of spontaneous release was measured in the absence of stimulation. This measure of release is also significantly reduced in the mutants. Quantal content was measured and it was found that wit mutant synapses release 7-fold less neurotransmitter than wild-type synapses. Several allelic combinations of wit mutants gave similar results. These results suggest that Wit is required for normal synaptic function. All defects of neurotransmission were rescued when Wit was overexpressed either pan-neuronally or in motoneurons with elav-Gal4 or OK6-Gal4, respectively. This indicates that wit loss of function disrupts a presynaptic process, consistent with the genetic results for rescuing lethality and structural defects (Aberle, 2002).
Finally, the electrophysiological defects in wit mutants were examined by measuring the Ca2+ dependence of neurotransmitter release. Loss of function of wit does not appear to affect the Ca2+ cooperativity at the neuromuscular junction. The electrophysiological examination implies that upon depolarization, wit mutant motor terminals release significantly less neurotransmitter (Aberle, 2002).
Given the dramatic light microscopic and electrophysiological defects, it was of interest to see whether the ultrastructure of wit neuromuscular synapses was also affected. Serial cross-sections of both wild-type and wit mutant boutons at muscles 6 and 7 of third instar larvae were examined by electron microscopy. It was found that wit mutant synapses have active zones with T-bars, synaptic vesicles both free within the bouton and clustered around the T-bars, and other features that appear wild-type. In addition to these features, also observed were a smaller population of vesicles larger than and distinct from the synaptic vesicles, an unusual ruffling of the presynaptic membrane in active zones, and a novel electron dense structure associated with a tight cluster of vesicles. The postsynaptic specialization of the muscle membrane, the subsynaptic reticulum (SSR), was indistinguishable from wild-type (Aberle, 2002).
A number of bouton characteristics was quantified from serial sections of 17 wit boutons and they were compared with similar data from 18 wild-type boutons. Although wit and wild-type boutons are approximately the same size, the bouton surface area per active zone is 2-fold greater in wit mutants. In addition, the number of T-bars per active zone was 2-fold greater in wit mutants. The bouton surface area per T-bar (an indicator of an active vesicle release site) was not statistically different between wit mutants and wild-type (Aberle, 2002).
The ultrastructural observations reveal two significant defects in wit mutant boutons. In wit mutants, the postsynaptic membrane and the electron dense material in the cleft appear normal, but the presynaptic membrane within the active zone displays regions of ruffling or detachment from the postsynaptic membrane/cleft material. These membrane detachments do not include the entire active zone since some regions maintain close apposition of the pre- and post-synaptic membranes. This adhesion defect is observed in nearly all active zones of wit mutants and is similar in severity in sections with, and without, T-bars. This defect was quantified by comparing the ratio of the length of presynaptic membrane in close apposition to the postsynaptic membrane to the overall length of postsynaptic membrane in active zones of wild-type and wit mutants. In wild-type boutons, this ratio is 0.97 ± 0.01. By contrast, this ratio in wit mutants is 0.56 ± 0.03, indicating that 44% of the presynaptic membrane is detached. In no instance, however, was a T-bar observed associated with a detached section of presynaptic membrane, even when the immediately adjacent presynaptic membrane was detached (Aberle, 2002).
The second defect in wit mutant boutons is the presence of a novel electron dense structure and its associated cluster of vesicles within the cytoplasm of the bouton. An average of 2.4 ± 0.7 such structures per bouton has been observed in wit mutants, whereas none were observed in wild-type boutons. These structures were named T-bodies, due to their shared ability to cluster synaptic vesicles in a similar fashion to T-bars. Unlike T-bars, T-bodies appear to float freely within the bouton with no attachment to the presynaptic membrane. T-bodies often spanned 2-3 adjacent 80 nm serial sections, had a 'C' or closed ring shape in 3D reconstruction, and did not resemble conventional T-bars seen in either cross-section or tangential section. Their association with tight clusters of synaptic vesicles suggests, however, that they contain vesicle-docking components similar, if not identical, to those in T-bars. The wit loss of function therefore appears to disrupt the proper adhesion of the presynaptic membrane to the postsynaptic membrane/cleft material within active zones, which could underlie the defects observed in synaptic function (Aberle, 2002).
The 64A region had been screened previously by EMS mutagenesis for the presence of lethal complementation groups. Transgenic lines that contain the entire wit transcription unit, as well as 1.8 kbp or more of upstream sequence, rescue the l(3)64Aa complementation group, while P lines that do not completely cover wit do not rescue this group of mutants. To confirm that l(3)64Aa does indeed contain lesions within wit, genomic DNA from five l(3)64Aa alleles was sequenced. In each case, a molecular lesion was identified in the wit coding sequence. Since all alleles of l(3)64Aa contain mutations in the wit transcription unit, they will henceforth be referred to as wit alleles. Among the various lesions isolated, witA12 likely represents a simple null allele since the C684T change introduces an early stop codon prior to the transmembrane segment. The proximal breakpoint of the small deficiency Df (3L) C175 was localized to the 3' portion of the wit transcription unit. Interestingly, no mutations were found in the carboxy-terminal extension of Wit, despite the fact that this region represents over 40% of the coding sequence and that mutations in the tail of human BMPRII account for some cases of primary pulmonary hypertension (Marqués, 2002).
To determine the wit lethal phase, the number of wit mutant embryos that survived to the first instar, third instar, and pharate stage was quantitated. All transheterozygous combinations of wit alleles are 100% lethal, with no adult escapees. Approximately 50% of the mutant animals survive to the pharate stage. The remaining animals die during larval and early pupal stages. The animals that do survive to the pharate stage exhibit the head expansion and abdominal peristaltic movements characteristic of normal pre-eclosion behavior, but neither rupture nor escape from the pupal case. Removing the operculum and underlying pupal cuticle is not sufficient to allow the animals to eclose. If the animals are dissected free of the pupal case and cuticle, they remain alive for several days but, for the most part, exhibit only sporadic leg movements. None of these animals inflate their wings. However, when the wings are artificially inflated by immersion in water, they show no patterning defects. Many of the dissected animals showed a bend in the femur or tibia, the phenotype being more pronounced and frequent in the metathoracic leg. A small fraction of the animals also exhibited truncations of the tarsal segments. These defects are unlikely to be the cause of the inability to eclose since the penetrance of such defects is incomplete, and animals rescued to viability by expression of Wit in the nervous system still showed leg defects yet were able to escape the pupal case (Marqués, 2002).
The general morphology of the wit pharates was analyzed by silver staining of 10 µm paraffin sections. General organization of neural centers and muscle groups is essentially normal, and neither misrouted axonal tracts nor fasciculation defects were found (Marqués, 2002).
Germline clones were generated in an attempt to uncover earlier phenotypes. Maternally depleted animals derived from null alleles showed the same lethal phase and range of defects as those found in the zygotic mutants, indicating that wit is not required in the germline or in the early embryo (Marqués, 2002).
The inability of wit mutants to eclose, together with the locomotion defects exhibited by dissected animals and the strong neural expression of wit, indicates that Wit's function is likely to be required in neurons. To determine if the nervous system is the primary tissue requiring wit activity, wit cDNA was expressed using a number of different promoters as well as with the conditional Gal4-UAS expression system. Several independent lines of both Hsp70-wit and ubiquitin-wit were able to rescue different wit allelic combinations indicating that low level, ubiquitous expression of wit is sufficient to ensure its function. Using the Gal4 system, it was found that drivers such as nervana and elav that are expressed in most or all differentiated neurons completely rescue wit mutant animals to full viability. Other drivers such as twist (pan mesoderm) or Mhc (myosin heavy chain promoter, muscles) showed no rescuing ability (Marqués, 2002).
Since wit mutants exhibit defects in locomotion, whether more limited expression of wit in motor neurons could rescue wit lethality was examined. Viability is restored to wit mutants when two different Gal4 drivers, BG380 and OK6 are used: their expression in larvae is primarily restricted to motor neurons. These results demonstrate that Wit activity is primarily required in differentiated motoneurons (Marqués, 2002).
Since mutations in the tail of the human BMPR type II receptor have been isolated that affect its function (Deng, 2000; Lane, 2000; Thomson, 2000), whether the carboxy-terminal extension of Wit is also required for viability was examined. To address this issue, a genomic rescue construct was generated in which the tail was truncated from 351 to 88 aa. This construct was able to rescue wit mutant animals to full viability. Using the Gal4-UAS system, a truncated form of Wit that had only 24 aa of the carboxy-terminal tail was expressed in the nervous system and similar results were observed. It is concluded that the carboxy-terminal extension is not required for viability, possibly accounting for the absence of lethal mutations mapping to this part of the protein (Marqués, 2002).
Since in many tissues Punt appears to be the primary type II receptor for mediating BMP signals, whether overexpression of Punt in neurons would rescue wit mutants was examined. Expression of UAS-punt using elav-Gal4 did not rescue wit mutant animals nor did ubiquitous expression of wit rescue punt mutations, suggesting that these two receptors are not functionally interchangeable for mediating TGF-ß type signals during development (Marqués, 2002).
Since mutations in the C. elegans unc-129 gene, which encodes a BMP type ligand, have been shown to exhibit axon guidance defects (Colavita, 1998), the pattern of axon tracts was examined in wit mutant embryos using the 22C10, BP102, and Fas II antibodies. No evidence was found that wit mutant embryos have defects in axon guidance, and the synapses of all larval motoneurons appear to be established with their proper targets (Marqués, 2002).
The lack of any evidence for cell fate or axon guidance problems within the nervous system of wit mutants prompted an examination of these mutants for alterations in their electrophysiological properties. Since specific expression of wit in motoneurons is able to rescue wit mutant animals and lack of wit function results in specific loss of P-Mad accumulation in embryonic motoneurons, attention was focused on analysis of glutamatergic motoneuron synapses in third instar larva NMJs. wit mutants show a dramatic decrease in transmitter release without any change in the average quantal size. The reduction in evoked junctional synaptic currents (EJCs) in wit mutants is significantly rescued by expression of Wit in neurons using elav-Gal4. This observation is consistent with the finding that Wit is required in nerve cells and not in muscle. A proportional decrease in quantal content was observed at multiple calcium concentrations, indicating that there is not a substantial alteration in the Ca2+ cooperativity of transmitter release in the mutant. In addition, failure analysis demonstrates that most nerve stimuli fail to evoke transmitter release at low extracellular [Ca2+] in wit mutants compared to wild-type and rescued wit larvae. Consistent with a high rate of failures, the average frequency of spontaneous vesicle release was dramatically reduced in mutants by nearly 4-fold compared to control animals. The amplitude of spontaneous miniature excitatory junction potentials (mEJPs), however, was not significantly altered in wit mutants. Taken together, these results indicate that wit mutants are defective for vesicle release from the presynaptic terminal (Marqués, 2002).
To determine if the reduced transmitter release observed in wit mutants is caused by changes in synapse morphology and/or ultrastructure, overall synapse size, bouton number, and active zone morphology were examined at several different muscles of segment A3 in third instar larvae. The number of synaptic boutons in ventral longitudinal muscles 6 and 7 was 57 ± 18 for wild-type and 31 ± 9 for wit mutants (p < 0.0001). Since it was noticed that wit mutant larvae are slightly smaller than wild-type animals, the decrease in synapse size must be corrected to account for a 25% decrease in ventral muscle volume. After correcting for decreased muscle size, the number of synaptic boutons in wit mutant animals is about 70% of wild-type. The synapse length in dorsal longitudinal muscle 9 was 88 ± 28 µm (all data expressed as mean ± standard deviation) for wild-type and 44 ± 23 µm for wit mutants (Marqués, 2002).
Although synapse size and bouton number are reduced in wit mutants, the overall percentage reduction was not proportionately as severe as the reduction in EJC magnitude. Therefore, the ultrastructure of the synapses was examined by serial section electron microscopy (EM). In wild-type animals, sites of presumptive transmitter release (active zones) are characterized by tight apposition of muscle and membrane at a number of discrete sites in each bouton and the presence of a specialized structure referred to as T-bar. Bouton size and the overall synaptic structure, including the subsynaptic reticulum, number of synaptic vesicles, and vesicle docking at T-bars, appears normal in wit mutants. However, it was noted that there is a dramatic alteration in the morphology of the active zone. In wit mutants, the pre- and post-synaptic membranes in certain areas are no longer closely aligned. The presynaptic membrane detaches completely from the postsynaptic membrane, forming large curvatures bending toward the cytosol of nerve terminals. Such membrane detachments also occur in areas near the active zones. Natural membrane curvatures are also observed in wild-type controls. However, presynaptic membranes do not separate from postsynaptic membranes. This ultrastructural defect in the active zone morphology has not been previously described and is consistent with the altered transmitter release properties exhibited by wit mutants, and suggests that BMP signaling plays an important role in synapse maturation or maintenance during development (Marqués, 2002).
spinster (spin), which encodes a multipass transmembrane protein, has been identified in a genetic screen for genes that control synapse development. spin mutant synapses reveal a 200% increase in bouton number and a deficit in presynaptic release. spin is expressed in both nerve and muscle and is required both pre- and postsynaptically for normal synaptic growth. Spin has been localized to a late endosomal compartment and evidence is presented for altered endosomal/lysosomal function in spin mutants. Evidence is presented that synaptic overgrowth in spin is caused by enhanced/misregulated TGF-ß signaling. TGF-ß receptor mutants show dose-dependent suppression of synaptic overgrowth in spin. Furthermore, mutations in Dad, an inhibitory Smad, cause synapse overgrowth. A model is presented for synaptic growth control with implications for the etiology of lysosomal storage and neurodegenerative disease (Sweeney, 2002).
To determine whether synaptic overgrowth in spin is caused by enhanced TGF-ß signaling, it was asked whether TGF-ß signaling is necessary for synaptic overgrowth in spin. The type II receptor mutation wishful thinking causes a severe decrease in bouton number at the NMJ. Type I TGF-ß receptors are known to function in concert with type II receptors, and the type I receptors tkv and sax participate in synaptic growth regulation in this system. Third instar larva mutant for sax or tkv have smaller neuromuscular synapses. This study confirms that there is a significant decrease in bouton number in wit, and that there is a similar decrease in bouton number in both tkv and sax. These receptors are shown to function in the larval motoneurons by demonstrating that pMAD staining in the cell bodies of larval motoneurons requires wit or sax. In this experiment, the larval CNS was costained with pMAD and anti-evenskipped, which labels a subset of motoneurons (Sweeney, 2002).
A genetic analysis of the TGF-ß receptor mutations wit, tkv, and sax in combination with spin demonstrates that TGF-ß signaling is necessary for synaptic overgrowth in spin. Heterozygous mutations in tkv, sax, and wit do not alter synaptic bouton numbers at the NMJ. Heterozygous mutations in tkv, sax, and wit suppress synaptic overgrowth when placed in the spin mutant background. Bouton numbers are significantly reduced in each case where a single copy of a receptor is mutated in combination with spin. Bouton numbers were quantified in each of the double mutant combinations of tkv, sax, or wit with spin. In each case, when both copies of a receptor were removed, synaptic overgrowth was suppressed in the spin mutant background further than when only a single copy of a receptor was mutated. These data demonstrate that TGF-ß receptor mutations suppress synaptic overgrowth in spin in a dose-dependent manner. Furthermore, since bouton numbers return to wild-type, or below wild-type levels, it demonstrates that TGF-ß signaling is necessary for synaptic overgrowth in spin. Taken together with the increase in bouton numbers seen in dad, these data support the conclusion that enhanced or misregulated TGF-ß signaling is a major determinant of synaptic overgrowth in spin. It is hypothesized that altered endosomal function due to loss of Spin causes enhanced TGF-ß signaling and subsequent synaptic overgrowth. Future experiments will be necessary to determine whether enhanced signaling is due to increased receptor number at the plasma membrane, or an inability to stop signaling within the late endosomal system (Sweeney, 2002).
The BMP ortholog Gbb can signal by a retrograde mechanism to regulate synapse growth of the Drosophila neuromuscular junction (NMJ). gbb mutants have a reduced NMJ synapse size, decreased neurotransmitter release, and aberrant presynaptic ultrastructure. These defects are similar to those observed in mutants of BMP receptors and Smad transcription factors. However, whereas these BMP receptors and signaling components are required in the presynaptic motoneuron, Gbb expression is required in large part in postsynaptic muscles; gbb expression in muscle rescues key aspects of the gbb mutant phenotype. Consistent with this notion, blocking retrograde axonal transport by overexpression of dominant-negative p150/Glued in neurons inhibits BMP signaling in motoneurons. These experiments reveal that a muscle-derived BMP retrograde signal participates in coordinating neuromuscular synapse development and growth (McCabe, 2003).
Of particular relevance to the involvement of endosomes in mediating BMP signaling at synapses is the recent observation that mutations in the spinster (spin) gene greatly enhance synaptic growth. Spin is a putative multipass transmembrane protein localized to the late endosomes/lysosomes. Mutations in spin disrupt the morphology of late endosomes and lead to a 2-fold overgrowth in bouton number at the NMJ. This overgrowth can be suppressed by mutations in wit, suggesting that elimination of Spin leads to enhanced BMP signaling. Although the molecular mechanism responsible for the increased signal is unclear, the combined data support the idea that BMP signaling is likely modulated by intracellular trafficking of various signaling components at the level of the endosome. If these endosomes are capable of binding to dynein motors, then, like the Trk containing endosomes, they may also be transported along microtubules back to the cell body. Other models, such as wave activation of receptors, may also be possible, although if this is the case, then it is not clear why overexpression of ΔGl should interfere with BMP signaling as described above (McCabe, 2003).
Although loss-of-function and rescue experiments clearly demonstrate that Gbb is required for proper synaptic development at the NMJ, it is not certain that it is the only TGF-β-type ligand or indeed the primary ligand that regulates this process. The electrophysiological and ultrastructure defects observed in gbb mutant synapses are not as severe as those found in wit null mutants. This could simply reflect an inability to produce true null animals that survive to the third instar stage, or it may indicate that another ligand also provides a signal. In support of this view is the observation that P-Mad accumulation is not totally eliminated in gbb1/gbb2 null mutant embryos as it is in wit mutants. In addition, it is noted that while overexpression of Gbb in the CNS only weakly rescued P-Mad accumulation in the CNS, the pattern of accumulation does not appear to change. That is, P-Mad still seems to be found primarily in motoneurons. Thus, other neurons do not appear to be competent to respond to BMP-type ligands, perhaps because a specific cosignal is absent or because they do not express the right combination of receptors. It is interesting to note that in several other developmental contexts in Drosophila, it appears that at least two BMP ligands provide regulatory inputs into a common process (McCabe, 2003).
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).
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).
Mutations in the human dystrophin gene cause the Duchenne and Becker muscular dystrophies. The Dystrophin protein provides a structural link between the muscle cytoskeleton and extracellular matrix to maintain muscle integrity. Recently, Dystrophin has also been found to act as a scaffold for several signaling molecules, but the roles of dystrophin-mediated signaling pathways remain unknown. To further an understanding of this aspect of the function of dystrophin, Drosophila mutants that lack the large dystrophin isoforms were generated and their role in synapse function at the neuromuscular junction was analyzed. In expression and rescue studies, lack of the large dystrophin isoforms in the postsynaptic muscle cell were shown to lead to elevated evoked neurotransmitter release from the presynaptic apparatus. Overall synapse size, the size of the readily releasable vesicle pool as assessed with hypertonic shock, and the number of presynaptic neurotransmitter release sites (active zones) are not changed in the mutants. Short-term synaptic facilitation of evoked transmitter release is decreased in the mutants, suggesting that the absence of dystrophin results in increased probability of release. Absence of the large dystrophin isoforms does not lead to changes in muscle cell morphology or alterations in the postsynaptic electrical response to spontaneously released neurotransmitter. Therefore, postsynaptic glutamate receptor function does not appear to be affected. These results indicate that the postsynaptically localized scaffolding protein Dystrophin is required for appropriate control of neuromuscular synaptic homeostasis (van der Plas, 2006).
This study finds that the presynaptically localized type II BMP receptor wit is required for the increased QC observed in the dystrophin mutant, as also shown for increases in QC induced by the postsynaptic inhibition of CaMKII or glutamate receptor function (Haghighi, 2003). Dystrophin and CaMKII are unlikely, however, to signal through the PMad-dependent BMP signaling pathway, because the expression levels and domains of PMad are unchanged when Dystrophin or CaMKII levels are either decreased or increased. The results address the question as to whether the retrograde BMP signal directly participates in homeostatic signaling ('instructive') or is required for the overall development of the synapse ('permissive'). The findings indicate that it is likely permissive, at least for the homeostatic mechanisms induced by perturbation of Dystrophin or CaMKII levels. BMP signaling may simply be required for the development of the presynaptic apparatus to a point at which it can respond to muscle-derived cues (van der Plas, 2006).
The BMP pathway plays an important role in retrograde signaling at the larval NMJ and in the Drosophila CNS. To evaluate the role of BMP signaling in the increase in neurotransmitter release induced by the absence of dystrophin, electrophysiological recordings were performed at NMJs lacking both dystrophin and wit. The transgenic RNA interference approach was used to reduce postsynaptic dystrophin levels (van der Plas, 2006).
Homozygous wit mutant NMJs postsynaptically expressing dsRNA directed against the Dystrophin large isoforms (UAS-RNAi-dysNH2/+; G14-Gal4/+; wit) displayed mEJP amplitude, EJP amplitude, and QC values similar to the wit mutant alone. The homozygous wit NMJ has very low EJP amplitudes but maintains wild-type level mEJP amplitudes. Thus, the absence of dystrophin failed to elicit an increase in QC in the wit background. Wit function and, by extension, BMP signaling is therefore required at the NMJ for the increase in neurotransmitter release elicited by the absence of dystrophin (van der Plas, 2006).
To evaluate whether the absence of postsynaptic dystrophin affects a known BMP target, immunofluorescence analysis was performed of mutant and wild-type embryonic ventral nerve cords, using an anti-phospho-Mad (PMad) antibody that recognizes the activated form of the Mad downstream effector of wit signaling. No differences were observed in PMad expression levels or domains between the dystrophin mutants or individuals postsynaptically overexpressing DLP2 and controls, whereas homozygous wit mutants showed significantly decreased levels of the protein (van der Plas, 2006).
The dystrophin mutants show similar electrophysiological and morphological phenotypes as larvae postsynaptically expressing CaMKII inhibitors. Both display increased QC and an increase in the ratio of active zones with T-bars versus the total number of active zones, without additional significant changes in synaptic morphology. Furthermore, the increase in QC when CaMKII is reduced postsynaptically is dependent on wit function. Therefore, embryonic ventral nerve cords that have reduced or elevated levels of CaMKII, UAS-Ala, and UAS-T287D, respectively, were examined for PMad staining. No changes in the expression of PMad was observed. These results suggest that, although wit is required for increased QC at NMJs postsynaptically deficient for either Dystrophin or CaMKII, alteration of either Dystrophin or CaMKII levels does not result in changes in embryonic PMad expression. Thus, the interaction of dystrophin or CaMKII and wit in NMJ homeostasis is unlikely to involve regulation of PMad expression (van der Plas, 2006).
Inhibition of postsynaptic glutamate receptors at the Drosophila NMJ initiates a compensatory increase in presynaptic release termed synaptic homeostasis. BMP signaling is necessary for normal synaptic growth and stability. It remains unknown whether BMPs have a specific role during synaptic homeostasis and, if so, whether BMP signaling functions as an instructive retrograde signal that directly modulates presynaptic transmitter release. This study demonstrates that the BMP receptor (Wit) and ligand (Gbb) are necessary for the rapid induction of synaptic homeostasis. Evidence is provided that both Wit and Gbb have functions during synaptic homeostasis that are separable from NMJ growth. However, further genetic experiments demonstrate that Gbb does not function as an instructive retrograde signal during synaptic homeostasis. Rather, the data indicate that Wit and Gbb function via the downstream transcription factor Mad and that Mad-mediated signaling is continuously required during development to confer competence of motoneurons to express synaptic homeostasis (Goold, 2007).
These data advance understanding of BMP signaling at the Drosophila NMJ in several important ways. First, it was demonstrated that BMP signaling is essential for the rapid, protein-synthesis-independent, induction of synaptic homeostasis identified at this NMJ. Because expression of UAS-wit in motoneurons restores synaptic homeostasis in the wit mutant and because suppression of Mad-mediated signaling in neurons blocks synaptic homeostasis, it is concluded that BMP signaling acts upon the motoneuron to enable the rapid induction of synaptic homeostasis. Next, it was shown that the requirement for BMP signaling during synaptic homeostasis is separable from BMP-dependent support of synaptic growth and baseline neurotransmission. Finally, the temporal and spatial requirements for BMP signaling was dissected. The data support the conclusion that Mad-mediated signaling is required constitutively, downstream of the Wit receptor, in order to maintain the competence of motoneurons to express homeostatic plasticity. Further, the data argue that Gbb is not the retrograde signal that directly acts upon the presynaptic motoneuron terminal to homeostatically modulate presynaptic release (Goold, 2007).
It has been hypothesized that Gbb could function as a homeostatic retrograde signal at the Drosophila NMJ. According to this model, Gbb would be released in proportion to the perturbation of postsynaptic muscle excitation in a glutamate receptor mutant and, thereby, instruct the degree of homeostatic compensation expressed by the presynaptic motoneuron terminal. In favor of this model, homeostatic compensation observed in a glutamate receptor mutant is blocked by the wit mutation. This study present two lines of evidence that are consistent with the necessity of BMP signaling for homeostatic compensation. First, it was confirmed that the rapid induction of homeostatic compensation following application of a use-dependent glutamate receptor antagonist, Philanthotoxin (PhTx) is blocked by null mutations in both wit and gbb. Furthermore, it was shown that muscle-specific rescue of the gbb null mutation is sufficient to restore the rapid induction of homeostatic compensation (Goold, 2007).
Despite these compelling genetic data, several experiments now argue against the possibility that Gbb functions as an instructive, retrograde signal that directly modulates presynaptic release during synaptic homeostasis. First, it was found that although muscle-specific rescue of the gbb null mutation is sufficient to restore synaptic homeostasis, so is neuron-specific rescue of the gbb null mutation. Thus, homeostatic compensation can occur even in the absence of muscle-derived Gbb. These data argue against a model in which Gbb functions as the instructive retrograde signal that directly modulates presynaptic release during synaptic homeostasis (Goold, 2007).
Next, it was demonstrated that homeostatic signaling is blocked by expression of DN-Glued in neurons, which disrupts retrograde axonal transport. In this experiment, Gbb signaling at the NMJ should, in theory, persist. Furthermore, it was established that an intact motor axon is not required for the rapid induction of synaptic homeostasis. Thus, it can be concluded that trans-synaptic Gbb signaling from muscle to nerve is not sufficient for the rapid induction of synaptic homeostasis (Goold, 2007).
Given that Wit and Gbb are necessary for synaptic homeostasis, how do they participate in the process if Gbb is not the instructive retrograde signal? This study demonstrates that Mad is necessary for synaptic homeostasis, and evidence is providied that Mad-mediated signaling is required in the motoneuron. In addition, neuronal expression of UAS-Gbb restores homeostatic compensation in the presence of the DN-Glued transgene. These results suggest that the reason DN-Glued disrupts synaptic homeostasis is because it interferes with the retrograde axonal transport of P-Mad downstream of the Wit receptor. This is consistent with the prior demonstration that neuronal expression of Gbb can restore nuclear P-Mad in the presence of UAS-DN-Glued. Because the induction of synaptic homeostasis does not require the motoneuron soma, it is concluded that Gbb does not function as an acute, retrograde signal. Rather, Gbb may be a muscle-derived signal that acts developmentally to confer the competence of motoneurons to express synaptic homeostasis. Thus, the identity of the homeostatic retrograde signal at the NMJ remains unknown. It remains possible that other TGF-β superfamily signaling molecules could function at the NMJ in this capacity, including myoglianin and maverick, though it has been shown that synaptic homeostasis is intact in the baboon receptor mutant (Goold, 2007).
There are several possible ways in which BMP signaling could confer competence for motoneurons to express homeostatic plasticity. One possibility is that the BMPs control a transcriptional program that is necessary for synaptic homeostasis. For example, BMPs are potent regulators of cell fate during embryonic development. Perhaps the ability of motoneurons to express synaptic homeostasis is related to the maintenance of their cellular or electrical identity. An alternate possibility is that BMPs control the expression of essential presynaptic proteins that are required for synaptic homeostasis. For example, it has been shown in other systems that target-dependent TGF-β signaling can modulate neuronal ion channel expression. It has been demonstrated that CaV2.1 calcium channels are required for synaptic homeostasis at the Drosophila NMJ. However, it is considered unlikely that BMPs control synaptic homeostasis through the regulation of CaV2.1 channel expression because there is not a strong correlation between altered baseline synaptic transmission and the expression of synaptic homeostasis. Furthermore, overexpression of a GFP-tagged CaV2.1 calcium channel (cacophony-GFP) is unable to restore synaptic homeostasis when coexpressed with UAS-dad. Finally, BMP signaling could influence the expression of synaptic homeostasis by targeting the rate of spontaneous miniature release. Spontaneous release events that persist in the absence of evoked neurotransmission are sufficient to induce homeostatic compensation at the Drosophila NMJ. However, no strong correlation is found between baseline mEPSP frequency and whether or not a mutant NMJ is able to express synaptic homeostasis. Although the wit mutants show a severe decrease in mEPSP rate compared to wild-type, the expression of UAS-dad or UAS-DN-Glued both block synaptic homeostasis without severely impairing baseline mEPSP rate. Ultimately, continued forward genetic investigation of homeostatic signaling may be required to identify the BMP-dependent mechanisms that control the expression of synaptic homeostasis (Goold, 2007).
BMP signaling is required for NMJ growth, baseline neurotransmission, and NMJ stability in addition to being required for synaptic homeostasis. It is a challenge, therefore, to determine whether BMP signaling has a specific function during synaptic homeostasis versus a more general role during synapse development. This study presents several lines of evidence that BMP signaling may have a separable function during synaptic growth versus synaptic homeostasis. First, it was demonstrated that synaptic homeostasis can occur at BMP mutant synapses that show severely impaired synaptic growth. For example, the gbb hypomorphic mutant has a decrease in bouton number that is just as severe as the gbb null mutant, but the gbb hypomorphic mutant shows normal homeostatic compensation. As another example, animals in which UAS-gbb and UAS-DN-Glued are coexpressed have a severe decrease in bouton number but normal homeostatic compensation. Thus, it is concluded that normal BMP-dependent synaptic growth is not required for the expression of synaptic homeostasis (Goold, 2007).
It was also possible to dissociate BMP-dependent baseline transmission from both synaptic growth and synaptic homeostasis. (1) Muscle-specific rescue of the gbb null mutation significantly restores synaptic growth and rescues synaptic homeostasis, but baseline transmission remains at levels observed in the null mutant. (2) Motoneuron-specific rescue of the wit mutation (OK371-GAL4) similarly rescues bouton number and synaptic homeostasis, although baseline transmission remains severely impaired. (3) Animals in which UAS-gbb and UAS-DN-Glued are coexpressed have a severe decrease in baseline transmission but normal homeostatic compensation. (4) Results were obtained that show the converse effect. When UAS-dad is expressed for 1.5 days at the end of larval development, both synaptic homeostasis and baseline transmission are significantly impaired, but synaptic bouton numbers remain wild-type. From these data it is concluded that impaired synaptic homeostasis is not a secondary consequence of BMP-dependent functional NMJ development. It also appears that there may be distinct effects of BMP signaling on the anatomical versus functional development of the NMJ. One possibility, consistent with BMPs being a classical morphogen, is that different levels of the ligand could initiate specific transcriptional programs with distinct effects on bouton number, baseline transmission, and homeostatic plasticity. It is also possible that the site of action of BMP signaling will play an important role in specifying signaling outcome (Goold, 2007).
It has been speculated that synaptic homeostasis might function, over the course of development, to ensure that the muscle cell is normally depolarized by the NMJ. How can one explain the observation that csp and syx/+ mutations have decreased baseline neurotransmitter release but normal acute synaptic homeostasis in response to PhTx application, or other genotypes explored in this study that show impaired baseline transmission and normal acute synaptic homeostasis? It has been demonstrated that the acute induction of synaptic homeostasis is independent of evoked neurotransmission. Thus, synaptic homeostasis may not function to modulate the absolute amplitude of evoked neurotransmitter release. Rather, synaptic homeostasis might be a rapid system to offset acute perturbations of postsynaptic receptor function. In this case, developmental programs that specify NMJ anatomy and active zone addition would achieve the reproducible development of the NMJ. Alternatively, the mechanisms of acute homeostatic compensation following PhTx application may be separable, either temporally or molecularly, from the other potential mechanisms that monitor and homeostatically control evoked EPSP amplitudes (Goold, 2007).
The data also suggest a possible link between the expression of homeostatic plasticity and the mechanisms of neuromuscular degenerative disease. Genetic mutations that impair retrograde axonal transport have been shown to cause familial amyotrophic lateral sclerosis. It has also been shown that, in Drosophila and mice, mutations that disrupt dynein-dynactin complex function lead to neuromuscular synapse degeneration. It is hypothesized that impaired retrograde axonal transport deprives motoneurons of muscle-derived trophic support leading to motoneuron degeneration. This study has demonstrated that impaired retrograde axonal transport blocks the expression of homeostatic plasticity at the NMJ. This deficit can be restored by expression of BMPs in the central nervous system, bypassing retrograde axonal transport as the source of BMPs to the motoneuron cell body. It is tempting to speculate that impaired synaptic homeostasis at the NMJ may play a role in the progression of motoneuron disease associated with impaired retrograde axonal transport (Goold, 2007).
Finally, the data could have relevance to the sustained expression of homeostatic plasticity in regions of the adult nervous system. BMPs and downstream signaling proteins such as the Smads continue to be expressed in the adult nervous system. In particular, BMPs are secreted into the cerebral spinal fluid at concentrations that are relevant for neuronal signaling. It is, therefore, interesting to speculate that circulating levels of BMPs might sustain the competence of neurons to express homeostatic plasticity without driving morphological plasticity in the adult nervous system (Goold, 2007).
Reference names in red indicate recommended papers.
Aberle, H., et al. (2002). wishful thinking encodes a BMP type II receptor that regulates synaptic growth in Drosophila. Neuron 33: 545-558. 11856529
Allan, D., St. Pierre, S. E., Miguel-Aliaga, I. and Thor, S. (2003). Specification of neuropeptide cell identity by the integration of retrograde BMP signaling and a combinatorial transcription factor code. Cell 113: 73-86. 12679036
Arber, S., et al. (1998). Regulation of actin dynamics through phosphorylation of cofilin by LIM-kinase. Nature 393: 805-809. 9655397
Baarends, W. M., et al. (1994). A novel member of the transmembrane serine/threonine kinase receptor family is specifically expressed in the gonads and in mesenchymal cells adjacent to the mullerian duct. Development 120: 189-197. 8119126
Beppu, H., Kawabata, M., Hamamoto, T., Chytil, A., Minowa, O., Noda ,T. and Miyazono, K. (2000). BMP type II receptor is required for gastrulation and early development of mouse embryos. Dev. Biol. 221: 249-258. 10772805
Beppu, H., Lei. H., Bloch, K. D. and Li, E. (2005). Generation of a floxed allele of the mouse BMP type II receptor gene. Genesis 41(3): 133-7. 15736264
Cameron, J. S., Lhuillier, L., Subramony, P. and Dryer, S. E. (1998). Developmental regulation of neuronal K+ channels by target-derived TGFß in vivo and in vitro. Neuron 21: 1045-1053. 9856460
Cameron, J. S., Dryer, L. and Dryer, S. E. (1999). Regulation of neuronal K(+) currents by target-derived factors: opposing actions of two different isoforms of TGFß. Development 126: 4157-4164. 10457024
Charytoniuk, D. A., Traiffort, E., Pinard, E., Issertial, O., Seylaz, J. and Ruat, M. (2000). Distribution of bone morphogenetic protein and bone morphogenetic protein receptor transcripts in the rodent nervous system and up-regulation of bone morphogenetic protein receptor type II in hippocampal dentate gyrus in a rat model of global cerebral ischemia. Neuroscience 100: 33-43. 10996456
Chin, J., Angers, A., Cleary, L. J., Eskin, A. and Byrne, J. H. (1999). TGF-1 in Aplysia: role in long-term changes in the excitability of sensory neurons and distribution of TR-II-like immunoreactivity. Learn. Mem. 6: 317-330. 10492013
Colavita, A., Krishna, S., Zheng, H., Padgett, R. W. and Culotti, J. G. (1998). Pioneer axon guidance by UNC-129, a C. elegans TGF-ß. Science 281: 706-709. 9685266
Délot, E. C., et al. (2003). BMP signaling is required for septation of the outflow tract of the mammalian heart. Development 130: 209-220. 12441304
Deng, Z., et al. (2000). Familial primary pulmonary hypertension (gene PPH1) is caused by mutations in the bone morphogenetic protein receptor-II gene. Am. J. Hum. Genet. 67: 737-744. 10903931
Eaton, B. A., Fetter, R. D. and Davis, G. W. (2002). Dynactin is necessary for synapse stabilization. Neuron 34: 729-74. 12062020
Eaton, B. A. and Davis, G. W. (2005). LIM kinase1 controls synaptic stability downstream of the type II BMP receptor. Neuron 47: 695-708. 16129399
Endo, M., et al. (2003). Control of growth cone motility and morphology by LIM kinase and Slingshot via phosphorylation and dephosphorylation of cofilin. J. Neurosci. 23: 2527-2537. 12684437
Foletta, V. C. et al., (2003). Direct signaling by the BMP type II receptor via the cytoskeletal regulator LIMK1. J. Cell Biol. 162: 1089-1098. 12963706
Frisch, A. and Wright, C. V. (1998). XBMPRII, a novel Xenopus type II receptor mediating BMP signaling in embryonic tissues. Development 125: 431-442. 9425138
Goda, Y. and Davis, G. W. (2003). Mechanisms of synapse assembly and disassembly. Neuron 40: 243-264. 14556707
Goold, C. P. and Davis, G. W. (2007). The BMP Ligand Gbb gates the expression of synaptic homeostasis independent of synaptic growth control. Neuron 56: 109-123. Medline abstract: 17920019
Guo, X., Rueger, D. and Higgins, D. (1998). Osteogenic protein-1 and related bone morphogenetic proteins regulate dendritic growth and the expression of microtubule-associated protein-2 in rat sympathetic neurons. Neurosci. Lett. 245: 131-134. 9605473
Haghighi, A. P., et al. (2003). Retrograde control of synaptic transmission by postsynaptic CaMKII at the Drosophila neuromuscular junction. Neuron 39: 255-267. 12873383
Hassel, S., et al. (2006). Interaction and functional cooperation between the serine/threonine kinase bone morphogenetic protein type II receptor with the tyrosine kinase stem cell factor receptor. J. Cell Physiol. 206(2): 457-67. 16155937
Jiang, Y., McLennan, I. S., Koishi, K. and Hendry, I. A. (2000). Transforming growth factor-beta 2 is anterogradely and retrogradely transported in motoneurons and up-regulated after nerve injury. Neuroscience 97: 735-742. 11107165
Kawabata, M., Chytil, A. and Moses, H. L. (1995). Cloning of a novel type II serine/threonine kinase receptor through interaction with the type I transforming growth factor-ß receptor. J. Biol. Chem. 270: 5625-5630. 7890683
Kirsch, T., Nickel, J. and Sebald, W. (2000). BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-II. EMBO J. 19: 3314-3324
Lane, K. B., et al. (2000). Heterozygous germline mutations in BMPR2, encoding a TGF- receptor, cause familial primary pulmonary hypertension. The International PPH Consortium. Nat. Genet. 26: 81-84
Lee-Hoeflich, S. T., et al. (2004). Activation of LIMK1 by binding to the BMP receptor, BMPRII, regulates BMP-dependent dendritogenesis. EMBO J. 23: 4792-4801. 15538389
Lein, P., Johnson, M., Guo, X., Rueger, D. and Higgins, D. (1995). Osteogenic protein-1 induces dendritic growth in rat sympathetic neurons. Neuron 15: 597-605. 7546739
Liu, F., et al. (1995). Human type II receptor for bone morphogenic proteins (BMPs): extension of the two-kinase receptor model to the BMPs. Mol. Cell. Biol. 15(7): 3479-3486.
Liu, Q. R., et al. (1997). A developmental gene (Tolloid/BMP-1) is regulated in Aplysia neurons by treatments that induce long-term sensitization. J. Neurosci. 17: 755-764. 8987797
Lo, P. C. and Frasch, M. (1999). Sequence and expression of myoglianin, a novel Drosophila gene of the TGF-ß superfamily. Mech. Dev. 86: 171-175. 10446278
Lorentzon, M., Hoffer, B., Ebendal, T., Olson, L. and Tomac, A. (1996). Habrec1, a novel serine/threonine kinase TGF-ß type I-like receptor, has a specific cellular expression suggesting function in the developing organism and adult brain. Exp. Neurol. 142: 351-360. 8934566
Machado, R. D., et al. (2003). Functional interaction between BMPR-II and Tctex-1, a light chain of Dynein, is isoform-specific and disrupted by mutations underlying primary pulmonary hypertension. Hum. Mol. Genet. 12(24): 3277-86. 14583445
Marqués, G., et al. (2002). The Drosophila BMP type II receptor Wishful thinking regulates neuromuscular synapse morphology and function. Neuron 33: 529-543. 11856528
Marqués, G., et al. (2003). Retrograde Gbb signaling through the Bmp type 2 receptor Wishful thinking regulates systemic FMRFa expression in Drosophila. Development 130: 5457-5470. 14507784
McCabe, B. D., et al. (2003). The BMP homolog Gbb provides a retrograde signal that regulates synaptic growth at the Drosophila neuromuscular junction. Neuron 39: 241-254. 12873382
McLennan, I. S. and Koishi, K. (1994). Transforming growth factor-ß-2 (TGF-ß2) is associated with mature rat neuromuscular junctions. Neurosci. Lett., 177: 151-154. 7824170
McPherron, A. C., Lawler, A. M. and Lee, S. J. (1997). Regulation of skeletal mu