Several lines of evidence suggested that Nwk might interact with Wasp: (1) the SH3a domain of Nwk is between 35% and 45% identical to the three SH3 domains of Dock. Nck, a mammalian homolog of Dock, binds to Wasp via its SH3 domains. (2) Bzz1p, the yeast ortholog of Nwk, binds to Las17p, the yeast ortholog of Wasp (Soulard, 2002), via its SH3 domains. (3) wsp mutants have a morphological phenotype similar to nwk at larval NMJs (Coyle, 2004).
To evaluate whether Nwk and Wsp interact biochemically, affinity chromatography was performed using GST-fusion proteins containing each of the two Nwk SH3 domains. The construct containing the SH3a domain, but not the one containing the SH3b domain, precipitated Wsp immunoreactivity from Drosophila head homogenates with high affinity (Coyle, 2004).
A yeast two-hybrid protein binding assay further indicated that Nwk and Wsp can interact directly. Two-bait constructs expressing either full-length Nwk or a subfragment consisting of the two SH3 domains plus the C terminus stimulates transcription of reporter genes in yeast cells cotransformed with a full-length Wsp prey. Colonies grew rapidly on selectable media and expressed β-gal. In contrast, bait constructs containing the FCH and ARNEY domains of Nwk but lacking the SH3 domains did not activate reporter genes when cotransformed with Wsp prey, indicating that the SH3 domains are required for binding with Wsp. Single transformation of any Nwk bait or Wsp prey alone did not stimulate transcription of reporter genes at substantial levels (Coyle, 2004).
A previous study of Drosophila wsp mutants revealed defects in sensory cell proliferation during PNS development, but a role at NMJs had not been investigated (Ben-Yaacov, 2001). Therefore, to determine whether endogenous Nwk and Wsp are capable of interacting in vivo, the distribution of Wsp at the NMJ was examined by immunocytochemistry. Wsp was found to be localized diffusely at synaptic boutons in irregular patches. Postsynaptically, Wsp is associated with the subsynaptic reticulum (ssr), which was revealed by colocalization with the ssr marker, Dlg, and by reduction of immunoreactivity in pak mutants that reduce the ssr. There is also a significant presynaptic component of Wsp immunoreactivity that is most evident in thin (0.25 μm) optical sections through the center of boutons. Wsp is clearly present both inside the bouton as well as in association with the surrounding ssr. Although the presynaptic distribution of Wsp is not continuous throughout the periactive zone, regions of overlap with Nwk can be observed in thin optical confocal sections through double-labeled boutons. Thus, Nwk and Wsp colocalize in spatially restricted regions of the periactive zone. Nwk is not required for Wsp localization since Wsp immunoreactivity is not obviously disrupted in nwk mutants (Coyle, 2004).
Wasp promotes the nucleation and branching of F-actin. Phalloidin staining revealed that F-actin is enriched around synaptic boutons and generally colocalizes with Wsp. The overall abundance and distribution of F-actin appear normal in wsp mutants, indicating that Wsp is not the only factor promoting actin polymerization at the synapse. The Drosophila genome contains one other wsp family member, scar, whose function may overlap with wsp, although its role at synapses has not been investigated. Thus, loss of Wsp may have only a limited, localized effect on actin dynamics and architecture without completely abolishing the cortical cytoskeleton (Coyle, 2004).
In situ hybridization of embryos has shown that nwk is abundantly transcribed throughout the CNS and PNS, whereas its expression was not detected in other tissues, including muscles. A polyclonal anti-Nwk antiserum stained motorneuron axons and terminals, but not muscles. Confocal microscopy revealed a distinctive distribution of Nwk within synaptic boutons. Nwk is distributed in a patchwork of rings in the plane of the bouton membrane. Pak, a marker of synaptic densities, has a complementary punctate distribution, such that each Nwk ring surrounds one Pak punctum. This distribution of Nwk corresponds to the periactive zone, described by for Highwire. These proteins also regulate NMJ morphology: this led to the proposal that the periactive zone is specialized for synaptic growth regulation. Thus, the localization of Nwk to this region is consistent with the morphological defects in nwk mutants (Coyle, 2004).
To identify mutations that perturb function and development of the nervous system in Drosophila, a large collection of temperature-sensitive (ts) paralytic mutants was isolated following mutagenesis by ethylmethane sulfonate (EMS). The ts paralytic phenotype often results from the absence of an affected protein and not from a thermolabile variant. Such mutants can have defective synaptic morphology or function even at nominally permissive temperatures. Therefore, ts mutants were screened to identify those with abnormal morphology at the larval NMJ. nwk is the first of these lines to be so characterized (Coyle, 2004).
nwk adults behave normally at room temperature, but rapidly (10-40 s) lose coordination at 38°C and undergo seizure-like spasms that gradually diminish until the flies become paralyzed within 3 min. When returned to room temperature after a 5 min exposure to 38°C, nwk mutants gradually recover, and normal behavior is fully restored after 5 to 10 min. Wild-type flies exhibit no locomotor defects at 38°C for at least 20 min (Coyle, 2004).
The glutamatergic synapses of the Drosophila larval NMJ have become a favored genetic model for studies of synaptic development and plasticity. Each motorneuron forms distinctive synaptic junctions on its target muscle(s), containing stereotypic numbers of branches and boutons. Initial observations revealed that NMJs in nwk larvae were more extensive than in wild-type and that this phenotype was 100% penetrant. No defects in axon pathfinding were observed (Coyle, 2004).
To quantify the nwk phenotype, bouton number, NMJ length, branch number, and branch complexity at NMJ 6/7 and NMJ 4 were measured. Compared with controls, the number of boutons at NMJ 6/7 in nwk1 larvae is increased by 50%. A similar result was also observed at NMJ 4. Because NMJ growth varies in proportion to muscle size, measurements were normalized to muscle surface area. At NMJ 6/7, normalized bouton number remains 50% larger in nwk1 compared with control larvae, whereas at NMJ 4, it becomes 22% larger. The increase in bouton number was accompanied by an expansion in NMJ length as measured by summing the lengths of all axon branches within NMJ 6/7. After normalization to muscle area, NMJ length was increased by 30% (Coyle, 2004).
nwk mutants also display increased frequency and complexity of branch formation. Branches form from pre-existing boutons in a process that superficially resembles budding or asymmetric division in yeast cells. Typically, a branching bouton bifurcates to yield just two new branches. In agreement with previous studies, only a small percentage of boutons in wild-type larvae contained branch points at NMJ 6/7 and NMJ 4 and virtually all of these branches were bifurcations. Among all 36 wild-type NMJs examined, only a single instance was found of a bouton extending more than two new branches. In nwk1, the proportion of branching boutons was increased by 50% at NMJ 6/7 and by 125% at NMJ 4. Moreover, there was a striking increase in the incidence of hyperbranched boutons, those extending three or more new branches. At least one such hyperbranched bouton was observed at all NMJs 6/7 and among about half of all NMJs 4 examined in nwk1 larvae. Most of these hyperbranched boutons extended three new branches but 10% extended even four, suggesting that the underlying mechanism of bouton division is misregulated in nwk mutants (Coyle, 2004).
Four additional nwk alleles (nwk2-5) were obtained following EMS and γ ray mutagenesis in screens for failure to complement the ts paralytic phenotype of nwk. All of these mutants are thought to be functional nulls, and all display similar NMJ overgrowth phenotypes, confirming that nwk is responsible for both morphological defects in larvae and ts paralysis in adults. To further verify that loss of Nwk is responsible for the observed phenotypes, the phenotype of nwk2 heterozygous were evaluated with the deficiency Df(3L)Rdl2, which uncovers nwk. These nwk2 hemizygotes had synaptic defects that were nearly identical with those observed in nwk2 homozygotes (Coyle, 2004).
To investigate the effects of nwk on synaptic growth at the ultrastructural level, electron microscopic analysis of serially sectioned type Ib boutons from NMJ 6/7 in nwk2 and wild-type larvae was performed. General bouton anatomy in nwk2 appeared normal. Postsynaptically, boutons were surrounded by an extensive subsynaptic reticulum (ssr) that resembled wild-type with respect to thickness and complexity. Individual synaptic contacts (synaptic densities), which are identified by dark electron-dense areas in the cell membrane, were present in mutant boutons and contained active zones, marked by the presence of T bars surrounded by clouds of synaptic vesicles. The size and abundance of these vesicles appeared normal. In addition, the distribution of T bars (ranging from 0 to 3) per synaptic contact was similar in mutant and wild-type boutons. In wild-type, 25% of synaptic contacts had no T bars, 63% had one T bar, 10% had two T bars, and 2% had three T bars. In nwk2, 20.6% had no T bars, 63.7% had one T bar, 13.7% had two T bars, and 2% had three T bars. Comparable results were obtained for nwk2/Df (Coyle, 2004).
Nonetheless, the general shape of nwk boutons was abnormal. Midline cross-sections of wild-type boutons were circular in outline, but those of nwk2 were consistently elliptical. Morphometric analysis indicated that the surface area and volume of nwk2 boutons were reduced by 42% and 60%, respectively. These changes were accompanied by a reduction in the size and number of individual synaptic contacts. Each wild-type bouton contained an average of 20 synaptic contacts, with an average surface area of 0.3673 ± 0.0202 μm2 each. In contrast, nwk2 boutons contained an average of only 10 synaptic contacts, 50% percent fewer, with an average surface area of 0.2817 ± 0.0181 μm2 (p < 0.005), 23% smaller (Coyle, 2004).
It is concluded that the total area of synaptic contact is reduced in each nwk NMJ. Because type Ib boutons contain the majority of the synaptic contacts within a given NMJ 6/7, having almost 3 times more than type I boutons per μm length, the 50% reduction in synaptic contact number observed in nwk2 indicates a large reduction in the total synaptic complement. This reduction would not be rectified by the increased synaptic bouton number in nwk mutants. Although there are 50% more boutons at NMJ 6/7 in nwk larvae, at least a 100% increase would be required to compensate for the 50% reduction in synaptic contacts per bouton. Since T bars (active zones) are distributed normally among nwk synapses, nwk NMJs contain fewer total active zones. Furthermore, the average surface area of each synaptic contact is reduced in nwk. Taken together, these data demonstrate that motorneuron terminals in nwk NMJs contain less total surface area of synaptic contact with the postsynaptic muscle (Coyle, 2004).
Nerve-evoked excitatory junctional currents (EJCs) were recorded from NMJ 6/7 to determine if nwk causes defects in synaptic function that parallel the morphological defects described above. Both nwk1 and nwk2 reduced mean EJC amplitude by >50% compared with wild-type. Although EJC amplitudes were reduced, the amplitude of miniature EJCs (mEJCs), or quantal size, was slightly increased in nwk1 and nwk2 compared with wild-type. mEJC frequency was not significantly different in nwk2, although it was slightly elevated in nwk1. Taken together, these results demonstrate that synaptic transmission is impaired in nwk mutants. The large reduction in quantal content, despite a small increase in quantal size, suggests that presynaptic vesicle release is reduced (Coyle, 2004).
Transgenic flies were generated expressing a cDNA that encodes a complete nwk ORF plus 5′ and 3′ untranslated sequences (UTRs) linked to a yeast UAS promoter (UAS-nwk+). Neural-specific expression of UAS-nwk+ using an elaV-Gal4 driver in a homozygous nwk2 background restored the presence of Nwk at motor terminals with normal periactive zone localization. The ts paralytic phenotype of nwk2 adults was also rescued: after 15 min at 38°C, more than 90% of the transgenic flies continued to walk and climb normally (Coyle, 2004).
NMJ overgrowth phenotypes were fully or partially rescued by neural-specific expression of UAS-nwk+. The normalized numbers of boutons and branches were fully rescued at NMJ 6/7 and NMJ 4. The incidence of hyperbranching boutons per NMJ was not significantly different between the rescue and control lines. Total synapse length of NMJ 6/7 was substantially reduced compared with nwk2. At the electrophysiological level, EJC amplitude was fully rescued by neural-specific expression of the UAS-nwk+ transgene, and mEJC amplitude was partially rescued (Coyle, 2004).
The inability to achieve perfect rescue could result from failure of the transgene to exactly recapitulate the timing and amplitude of nwk expression, or from the lack of particular splice variants. Nonetheless, all nwk phenotypes were at least partially rescued by neural-specific expression of UAS-nwk+. In contrast, postsynaptic expression of UAS-nwk+ using a muscle driver did not rescue the ts or morphological phenotypes. In summary, these results demonstrate that the behavioral, developmental, and electrophysiological phenotypes associated with nwk all result from mutation of a single gene that acts presynaptically (Coyle, 2004).
The physical interaction between Nwk and Wsp suggest that they might participate in a common regulatory pathway affecting F-actin dynamics in synapses. If so, then wsp loss-of-function mutations should cause increased NMJ growth and branching similar to nwk. Indeed, it was found that bouton number, branch number, total NMJ length, and hyperbranching are all increased in wsp1/Df and wsp1. wsp1 is a putative null allele, with occasional escapers surviving to the pharate adult stage. Although there is maternal contribution of Wsp in embryos, no Wsp immunoreactivity was detected at wsp NMJs. Thus, Wsp and Nwk have a similar function in the regulation of synaptic growth and bouton branching (Coyle, 2004).
To further investigate the functional relationship between Nwk and Wsp in vivo, double mutants were examined. The phenotype of homozygous nwk wsp double mutants is more severe than that of either single mutant with respect to bouton number, branch formation, and NMJ length. Also, hyperbranch complexity was dramatically increased, producing distinct synaptic morphologies. At NMJ 6/7 in double mutants, 3.2 ± 0.5 hyperbranched boutons were observed per NMJ, compared with only 0.56 ± 0.16 hyperbranches per NMJ in nwk2 or 0.86 ± 0.35 in wsp1/Df (p < 0.05). On average, each double mutant NMJ contained one bouton that emitted four new branches, while this type of hyperbranch was observed in only 10% of NMJs in single mutants. Boutons extending 5 or 6 new branches are found in more than 10% of the double mutant larvae (Coyle, 2004).
These results suggest that Nwk and Wsp interact presynaptically to regulate NMJ growth. Consistent with this interpretation, the morphological phenotype of wsp mutants was partially rescued by driving expression of a UAS-wsp+ transgene using a panneuronal Gal4 driver. Normalized bouton number, NMJ length, and branch number were significantly reduced toward wild-type values. Thus, presynaptic expression of Wsp, like Nwk, normally suppresses NMJ growth and bouton branching (Coyle, 2004).
In summary, the striking similarity in NMJ phenotypes exhibited by nwk and wsp mutants and the enhanced phenotypes observed in double mutants strongly support the idea that Nwk and Wsp act in a common regulatory pathway. The more extreme phenotype observed in double mutants compared with either single mutant most likely indicates that either mutation alone does not completely abolish the pathway as would be expected if there is some functional redundancy. For example, Wsp is known to be activated by many different binding partners and Nwk may act in part via other effectors besides Wsp (Coyle, 2004).
These data above suggest that nwk and wsp contribute to a common regulatory mechanism in a gene dosage-dependent manner, as do several other cytoskeletal effectors required for axon growth. To verify this, focus was placed on the incidence of hyperbranching, which is the most distinctive trait observed in these mutants. An increase was observed in the frequency and complexity of these structures; the number of wild-type nwk and wsp alleles decreased. Wild-type larvae exhibited no hyperbranches. Heterozygotes for either nwk or wsp alone displayed a low incidence of hyperbranching, but boutons with four or more new branches were never observed. In contrast, double heterozygotes (i.e., nwk2 wsp1/+ +) were intermediate between wild-type and the respective homozygotes, including the appearance of boutons that gave rise to four new branches. In hemizygotes for nwk or wsp, loss of one wild-type allele of the other gene caused a further enhancement of the mutant phenotype. Double homozygotes displayed the highest incidence of hyperbranch formation with boutons emitting up to six new branches. Similar trends were observed for bouton number and NMJ length. These data provide additional evidence that Nwk and Wsp interact within a common regulatory complex to regulate bouton proliferation and branching (Coyle, 2004).
The mechanism of bouton branching is unresolved at present, but it involves the reorganization of cytoskeletal components including microtubules (MTs). Overall MT organization appeared normal in nwk and wsp mutants, including the presence of microtubule loops within a subset of terminal boutons. In wild-type larvae, these loops are associated with stable boutons and disintegrate during branch formation. In addition, the patterns of MTs within branching boutons in wild-type and double mutants were indistinguishable. Thus, there is no major disorganization of MT structure in nwk and wsp mutants. Rather, abnormal proliferation and branching of boutons in these mutants is likely caused by the disruption of some other pathway.
Reference names in red indicate recommended papers.
Ben-Yaacov, S., Le Borgne, R., Abramson, I., Schweisguth, F. and Schejter, E.D. (2001). Wasp, the Drosophila Wiskott-Aldrich syndrome gene homolog, is required for cell fate decisions mediated by Notch signaling. J. Cell Biol. 152: 1-13. 11149916
Coyle, I. P., et al. (2004). Nervous wreck, an SH3 adaptor protein that interacts with Wsp, regulates synaptic growth in Drosophila. Neuron 41: 521-534. 14980202
Endris, V., et al. (2002). The novel Rho-GTPase activating gene MEGAP/ srGAP3 has a putative role in severe mental retardation. Proc. Natl. Acad. Sci. 99: 11754-11759. 12195014
Gavin, A.C., Bosche, M., Krause, R., Grandi, P., Marzioch, M., Bauer, A., Schultz, J., Rick, J. M., Michon, A. M. and Cruciat, C. M., et al. (2002). Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415: 141-147. 11805826
Soulard, A., Lechler, T., Spiridonov, V., Shevchenko, A., Shevchenko, A., Li, R., and Winsor, B. (2002). Saccharomyces cerevisiae Bzz1p is implicated with type I myosins in actin patch polarization and is able to recruit actin-polymerizing machinery in vitro. Mol. Cell. Biol. 22: 7889-7906. 12391157
Wong, K., Ren, X. R., Huang, Y.Z., Xie, Y., Liu, G., Saito, H., Tang, H., Wen, L., Brady-Kalnay, S. M. and Mei, L. et al. (2001). Signal transduction in neuronal migration: roles of GTPase activating proteins and the small GTPase Cdc42 in the Slit-Robo pathway. Cell 107: 209-221. 11672528
date revised: 23 June 2004
Home page: The Interactive Fly © 2003 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.