Tricornered Kinase Regulates Synapse Development by Regulating the Levels of Wiskott-Aldrich Syndrome Protein

Precise regulation of synapses during development is essential to ensure accurate neural connectivity and function of nervous system. Many signaling pathways, including the mTOR (mechanical Target of Rapamycin) pathway operate in neurons to maintain genetically determined number of synapses during development. mTOR, a kinase, is shared between two functionally distinct multi-protein complexes- mTORC1 and mTORC2, that act downstream of Tuberous Sclerosis Complex (TSC). We and others have suggested an important role for TSC in synapse development at the Drosophila neuromuscular junction (NMJ) synapses. In addition, our data suggested that the regulation of the NMJ synapse numbers in Drosophila largely depends on signaling via mTORC2. In the present study, we further this observation by identifying Tricornered (Trc) kinase, a serine/threonine kinase as a likely mediator of TSC signaling. trc genetically interacts with Tsc2 to regulate the number of synapses. In addition, Tsc2 and trc mutants exhibit a dramatic reduction in synaptic levels of WASP, an important regulator of actin polymerization. We show that Trc regulates the WASP levels largely, by regulating the transcription of WASP. Finally, we show that overexpression of WASP (Wiskott-Aldrich Syndrome Protein) in trc mutants can suppress the increase in the number of synapses observed in trc mutants, suggesting that WASP regulates synapses downstream of Trc. Thus, our data provide a novel insight into how Trc may regulate the genetic program that controls the number of synapses during development.


Introduction
Synapses are the fundamental communication links between neurons and their targets.Accurate neuronal circuit function is partly determined by the number of synapses; therefore, synapse numbers are precisely regulated during development [1].Altered synapse development is associated with some of the neurodevelopmental disorders, such as autism spectrum disorders (ASDs) [2].Various signaling cascades act in concert to accurately establish an appropriate number of synapses.Among them, the ubiquitous mechanistic target of rapamycin (mTOR) pathway is known to play a crucial role at synapses, although its molecular mechanism of action remains to be elucidated [3].
Many patients with mutations in the tuberous sclerosis complex (TSC) gene suffer from various neurocognitive defects, including ASD and epilepsy [4].Several recent studies in vertebrates have demonstrated an essential role for TSC in regulating the neuronal morphology and the function of excitatory glutamatergic synapses [5][6][7][8][9].Consistent with these studies, we reported that in a fly model of TSC, loss of TSC function leads to increased number of synapses [10].Furthermore, we suggested that mTORC2 [11,12] and not mTORC1 might be largely responsible for the increased number of synapses.
Little is known about the functionally distinct roles of mTORC1 and mTORC2, especially during synapse development.Increasingly, it is becoming apparent that mTORC2 plays important roles in neurons [13][14][15][16][17][18], including regulating synapse development and physiology [10].Therefore, at least some neurological deficits observed in TSC patients could be due to impaired mTORC2 signaling.Taken together, these data highlight the emerging role for the mTORC2 pathway in regulating neuronal morphology and physiology during development.However, the molecular mechanisms by which mTORC2 regulates these processes remain poorly understood.
In order to understand the molecular mechanisms by which mTORC2 may regulate synapse development, especially in the context of TSC, we used a candidate screening approach and identified Trc kinase as a possible downstream genetic effector of this pathway.The tricornered (Trc) serine/threonine kinase belongs to the mammalian NDR-1 (nuclear Dbf-2 related) subgroup of AGC family of kinases and are functionally conserved from yeast to humans [19].NDR kinases have important roles in mitotic exit, cytokinesis, proliferation, and apoptosis [19,20], all of which require major rearrangements of the actin cytoskeleton.In sensory neurons, mTORC2 regulates Trc activity to regulate dendritic tiling [21].A recent study hinted at the possibility that Trc kinase may function at the fly NMJs [22], but its mechanism of action and more importantly, whether it can mediate TSC signaling in neurons remains unknown.
In this study we show that trc mutants phenocopy the synaptic overgrowth exhibited by TSC pathway mutants and interact genetically with Tsc2 and rictor-an essential component of mTORC2.Furthermore, synaptic Trc levels are significantly decreased in Tsc2 mutants, indicating that Trc may act downstream of TSC.Interestingly, both Tsc2 and trc mutants exhibited dramatic decreases in synaptic WASP levels [23]-a potent regulator of actin cytoskeleton [23,24].Importantly, overexpression of WASP in trc mutants and transheterozygotes of trc and Tsc2 (which also show synaptic overgrowth) suppressed their synaptic overgrowth phenotypes.Thus, we propose that Trc kinase likely acts downstream of the TSC-mTORC2 pathway to restrict synapse numbers by regulating the synaptic WASP levels.

Results
Trc kinase restricts synapse growth at the Drosophila NMJ Our previous study suggests that TSC restricts the number of synapses at the Drosophila NMJ via the mTORC2-Akt pathway [10].To determine how this pathway regulates synapse development, we performed a candidate genetic screen.
Among the known/predicted genetic interactors of mTORC2/Akt, we chose those that are enriched in larval/adult Drosophila central nervous system (CNS) (S1 Table ).Mutants of top ~95 candidate genes were obtained from the Bloomington Stock Center (http://flystocks.bio.indiana.edu),balanced over a green fluorescent protein (GFP) balancer to identify the homozygous mutants, and labeled with antibodies against Bruchpilot (BRP, presynaptic marker) [25], glutamate receptor (DGluRIII, postsynaptic marker) [26], and horseradish peroxidase (HRP, to mark neuronal membranes).The homozygous (non-GFP fluorescent) larvae were identified and tested for any alterations in synapse numbers.Using this strategy, we isolated three independent lines that exhibited alteration of synapse numbers at the NMJs.Of those, arfaptin [27] and lim kinase [28] are known to be required for synapse development/ maintenance, and served as controls for our screen.The third candidate was Trc kinase, which had been previously implicated in regulating dendritic tiling downstream of mTORC2 in sensory neurons [18,21].However, several questions still remain unanswered: Does Trc function presynaptically?How does it regulate synapse development?Does it function in the TSC pathway?Before proceeding further, we first confirmed our observation that trc is essential for synapse development using various genetic tests as described below.
Homozygous trc 1 mutants, a previously characterized loss-of-function point mutant allele [29], do not typically survive beyond the late second or early third-instar stage [22] and are also much smaller compared to age-matched wild-type (WT) larvae.This is likely because Trc regulates multiple processes that are important for survival [19].However, upon rearing the mutant flies at 18°C we were able to find occasional escapers that barely made it to the late third-instar stage.To test its role in NMJ development, we dissected trc 1 mutants and labeled them with antibodies against BRP, DGluRIII, and HRP (Fig 1).Comparing trc 1 mutants to WT larvae revealed a significant increase in the number of synaptic boutons at the neuromuscular junctions (NMJs).However, we may have underestimated the increase because the muscle area of trc 1 larvae is significantly smaller than that of age-matched WT larvae (S1 Fig) .Indeed, when normalized to the muscle area, trc 1 mutants have roughly three-fold increase in the number of synaptic boutons compared to the WT larvae (Fig 1A -1C).We also confirmed that the number of active zones (as marked by BRP) in trc 1 mutants was significantly more than the WT (S2 Fig) .To confirm that loss of trc was responsible for the increase in synaptic boutons, we performed a genetic complementation analysis wherein homozygous trc 1 mutants were crossed to an independent deficiency line in which the trc gene (trc Df ) is completely deleted.The resultant progeny (trc 1 /trc Df ) also exhibited an increase in the number of synaptic boutons (Fig 1A -1C), similar to the homozygous mutants.Thus, we concluded that Trc kinase is required to restrict the number of synaptic boutons at the NMJ.
Beside the increase in synaptic boutons, we also quantified other parameters of synapse growth that include, branching of axons and synaptic span.We found that there was a significant difference in the number of branches and synaptic span between trc mutants and WT (P<0.01)(S2 trc acts in the TSC pathway to restrict synapse number Similar increases in synaptic bouton numbers between TSC mutants and trc mutants, previous reports that mTORC2 interacts with trc [21]; and our previous data that suggest that TSC regulates synapse development mostly via mTORC2 [10], together suggested that Trc may function in the TSC pathway.An alternate possibility is that loss of trc or TSC pathway function might increase synaptic bouton numbers through independent mechanisms.To distinguish between these possibilities, we performed transheterozygous analysis between trc and Tsc2/TORC2/ Akt1 using the number of synaptic boutons as the readout.Unfortunately, we could not perform the double mutant analysis because single mutants of trc have very few larvae that can survive to the third-instar stages and the double mutant combination(s) are likely embryonic lethal because we do not find any homozygous larvae.Transheterozygous analysis is an alternate method to test if two genes affect the same pathway.We reasoned that if the transheterozygous combinations (e.g., Tsc2/+, trc 1 /+) led to significant increases in the number of synaptic boutons, it would suggest that trc and Tsc2 affect synapse growth via the same genetic pathway.
We independently crossed trc 1 mutants with Tsc2 (gig 109 [30], rictor (mTORC2), and Akt1 mutants respectively.Synaptic bouton numbers of the individual heterozygotes crossed to WT (Canton S (CS)) were used for comparison.We noted that unlike homozygous mutants, the transheterozygotes were comparable to the WT in overall size and muscle area (S1 Fig).

Presynaptic but not postsynaptic Trc kinase is required to regulate NMJ synapse number
Our previous data indicated that Tsc2 acts presynaptically to regulate synapse growth [10].In addition, Mob2, an interactor of Trc 1 , can regulate synapse growth presynaptically [22].These data suggest that the presynaptic pool of Trc may be required to regulate synapse number.To test this hypothesis, we knocked down Trc specifically in neurons or in muscles using tissuespecific Gal4 drivers [34] to drive Trc RNAi expression.Neuron-(BG380.Gal4 [35]) or muscle (G7.Gal4 [36])-specific drivers were used to drive the trc RNAi expression in presynaptic neurons or postsynaptic muscles, respectively.BG380.Gal4 or G7.Gal4 drivers crossed to WT flies served as their respective controls.Using RT-PCR, we confirmed that expression of trc RNAi in neurons and muscles decreased the trc transcripts to less than 20% of WT levels (n = 4 independent experiments with 3 larvae per genotype, p = 0.03).Presynaptic expression of two independent trc RNAi lines significantly increased the number of synaptic boutons (Fig 3A and  3B).On the other hand, when trc was knocked down in the muscles, synaptic bouton numbers were not significantly different than that of WT (Fig 3A and 3B).These data suggest that presynaptic, but not the postsynaptic, Trc kinase is likely responsible for regulating NMJ bouton numbers.
To confirm these observations, we performed rescue analyses, wherein we added back wild type Trc kinase in neurons or muscles in trc 1 mutants using the UAS-GAL4 system [34].This allowed us to test if adding back the pre-or postsynaptic Trc in trc mutants was sufficient to restore its normal function.A full-length wild type UAS-trc transgene [37] was expressed either pre-or postsynaptically in trc 1 mutants.WT and trc 1 mutant synapses were used as controls for the comparison of synaptic bouton numbers.As expected from the RNAi experiments, presynaptic overexpression of Trc kinase restored the number of synaptic boutons close to WT bouton numbers (Fig 3C and 3D).However, postsynaptic expression of the same transgene in a trc 1 mutant background failed to restore the number of synaptic boutons back to WT levels (Fig 3C and 3D).We noticed that neither pre-nor postsynaptic expression of trc was sufficient to overcome the severe lethality or small size of trc 1 mutant larvae, suggesting that loss of Trc function in other tissues may contribute toward the lethality observed in trc 1 mutant larvae.To test the localization of Trc at the NMJ, we stained NMJs of WT larvae with anti-Trc antibody [38].However, we could not detect any specific staining of Trc at the NMJ synapses.To test if Trc could localize to larval NMJs, we expressed a trc transgene tagged with GFP [39]

WASP interacts genetically with trc and Tsc2
Trc kinase and mTORC2 control polarized outgrowths in various developmental contexts by regulating actin polymerization [19,40,41].Interestingly, WASP, a potent regulator of actin polymerization [42], has also been implicated in regulating synapse development at the Drosophila NMJ [43][44][45].In addition, akin to trc 1 and other Tsc2 pathway mutants, wasp mutants also exhibit increased number of synaptic boutons [43,44].However, it is not known if WASP and Tsc2/Trc pathway act together to regulate synapse growth.To further explore the molecular mechanism of how Trc regulates synapse numbers, we asked if wasp interacted genetically with Tsc2/trc to regulate the number of synaptic boutons at NMJ.We first analyzed trc 1 /wasp 1 and trc 1 /Tsc2 transheterozygotes.Corresponding heterozygotes (Trc1/+ and wsp 1 /+) and WT larvae were used as controls for comparison.As expected, a 50% decrease in trc 1 (trc 1 /+), Tsc2 (Tsc2/+), or wasp 1 (wasp 1 /+) did not alter the number of synaptic boutons compared to WT.However, Tsc2/wasp and Trc/wasp double transheterozygotes exhibited significant increases in synaptic bouton numbers (Fig 4A -4D), suggesting that wasp interacts genetically with Tsc2 and trc to restrict the growth of synaptic boutons.
Synaptic WASP levels are reduced in trc, Tsc2, and Akt1 mutants WASP is localized to NMJ synapses in Drosophila [43], and it was previously demonstrated that reduced synaptic WASP levels leads to an increase in synaptic boutons [44,45].Therefore, we hypothesized that Trc might regulate synapse growth by modulating synaptic WASP levels.To test our hypothesis, we first labeled the NMJ synapses of trc 1 mutants and WT larvae with antibodies against HRP, Discs large [46] (Dlg, which largely labels postsynaptic sites on muscles [47]) and WASP [48].Consistent with previous data [43], WASP localized to pre-and postsynaptic compartments in both WT flies and trc 1 mutants.However, synaptic levels of WASP were dramatically reduced in trc 1 mutant synapses (Fig 5A and 5B) compared to WT synapses.These data suggest that Trc regulates synaptic WASP levels.Interestingly, we also found reduction in synaptic WASP in homozygous mutants of Tsc2 and Akt1 (Fig 5C and 5D) as well as Tsc2/ trc 1 transheterozygotes (Fig 5E ), which is consistent with the idea that Tsc2 and Trc may regulate synaptic WASP levels to restrict synapse growth.

Presynaptic loss of Trc kinase decreases synaptic WASP
We next sought to test if synaptic WASP levels correlate with number of synaptic boutons.To address that question, we compared the effects of specific reduction of trc in presynaptic versus postsynaptic compartments because only presynaptic trc knockdown resulted in increased bouton numbers.We specifically expressed trc RNAi presynaptically using BG380.GAL4 [35] or postsynaptically using G7.GAL4 driver [36].As seen in

Trc regulates the transcription of WASP
Trc kinase is a member of the Hippo pathway, whose final readout is Yorkie-mediated transcriptional regulation of genes [49].Therefore, we wondered if altered levels of WASP in trc mutants could be attributed to changes in transcription of wasp mRNA.To compare the levels of wasp mRNA in trc mutants to WT larvae, we performed reverse transcriptase-PCR (RT-PCR).We measured the levels of wasp mRNA extracted from trc mutants or wild type larval ventral nerve cords (VNC) and the body wall, which contains most of the muscle.We used 3 larvae from each genotype per experiment and performed 4 independent experiments to compare the levels of wasp mRNA (Fig 5F).Compared to WT, wasp mRNA levels were down to less than 25% in both the VNC and less than 14% in body walls (muscle) (n = 4 independent experiments, p<0.05) of the trc 1 mutants.These data indicate that Trc kinase may regulate transcription of WASP.

Presynaptic expression of wsp suppresses excess synaptic boutons in trc mutants
Previous studies have described a function for pre-and post-synaptic pools of WASP in regulating synapse development, presumably by independent mechanisms [43][44][45].Loss-of-function trc 1 mutants and presynaptic knockdown of Trc kinase exhibit excess synapses and reduced WASP levels in both pre-and post-synaptic compartments.Therefore, we wondered whether either or both pools of WASP may be essential for synaptic growth downstream of Trc.To test if TSC-Trc pathway regulated synapse growth via pre-or post-synaptic WASP, we overexpressed wasp presynaptically or postsynaptically using tissue-specific Gal4 drivers in trc 1 mutants or in trc 1 /Tsc2 transheterozygotes, both of which exhibited significant increases in synaptic bouton numbers (Figs 1 and 2).We reasoned that significant reversion to WT number of boutons following pre-or postsynaptic wasp overexpression would indicate that WASP function was necessary to restrict synapse growth in that compartment.In addition, this would also confirm that WASP acts downstream of Tsc2 and Trc kinase.On the other hand, if both pools of WASP need to function together to restrict synapses, we expected to see negligible or partial suppression in synaptic bouton numbers following the expression of wasp transgene in either neurons or muscles.We observed that presynaptic wasp overexpression in trc 1 mutants led to significant decrease in the number of synaptic boutons-almost similar to that of WT (Fig 7A and 7B)-but postsynaptic wasp overexpression did not (Fig 7A and 7B).Similar suppression was also observed in Tsc2/trc 1 transheterozygotes (Fig 7C and 7D).Finally, there was no change in the number of boutons when wasp was overexpressed in an otherwise WT background {WT = 18.5 ± 2.7; Overexpression of wasp in WT = 20.3 ± 3.5, n = 10}.Together, these data suggest that presynaptic WASP plays an essential role in restricting synapse growth at the NMJs via the TSC-Trc pathway (Fig 8).

Discussion
Synapse development is a highly coordinated process that is precisely programmed and regulated by various signaling pathways.Dysregulation of these processes leads to alterations in synaptic function and can result in cognitive impairments, including ASDs [2,50].TSC signals via the mTOR kinase to regulate various aspects of neuronal morphogenesis, including proliferation, autophagy and synapse development [5,10,11,51].However, the precise mechanisms by which the TSC-mTOR pathway regulates synapses remains to be elucidated.In this study, we identified a novel role for Trc kinase, a member of the AGC family kinases [20], in the TSCmediated growth of NMJ synapses.Although a previous study suggested that Trc kinase has a role in synapse regulation at the NMJs [22], its molecular mode of action remains unknown.In this report we demonstrate that Trc acts downstream of TSC-mTORC2 pathway.Moreover, we demonstrate that Trc restricts synaptic bouton numbers likely by decreasing the levels of presynaptic WASP, a known regulator of actin polymerization [42].

Trc kinase acts downstream of TSC-mTOR pathway
Recent studies have shown that mTORC2 can regulate Trc kinase in sensory neurons [18,52,53].Here, we suggest a novel functional link between the TSC-mTORC2 pathway and Trc kinase for the proper development of NMJ synapses.We demonstrate that Tsc2-mTORC2 largely acts through Trc kinase to regulate the levels of presynaptic WASP, presumably by affecting actin polymerization.
Trc/NDR1 kinases have been implicated in regulating polarized outgrowths in various developmental contexts, including budding in yeast, morphogenesis of sensory bristles, and dendritic patterning in Drosophila [19].While the mechanistic details of how NDR1 kinases act in different developmental contexts are not clear, it is known that Hippo and Furry regulate dendritic tiling pattern establishment in sensory neurons.Interestingly, neither hpo [54] nor fry [21] mutants exhibit the synaptic overgrowth phenotype that we observe in trc mutants (Natarajan and Wairkar, S1 Table and unpublished data).This is consistent with our hypothesis that non-canonical upstream activators and/or signaling pathways (TSC/mTOR) may regulate Trc kinase at NMJs.
Interestingly, a recent study identified a role for mob2, a gene encoding an NDR kinase activator, in NMJ development [22].Furthermore, it has been demonstrated that Mob2 physically interacts with Trc kinase in Drosophila epidermal hairs during development [38].So, it is possible that Mob2 rather than Fry is the activator of Trc at the NMJ synapses.Consistent with this idea, our data and a previous report 14 suggest that both Mob2 and Trc act presynaptically to regulate synapse growth.It would be interesting to test if Mob2 functions in the TSC pathway and whether Trc kinase can be directly activated by mTORC2, similar to its role in dendrite development [18].

Trans-synaptic regulation of WASP by Trc
Actin cytoskeleton is important in regulating synapse development, as well as for the formation and storage of long-term memory [55].A recent study suggests that mTORC2 might play a role in regulating long-term potentiation via regulation of the actin cytoskeleton [16].However, mechanism by which mTORC2 regulates actin cytoskeleton remains unknown.Our data suggest that Tsc2 somehow activates Trc kinase, which in turn, regulates the synaptic levels of the actin polymerization protein, WASP.Interestingly, presynaptic knockdown of Trc caused a significant decrease in total synaptic WASP levels and increased synapse numbers while postsynaptic WASP knockdown did not.Since WASP is present both presynaptically and postsynaptically, this raises an interesting possibility that Trc may regulate the levels of WASP transsynaptically.One of the best-studied anterograde pathways know to function at the fly NMJ synapse is the Wnt/wingless signaling pathway [56].Indeed, there are instances where the TSC pathway has been shown to interact with Wnt/wingless signaling pathway [57,58].Intriguingly, in at least in one of those cases, Tsc2 is known to act upstream of Wnt signaling pathway [57].

Role for Trc in muscle
While experimental evidence suggests that WASP may regulate synapses both pre- [43,44] and postsynaptically [45], our data suggest that Trc only requires presynaptic WASP to regulate fly NMJ synapses.So what might be the role for WASP in the muscles?Interestingly, studies in Drosophila and mice have shown that Neuronal-WASP (N-WASP) and Arp2/3-mediated actin polymerization are essential for myoblast fusion during myogenesis [59][60][61].Therefore, we speculate that perhaps the increased fragility of muscles that we observed in trc mutants is due to a reduction in postsynaptic WASP levels.Given the significantly smaller (S1 Fig) and extremely fragile muscles in trc 1 mutants, it is possible that postsynaptic Trc kinase may have an important hitherto unidentified function in the development of musculature.

How does Trc regulate WASP levels?
Trc could regulate WASP levels in the following ways: by affecting its transcription, translation or by influencing the overall stability or localization of WASP.The last alternative is less likely because although we observed a decrease in WASP levels, we did not find any evidence of gross mislocalization of WASP protein at the NMJ synapses.
So, how does Trc regulate WASP levels?Trc is a member of the Hippo pathway, which regulates processes such as cell growth via transcription of downstream targets [49].It does so via a protein called Yorkie (YAP in vertebrates), which upon activation, translocates to the nucleus to initiate specific transcription programs [49].
Our data shows that wasp transcripts are significantly decreased in trc mutants as compared to the WT, suggesting that Trc may regulate a transcriptional program for the regulation of WASP levels.Further studies are needed to test whether this regulation is achieved in a Yorkiedependent manner or via other independent mechanisms.

Upstream modulators and downstream effectors of Trc neurons
Our data suggest that neuronal Trc can be regulated by Tsc2 to regulate synapse development via WASP.Phosphatidylinositol 4,5-bisphosphate (PIP 2 ) and nervous wreck (Nwk) are other effectors that are known to influence synapse development via WASP.Both of them are thought to regulate presynaptic WASP leading to alterations in NMJ growth [43,44].Therefore, we asked if these proteins act via the Tsc2 pathway.Our preliminary data does not show any transheterozygous genetic interaction between Trc and Nwk or Tweek [62], an upstream regulator of PIP 2 (Natarajan and Wairkar, unpublished data), indicating that Nwk/PIP2 affect synapse growth independent of Tsc2-Trc pathway.However, absence of transheterozygous interactions are not conclusive and therefore, we cannot completely rule out an interaction between these genes.Further analysis is needed to confirm whether Nwk and Tweek interact with Trc.
What are the downstream targets of Trc?A recent chemical genetic screen, which sought to identify downstream regulators of Trc in dendrites, found that the mammalian homolog of Trc (NDR-1 kinase) regulates dendritic arborization and excitatory synaptic function via its substrates AAK1 (AP-2 associated kinase) and Rabin8, a GTP/GDP exchange factor of Rab8 [53].Interestingly, both these substrates appear to regulate distinct aspects of Trc-mediated dendritic growth-AAK1 is important for the regulation of dendritic branching, and Rabin8 is critical for spine morphogeneis.Our data add to this information and suggest that WASP is another potential downstream mediator of Trc signaling pathway, with perhaps, a specific role in synapse development.

Potential role of Trc kinase in neurodevelopmental disorders
The loss of Trc leads to abnormal dendritic tiling in flies, and it has also been implicated in dendrite maintenance [21].These functions of Trc seem to be conserved in mice because loss of NDR1 kinase leads to defects in spine morphogenesis and dendritic arborization [53], which are reminiscent of some neurodevelopmental disorders [63][64][65].These data support the idea that abnormal synapse development might be one of the major contributors to neurodevelopmental disorders.Previous studies [18,21,53,66] and our data support this hypothesis.We found that Trc plays a role in synapse development downstream of Tsc2, which has been strongly implicated in Autism Spectrum disorders [67,68].Thus, a growing body of evidence indicates that Trc/NDR1 kinase signaling maybe important in neurodevelopment, and future studies are likely to address the precise molecular mechanisms that regulate Trc kinase and its role in neurodevelopmental disorders.

Imaging and analysis
All the NMJ imaging was done at muscle 4, segment A2-A4.Confocal imaging and analysis were performed as described previously [26].Bouton numbers were manually counted and normalized to the muscle size.Although bouton quantifications for muscle 4 are shown in the figures, synaptic overgrowth was apparent and was qualitatively similar at all muscles observed, including muscle 6/7.NMJ labeling intensities were quantified with MetaMorph software (Molecular Devices, Sunnyvale, CA, USA).Student's t-tests were used to compare two samples, and one-way analyses of variance (ANOVAs) followed by Dunnett's or Tukey's multiple comparison tests were performed to identify statistically significant differences within a group.

RT-PCR Experiments
RNA was extracted using Trizol (Ambion, Life Technology) from wandering third instar larvae (3 larvae per sample, each experiment repeated four times) according to the manufacturer's instructions.For semi quantitative RT-PCR, 1μg of pure RNA of each genotype was reverse transcribed using Superscript III RT kit (invitrogen) primed with oligo dT.cDNA thus obtained was amplified using the following primers: Fig).Interestingly, only synaptic span (not branching) was significantly altered in Tsc2 mutants (S2 Fig).This suggests that Tsc2 may regulate some of the aspects of synapse growth via Trc other than the number of synaptic boutons (see below).

Fig 1 .
Fig 1. Trc kinase is required to restrict the number of synaptic boutons at the NMJ. A. Representative confocal stacks of muscle 4 NMJ synapses of wild type (WT), trc mutant (trc 1 ) and trc mutant crossed to an independent deficiency line (trc 1 /trc Df ) that deletes the entire trc gene.Larval NMJs were dissected and stained with anti-Bruchpilot (anti-BRP), anti-glutamate receptor (DGluRIII) and anti-horseradish peroxidase (HRP) antibodies.B & C. Quantification of bouton numbers and bouton number per muscle area of muscle 4 from the genotypes in A. One-way ANOVA followed by Tukey's post-hoc test was performed.n>10 and ***p<0.001.Error bars represent S.E.M. doi:10.1371/journal.pone.0138188.g001 Finally, to test whether Trc kinase is sufficient to regulate the synapse numbers, we overexpressed full-length Trc transgene presynaptically in a WT background.We did not find a significant difference between the numbers of synaptic boutons in WT larvae or WT larvae overexpressing Trc transgene presynaptically {WT = 22±3; WT with overexpression of presynaptic Trc = 23.4±2,n = 10, p = 0.4}.

Fig 3 ,
presynaptic knockdown of Trc kinase increased the number of synapses and it also resulted in significant reduction in synaptic WASP levels (Fig 6A and 6B).Conversely, postsynaptic knockdown of Trc did not significantly change the number of synaptic boutons (Fig 3) or decrease synaptic WASP levels (Fig 6A and 6B).Interestingly, decrease in synaptic WASP levels might be specific to synaptic growth via

Fig 5 .
Fig 5. Synaptic WASP levels are reduced in trc and Tsc2 mutant synaptic boutons.A. Representative confocal images of wild type (WT) and trc mutant synapses (trc 1 ) stained with anti-HRP, anti-Discs-large (Dlg) and anti-WASP antibodies.B. Quantification of total synaptic WASP levels in WT, trc 1 mutants and transheterozygote of trc 1 crossed to an independent deficiency line that deletes the entire trc gene (trc 1 /trc Df ).C. Representative confocal images showing NMJ synapses of wild type (WT) and Tsc2 mutants and Akt1 mutants stained with antibodies against DGluRIII and WASP.D. Quantification of the synaptic WASP levels measured in genotypes described in C. E. Quantification of the synaptic WASP levels in WT (CS), heterozygotes of Tsc2 and trc 1 (Tsc2/+, trc1/+) and in their respective transheterozygotes (Tsc2/trc1).One-Way ANOVA followed by Tukey's post-hoc test was performed.*p<0.01,**p<0.001,***p<0.0001.Error bars represent S.E.M. F. Representative gel showing wasp transcripts from ventral nerve cords (VNC) and muscles of WT and trc 1 mutants.Ribosomal protein 49 (RP49) is used as a control.doi:10.1371/journal.pone.0138188.g005

Fig 8 .
Fig 8. Schematic model of how Trc may regulate synapse development at Drosophila NMJ.Based on our data, Trc likely acts downstream of mTORC2 in the presynaptic compartment to regulate synaptic WASP, largely by regulating the transcription of WASP.Since WASP is known to regulate the polymerization of actin, we hypothesize that Trc ultimately regulates the actin cytoskeleton to regulate synapse development at the Drosophila NMJ synapses.doi:10.1371/journal.pone.0138188.g008