Plexin-Semaphorin Signaling Modifies Neuromuscular Defects in a Drosophila Model of Peripheral Neuropathy

Abstract

Dominant mutations in GARS, encoding the ubiquitous enzyme glycyl-tRNA synthetase (GlyRS), cause peripheral nerve degeneration and Charcot-Marie-Tooth disease type 2D (CMT2D). This genetic disorder exemplifies a recurring paradigm in neurodegeneration, in which mutations in essential genes cause selective degeneration of the nervous system. Recent evidence suggests that the mechanism underlying CMT2D involves extracellular neomorphic binding of mutant GlyRS to neuronally-expressed proteins. Consistent with this, our previous studies indicate a non-cell autonomous mechanism, whereby mutant GlyRS is secreted and interacts with the neuromuscular junction (NMJ). In this Drosophila model for CMT2D, we have previously shown that mutant gars expression decreases viability and larval motor function, and causes a concurrent build-up of mutant GlyRS at the larval neuromuscular presynapse. Here, we report additional phenotypes that closely mimic the axonal branching defects of Drosophila plexin transmembrane receptor mutants, implying interference of plexin signaling in gars mutants. Individual dosage reduction of two Drosophila Plexins, plexin A (plexA) and B (plexB) enhances and represses the viability and larval motor defects caused by mutant GlyRS, respectively. However, we find plexB levels, but not plexA levels, modify mutant GlyRS association with the presynaptic membrane. Furthermore, increasing availability of the plexB ligand, Semaphorin-2a (Sema2a), alleviates the pathology and the build-up of mutant GlyRS, suggesting competition for plexB binding may be occurring between these two ligands. This toxic gain-of-function and subversion of neurodevelopmental processes indicate that signaling pathways governing axonal guidance could be integral to neuropathology and may underlie the non-cell autonomous CMT2D mechanism.

Keywords: Charcot-Marie-Tooth disease type 2D (CMT2D); GARS; aminoacyl-tRNA synthetase (ARS); axonal guidance; distal spinal muscular atrophy type V (dSMA-V); glycyl-tRNA synthetase; neurodevelopment; neuromuscular disease.

Figure 1

gars^P234KY expression phenocopies plexin mutants and subverts plexin-mediated axonal branching. (A) Schematic of the larval ventral body wall muscles in one hemisegment, showing the transverse nerve (TN), and branches of segmental nerves b (SNb) and d (SNd). The ectopic synaptic contacts often observed on muscles 6 and 7 in gars^P234KY flies are shown in red. (B) In wild-type flies, SNb innervates muscle 12, while the TN entirely bypasses muscles 6, 7, 12 and 13. (C,D) Ectopic axons (arrows) and synaptic contacts can be observed from the SNb and the TN when gars^P234KY is expressed ubiquitously (C) or in muscle (D), phenocopying loss-of-function plexB homozygotes (Carrillo et al., 2010). (E) Ubiquitous (1032-GAL4) and muscle (MHC-GAL4) gars^P234KY expression, but not mutant neuronal (elav-GAL4) or ubiquitous wild-type gars expression, lead to an increased number of ectopic contacts in L3 larvae. This is not seen in a model of spinal muscular atrophy (smn^x7/smn^x7), a second, unrelated neuromuscular condition. Ectopic branches are scored from both the TN and the SNb nerve. (F) Expressing gars^P234KY with a ubiquitous driver in either a plexA or plexB heterozygous knockout background significantly enhances the branching defects from the TN nerve. *P < 0.05, **P < 0.01, Dunn’s multiple comparison test. N.b., gars^P234KY is not expressed in the following flies: control, Smn^x7/Smn^x7, plexA/+, plexB/+, Smn^x7/Smn^x7; plexA/+ and Smn^x7/Smn^x7; plexB/+ (E,F). Scale bars = 10 μm. For all experiments, n > 16 larvae per genotype. Error bars represent ± standard error of the mean (SEM). See also Supplementary Figure S2.

Figure 2

Plexin mutants modify glycyl-tRNA synthetase (GlyRS)^P234KY-associated viability and motor defects. (A) Expressing gars^P234KY with a ubiquitous driver (1032-GAL4) in a plexB heterozygous loss-of-function background significantly alleviates adult viability defects, while plexB overexpression reciprocally causes full lethality. Contrastingly, gars^P234KY expression in a plexA loss-of-function background leads to a decrease in viability, while its overexpression has no effect. (B) gars^P234KY expression in plexB heterozygous larvae partially rescues muscle contractions, whilst gars^P234KY expression in a plexA background enhanced the motor defect. plexA overexpression partially restores movement, while plexB overexpression enhances the motor defect. The larval motor defects are thus analogous to the adult viability defects in (A). *P < 0.05, **P < 0.01, ***P < 0.001 Dunn’s multiple comparison test. N.b., gars^P234KY is not expressed in the plexA/+ and plexB/+, control flies (third and fourth bars from the y-axes, respectively). For viability assays, n > 100 flies per genotype; for contraction assays, n > 20 larvae per genotype. Error bars represent ± SEM.

Figure 3

plexA and plexB dosage changes modify GlyRS^P234KY-associated axonal defects and mutant GlyRS accumulation at the neuromuscular junction (NMJ). (A–C) Digitally inverted images of the TN showing increased axonal bruchpilot (Brp) compared to control (A) upon ubiquitous (B) and muscle-specific (C) gars^P234KY expression. (D) The build-up of Brp in axons of the TN caused by ubiquitous GlyRS^P234KY expression is modulated by alterations in PlexA or PlexB levels. Greater Brp accumulation is observed upon PlexA reduction and PlexB overexpression, whereas an increase in PlexA and a decrease in PlexB ameliorate Brp build-up. N.b., gars^P234KY is not expressed in the plexA/+ and plexB/+ control flies (third and fourth bars from the y-axis, respectively). (E) An example of GlyRS^P234KY (HA staining) accumulation at the presynapse when ubiquitously expressed. (F) GlyRS^P234KY accumulates at the neuronal membrane, and correlates with PlexB, but not PlexA, levels, i.e., increased and decreased abundance of PlexB causes increased and decreased mutant GlyRS build-up, respectively. (G) When expressed with a ubiquitous driver, PlexB localizes at the NMJ with the neuronal marker HRP. (H) PlexB and GlyRS^P234KY localize to the NMJ when co-expressed using a ubiquitous driver. *P < 0.05, **P < 0.01, ***P < 0.001 Dunn’s/Bonferroni’s multiple comparison test. #: displayed a significant increase when compared to the mutant GlyRS background alone (Mann-Whitney U test, P = 0.03). All experiments are performed expressing gars^P234KY with a ubiquitous driver (1032-GAL4). Scale bar = 10 μm. For all experiments, n > 20 larvae per genotype. Error bars represent ± SEM. See also Supplementary Figure S4.

Figure 4

Overexpression of the Plexin B ligand Semaphorin-2a (Sema2a) suppresses mutant GlyRS-mediated neuropathology. (A) Expressing Sema2a, but not Sema1a, in gars^P234KY flies (using the muscle driver MHC-GAL4) significantly alleviates mutant motor defects (compare gars^P234KY with gars^P234KY, Sema2a OE). In the plexB heterozygous loss-of-function background, this amelioration is further improved (compare gars^P234KY, Sema2a OE with gars^P234KY, Sema2a OE; plexB/+). In contrast, reduction of plexA levels abrogates the amelioration caused by Sema2a overexpression in gars^P234KY flies (compare gars^P234KY, Sema2a OE; plexA/+ with gars^P234KY and gars^P234KY, Sema2a OE). (B,C) Consistent phenotype correlations were seen for both axonal Brp build-up (B) and synaptic accumulation of GlyRS^P234KY (C). *P < 0.05, **P < 0.01, ***P < 0.001 Dunn’s/Bonferroni’s multiple comparison test. N.b., gars^P234KY is not expressed in the Sema1a OE and Sema2a OE control flies (third and seventh bars from the y-axes, respectively). All experiments are performed expressing gars^P234KY with a ubiquitous driver (1032-GAL4). For all experiments, n > 20 larvae per genotype. Error bars represent ± SEM. See also Supplementary Figure S3.

Figure 5

Synaptic accumulation of mutant GlyRS correlates with neuropathology. (A,B) Linear regression analyses identifying correlations between synaptic mutant GlyRS build-up with both muscular contractions (A, r² = 0.727; P = 0.0015) and axonal Brp build-up (B, r² = 0.945; P < 0.001).