Neuropilin 1 sequestration by neuropathogenic mutant glycyl-tRNA synthetase is permissive to vascular homeostasis

The mechanism by which dominantly inherited mutations in the housekeeping gene GARS, which encodes glycyl-tRNA synthetase (GlyRS), mediate selective peripheral nerve toxicity resulting in Charcot-Marie-Tooth disease type 2D (CMT2D) is still largely unresolved. The transmembrane receptor protein neuropilin 1 (Nrp1) was recently identified as an aberrant extracellular binding partner of mutant GlyRS. Formation of the Nrp1/mutant GlyRS complex antagonises Nrp1 interaction with one of its main natural ligands, vascular endothelial growth factor-A (VEGF-A), contributing to neurodegeneration. However, reduced extracellular binding of VEGF-A to Nrp1 is known to disrupt post-natal blood vessel development and growth. We therefore analysed the vascular system at early and late symptomatic time points in CMT2D mouse muscles, retina, and sciatic nerve, as well as in embryonic hindbrain. Mutant tissues show no difference in blood vessel diameter, density/growth, and branching from embryonic development to three months, spanning the duration over which numerous sensory and neuromuscular phenotypes manifest. Our findings indicate that mutant GlyRS-mediated disruption of Nrp1/VEGF-A signalling is permissive to maturation and maintenance of the vasculature in CMT2D mice.

development by promoting extracellular matrix remodeling downstream of integrins 19 and by suppressing excessive TGF-β signalling 20 . Although mutating the VEGF-A-binding site of Nrp1 does not phenocopy the severity of full Nrp1 knockout mice, it causes a post-natal impairment in angiogenesis and arteriogenesis 21,22 , resembling the milder phenotype of mice expressing VEGF-A isoforms lacking the Nrp1-binding site 23 . This work suggests that Nrp1 has both VEGF-dependent and VEGF-independent functions in blood vessels, and that Nrp1/VEGF-A signalling is perhaps more important for vascularisation post-than pre-birth 21 . Consistent with this, blocking VEGF-A binding to Nrp1 with a monoclonal antibody impairs vascular remodeling in the peri-natal mouse retina 24 .
Given that 1) mutant GlyRS competes with VEGF-A for Nrp1 binding 10 , 2) GlyRS is secreted 7, 10, 25 and found circulating in serum of humans and mice 25 , and 3) vascular impairment can contribute to neurodegeneration 26 , we decided to assess the effect of mutant GlyRS on the vascular system of the Gars C201R/+ murine model of CMT2D.

Results
GlyRS C157R aberrantly associates with Nrp1. Six different mutations in GARS have been shown to alter protein conformation -five CMT2D-associated mutants 9 and P234KY corresponding to murine P278KY 10 , which is the spontaneous mutation found in Gars Nmf249/+ , the more severe CMT2D mouse model 4 . Neomorphic Nrp1 binding and disturbance of Nrp1/VEGF-A signalling has been confirmed for a number of GlyRS mutants, including P234KY 10 . In order to verify a common mechanism in the mild Gars C201R/+ mouse 27 , we performed co-immunoprecipitation assays in NSC-34 cells expressing V5-tagged wild-type, P234KY, L129P, and C157R (human equivalent of C201R) GlyRS. GlyRS L129P was included because it is one of the GARS mutants with the best evidence linking it to human neuropathy 28 . As expected, wild-type GlyRS did not bind Nrp1, while GlyRS C157R did, albeit at a lower affinity than the other two mutant proteins (Fig. 1B). The intensity of the aberrant interaction between mutant GlyRS and Nrp1 thus correlates with the overall severity of the mouse model. These experiments confirm that C157R GlyRS spuriously binds to Nrp1 and, extrapolating from findings in the Gars Nmf249/+ mouse 10 , suggest that the antagonisation of Nrp1/VEGF-A signalling could contribute to the neuropathic phenotype of Gars C201R/+ mice. Nrp1 is highly expressed in blood vessels of skeletal muscles but not motor neurons. Nrp1 is widely expressed in a range of different tissues 29 , localising to the vascular system 30, 31 and many neuronal types, including motor neurons [32][33][34] . As mutated GlyRS binds to Nrp1 impacting the motor nervous system of Gars mice 10 , we examined the expression pattern of Nrp1 in wholemount distal, fast-twitch lumbrical and proximal, slow-twitch transversus abdominis (TVA) muscles from one and three month old wild-type mice. These muscles are thin and flat, permitting the visualisation of the entire vascular and nervous systems within the tissue 35,36 . In these muscles at both time points, Nrp1 (white) did not specifically localise to SV2/2H3 + (green) motor neurons or S100 + (blue) Schwann cells ( Fig. 2A,B). Rather, its expression consistently coincided with the endothelium-binding isolectin B 4 (IB4, red) (Figs 2C and S1A,B). Nrp1 could be observed faintly surrounding motor neurons, but the staining was wider than that of both S100 (Fig. 2C) and myelin basic protein (Mbp, yellow) (Fig. S1C), suggestive of localisation to blood vessels or closely juxtaposed extracellular matrix encasing the motor neurons and Schwann cells. These results indicate that Nrp1 predominantly localises to the vascular system in post-natal skeletal muscle. and one of its principal co-receptors, VEGFR2, which together bind to the secreted glycoprotein VEGF-A 165 (on the left) 11 . VEGF-A functions in vasculogenesis, angiogenesis, and arteriogenesis, with the latter two processes occurring through Nrp1 and VEGFR2 signalling 17,18 . VEGF-A is also critical for nervous system development and maintenance 15 . VEGF-A 165 binding to Nrp1 is mainly dependent by the b1 domain 12 . CMT2D-associated mutations in GARS cause a conformational opening of GlyRS, allowing the aberrant binding of mutant GlyRS to the b1 domain of Nrp1 (on the right) 10 . This competitively antagonises Nrp1/VEGF-A signalling, which contributes to motor deficits observed in CMT2D mice 10 . Schematics are not drawn to scale and adapted from Plein et al. 18 and He et al. 10 . (B) Co-immunoprecipitation of endogenous Nrp1 showing aberrant interaction with P234KY, L129P, and C157R GlyRS (ectopically expressed with a V5-tag) in NSC-34 cells. Wild-type GlyRS shows no significant binding to Nrp1. The aberrant interaction is weaker with C157R than with P234KY GlyRS, which correlates with the severity of CMT phenotypes in Gars C201R/+ and Gars Nmf249/+ mice, respectively.
SCIeNTIfIC RepoRts | 7: 9216 | DOI:10.1038/s41598-017-10005-w Blood vessels in Gars C201R/+ muscle are unaffected. Having shown that C157R mutant GlyRS binds to Nrp1 in vitro (Fig. 1B), we assessed structural features of the vascular bed in wild-type and Gars C201R/+ muscles using IB4 staining (green, Fig. S2). Confocal Z-stack images were taken throughout the entire depth of one and three month old lumbrical and TVA muscles (Fig. 3A,B), and capillary diameter (Fig. 3C,D), density (Fig. 3E,F), and branching (Fig. 3G,H) assessed. These time points represent early and late symptomatic stages of disease in Gars C201R/+ mice, and were chosen to complement previously performed, in-depth motor and sensory phenotypic analyses 8,37 . We saw no significant difference between wild-type and mutant Gars capillary beds in any of the parameters analysed, suggesting that vasculature is unimpaired in fast and slow twitch skeletal muscles of Gars C201R/+ mice. The capillary diameter was consistent across muscles and varied little over time, while capillary and branching densities showed variability between muscles, and appeared to decline with age, presumably due to muscle growth. These differences confirm the suitability of this strategy for detecting changes in small blood vessel structure.
Disruption of VEGF-A binding to Nrp1 has previously been linked with reduced Nrp1 expression in various different mouse tissues 21 . We therefore looked at Nrp1 (white) and VEGFR2 (red) protein levels and localisation in lumbrical and TVA muscles, but found no obvious discrepancies between genotypes (Fig. 4A). This was confirmed at the mRNA level in disease-susceptible 27 tibialis anterior muscles using quantitative RT-PCR (qPCR) with either glyceraldehyde 3-phosphate dehydrogenase (Gapdh) or endothelium-specific platelet endothelial cell adhesion molecule 1 (Pecam1) as the reference gene (Fig. 4B). Total VEGF-A expression was not significantly increased (Fig. 4B). To rule out the possibility that alternative pathways integral to vascular homeostasis may be contributing to the maintenance of mutant Gars blood vessel stability, we also assessed the expression of Nrp2 38 , Figure 2. Nrp1 localises to skeletal muscle blood vessels but not motor neurons. Representative Nrp1 staining in one month old wild-type lumbrical muscles. (A) Nrp1 (white) localises to structures surrounding, but not perfectly overlapping, lower motor neurons (SV2/2H3, green), and in SV2/2H3 negative areas. α-bungarotoxin (α-BTX, magenta) identifies post-synaptic acetylcholine receptors at the neuromuscular junction. (B) Nrp1 staining is not wholly contiguous with the Schwann cell marker S100 (blue) either. This was confirmed with a second glial cell marker, myelin basic protein (Mbp) in Fig. S1C. (C) Nrp1 co-localises with IB4 (red), an endothelial cell marker, indicating that Nrp1 is found in muscle blood vessels. Arrows highlight Nrp1 staining associated with motor neurons/myelin, while arrowheads indicate Nrp1 + blood vessels. A similar staining pattern was observed at three months and in the TVA muscle at both time points (data not shown). A-B are collapsed Z-stack images and C is a single plane image. Scale bars = 20 μm. See also Fig. S1. total and soluble VEGFR1 (also known as Flt1 and sFlt1, respectively) 14 , and genes integral to angiopoietin-Tie signalling (Angpt1, Angpt2, Tie1, and Tie2) 39 . We saw no difference between genotypes in the expression of any of these genes (Fig. 4C), indicating that genetic compensation is unlikely to be playing a crucial role in the maintenance of Gars C201R/+ vasculature. GlyRS C157R /Nrp1 binding is permissive to vascular development and homeostasis. Nrp1 is indispensable for vascularisation of the mouse central nervous system, but is thought to be less critical, although still important 40 , for the vasculature of other tissues such as muscle 41 . In vivo disruption of VEGF-A binding to Nrp1 has previously been shown to disturb blood vessel growth and patterning of the mouse retina 21,22 . We therefore assessed the IB4 + (green) capillary network in one and three month old retinas (Fig. 5A) using an approach similar to that implemented in the muscle analyses (Fig. 3). Gars mice had similar retinal capillary diameters (Fig. 5B), densities (Fig. 5C), and branching (Fig. 5D) as wild-type animals at both time points. There was also no difference in the number of major radial branches (arteries and veins) emanating from the central retina (Fig. 5E). Similar to skeletal muscles, no obvious differences in Nrp1 and VEGFR2 staining were seen in mutant Gars retinas at either time point (data not shown).
Fly and mouse models of CMT2D display early developmental defects in the nervous system 7,8,10 . We therefore assessed parameters of the capillary network in E13.5 hindbrains of wild-type and Gars C201R/+ mice. Mutant brains showed no differences in blood vessel structures between genotypes (Fig. 6A,C-E), nor in Nrp1 (B,C) qPCR analysis of a series of genes integral to the maintenance of blood vessels indicates very little perturbation in gene expression in one month Gars C201R/+ tibialis anterior muscles. This was confirmed using either Gapdh (top graphs) or endothelium-specific Pecam1 (bottom graphs) as the reference gene. All individual gene data sets were separately analysed with unpaired, two-tailed t-tests. n = 6 wild-type and 4 Gars C201R/+ . and VEGFR2 staining (data not shown), which is consistent with western blotting data showing that Nrp1 and VEGFR2 protein levels remain unaltered in embryonic Gars neural tissues 10 . This lack of an embryonic vascular phenotype mirrors the result in mice with deficient VEGF-A binding to Nrp1 22 . Finally, we assessed vascular density in sectioned sciatic nerves from one month old mice, in order to determine whether post-natal neuronal tissue was perturbed. Anti-Pecam1 (green) was used instead of IB4 due to superior staining of the vasculature in this tissue (Fig. 6B). Mutant mice showed a non-significant trend towards increased capillary density (Fig. 6F); however, this might simply be caused by the reduced axon calibres (without axon loss) of mutant mice 27 , i.e. the capillaries command a greater percentage area of the sciatic nerve due to the axons being smaller.

Discussion
Mutant GlyRS appears to compete with VEGF-A for extracellular binding to the transmembrane receptor protein Nrp1 contributing to the peripheral nerve pathology observed in CMT2D 10 . Reduced binding of VEGF-A to Nrp1 has previously been shown to impair post-natal angiogenesis and arteriogenesis 21,22 . We thus set out to determine whether the capillary network of mutant Gars mice is altered either in the post-natal period or during development. First, we confirmed that the human orthologue of murine GlyRS C201R (GlyRS C157R ), but not wild-type, was capable of aberrantly binding to Nrp1 (Fig. 1B). This interaction is weaker than for GlyRS P234KY , which is expressed in Gars Nm249/+ mice, indicating that mutant disease severity correlates with neomorphic binding affinity. We then found that Nrp1 was highly expressed in post-natal skeletal muscle blood vessels, but not in motor neurons (Fig. 2); in spite of previously published results 43 . This discrepancy may be due low Nrp1 abundance in motor neurons compared to blood vessels and/or differential Nrp1 expression between muscle types (e.g. gastrocnemius versus TVA and lumbricals).
The fundamental requirement for Nrp1 in motor nervous system development is undisputed [32][33][34]44 , but its post-natal function and localisation remains less well defined. Motor neuron-specific ablation of Nrp1 using the Olig2 promoter is reported to cause post-natal motor axon loss and muscle atrophy 45 , but Nrp1 was deleted developmentally 46 rather than at or post-birth. It is possible that low levels of post-natal Nrp1 expression in motor neurons allows aberrant mutant GlyRS binding and signalling sufficient to drive peripheral nerve degeneration in CMT2D mice. However, given that embryonic defects are observed in Gars mice 8,10 , and that Nrp1 expression is developmentally downregulated in neuronal tissue 47, 48 , we should not rule out that mutant GlyRS-mediated inhibition of Nrp1/VEGF-A signalling during development may be predisposing the motor system to the subsequent degeneration observed at later, post-natal stages 37,49 .
Given the high levels of Nrp1 expression present in the vasculature of skeletal muscles (Fig. 2), and previously reported in retinas and embryonic hindbrains 30, 31 , we decided to assess capillary architecture in a range of tissues covering the developmental to late symptomatic period. Capillary diameter, density, and branching were unaltered by mutant GlyRS in both distal and proximal skeletal muscles (Fig. 3), as was expression of numerous genes that play a key role in vascular homeostasis (Fig. 4). Lack of a vascular phenotype was replicated in adult retinas (Fig. 5A-E), one month sciatic nerves (Fig. 6F), and embryonic hindbrains (Fig. 6C-E), while P6-9 retinal defects caused by mutating the VEGF-A binding site of Nrp1 21, 22 were also not present in mutant Gars vasculature (Fig. 5F-H). We have therefore shown that the vascular system is unaffected in Gars C201R/+ mice from embryonic development to adulthood. Extra-neuronal tissue pathology and disease mechanisms have been reported in mouse models of several peripheral nerve conditions including spinal muscular atrophy 50 , Kennedy's disease 51,52 , and amyotrophic lateral sclerosis 53 , appearing to mirror the human conditions, at least in some of their most debilitating incarnations 54 . Nevertheless, we have convincingly ruled out that vascular pathology extends to CMT2D, which is in keeping with the clinical presentation of patients.
So why does antagonising Nrp1/VEGF-A signalling affect the peripheral nervous system, but not the vascular system, when this pathway is vital for the functioning of both? First of all, timing and location of Nrp1 and GlyRS expression could conspire to ensure neuronal specificity. GlyRS is secreted and present in serum, but it may not be found in proximity of Nrp1 on blood vessels. For instance, if GlyRS is secreted by terminal Schwann cells at the neuromuscular synapse, mutant GlyRS may only gain access to Nrp1 on motor nerve terminals. Alternatively, the lack of a Gars vascular phenotype may reflect that blood vessels require lower levels of Nrp1/VEGF-A signalling than neurons. Finally, mutant GlyRS may be competing with VEGF-A for Nrp1 binding and simply impinging upon VEGF-A-initiated signalling events integral to the nervous, but not vascular, system. Given that VEGF-A is capable of regulating the two vascular-specific processes of angiogenesis and arteriogenesis via distinct pathways 40 , this latter scenario is highly conceivable. Indeed, it is corroborated by the previous observation that VEGF-A-deficient mice display selective degeneration of motor neurons 16 .
In summary, we have demonstrated that CMT2D mice display a pathological phenotype restricted to the nervous system, and that GlyRS-mediated disruption of Nrp1/VEGF-A signalling appears to be permissive to capillary maturation and maintenance. This is consistent with mutant GlyRS binding to Nrp1 affecting a nervous system-specific function of the Nrp1/VEGF-A pathway. Co-immunoprecipitation and western blotting. For cell lysate preparation, 36 h post-transfection NSC-34 cells were washed twice in phosphate-buffered saline (PBS, 137 mM NaCl, 10 mM Na 2 HPO 4 , 2.7 mM KCl, 1.8 mM KH 2 PO 4 ), scraped into PBS, pelleted, re-suspended in Pierce IP Lysis Buffer (Thermo Fisher, 87787) for 30 min, and the extract cleared for 7 min at 12,000 × g. Protein G beads (Invitrogen) were pre-incubated with rabbit anti-Nrp1 (Abcam, Cambridge, UK, ab81321) or rabbit IgG (Cell Signaling Technology, Danvers, MA) for 30 min before mixing with cell lysates overnight. The beads were then washed three times with buffer (100 mM NaCl, 50 mM Tris-HCl, pH 7.5, 0.1% (v/v) Triton X-100, 5% (v/v) glycerol), and immunoprecipitates probed by western blot. Immunoprecipitates were fractionated by 4-12% (v/v) Bis-Tris-Plus SDS-PAGE gels (Invitrogen) and transferred to PVDF membranes using the iBlot Dry Blotting System (Invitrogen). Membranes were blocked for 1 h with Tris-buffered saline with 0.1% (v/v) Tween 20 containing 5% (w/v) non-fat dry milk. Wild-type and mutant GlyRS proteins were detected using mouse anti-V5 (1/3000, Invitrogen, R960CUS) and Nrp1 using the same antibody as for co-immunoprecipitation (1/1000). After primary antibody incubation, membranes were washed and probed with HRP-conjugated anti-mouse or anti-rabbit secondary antibodies (Cell Signaling Technology), followed by detection using ECL western blotting substrate (Thermo Fisher) using the FluorChem M imager (ProteinSimple, San Jose, CA).

NSC
Animals. Gars C201R/+ mice were maintained as heterozygote breeding pairs on a predominantly C57BL/6 background and genotyped as described previously 27 . Mice sacrificed at one and three month time points were 28-34 and 89-97 days old, respectively. Multiple tissues were simultaneously harvested from both males and females. Mouse handling and experiments were performed under license from the UK Home Office in accordance with the Animals (Scientific Procedures) Act (1986), and approved by the University College London (UCL) -Institute of Neurology Ethics Committee.
Tissue preparation and immunohistochemistry. All steps were performed at room temperature, apart from overnight incubations conducted at 4 °C. Lumbrical and TVA muscles were dissected and immunohistochemically labelled as previously described [35][36][37] . Eyes were removed and fixed in 4% (w/v) paraformaldehyde (PFA, Electron Microscopy Sciences, Hatfield, PA) for 2 h, before retinas were dissected and stained as reported previously 55 . Sciatic nerves were dissected from mice transcardially perfused with 4% PFA, post-fixed for 2 h, and 10 μm sections processed and stained as previously described 8 . E13.5 hindbrains were dissected and stained using published protocols 56 .