Wnt4 participates in the formation of vertebrate neuromuscular junction

Abstract

Neuromuscular junction (NMJ) formation requires the highly coordinated communication of several reciprocal signaling processes between motoneurons and their muscle targets. Identification of the early, spatially restricted cues in target recognition at the NMJ is still poorly documented, especially in mammals. Wnt signaling is one of the key pathways regulating synaptic connectivity. Here, we report that Wnt4 contributes to the formation of vertebrate NMJ in vivo. Results from a microarray screen and quantitative RT-PCR demonstrate that Wnt4 expression is regulated during muscle cell differentiation in vitro and muscle development in vivo, being highly expressed when the first synaptic contacts are formed and subsequently downregulated. Analysis of the mouse Wnt4⁻/⁻ NMJ phenotype reveals profound innervation defects including motor axons overgrowing and bypassing AChR aggregates with 30% of AChR clusters being unapposed by nerve terminals. In addition, loss of Wnt4 function results in a 35% decrease of the number of prepatterned AChR clusters while Wnt4 overexpression in cultured myotubes increases the number of AChR clusters demonstrating that Wnt4 directly affects postsynaptic differentiation. In contrast, muscle structure and the localization of several synaptic proteins including acetylcholinesterase, MuSK and rapsyn are not perturbed in the Wnt4 mutant. Finally, we identify MuSK as a Wnt4 receptor. Wnt4 not only interacts with MuSK ectodomain but also mediates MuSK activation. Taken together our data reveal a new role for Wnt4 in mammalian NMJ formation that could be mediated by MuSK, a key receptor in synaptogenesis.

Figure 1. Wnt4 expression during neuromuscular junction development.

(A) Table showing results of Affymetrix microarrays data comparing relative Wnt4 mRNA expression during myotube differentiation, between stages T1/T2 and T2/T3 (see Materials and Methods). Relative Wnt4 mRNA is upregulated more than three fold between stage T1/T2 and downregulated more than one fold between stages T2/T3. (B and C) Real time RT-PCR quantification of relative Wnt4 mRNA expression during myotube differentiation (B, stages T1, T2 and T3) and hind limb development (C, embryonic stages E13.5, E14, E16 and newborn mice P0, N = 6 embryos tested for each stage). Relative Wnt4 mRNA expression is significantly increased between stages T1/T2 and further downregulated between stages T2/T3 and decreases as the limb developed. (D) Real time RT-PCR quantification of relative MuSK and Wnt4 mRNA expression in synaptic and extrasynaptic regions of diaphragms from stage E18.5 embryos. Relative MuSK and Wnt4 expression are three and two fold increased in synaptic compared to extrasynaptic regions respectively. Results are represented as relative expression (2^−ΔCt versus reference gene ×100, N = 3). (E and F) In situ hybridization with probes for Wnt4 mRNAs in E11.5 and E13.5 spinal cord sections (thoracic level) of wild type mice embryos (N = 3 embryos tested for each condition). Wnt4 mRNA is expressed in the floor plate and dorsal spinal cord but not in motoneurons (MN). Error bars show means ± SEM from three independent experiments. *P<0.05; **P<0.001; Mann-Whitney U test. Scale bar: in E, 20 µm for E and F.

Figure 2. Aberrant neuromuscular junction innervation in muscles of Wnt4−/− embryos.

(A–F) Confocal images of whole mount diaphragm (A and B), intercostal muscles (C and D) or cross sections of hind limb muscles (E and F) from stage E18.5 control littermates (wild type, A, C and E) or Wnt4−/− embryos (B, D and F) stained with neurofilament (NF, red) and synaptophysin (Syn, red) antibodies together with α-bungarotoxin (AChRs, green). Examples of nerve terminals passing through and projecting beyond the central band of AChR clusters in mutant diaphragm or intercostal muscles are indicated by white arrows in the merged image in B and D. Examples of non innervated synapses in mutant limb muscles are indicated by white stars in the merged image in F. (G) Measurement of AChR endplate band width, AChR clusters surface, α-bungarotoxin fluorescence signal intensity (numbers of AChR clusters tested: 95 in control and 76 in Wnt4−/−), AChR cluser number and number of non innervated synapses (%) in limb muscle cross sections (numbers of synapses counted: 35 in control and 28 in Wnt4−/−; N = 3 for Wnt4 mutants and N = 4 for control littermates embryos). Error bars show means ± SEM. *P<0.05; **P<0.001; Mann-Whitney U test. NS, non significant. Scale bars: in the merged image in A, 60 µm for A and B; in the merged image in D, 30 µm for C, D, E and F.

Figure 3. Synaptic markers are localized at the NMJ in Wnt4−/− embryos.

(A–F) Hind limb muscle cross sections from stage E18.5 control littermates (A–C) or Wnt4−/− embryos (D–F) stained with AChE (red, A and D), MuSK (red, B and E) or rapsyn (red, C and F) antibodies together with α-bungarotoxin (AChR, green). AChE, MuSK and rapsyn colocalized with AChR at the NMJ of wild type and Wnt4−/− mutant embryos (15 cross sections from 2 Wnt4 mutants and control littermates were analyzed for each condition). (G) Confocal images of hind limb muscles cross sections from stage E18.5 Wnt4−/− embryos stained with neurofilament (NF, red), synaptophysin (Syn, red) and rapsyn (blue) antibodies together with α-bungarotoxin (AChRs, green). Examples of innervated and non innervated synapses are indicated by white stars and arrowhead respectively. Non innervated synapses still expressed the rapsyn protein (15 cross sections from 2 for Wnt4 mutants and control littermate embryos). Scale bar: in the merged image in A, 20 µm.

Figure 4. Wnt4 affects muscle prepatterning and AChR clustering in muscle cells.

(A and B) Confocal images of whole mount intercostal muscles from stage E14 control littermates (wild type, A) or Wnt4−/− embryos (B) stained with neurofilament (NF, red) and synaptophysin (Syn, red) antibodies together with α-bungarotoxin (AChRs, green). Both in wild type and Wnt4−/− mutant embryos, AChR clusters were detected (N = 2 for Wnt4 mutants and N = 2 for control littermate embryos). (C) Quantification analysis of the AChR endplate band width and number of prepatterned AChR clusters. (D and E) Hind limb muscle cross sections from stage E14 control littermates (C) or Wnt4−/− embryos (D) stained with MuSK (red) antibodies together with α-bungarot oxin (AChR, green). MuSK colocalized with AChR at the NMJ of wild type and Wnt4−/− mutant embryos (10 cross sections from 2 Wnt4 mutants and control littermates were analyzed for each condition). (F) Examples of myotubes stained with α-bungarotoxin (AChR) upon control or Wnt4 treatment. (G) Measurements of the myotube area/field, the number of AChR clusters/myotube, the AChR cluster fluorescence signal intensity and the average AChR cluster area (50 AChR clusters for control and 65 for Wnt4 treated myotubes were analyzed). Wnt4 treatment induced an increase in the number of AChR clusters/myotube. However, AChR cluster fluorescence signal intensity was significantly reduced in Wnt4 treated myotubes. Error bars show means ± SEM. *P<0.05; **P<0.001; Mann-Whitney U test. NS, non significant. Scale bar: in A, 100 µm for A and B; in D, 30 µm for D and E; in F, 20 µm.

Figure 5. Wnt4 does not alter muscle structure but modifiy fiber type composition.

(A) Histological analysis of hind limb muscle cross sections from stage E18.5 control littermate (wild type) or Wnt4−/− embryos stained with heamatoxylin/eosin. The muscle gross organization was not affected in the Wnt4−/− embryos (N = 3 for Wnt4 mutants and N = 4 for control littermate embryos). (B) Measurement of muscle fibers perimeter. No significant difference in limb muscle section perimeter from Wnt4−/− mutant compared to wild type was detected. (C) Hind limb muscle cross sections (soleus level) from stage E18.5 control littermates (wild type) or Wnt4−/− embryos stained with myosin heavy chain I (MyHCI) antibodies (N = 2 for Wnt4 mutants and control littermate embryos). (D) Measurement of the number of MyHCI postive cells. The number of MyHCI positive cells was increased in Wnt4−/− mutant compared to wild type. Error bars show means ± SEM. **P<0.001; Mann-Whitney U test. NS, non significant. Scale bars: in A, 30 µm; in C, 10 µm.

Figure 6. Wnt4 interacts with MuSK and increases MuSK level of phosphorylation.

(A) Domain structure of MuSK, MuSKΔCRD and Wnt4-HA proteins. SS, signal sequence; TM, transmembrane. (B) Quantification of AChR cluster numbers in control or agrin treated myotubes transfected with MuSK or MuSKΔCRD. The deletion of MuSK CRD domain did not affect agrin-induced AChR clustering. (C, D) Coimmunoprecipitation of MuSK/Wnt4 in COS 7 cells. COS 7 cells were cotransfected with Wnt4-HA and MuSK or MuSKΔCRD. Western blot using HA antibodies was performed on cell lysates to assess the expression of Wnt4-HA (C, WCL, Whole Cell Lysate). Western blot of MuSK or HA immunoprecipitates probed with HA or MuSK antibodies showed that Wnt4 interacted with MuSK but not with MuSKΔCRD. (E) MuSK phosphorylation induced by Wnt4. HEK 293T cells were cotransfected with HA-MuSK or HA-MuSKΔCRD with or without Wnt4-HA. HA-MuSK, HA-MuSKΔCRD and Wnt4-HA were immunoprecipitated with HA antibodies. Western blots of HA immunoprecipitates were probed with HA or phosphotyrosine (pTyr) antibodies to assess HA-MuSK or HA-MuSKΔCRD tyrosine phosphorylation level. (F) Quantification of HA-MuSK or HA-MuSKΔCRD phosphorylation levels normalized to the total amount of MuSK or MuSKΔCRD proteins expressed as the +Wnt4/−Wnt4 ratio. Wnt4 induced MuSK but not MuSKΔCRD phosphorylation. Error bars show means ± SEM from three independent experiments. *: non phosphorylated MuSK, **: phosphorylated MuSK.