Click here to close
Hello! We notice that you are using Internet Explorer, which is not supported by Xenbase and may cause the site to display incorrectly.
We suggest using a current version of Chrome,
FireFox, or Safari.
???displayArticle.abstract???
During vertebrate neuromuscular junction (NMJ) development, contact between motor axons and muscle fibers is followed by pre- and post-synaptic specialization. Using Xenopus nerve-muscle cocultures, we recently showed that spinal neurons initially contacted muscle cells by means of filopodial processes, and that muscle-derived basic fibroblast growth factor induced axonal filopodia and slowed axonal advance to promote nerve-muscle interaction and NMJ establishment. In contrast, neurotrophins enhanced axonal growth but suppressed the extension of axonal filopodia and blocked NMJ formation.Here we report that hepatocyte growth factor (HGF), which also supports motor neuron survival, was expressed by Xenopus muscle cells, and that forced expression of HGF in Xenopus spinal neurons inhibited the extension of axonal filopodia. Overexpression of the HGF-receptor c-Met in neurons also blocked the formation of axonal filopodia and furthermore sped up axonal growth, but a kinase-dead form of c-Met was unable to effect these changes. Importantly, treatment of nerve-muscle cocultures with recombinant HGF or the expression of HGF or active c-Met in neurons, or that of excess HGF in muscle, inhibited nerve-induced AChR clustering in muscle.Our results suggest that HGF/c-Met signaling in neurons promotes axonal growth but suppresses filopodial assembly in neurons and hinders NMJ establishment.
Figure 1. Expression of HGF and c-Met mRNAs in Xenopus nerve and muscle. Fragments of HGF and c-Met mRNAs were amplified by RT-PCR and separated using agarose gels. More HGF mRNA was detected in Xenopus myotomal tissue (M) than in neural tubes (N), while the opposite was the case for c-Met mRNA. GAPDH was used as a control for PCR amplification and gel-loading.
Figure 2. Inhibition of filopodial assembly in neurons by HGF. Phase-contrast (A) and fluorescence images (A′′) of cultured Xenopus spinal neurons expressing GFP (A′) or GFP and HGF (B′, C′). Filopodia along the axon are indicated by arrows in A. Compared to GFP-neurons (A, A′), neurons overexpressing HGF (B, B′) had fewer filopodia. However, bath application of 100 nM SU11274, a specific inhibitor of c-Met, rescued filopodial formation in HGF-expressing neurons (C, C′) almost to the control level. D: Filopodial densities calculated for axon segments from GFP- and HGF-neurons treated without or with SU11274; mean SEM, t-test, *P < 0.05.
Figure 3. Suppression of filopodial formation and enhancement of axonal growth by c-Met overexpression in neurons. Spinal neurons expressing GFP (A, A′), GFP and wild-type c-Met (WT-c-Met; B, B′), or GFP and kinase-dead c-Met (K1110A-c-Met; C, C′) were examined for filopodia. Compared to GFP-neurons (A), neurons expressing active c-Met (B) but not inactive c-Met (C) had fewer filopodia (arrows). This is quantified as filopodial densities in panel E; mean SEM; t-test, *P < 0.05 compared to control. D: HEK293T cells were transfected with cDNAs encoding HA-tagged WT-c-Met and K1110A-c-Met and the proteins were immunoprecipitated (IP) from cell extracts using anti-HA antibody. Immunoblotting (IB) of IP samples with anti-HA (upper blot) and anti-pY (lower blot) showed that WT-c-Met was tyrosine-phosphorylated (pY-c-Met) but kinase-dead c-Met was not. F: Spinal neurons expressing GFP, GFP plus WT-c-Met, and GFP plus K1110A-c-Met were monitored by time-lapse imaging (pictures not shown) and axonal growth rates were calculated and normalized relative to GFP-neurons. Neurons overexpressing WT-c-Met, but not K1110A-c-Met, grew faster than control neurons; mean SEM; t-test, *P < 0.05 compared to control.
Figure 4. Inhibition of NMJ assembly by HGF bath application. Nerve-muscle cocultures were incubated in control medium (A, B) or medium containing 100 ng/ml HGF (C, D) and AChR clustering at innervation sites was examined by R-BTX labeling. Spinal axons induced less AChR aggregation in HGF-treated cocultures (D) than in the control (B). E: The percentages of nerve-muscle contacts with AChR clusters. Mean SEM shown; t-test, **P < 0.01 compared to control. Arrows point to nerve tracks, arrowheads to nerve-induced AChR clusters.
Figure 5. Inhibition of NMJ assembly by c-Met or HGF overexpression in neurons. Neurons expressing GFP (A) or GFP plus WT-c-Met (D), K1110A-c-Met (G), or HGF (J) were cultured with normal muscle cells. After 1 day, the cocultures were labeled with R-BTX to examine muscle AChR clustering. WT-c-Met-neurons (F) and HGF-neurons (L) poorly induced AChR clustering at innervation sites in muscle compared to GFP-neurons (C) and K1110A-c-Met (I). Pooled data from multiple cocultures were used to calculate percentages of nerve-muscle contacts with AChR clusters (M); mean SEM shown; t-test, **P < 0.01 compared to control. Arrows point to nerve tracks and arrowheads to nerve-induced AChR clusters; .s.marks pre-existing AChR clusters (hot spots) in muscle cells innervated by neurons expressing WT-c-Met (F) and HGF (L).
Figure 6. Increase in axonal growth by HGF overexpression in muscle. Axons grew slowly along GFP-positive muscle cells after contacting them (A), as shown by these two sequential images taken 30 min apart. Compared to GFP muscle cells, HGF-expressing muscle cells allowed unabated axonal growth after contact (D). Green fluorescence indicates GFP or HGF/GFP expression, and the arrows indicate the initial (0′, black) and final positions (30′, white) of the distal ends of the axons. G: The average growth speed of axons in contact with HGF-expressing muscle cells was calculated and normalized relative to the growth speed of axons touching control GFP-muscle cells. Mean SEM shown; t-test, ***P < 0.001 compared to control.
Figure 7. Suppression of NMJ formation by HGF overexpression in muscle. Cocultures were prepared using normal (A) or K1110A-c-Met-expressing (G) neurons and muscle cells expressing GFP (A) or GFP plus HGF (D). R-BTX-labeling showed that neurons induced less AChR aggregation in HGF-overexpressing muscle cells (F) than in control muscle cells (C). J: Expression of excess HGF in muscle cells resulted in ∼40% reduction in nerve-induced AChR clustering. However, compared with Ctl neurons (D), K1110A-c-Met-expressing neurons (G) were able to induce more AChR clusters (I) at nerve-muscle contact sites on HGF-expressing muscle cells; mean SEM; t-test, **P < 0.01. < 0.001. Arrows mark nerve tracks and the arrowheads show AChR clustering along innervation sites; .s.indicates an AChR hot spot in a cell expressing exogenous HGF in which nerve failed to induce AChR aggregation.