Semaphorin3A Signaling Is Dispensable for Motor Axon Reinnervation of the Adult Neuromuscular Junction

Sema3A and its receptor components are expressed in skeletal muscle and spinal cord. Quantitative real-time PCR (RT-qPCR) was performed on RNA isolated from the gastrocnemius/plantaris (GP), extensor digitorum longus (EDL), soleus (Sol), and spinal cord (SC) tissues of C57BL/6J mice (n = 6). Individual gene expression was normalized to the β2-Microglobulin (B2M) housekeeping gene. Myosin heavy chain (MyHC) isoforms (A), Semaphorin3A (Sema3A, B), class A plexin family members (PlxA1-4, C), and neuropilin family members (Npn1-2, D) were examined. Error bars represent the mean ± SE. *, p ≤ 0.05; ** p ≤ 0.01; ***, p ≤ 0.001; **** p ≤ 0.0001.

Sema3A and Npn1 gene expression decrease in the EDL following common peroneal nerve crush. A common peroneal nerve crush was performed and RNA was isolated at D7 (n = 9), D21 (n = 9), D30 (n = 4), and D50 (n = 4) from denervated and the contralateral control EDL muscles. In some cases, a nerve cut was performed instead of a nerve crush, and tissues were collected at the D21 time point (n = 8). A–D, Gene expression for the postsynaptic gene Muscle-specific kinase (MuSK, A), the regeneration marker Growth-associated protein 43 (GAP43, B), and two myelin markers Myelin protein zero (MPZ, C) and S100 calcium binding protein B (S100B, D) were used as controls to demonstrate the effectiveness of the nerve crush and how the regenerative response changes over the examined time course. E, F, Gene expression for Neuropilin1 (Npn1, E) and Semaphorin3A (Sema3A, F) decrease in response to nerve crush injury. Npn1 levels return to baseline over the time course, while Sema3A remains low even at the latest time point examined. For all graphs, error bars represent the mean ± SE, and letters represent significant differences (a–b, a–e: p < 0.0001; b–c: p ≤ 0.001; a–d, b–d: p < 0.01; b–e, a–c: p < 0.05).

Loss of Npn1 modestly impairs NMJ reinnervation after injury. A, Graphical depiction of the experimental paradigm. Tamoxifen (TMX) was administered for 5 d, followed by an at least 10-d rest period to allow for Cre-mediated recombination. At D0, a common peroneal crush was performed, and tissues were collected at the various time points. B, RT-qPCR was used to assess the level of Npn1 knockdown in GP muscle and SC from all mice used to examine reinnervation in subsequent experiments. The level of Npn1 transcript detected in Npn1^UBC mice (n = 21) is graphed relative to that observed in Npn1^WT control littermates (n = 20), with the average percentage knockdown in Npn1^UBC mice displayed above the bar graph. C, D, Control or Npn1 IP, followed by immunoblotting for Npn1, was used to assess the amount of Npn1 protein in Npn1^WT and Npn1^UBC mice after TMX-mediated deletion. Representative Western blots from the D30 time point demonstrate that Npn1^UBC mice have no detectable levels of Npn1 protein in either the GP muscle (C) or the whole SC (D). Supernatants immunoblotted for α-tubulin (αTub) or actin are provided as loading controls. E, Npn1 deletion does not alter the total number of scored NMJs observed at different time points after nerve crush injury (n.s., one-way ANOVA). F, G, The innervation status of sham-operated uninjured EDLs from Npn1^UBC (n = 4) and Npn1^WT (n = 4) mice was examined at the D21 time point. Scoring performed on the basis of nerve fiber (βIII-tubulin, F) or presynaptic terminal (synapsin, G) staining demonstrates that loss of Npn1 does not perturb the normal maintenance of the NMJ. H–K, Reinnervation following a nerve crush injury was examined in Npn1^WT littermate controls (H, J) and Npn1^UBC (I, K) mice. Innervation scores based on nerve βIII-tubulin staining (H, I) or presynaptic synapsin staining (J, K) produced similar trends. Despite loss of Npn1, early reinnervation (D7–D21) occurs normally. However, at D30, there appears to be a delay in motor nerve maturation as the number of NMJs scored as a 5 are decreased. This effect is only transient, as reinnervation is largely complete by the D50 time point. L, M, Reinnervation data were replotted using only scores 3–5 to directly compare differences between the two genotypes. Error bars represent the mean ± SE. **** p ≤ 0.0001, **, p < 0.01.

Loss of Sema3A does not impact NMJ reinnervation after injury. A, RNA was isolated from the GP muscle and SC of mice used for subsequent reinnervation studies. The level of Sema3A transcript detected in Sema3A^UBC mice treated with corn oil (n = 10) or TMX (n = 12) is graphed relative to that observed in Sema3A^WT control littermates (n = 12). The average percentage knockdown in Sema3A^UBC mice is displayed above each bar graph. B, GP muscle lysates were isolated from Sema3A^WT and Sema3A^UBC mice at various time points after nerve crush (D14–D30) and subjected to Western blotting. No detectable Sema3A protein was observed after TMX-mediated deletion in Sema3A^UBC mice. Supernatants immunoblotted for actin are provided as a loading control. C, Loss of Sema3A does not alter the number of NMJs scored at different time points (n.s., one-way ANOVA). D–G, Reinnervation after a common peroneal crush was analyzed in Sema3A^WT control littermates (D, F) and Sema3A^UBC (E, G) mice. Innervation scores based on nerve βIII-tubulin staining (D, E) or presynaptic synapsin staining (F, G) followed similar trends. No differences in reinnervation were detected at any time point. (H, I) Reinnervation data were replotted using only scores 3–5 to directly compare differences between the two genotypes. Error bars represent the mean ± SE. **** p ≤ 0.0001, *, p < 0.05.

Npn1 deletion does not affect emyelination after a denervating injury. Myelination in sham-injured and denervated EDLs was examined. The nerve and endplates were visualized by staining for βIII-tubulin and BTX (green), while the myelin was visualized by staining for S100b (blue) and MBP (magenta). A, Normal innervation and myelination is observed in sham-injured EDL sections. B, C, At D7 after nerve crush, there is a slight reduction in S100b and a near-total loss of MBP staining in both Npn1^WT (B) and Npn1^UBC (C) mice. D, E, Myelination was observed to have largely taken place by the D21 time point in both Npn1^WT (D) and Npn1^UBC (E) mice. Scale bars represent 100 µm.