β-Catenin stabilization in skeletal muscles, but not in motor neurons, leads to aberrant motor innervation of the muscle during neuromuscular development in mice

Abstract

β-Catenin, a key component of the Wnt signaling pathway, has been implicated in the development of the neuromuscular junction (NMJ) in mice, but its precise role in this process remains unclear. Here we use a β-catenin gain-of-function mouse model to stabilize β-catenin selectively in either skeletal muscles or motor neurons. We found that β-catenin stabilization in skeletal muscles resulted in increased motor axon number and excessive intramuscular nerve defasciculation and branching. In contrast, β-catenin stabilization in motor neurons had no adverse effect on motor innervation pattern. Furthermore, stabilization of β-catenin, either in skeletal muscles or in motor neurons, had no adverse effect on the formation and function of the NMJ. Our findings demonstrate that β-catenin levels in developing muscles in mice are crucial for proper muscle innervation, rather than specifically affecting synapse formation at the NMJ, and that the regulation of muscle innervation by β-catenin is mediated by a non-cell autonomous mechanism.

Figure 1. Stabilization of β-catenin in skeletal muscles results in neonatal lethality

A-G: LacZ staining of wholemount diaphragm (A & B) and a cross-section of the eye (C), the spinal cord (D & E) and the heart (F & G) from a ROSA26-LacZ;Myo-Cre+ embryo (E16.5). LacZ expression was detected in the diaphragm muscle (A & B; A, low power; B, high power), extraocular muscle (C), but absent in the spinal cord (D & E; D, low power; E, high power) and the cardiac muscles (F: atrium; G: ventricle). H: Immunoblot analysis of tissue homogenates from two pairs of Ctnnb1^lox(ex3)/+;Myo-Cre+ (Cre+) and Ctnnb1^lox(ex3)/+; Myo-Cre- (Cre-) embryos (E18.5) probed with anti-β-catenin antibodies and anti-α-tubulin, which served as a loading control. Full-length β-catenin (WT, upper band) was detected in spinal cord and diaphragm muscle in both genotypes, whereas a truncated β-catenin (Mut, lower band) was detected only in the diaphragm muscle in Ctnnb1^lox(ex3)/+;Myo-Cre+ mice. I. External phenotype of Ctnnb1^lox(ex3)/+;Myo-Cre- (control) and Ctnnb1^lox(ex3)/+;Myo-Cre+ (mutant) mice (P0). Mutant pup died within 24 hr after birth. Arrowhead in I points to the white patch of milk in the stomach in the control pup. Mutant pup exhibited a lack of milk in the stomach (J, bottom picture). Scale bars: A: 500 µm; B: 50 µm; C: 100 µm; D: 100 µm; E: 50 µm; F-G: 30 µm.

Figure 2. β-catenin stabilization in skeletal muscles leads to excessive nerve defasciculation and branching

A-F: Hemi-diaphragm muscles were labeled by anti-NF150 and anti-Syt2 antibodies to reveal innervation pattern. The primary phrenic nerve (p) enters the diaphragm muscle at mid-costal level and extends secondary (s) and tertiary (t) nerve branches prior to making synaptic contacts with the muscle fibers. Excessive nerve defasciculation and branching are evident in mutant muscles (D-F, Ctnnb1^lox(ex3)/+;Myo-Cre) compared with the controls (A-C). G-H: High-magnification views of E18.5 diaphragm muscles double-labeled with antibodies against Syt2 and NF150 and Texas Red conjugated α-bgt. Nerve terminal sprouts (arrowheads) are markedly increased in mutant muscles (H) compared with controls (G). Scale bars: A, D: 200 µm; B, E: 400 µm; C, F, 1000 µm; G-H, 100 µm.

Figure 3. β-catenin stabilization in muscle leads to expansion of the end-plate band

A-D: hole mounts of embryonic diaphragm muscles (A-B: E14.5; C-D: E18.5) labeled with α-bgt for AChRs. In both control (Ctnnb1^lox(ex3)/+) and mutant (Ctnnb1^lox(ex3)/+;Myo-Cre), AChR clusters (white arrowheads) are aligned along the central region of the muscle, forming a centrally located end-plate band. However, the end-plate band occupied a broader region in the mutant muscle (B, D: Ctnnb1^lox(ex3)/+;Myo-Cre) than in the control muscle (A, C: Ctnnb1^lox(ex3)/+). Insets in C and D show high power views of individual AChR clusters. Scale bars: A-B, 200µm; C-D, 200µm; insets in C and D, 20 µm.

Figure 4. Normal morphology of the NMJ in mice expressing stabilized β-catenin in muscles

A-B: Confocal images of E18.5 diaphragm muscle (A) and triangulris sterni muscle (B), immunostained by a mixture of anti-NF 150 and anti-Syt2 antibodies (green) and α-bgt (red), and the merged images are shown at the bottom row in both A and B. The pre-synaptic nerve terminals (white arrowheads) were juxtaposed with post-synaptic AChR clusters (white arrows) in both controls (left column) and Ctnnb1^lox(ex3)/+;Myo-Cre muscles (right column). C-D: Electron micrographs of the NMJ in E18.5 diaphragm muscles from control (C) and Ctnnb1^lox(ex3)/+;Myo-Cre (D) muscles. Pre-synaptic nerve terminals (Nt), capped by a terminal Schwann cell, were filled with an abundance of synaptic vesicles in both control and Ctnnb1^lox(ex3)/+;Myo-Cre muscles. Black arrow points to the basal lamina between the pre- and post-synaptic membranes, and black arrowhead points to the postsynaptic junctional fold. Scale bars: A-B, 20 µm; C-D, 1 µm.

Figure 5. β-catenin stabilization in muscles leads to formation of the NMJ within the central tendon region of the diaphragm

A: Low power views of right hemi-diaphragm (E15.5) from control (left panel) and Ctnnb1^lox(ex3)/+;Myo-Cre (right panel). B: High power views of the central tendon region bordered by the square, labeled by anti-NF150 and anti-Syt2 antibodies (green) and α-bgt (red). The merged image is shown at the bottom. Arrowheads point to the NMJ. C: A whole diaphragm muscle in from Ctnnb1^lox(ex3)/+;Myo-Cre (E18.5) illustrating the aberrant nerves in the central tendon of the diaphragm (arrowhead in C). D: Individual NMJs (arrowheads) in the central tendon regions of the diaphragm in Ctnnb1^lox(ex3)/+;Myo-Cre embryos (E18.5), labeled by a mixture of anti-NF150 and anti-Syt2 antibodies (green) and α-bgt (red) (top three panels), or labeled by anti-synaptophysin antibodies (green) and α-bgt (red) (bottom three panels). E: Wholemount AChE staining of E18.5 diaphragm muscle. AChE clusters were detected in extra muscle patches (arrowheads, left panel) within the central tendon of the diaphragm muscle in Ctnnb1^lox(ex3)/+;Myo-Cre mice, but not in the control. Scale bars: A, C: 400 µm; B: 10 µm; D: 20 µm; E, 2000 µm.

Figure 6. β-catenin stabilization in muscles leads to an increase in motor axon numbers in the phrenic nerve

Cross-sections of the phrenic nerve (E18.5) were examined under an electron microscope: A-B: Low power views of the entire nerve cross-section. The number of axons was significantly increased (t-test, P < 0.01) in the mutants (261 ± 7 per nerve, n = 6 nerves) compared with the controls (228 ± 9 per nerve, n = 7). C-D: High power views illustrating individual axons (Ax) wrapped by Schwann cell (SC) processes (arrow). Scale bars: A-B: 10 µm; C-D: 1 µm.

Figure 7. β-catenin stabilization in muscles leads to a reduction in muscle fiber caliber

A-B: Cross-section (semi-thin, 1 µm) of the mid-costal regions of E18.5 diaphragm muscle stained with toluidine blue. Muscle nuclei (arrows) are peripherally localized in both Ctnnb1^lox(ex3)/+;Myo-Cre and control mice. C: Quantification of the average area of muscle fiber size (left bar graph) and size distribution (right bar graph) in the control and Ctnnb1^lox(ex3)/+;Myo-Cre mice. The average size of muscle fibers was 139.6 ± 4.1 µm² (n = 407 fibers) in Ctnnb1^lox(ex3)/+;Myo-Cre mice, vs. 167.9 ± 3.9 µm² (n = 408 fibers, P < 0.001) in the control mice. D-E: Longitudinal sections of the diaphragm muscle under EM. The organization of the Z-lines (arrowheads) and sarcomeres (unit between Z-lines) is similar between the control (D) and Ctnnb1^lox(ex3)/+;Myo-Cre mice (E). Scale bars: A-B: 30 µm; D-E: 1 µm.

Figure 8. Normal synaptic transmission at the NMJ in mice expressing stabilized β-catenin in muscles

A: Sample mEPP traces (superimposed 120 X 1-second traces) from E18.5 diaphragm muscle in control and Ctnnb1^lox(ex3)/+;Myo-Cre mice. B: Quantification of mEPP frequencies, amplitudes, 10–90% rise time and half width – mEPP frequencies (event/minute): 0.90 ± 0.15 (Ctnnb1^lox(ex3)/+;Myo-Cre) vs. 0.93 ± 0.15 (control); mEPP amplitudes 2.60 ± 0.25 mV (Ctnnb1^lox(ex3)/+;Myo-Cre) vs. 2.12 ± 0.26 mV (control); mEPP rise time: 3.87 ± 0.24 ms (Ctnnb1^lox(ex3)/+;Myo-Cre) vs. 3.84 ± 0.56 ms (control); mEPP half width: 12.82 ± 0.24 ms (Ctnnb1^lox(ex3)/+;Myo-Cre) vs. 11.74 ± 0.74 ms (control). No statistical differences were found between the Ctnnb1^lox(ex3)/+;Myo-Cre and the controls. C-D: Sample traces and quantification of EPP amplitudes. EPP amplitudes were indistinguishable between controls (10.67 ± 0.96 mV) and Ctnnb1^lox(ex3)/+;Myo-Cre (10.39 ± 1.3 mV). Number of sample analyzed: Ctnnb1^lox(ex3)/+;Myo-Cre (n = 25 cells, N = 4 mice); control (n = 18 cells, N = 3 mice). E: Action potentials were evoked by electrical stimulation of the phrenic nerves in both control and Ctnnb1^lox(ex3)/+;Myo-Cre diaphragm muscles.

Figure 9. Normal muscle contractile properties in mice expressing stabilized β-catenin in muscles

A: Muscle twitch responses at 4, 10, 20 and 50 Hz in control and Ctnnb1^lox(ex3)/+;Myo-Cre diaphragm muscles (E18.5). B: Plot of twitch force (expressed as % of maximal force) against stimulation frequency. In both control and Ctnnb1^lox(ex3)/+;Myo-Cre diaphragm muscles, muscle twitch force increases as stimulation frequency increases. C: Normalized maximal twitch force by muscle weight. No significant difference was detected between control (0.78 ± 0.06 g/mg, N = 3 embryos) and Ctnnb1^lox(ex3)/+;Myo-Cre (0.70 ± 0.07 g/mg, N = 3 embryos).

Figure 10. Neuromuscular synapses develop normally in mice with β-catenin stabilization in motor neurons

Hemi-diaphragm muscles (E18.5) were double-labeled with anti-NF150 and anti-Syt2 antibodies and α-bgt. Similar innervation patterns were seen in the control (A) and the mice expressing stabilized β-catenin specifically in motor neurons (B, Ctnnb1^lox(ex3)/+;HB9^cre). High-power magnification of the diaphragm muscle in A and B, demonstrating individual NMJs were developed in control (C-E) and Ctnnb1^lox(ex3)/+;HB9^cre mice (F-H). I-N: NMJs in triangularis sterni muscles (P45) of control (I-K) and Ctnnb1^lox(ex3)/+;HB9^cre mice (L-N). Scale bars: A-B, 400 µm; C-H, 20µm; I-N, 20 µm.

Figure 11. Neuromuscular synapses function normally in mice with β-catenin stabilization in motor neurons

A: Typical mEPP traces (superimposed 20 X 0.5-second traces) from the lumbrical muscles at P45 in control and Ctnnb1^lox(ex3)/+;HB9^Cre mice. B: Quantification of mEPP frequencies, amplitudes, 10–90% rise time and half width – mEPP frequencies (Hz) were 0.62 ± 0.08 in control and 0.62 ± 0.12 in Ctnnb1^lox(ex3)/+;HB9^Cre; mEPP amplitudes: 1.63 ± 0.13 mV (control), 1.68 ± 0.16 mV (Ctnnb1^lox(ex3)/+;HB9^Cre); mEPP rise time: 2.20 ± 0.24 ms (control), 2.26 ± 0.26 ms (Ctnnb1^lox(ex3)/+;HB9^Cre); mEPP half width: 4.12 ± 0.24 ms (control), 4.33 ± 0.29 ms (Ctnnb1^lox(ex3)/+;HB9^Cre). No statistical differences were found between control and Ctnnb1^lox(ex3)/+;HB9^Cre mice. C-D: Sample traces (C) and quantification of EPP (D). EPP amplitudes (control: 23.96 ± 1.44 mV; Ctnnb1^lox(ex3)/+;HB9^Cre: 22.67 ± 1.41 mV), EPP rise time (control: 1.24 ± 0.04 ms; Ctnnb1^lox(ex3)/+;HB9^Cre: 1.30 ± 0.05 ms) and EPP half width (control: 3.92 ± 0.09 ms; Ctnnb1^lox(ex3)/+;HB9^Cre: 4.09 ± 0.14 ms) were indistinguishable between control and Ctnnb1^lox(ex3)/+;HB9^Cre mice. E-F: Sample EPP traces evoked by electrical stimulation of the nerve at 70 Hz (E) and quantification of EPP rundown ratio. The rate of EPP rundown is similar between controls and Ctnnb1^lox(ex3)/+;HB9^cre mice. Number of sample analyzed: control (n = 13); Ctnnb1^lox(ex3)/+;HB9^Cre (n = 11).