Figure 4
EphA4 KO does not protect against NMJ denervation, motor neuron loss or gliosis in SOD1G93A mice. (a–c) The early phenotype of NMJ loss was examined in 9-week TA muscle tissue from the cohort of mice with EphA4 KO initiated at 4 weeks (Fig. 2). (a) Examples of innervated and denervated NMJs co-stained with anti-VACHT (green) and α-Bungarotoxin (red). (b) One-way ANOVA comparing total number of synapses found no differences between the four genotypes. (n = 8 in WT groups, 10 in SOD1 groups). (c) One-way ANOVA comparing % denervation found an overall significant effect of genotype (F(3, 32) = 112.8; p < 0.0001). Follow up with Tukey’s t-test, revealed a significant effect of SOD1 genotype but not EphA4 genotype. (n = 8 in WT groups, 10 in SOD1 groups). (d–f) The later phenotypes of spinal cord motor neuron loss and gliosis was examined in 15-week-old tissue from the cohort of mice with EphA4 KO initiated at 7 weeks (Fig. 3). (d) Sample images of ChAT stained motor neurons and Iba1 stained microglia in lumbar spinal cord in the different genotypes. (e) L3-L5 motor neuron counts analyzed by one-way ANOVA showed an overall genotype group effect (F(3, 51) = 19.17; p < 0.0001), with post-hoc Tukey’s t-test revealing significant effects of SOD1 but not EphA4 genotype. (n = 8–10 in WT groups, 18 in SOD1 groups). (f) One-way ANOVA comparing Iba1 stained microglia in lumbar spinal cord shows an overall effect across groups (F(3, 51) = 27.13; p < 0.001) with post-hoc Tukey’s t-test showing increased gliosis in the SOD1 groups but no changes when EphA4 is knocked down. (n = 8–10 in WT groups, 18 in SOD1 groups).
