Fig 2: Neuromuscular junction defects in homozygous S85C mice at endpoint.a Representative images of lower lumbar spinal cords (L4–L6) with magnified images showing ChAT-and NeuN-positive α-motor neurons (MNs) and ChAT-positive, NeuN-negative γ-MNs in the ventral horns of the spinal cord. Scale bars indicate 500 µm (left column) and 100 µm (magnified images). b Graph shows the number of α- and γ-MNs in the lumbar spinal cord at disease end-stage (n = 10 Matr3+/+, 10 Matr3S85C/+, 13 Matr3S85C/S85C). c Images showing synaptophysin and synapsin staining of presynaptic terminals, neurofilament H staining of presynaptic axons and α-bungarotoxin (BTX) staining of post-synaptic terminals in the TA muscles at end-stage. White arrows indicate partially denervated NMJs. Scale bar indicates 50 µm. d Graph shows the percentage of partially denervated NMJs (n = 3 Matr3+/+, 3 Matr3S85C/S85C, **p = 0.0053). e Representative images showing synaptophysin and synapsin staining of presynaptic terminals, neurofilament H staining of presynaptic axons and α-bungarotoxin staining of post-synaptic terminals in the TA muscles at end-stage. White arrow indicates axonal bleb proximal to the NMJ. Scale bar indicates 50 µm. f Graph shows the percentage of neurofilament-positive axon swelling (n = 9 Matr3+/+, 9 Matr3S85C/S85C, **p = 0.0017). g Representative images showing α-bungarotoxin staining of endplates in the TA muscles at end-stage. Scale bar indicates 60 µm. h Graph shows average TA endplate size (n = 6 Matr3+/+, 7 Matr3S85C/S85C, **p = 0.0019). Data shown in b, d, f and h are displayed as mean ± SEM and each dot represents a single animal. Significance was determined by unpaired two-tailed t-test. Source data are provided as a Source Data file.

Fig 3: Homozygous S85C knock-in mice reach humane endpoint at over one year of age.a Homozygous S85C animals (both males and females) exhibited significant differences in weight starting at around 15 and 21 weeks of age for males and females, respectively. Each dot represents mean ± SEM, where males: n = at least 10 Matr3+/+, 10 Matr3S85C/+, 5 Matr3S85C/S85C; females: n = at least 8 Matr3+/+, 10 Matr3S85C/+, 8 Matr3S85C/S85C (exact n numbers for each age are presented in Supplementary Information). Two-way ANOVA, Dunnett correction for multiple comparisons, see Supplementary Information for p-values at each age (*p < 0.05). b The phenotype score for wild-type, heterozygous and homozygous S85C male and female animals at over 1 year of age. Mice with a score of 60 were considered to be at the endpoint. Each dot represents a single animal (males: n = 10 Matr3+/+, 12 Matr3S85C/+, 6 Matr3S85C/S85C; females: n = 10 Matr3+/+, 12 Matr3S85C/+, 11 Matr3S85C/S85C), with data presented as mean ± SEM. Significance was determined by unpaired two-tailed t-test; ****p < 0.0001. c Size difference between wild-type and homozygous S85C mice at endpoint. Source data are provided as a Source Data file.

Fig 4: A striking loss of Purkinje cells in homozygous S85C mice.a Representative images of the whole cerebellum (upper row) and magnified images of calbindin-positive Purkinje cells (bottom row) of 6-week-old mice. Scale bars indicate 800 µm (upper row) and 50 µm (bottom row). b, c Graphs show (b) cerebellum size (n = 3 Matr3+/+, 3 Matr3S85C/+, 3 Matr3S85C/S85C, ns = not significant) and (c) the total number of calbindin-positive Purkinje cells in the whole cerebellum (n = 3 Matr3+/+, 3 Matr3S85C/+, 3 Matr3S85C/S85C, ns = not significant) of 6-week-old mice. d Representative images of the whole cerebellum (upper row) and magnified images of calbindin-positive Purkinje cells (bottom row) of endpoint mice (or over 60 weeks of age). Scale bars indicate 800 µm (upper row) and 50 µm (bottom row). e, f Graphs show (e) cerebellum size (n = 3 Matr3+/+, 3 Matr3S85C/+, 3 Matr3S85C/S85C, *p = 0.0194) and (f) the total number of calbindin-positive Purkinje cells in the whole cerebellum (n = 3 Matr3+/+, 3 Matr3S85C/+, 3 Matr3S85C/S85C, **p = 0.0015) of endpoint mice. Data shown in b, c, e and f are displayed as mean ± SEM and each dot represents a single animal. Significance was determined by unpaired two-tailed t-test. Source data are provided as a Source Data file.

Fig 5: Upregulation of immune response genes in homozygous S85C mice at early disease stage.a–c Volcano plots showing differentially expressed genes in the a cerebellum, b spinal cord, and c cortex of wild-type and homozygous S85C mice at 8–10 weeks old. FDR p < 0.05, absolute fold change (abs(FC)) > 1.5. The top five genes with significant expression changes are labeled with gene names. c In the cortex, no genes are significantly differentially expressed including immune response genes. d, e GO analysis for upregulated (red) or downregulated (blue) genes in the d cerebellum and e spinal cord. f Representative images showing IBA1 and GFAP staining of the cerebellum at 30 weeks of age. Scale bars indicates 300 µm. g, h Graphs show the integrated density of g IBA1 (n = 3 Matr3+/+, 3 Matr3S85C/+, 3 Matr3S85C/S85C, *p = 0.023) and h GFAP (n = 3 Matr3+/+, 3 Matr3S85C/+, 3 Matr3S85C/S85C, *p = 0.0157) staining in the cerebellum at 30 weeks old. Data shown in g and h are displayed as mean ± SEM and each dot represents a single animal. Significance was determined by unpaired two-tailed t-test. Source data are provided as a Source Data file.