a p-Drp1 S616 and Fis1 protein expression levels in the lumbar spinal cord of SOD1 G93A mice, and quantification of the relative expression of p-Drp1 S616 and Fis1, compared to WT. WT:p-DRP^S616 = 1.00 ± 0.12 (n = 3 mice), G93A: p-DRP^S616 = 0.64 ± 0.07 (n = 5 mice), WT:DRP1 = 1.00 ± 0.12 (n = 3 mice), G93A: DRP1 = 0.96 ± 0.06 (n = 5 mice), WT:Fis1 = 0.24 ± 0.07 (n = 3 mice), G93A: Fis1 = 0.70 ± 0.12 (n = 5 mice). Values are mean ± S.E.M. F(1,6) = 2.80, *P = 0.0311 for p-DRP^S616; F(1,6) = 0.2721, P = 0.7946 for DRP1; F(1,6) = 2.691, *P = 0.0.0360 for Fis1 graph by two-sided Student’s t-test. n.s.: not significant. b p-Drp1 S616 immunostaining in the spinal MN of G93A and WT mice. The graph shows the ratio of intensity between p-Drp1 (green) and Drp1 (red) in ChAT-positive (blue; MN marker) neurons. WT = 1.56 ± 0.04 (n = 316 ChAT+ neurons), G93A = 0.81 ± 0.01 (n = 346 ChAT+ neurons. Values are mean ± S.E.M. F(1,660) = 18.4, ****P < 0.0001 by two-sided Student’s t-test. Scale bar, 100 μm. c Expression of p-Drp1 S616 in GFP-positive cortical neurons, following G93A or Q331K overexpression. The graph shows the quantification of p-Drp1 S616 intensity. Arrowhead indicates a GFP-positive neuron. GFP = 100.00 ± 8.40 (n = 24 GFP+ neurons), G93A = 45.30 ± 3.73 (n = 25 GFP+ neurons), Q331K = 42.95 ± 4.2 (n = 20 GFP+ neurons). Values are mean ± S.E.M. F(2,66) = 29.75, ****P < 0.0001 by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons. Scale bar, 20 μm.

Analysis of mitochondrial length ((a) N = 10 per group) and active caspase-3-positive neurons (b–d) after 3 days following shRNA co-transfection or 25 μM Mdivi-1 treatment with G93A or Q331K clone. Mitochondrial images have been embossed. GFP = 100.1 ± 6.54 (n = 31 GFP+ neurons), SOD G93A:CON = 74.55 ± 6.54 (n = 24 GFP+ neurons), SOD G93A:shDRP1 = 197.3 ± 14.96 (n = 7 GFP+ neurons), SOD G93A:Mdivi = 105.5 ± 8.79 (n = 10 GFP+ neurons), SOD G93A:shFis1 = 143.7 ± 17.38 (n = 9 GFP+ neurons). Values are mean ± S.E.M. F(1,53) = 3.239, **P = 0.0021 for GFP versus SOD G93A:CON by two-sided Student’s t-test. F(3,46) = 33.54, *P = 0.0440, ****P < 0.0001 by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons in (a). GFP:CON = 100 ± 0.00 (n = 3 independent primary cultures), GFP:shDRP1 = 95.25 ± 36.86 (n = 3 independent primary cultures), G93A:CON = 218 ± 25.18 (n = 3 independent primary cultures), G93A:shDRP1 = 116.4 ± 7.52 (n = 3 independent primary cultures), Q331K:CON = 370 ± 38.99 (n = 3 independent primary cultures), Q331K:shDRP1 = 247.6 ± 18.38 (n = 3 independent primary cultures). Values are mean ± S.E.M. F(2,6) = 25.58, *P = 0.0473, ***P = 0.009 by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons. F(1,4) = 0.1289, P = 0.9037 for GFP:CON versus GFP:shDRP1; F(1,4) = 3.869, ^#P = 0.0180 for G93A:CON versus G93A:shDRP1; F(1,4) = 2.846, ^#P = 0.0466 for Q331K:CON versus Q331K:shDRP1 by two-sided Student’s t-test in (b). GFP:CON = 99.98 ± 0.03 (n = three independent primary cultures), GFP:Mdivi-1 = 86.85 ± 24.57 (n = 3 independent primary cultures), G93A:CON = 257 ± 15.85 (n = 3 independent primary cultures), G93A: Mdivi-1 = 119.9 ± 17.42 (n = 3 independent primary cultures), Q331K:CON = 372 ± 60.06 (n = 3 independent primary cultures), Q331K: Mdivi-1 = 171.9 ± 32.03 (n = 3 independent primary cultures). Values are mean ± S.E.M. F(2,6)=14.53, *P = 0.0483, **P = 0.0041 by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons. F(1,4) = 0.5345, P = 0.6213 for GFP:CON versus GFP:shDRP1; F(1,4) = 5.819, ^##P = 0.0043 for G93A:CON versus G93A:shDRP1; F(1,4) = 2.922, ^#P =0 .0422 for Q331K:CON versus Q331K:shDRP1 by two-sided Student’s t-test in (c). GFP:CON = 100 ± 0.03 (n = 3 independent primary cultures), GFP:shFis1 = 94.73 ± 27.67 (n = 3 independent primary cultures), G93A:CON = 244 ± 4.88 (n = 3 independent primary cultures), G93A:shFis1 = 125.8 ± 36.28 (n = 3 independent primary cultures), Q331K:CON = 324.9 ± 18.34 (n = 3 independent primary cultures), Q331K: shFis1 = 146 ± 38.89 (n = 3 independent primary cultures). Values are mean ± S.E.M. F(2,6) = 108.1, ***P = 0.0002, ****P < 0.0001 by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons. F(1,4) = 0.1922, P = 0.8569 for GFP:CON versus GFP:shDRP1; F(1,4) = 3.24, ^#P = 0.0317 for G93A:CON versus G93A:shDRP1; F(1,4) = 2.922, ^#P = 0.0141 for Q331K:CON versus Q331K:shDRP1 by two-sided Student’s t-test in (d). Scale bar, 5 μm. e Zebrafish expressing G93A and Q331 with 2.5 μM Mdivi-1 daily treatment over 2 days, and the number of axonal defects of MNs counted for each of the four segments. Normal = 0.14 ± 0.1 (n = 14 zebrafish), G93A:CON = 2.93 ± 0.29 (n = 14 zebrafish), G93A:Mdivi-1 = 0.2 ± 0.2 (n = 5 zebrafish), Q331K:CON = 3.43 ± 0.33 (n = 14 zebrafish), Q331K:Mdivi-1 = 0.45 ± 0.28 (n = 11 zebrafish). Values are mean ± S.E.M. F(2,39) = 47.44, ****P < 0.0001 by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons.; F(1,17) = 5.445, ^####P < 0.0001 for G93A:CON versus G93A:Mdivi-1, F(1,23) = 6.671, ^####P < 0.0001 for Q331K:CON versus Q331K: Mdivi-1 by two-sided Student’s t-test.

a Western blot analysis of p-Drp1 S616 and Drp1 expression level in primary cortical neuron with Cantharidin (for 1 h) or I-2 (for 1 day) treatment. b Representative immunostaining image of p-Drp1 S616 (red) in GFP+ primary cortical neuron with or without 80 nM I-2 treatment for 24 h. Arrowhead indicates the GFP-positive neuron. Scale bar, 20 μm. c PP1 activity in the lumbar spinal cord of G93A and WT mice at 60, 100, and 200 days of age. WT:Pre = 100 ± 7.79 (n = 3 mice), G93A:Pre = 125.9 ± 5.51 (n = 4 mice), WT:Onset = 100 ± 17.35 (n = 3 mice), G93A: Onset = 166 ± 14.32 (n = 5 mice), WT:Post = 100 ± 18.35 (n = 5 mice), G93A:Post = 148.7 ± 9.58 (n = 5 mice). Values are mean ± S.E.M. F(1,5) = 2.811, *P = 0.0375 for WT:Pre versus G93A:Pre; F(1,6) = 2.88, *P = 0.0281 for WT:Onset versus G93A:Onset; F(1,8) = .354, *P = 0.0464 for WT:Post versus G93A:Post by two-sided Student’s t-test. d Western blot analysis of p-PP1 and PP1 expression in the lumbar spinal cord of G93A and WT mice at the onset period. The graph shows the quantification of p-PP1 vs. PP1 intensity. WT = 1.00 ± 0.04 (n = 3 mice), G93A = 0.76 ± 0.05 (n = 3 mice). Values are mean ± S.E.M. F(1,4) = 3.832, *P = 0.0186 by two-sided Student’s t-test.

a The representative image of mitochondria in the neurite of GFP+ primary cortical neuron with 40 nM I-2 treatment or shRNA lenti-viral infection after transfection of dsRed-mito and mALS genes, over 72 h. The mitochondrial image has been embossed using Photoshop and measured on the GFP⁺ cortical neuron. Scale bar, 5 μm. b Analysis of mitochondrial length in (a). GFP:CON = 1.00 ± 0.06 (n = 29 GFP+ neurons), GFP:shPP1α = 1.1 ± 0.09 (n = 17 GFP+ neurons) GFP:shPP1β=0.79 ± 0.87 (n = 7 GFP+ neurons), GFP:shPP1γ = 1.08 ± 0.05 (n = 27 GFP+ neurons), GFP:I-2 = 0.96 ± 0.05 (n = 30 GFP+ neurons), G93A:CON = 0.7 ± 0.02 (n = 30 GFP+ neurons), G93A:shPP1α = 1.00 ± 0.4 (n = 33 GFP+ neurons), G93A:shPP1β = 0.66 ± 0.02 (n = 30 GFP+ neurons), G93A:shPP1γ = 0.99 ± 0.05 (n = 20 GFP+ neurons), G93A:I-2 = 1.04 ± 0.06 (n = 30 GFP+ neurons), Q331K:CON = 0.65 ± 0.02 (n = 30 GFP+ neurons), Q331K:shPP1α = 1.1 ± 0.06 (n = 31 GFP+ neurons), Q331K:shPP1β = 0.74 ± 0.03 (n = 30 GFP+ neurons), Q331K:shPP1γ = 0.89 ± 0.6 (n = 19 GFP+ neurons), Q331K:I-2 = 0.98 ± 0.05 (n = 35 GFP+ neurons). Values are mean ± S.E.M. F(2,86) = 26.64, ****P < 0.0001 for CON groups; F(4,105) = 1.710, n.s.: not significant for GFP groups; F(4,138) = 20.02, ^###P = 0.0001, ^####P < 0.0001 for G93A groups; F(4,138) = 20.02, ^##P = 0.0089, ^####P < 0.0001 for Q331K groups by one-way ANOVA with Tukey’s post hoc analysis for multiple comparisons. c Quantification of cleaved caspase-3-positive neurons in GFP+ neurons with 40 nM I-2 treatment or shRNA lenti-viral infection following transfection of mALS genes, over 72 h. GFP:CON = 7.67 ± 2.06 (n = 6 independent primary culture), GFP:shPP1α = 35.61 ± 12.54 (n = 2 independent primary culture), GFP:shPP1β=8.28 ± 0.75 (n = 3 independent primary culture), GFP:shPP1γ = 17.11 ± 4.65 (n = 3 independent primary culture), GFP:I-2 = 7.5 ± 2.51 (n = 5 independent primary culture), G93A:CON = 26.59 ± 2.31 (n = 5 independent primary culture), G93A:shPP1α = 37.09 ± 5.51 (n = 2 independent primary culture), G93A:shPP1β = 18.39 ± 2.45 (n = 2 independent primary culture), G93A:shPP1γ = 12.09 ± 2.63 (n = 3 independent primary culture), G93A:I-2 = 13.79 ± 0.84 (n = 4 independent primary culture), Q331K:CON = 34.83 ± 1.370 (n = 3 independent primary culture), Q331K:shPP1α = 32.39 ± 2.39 (n = 2 independent primary culture), Q331K:shPP1β = 17.03 ± 6.39 (n = 3 independent primary culture), Q331K:shPP1γ=17.07 ± 1.44 (n = 3 independent primary culture), Q331K:I-2 = 12.57 ± 2.60 (n = 3 independent primary culture). Values are mean ± S.E.M. F(2,11) = 41.29, ****P < 0.0001 for CON groups; F(4,14) = 7.230, ***P = 0.0007 for GFP groups; F(4,11) = 14.03, ^##P = 0.0096 for G93A:CON versus G93A: shPP1γ, ^##P = 0.0073 for G93A:CON versus G93A: I-2; F(4,9) = 7.887, ^#P = 0.0327 for Q331K:CON versus Q331K: shPP1β, ^#P = 0.0327 for Q331K:CON versus Q331K: shPP1γ, ^##P = 0.0089 by one-way ANOVA followed by Tukey’s multiple comparisons test. d Quantification of pyknotic cell in GFP+ neurons with 40 nM I-2 treatment or shRNA lenti-viral infection following transfection of mALS genes, over 72 h. GFP:CON = 9.39 ± 2.33 (n = 6 independent primary culture), GFP:shPP1α = 45.01 ± 14.25 (n = 2 independent primary culture), GFP:shPP1β = 11.08 ± 2.31 (n = 3 independent primary culture), GFP:shPP1γ = 15.40 ± 5.19 (n = 3 independent primary culture), GFP:I-2 = 6.64 ± 2.67 (n = 5 independent primary culture), G93A:CON = 34.85 ± 4.25 (n = 6 independent primary culture), G93A:shPP1α=41.18 ± 6.97 (n = 2 independent primary culture), G93A:shPP1β = 15.78 ± 0.16 (n = 2 independent primary culture), G93A:shPP1γ = 13.13 ± 4.84 (n = 3 independent primary culture), G93A:I-2 = 17.98 ± 3.12 (n = 5 independent primary culture), Q331K:CON = 42.76 ± 7.15 (n = 3 independent primary culture), Q331K:shPP1α = 38.71 ± 6.21 (n = 2 independent primary culture), Q331K:shPP1β = 21.69 ± 2.95 (n = 3 independent primary culture), Q331K:shPP1γ = 22.14 ± 1.44 (n = 3 independent primary culture), Q331K:I-2 = 14.45 ± 2.93 (n = 3 independent primary culture). Values are mean ± S.E.M. F(2,12) = 17.58, **P = 0.0012, ***P = 0.0007 for each CON group; F(4,14) = .101, ***P = 0.0007 for each GFP group; F(4,13) = 6.49, ^#P = 0.0245 for G93A:CON versus G93A:shPP1γ, ^#P = 0.0441 for G93A:CON versus G93A:I-2; F(4,9) = 7.405, ^#P = 0.0433 for Q331K:CON versus Q331K: shPP1β, ^#P = 0.0483 for Q331K:CON versus Q331K: shPP1γ, ^##P = 0.0080 by one-way ANOVA followed by Tukey’s multiple comparisons test. e Analysis of total neurite length in (a). GFP:CON = 100 ± 8.70 (n = 57 GFP+ neurons), GFP:I-2 = 115.90 ± 9.90 (n = 22 GFP+ neuron), GFP: shPP1α=130.3 ± 18.51 (n = 9 GFP+ neuron), GFP:shPP1β=72.12 ± 8.25 (n = 16 GFP+ neuron), GFP:shPP1γ=60.66 ± 5.05 (n = 29 GFP+ neuron), G93A:CON = 61.61 ± 3.10 (n = 91 GFP+ neuron), G93A:I-2 = 106.3 ± 7.51 (n = 27 GFP+ neuron), G93A:shPP1α = 93.39 ± 13.05 (n = 21 GFP+ neuron), G93A:shPP1β = 57.56 ± 4.52 (n = 23 GFP+ neuron), G93A:shPP1γ = 88.06 ± 11.45 (n = 26 GFP+ neuron), Q331K:CON = 42.39 ± 3.47 (n = 25 GFP+ neuron), Q331K:I-2 = 122.2 ± 10.61 (n = 21 GFP+ neuron), Q331K:shPP1α=128.8 ± 17.3 (n = 23 GFP+ neuron), Q331K:shPP1β = 60.97 ± 6.15 (n = 27 GFP+ neuron), Q331K:shPP1γ = 94.81 ± 11.55 (n = 37 GFP+ neuron),. Values are mean ± S.E.M. F(2,170) = 19.98, ****P < 0.0001 for each CON group;F(4,128) = 7.415, ^#P = 0.0106, ^##P = 0.0025 for GFP:I-2 versus GFP:shPP1γ, ^##P = 0.0057 for GFP:shPP1α versus GFP:shPP1γ; F(4,183)=9.91, ^#P = 0.0235, ^#P = 0.0089, ^###P = 0.0008 for each G93A group; F(4,128) = 10.63, ^##P = 0.0044, ^####P < 0.0001 for each Q331K group by one-way ANOVA followed by Tukey’s multiple comparisons test. f Analysis of total axonal length in (a). GFP:CON = 100 ± 10.33 (n = 56 GFP+ neuron), GFP:I-2 = 119.4 ± 11.51 (n = 22 GFP+ neuron), GFP: shPP1α = 141.4 ± 24.44 (n = 9 GFP+ neuron), GFP:shPP1β = 73.71 ± 9.06 (n = 16 GFP+ neuron), GFP:shPP1γ = 59.58 ± 6.13 (n = 29 GFP+ neuron), G93A:CON = 54.08 ± 3.02 (n = 86 GFP+ neuron), G93A:I-2 = 80.63 ± 4.57 (n = 27 GFP+ neuron), G93A:shPP1α = 02.8 ± 15.10 (n = 21 GFP+ neuron), G93A:shPP1β = 58.73 ± 6.10 (n = 23 GFP+ neuron), G93A:shPP1γ = 95.06 ± 13.3 (n = 26 GFP+ neuron), Q331K:CON = 49.70 ± 7.46 (n = 28 GFP+ neuron), Q331K:I-2 = 131.7 ± 15.21 (n = 21 GFP+ neuron), Q331K:shPP1α = 145.8 ± 20.67 (n = 23 GFP+ neuron), Q331K:shPP1β = 60.02 ± 7.01 (n = 27 GFP+ neuron), Q331K:shPP1γ = 98.4 ± 12.79 (n = 37 GFP+ neuron). Values are mean ± S.E.M. F(2, 167) = 15.91, ***P = 0.0001, ****P < 0.0001 for each CON group; F(4,127) = 5.083, ^#P = 0.0386 for each GFP group;F(4,178) = 9.565, ^#P = 0.0356, ^###P = 0.0002, ^####P < 0.0001 for each G93A group;F(4,131) = 4.236, ^#P = 0.0395, ^###P = 0.0005, ^####P < 0.0001 for each Q331K group by one-way ANOVA followed by Tukey’s multiple comparisons test.

a The histological image of NADH dehydrogenase activity and biochemical measurement of complex I activity on the MNs of the lumbar spinal cord in G93A and WT mice at 60 days of age. The inset graph shows the average of the complex I activity value. WT = 99.74 ± 3.83 (n = 4 mice), G93A = 62.9 ± 12.37 (n = 4 mice). Values are mean ± S.E.M. F(1,6) = 2.87, *P = 0.0284 by two-sided Student’s t-test. Scale bar, 100 μm. b The activity of mitochondrial complex I in primary cortical neurons cultured from WT and G93A mice following treatment with 25 µM Mdivi-1 or 80 nM I-2 for 72 h. WT:CON = 99.69 ± 6.81 (n = 3 independent primary culture), WT:Mdivi-1 = 99.37 ± 8.55 (n = 3 independent primary culture), WT:I-2 = 113.9 ± 9.12 (n = 4 independent primary culture), G93A:CON = 77.36 ± 5.33 (n = 5 independent primary culture), G93A:Mdivi-1 = 88.09 ± 7.70 (n = 8 independent primary culture), G93A:I-2 = 113.7 ± 7.09 (n = 7 independent primary culture). Values are mean ± S.E.M. F(1,6)=2.58, *P = 0.0421 for WT:CON versus G93A:CON by by two-sided Student’s t-test; F(2,17)=6.22, ^#P = 0.0110 for G93A:CON versus G93A:I-2, n.s. = not significant group by one-way ANOVA followed by Tukey’s multiple comparisons test. c The representative images and measurement of MMP in GFP+ primary cortical neurons following mALS gene transfection with 25 µM Mdivi-1 or 80 nM I-2 treatment for 72 h. GFP:CON = 1 ± 0.07 (n = 15 GFP+ primary cortical neurons), GFP:Mdivi-1 = 0.73 ± 0.08 (n = 16 GFP+ primary cortical neurons), GFP:I-2 = 0.92 ± 0.06 (n = 16 GFP+ primary cortical neurons), G93A:CON = 0.66 ± 0.06 (n = 23 GFP+ primary cortical neurons), G93A:Mdivi-1 = 0.64 ± 0.07 (n = 17 GFP+ primary cortical neurons), G93A:I-2 = 0.96 ± 0.07 (n = 22 GFP+ primary cortical neurons), Q331K:CON = 0.64 ± 0.07 (n = 17 GFP+ primary cortical neurons), Q331K:Mdivi-1 = 0.57 ± 0.04 (n = 11 GFP+ primary cortical neurons), Q331K:I-2 = 0.97 ± 0.10 (n = 16 GFP+ primary cortical neurons). Values are mean ± S.E.M. F(2,52) = 7.529, **P = 0.0030 for GFP:CON versus G93A:CON, **P = 0.0037 for GFP:CON versus Q331K:GFP; F(2,44) = 3.86, *P = 0.0268 for GFP:CON versus GFP:Mdivi-1; F(2,59) = 7.184, **P = 0.0048 for G93A:CON versus G93A:I-2; F(2,41) = 6.87, *P = 0.0119 for Q331K:CON versus Q331K:I-2 by one-way ANOVA followed by Tukey’s multiple comparisons test. Scale bar, 20 μm.

a Motor axons of zebrafish treated with 200 nM I-2 or 100 nM OA daily for 2 days following transient expression of G93A or Q331K mutations. The red dashed-line square indicates a magnified image of motor axons. The blue arrowhead indicates the ventral projection of the motor axon, and the yellow arrowhead indicates neuromuscular junction. b Analysis of axonal defects in (a). CON = 0.14 ± 0.36 (n = 14 zebrafish), G93A:Non = 2.93 ± 0.29 (n = 14 zebrafish), G93A:I-2 = 0.86 ± 0.23 (n = 14 zebrafish), G93A:OA = 0.73 ± 0.21 (n = 30 zebrafish), Q331K:Non = 3.71 ± 0.19 (n = 14 zebrafish), Q331K:I-2 = 1.00 ± 0.27 (n = 14 zebrafish), Q331K:OA = 1.43 ± 0.39 (n = 14 zebrafish). Values are mean ± S.E.M. F(2,39) = 81.83, ****P < 0.0001 for CON versus G93A:Non and CON versus Q331K:Non;F(2,55) = 21.78, ^####P < 0.0001 for G93A:CON versus G93A:I-2 and G93A:CON versus G93A:OA; F(2,39) = 25.12, ^&&&&P < 0.0001 for Q331K:CON versus Q331K:I-2 and Q331K:CON versus Q331K:OA by one-way ANOVA followed by Tukey’s multiple comparisons test. c Analysis of axonal length in (a). CON = 162.5 ± 2.77 (n = 6 zebrafish), G93A:Non = 122 ± 2.96 (n = 13 zebrafish), G93A:Mdivi-1 = 141.4 ± 2.45 (n = 13 zebrafish), G93A:I-2 = 151.1 ± 1.56 (n = 14 zebrafish), Q331K:Non = 23.3 ± 5.82 (n = 10 zebrafish), Q331K:Mdivi-1 = 141.3 ± 3.37 (n = 14 zebrafish), Q331K:I-2 = 145.3 ± 3.14 (n = 14 zebrafish). Values are mean ± S.E.M. F(2,26) = 21.27, ****P < 0.0001 for CON versus G93A:Non and CON versus Q331K:Non;F(2,37) = 39.83, ^####P < 0.0001 for G93A:CON versus G93A:Mdivi-1 and G93A:CON versus G93A:I-2, ^#P = 0.0156 for G93A:Mdivi-1 versus G93A:I-2;F(2,35 )= 7.755, ^&P = 0.0104 for Q331K:CON versus Q331K:Mdivi-1, ^&&P = 0.0017 for Q331K:CON versus Q331K:I-2 by one-way ANOVA followed by Tukey’s multiple comparisons test. d Immunostaining of cleaved caspase-3 and ChAT in 3-week-old MNs of control and mTDP-43 lines treated with 80 nM I-2 every other day for 2 weeks. Scale bar, 25 μm. e Measurement of mitochondrial length in 3-week-old MNs of control and mTDP-43 lines treated with 80 nM I-2 every other day for 2 weeks. CON:I-2(−) = 2.42 ± 0.07 (n = 38 ChAT+ neurons), CON:I-2(+) = 3.24 ± 0.13 (n = 37 ChAT+ neurons), mTDP-43:I-2(−) = 1.75 ± 0.06 (n = 38 ChAT+ neurons), mTDP-43:I-2(+) = 2.63 ± 0.11 (n = 36 ChAT+ neurons). Values are mean ± S.E.M. F(1,74) = 6.837, ****P < 0.0001 for CON:I-2(−) versus mTDP-43:I-2(−); F(1,73) = 5.538, ****P < 0.0001 for CON:I-2(−) versus CON:I-2(+); F(1,72) = 6.840, ****P < 0.0001 for mTDP-43:I-2(−) versus mTDP-43:I-2(+) for by two-sided Student’s t-test. f Quantification of cleaved caspase-3⁺ cell in human iPS-derived ChAT⁺ neuron in (d). CON:I-2(−) = 7.86 ± 0.42 (n = 3 iPS cell lines), CON:I-2(+) = 7.96 ± 0.89 (n = 3 iPS cell lines), mTDP-43:I-2(−) = 31.47 ± 3.34 (n = 3 iPS cell lines), mTDP-43:I-2(+) = 15.67 ± 2.06 (n = 3 iPS cell lines). Values are mean ± S.E.M. F(1,4) = 7.01, **P = 0.0022 for CON:I-2(−) versus mTDP-43:I-2(−); F(1,4) = 4.03, *P = 0.0158 for mTDP-43:I-2(−) versus mTDP-43:I-2(+) for by two-sided Student’s t-test. g Quantification of pyknotic cell in ChAT⁺ neuron in (d). Values are mean ± S.E.M. (N = 3 lines in CON and mTDP-43; *p < 0.05, **p < 0.01 in two-sided Student’s t-test). CON:I-2(−) = 14.37 ± 0.84 (n = 3 iPS cell lines), CON:I-2(+) = 10.43 ± 0.75 (n = 3 iPS cell lines), mTDP-43:I-2(−) = 7.11 ± 4.04 (n = 3 iPS cell lines), mTDP-43:I-2(+) = 18.99 ± 1.71 (n = 3 iPS cell lines). Values are mean ± S.E.M. F(1,4) = 5.514, **P = 0.0053 for CON:I-2(−) versus mTDP-43:I-2(−); F(1,4) = 4.131, *P = 0.0145 for mTDP-43:I-2(−) versus mTDP-43:I-2(+) for by two-sided Student’s t-test.

Also, abnormal PP1 activity decreases the mitochondrial complex I activity, leading to axonal degeneration. Thus, the blockade of PP1 activity reduces the ALS-induced mitochondrial fission and improves mitochondrial functions such as the complex I activity of the mitochondrial electron transport chain and membrane potential, resulting in improvement of axonal impairment. Therefore, inhibition of excessive PP1 activity may promote MN survival by improving mitochondrial function.