Additional Supporting Information may be found in the online version of this article.

CNE_23266_sm_SuppFig1.tif13933KSupplementary Figure 1: (magenta-green version of Figure 1 for the assistance of color-blind readers): Decreases in the number of NeuN-IR motor neurons occur before and are more intense than decreases in calbindin-IR Renshaw cells. (a-d) Low magnification confocal microscopy images of NeuN-IR cells and calbindin-IR cells in the lumbar 5 segment of a WT (a-b) and a Sod1G93A spinal cord (c-d), both 3 months of age. Laminar boundaries are drawn according to NeuN-immunoreactivity and established cytoarchitectonic criteria. Calbindin-IR Renshaw cells (RCs) are indicated in ventral LVII and LIX (arrows in b and d). (e-g) Ventral horn of Sod1G93A mice of 1, 3 and 4 months showing NeuN (green) and calbindin (white) immunoreactivities. NeuN-IR motor neurons are depleted in 3 and 4 months animals, but most calbindin-IR Renshaw cells remain. (h) Quantification of the number of NeuN-IR motor neurons, (i) calbindin-IR Renshaw cells and (j) ventral horn NeuN-IR interneurons per ventral horn in 50 μm thick sections. Black and grey bars are wildtype and Sod1G93A averages obtained by pooling data from 2-4 animals at each age. Significant depletions in the number of NeuN-IR motor neurons were found at 3 and 4 months of age (asterisks, p<0.001 one-way ANOVA, post-hoc Bonferroni corrected tests). Depletions in the number of Renshaw cells and interneurons are smaller and are significant at 4 months of age (asterisks, p<0.05). Error bars indicate SEMs. Scale bars: in d, 500 μm (also for a-c); in g, 200 μm (also for e and f).
CNE_23266_sm_SuppFig2.tif11564KSupplementary Figure 2: (magenta-green version of Figure 2 for the assistance of color-blind readers): Expression of calbindin and chrna2 downregulate with different time courses in Renshaw cells as disease progresses. (a-f) Images of ISH/IHC double staining using a calbindin probe (green) and calbindin antibodies (purple) in the ventral horn Renshaw cell area of lumbar spinal cord sections from WT and Sod1G93A mice at day 40, day 60, day 80, day 100 and end stage (120-150 days). (g-l) Brightfield images of ISH using a chrna2 probe at the same ages. Quantifications demonstrate that the number of cells expressing calbindin mRNA (m) as well as the number of cells expressing calbindin protein (n) are only decreased at disease end stage compared to WT (p<0.001 one-way ANOVA, post-hoc Bonferroni tests of all ages). In contrast, the number of chrna2 mRNA expressing cells (o) are significantly decreased by day 80 and continued to decrease at day 100 and at end stage compared to WT (all p<0.001, tests as above). Note a more profound downregulation of calbindin mRNA (m) compared to calbindin immunoreactivity (n). (p-q) Double IHC/ISH of calbindin (pink, IHC) and chrna2 (green, ISH) on lumbar spinal cord from WT (p) and Sod1G93A mice at end stage (120-150 days old) (q). High magnification images of WT (p1) and Sod1G93A (q1) show Renshaw cells co-expressing calbindin and chrna2. Note decreased expression of chrna2 in Sod1G93A and thus fewer co-localization with calbindin immunoreactivity. (r) Venn-diagrams illustrate the relative sizes of cell populations expressing either calbindin (pink circles), chrna2 (green circles) or both (yellow circles) for WT or Sod1G93A mice at end stage. The number of calbindin-IR cells in the ventral Renshaw area significantly decreased by 42% in Sod1G93A mice compared to WT, chrna2 positive cells decreased by 77% and thus cells showing co-localization between both markers decreased by 83%; from 93.5% in WT to 28% in Sod1G93A mice (p<0.001 for all markers, t-tests). At end stage, the number of calbindin-IR cells in the Renshaw area was significantly higher than the number chrna2 positive cells (p<0.05) or cells with co-localization of both markers (p<0.01, not indicated in the figure one-way ANOVA, post-hoc Bonferroni tests). No significant differences were detected in WTs. Error bars indicate SEMs. Error bars indicate SEMs. Scale bars: in d, 500 μm (also for a-c); in g, 200 μm (also for e and f); in p, 100 μm (applies also to q); in p1, 25 μm (applies also to q1).
CNE_23266_sm_SuppFig4.tif7692KSupplementary Figure 3: (magenta-green version of Figure 4 for the assistance of color-blind readers): VAChT-IR synapses on Renshaw cells are profoundly altered in Sod1G93A mice, starting at 2 months of age. (a) Low and (b) high magnification confocal images of VAChT immunoreactivity (green) and Renshaw cells (red) in WT animals. (c-e) Images of calbindin-IR Renshaw cell dendrites in 1, 2 and 3 month old Sod1G93A mice. VAChT-IR fills most boutons in contact with WT Renshaw cells VAChT-IR boutons of 2 month Sod1G93A animals appear hollow and the immunoreactivity concentrates at the periphery of the bouton (e). Few boutons remain in 3 month Sod1G93A mice. (f) Quantification of VAChT-IR contacts on Renshaw cell dendrites. No significant differences in contact density were found in WTs of different ages (p=0.653, One-Way ANOVA). Contact density was significantly reduced in 3 month (p<0.01, Bonferroni corrected t-test) and 4 month (p<0.001, Bonferroni corrected t-test) old Sod1G93A animals compared to their age-matched controls, but not in 1 and 2 month Sod1G93A animals. Error bars indicate SEMs. Scale bars: in a, 250 μm; in b-e, 10 μm. (A version of this figure in which white-green has been changed for magenta-green is supplied as supplementary figure 3).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.