Fig. 4

Morphological changes in spiny excitatory neurons of S1 in EAE, and PLZ-treated EAE, at clinical onset. a Representative maximum z-projected images showing appearance and density of spines on dendritic segments from spiny excitatory neurons in S1. Spines were visualized by reflectance-mode (488 nm) laser-scanning confocal microscopy on Golgi-Cox stained brains from VEH-treated CFA (CFA), VEH-treated EAE (VEH), and PLZ-treated EAE (PLZ) animals at clinical onset. b Mean (±S.E.) dendritic-spine densities assessed from the branches of spiny neurons in layers 2/3 of S1 from CFA-VEH (n  = 42 dendritic segments, 4 animals), EAE-VEH (n  = 66 dendritic segments, 8 animals), and EAE-PLZ mice (n  = 78 dendritic segments, 9 animals). Dendritic segments from EAE-VEH animals exhibited significantly increased spine-densities compared to segments from CFA-VEH mice. This increase was localized almost exclusively to the tertiary basilar branches (CFA-VEH n  = 12, EAE-VEH n  = 13, EAE-PLZ n  = 14 dendritic segments). Daily treatment with PLZ from 7 dpi prevented or reversed this increase—mean spine-densities along segments from EAE-PLZ animals did not significantly differ from CFA controls (Kruskal-Wallis one-way ANOVA on ranks; “all-branches” p = 0.032; tertiary-basilar branches p = 0.007, all post hoc comparisons vs. CFA controls by Dunn’s method). c Mean (±S.E.) dendritic-spine densities assessed from the branches of spiny neurons in layer 4 of S1 from CFA-VEH (n  =  36 dendritic segments, 4 animals), EAE-VEH (n  = 58 dendritic segments, 8 animals), and EAE-PLZ mice (n  = 83 dendritic segments, 10 animals). Dendritic segments from EAE-VEH animals exhibited significantly increased spine-densities compared to segments from CFA-VEH mice. This increase was also specifically significant for second-order branches (CFA-VEH n  = 23, EAE-VEH n  = 43, EAE-PLZ n  = 54 dendritic segments). Daily treatment with PLZ prevented or reversed this increase—mean spine densities along segments from EAE-PLZ animals did not significantly differ from CFA-controls but were significantly reduced vs. the EAE-VEH group (“all-branches” analyzed by Kruskal-Wallis one-way ANOVA on ranks, p < 0.001; post hoc comparisons vs. CFA controls by Dunn’s method. Secondary branches analyzed by one-way ANOVA, p < 0.001, all pairwise post hoc comparisons by SNK test)