AAV carrying the above construct (shown in A) were injected into the corpus callosum of the newborn rat brain (at P6). Left: transmission electron microscopy (TEM) micrographs from rat brains (P21) injected with shmiR-Scrambled or shmiR-RNLTR12-int. Scale bars, 1 μm. Right (top): scatter plots representing G-ratio in relation to axon diameter. n = 183 (for shmiR-Scrambled) to 203 (shmiR-RNLTR12-int) axons from three different rats. The fitted linear regression lines were shown, and the statistical significance (ANCOVA) of their slope and intercepts difference were determined. For slope difference: p > 0.05, indicating that the relation between G-ratio and the axon diameter does not change due to inhibition of RNLTR12-int. For intercept difference:**p < 2 × 10−16, indicating that axons of a given diameter have a significantly higher G-ratio in shmiR-RNLTR12-int compared with shmiR-Scrambled cases. Right (bottom): myelinated and unmyelinated axons were counted and plotted as a percentage. n = 3 different rats. Mean ± SEM**adjusted p < 0.01, two-way ANOVA (Bonferroni post test). See alsoFigure S5.
