Fig. 4

Inhibition of CXCR4 blocks testosterone-dependent remyelination. a–c Double labeling of MBP and GFAP on spinal cord cross sections of castrated males at 30 dpl. Cell nuclei were counterstained in blue with DAPI. Within the LPC lesion, delimited by the dotted line, both MBP (a, b) and GFAP (a, c) staining were nearly absent in males receiving an empty implant (LPC), but MBP and GFAP co-occurred in males receiving an implant filled with T (LPC + T). AMD3100 blocked the T-dependent colocalization of MBP and GFAP inside the lesion (LPC + T + AMD3100). d and f, T also restored MBP (d) and GFAP (e) immunostaining within the LPC lesion of ovariectomized females, and its effect was inhibited by AMD3100. Data are presented as means ± S.E.M. (one-way ANOVA with Tukey's multiple comparisons tests). Asterisks mark significant differences. ****P < 0.0001 ***P < 0.001, **P < 0.01. Scale bar: 20 µm. f Electron microscopy images of axons and their myelin sheaths within the ventral spinal white matter of gonadally intact males (Control), or within the LPC lesion at 30 dpl of castrated males treated or not with T. Yellow stars indicate myelinated axons. Scale bar: 0.5 µm. g Correlations between g-ratios and axon diameters in control males and within the LPC lesion of castrated males treated with T or T + AMD3100