Efficient Remyelination Requires DNA Methylation

DNA methyltransferases are differently expressed in adult OPCs during remyelination. A, Schematic of the lysolecithin-induced focal demyelination and of the area of NWM used for quantification. B, Representative DNMT1, NKX2.2, and CC1 stainings in NWM and at 5, 14, and 21 dpl (white arrowheads indicate double-positive cells). C, Quantification of the number of double DNMT1⁺ and NKX2.2⁺ cells at 5, 14, and 21 dpl, compared with NWM^a. D, Quantification of the percentage of double DNMT1⁺ and NKX2.2⁺ or CC1⁺ cells at 5, 14, and 21 dpl, compared with NWM^b. E, Representative DNMT3A, NKX2.2, and CC1 stainings in NWM and at 5, 14, and 21 dpl (white arrowheads indicate double-positive cells). F, Quantification of the number of double DNMT3A⁺ and CC1⁺ cells at 5, 14, and 21 dpl, compared with NWM^c. G, Quantification of the percentage of double DNMT3A⁺ and NKX2.2⁺ or CC1⁺ cells at 5, 14, and 21 dpl, compared with NWM^d. H, Representative 5mC and OLIG2 staining in NWM and at 5, 14, and 21 dpl (white arrowheads indicate high-5mC⁺/OLIG2⁺ cells). Representative low-, medium-, and high-5mC cells are shown below. I, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells at 5, 14, and 21 dpl, compared with NWM^e. Scale bar = 20 µm. Data are mean ± SEM. n = 4–6 animals, three sections per animal. *p < 0.05, **p < 0.01, ***p < 0.001 (ANOVA).

Ablation of Dnmt1 or Dnmt3a does not impair oligodendrocyte differentiation or their methylation levels in control conditions. A, Representative OLIG2 and CC1 staining in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl NWM spinal cords. B, Quantification of OLIG2⁺ and CC1⁺ cell densities in NWM^f (p = 0.3537, p = 0.3803). C, Representative OLIG2 and CC1 staining in tamoxifen-treated Plp^{+ / +};Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt3a^fl/fl NWM spinal cords. D, Quantification of OLIG2⁺ and CC1⁺ cell densities in NWM^g (p = 0.8926, p = 0.5109). E, Representative OLIG2 and CC1 staining in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl;Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl;Dnmt3a^fl/fl NWM spinal cords. F, Quantification of OLIG2⁺ and CC1⁺ cell densities in NWM^h (p = 0.3136, p = 0.2173). G, Representative 5mC and OLIG2 stainings in tamoxifen-treated Plp^{creER(t) / +};Dnmt1^fl/fl, Plp^{creER(t) / +};Dnmt3a^fl/fl, and Plp^{creER(t) / +};Dnmt1^fl/fl;Dnmt3a^fl/fl NWM spinal cords (white arrowheads indicate high-5mC⁺/OLIG2⁺ cells). H, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl NWM^j. I, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells in tamoxifen-treated Plp^{+ / +};Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt3a^fl/fl NWM^jj. J, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl;Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl;Dnmt3a^fl/fl NWM^k. Scale bar = 50 µm. Data are mean ± SEM. n = 4–6 animals, three sections per animal (Student’s t test, ANOVA).

Ablation of Dnmt3a and both Dnmt1 and Dnmt3a impairs oligodendrocyte differentiation during remyelination. A, Representative OLIG2 and CC1 staining at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl spinal cords. B, Quantification of OLIG2⁺ and CC1⁺ cell densities and CC1⁺/OLIG2⁺ cells percentage at 14 dpl^l (p = 0.7955, p = 0.3573, p = 0.9689). C, Representative OLIG2 and CC1 staining at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt3a^fl/fl spinal cords. D, Quantification of OLIG2⁺ and CC1⁺ cell densities and CC1⁺/OLIG2⁺ cells percentage at 14 dpl^m (p = 0.7851, p = 0.0550, p = 0.0149). E, Representative OLIG2 and CC1 staining at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl;Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl;Dnmt3a^fl/fl spinal cords. F, Quantification of OLIG2⁺ and CC1⁺ cell densities and CC1⁺/OLIG2⁺ cells percentage at 14 dplⁿ (p = 0.1510, p = 0.0357, p = 0.0006). G, Quantification of DNMT1 and DNMT3A expression in CC1⁺ cells at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl, Plp^{+ / +};Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt3a^fl/fl, Plp^{+ / +};Dnmt1^fl/fl;Dnmt3a^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl;Dnmt3a^fl/fl spinal cords, to detect eventual compensation between DNMTs at the protein level° (p = 0.0075, p = 0.0505, p = 0.0074, p = 0.0053, p = 0.0072, p = 0.0012). Scale bar = 20 µm. Data are mean ± SEM. n = 4–6 animals, three sections per animal. *p < 0.05, **p < 0.01, ***p < 0.001 (Student’s t test).

Ablation of Dnmt3a and both Dnmt1 and Dnmt3a impairs methylation levels in oligodendroglial cells during remyelination. A, Representative 5mC and OLIG2 staining at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl spinal cords (white arrowheads indicate high-5mC⁺/OLIG2⁺ cells). B, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells at 14dpl^p. C, Representative 5mC and OLIG2 staining at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt3a^fl/fl spinal cords (white arrowheads indicate high-5mC⁺/OLIG2⁺ cells). D, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells at 14dpl^q. E, Representative 5mC and OLIG2 staining at 14 dpl in tamoxifen-treated Plp^{+ / +};Dnmt1^fl/fl and Plp^{creER(t) / +};Dnmt1^fl/fl;Dnmt3a^fl/fl spinal cords (white arrowheads indicate high-5mC⁺/OLIG2⁺ cells). F, Quantification of low-, medium-, and high-5mC levels in OLIG2⁺ cells at 14 dpl^r. Scale bar = 100 µm. Data are mean ± SEM. n = 4–6 animals, three sections per animal. **p < 0.01, ***p < 0.001 (ANOVA).