Extended Data Fig. 9: De novo activity of DNMT1 independent of DNMT3A in the Stella^−/− oocyte.

a, Stella^−/−Dnmt3a^−/− and Stella^−/−Dnmt1^−/− germinal vesicle oocytes stained with anti-DNMT3A (left) and anti-DNMT1 (M377) (middle). Images shown are representative of two independent experiments. Scale bars, 10 μm. b, c, Venn diagrams show the Uhrf1, Dnmt1 and Dnmt3a-dependent ooAMRs in Stella^+/− oocytes (b, ooAMRs that methylated in Stella^+/− oocytes but unmethylated in Stella^+/−Uhrf1^−/−, Stella^+/−Dnmt1^−/− and Stella^+/−Dnmt3a^−/− oocytes) and in Stella^−/− oocytes (c, ooAMRs that methylated in Stella^−/− oocytes but unmethylated in Stella^−/−Uhrf1^−/−, Stella^−/−Dnmt1^−/− and Stella^−/−Dnmt3a^−/− oocytes). ooAMRs that have methylation level >75% or <25% are defined as methylated or unmethylated, respectively. d, Hierarchical clustering heat map shows the methylation status of Dnmt1-dependent ooAMRs in Stella^−/− oocytes (definition as in c) from D5 to MII and in all types of MII oocytes. e, Violin plot of the methylation levels of 100-bp CG tiles for Stella^+/− oocytes and Stella^−/−Dnmt3a^−/− oocytes. The violin plot is shown as the combination of a box plot (midline, median; outer edges, first and third quartiles) and a kernel density. Data analysed from two (Stella^−/−) or three (Stella^−/−Dnmt3a^−/−) biological replicates. f, Distribution of ooAMRs in active or inactive genomic regions. g, DNA methylation of normal ooAMRs and extra ooAMRs at different stages of oocyte development. The median methylation levels of 100-bp tiles are connected by lines, and black bars represent the third and first quartiles. For both Stella^+/− and Stella^−/− oocytes: D5, n = 3; D10, D20, MII, n = 2 independent experiments. h, Box plots showing RPKM values from RNA-seq of normal ooAMRs that are categorized by their methylation level in Stella^−/− Dnmt3a^−/− MII oocytes. Gene expression levels in Stella^+/− Dnmt3a^−/− and Stella^−/−Dnmt3a^−/− MII oocytes are shown. The midline of the box indicates the median, the outer edges represent the first and third quartiles, and values outside 1.5× IQR are designated outliers and are not shown. The upper and lower whiskers show the maximum and minimum values that are not outliers, respectively. Two-way ANOVA (RPKM ~ methylation × genotype) was performed, methylation level has a statistically significant effect on the RPKM level (degrees of freedom (Df) = 2, F = 74, P < 2 × 10⁻¹⁶), but not the genotype (Df = 1, F = 1.5, P = 0.22) or its interaction with the methylation level (Df = 2, F = 0.13, P = 0.88). Data analysed from two biological replicates.