Centromere damage is microtubule tension–dependent. (A and B) Representative images show centromeres in oocytes treated with 40 µM 5MM-ASO or Cen-ASO. Yellow boxes indicate the bivalents enlarged in the inset (blue: Hoechst; green: MajSAT-mClover; red: MajSAT-mRuby or mCherry-CenpC). Scale bars: 5 µm (yellow) and 2 µm (white). (C) The proportion of centromeres classified by damage location in B (287 bivalents were imaged at 6 h after NEBD; data from three independent experiments; ****, P < 0.0001, ANOVA with Tukey’s post hoc test). (D) Representative time-lapse images showing the time of centromere damage initiation (gray: H2B-mCherry; green: MajSAT-mClover). Monastrol was added at the beginning of oocyte maturation in culture. Yellow arrow: damaged centromeres. Time from NEBD. Scale bar: 5 µm. (E) Time-lapse images show the process of a single representative centromere signal splitting (gray: H2B-mCherry; green: MajSAT-mClover; yellow arrows: paired centromere signals; Video 2). Time from NEBD. Scale bar: 2 µm. (F) Number of damaged centromeres per oocyte are plotted by time (A versus B, P = 0.0235; B versus C, P = 0.0023; A versus C, P < 0.0001; ANOVA with Tukey’s post hoc test). 35 oocytes were tested in three independent experiments. (G) Number of damaged centromeres per oocyte (5MM-ASO: n = 29; Cen-ASO: n = 35; monastrol: n = 16; ****, P < 0.0001, ANOVA with Tukey’s post hoc test). (H) Schematic showing the mechanism of MajSAT DNA damage. The Cen-ASO activates RNaseH, which cleaves the DNA–RNA heteroduplex and therefore leads to degradation of Cen-RNA, inducing the instability of centromeric DNA. Data from three independent experiments. Error bars indicate SD.