Fig 2: CENP-B drives de novo CENP-A and CENP-C assembly at endogenous centromeres Schematic representation of experiments in B–E.Timeline of experiment in C-E.Representative immunofluorescence images showing CENP-A and CENP-C reloading in the indicated conditions. Cells with centromeric CENP-A/CENP-C are marked with a red dashed contour line, while a yellow contour lines mark cell without centromeric CENP-A. The centromere position is marked using immunofluorescence staining for CENP-B. White dashed lines contour DAPI stained nucleus of representative cells. Scale bar, 5 µm.Bar plot showing the percentage of cells with CENP-A and CENP-C co-localization after a CENP-A/COFF/ON cycle in the indicated conditions. Error bars show SD from 5 independent experiments. Unpaired t-test, ***P = 0.0002. Each dot represents one experiment (> 30 cells analyzed for experiment).Box plot of CUT&RUN–qPCR for CENP-A (green) or CENP-C (orange). Box plot shows median, 25th and 75th percentiles, and whiskers show minimum and maximum values. Results from 3 independent experiments including combined data from qPCRs using alpha-satellite primers and primers binding at the centromere of chromosome 4. Enrichment is measured relative to the IgG control and normalized to Alu repeats. Mann–Whitney test, CENP-A: **P = 0.0087, *P = 0.0411, CENP-C: **P = 0.0022.

Fig 3: Sensitivity to X-irradiation upon CENP-A overexpression is dependent on functional p53.a Scheme of colony formation assays (CFAs) and change of p53 status with relative CENP-A protein levels over time for TetOn-CENPA-FLAG-HA cell lines. WT p53 is transiently activated in HeLa cells by a short heat shock, while p53 is stably inactivated in MCF10-2A cells by expression of a p53 dominant-negative peptide (p53-DN). b Temporary change of p53 status in HeLa cells by 1 h heat shock (HS). Western blot of HeLa cell TCEs at indicated time points following 1 h at 42 or 37 °C control. Temporary increase of p53 and subsequently p21. 1x load = ~33,000 cells. Primary antibodies are indicated on the right. # = high sensitivity ECL exposure. H4 used as loading control. c Increase of p53 levels in HeLa cells by heat shock. Western blot of TCEs after 1 h at 42 or 37 °C control. Loaded at 8x, 4x, 2x, 1x concentrations, where 1x load = ~12,500 cells. Primary antibodies are indicated on the right. # = high sensitivity ECL exposure. H4 used as loading control. Representative western blot from three independent experiments is shown. d HeLa S3 CFAs pertaining to scheme in (a). Each dot represents a single biological replicate. Plots show mean and 95% confidence interval. Statistical significance tested by two-tailed Welch’s t test with Bonferroni cutoff at a p value of 0.0125 (a = 0.05). * = p value < 0.01. e Change of p53 status in MCF10-2A TetOn-CENPA-FLAG-HA cells. Western blot of TCEs after 24 h of 10 µM Nutlin-3 treatment or DMSO control, as indicated. Dominant negative p53 peptide stabilizes p53 and suppresses its activation of p21. 1x load corresponds to total protein extract from ~30,000 cells. Primary antibodies are indicated on the right. # = high sensitivity ECL exposure. H4 used as loading control. Representative western blot from three independent experiments is shown. f MCF10-2A CFAs pertaining to scheme in (a). Each dot represents a single biological replicate. Plots show mean and 95% confidence interval. Statistical significance tested by two-tailed Welch’s t test with Bonferroni cutoff at a p value of 0.0125 (a = 0.05). *** = p value < 0.0001. See also Supplementary Fig. 1C, D for CFAs following range of Dox and X-irradiation doses and Supplementary Fig. 1E for representative CFA images.

Fig 4: PCR screening and Western verification of EYFP-knockin clones.(A) Schematic outlining PCR screening strategy. Grey regions indicate sequences outside the targeting construct. A primer set in which one primer complementary to a region outside the targeting construct and one primer inside the targeting construct (EYFP) was used to amplify a 2088 bp fragment unique to targeted clones. 47 monoclonal lines produced in the pipeline outlined in Figure 1B were screened for targeting. Water and a previously targeted HeLa clone (c+) were used as negative and positive clones respectively. (B) PCR verification of targeting. Representative clones are shown in which the left and right flanks are amplified using primer set in which one primer anneals outside the targeted area and one to EYFP sequences. Sizes of marker fragments (M) and of expected products are indicated. WT, wild type uninfected parent cells. KI, knockin clones. Asterisks indicate non-specific background products (C) Wild type hTERT-RPE or HeLa or targeted knockin clones were processed for SDS-PAGE and immunoblotting with an anti-CENP-A antibody revealing novel high molecular weight CENP-A-EYFP fusion protein in targeted clones only. Proteins from both the targeted CENP-A-EYFP and untagged CENP-A alleles are detected.

Fig 5: Aneuploidy impairs centromere mechanical maturation. a Live cell imaging of ANRPE-1-GI cells expressing CenpA-GFP at the outer centromere. Images are maximum intensity projections from full-cell-volume image series, and are representative of the described mitotic stages: early-prometaphase (left), late-prometaphase (center), and metaphase (right). b Histogram of chromosome numbers for a sample of RPE-1-GI cells differentiating between diploid cells and aneuploid cells. ANRPE-1-GI cells are also subgrouped by quartile (Q1, light blue–Q4, dark blue\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$;n_{{\mathrm{eP}}}^{{\mathrm{Ce}}}$$\end{document};nePCe = 48, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n_{{\mathrm{lP}}}^{{\mathrm{Ce}}}$$\end{document}nlPCe = 80, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n_M^{{\mathrm{Ce}}}$$\end{document}nMCe = 110). c–f Mean spring constant by mitotic stage for ANRPE-1-GI chromosomes by quartile of increasing chromosome numbers as in panel b (Q1, light blue filled data point–Q4, dark blue filled data point). Data from DIPRPE-1-GI cells included for comparison (green unfilled data points; n = 131 EP, 136 LP, 178 M, all panels). Error bars represent standard error. All n-values listed are chromosome numbers. g, h Relationship between percent change in mean stiffness from g early to late-prometaphase or h late-prometaphase to metaphase and cell ploidy for RPE-1-GI cells. Data points include DIPRPE-1-GI cells (green) and ANRPE-1-GI cells by quartile (Q1, light blue–Q4, dark blue). X-axis error bars represent IQR; Y-axis error bars represent standard error. Fit to an exponential regression model is indicated by the dashed-line and the listed R2, RMSE, and p-values. i Comparison of mean mitotic timing between DIPRPE-1-GI cells (teal) and ANRPE-1-GI cells (blue). Sample sizes (n) reflect the number of cells in each group. Error bars represent standard error; dots represent individual cells. See also Supplementary Figs. 5 and 7