Centrosome amplification primes ovarian cancer cells for apoptosis and potentiates the response to chemotherapy

Abstract

Centrosome amplification is a feature of cancer cells associated with chromosome instability and invasiveness. Enhancing chromosome instability and subsequent cancer cell death via centrosome unclustering and multipolar divisions is an aimed-for therapeutic approach. Here, we show that centrosome amplification potentiates responses to conventional chemotherapy in addition to its effect on multipolar divisions and chromosome instability. We perform single-cell live imaging of chemotherapy responses in epithelial ovarian cancer cell lines and observe increased cell death when centrosome amplification is induced. By correlating cell fate with mitotic behaviors, we show that enhanced cell death can occur independently of chromosome instability. We identify that cells with centrosome amplification are primed for apoptosis. We show they are dependent on the apoptotic inhibitor BCL-XL and that this is not a consequence of mitotic stresses associated with centrosome amplification. Given the multiple mechanisms that promote chemotherapy responses in cells with centrosome amplification, we assess such a relationship in an epithelial ovarian cancer patient cohort. We show that high centrosome numbers associate with improved treatment responses and longer overall survival. Our work identifies apoptotic priming as a clinically relevant consequence of centrosome amplification, expanding our understanding of this pleiotropic cancer cell feature.

Fig 1. Centrosome amplification enhances cell death in response to combined chemotherapy.

(A) Single-cell live-imaging workflow. OVCAR8 cells expressing H2B-RFP and inducible for PLK4 overexpression are exposed to DMSO or 1 μg/ml Doxycycline (DOX) for 72 h to induce centrosome amplification. PLK4Ctl and PLK4OE cells are then filmed during 72 h of chemotherapy and lineages are tracked over multiple generations. Representative images of mitotic behaviors and cell fates are shown with the color-coded legends used in the subsequent panels. White arrows point to defects. Lineage analysis consists in counting for each starting cell, the number of cells adopting the different fates (see panel B). Generation analysis consists in determining the percentage of a generation that will adopt the different fates (see panel C). Mitosis and fate correlation consist in determining the percentage of cells adopting the different fates, depending on the behavior of the mother cell during mitosis (see panel D). (B) Bar graphs showing the averages and SEM of the number of cells per lineage adopting the indicated fates (legends in panel A). A minimum of 20 lineages were analyzed from 2 independent experiments, statistical tests: Fisher’s exact test on the number of cell death events (pooling death in interphase and in mitosis). Numbers on the top of each graph represent the number of cells analyzed per condition. (C) Bar graphs showing the average and SEM of the percentages of cells undergoing indicated fates (legends in panel A). Two independent experiments, statistical tests: Fisher’s exact test on the number of cell death events (pooling death in interphase and death in mitosis). Numbers on the top of each graph represent the number of cells analyzed per condition. (D) Vertical axis: Bar graphs showing the average and SEM of the percentages of mitotic phenotypes (legends in panel A); 137 and 133 cell divisions were analyzed from 2 independent experiments, statistical test: Fisher’s exact test on the number of multipolar divisions. Horizontal axis: Bar graphs showing the average and SEM of the percentages of cells undergoing indicated fates (legends in panel A) according to the mitotic behavior of mother cells, with bar width depending on the proportion of cells displaying a given mitotic phenotype. Two independent experiments, statistical test: Fisher’s exact test on the number of No mis-segregation progeny (progeny of blue mitosis) dying in mitosis and interphase, p-value = 0,0005. (E, F) Single-cell profiles of PLK4Ctl (E) and PLK4OE (F) undergoing Carboplatin + Paclitaxel exposure. Each row corresponds to 1 cell (legends in panel A). Data for Fig 1 can be found in S1 Data.

Fig 2. Centrosome amplification enhances cell death in response to Carboplatin independently of chromosome mis-segregation.

(A, B) Single-cell profiles of PLK4Ctl (A) and PLK4OE (B) undergoing Carboplatin exposure. For comparison with untreated controls—S2F Fig. Color coding of mitosis and fates legends in panel C. (C) Legends for panels A, B, D, E, F, and G, as defined in Fig 1A. (D) Bar graphs showing the averages and SEM of the number of cells produced per lineage, adopting the indicated fates (legends in panel 1A). A minimum of 31 lineages were analyzed from 2 independent experiments, statistical tests: Fisher’s exact test on the number of cell death events (pooling death in interphase and in mitosis). Numbers on the top of each graph represent the number of cells analyzed per condition. (E) Bar graphs showing the average and SEM of the percentages of cells undergoing indicated fates (legends in Fig 1A and 1C). Two independent experiments, statistical tests: Fisher’s exact test on the number of cell death events (pooling death in interphase and death in mitosis). Numbers on the top of each graph represent the number of cells analyzed per condition. (F) Vertical axis: Bar graphs showing the average and SEM of the percentages of mitotic phenotypes (legends in Fig 1A and 1C); 142 and 146 cell divisions analyzed from 2 independent experiments, statistical test: Chi-square test. Horizontal axis: Bar graphs showing the averages and SEM of the percentages of cells undergoing indicated fates (legends in Fig 1A and 1C) according to the mitotic behavior of the mother cell, with bar width depending on the proportion of cells according to their mitotic behavior. Two independent experiments, statistical test: Fisher’s exact test on the number of No mis-segregation progeny (progeny of blue mitosis) dying in mitosis and interphase, p-value = 0,0003. (G) Scatter dot plot graphs showing time of mitotic entry, with median and interquartile range. Cells were classified depending on mitotic phenotypes with color-code defined in panel C. Two independent experiments with a minimum of 48 mitosis analyzed per category. Statistical tests: Kruskal–Wallis with Dunn’s multiple comparisons tests. (H, I) Single-cell profiles of FUCCI PLK4Ctl (H) and PLK4OE (I) cells undergoing Carboplatin exposure. Times in G1, S/G2 and mitosis, as well as death in each of these phases are color-coded as indicated. (J) Scatter dot plots of the time of mitotic entry depending on cell cycle phase at movie start, with median and interquartile range. Two independent experiments with a minimum of 10 times analyzed per category. Statistical tests: Kruskal–Wallis with Dunn’s multiple comparisons tests. (K) Bar graphs showing the average and SEM of the percentage of cells adopting the indicated fates (legends in panel I). Two independent experiments, statistical test: Fischer’s exact test on the number of cells dying (irrespective of the cell cycle phase). Numbers on the top of each graph represent the number of cells analyzed per condition. Data for Fig 2 can be found in S2 Data.

Fig 3. Centrosome amplification enhances mitochondrial outer membrane permeabilization in response to Carboplatin.

(A) Representative western blot detecting Caspase-3 cleavage. (B) Average and SEM of cleaved caspase 3 protein levels from 2 independent experiments, normalized to levels measured in Carboplatin treated PLK4Ctl cells at 72 h. (C) Bar graphs showing the average and SEM of the percentage of cells in specified Annexin V-APC/PI gates analyzed by flow cytometry. Four replicates obtained from 2 independent experiments, with a minimum of 60,000 cells analyzed per condition and replicate. Statistical test: comparison of the percentage of Annexin V positive cells, using ANOVA with Sidak’s multiple comparison test. Representative cytometry profiles can be found in the S1_Appendix. (D) Representative images of cells labeled with DAPI (gray) and antibodies against Cytochrome C (cyan) and centrosomes in pink. White arrows indicate dead cell debris, pink arrows indicate cells that have released Cytochrome C in the cytoplasm, M indicates mitotic cells. Representative insets are shown for individual Q-VD-Oph treated cells. (E) Bar graphs showing the average and SEM of the percentages of indicated cell populations. Two independent experiments, statistical test: Fisher’s exact test on the number of cells releasing Cytochrome C. Numbers on the top of each graph represent the number of cells analyzed per condition. (F) Left: Bar graphs showing the average and SEM of the percentages of cells with indicated nuclear phenotypes, within the populations releasing Cytochrome C. Two independent experiments, statistical test: Fisher’s exact test on the number of cells with fragmented nuclei. Numbers on the top of each graph represent the number of cells analyzed per condition. Right: Representative images of DAPI-stained nuclei with indicated phenotypes. Data for Fig 3 can be found in S3 Data.

Fig 4. Centrosome amplification primes for mitochondria outer membrane permeabilization.

(A) Dose-response of PLK4Ctl and PLK4OE cells to WEHI-539 normalized to their respective untreated conditions, obtained from MTT viability assays. Mean and SEM of 3 independent experiments each obtained from averaging 3 technical replicates. (B) Schematic of the induction of apoptosis by WEHI-539. When BAX and BAK channel formation are inhibited by BCL-XL, Cytochrome C is present in the mitochondria intermembrane space (left). WEHI-539 inhibits BCL-XL which relieves the inhibition of channel formation by BAX and BAK, leading to Cytochrome C release, Apoptosome activation, and cleavage of Caspase 3 (right). (C) Bar graphs showing the average and SEM of the percentage of cells in specified Annexin V-APC/PI gates analyzed by flow cytometry. Four replicates obtained from 2 independent experiments with a minimum of 15,000 cells analyzed per condition and replicate. Statistical test: comparison of the percentage of Annexin V positive cells, using ANOVA with Sidak’s multiple comparison test. Representative cytometry profiles can be found in the S1 Appendix. (D) Representative images of cells labeled with DAPI (gray) and antibodies against Cytochrome C (cyan), CEP192 (magenta), and Pericentrin (magenta). White arrows indicate dead cell debris, pink arrows indicate cells that have released Cytochrome C in to the cytoplasm, M indicates mitotic cells. Representative insets are shown in panel (G). (E) Bar graphs showing the average and SEM of the percentages of indicated cell populations. Two independent experiments, statistical test: Fisher’s exact test on the number of cells releasing Cytochrome C. Numbers on the top of each graph represent the number of cells analyzed per condition. (F) Bar graphs showing the average and SEM of the percentage of cells with the indicated number of centrosomes (determined by the co-localization of CEP192 and Pericentrin). Two independent experiments, statistical test: comparison of the percentage of cells with more than 2 centrosomes, using ANOVA with Sidak’s multiple comparison test. Numbers on the top of each graph represent the number of cells analyzed per condition. (G) Insets from panel (D) with inverted grayscale insets zooming on the centrosomes showing CEP192 (tick magenta border) and Pericentrin (light magenta border). For PLK4OE, WEHI-539, Q-VD-OPH a “Fire-lut” inset is shown. (H) Bar graphs showing the average and SEM of the percentage of cells in specified Annexin V-APC/PI gates analyzed by flow cytometry. Four replicates obtained from 2 independent experiments with a minimum of 10,000 cells analyzed per condition and replicate. Statistical test: comparison of the percentage of Annexin V positive cells, using ANOVA with Sidak’s multiple comparison test. Representative cytometry profiles can be found in the S1 Appendix. (I) Scatter dot plots showing the ratio between the percentages of Annexin V positive cells observed in presence and absence of WEHI-539 300 nM. Average and SEM of 3 replicates from 2 independent experiments. Statistical test: ANOVA with Dunnett’s multiple comparison test. Data for Fig 4 can be found in S4 Data.