Article Figures & Data
Figures
- Figure 1.
Identification of APC/C subunits in a genome-wide siRNA screen for CIN survival. A, Chromosome segregation error rates determined by FISH in postmitotic daughter cells. RPE1 cells were treated with reversine for 2 hours, then mitotic cells were collected by shake-off and allowed to reattach in the presence of reversine, before fixation. Graph represents the average rate measured for chromosomes 6, 7, 8, and 10 (bars, average ± 95% CI). B, HCT116 wild-type and p53−/− isogenic lines were imaged for 72 hours by live-cell imaging. Cell density for each drug concentration was normalized to DMSO for wild-type and p53−/− cell separately. Fold difference in cell density of HCT116 p53−/− relative to wild-type is displayed for each drug concentration. C, Clonogenic assay using isogenic HCT116 cells grown for 10 days in the presence or absence or reversine, as indicated. D, Genome-wide RNAi screen for synthetic viability with MPS1 inhibition. RPE1 cells were synchronized in G0–G1 by contact inhibition, trypsinized, and reverse-transfected at low density in triplicate to allow uniform passage through mitosis. Cells were exposed to 250 nmol/L reversine for 96 hours and fixed. Automated image acquisition and analysis were performed for various parameters, and Z-scores were derived based on median plate normalization for each siRNA pool. E, DAPI staining of fixed cells 48 hours following the indicated siRNA treatments with 250 nmol/L reversine where indicated. F, Chromosome segregation error rates measured by FISH as in A. Cells were transfected with a nontargeting or a CDC16 siRNA pool, and 48 hours later, cells were synchronized with a single thymidine block. Reversine was added 10 hours after thymidine release, prior to mitotic entry. Mitotic cells were collected at 12 hours by shake-off, allowed to reattach on coverslips still in the presence of reversine, and fixed (bars, average ± 95% CI). G, Time-lapse fluorescence microscopy of RPE1 cells expressing H2B-mCherry imaged every 3 minutes, 1 hour following the addition of 350 nmol/L reversine +/− proTAME (pT), as indicated. The duration from NEBD to metaphase and metaphase to anaphase is shown for cells in which all chromosomes congressed to form a metaphase plate. Each row corresponds to a single cell (n > 60 cells each; bars, mean ± 95% CI). H, Sample images of segregation errors scored in G. Maximum intensity projections are shown for timeframes immediately preceding and following anaphase onset. Shown are examples of a correct division (no error), an example of a lagging chromosome following proper congression at metaphase (middle), and an example where anaphase occurred before all chromosomes congressed to the metaphase plate (congression defect, hence NEBD–metaphase could not be determined). Scale bar, 10 μm.
- Figure 2.
APC/C mutations in human cancer. A, Lollipop plot representing the distribution, frequency, and type of mutations reported for CDC27 across 30 cancer types (see Supplementary Table S3). Green dots indicate missense mutations and red dots truncating mutations (nonsense, frameshift, and splice site). All lollipop plots were retrieved and adapted from cBioPortal (77, 78). The locations of the CRISPR/Cas9 guide RNA (gRNA) targeting sites used in the following experiments are indicated. B, Schematics describing the reversine adaptation assay. RPE1 populations were generated by infecting cells at low titre with lentiCRISPR vectors expressing either one of three different gRNAs targeting CDC27 (to control for possible off-target cleavage). Cells were grown in 250 nmol/L reversine by passaging cells when the dish became confluent or every 2 weeks, whichever came first. Approximate doubling times were derived from counting cells at every passage. For vector cells, additional plates were maintained in parallel to isolate rare resistant colonies (Supplementary Fig. S4). C, Doubling times of the RPE1/lentiCRISPR populations maintained in reversine as determined at days 14 and 28. D, Stills from time-lapse movies of the RPE1/lentiCRISPR/CDC27-2 population after 4 weeks in 250 nmol/L reversine. Note the actively dividing cells in reversine. Upon 24 hours following reversine washout, many round mitotic cells are visible, but these cells are delayed only in mitosis rather than arrested, as shown below in E (scale bar, 100 μm). E, Mitotic duration from NEBD–anaphase in three independent RPE1/lentiCRISPR CDC27 populations exposed to 250 nmol/L reversine for 4 weeks. Cells were imaged in parallel while still in 250 nmol/L reversine or 24 hours following reversine washout (bars, mean ± 95% CI; ***, P < 0.0001). F, Tetrad dissection of S. pombe strains and the genotype of each haploid spore bearing an extra allele (in red, mutated where indicated; see Supplementary Fig. S5A for mating scheme). Both the truncating mutation homologous to E736* (S. pombe E600*) and the missense mutant G506E (S. pombe G370E) failed to rescue NUC2 deletion lethality, indicating that they are deleterious mutations. WT, wild-type. G, Precise genome editing was performed in RPE1 cells to delete a single base pair at codon L454 (CTA) in one CDC27 allele, thereby creating a frameshift and stop codon 9 amino acids downstream. Mitotic duration from NEBD–anaphase was determined for the heterozygous L454Hfs*9 clone in parallel with the parental population and three nonedited clones isolated in parallel, all of which were never exposed to reversine. H, NEBD–anaphase duration of HCT116 clones isolated based on their resistance to the MPS1 inhibitor Cpd-5. Mitotic timings were either determined while cells were still maintained in 30 nmol/L Cpd-5 or following Cpd-5 washout (bars, mean ± 95% CI; *, P < 0.05; **, P < 0.01; ***, P < 0.0001; ns, nonsignificant).
- Figure 3.
APC/C dysfunction reduces CIN caused by merotely and lagging chromosomes in cancer cell lines. A, Genome-doubling status of tumors bearing a mutation in one of the APC/C subunits. Derived from 2,694 tumors from 9 cancer types for which mutational and copy-number data were available to determine genome-doubling status (**, P = 0.0014, Fisher exact test). B and C, Genome doubling (tetraploidization) was induced in RPE1/p53−/− cells by a 12-hour treatment with 4 μmol/L DCB to block cytokinesis. DCB was washed off, and cells were allowed to recover for 2 hours, after which 3 μmol/L proTAME or DMSO was added to the media and cells were imaged every 3 minutes to capture the first mitosis following tetraploidization. Mitotic duration from NEBD–anaphase was measured exclusively for binucleated cells, and the fraction of bipolar or multipolar divisions was scored (scale, 50 μm; bars, mean ± 95% CI; **, P = 0.01 for division outcome, Fisher exact test). D and E, Cells were grown on glass coverslips, treated with DCB using the same protocol as described in C, and fixed for immunostaining. Genome-doubled cells were identified on the basis of having four centrosomes (identified using pericentrin staining), and the frequency of anaphase lagging chromosomes was scored (***, P < 0.0001, Fisher exact test).
- Figure 4.
APC/C dysfunction results in CIN buffering and adaptation to extreme CIN. A and B, RPE1/H2B-mCherry cells were arrested in mitosis using 40 nmol/L STLC, collected by shake-off, and released into media containing 3 μmol/L proTAME or DMSO (t = 0h). Stacks were acquired every 3 minutes to determine anaphase onset (A) and the presence of lagging chromosomes (B; error bars, mean ± 95% CI). Sample images from movies are shown, and the arrow indicates a lagging chromosome in a cell without proTAME. Representative images are shown (scale bar, 10 μm). C, Degradation kinetics of cyclin A2-Venus fluorescence quantified from unsynchronized single cells as they progress through mitosis. 1.5 μmol/L proTAME or DMSO was added 2 hours before imaging. Total cell fluorescence was quantified and normalized to the level at NEBD. Curves end at anaphase onset (n = 24 cells per condition; error bars indicate SD; ***, P < 0.0001 Student t test). Time to reach 50% maximum intensity was used for statistical analysis.
- Figure 5.
Acquired CIN attenuation through APC/C dysfunction in aneuploidy-tolerant p53-null cells. A, Fraction of TP53-mutated tumors among samples that are either wild-type for all APC/C subunits or in which at least one APC/C subunit is mutated. Derived from 2,694 tumors from 9 cancer types (bladder, breast, colon, and head and neck cancers, glioblastoma multiforme, kidney renal clear cell carcinoma, lung adenocarcinoma, lung squamous cell carcinoma, and melanoma; **, P = 0.005, Fisher exact test). See Methods for the number of samples per tumor types. B, NEBD–anaphase duration of RPE1/p53−/− cells after acute treatment with 250 nmol/L reversine (added 2 hours before imaging). In parallel, RPE1/p53−/− grown for 8 weeks in reversine (adapted) were imaged while still in 250 nmol/L reversine and 1 hour after reversine washout, as indicated. The last column represents RPE1/p53−/− grown for 4 weeks in reversine, followed by 5 weeks without reversine (bars, average ± 95% CI; ***, P < 0.0001). C, Cyclin B degradation kinetics determined by western blot from cells that were maintained in mitosis for the indicated duration with 50 nmol/L nocodazole (a microtubule poison). RPE1/p53−/− cells were maintained in nocodazole ± 3 μmol/L proTAME, whereas reversine-adapted RPE1/p53−/− cells (growing over 4 weeks in the absence of reversine) were maintained in nocodazole only. ImageQuant was used to normalize cyclin B levels to GAPDH, and % cyclin B is expressed relative to t = 0h for each cell line. D, The fraction of anaphase cells with lagging chromosomes was determined for RPE1/p53−/− cells acutely treated with 250 nmol/L reversine, and RPE1/p53−/− reversine-adapted cells while maintained in reversine or grown without reversine 24 hours before fixation (***, P < 0.0001, Fisher exact test). E, H2B-mCherry was introduced in RPE1/p53−/− cells and RPE1/p53−/− reversine-adapted that had been cultured at least 5 weeks without reversine. Merotelic attachments were induced using the STLC arrest/washout protocol, and the frequency of segregation errors was determined by time-lapse fluorescence microscopy (from three experiments; ***, P < 0.0001, Fisher exact test).
- Figure 6.
APC/C subunit mutational status affects CIN in cancer cells. A and B, H2B-mCherry was introduced in H2030, U251, and SW480 cell lines and NEBD–anaphase duration (A) as well as the frequency of anaphase segregation errors (B) with and without proTAME was determined by time-lapse fluorescence microscopy. Stacks were acquired every 3 minutes, and an example of segregation error is shown for each cell line (scale bar, 10 μm; in A, bars, mean ± 95% CI of a representative experiment; in B, P values from Fisher exact test). C, Experimental procedure used to generate CRISPR/Cas9 edited H2030 and HT29 cells used for plots D to I. In each case, a single clone was infected (lentiCRISPR/CDC27 for H2030) or transfected (espCas9(1.1)/CDC23 + ssDNA donor) in a well of a 96-well plate before the colony reached confluency. Following transfection or transduction, the colony was dispersed by limiting dilution into 96-well plates. Clones were then screened for heterozygous disruption of CDC27 in H2030 cells or correction of the heterozygous E245 nonsense mutation in HT29 cells. Nonedited clones identified during screening were used as controls. Phenotypic analysis of all newly derived cell lines was performed following minimal clonal expansion to limit phenotypic diversity that may be acquired due to ongoing CIN. D, Lollipop plot of CDC27 showing only truncating mutations reported in TCGA and the location of the guide RNA used to disrupt CDC27 in H2030 cells. The clone isolated contained a heterozygous 35-bp deletion creating the truncation I442Sfs*15. E, NEBD–anaphase duration was determined for the H2030 clones using phase–contrast time-lapse microscopy (3 minutes/frame; bars, average ± 95% CI). F, The frequency of anaphase lagging chromosomes was determined on fixed cells by indirect IF microscopy (*, P < 0.05; Fisher exact test). G, Lollipop plot of CDC23 showing only truncating mutations reported in TCGA and the HT29 nonsense mutation from HT29 cells. H, NEBD–anaphase duration was determined for HT29 clones using phase–contrast time-lapse microscopy (3 minutes/frame; bars, average ± 95% CI). I, The frequency of anaphase lagging chromosomes was determined on fixed cells by indirect IF microscopy (*, P < 0.05; Fisher exact test).
Additional Files
Supplementary Data
- Supplementary Figure Legends, Table Legends - Supplementary Legends
- Supplementary Figures S1 - S10 - Supplementary Figure S1: Using reversine to cause chromosome segregation errors and p53-dependent aneuploidy tolerance. Supplementary Figure S2: Validation of APC/C subunits using siRNA deconvolution experiments in RPE1 and HCT116 cells, and how subunit knock-down rescues proliferation when the spindle assembly checkpoint is impaired. Supplementary Figure S3: APC/C subunit knock-down reduces segregation errors caused by SAC defects. Supplementary Figure S4: CRISPR-mediated disruption of TP53 and CDC27, and resistance to Mps1 inhibitors. Supplementary Figure S5: Functional testing of CDC27 mutations in Schizosaccharomyces pombe. Supplementary Figure S6: Cell division, microtubule attachment errors and tetraploidy. Supplementary Figure S7: CRISPR-mediated disruption of TP53 in RPE1 cells. Supplementary Figure S8: APC/C partial inhibition does not rescue structural anaphase defects caused by pre-mitotic replicative stress. Supplementary Figure S9: Spindle microtubule stability using photoactivation. Supplementary Figure S10: Adaptation of RPE1/p53ko populations to Mps1 inhibitors is accompanied by a mitotic delay.
- Supplementary Table S1 - Supplementary Table S1: Chromosome segregation error rates per chromosome used to determine average missegregation rates in our studies.
- Supplementary Table S2 - Supplementary Table S2: Gene ontology enrichment analysis from whole-genome RNAi screen identifies APC/C subunits.
- Supplementary Table S3 - Supplementary Table S3: 30 TCGA studies used in this paper.
- Supplementary Table S4 - Supplementary Table S4: APC/C mutation frequencies in TCGA.
- Supplementary Table S5 - Supplementary Table S5: Structural insight into the impact of cancer-associated APC/C mutations.
Volume 7, Issue 2
