Mitotic clustering of pulverized chromosomes from micronuclei – Nature

Cell lines and reagents

All of the cell lines were maintained at 37 °C under 5% CO₂ and atmospheric oxygen. DLD-1, HeLa, PC3 (a gift from S. Wu), HEK293T and 293GP cells were cultured in Dulbecco’s modified Eagle medium (DMEM, Gibco) supplemented with 10% tetracycline-free fetal bovine serum (Omega Scientific) and 100 U ml⁻¹ penicillin–streptomycin. RPE-1 and RPE-1 H2AX^–/– cells (a gift from S. Jackson) were cultured in DMEM/F-12 (Gibco) supplemented with 10% fetal bovine serum and 100 U ml⁻¹ penicillin–streptomycin. RPTECs (a gift from D. Marciano) expressing a short hairpin RNA against TP53 were cultured in renal epithelial cell growth basal medium (Lonza) supplemented with 0.5% tetracycline-free fetal bovine serum (Omega Scientific), 100 U ml⁻¹ penicillin–streptomycin, 10 ng ml⁻¹ human recombinant epidermal growth factor, 10 μg ml⁻¹ human transferrin, 1 μg ml⁻¹ hydrocortisone, 10 μM adrenaline, 50 ng ml⁻¹ triiodo-l-thyronine, 5 μg ml⁻¹ insulin, 30 μg ml⁻¹ gentamicin and 15 ng ml⁻¹ amphotericin B. Cell lines were authenticated by karyotyping and were routinely confirmed to be free of mycoplasma contamination using the Universal Mycoplasma Detection Kit (ATCC).

DOX and IAA (Millipore-Sigma) were dissolved in cell-culture-grade water and used at 1 μg ml⁻¹ and 500 μM, respectively, in DLD-1 cells. Geneticin (G418 sulfate) and zeocin (InvivoGen) were used at selection concentrations of 300 and 50 μg ml⁻¹, respectively. For cell-cycle-arrest experiments, 100 ng ml⁻¹ nocodazole (Millipore-Sigma), 100 ng ml⁻¹ Colcemid (KaryoMAX, Thermo Fisher Scientific) or 10 μM MG132 (a gift from J. Seemann) was used for mitotic arrest, and 10 μM of the CDK1 inhibitor RO-3306 (Millipore-Sigma) was used for G2 arrest, all of which were dissolved in dimethylsulfoxide. A total of 50 nM CENP-E inhibitor (GSK-923295, Cayman Chemical) and 480 nM MPS1 inhibitor (NMS-P715, Cayman Chemical) was used to induce chromosome segregation errors and micronuclei formation in HeLa and RPE-1 cells. dTAGv-1 (500 nM; a gift from B. Nabet) was used to induce degradation of FKBP12(F36V) fusion proteins. For ionizing-radiation experiments, cells were irradiated with γ-ray (2 Gy) generated by a Mark I ¹³⁷Cs irradiator (JL Shepherd) and fixed 1 h after irradiation for immunofluorescence analysis. Small interfering RNA (siRNA) transfections were conducted with Lipofectamine RNAiMAX reagent (Thermo Fisher Scientific). siRNAs were synthesized (Thermo Fisher Scientific) and used at a final concentration of 20 nM. A list of all of the siRNA sequences used in this study is provided in Supplementary Table 2.

Cell line engineering

To generate CIP2A-KO cells, target sequences for guide RNAs were designed using CRISPick (Broad Institute). Oligonucleotides encoding guide RNAs targeting exon 1 of CIP2A (sg3 and sg4; Supplementary Table 2) were cloned into the BsmBI restriction site of the Lenti-Cas9-gRNA-TagBFP2 vector (Addgene, 124774) and packaged in HEK293T cells by co-transfection with pMD2.G (Addgene 12259) and psPAX2 (Addgene, 12260) using X-tremeGENE 9 (Millipore-Sigma). Viral supernatants after 48 h or 72 h transfection were filtered (0.45 μm), and cells were infected in the presence of 5 μg ml⁻¹ polybrene (Millipore-Sigma) for around 24 h. Fluorescent cells were isolated using fluorescence-activated cell sorting (FACS) into 96-well plates (BD FACSAria II). For RPTECs (CIP2A sg3 and sg4) and HeLa (CIP2A sg1 and sg2) cells, KO populations were established by pooling together virus-infected cells. To establish KO clones, DLD-1 cells were sorted, expanded and verified by both Sanger sequencing and immunoblotting.

To generate DLD-1 KO cells by ribonucleoprotein (RNP)-mediated CRISPR genome editing, two sgRNAs per gene were synthesized (Synthego) and co-transfected with TrueCut Cas9 protein v2 (Invitrogen) using the Lipofectamine CRISPRMAX Cas9 transfection reagent (Invitrogen). A list of all of the sgRNA sequences used in this study is provided in Supplementary Table 2. After transfection, cells were plated by limiting dilution into 96-well plates. Single-cell-derived clones were expanded, screened by PCR for targeted deletions and confirmed to harbour frameshift deletion mutations by Sanger sequencing. A list of all PCR primers (Millipore-Sigma) is provided in Supplementary Table 2.

For complementation experiments, CIP2A cDNA (a gift from Q. Zhang) and HaloTag (Addgene, 112852) were cloned into pBABE-zeo (Addgene, 1766) and packaged in 293GP cells by co-transfection with pVSV-G using X-tremeGENE 9. CIP2A-KO cells generated by RNP-mediated gene editing were infected with retroviruses encoding full-length CIP2A, delta NES mutant (lacking amino acids 561–625) or CIP2A–FKBP12(F56V) fused to an N terminus HaloTag for 24 h and selected with zeocin for 10 days. For the expression of other exogenous genes, the H2B-mCherry (a gift from H. Yu) and cGAS-GFP constructs (a gift from Z. Chen) were used to generate viruses for transduction of DLD-1 cells, as described above.

Chromosome labelling using dCas9–SunTag

To label the Y chromosome in live cells, the SunTag labelling system was adopted and modified as described below. DYZ1 repeats (3,584 bp, sequence information provided by H. Skaletsky) were analysed by CRISPick (Broad Institute) and five sgRNA sequences were selected for targeting DYZ1 repeats. scFv-GCN4-sfGFP-GB1-NLS from SunTag plasmid (Addgene, 60906) was cloned into a lentiGuide-puro vector (Addgene, 52963). Lentiviral supernatants, which were packaged in HEK293T cells by co-transfection with pMD2.G and psPAX2 with either lentiGuide-scFv-GCN4-sfGFP-GB1-NLS or pHRdSV40-dCas9-10xGCN4_v4-P2A-BFP (SunTag plasmid, Addgene, 60903), and retroviral supernatants, which were packaged in 293GP cells by co-transfection of pBABE-H2B-mCherry with pVSV-G after 48 h or 72 h transfection, were filtered (0.45 μm) and DLD-1 cells were infected in the presence of 5 μg ml⁻¹ polybrene (Millipore-Sigma) for around 24 h. Fluorescent cells were isolated by FACS (BD FACSAria II) and plated by limiting dilution into 96-well plates. Single-cell-derived clones were expanded and screened for expected SunTag signals.

Live-cell imaging

DLD-1 cells expressing dCas9–SunTag and H2B–mCherry were plated into Nunc Lab-Tek chambered cover glasses. Images were acquired on the DeltaVision Ultra microscope (Cytiva) in a humidity- and temperature-controlled (37 °C) environment supplied with 5% CO₂ at 5 min intervals for 16 h using a ×60 objective with 11 × 0.5 μm z-sections under low power exposure. For CIP2A–HaloTag imaging, cells were labelled with 200 nM JF646 ligand (Promega) for 15 min and washed with fresh medium before imaging. Images were deconvolved and maximum-intensity quick projections were generated using softWoRx (v.7.2.1, Cytiva), and videos were analysed using Fiji (v.2.1.0/1.53c).

For long-term live-cell imaging, DLD-1 cells expressing H2B-mCherry were transfected with siRNAs and seeded in 96-well glass-bottom plates (Cellvis, P96-1.5H-N). The next day, cells were treated with DOX/IAA for 72 h. One day before image acquisition, cells were retransfected with siRNAs to ensure depletion of the target protein throughout the duration of the experiment. Images were acquired on the ImageXpress Confocal HT.ai High-Content Imaging System (Molecular Devices) in a humidity- and temperature-controlled (37 °C) environment in CO₂-independent medium at 15 min intervals for 48 h using a ×40 objective with 7 × 1.5 μm z-sections under low-power exposure. Maximum-intensity projections were generated using MetaXpress (Molecular Devices), and videos were analysed using Fiji (v.2.1.0/1.53c).

Immunofluorescence analysis

DLD-1 cells were plated onto CultureWell gaskets (Grace Bio-Labs) and assembled glass slides were fixed with 4% formaldehyde for 10 min. For dispersion analysis, cells were arrested in mitosis for 4 h using Colcemid and collected by shake-off. Cell suspensions were concentrated to 1 × 10⁶ cells per ml in PBS and centrifuged onto glass slides using a Cytospin 4 cytocentrifuge (Thermo Fisher Scientific). Fixed cells were permeabilized with 0.3% Triton X-100 in PBS for 5 min, incubated with Triton Block (0.2 M glycine, 2.5% fetal bovine serum, 0.1% Triton X-100, PBS) and then incubated with primary antibodies. The following primary antibodies were used at the indicated dilutions in Triton Block: 1:500 anti-CIP2A (sc-80659, Santa Cruz), 1:500 anti-TOPBP1 (sc-271043, Santa Cruz), 1:300 anti-TOPBP1 (ABE1463, Millipore), 1:1,000 anti-phosphorylated H2AX (Ser139) (05-636, Millipore), 1:1,000 anti-phosphorylated H2AX (Ser139) (2577, Cell Signaling), 1:1,000 anti-53BP1 (NB100-304, Novus) antibodies. Cells were washed with 0.1% Triton X-100 in PBS, incubated with 1:1,000 dilutions of Alexa Fluor-conjugated donkey anti-rabbit or donkey anti-mouse secondary antibodies (Invitrogen) for 1 h at room temperature, and washed with 0.1% Triton X-100 in PBS. Immunostained cells were fixed with Carnoy’s fixative for 15 min and rinsed with 80% ethanol. Air-dried cells were then used for DNA FISH, as described below.

For micronuclei analysis, DLD-1, HeLa and RPE-1 cells were grown on glass coverslips and fixed with PTEMF (0.2% Triton X-100, 0.02 M PIPES pH 6.8, 0.01 M EGTA, 1 mM MgCl₂ and 4% formaldehyde) for 10 min, followed by two washes in 1× PBS. The samples were blocked with 3% bovine serum albumin diluted in PBS. Cells were incubated for 1 h at room temperature with the following primary antibodies diluted in 3% BSA: 1:500 anti-CIP2A (sc-80659, Santa Cruz), 1:1,000 anti-CIP2A (14805, Cell Signaling), 1:500 anti-TOPBP1 (sc-271043, Santa Cruz), 1:500 anti-TOPBP1 (ABE1463, Millipore), 1:1,000 anti-phosphorylated histone H2AX (Ser139) (2577, Cell Signaling), 1:1,000 anti-acetyl-histone H3 (Lys9) (9649, Cell Signaling) and 1:1,000 anti-CGAS (15102, Cell Signaling). After three 5 min washes, Alexa-Fluor-conjugated donkey anti-rabbit or donkey anti-mouse secondary antibodies (Invitrogen) were diluted 1:1,000 in 3% BSA and applied to cells for 1 h at room temperature, followed by two 5 min washes with 1× PBS. DNA was counterstained with DAPI and cells were mounted in ProLong Gold antifade mounting solution.

For cytosolic dsDNA staining, cells were fixed with 4% formaldehyde for 10 min and then treated with 0.02% saponin in PBS for 5 min. Semi-permeabilized cells were incubated with blocking solution (2.5% fetal bovine serum in PBS) followed by incubation with anti-dsDNA antibodies (1:250 in blocking solution, sc-58749, Santa Cruz) at 4 °C overnight. After washing with PBS, cells were incubated with 1:1,000 dilutions of an Alexa-Fluor-conjugated donkey anti-mouse secondary antibody (Invitrogen) in blocking solution for 1 h and washed with PBS. Cells were then fully permeabilized with 0.3% Trion X-100 in PBS for 5 min and washed with PBS. Permeabilized cells were incubated with 5 U ml⁻¹ of fluorescent phalloidin (Biotium) in PBS for 20 min and washed with PBS.

Metaphase spread preparation

Cells were treated with 100 ng ml⁻¹ Colcemid (KaryoMAX, Thermo Fisher Scientific) for 4–5 h before collection by trypsinization and centrifugation. Cell pellets were resuspended in 500 μl PBS followed by adding 5 ml of 75 mM KCl solution dropwise while gently vortexing. Cells were incubated for 6 min in 37 °C water bath and fixed using freshly prepared, ice-cold Carnoy’s fixative (3:1 methanol:acetic acid), followed by centrifugation and resuspension in Carnoy’s. Cells were subsequently dropped onto slides and air dried for further processing.

DNA FISH

DNA FISH probes (MetaSystems) were applied to metaphase spreads and sealed with a coverslip using rubber cement. Slides were co-denatured on a heat block at 75 °C for 2 min and then hybridized at 37 °C in a humidified chamber overnight. The next day, the coverslips were removed, and the slides were washed with 0.4× SSC at 72 °C for 2 min and rinsed with 2× SSC with 0.05% Tween-20 at room temperature for 30 s. After washing, the slides were counterstained with DAPI, air dried and mounted in ProLong Gold antifade mounting solution.

Fixed-cell microscopy

Immunofluorescence images were captured on a DeltaVision Ultra (Cytiva) microscope system equipped with a 4.2 Mpx sCMOS detector. Interphase nuclei and micronuclei images were acquired with a ×100 objective (UPlanSApo, 1.4 NA) and 1 × 0.2 μm z-section. Quantitative fluorescence image analyses were performed using Fiji (v.2.1.0/1.53c). IF–FISH images were acquired with a ×60 objective (PlanApo N 1.42 oil) and 15 × 0.2 μm z-sections. Deconvolved maximum intensity projections were generated using softWoRx (v.7.2.1, Cytiva).

Metaphase FISH images were acquired on the Metafer Scanning and Imaging Platform microscope (Metafer 4, v.3.13.6, MetaSystems). The slides were first scanned for metaphases using M-search with a ×10 objective (ZEISS Plan-Apochromat 10x/0.45), and metaphases were automatically imaged using Auto-cap with a ×63 objective (ZEISS Plan-Apochromat 63x/1.40 oil). Images were analysed using the Isis Fluorescence Imaging Platform (MetaSystems) and Fiji (v.2.1.0/1.53c).

Chromosome distribution between daughter cells

DLD-1 cells were seeded in four-well chamber slides and treated with or without DOX/IAA for 48 h. Cells were then arrested in G2 with 10 µM CDK1 inhibitor RO-3306 (Millipore-Sigma) for 16 h, washed with PBS three times and released into mitosis in fresh medium. After 90 min, cells were fixed with 4% formaldehyde followed by IF–FISH, as described above, and hybridized to X- and Y-chromosome paint probes (MetaSystems). For analysis of chromosome inheritance between daughter cells, pairs of newly formed daughter cells were imaged on the DeltaVision Ultra (Cytiva) microscope system. Images were split into separate channels for quantification using the ImageJ plugin Segmentation (Robust Automatic Threshold Selection) to create a mask for the FISH signals. Particles of the mask were analysed to generate a list of regions of interest for intensity measurements. FISH signal intensities were then measured in each pair of daughter cells for both the X and Y chromosomes. The distribution of FISH signal was calculated by the ratio of the daughter cell with the lower signal compared to the daughter cell with the higher signal.

Mitosis-specific depletion of FKBP fusion proteins

CIP2A-KO DLD-1 cells complemented with CIP2A-FKBP12(F36V) were seeded in T75 flasks and treated with DOX/IAA for 72 h. Cells were than arrested in mitosis with 100 ng ml⁻¹ nocodazole for 6 h and mitotic cells were collected by mitotic shake-off. Mitotic cells were then reseeded in 24-well plates and treated with or without 500 nM dTAGv-1 for 4 h in the presence of 100 ng ml⁻¹ nocodazole. Cells were centrifuged onto glass slides using the Cytospin 4 cytocentrifuge and processed for IF–FISH.

Chromosome fragment dispersion

Metaphase spreads were prepared as described and hybridized to Y-chromosome paint probes (MetaSystems). Metaphases with fragmented Y chromosomes were identified and split into separate channels. Fragment dispersion was analysed using the ImageJ plugin HullAndCircle to measure the convex hull of the Y-chromosome fragments relative to all DAPI-stained chromosomes. Dispersion indices were calculated by dividing the area of Y-chromosome fragments by the overall DAPI area followed by minimum–maximum normalization of all data points within each sample.

Chromosome-specific micronuclei in RPE-1 cells and RPTECs

RPE-1 cells were seeded in T175 flasks and transfected with siRNAs the next day. One day after transfection, cells were arrested in mitosis with 100 ng ml⁻¹ nocodazole for 8 h. Mitotic cells were collected by mitotic-shake off, washed three times with culture medium and reseeded onto coverslips and T75 flasks. Cells growing on coverslips were fixed at 20 h after releasing from mitosis for analysis of chromosome 1 micronuclei by FISH. For dispersion analysis, after 20 h release from nocodazole, cells growing on T75 flasks were arrested in mitosis with Colcemid for 4 h. Mitotic cells were collected by mitotic-shake off, centrifuged onto glass slides using the Cytospin 4 cytocentrifuge and processed for FISH.

RPTECs were seeded in T75 flasks and transfected with Cas9 (TrueCut Cas9 protein v2, Thermo Fisher Scientific) in a complex with an sgRNA targeting chromosome 3p using Lipofectamine CRISPRMAX Cas9 transfection reagent (Thermo Fisher Scientific). During transfection, 3 μM of the DNA-PK inhibitor AZD7648 (MedChemExpress) was added for 24 h and washed out. Three days later, cells were arrested in mitosis with Colcemid for 6 h. Mitotic cells were collected by mitotic-shake off, centrifuged onto glass slides using the Cytospin 4 cytocentrifuge and processed for FISH.

Cell cycle profiling

Cells were trypsinized, washed with PBS and fixed with 70% ethanol in PBS at −20 °C for 2 h. Fixed cells were washed with PBS twice and incubated with staining solution (0.1 mg ml⁻¹ RNase A, 0.1% Triton X-100, 10 μg ml⁻¹ propidium iodide). Cells were analysed using a FACSCalibur (BD Biosciences) flow cytometer, and cell cycle profiles were generated using FlowJo (v.10.8.2, BD Biosciences) software.

Immunoblotting

Whole-cell extracts were collected in Laemmli SDS sample buffer and boiled for 5 min. The samples were resolved by SDS polyacrylamide gel electrophoresis, transferred to polyvinylidene fluoride membranes and blocked with 5% milk diluted in PBST (PBS, 0.1% Tween-20). The following primary antibodies were diluted in PBST and used: 1:1,000 anti-CIP2A (sc-80659, Santa Cruz), 1:5,000 anti-α-tubulin (3873, Cell Signaling), 1:1,000 anti-TOPBP1 (sc-271043, Santa Cruz), 1:1,000 anti-phosphorylated histone H3 (Ser10) (06-570, Millipore) and 1:5,000 anti-MDC1 (ab11171, Abcam). The blots were incubated with 1:4,000 dilutions of horseradish peroxidase-conjugated goat anti-rabbit or donkey anti-mouse secondary antibodies (Invitrogen), incubated with SuperSignal West Pico Plus chemiluminescent substrate (Thermo Fisher Scientific) and processed using the ChemiDoc MP imaging system (Bio-Rad).

Cell proliferation assays

To measure cell proliferation, 1 × 10⁵ cells were seeded onto p60 mm dishes in triplicate, treated with or without DOX/IAA the next day and counted at three-day intervals. For clonogenic survival assays, 1,000 cells were plated in p60 mm dishes in triplicate for 15 days. Colonies were fixed in ethanol, stained with 0.5% crystal violet/70% ethanol solution and manually counted.

Analysis of Y-chromosome rearrangements

Two-colour DNA FISH probes (MetaSystems) were applied to metaphase spreads and captured on the Metafer Scanning and Imaging Platform (Metafer 4, v.3.13.6, MetaSystems), as described above. Distinct types of structural rearrangements were manually inspected using previously described criteria^3,51. To determine the size of Y chromosomes with complex rearrangements, images were split into separate channels using Fiji (v.2.1.0/1.53c) followed by creation of a mask by segmentation of the DAPI channel using Threshold adjust. Particles of the mask were analysed to generate regions of interest for area measurement. The number of DAPI-occupied pixels of rearranged Y chromosomes were measured and normalized to the X chromosome from the same metaphase spread.

RNA-seq analysis

HeLa cells were transduced with a control sgRNA (sgNTC), sgCIP2A-1 or sgCIP2A-2 and selected with puromycin. Total RNA from three independent biological replicates was collected using the RNeasy Total RNA kit (Qiagen), and libraries were sequenced on the Illumina NovaSeq 6000 platform (Novogene). Sequencing reads were aligned to the transcriptome using STAR (v.2.7.4a)⁵². Gene expression counts were generated using HTSeq (v.0.6.1p1)⁵³ and normalized to transcripts per kilobase million. GENCODE (v.22) was used as the gene annotation reference⁵⁴. Gene Set Enrichment Analysis (GSEA, v.4.3.2)⁵⁵ was performed using the weighted enrichment statistic on normalized gene counts computed using DESeq2⁵⁶. We used Hallmark gene sets containing between 15 and 500 genes from the Human Molecular Signatures Database (MSigDB)⁵⁷.

Whole-genome sequencing analyses

To detect copy-number-balanced chromothripsis events, we applied ShatterSeek⁸ (v.1.1; https://github.com/parklab/ShatterSeek) to 2,575 tumour–normal pairs from PCAWG that passed quality-control criteria. We considered all chromosomes with a cluster of at least five structural variants (SVs). We considered all clusters irrespective of the number of copy-number oscillations in the cluster. To call a cluster of SVs a copy-number-balanced chromothripsis event, we required: (1) at least five intrachromosomal SVs; (2) no translocation mapping to the genomic region encompassed by the cluster of SVs; we included this filter to distinguish balanced chromothripsis from chromoplexy events, which are characterized by chains of interchromosomal SVs with limited genomic DNA loss and could therefore be misclassified as balanced chromothripsis if this filter was not applied; (3) no overlap with chromoplexy calls generated for these tumours using ChainFinder⁵⁸ as previously reported⁸; and (4) that less than 1% of the genomic region encompassed by the cluster of SVs shows a copy number of less than the modal copy number of the chromosome. We applied this filter to ensure that balanced chromothripsis calls do not contain canonical chromothripsis events. Finally, all cases that passed these filters were examined manually by visualizing genomic rearrangement plots using ReConPlot⁵⁹.

To find gene disruptions within the balanced chromothripsis clusters, we first downloaded gene coordinates from Ensembl⁶⁰ (GRCh37) using biomaRt⁶¹. We next intersected the coordinates of the breakpoints and genes using bedtools⁶². We considered a gene to be disrupted if a breakpoint mapped within the region defined by the start and end coordinates of the gene ±5 kilobases. We determined putative cancer-driver genes using the pan-cancer driver catalogue from the Hartwig Medical Foundation cancer whole-genome sequencing analysis pipeline (https://github.com/hartwigmedical/hmftools/blob/master/purple/DriverCatalog.md).

Statistics and reproducibility

Statistical tests were performed as described in the figure legends using GraphPad Prism (v.9.5.0). Sample sizes, statistical analyses and significance values are reported in the figure legends, denoted in the figure panel or described in the text. P ≤ 0.05 was considered to be statistically significant. Error bars represent s.e.m. unless otherwise stated. Experiments showing representative images were independently repeated two (Fig. 4g and Extended Data Figs. 1a,e,g and 7a,b), three (Extended Data Fig. 5i) or four (Fig. 4j) times with similar results.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.