A transcription-based mechanism for oncogenic β-catenin-induced lethality in BRCA1/2-deficient cells

Abstract

BRCA1 or BRCA2 germline mutations predispose to breast, ovarian and other cancers. High-throughput sequencing of tumour genomes revealed that oncogene amplification and BRCA1/2 mutations are mutually exclusive in cancer, however the molecular mechanism underlying this incompatibility remains unknown. Here, we report that activation of β-catenin, an oncogene of the WNT signalling pathway, inhibits proliferation of BRCA1/2-deficient cells. RNA-seq analyses revealed β-catenin-induced discrete transcriptome alterations in BRCA2-deficient cells, including suppression of CDKN1A gene encoding the CDK inhibitor p21. This accelerates G1/S transition, triggering illegitimate origin firing and DNA damage. In addition, β-catenin activation accelerates replication fork progression in BRCA2-deficient cells, which is critically dependent on p21 downregulation. Importantly, we find that upregulated p21 expression is essential for the survival of BRCA2-deficient cells and tumours. Thus, our work demonstrates that β-catenin toxicity in cancer cells with compromised BRCA1/2 function is driven by transcriptional alterations that cause aberrant replication and inflict DNA damage.

Fig. 1. GSK3 chemical inhibition triggers β-catenin accumulation, which decreases cell survival in the absence of BRCA1 or BRCA2.

a Diagrammatic representation of GSK3 function in stabilising β-catenin and the effect of GSK3 inhibitor LY2090314. b Human H1299 cells carrying a doxycycline (DOX)-inducible BRCA2 shRNA were grown in the presence (−BRCA2) or absence (+BRCA2) of 2 μg/mL DOX, and treated or not with 250 nM LY2090314. Whole-cell extracts prepared at indicated times after LY2090314 addition were immunoblotted as shown. SMC1 was used as a loading control. Phosphorylation sites are indicated in red. Data are representative of n = 3 independent experiments. c, d Clonogenic survival assays performed in BRCA1/2-proficient or -deficient H1299 (c) and RPE1 (d) human cells treated with LY2090314 at the indicated doses. Whole-cell extracts prepared at the time of cell seeding were immunoblotted as shown. SMC1 was used as a loading control. Error bars represent standard error of the mean (SEM) of n = 3 independent experiments, each performed in triplicate. Statistical significance was determined by an unpaired two-tailed t-test. e H1299 cells carrying a DOX-inducible BRCA2 shRNA were grown in the presence (−BRCA2) or absence (+BRCA2) of DOX and transfected with control or β-catenin siRNAs. Two days later, cells were treated with LY2090314 at the indicated concentrations for 6 days and processed for cell viability assays. Whole-cell extracts prepared at the start of treatment were immunoblotted as shown. SMC1 was used as a loading control. Error bars represent SEM of n = 3 independent experiments, each performed in triplicate. Statistical significance was determined by an unpaired two-tailed t-test. Source data for (b–e) are provided in the Source data file.

Fig. 2. Oncogenic β-catenin deregulates transcription of E2F gene targets and CDKN1A in BRCA2-deficient cells.

a BRCA2-deficient H1299 cells transfected with control or β-catenin siRNAs were treated with DMSO or 250 nM LY2090314 for 24 h before processing for RNA-seq (n = 3 independent experiments). Volcano plots show differentially expressed genes (FDR < 0.05) in the samples indicated. Gene subsets significantly downregulated (log₂(Fold Change) < −0.5) or upregulated (log₂(Fold Change) > 0.5) are marked by dotted lines. b Venn diagram of genes significantly upregulated by LY2090314 (FDR < 0.05 and log₂(Fold change) > 0.5) and downregulated by β-catenin siRNA (FDR < 0.05 and log₂(Fold change) < −0.5) in H1299 BRCA2-proficient (purple) and -deficient (green) cells. β-catenin target genes overlapping in the two samples of Venn diagram are indicated in blue (241 genes). c Gene set enrichment analysis based on functional annotation (Hallmark gene sets) of β-catenin upregulated genes common between BRCA2-proficient and -deficient cells (blue), or unique to BRCA2-deficient cells (green). d H1299 cells expressing a DOX-inducible BRCA2 shRNA were transfected with control or β-catenin siRNAs and grown in the presence (−BRCA2) or absence (+BRCA2) of DOX and LY2090314. Whole-cell extracts were immunoblotted as shown. SMC1 and GAPDH were used as loading controls. p21 expression was quantified relative to GAPDH control. Data are representative of n = 3 independent experiments. e FACS analyses of DNA content of cells treated as in (d), error bars represent SEM of n = 3 independent experiments. Statistical significance was determined by an unpaired two-tailed t-test. Source data for (d, e) are provided in the Source data file.

Fig. 3. β-catenin activation accelerates the G1/S transition and triggers the ATR-dependent checkpoint in BRCA2-deficient cells.

a H1299 cells expressing a DOX-inducible BRCA2 shRNA were grown in the presence (-BRCA2) or absence (+BRCA2) of DOX and LY2090314 and arrested in mitosis with nocodazole treatment. Mitotic cells were collected and released in fresh medium containing 25 μM EdU, in the presence or absence of LY2090314. The % of EdU-positive cells was determined using FACS analyses at the indicated time points after mitotic shake-off. Error bars represent SEM of n = 4 independent experiments. Statistical significance was determined by an unpaired two-tailed t-test. Whole-cell extracts prepared at the time of mitotic shake-off were immunoblotted as shown. SMC1 was used as a loading control. b Whole-cell extracts prepared from cells treated as in (a) and collected at the indicated time points after release from mitotic arrest were immunoblotted as indicated. GAPDH was used as a loading control. Cyclin E and CDC6 expression were quantified relative to GAPDH control. Data are representative of n = 2 independent experiments. c BRCA2-deficient H1299 cells were transfected with control, β-catenin or p21 siRNA and grown in the presence or absence of LY2090314 for 24 h. Mitotic cells were collected and released as in (a). The % of EdU-positive cells was determined using FACS analyses at 14 h after mitotic shake-off. Mean of n = 2 independent experiments. Whole-cell extracts prepared at the time of mitotic shake-off were immunoblotted as shown. d BRCA2-proficient or -deficient H1299 cells were grown in the presence or absence of LY2090314 for 24 h and immunoblotted as shown. Phosphorylation sites are indicated in red. Data are representative of n = 3 independent experiments. e Quantification of frequency of cells with 20 or more 53BP1 foci following treatment as in (d). Error bars represent SEM of n = 3 independent experiments. Statistical significance was determined by an unpaired two-tailed t-test. Source data for (a–e) are provided in the Source data file.

Fig. 4. β-catenin activation triggers aberrant origin firing in BRCA1/2-deficient human cells.

a, b BRCA2-deficient H1299 (a) and BRCA1^−/− RPE1 (b) cells were grown in the presence (red) or absence (blue) of LY2090314 for 24 h and 100 ng/mL nocodazole was added during the final 8 h of treatment. Mitotic cells were collected and released in fresh medium containing 25 μM EdU and 2 mM HU. EdU-labelled DNA was isolated from cells at 20 h after mitotic release and analysed using high-throughput sequencing. RT, replication timing: blue, early; green, mid; yellow, late S-phase. Ge, genes (green, forward direction of transcription; purple, reverse direction of transcription); iG, intergenic regions (grey). β-catenin-induced origins firing within genes (indicated with white stars) are shown by red arrows. c Origin classification based on adjusted σ value ratios in BRCA2-deficient H1299 cells released from mitotic arrest for 16 or 20 h. d Scatter plots of EdU-seq σ-values at 16 and 20 h after BRCA2-deficient H1299 cells release from mitotic arrest, colour-coded as in (c).

Fig. 5. β-catenin-induced fork collapse and origin firing in genic vs inter-genic regions.

a Genome-wide average fork progression in BRCA2-deficient H1299 cells treated with 25 μM EdU 30 min before cell collection. b Genome-wide distribution of β-catenin-induced or cyclin E-induced origins in BRCA2-proficient and -deficient H1299 cells, BRCA1-deficient RPE1 cells and U2OS cells, based on gene annotation. c Size frequency distribution of all protein-encoding genes in H1299 cells (grey) and of protein-encoding genes containing β-catenin-induced origins in BRCA2-deficient H1299 cells (16 and 20 h after mitotic release; red). Dotted lines indicate mean values for each gene set. Statistical significance was determined by a two-tailed Mann–Whitney test. Source data for (b) is provided in the Source data file.

Fig. 6. Oncogenic β-catenin enhances transcription and R-loop formation in BRCA2-deficient cells.

a HeLa cells transfected with control, BRCA2 or β-catenin siRNAs were grown in the presence or absence of 250 nM LY2090314 for 16 h, with 75 μM DRB added during the final 2 h as indicated, and pulsed with 1 mM EU for 1 h. Whole-cell extracts prepared from each sample were immunoblotted as shown. SMC1 was used as a loading control. EU incorporation was quantified as immunofluorescence intensity per nucleus (n = 3 independent experiments; ≥200 nuclei analysed per experiment). Boxes indicate the median value and 25th–75th percentile; whiskers indicate the 10th and 90th percentiles. Statistical significance was determined by a two-tailed Mann–Whitney test. b HeLa cells carrying a DOX-inducible FLAG-tagged RNase H1 were transfected with control, BRCA2 or β-catenin siRNAs. Cells were treated with 50 nM LY2090314 for 16 h followed by quantification of S9.6 immunofluorescence per nucleus (n = 3 independent experiments; ≥100 nuclei were analysed per experiment). Nucleoli were excluded from the analysis of S9.6 signal. Whole-cell extracts prepared from each sample were immunoblotted as indicated. SMC1 was used as a loading control. Boxes indicate the median value and 25th–75th percentile; whiskers indicate the 10th and 90th percentiles. Statistical significance was determined by a two-tailed Mann–Whitney test. c HeLa cells treated as in (b) were incubated with 2 μg/mL DOX to induce FLAG-tagged RNase H1 expression. Quantification of S9.6 immunofluorescence was performed as described in (b). d Clonogenic survival assays were performed in HeLa cells treated as in (c) and then incubated with 0.1 µM LY2090314 for 24 h. Survival is expressed relative to DMSO-treated cells. Error bars represent SEM of n = 3 independent experiments. Statistical significance was determined by an unpaired one-tailed t-test. Whole-cell extracts prepared at the time of cell seeding were immunoblotted as shown. SMC1 was used as a loading control. Source data for (a–d) is provided in the Source data file.

Fig. 7. Oncogenic β-catenin accelerates replication fork progression by downregulating p21 expression.

a H1299 cells expressing a DOX-inducible BRCA2 shRNA were transfected with control or β-catenin siRNAs and grown in the presence (−BRCA2) or absence (+BRCA2) of DOX, or LY2090314. Cells were pulse labelled with CldU followed by IdU for 20 min each. CldU + IdU track length was used to determine fork velocity. Whole-cell extracts prepared at the time of analysis were immunoblotted with SMC1 as a loading control. Fibres quantified from n = 3 independent experiments. b qRT-PCR analyses of H1299 cells grown as in (a), for CDKN1A. mRNA levels expressed relative to the gene encoding GAPDH and untreated BRCA2-proficient cells (2^−ΔΔCT). Error bars represent SEM of n = 5 independent experiments. Statistical significance was determined by an unpaired two-tailed t-test. c DNA fibre analysis as in (a) of cells also transfected with p21 siRNA. Whole-cell extracts were immunoblotted with SMC1 and GAPDH as loading controls. p21 expression was quantified relative to GAPDH control. Fibres quantified from n = 2 independent experiments. d qRT-PCR analyses as described in (b) for H1299 and MDA-MB-231 human cells carrying a DOX-inducible BRCA2 shRNA grown in the presence (BRCA2-deficient) or absence (BRCA2-proficient) of DOX for 28 days. Error bars represent SEM of n = 3 independent experiments. Statistical significance was determined by an unpaired two-tailed t-test. e CDKN1A mRNA expression in TCGA breast tumours (n = 459 for WT; n = 22 for BRCA2-mutated;). Each dot represents a single tumour. Middle line, mean. Statistical significance was determined by a two-tailed Mann–Whitney test. f Quantification of 53BP1 foci in BRCA2-deficient H1299 cells, 7 days after transfection with control or p21 siRNAs (n = 3 independent experiments; ≥200 cells per experiment). g Clonogenic survival assays in BRCA2-deficient H1299 cells transfected with control or p21 siRNAs. Error bars represent SEM of n = 3 independent experiments. Statistical significance was determined by an unpaired two-tailed t-test. For a, c and f boxes indicate the median value and 25th–75th percentile, whiskers indicate the 10th and 90th percentiles, and statistical significance was determined by a two-tailed Mann–Whitney test. Source data for (a–d) and (f, g) is provided in the Source data file.

Fig. 8. Mechanism of action of oncogenic β-catenin in BRCA1/2-proficient and -deficient cells.

Oncogenic β-catenin activation accelerates G1/S transition and fork progression, both mediated by p21 transcriptional downregulation and leading to fork stalling and DSBs accumulation. BRCA1/2-proficient cells can repair oncogene-induced DSBs via homologous recombination. In contrast, in BRCA1/2-deficient cells this homologous recombination repair is abrogated, as well as the mechanisms of fork recovery, leading to DSB accumulation to levels incompatible with cell survival.