A Synthetic Lethal Screen Identifies DNA Repair Pathways that Sensitize Cancer Cells to Combined ATR Inhibition and Cisplatin Treatments

Abstract

The DNA damage response kinase ATR may be a useful cancer therapeutic target. ATR inhibition synergizes with loss of ERCC1, ATM, XRCC1 and DNA damaging chemotherapy agents. Clinical trials have begun using ATR inhibitors in combination with cisplatin. Here we report the first synthetic lethality screen with a combination treatment of an ATR inhibitor (ATRi) and cisplatin. Combination treatment with ATRi/cisplatin is synthetically lethal with loss of the TLS polymerase ζ and 53BP1. Other DNA repair pathways including homologous recombination and mismatch repair do not exhibit synthetic lethal interactions with ATRi/cisplatin, even though loss of some of these repair pathways sensitizes cells to cisplatin as a single-agent. We also report that ATRi strongly synergizes with PARP inhibition, even in homologous recombination-proficient backgrounds. Lastly, ATR inhibitors were able to resensitize cisplatin-resistant cell lines to cisplatin. These data provide a comprehensive analysis of DNA repair pathways that exhibit synthetic lethality with ATR inhibitors when combined with cisplatin chemotherapy, and will help guide patient selection strategies as ATR inhibitors progress into the cancer clinic.

Fig 1. ATR inhibition sensitizes cells to cisplatin.

(A and B) U2OS cells were treated with increasing doses of cisplatin or the ATR inhibitor (ATRi) alone or in combination for 72 hours. Cell viability was determined with alamar blue and reported as a percentage of the untreated control cells. Analysis of synergy between cisplatin and ATRi using Bliss Independence (B) and isobologram analysis (C) as described in the materials and methods. (D) U2OS cells were treated with 1μM cisplatin, 1μM ATRi, or both (ATRi + Cis); cells were released into media without drugs after 24, 48, or 72 hours and allowed to form colonies. (E) Cells were treated with 1μM cisplatin or 1μM ATRi for 24hr, ATRi (24hr) followed by cisplatin (24hr) A→C, or cisplatin (24hr) followed by ATRi (24hr) C→A. Error bars in all panels are standard deviation (n = 3).

Fig 2. ATR inhibition resensitizes cisplatin-resistant cancer cells to cisplatin.

(A) MDA-MB-468 (468) and MDA-MB-468 cisplatin-resistant (468-CR) cells were treated with increasing doses of cisplatin for 96h prior to measuring cell viability with alamar blue. (B and C) MDA-MB-468 and MDA-MB-468 cisplatin-resistant cells were treated with ATRi alone or ATRi and cisplatin at 0.5μM (B) or 3μM (C) for 96h prior to measuring cell viability with alamar blue. (D and E) Bliss independence synergy between cisplatin and ATRi in MDA-MD-468 (D) and MDA-MB-468 cisplatin-resistant cells (E). Error bars in all panels are standard deviation (n = 3).

Fig 3. siRNA screen identifies synthetic lethal interactions with cisplatin and ATR inhibitor treatment.

(A) Schematic of the siRNA screen. U2OS cells were transfected with siRNAs and then either left untreated or treated with 1μM ATRi, 0.5μM cisplatin, or ATRi and cisplatin. Cell viability was measured with alamar blue. (B) Viability of the non-targeting (NT) and ATR siRNA controls using the screen conditions. The values represent the mean ± SD of the three independent replicates of the screen. (C-E) The robust z-scores of treated compared with untreated cell viability were determined for each siRNA in the library for ATRi (C), cisplatin (D), and ATRi and cisplatin (E). The red circles represent the 8 unique siRNAs targeting ATR and ATRIP present in the library and the red line indicates a robust z score of -1.3. (F) Summary of the overlap of synthetic lethal relationships with ATRi, cisplatin, or ATRi and cisplatin. (G) Complete list of genes that exhibit a synthetic lethal relationship with ATRi, cisplatin, or ATRi and cisplatin. Genes with 2 or more siRNAs with robust z-scores of less than -1.3 are considered as synthetic lethal. The ATRi single-agent screen was previously published and included for comparison to the other drug treatments [18].

Fig 4. Loss of homologous recombination is not synthetic lethal with ATRi/cisplatin.

BRCA2-deficient VC8 cells or VC8 cells complemented with a BRCA2 expression vector were treated with ATRi, cisplatin, and ATRi/cisplatin combination. (A) Sensitivity of BRCA2-deficient cells to cisplatin. (B and C) Sensitivity of BRCA2-deficient cells to ATRi alone and ATRi with either 0.05 or 0.1μM cisplatin. Cell viability was measured with alamar blue after 72 hours and reported as a percent of the untreated control cells. Error bars are standard deviation (n = 3).

Fig 5. Loss of mismatch repair is not synthetic lethal with ATRi/cisplatin.

(A and C) Cells were treated with increasing doses of ATRi alone or in combination with 0.5μM cisplatin for 72 hours. Cell viability was measured with alamar blue and reported as a percent of the untreated control cells for each cell line. (B and D) Cells were treated with 1μM ATRi, 0.5μM cisplatin, or both (A + C); cells were released into media without drugs after 24 hours and allowed to form colonies. (A and B) MLH1-deficient HCT-116 cells and HCT-116 cells complemented with MLH1. (C and D) MSH2-deficient HEC59 cells and HEC59 cells complemented with MSH2. Error bars in all panels are standard deviation (n = 3).

Fig 6. ATR inhibition is synthetic lethal with pharmacologic inhibition of PARP.

(A) U2OS cells were treated with increasing doses of ATR and PARP inhibitors for 96 hours. Cell viability was measured with alamar blue and reported as a percent of the untreated control. Synergy between ATR and PARP inhibition using Bliss Independence (B) and isobologram analysis (C) as described in the materials and methods. (D) Cells were treated with 1μM ATRi, 2nM PARPi, or both (A + P) and cells were released into media without drugs after 72 hours and allowed to form colonies. Error bars in all panels are standard deviation (n = 3).

Fig 7. Loss of REV3 is synthetic lethal with ATRi and cisplatin.

(A-F) H157 NSCLC cells were transfected with non-targeting siRNA (siNT) or two siRNAs targeting REV3 (number 2 and 4 refer to specific sequences described in the materials and methods). Cells were then treated with ATRi, cisplatin, and ATRi and cisplatin. Cell viability was determined with alamar blue and reported as a percent of the untreated control cells. (A) Sensitivity of REV3 knockdown cells to cisplatin. (B and C) Sensitivity of REV3 knockdown cells to ATRi and ATRi with 0.1μM cisplatin. Bliss independence synergy between ATRi and cisplatin in control (D) and REV3 knockdown cells (E and F). Error bars in all panels are standard deviation (n = 3).

Fig 8. Loss of 53BP1 is synthetic lethal with ATRi and cisplatin.

(A-F) H157 NSCLC cells were transfected with non-targeting siRNA (siNT) or two siRNAs targeting 53BP1 (number 2 and 3 refer to specific sequences described in the materials and methods). Cells were then treated with ATRi, cisplatin, and ATRi and cisplatin. Cell viability was determined with alamar blue and reported as a percent of the untreated control cells. (A) Sensitivity of 53BP1 knockdown cells to cisplatin. (B and C) Sensitivity of 53BP1 knockdown cells to ATRi and ATRi with 0.5μM cisplatin. Bliss independence synergy between ATRi and cisplatin in control (D) and 53BP1 knockdown cells (E and F). Error bars in all panels are standard deviation (n = 3).

Fig 9. ATR is required to maintain genome integrity in 53BP1-deficient cells.

H157 cells were transfected with non-targeting (siNT) or 53BP1 siRNA and then treated with 2.5μM ATRi for 24 hours. Cells were then fixed and stained with DAPI to visualize nuclei (A) and γH2AX to identify sites of DNA damage (B). (C) Quantification of the γH2AX intensity of cells shown in B. Box and whiskers plot shows the mean and the range of the samples, * p<0.01, ns (not significant).