ATR inhibition reverses the resistance of homologous recombination deficient MGMTlow/MMRproficient cancer cells to temozolomide

Abstract

The therapeutic efficacy of temozolomide (TMZ) is hindered by inherent and acquired resistance. Biomarkers such as MGMT expression and MMR proficiency are used as predictors of response. However, not all MGMT^low/-ve/MMR^proficient patients benefit from TMZ treatment, indicating a need for additional patient selection criteria. We explored the role of ATR in mediating TMZ resistance and whether ATR inhibitors (ATRi) could reverse this resistance in multiple cancer lines. We observed that only 31% of MGMT^low/-ve/MMR^proficient patient-derived and established cancer lines are sensitive to TMZ at clinically relevant concentrations. TMZ treatment resulted in DNA damage signaling in both sensitive and resistant lines, but prolonged G₂/M arrest and cell death were exclusive to sensitive models. Inhibition of ATR but not ATM, sensitized the majority of resistant models to TMZ and resulted in measurable DNA damage and persistent growth inhibition. Also, compromised homologous recombination (HR) via RAD51 or BRCA1 loss only conferred sensitivity to TMZ when combined with an ATRi. Furthermore, low REV3L mRNA expression correlated with sensitivity to the TMZ and ATRi combination in vitro and in vivo. This suggests that HR defects and low REV3L levels could be useful selection criteria for enhanced clinical efficacy of an ATRi plus TMZ combination.

Keywords: ATR; HR; MMR; REV3L; TMZ.

Figure 1. Response of cancer lines to temozolomide in the context of MGMT and MMR.

(A) Thirty-nine cell lines were treated with TMZ (40 μM) for 96 h. MGMT^+ve cells were exposed to TMZ with or without AZD5896 (20 μM). Heat maps represent the mean percent growth inhibition of at least 3 independent experiments. (B) MGMT expression was verified by Western blotting (left) from lentivirally transduced BL0293-EV (empty vector control) and BL0293-MGMT. The cells were treated with increasing concentrations of TMZ for 96 h. BL0293-MGMT were exposed to TMZ with or without AZD5896 (20 μM). Percent proliferation relative to the vehicle control is plotted as the mean ± SEM of 2 independent experiments (^*** p < 0.001 relative to BL0293 + MGMT). (C) DLD-1 and DLD-1 + Chr2 (left), HCT116 and HCT116 + Chr3 (right) were treated with increasing concentrations of TMZ for 96 h ± AZD5896 (10 μM). Percent proliferation relative to the vehicle control is plotted as the mean ± SEM of 3 independent experiments. (D) BL0293 and BL0479-72 were treated with TMZ (40 μM) and the cell cycle was analyzed at indicated time points. The percentage of cells in the G₂/M phase are plotted as mean ± SEM from 2 independent experiments (compared to 24 h, ns (non-significant), ^*** p < 0.001, ^**** p < 0.0001). (E) BL0293 and BL0479-72 were treated with TMZ for the indicated time points. The activation and expression of DDR pathway components were assessed by Western blot and the average fold expression ± SEM of pCHK1, pTIF1β, Cleaved (Cl.) PARP and γH2AX in lysates of BL0293 (blue lines) and BL0479-72 (red lines) treated with TMZ (40 μM) over vehicle treated cells at each time point has been quantified from at least 3 experiments repeats and shown in (F).

Figure 2. ATR inhibition, in combination with TMZ, induces persistent DNA damage, growth inhibition and cell death in the majority of MGMT^−ve/MMR^proficient models.

(A) BL0479-72 cells were exposed to 6 concentrations of TMZ in combination with 5 concentrations of VE821 for 96 h (left), while BL0293 cells were treated with the same 6 concentrations of TMZ only. Results were plotted as line graphs with average % T/V calculated as 100× (blanked luminescence of wells at each dose/blanked luminescence of vehicle control wells), ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001 compared to same dose of TMZ as a single agent (B) BL0479-72 cells were treated with vehicle or TMZ (40 μM) ± VE821 (1 μM) for 48 h and DDR pathway components were assessed by Western blot (C) Early and late apoptosis (Annexin V + PI positive cells) was measured via Annexin V^+ve/PI^+ve staining after BL0479-72 cells were treated with vehicle or TMZ (40 μM) ± VE821 (1 μM) for 72 h. The percentage of early and late apoptotic cells are plotted as mean ± SEM of 3 independent experiments (p < 0.01 compared to ^**DMSO, ^##TMZ, or ^{^^}VE821). (D) BL0479-72 were treated with vehicle or TMZ (40 μM) ± VE821 (0.5 μM) for 96 h. The treatments were either removed after 96 h and replaced with inhibitor-free medium (Recovery) or retained for an additional 6 days. Proliferation was assessed by SRB at the indicated days beyond 96 h and plotted as mean ± SEM for 3 independent experiments. (E) Thirty-nine cell lines were treated with TMZ (40 μM) ± VE821 (1.25 μM) for 96 h. The 20 MGMT^+ve cells were also treated with AZD5896 (20 μM). Heat maps indicate the average percent growth inhibition from at least 3 independent experiments. Abbreviations B: breast cancer; M: melanoma; G: glioblastoma; A: astrocytoma; R: renal cancer; L: leukemia; LY: lymphoma; LG: lung cancer; HN: head and neck cancer; BL: bladder cancer; ME: mesothelioma; O: ovarian cancer; P: prostate cancer.

Figure 3. Similar to ATR inhibition, CHK1 inhibition induces DSB signaling, decreases HR-dependent DNA repair and sensitizes cells to TMZ.

(A–B) BL0479-72 cells were treated with vehicle or TMZ ± VE821 or ± LY2606318 for 72 h and assessed for (A) induction of DNA damage, PARP cleavage and mitotic entry by Western blot and (B) early and late apoptosis by Annexin V^+ve and PI^+ve staining, respectively. The results are plotted as mean ± SEM of two independent experiments (^* p < 0.05, ^** p < 0.01 compared to DMSO, TMZ or VE821; ^# p < 0.05 compared to LY2606318). (C–E) Growth inhibition of BL0479-72 after combination treatment with TMZ and (C) LY2606318 (D) LY2606368 and (E) MK8776. Results were plotted as line graphs with average % T/V calculated as 100× (blanked luminescence of wells at each dose/blanked luminescence of vehicle control wells), ^* p < 0.05, ^** p < 0.01, ^*** p < 0.001, ^**** p < 0.0001 compared to same dose of TMZ as a single agent. (F) The HR function of BL0479-72 cells was measured by the DRGFP assay following a 48 h treatment with inhibitors of ATR (VE821 and AZD6738), CHK1 (LY2606318, MK8776, LY2606368), RAD51 (B02), DNAPK (NU7441), ATM (KU55933) and pan-CDK (roscovitine). Results are plotted as the mean ± SEM of at least 2 independent experiments (^* p < 0.05, ^*** p < 0.001, ^**** p < 0.0001).

Figure 4. HR deficiencies sensitize cancer cell lines to the combination of TMZ and ATRi.

(A) Homologous recombination capacity of UWB1.289 ± BRCA1 cells by the DRGFP assay. B02 was used at 2.5 μM. The results are shown as mean ± SEM of two independent experiments. (B) UWB1.289 and UWB1.289-BRCA1 cells were exposed to increasing concentrations of TMZ (in the presence of 10 μM AZD5896) ± VE821 (0.5 μM) for 96 h. Cell proliferation relative to the vehicle control is plotted as the mean ± SEM of two independent experiments (^* p < 0.05 compared to UWB1.289-BRCA1 + VE821). (C) UWB1.289-BRCA1 cells and (D) UWB1.289 cells were exposed to increasing concentrations of TMZ (in the presence of AZD5896, 10 μM) ± VE821 (0.5 μM) and ± B02 (2.5 μM) for 96 h. Cell proliferation relative to the vehicle control is plotted as the mean ± SEM of two independent experiments (^* p < 0.05 compared to UWB1.289-BRCA1 + VE821). (E) BL0479-72 cells were treated with vehicle or TMZ (40 μM) ± VE821 (1 μM) ± B02 pre-treatment (1 h, 10 μM) and exposed for 72 h before assessing DDR signaling by Western blot. (F) BL0479-72 were treated with increasing concentrations of TMZ ± B02 pre-treatment (1 h, 10 μM) and VE821 (1.25 μM) and exposed for 96 h. Cell proliferation relative to the vehicle control is plotted as the mean ± SEM of two independent experiments. (^* p < 0.05 compared to VE821). (G) BL0479-72 cells stably expressing non-target (NT) shRNA or RAD51 shRNA 77 (SH-77) were treated with TMZ. After 72 h DDR signaling was assessed by Western blot. Densitometry was used to quantify RAD51 expression and the values from each condition were averaged from the two experiments. The ratio of RAD51 expression relative to the control (vehicle + shNT) is shown for each condition at the bottom of the figure. (H) BL0479-72 cells stably expressing NT shRNA or SH-77 were treated with increasing concentrations of TMZ ± VE821 (1.25 μM) for 96 h. Cell proliferation relative to the vehicle control is plotted as the mean ± SEM of two independent experiments.

Figure 5. Low REV3L mRNA expression is associated with sensitivity to the combination of TMZ + ATRi.

The mRNA expression of TLS polymerases and RRM2 was evaluated by qPCR. The values were calculated as 2^-[GOI Cts-GAPDH Cts]*10³) from qPCR data (GOI: gene of interest). (A) The expression of TLS polymerases and RRM2 in models sensitive (S, GI ≥ 50%) and resistant (R, GI < 50%) to TMZ. (B) The expression of TLS polymerases and RRM2 in models sensitive (S, GI ≥ 20%) and resistant (R, GI < 20%) to VE821. (C) The expression of TLS polymerases and RRM2 in models sensitive (S, GI ≥ 50%) and resistant (R, GI < 50%) to TMZ + VE821. (D) The expression of REV3L in models not sensitive to TMZ or ATRi as single agents but sensitized by TMZ + ATRi. Mean expression is reported ± SEM with ^* p < 0.05, ^** p < 0.01 by an unpaired t-test.

Figure 6. MGMT^low/MMR^proficient/LOH^high PDX models are responsive to the TMZ + ATRi combination.

Mice harboring (A–B) BL0382 or (C–D) LG0520 tumor fragments were treated for 2 cycles with vehicle, TMZ (50 mg/kg; QDx5, 2 days rest), VE822 (45 mg/kg; QDx4, 3 days rest) or TMZ + VE822. Treatments were administered between days 22 and 33 (BL0382) or days 28 and 39 (LG0520). (A, C) Mean tumor volume ± SD is shown (^* p < 0.05 by the Holm-Sidak method). (B–D) Survival graphs for PDXs (B) BL0382 and (D) LG0520. Arrowheads depict the end of treatment.