Published OnlineFirst April 2, 2014; DOI: 10.1158/1535-7163.MCT-13-0906-T Molecular Cancer Cancer Biology and Signal Transduction Therapeutics New Insights into PARP Inhibitors' Effect on Cell Cycle and Homology-Directed DNA Damage Repair Petar Jelinic and Douglas A. Levine Abstract In preclinical and clinical studies, olaparib and veliparib are the most represented PARP inhibitors (PARPi), which mainly target homologous DNA damage repair pathway-deficient cancer cells. Their off-target effects are not fully understood, especially with regard to cell cycle and homology-directed DNA damage repair. Our objective was to comparatively evaluate olaparib and veliparib in this context and correlate our findings with their therapeutic potential. We used a well-established direct repeat GFP (DR-GFP) reporter assay in U2OSDR-GFP and H1299DR-GFP cells and measured DNA damage repair activity upon drug treatment. Olaparib-treated U2OSDR-GFP cells showed a dramatic decrease in DNA damage repair versus veliparib irrespective of inhibitory potency. We demonstrate that this effect was a result of olaparib’s strong effect on the cell cycle. Unlike in veliparib-treated U2OSDR-GFP cells, in olaparib-treated cells S-phase decreased and G2-phase increased sharply, indicating a G2-phase arrest-like state and replicative stress. This was further confirmed by upregulation of p53 and p21 and accumulation of cyclin A. Lack of the same effect in p53-null H1299DR-GFP cells suggested that olaparib’s effect is p53 related, which was confirmed in p53-depleted U2OSDR-GFP and p53-null HCT116 cells. Importantly, we also demonstrate that olaparib, but not veliparib, induced a robust phosphorylation of Chk1, a crucial component of the replicative stress response pathway. Our data show olaparib and veliparib differ in their off-target effects; olaparib, unlike veliparib, mitigates DNA damage repair activity via G2 cell-cycle arrest-like effect in a p53-dependent manner. These off-target effects may add to PARPis’ anticancer properties. Mol Cancer Ther; 13(6); 1645–54. Ó2014 AACR. Introduction strand breaks (DSB), which are repaired predominantly PARP inhibitors (PARPi) target PARPs by suppres- by the homology-directed DNA damage repair (HDR) sing their enzymatic activity PARylation—the attach- pathway. If the HDR pathway is impaired, cells rely on ment of poly (ADP)-ribose polymers onto themselves more error-prone pathways, resulting in genomic insta- and other acceptor proteins (1, 2). PARP-1, the most bility and loss of viability (7, 8). PARP inhibition in abundant among all PARPs, is a stress-sensing protein HDR-compromised cells was the first demonstration of that acts mainly via the base excision repair (BER) the concept of synthetic lethality working in cancer pathway. One of the proposed mechanisms is that cells. This provided a strong rationale for the develop- PARP-1 binds to single-strand break (SSB) intermedi- ment of PARPis as a treatment for patients with cancer ates. After binding and self-PARylating, PARP-1 dis- with an impaired HDR pathway (9, 10). sociates from the SSB, which subsequently triggers the Olaparib and veliparib are the most extensively stud- downstream BER pathway to repair the lesion (3). It has ied PARPis in preclinical and clinical studies, showing been suggested that PARP inhibition impairs the BER greatest success in ovarian and breast cancers (11–14). pathway by trapping PARP-1 at the SSBs and blocking Although data from early clinical trials looking at PAR- further repair (4–6). When the replication fork encoun- Pis were promising, more recent results have failed to ters these SSBs, they are converted into DNA double- meet expectations, prompting a call for a better under- standing of these drugs’ mode of action (15). PARPis are þ nicotinamide (NAM) mimetics that compete with NAD for the NAM pocket within PARPs’ catalytic domain (16, Authors' Affiliation: Gynecology Service, Department of Surgery, Memo- rial Sloan Kettering Cancer Center, New York, New York 17). Although specifically developed as PARP-1 inhibi- tors, all PARPis exhibit a polypharmacology targeting Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/). other members of the PARP family (16, 18). Olaparib and veliparib fall within the same group of PARPs with the Corresponding Author: Douglas A. Levine, Gynecology Service, Depart- ment of Surgery, Memorial Sloan Kettering Cancer Center, 1275 York similar polypharmacology (16). It is well documented Avenue, New York, NY 10065, Phone: 212-639-8125; Fax: 212-717- that olaparib and veliparib have similar inhibitory effects 3214; E-mail: [email protected] on PARP-1’s enzymatic activity (19). However, their off- doi: 10.1158/1535-7163.MCT-13-0906-T target effects, which may contribute to their differential Ó2014 American Association for Cancer Research. therapeutic properties, are not fully understood. It has www.aacrjournals.org 1645 Downloaded from mct.aacrjournals.org on September 25, 2021. © 2014 American Association for Cancer Research. Published OnlineFirst April 2, 2014; DOI: 10.1158/1535-7163.MCT-13-0906-T Jelinic and Levine been demonstrated that some PARPis cause cell-cycle serial PARPi dilutions, U2OSDR-GFP (1 Â 106) cells and abrogation, which has been recognized as an off-target H1299DR-GFP (0.5 Â 106) cells were transfected by effect (20, 21). Only a few PARPis share this character- electroporation using Nucleofector (Amaxa) with 2 or istic, suggesting that it is not related to their enzymatic 1 mgpCMV-I-SceI,respectively. Cells were then split activity-targeted inhibitory properties. Considering that into 2 wells of a 6-well plate. After 72 hours, cells were the HDR pathway is active predominately during the harvested and prepared for either HDR or PARP replicative phase of the cell cycle, the concept of synthetic activity assessment. For HDR activity assessment, cells lethality caused by PARP inhibition works only in HDR- were harvested with 250 mL trypsin. An equal amount impaired cells in which cell cycle is not arrested. There- of media was added to neutralize trypsin, and per- fore, it is essential to evaluate PARPis’ therapeutic prop- centages of GFP-positive cells were determined by erties in the context of the cell cycle and HDR, which thus flow cytometry (FACSCalibur; Becton Dickinson). For far has not been addressed. The important question PARP activity assessment, we used Trevigen’s Univer- remains whether or not olaparib and/or veliparib, the sal Colorimetric PARP assay (#4677-096-K) according 2 most common PARPis in clinical trials, have an impact to the manufacturer’s directions, with minor modifica- on the cell cycle that could compromise HDR activity and tions. Briefly, cells were washed 3 times in ice-cold PBS render these inhibitors less effective in the context of and lysed with PARP lysis buffer supplemented with synthetic lethality. 400 mmol/L NaCl, 1% Triton X-100, 0.4 mmol/L phe- To address this hypothesis, we took advantage of a nylmethylsulfonylfluoride, and protease inhibitor well-established direct repeat GFP (DR-GFP) reporter cocktail (P8340; Sigma-Aldrich). After 30 minutes of assay to comparatively evaluate HDR activity in cells incubation on ice, with occasional flicking, the lysates treated with olaparib or veliparib. We report a significant were centrifuged for 15 minutes at 4C at 10,000 Â g. differential effect between the 2 PARPis on cell cycle and The supernatant was transferred to a fresh tube, and HDR activity irrespective of their catalytic inhibitory protein concentration was measured using a Bradford potency. This work provides a new insight into these 2 assay (#500-0006; Bio-Rad). Fifty micrograms of pro- PARPis’ mode of action, which may help define selection tein lysate was mixed with PARP Cocktail into a criteria when evaluating patients for treatment with ola- histone-coated 96-well dish. After 60 minutes of room parib, veliparib, or other PARPis. temperature incubation, theplatewaswashed4times in PBS, 50 mL of Strep-HRP was added, and it was Materials and Methods incubated for 20 minutes at room temperature. The Cell culture and drug preparation plate was washed 4 times in PBS, and then 50 mLof U2OSDR-GFP human osteosarcoma cell line was cultured TACS-Sapphire was added. After 15 minutes of incu- in high glucose (4.5 g/L) Dulbecco’s Modified Eagle bation in the dark, the reaction was stopped by adding Medium (DMEM-HG) supplemented with 10% fetal calf 50 mL of 0.2M HCl. The plate was read by a microplate DR-GFP serum (FCS) and 2 mmol/L L-glutamine. H1299 reader (Synergy HT; BioTek) at 450 nm. human non–small cell lung carcinoma and HCT116 À À [wild-type (WT) and p53 / ] colorectal carcinoma cell Cell-cycle analysis lines were cultured in RPMI media supplemented with After incubation in specific conditions, cells were har- 10% FCS. U2OSDR-GFP and H1299DR-GFP cell lines were vested and counted to ensure equal numbers of cells were kindly donated by Dr. S. Powell (Department of Radiation processed (1 Â 106). Cells were then washed with ice-cold À À Oncology, MSKCC). HCT116 (WT and p53 / ) cell lines PBS, fixed in 80% ethanol at À20C for a minimum of 2 were kindly donated by Dr. G. Schwartz (Department of hours to overnight. Subsequently, cells were washed in Medicine, MSKCC). All cell lines were authenticated by ice-cold PBS and stained with propidium iodide DNA the short tandem repeat DNA profiling method (Genetica staining solution containing RNase A (200 mg/mL), Triton DNA Laboratories) using the DSMZ database (http:// X-100 (0.1%, v/v), and propidium iodide (50 mg/mL) in www.dsmz.de/). All cells tested negative for mycoplas- PBS. After 20 minutes of incubation at 37C, the samples ma. Olaparib (#S1060), veliparib (#S1004), and PJ34 were stored at 4C and processed by flow cytometer (#S7300) were obtained from Selleck Chemicals.