Published OnlineFirst August 21, 2015; DOI: 10.1158/0008-5472.CAN-15-1107 Cancer Therapeutics, Targets, and Chemical Biology Research Identification of P450 Oxidoreductase as a Major Determinant of Sensitivity to Hypoxia-Activated Prodrugs Francis W. Hunter1, Richard J. Young2,3, Zvi Shalev4, Ravi N. Vellanki4, Jingli Wang1, Yongchuan Gu1,5, Naveen Joshi1, Sreevalsan Sreebhavan1, Ilan Weinreb4, David P. Goldstein6, Jason Moffat7, Troy Ketela8, Kevin R. Brown8, Marianne Koritzinsky4,7,9, Benjamin Solomon2,3,10, Danny Rischin3,10, William R. Wilson1, and Bradly G. Wouters4,7,11,12 Abstract Hypoxia is a prevalent feature of many tumors contributing SN30000 in three different genetic backgrounds. No other to disease progression and treatment resistance, and therefore genes consistently modified SN30000 sensitivity, even within constitutes an attractive therapeutic target. Several hypoxia- a POR-negative background. Knockdown or genetic knockout activated prodrugs (HAP) have been developed, including of POR reduced SN30000 reductive metabolism and clono- the phase III candidate TH-302 (evofosfamide) and the pre- genic cell death and similarly reduced sensitivity to TH-302 clinical agent SN30000, which is an optimized analogue of the under hypoxia. A retrospective evaluation of head and neck well-studied HAP tirapazamine. Experience with this therapeu- squamous cell carcinomas showed heterogeneous POR expres- tic class highlights an urgent need to identify biomarkers of sion and suggested a possible relationship between human HAP sensitivity, including enzymes responsible for prodrug papillomavirus status and HAP sensitivity. Taken together, our activation during hypoxia. Using genome-scale shRNA screens study identifies POR as a potential predictive biomarker of and a high-representation library enriched for oxidoreductases, HAP sensitivity that should be explored during the clinical we identified the flavoprotein P450 (cytochrome) oxidoreduc- development of SN30000, TH-302, and other hypoxia-directed tase (POR) as the predominant determinant of sensitivity to agents. Cancer Res; 75(19); 4211–23. Ó2015 AACR. Introduction ination (2) and resistance to ionizing radiation (3). Reflecting these effects, hypoxia has been compellingly linked to adverse Hypoxia is a prevalent feature of the tumor microenvironment outcome in malignancies for which radiotherapy is administered that has been implicated in neoplastic progression (1), dissem- with curative intent (4), and agents that target tumor hypoxia promise to improve outcomes in radiation oncology. Hypoxia-activated prodrugs (HAP) have been developed to 1 Auckland Cancer Society Research Centre, Faculty of Medical and target hypoxic tissue (5). HAPs are metabolized, typically by Health Sciences, University of Auckland, Auckland, New Zealand. 2Research Division, Peter MacCallum Cancer Centre, East Melbourne, flavoenzyme-catalyzed one-electron reduction, to a prodrug rad- Victoria, Australia. 3Sir Peter MacCallum Department of Oncology, ical from which an active cytotoxin is subsequently derived (6). In 4 University of Melbourne, Parkville, Victoria, Australia. Princess Mar- the presence of oxygen, the radical is back-oxidized to the initial garet Cancer Centre, University Health Network, Toronto, Ontario, Canada. 5AnQual Laboratories, School of Pharmacy, Faculty of Med- prodrug, and in this manner HAPs exploit hypoxia to achieve ical and Health Sciences, University of Auckland, Auckland, New selectivity (6). The most advanced clinical agent in this class is the 6 Zealand. Department of Otolaryngology–Head and Neck Surgery, nitroimidazole mustard evofosfamide (TH-302; refs. 7, 8), which University of Toronto,Toronto, Ontario, Canada. 7Department of Radi- ation Oncology, University of Toronto,Toronto,Ontario,Canada. 8Don- is in phase III trials for pancreatic adenocarcinoma and soft tissue nelly Centre and Banting and Best Department of Medical Research, sarcoma (NCT01746979 and NCT01440088). 9 University of Toronto, Toronto, Ontario, Canada. Institute of Medical The benzotriazine di-N-oxide tirapazamine (TPZ) is the most Science, University of Toronto, Toronto, Ontario, Canada. 10Depart- ment of Medical Oncology, Peter MacCallum Cancer Centre, East thoroughly studied HAP (9). Metabolic activation of TPZ releases Melbourne, Victoria, Australia. 11Department of Medical Biophysics, free radicals that damage DNA (10) and selectively kill hypoxic University of Toronto,Toronto, Ontario, Canada. 12Ontario Institute for cells (11). TPZ showed promising activity in early clinical studies Cancer Research, Toronto, Ontario, Canada. (12, 13). However, the addition of TPZ to chemoradiotherapy Note: Supplementary data for this article are available at Cancer Research for head and neck squamous cell carcinoma (HNSCC) failed to Online (http://cancerres.aacrjournals.org/). extend survival in a pivotal trial (14). This disappointing out- Corresponding Author: Bradly G. Wouters, Princess Margaret Cancer Research come has been attributed to poor compliance with radiotherapy Tower, MaRS Centre, 12th Floor Room 12-401, 101 College Street, Toronto, ON, protocols at some centers (15), differential impact of TPZ in M5G 1L7, Canada. Phone: 416-581-7861; Fax: 416-946-2984; E-mail: HPV-positive and -negative tumors (16) and failure to select pati- [email protected] ents for presence of tumor hypoxia (17, 18). Moreover, TPZ shows doi: 10.1158/0008-5472.CAN-15-1107 poor tumor penetration (19–21), a liability that is corrected in the Ó2015 American Association for Cancer Research. second-generation TPZ analogue SN30000 (Fig. 1A), which was www.aacrjournals.org 4211 Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2015 American Association for Cancer Research. Published OnlineFirst August 21, 2015; DOI: 10.1158/0008-5472.CAN-15-1107 Hunter et al. A O- B N+ N O Lentiviruses armed with NH+ barcoded shRNAs N+ O- SN30000 O - 2 Infection 1e, H+ reductases O2 - Puro SN30000 Recovery O 1-oxide O HKO1 Selection treatment N N+ HKO2 N N HT-29 PANC-1 N N SN30000 OH radical O- BDZ radical Barcode sequencing (cytotoxin) nor-oxide N+ N N N N N C HT-29 D PANC1 4 4 shRNA24 shRNA24 shRNA26 shRNA26 shRNA27 3 2 shRNA27 2 0 1 0 Z-score − Z-score 2 All genes −1 All genes POR POR −2 −4 −3 −6 −4 HKO2 E 4 F 4 shRNA24 All genes shRNA26 3 POR shRNA27 2 2 1 0 0 Z-score All genes − −2 1 POR Z-score (PANC-1) −2 −4 −3 − 4 −6 −6 −4 −2 420 Z-score (HT-29) Figure 1. High-throughput SN30000 reductase shRNA screens in carcinoma cell lines. A, chemical structure of the benzotriazine di-N-oxide HAP SN30000 and its reductive metabolism to the cytotoxic benzodiazinyl (BDZ) radical, as based on (27). B, schematic overview of the SN30000 screening workflow used in this study. C and D, waterfall plots of Z-scores for shRNA enrichment factors, arrayed by ascending magnitude of effect, for HT-29 and PANC-1. The colored fields indicate Z > 1.96. Data points corresponding to POR-targeted hairpins are highlighted in red with specific shRNAs labeled. E, scatter plot of Z-scores for all unique shRNAs in the HT-29 and PANC-1 reductase screens. Gray fields indicate Z > 1.96 for either line, with the red field indicating Z > 1.96 in both genetic backgrounds. As for B, POR hairpins are colored red with the three shRNA significantly selected in both cell lines labeled. F, waterfall plot of Z-scores for shRNA enrichment factors, arrayed by ascending magnitude of effect, for the HKO2 (PORÀ/À) reductase screen. 4212 Cancer Res; 75(19) October 1, 2015 Cancer Research Downloaded from cancerres.aacrjournals.org on September 24, 2021. © 2015 American Association for Cancer Research. Published OnlineFirst August 21, 2015; DOI: 10.1158/0008-5472.CAN-15-1107 POR Is a Major Determinant of Sensitivity to HAPs developed using a lead selection algorithm that explicitly con- months cumulative passage from frozen stocks confirmed to be sidered extravascular transport (21, 22). These agents are equiv- Mycoplasma-free by PCR-ELISA (Roche Diagnostics). Cell lines alent in mechanism of action (22–24) and nonclinical toxicology were authenticated by commercial suppliers (HCT116, PANC-1, (22); however, SN30000 provides superior plasma pharmacoki- HEK293T) or by short tandem repeat (STR) profiling (HT-29). netics in rodents, shows faster diffusion through three-dimen- sional cell cultures and is more active in multiple xenografts (22). High-throughput shRNA screens Experience with this therapeutic class suggests that selecting As a screening library, we used the Sigma MISSION TRC1 tumors that express the intended target is essential for optimizing lentiviral shRNA pool, which encompasses 82,017 barcoded their clinical use. However, biomarkers of sensitivity to HAPs are shRNAs targeting 16,019 genes (33). Generation of a custom poorly defined and this remains a major challenge in patient "reductase" library of 1,821 shRNAs covering 359 genes, includ- stratification. Prior studies have focused on known aspects of HAP ing the majority of annotated human flavoproteins, is described pharmacology, such as hypoxia (17), and the role of DNA- in Supplementary Methods. We designed pharmacologically rel- damage responses (10, 23, 25, 26). Reductive activation is also evant shRNA screens using SN30000 exposures consistent with a requirement for HAP effect; however, the identity of enzymes plasma pharmacokinetics achieved below murine MTD (22). that activate HAPs, their potential as sensitivity biomarkers, and Screens were calibrated to robustly detect (Z-scores > 1.96) hair- variation of their