Presentation Authors: Tomoaki Takai*, Xiao Bai, Takuya Tsujino, Adam Kibel, Li Jia, Boston, MA
Introduction: While next-generation anti-androgens have been developed, castration-resistant prostate cancer (CRPC) remains a lethal disease. There is an urgent need for new therapeutic options in advanced PCa. Inhibitors of poly (ADP-ribose) polymerase (PARP) are a new type of targeted therapy, which works by preventing the enzyme PARP from DNA damage site in tumor cells. Cancer cells lacking BRCA1 or BRCA2 are hypersensitive to PARP inhibitor. A recent clinical trial has shown promising results of the PARP inhibitor olaparib in metastatic CRPC patients who have DNA-repair defects. However, one of the major barriers to effective treatment in clinical trials is how to select patients who most likely benefit from the treatment. Resistance to PARP inhibitor represents a formidable clinical problem. Identification of synthetic lethality is critical to address the emergence of drug-resistant cancer. To this end, we perform a genome-wide CRISPR screen to determine the mechanisms of olaparib resistance and sensitivity in PCa cells. Our research goal is to identify candidate genes as potential biomarkers for drug response and resistance.
Methods: We performed a genome-wide CRISPR screen in PCa LNCaP and C4-2B cells using a pooled CRISPR library that targets about 18,000 genes with 10 single guide RNAs (sgRNAs) for each gene. Infected cells were treated with olaparib or DMSO for 30 days (10 passages), and collected for next-generation sequencing. The enrichment of sgRNA in olaparib-treated cells was calculated using MAGeCK pipeline by comparing to DMSO-treated cells and cells prior to treatment. Positive selection (enriched sgRNAs) identifies gene knockouts that result in resistance to olaparib. Negative selection (depleted sgRNAs), on the other hand, identifies gene knockouts that increase olaparib sensitivity.
Results: We detected more than a hundred candidate genes from both positive and negative selection. As expected, genes from negative selection are highly enriched for DNA repair pathways, including well-known BRCA1 and ATM. We have further validated a subset of novel DNA repair genes through generating knockout cell lines and testing their response to olaparib. We also identified a number of genes from the positive selection, knockout of which led to PCa cell resistance to olaparib. These genes include PARP1, PARP2 and PARG, which represent novel PARP inhibitor resistance mechanisms.
Conclusions: We have demonstrated the power of CRISPR screening for identification of novel genes as potential biomarkers of olaparib resistance and sensitivity in PCa.