Introduction: Clinical evidence from trials employed PARP inhibitor (PARPi) in prostate cancer (PCa) demonstrated that patients harboring BRCA1/2 mutations are sensitive to PARPi. However, whether and to what extent PARPis can be used to treat tumors with non-BRCA1/2 genomic alterations remains controversial. Here, we performed genome-wide CRISPR/Cas9 knockout (KO) screens in PCa cells and identified novel genes whose loss has a profound impact on the response to PARPi.
Methods: Genome-wide CRISPR-KO screens were carried out in PCa cell lines under the treatment with PARPi olaparib (Figure 1). Genomic DNA was then harvested, followed by library amplification and analysis by next generation sequencing. The abundance of single guide RNA (sgRNA) for each gene was assessed by ß-score and the differential ß-score was further calculated using the MAGeCKFlute pipeline by comparing olaparib treatment to vehicle treatment.
Results: We identified 216, 243, 153, and 211 negatively selected genes, loss of which sensitize cells to olaparib in LNCaP, C4-2B, 22Rv1, and DU145 cells, respectively (Figure 2A). Gene Ontology analyses revealed that negatively selected genes were enriched for DNA replication and repair (Figure 2B and C). On the other hand, we identified 280, 347, 222, and 281 positively selected genes, loss of which renders cells resistant to olaparib (Figure 2D). Analyses of positively selected genes showed that cell cycle and positive regulation of gene expression were largely enriched (Figure 2B and E). Finally, we ranked all genes based on the average of differential ß-scores from all four cell lines. Representative negatively and positively selected genes are shown in Figure 2F. Among the common negatively selected genes, RNASEH2B is most frequently deleted in PCa, whose loss was validated to confer hypersensitive to olaparib (Figure 2G and H).
Conclusions: Our findings may extend the benefit of PARP inhibition beyond BRCA1/2-dificient tumors and support reevaluation of currently used biomarkers for PARP inhibition in PCa.
Source of Funding: National Institutes of Health grant 1R21CA252578-01 (to L.J.) National Institutes of Health grant 1R01CA262524-01 (to L.J.)
.jpg)
.jpg)
