Increasing sensitivity to olaparib through inhibition of discoidin domain receptor 2 (DDR2) in homologous-recombination proficient ovarian cancer models

Abstract

<h3>Objectives:</h3> Discoidin Domain Receptor 2 (DDR2) is a tyrosine kinase receptor which binds the most common extracellular matrix protein, fibrillar collagen type 1. DDR2 expression is critical for invasion and migration of ovarian cancer tumor cells, making DDR2 a potential target for new treatments. This study aimed to determine whether genetic inactivation of DDR2 in ovarian cancer cells would increase sensitivity to treatment with PARP inhibitor. <h3>Methods:</h3> Two ovarian cancer cells lines, ES2 and COV362, were used for <i>in vitro</i> assays. Stable hairpin control and DDR2 knockdowns (shSCRM and shDDR2) were used for ES2 and transient small interfering RNA knockdown of DDR2 (siControl and siDDR2) were used for COV362. Homologous recombination (HR) status was determined by quantification of RAD-51 foci relative to controls using immunofluorescence after irradiation. Cell viability and survival after treatment with olaparib was quantified using MTS assays and clonogenics. The DNA damage response after treatment with olaparib was analyzed using immunofluorescence. <i>In vivo</i> studies to evaluate the sensitivity to olaparib in the presence or absence of DDR2 expression were performed in nude mice after intraperitoneal injection with ES2shSCRM or shDDR2 cells. <h3>Results:</h3> ES2 and COV362 cells were found to have a 2-fold increase in RAD51 foci after irradiation indicating HR proficiency. DDR2 genetically inactivated cells (ES2shDDR2) were found to be more HR deficient than DDR2 expressing cells (ES2shSCRM) which remained HR proficient (0.80 vs 4.78 relative RAD51 foci per cell). We found that genetic inactivation of DDR2 in the ES2s and COV362s led to increased sensitivity to olaparib with an 80% (15uM vs 73uM) and 68% (62uM vs 195uM) reduction in IC50s, respectively, compared to DDR2 expressing cells. Additionally, ES2shDDR2 cells had decreased cell survival by clonogenic assay after treatment with olaparib compared to ES2shSCRM cells as measured by absorbance (0.64 vs 0.80, p=0.005). We found that DNA damage and the repair response was altered in DDR2 genetically inactivated cells treated with olaparib with an increase in gamma-H2AX (8 vs 6 foci per cell, p<0.0001) and 53BP1 foci (3 vs 1 foci per cell, p<0.0001) when compared to DDR2 expressing cells. In an intraperitoneal metastatic mouse model with the HR-proficient cell line ES2, we found that genetic inactivation of DDR2 led to decreased tumor burden when treated with olaparib compared to DDR2-expressing cells treated with olaparib (1499mm³ vs 3236mm³, p=0.04). <h3>Conclusions:</h3> DDR2 knockdown sensitizes HR proficient ovarian cancer cells to treatment with PARP inhibitor in both <i>in vitro</i> and <i>in vivo</i> models through a mechanism of induced HR deficiency and increased DNA damage. Future experiments will explore whether a DDR2 inhibitor can improve sensitivity to PARP inhibitor in HR proficient ovarian cancer models.