In this study, we found the blockade of VEGF/VEGFR signaling suppressed the HR activity in EOC cells without obvious mutations in HR-related genes, resulting in the sensitization of the HRP EOC cells to PARPi. DNA damage stress induced by X-ray irradiation or PARPi activated the VEGF/VEGFR signaling pathway, which increased the expression of CRY1. CRY1 enhanced the HR activity. Thus, antiangiogenic agents may potentiate the therapeutic effect of PARPi via inhibition of the VEGFR-PI3K/AKT-CRY1 axis (Fig. 4).
The VEGF/VEGFR signaling pathway is a well-known regulator of angiogenesis. Bevacizumab or cediranib, antiangiogenic agents, inhibit the VEGF/VEGFR signaling in vascular endothelial cells 26. However, the VEGF/VEGFR signaling also directly regulates cell survival, proliferation, metastasis, and sensitivity to chemotherapeutics in cancer cells 27. VEGF is produced in various cells, including cancer cells or non-neoplastic stromal cells 27. This study showed that X-ray irradiation or PARPi treatment increased the VEGF concentration in the culture media (Fig. 3A). Although the precise mechanism is unknown, these data indicate that DNA damage stress stimulated VEGF production in cancer cells. The increased VEGF activated the downstream PI3K/AKT pathway, which was shown by the increase in the phosphorylated AKT fraction (Fig. 3B and C and Supplementary Figure S2). The activating mutations in PIK3CA are common in clear cell carcinoma and relatively rare in the other histological subtypes 28. Since the increase in phosphorylated AKT by DNA damage stress was observed not only in clear cell carcinoma cell lines but also in cell lines of other histological subtypes (Fig. 3B and C and Supplementary Figure S2), the activation of VEGFR-PI3K/AKT axis might be independent of PI3KCA mutation.
CRY1 is one of the transcriptional coregulators associated with circadian rhythm 29. Disturbance of the circadian rhythm has recently been identified as an independent risk for cancer and classified as a carcinogen 29. Furthermore, circadian rhythm affects several hallmark phenotypes of cancer, including alterations in cell proliferation, survival, DNA repair, and metabolic regulation 29. Recent research showed that DNA damage stabilized CRY1, and the stabilized CRY1 temporally regulated the expression of genes required for HR in cancer cells 30. This study reported an increase in the CRY1 expression by DNA damage, including X-ray irradiation and PARPi treatment, and the knockdown of CRY1 by RNAi significantly suppressed the HR activity (Figs. 2 and 3 and Supplementary Figure S2). These data support the contribution of CRY1 in the enhancement of DNA damage repair by HR. Additionally, bevacizumab decreased the CRY1 expression in irradiated cells and suppressed the HR activity (Fig. 2C and D) and the suppression of HR activity by bevacizumab was restored by the exogenous expression of CRY1 (Fig. 2D). These results showed that bevacizumab inhibited an increase in the CRY1 expression induced by DNA damage, resulting in the suppression of HR activity and enhancement of the PARPi effect. Furthermore, the PI3K inhibition or the knockdown of VEGFR2, which blocks the upstream signaling of PI3K decreased the CRY1 expression under DNA damage stresses as bevacizumab did (Fig. 3B and C). These data indicate that inhibition of the VEGFR-PI3K/AKT-CRY1 axis may be sufficient to suppress HR activation by the increase in CRY1. Several clinical trials reported that the combination of PARPi and PI3K/AKT inhibitors showed enhanced efficacy regardless of cancer type and HR status 31–33. These results are consistent with those of our study.
KS-15, an inhibitor of CRY1, increased the sensitivity of olaparib as bevacizumab did (Fig. 3D and E). Interestingly, KS-15 has different effects depending on the cell type. KS-15 exerted an antiproliferative effect and increased sensitivity to doxorubicin in the breast cancer cell line MCF7 but not in the non-transformed mammary epithelial cell line MCF10A 34. KS-15 showed a protective effect in non-neoplastic cells against cisplatin by promoting DNA repair and arresting the cell cycle 35. These results indicate that KS-15 selectively potentiates the therapeutic anticancer effect agents in transformed cells. Future studies are needed to investigate the mechanism of selective potentiation of therapeutic agents by KS-15. Interestingly, the enhancement of growth inhibition combined with olaparib was sustained for a longer period by KS-15 compared with bevacizumab (Supplementary Figure S3). The suppression of CRY1 by bevacizumab was attenuated as time went by X-ray-irradiated cells (data not shown). This may be due to the rhythmic nature of CRY1 expression regulation. Thus, KS-15 may be a better agent to inhibit HR activation in olaparib-treated cells and is a good candidate worth testing in combination with olaparib in clinical trials.
Preclinical studies showed that antiangiogenic agents affect HRR through various mechanisms, indicating synergy between PARPi and antiangiogenic agents. By blocking angiogenesis, antiangiogenic agents induce hypoxia in the microenvironment, and the hypoxic conditions lead to decreased expression of BRCA1/2 and RAD5136–38. Furthermore, VEGFR3 inhibition downregulates BRCA genes, and cediranib directly represses BRCA1/2 and RAD51 gene expression 36,39. In this study, the antiangiogenic agents, bevacizumab or cediranib, enhanced the effect of olaparib in HRP EOC cells through a mechanism that is not associated with hypoxia induced by antiangiogenic agents reported to date.
Although the underlying mechanisms of these combinations are still not fully understood, clinical trials have been conducted to evaluate the combination of PARPi and antiangiogenic agents. In two phase II studies on patients with platinum-sensitive recurrent EOC, the combination of PARPi and antiangiogenic agents significantly improved PFS compared with PARPi alone 12,13. A phase III study on patients with recurrent platinum-sensitive EOC, which compared the combination of cediranib and olaparib or olaparib alone with standard platinum-based chemotherapy, demonstrated that the median PFS was 10.4, 8.2, and 10.3 months for the combination, olaparib alone, and chemotherapy, respectively, and the results were similar in patients without germline BRCA (gBRCA) mutation 40. Another phase II study on heavily pre-treated patients with platinum-resistant recurrent EOC, which compared the combination of olaparib and cediranib or olaparib alone with weekly paclitaxel, demonstrated that the median PFS was 5.7, 3.8, and 3.1 months for the combination, olaparib alone, and weekly paclitaxel, respectively, and no significant difference in PFS was observed between the combination and weekly paclitaxel, and in the subgroup analysis of patients with wild-type gBRCA, the median PFS was 5.8, 3.8, and 2.1 months for the combination, olaparib alone, and weekly paclitaxel, respectively, indicating that the combination therapy showed a promising trend toward improved PFS compared with weekly paclitaxel 41. These results indicate that the combination of PARPi and antiangiogenic agents prolongs PFS compared with PARPi alone, but its efficacy has not been shown to be superior to standard platinum-based chemotherapy regardless of the gBRCA mutation status. Therefore, the combination of PARPi and antiangiogenic agents may be a viable alternative to chemotherapy for patients with recurrent EOC, particularly platinum-resistant recurrent EOC patients with wild-type gBRCA. The combination of PARPi and antiangiogenic agents was first evaluated in the phase III PAOLA-1 study as a maintenance treatment in the first-line setting, which reported a statistically significant improvement in the median PFS for olaparib and bevacizumab compared with placebo and bevacizumab in the overall population, and in the subgroup analysis, a substantial PFS benefit was observed with the combination treatment compared with bevacizumab alone in the HRD population but not in the HRP population 42. The lack of an olaparib alone arm makes it difficult to determine whether the combination has synergistic effects.
In conclusion, VEGF/VEGFR/PI3K signaling enhanced HR activity through the increase in the expression of CRY1. The study findings indicate that the antiangiogenic agents and the CRY1 inhibitors are promising combination partners to overcome primary resistance to PARPi by turning HRP cells into HRD cells. Furthermore, antiangiogenic agents and CRY1 inhibitors may concur with the secondary resistance to PARPi due to the activation of HR. These data provide an important molecular basis for the development of new therapeutic strategies for EOC.