Targeting oestrogen receptor subtypes in ovarian cancer treatment

Abstract

Ovarian cancers are known to express oestrogen receptors. For many years, Tamoxifen, a selective oestrogen receptor modulator, has been used in the treatment of ovarian cancers. It has a favourable side- effect profile, but the response rate is modest. It is mostly used in a palliative setting, and the overall clinical effectiveness is unclear. We reviewed our 15-year experience in using Tamoxifen in ovarian cancer patients in our unit. Tamoxifen was used mostly after failing two or more lines of chemotherapy. As expected, we found that the response rate was only about 10%, but 46% of patients had static disease, indicating that Tamoxifen could keep the disease under control, even if it could not eliminate it. This encouraging finding prompted us to explore ways to optimise the response to hormonal treatment. Oestrogen receptor ( ER ) has two subtypes, ERα and ERβ, with different tissue distribution and may have opposing actions. Our previous study demonstrated that both subtypes were present in ovarian cancer tissues. Here, we further demonstrated that ERβ expression level, but not ERα, correlated with patients’ survival, suggesting a particular clinical significance of ERβ. In a cohort of 175 patients, by categorising ERβ expression into four quartiles, higher ERβ was found to be significantly associated with better disease-free and overall survival. Alternative splicing leads to the presence of different ERβ variants ( ERβ1-5 ). We further characterised the subcellular distribution of ERβ splice variants by immunostaining in 106 ovarian cancer tissue blocks. We found that three splice variants ( ERβ1, ERβ2 and ERβ5 ) were present in the nucleus in all the samples and over 90% of samples also had ERβ1 and ERβ2 in the cytoplasm. Furthermore, ERβ5 expression correlated with poorer overall survival in multivariate analysis. With this finding, we focused on ERβ5 and found that overexpression of ERβ5 in ovarian cancer cell lines increased cell migration, invasion and proliferation. We explored the underlying molecular mechanisms. We found that ERβ5 induced FAK and c-Src activation and inhibition of FAK by FAK inhibitors blocked ERβ5 induced cell migration, invasion and proliferation. Functional studies with specific ER subtype agonists and antagonists showed that very selective ERβ agonist reduced cell proliferation while ERβ antagonist increased cell proliferation. The opposite effects were seen with selective ERα agonists and antagonists. In addition, we found a synergistic effect with ERα antagonist with ERβ agonist, both in vitro and in vivo. These results suggest that there may be a role for selective ER subtype agonists/ antagonists in the treatment of ovarian cancer. We further explored the effects of Genistein and Daidzein. These are two naturally occurring phytoestrogens found in soy, with much higher ERβ affinity. We demonstrated that they also significantly inhibited ovarian cancer cell migration, invasion, proliferation, as well as induced cell cycle arrest and apoptosis. Overall, our findings suggest that targeting ERβ subtype may be a therapeutic option in ovarian cancer and should be further explored.