In the present study, we investigated the expression of the nuclear co-repressor NCOR2 in EOC and its potential role in estrogen-dependent physiological and pathophysiological processes promoted by the G-protein coupled estrogen receptor GPER through non-genomic signaling. Co-expression of both GPER and NCOR2 showed significant changes in clinical outcome in EOC patients, highlighting not only the molecular, but also the prognostic significance of the transcriptional regulation altered by NCOR2. In accordance to differences in the pathogenesis of EOC (Kurman and Shih, 2011), the NCOR2 expression varied significantly accordingly to histological subtypes and was independent from other clinical-pathological data. Highest NCOR2 immunoreactivity was observed in serous carcinomas and was dependent on tumor grading. Since subtype-specific impact of NCOR2 expression on overall survival by Kaplan-Meier estimates was not statistically significant, the present study further evaluated general molecular mechanisms of ovarian cancer biology rather than focusing on subtype specific analyses. Isolated evaluation of high NCOR2 co-repressor as well as high GPER expressing tumors could not reveal significant impact on OS, whereas evaluation of both factors together indicated that the tumors exhibiting the combination of both, high NCOR2 and high GPER expression, seem to have significantly improved OS with a median improvement of 54.2 months. Consequently, these results suggest an association with a possible regulating role of the nuclear co-repressor NCOR2 in GPER mediated signaling, which merits further evaluation.
As a transcription co-regulator of metabolic processes, NCOR2 is involved in the development of various cancer entities by unbalancing pro- and anti-inflammatory signaling pathways (Mottis et al., 2013). For primary ovarian cancers, its expression was found in up to 70% of tumor tissues (Havrilesky et al., 2001). Structural investigations detected NCOR2 appearing in a large co-regulatory complex that assembles multiple and context-dependent partner proteins (Oberoi et al., 2011). The NCOR2 co-repressor complex disrupts transcription of various target genes through chromatin remodeling, leading to altered cellular processes and to malignancy through induced transcriptional arrest (Battaglia et al. 2010). Some of the complex´ proteins represent substantial core components as they account for the repressive function (Oberoi et al., 2011). Each of these components was shown to get involved in carcinogenesis. Transducing β-like related 1 (TBLR1), for instance, is taking an ambiguous role, having either co-repressing or co-activating function depending on the affected cell type and interacting hormone receptor (Wu et al., 2016). It appeared to be significantly upregulated in ovarian cancer cells and served as predictor for the clinical outcome of EOC patients (Havrilesky et al., 2001; Wu et al., 2016; Ma and Yu, 2017). Showing a significantly higher expression in both the nucleus and cytoplasm of breast and ovarian cancer cells compared to benign tissues, it activates cell proliferation and accelerates migration and invasion through ER-independent and ER-dependent pathways (Wu et al., 2016). However, in EOC, granulosa cell tumors of the ovary and healthy ovarian tissue, no significant correlation of NCOR1 and NCOR2 with ERβ was found, consistent with our results of NCOR2 expression and their correlation to ERα/ERβ (Hussein-Fikret and Fuller, 2005).
Depending on the hormonal signaling context, the NCOR2 complex gets involved into the transcriptional activity of different nuclear receptors, including steroid hormone receptors such as estrogen, androgen and progesterone receptors and is incorporated in different regulation processes (Jepsen and Rosenfeld, 2002; Wong et al., 2014). Various studies affirm that an appropriate suppression of NCOR2 expression is a key aspect for proper cell signaling. In contrast, aberrant function can promote cancer and disease progression. In breast cancer cells, NCOR2 was proposed as a candidate to initiate cancer cell growth by modifying the transcriptional activity of ERα or directly influencing the ERα expression (Ciriello et al., 2013; Dobrzycka, 2003; Kurebayashi et al., 2000). In prostate cancer, loss of the repressive function of NCOR2 altered the AR response to ligands and contributed to cancer development (Godoy et al., 2012). By interacting with other co-repressing proteins, NCOR2 was identified as a fundamental modulator of the estrogen receptor in breast cancer patients treated with tamoxifen as adjuvant antihormonal therapy. In these, aberrant NCOR2 appeared to modify the hormone receptor response to tamoxifen while low NCOR2 levels predicted drug resistance against tamoxifen (Gong et al., 2018). For serous ovarian cancer patients, NCOR2 was identified as a possible predictive biomarker of chemotherapy response among seven other genes in a gene array study, correlating with resistance to chemotherapy (Fekete et al., 2020). By epigenetic regulation, co-repressors play a context-dependent role in biological processes and deregulated function might unbalance homeostasis and therefore accelerate malignant degeneration (Mottis et al., 2013).
Estrogen-dependent signaling regulated by GPER was found in several tumor entities (Chevalier et al., 2012; Fujiwara et al., 2012; He et al., 2009; Prossnitz and Barton, 2011; Revankar et al., 2005). Thereby, the role of GPER expression in ovarian cancer tumorigenesis either as promoting or as suppressing factor remains still subject to current research: For either ERα negative or ERα positive ovarian cancer cells, GPER- mediated signaling pathways were found to promote ovarian cancer cell proliferation (Liu et al., 2014; Albanito et al., 2008). In contrast, GPER expression might inhibit cell growth and therefore serve as tumor suppressor and as a positive prognostic factor for disease-free survival (Ignatov et al., 2013). While comparable nuclear and cytoplasmatic occurrence was described, the specific intracellular localization and signaling pathway has not been identified yet. On a differentiated analysis of a previous study, only nuclear GPER expression was identified as an independent negative prognostic factor (Zhu et al., 2018). The study assumed that nuclear rather than cytoplasmatic occurrence drives carcinogenesis and therefore leads to impaired outcome (Zhu et al., 2018). Further investigation with a discrimination of nuclear and cytoplasmatic expression of GPER in EOC might serve as a future approach to explain these contradictory findings.
However and in accordance with our results, effects on overall survival are not depending on GPER expression alone. The present results indicate a possible regulating role of the NCOR2 nuclear co-repressor complex on GPER-regulated pathways in EOC. High expression of both immunophenotypes was shown to be significantly correlated with a positive impact on OS in ovarian cancer patients possibly indicating a role to serve as a reliable prognostic marker. Approximately 34% of drugs approved by the American Food and Drug Administration (FDA) exert their therapeutic effects by targeting G protein-coupled receptors (Hauser et al., 2018). Our new insights to the previously relatively unknown role of GPER in EOC carcinogenesis make GPER a potential novel target for treatment strategies. Furthermore and more importantly, drugs with epigenetic targets such as Small inhibiting molecules, Histone deacetylase inhibitors and demethylation agents might be able to modulate the function of the NCOR2 complex and might therefore alter response to chemotherapy and affect the clinical outcome of EOC patients (Graham et al., 2009).
Based on the results of the present study, we hypothesize that NCOR2 mediated modulation of nuclear GPER-mediated signaling pathways is responsible for the improved prognosis. Some limitations of our study should be taken into consideration for the interpretation and evaluation of the present data. Since EOC comprise a heterogenous group of several histological subtypes that differ in biological, clinical and prognostic behavior, our study is limited by a relatively small sample subset that allows rather general than subtype-specific conclusions regarding the ovarian cancer biology. As a retrospective dataset, additional data on patient characteristics are lacking for a deeper exploration considering factors such as menopausal status, estrogen levels and exposure to hormonal replacement therapy. Moreover, further studies and experimental methods will be necessary to elucidate not only the static receptor expression by immunohistochemistry, but also mechanisms regarding the regulation of gene transcription, modulation of receptor dynamics and ultimately its exact biological function. Based on the presented results, these consecutive experiments are already in preparation by our group and will be prospectively followed. Accordingly, the current study might serve as a potential starting point to further explore the complex molecular implications of the epigenetic regulator NCOR2 and its components on processes involved in ovarian cancer development, such as GPER mediated signaling pathways. A comprehensive understanding of the mechanisms of chromatin regulation and transcription activity in ovarian cancer development enables future patient specific therapeutic approaches.