Transcriptome analysis reveals that Müllerian inhibiting substance regulates signaling pathways that contribute to endometrial carcinogenesis

Müllerian inhibiting substance (MIS) has been shown to inhibit growth of a number of tumors in vitro and/or in vivo, but the downstream pathways which it regulates are not fully understood. In the present study we show that MIS type II receptor was highly expressed in AN3CA cells, a cell line derived from human endometrial cancer cell in which MIS-treatment caused a reduction of cell viability, and induced cellular apoptosis and genes involved cell cycle arrest. To understand the genome-wide effects of MIS on gene regulation, we performed serial gene expression analyses from 0 to 96 h at 24 h intervals after treating AN3CA cells with MIS. Transcriptomic analysis of molecular changes induced by MIS identified 2,688 differentially expressed genes that were significantly up- or down-regulated during the 96 h study period. When the 2,688 differentially expressed genes were mapped to known biological processes, Wnt-, cancer-, proteolysis-, cytoskeleton-, cell cycle-, apoptosis-, and MAPK-signaling pathways emerged as the functions most significantly changed by MIS in AN3CA cells. Furthermore, western blot analysis validated that protein expression of cell cycle inhibitory genes, apoptotic protease activating factor-1 (APAF-1), β-catenin-interacting protein (ICAT), Rb related protein 130 (p130), and inhibitor of disheveled Dvl and Axin complex (IDAX), were gradually increased over the time of the study, whereas downstream cell cycle activating genes, cyclin-dependent kinase 2 (CDK2) and phospho-c-Jun were downregulated in MIS-treated AN3CA cells. These transcriptome analyses support previous observations that MIS functions as a tumor suppressor, potentially by regulating signaling pathways that could contribute to endometrial carcinogenesis, and indicating that MIS should be considered as a potential treatment for endometrial cancer.


Introduction
Endometrial cancer is the most common invasive gynecologic cancer in the Western world and the fourth most common cancer in women after breast, lung and colorectal cancers (1). Endometrial cancer has been divided into two different types on the basis of distinct histology and clinical outcomes. Most endometrial cancers are type I (approximately 75%) estrogendependent adenocarcinomas with endometrioid morphology. Type I endometrial cancers are low-grade, early stage cancers that usually develop in perimenopausal women and coexist with, or are preceded by, complex and atypical endometrial hyperplasia, and have an excellent prognosis (2). In contrast, type II endometrial cancers are of poorly differentiated endometrioid and serous histologies that occur mainly at an older age and are associated with a lower 5-year survival rate (3). The definite role of estrogen in the development of most endometrial cancers is established; any factor that increases exposure to unopposed estrogen increases the risk for endometrial cancer. Endometrial cancers are characterized by various genetic alterations. Type I endometrial cancer harbors altered PI3K/AKT/mTOR signaling pathway constituents (4). Type II endometrial cancer have loss of function mutations in p53 (5) and gain of function mutations in epidermal growth factor receptor 2 (HER-2) (6).
Most gynecologic tumors originate from Müllerian ducts, which develop into the Fallopian tubes, uterus, cervix, proximal vagina, and from the surface epithelium of the ovary. Müllerian inhibiting substance (MIS), also known as anti-Müllerian hormone (AMH), has long been known for causing the regression of the embryonic Müllerian duct. MIS has been shown to inhibit tumor growth in vitro and in vivo, Transcriptome analysis reveals that Müllerian inhibiting substance regulates signaling pathways that contribute to endometrial carcinogenesis but its downstream regulated genes have not been fully elucidated. MIS initiates its effect by binding to the MIS type I and type II receptors (7). Several studies showed that MIS exerts antiproliferative and apoptotic effects on gynecologic malignancies such as cervical and ovarian cancers (8,9). In ovarian cancer, the MIS caused a suppressive effect on Cyclin Dependent Kinases (CDKs) and the E2F transcription/ dimerization partner (E2F/DP1) complex, with transcriptional enhancement of certain histone deacetylases, which act as co-suppressors of the E2F/DP complex (9). In cervical cancer, MIS suppresses cell division by inducing an increase in the expression of CDK inhibitors such as p16 INK4a which results in a decrease in activity of the CDK complex leading to inhibition of E2F activity (8). Since endometrial cancer is also a tumor of Müllerian duct origin and prior studies (10) have reported antitumor properties of MIS against endometrial carcinoma cell lines, we sought to understand the molecular drivers of growth inhibition or apoptosis in endometrial cancer. To do so, we undertook transcriptome scans of AN3CA cells treated with MIS. Analysis of the differentially expressed genes that were further validated by western analysis indicated that MIS could elicit its tumor suppressive effects on endometrial cancer cells through control of components of pathways on which endometrial cancer progression is dependent.

Materials and methods
Endometrial cancer cell culture. Statistical analysis. Results are presented as mean ± SD of replicate analyses and are either representative of, or inclusive of, at least three independent experiments. Statistical comparisons between two experimental groups were performed using Student's t-test (paired) whilst multiple group comparisons were performed using analysis of variance (ANOvA). Data were regarded as statistically significant when P<0.05.
MTT results are presented as percentage of control, which was calculated using the following equation: (mean absorbance of treated cells/mean absorbance of control cells) x100. Data are expressed as mean ± SD from nine independent experiments, each with replicates of three. A P-value <0.05 was considered statistically significant when compared with corresponding vehicle control cells.
Cell cycle distributions after exposure of AN3CA endometrial cancer cells to MIS are presented as histograms of the mean ± SD from three independent experiments. Annexin v analysis was done to evaluate apoptosis. Quadrant rectangular dot grams from a representative of three independent experiments were analyzed.

Results
Treatment results of Müllerian inhibiting substance in growth retardation and cellular apoptosis of AN3CA endometrial cancer cells. It has been well established that proliferation of ovarian, cervical and breast cancer cells are suppressed by MIS treatment, and many of these studies mentioned the potential of MIS as a treatment of MIS-receptor expressing tumors. Therefore, before initiating experiments on the regulatory roles of MIS in endometrial cancer cells, it was necessary to confirm expression of the MIS receptor in AN3CA cells. To this end, we performed immunohistochemical staining with an anti-MIS type II receptor (MISR II) antibody on AN3CA cells which detected strong expression on their cell surface (Fig. 1A).
Next, to investigate the antitumor effects of MIS on endometrial cancer cells, AN3CA cells were treated with MIS at the indicated concentration for 0, 24, 28, 72 and 96 h, and MTT assays were performed. As shown in Fig. 1b, MIS treatment on AN3CA cells resulted in a 30% reduction of tumor cell growth (n=3, P<0.05, using Student's t-test) resulting from cell cycle arrest or apoptosis. Cell cycle analysis of PI-stained cells using flow cytometry at 72 h after MIS treatment (Fig. 1C) detected a slight 4.85% increase in cells in the G1 phase, with a concomitant decrease in S phase and G2/M phase by 3.88 and 4.61%, respectively, compared to controls (non-treatment) (n=3, P<0.05, using Student's t-test). Flow cytometric analysis at 72 h of Annexin v stained AN3CA cells (Fig. 1D), however, showed a statistically significant increase (n=3, P<0.0001, using Student's t-test) in early cellular apoptosis of 14.37% (from 2.31±0.18% at 0 h to 16.68±1.21% at 72 h) and in late cellular apoptotic and necrotic cells of 6.49% (from 0.04±0.03% at 0 h to 6.53±0.75% at 72 h) (Fig. 1D). These results suggest that MIS exerts its antitumor effects on endometrial cancer cells by causing significant cellular apoptosis.

Characterization of molecular and functional signatures after MIS treatment of endometrial cancer cells.
To identify a molecular signature induced by MIS, serial gene expression analysis was conducted. AN3CA cells were continuously treated with MIS for up to 96 h, and the RNAs were harvested for microarray analysis at multiple time points from 6 h. In Fig. 2A, 11,470 genetic elements, which passed the minimum selection and filtering criteria, are shown in unsupervised hierarchical clustering. Mathematical comparisons of genes between 0 h and the 6-96 h time points were performed and 2,688 genes showing at least 1.3-fold changes (Fig. 2B) were visualized as a heat map where red indicates that expression levels of genetic elements that are higher after MIS treatment and green indicates that expression levels of genetic elements that are lower than the mean value of no treatment.
To derive insight into the molecular pathways in which these 2,668 genes participate, we used the pathway tool in the Kyoto Encyclopedia of Genes and the Genomes (KEGG) pathway database (http://www.genome.jp/kegg/) which maps genes to known pathways and provides a summary of the biological processes affected. based on this database analysis, we identified 20 pathways containing at least 10 genetic elements from the 2,668 differentially expressed genes. AN3CA cells treated with MIS fall into cancer-related signaling pathways involving Wnt, cell cycle, MAPK, p53 and apoptosis pathways (Table I). From these, genes associated with Wnt, cell cycle regulation and apoptosis pathways were depicted as heat maps (Fig. 3A). In Wnt signaling pathways, CTNNbIP1 and CXXC4 genes were upregulated and JUN was downregulated compared to non-treatment. CDK2 gene was downregulated in cell cycle pathways. In apoptosis pathways, the APAF1 gene was up regulated while caspase-3 was slightly downregulated (Fig. 3A).
Differentially expressed genes in AN3CA cells, that were examined by western analysis include phospho-c-Jun, ICAT, IDAX, CDK2, p130, p107, APAF-1, caspase-3 and PARP, which confirmed upregulation of ICAT and IDAX and downregulation of phospho-c-Jun in Wnt signaling pathways, downregulation of CDK2 and slight upregulation of p130 and p107 in cell cycle pathways, and upregulation of APAF-1 and cPARP and downregulation of full length caspase-3 in the apoptosis pathway (Fig. 3b).

Discussion
In previous studies, MIS has been shown to inhibit the growth of ovarian, cervical and endometrial cancer cell lines. Treatment with recombinant human MIS induced up regulation of both p107 and p130 in non-HPv-related cervical cancer cell lines (13) and inhibited the growth of the ovarian cancer cell line through induction of the CDK inhibitor p16 at the protein level and downregulation of p130 and upregulation of E2F1 when Rb levels were not detectable (14). In endometrial cancer cell lines (AN3CA) lacking p16 and pRb, MIS-induced apoptosis was associated with an increase in p107 and p130 and a decrease in E2F1 expression at 72 h (10). Thus, gene regulation by MIS appears to be specific to tumor type, but in these studies involves regulation of cell cycle pathways.
The cornerstone of curative therapy for endometrial cancer is surgical treatment, including complete hysterectomy, adnexectomy and appropriate surgical staging in patients considered at high risk. The type of surgery and postoperative therapy depends on the stage and other clinicopathological risk factors and survival is strongly dependent on surgical stage. Recent advances in the understanding of the molecular and genetic basis of endometrial cancer have led to the development of targeted therapies that inhibit angiogenesis and the cellular signaling pathways involved in cell growth and proliferation. Trastuzumab, a human EGFR type II (HER2)related inhibitor that affects signal transduction, is currently a standard treatment for HER2-positive breast cancer. Certain serous adenocarcinomatous endometrial cancers which are characterized by overexpression of HER2 could be helped by trastuzumab (15). mTOR inhibition with temsirolimus has encouraging single-agent activity in endometrial cancer which has been shown to be more effective in chemotherapynaive patients than in chemotherapy-treated patients and is independent of PTEN status (16). Another mTOR inhibitor, ridaforolimus, has antitumor activity and acceptable tolerability in advanced endometrial cancer patients (17). Given the potential of MIS as a therapeutic for endometrial cancer, particularly the more aggressive Type II, it is important to understand the genes regulated by MIS.
To demonstrate the molecular mechanism by which MIS might inhibit the growth of an endometrial carcinoma cell line, we first evaluated the expression of MISRII on the cell surface of the endometrial carcinoma cell line, AN3CA. After confirmation of the MISRII expression on the AN3CA by immunohistochemistry, we examined the inhibitory effect of MIS on the growth of the endometrial cell line and showed 7-29% reduction in cell survival, an increase of cellular apoptosis and a small change in cell cycle parameters. Microarray and western blot analyses confirmed changes in pathways related to cell cycle, apoptosis and Wnt signaling.
For example, ICAT, the protein encoded by CTNNbIP1, a negative regulator of the Wnt signaling pathway (18) binds CTNNb1 (β-catenin) and prevents its intracellular inter action with TCF (T-cell transcription factor) family members. IDAX, the protein encoded by the CXXC4 gene also functions as a negative regulator of the Wnt signaling pathway by directly binding to the PDZ domain of Disheveled (DvL) in the nucleus (19). C-Jun transcribes a protein that is required for progression through the G1 phase of the cell cycle as c-Jun null cells show increased G1 arrest (20). C-Jun is also a target gene of β-catenin, which, if decreased, could result in tumor suppression, indicating that MIS could possibly result in tumor suppression, through regulation of Wnt signaling.
APAF1 encodes a cytoplasmic protein that forms one of the central hubs in the apoptosis regulatory network, and its activation of PARP induces necrotic and programmed cell death, thus supporting the most robust changes seen in these studies (Fig. 1D, i.e., apoptosis). In previous studies, MIS has been documented to increase both p107 and p130 in AN3CA cells (10), which was also correlated with inhibition of proliferation in cervical cancer cell lines (13). Slight upregulation of p107 and p130 was shown in the study. However, other genes associated with cell cycle demonstrated significant differences There were significant pathways including more than 10 genes by one pathway in outliers of 2,688 genes which showed expression more than 1.3-fold increase or decrease.
after MIS treatment; in cell cycle pathways, CDK2 is essential for the G1/S transition and reduction of CDK2 levels correlates with cell cycle arrest. In summary, these findings indicate that MIS regulates the proliferation of tumor cells through modulation of apoptosis and cell cycle pathways, and by causing blockade of the Wnt/ β-catenin signaling pathways. The findings support the hypothesis that MIS was able to function as a potential therapeutic for endometrial cancer by targeting these molecular pathways.