Induced Tamoxifen Resistance is Mediated by Increased Methylation of E-Cadherin in Estrogen Receptor-Expressing Breast Cancer Cells

Estrogen receptor-positive breast cancers are treated with tamoxifen, a drug that competitively inhibits the binding of estrogen to its receptor. Resistance to tamoxifen is a major hurdle in effective management of target breast cancer patient population. A number of dynamic changes within the tumor microenvironment, including the phenomenon of epithelial to mesenchymal transition (EMT), determine the response to endocrine therapy. EMT is marked by silencing or suppression of epithelial marker, E-Cadherin and we found significantly down-regulated E-Cadherin, among other epithelial markers, and a significantly up-regulated mesenchymal marker, Twist, among other mesenchymal markers, in a model system that comprised of tamoxifen sensitive MCF-7 cells and their tamoxifen-resistant counterparts, MCF-7-TAM, developed by chronic and escalating exposure of parental cells to tamoxifen. Further, E-cadherin, but not Twist, was differentially expressed in MCF-7-TAM cells because of differential methylation. Treatment with demethylating agent 5-azacytidine increased the expression of E-cadherin thus verifying a role of methylation in its silencing and, moreover, 5-azacytidine treatment also re-sensitized MCF-7-TAM cells to tamoxifen, as evaluated by assays for viability, apoptosis and migration potential. The 5-azacytidine effects were similar to effects of E-cadherin overexpression in MCF-7-TAM cells. This work describes novel mechanism of E-cadherin downregulation in tamoxifen resistant breast cancer cells. Further studies are needed to exploit this information for betterment of breast cancer therapy.


Material and Methods cell culture and other reagents. MCF-7 (ATCC HTB-22) cells were obtained from Beijing Zhongyuan
Limited (official ATCC distributors; Beijing, China) and MCF-7-TAM were generated by exposing MCF-7 cells to increasing concentrations of tamoxifen over a period of 6 months starting with a dose equivalent to IC-5 of tamoxifen. Cells were maintained in the culture medium and passaged, as per ATCC's instructions. All cells were cultured in 5% CO 2 humidified atmosphere at 37 °C within our laboratory. Tamoxifen and 5-azacytidine were purchased from Sigma Chemical Company (Shanghai, China).
Apoptosis detection. Cells were treated with increasing concentrations of tamoxifen, as indicated for individual figures, and harvested, fixed, and labeled exactly according to the instructions provided with the TiterTACS Kit (R&D Systems, Shanghai, China). Colorimetric readings were recorded on a Shimadzu spectrophotometer (Japan).

Measurement of migration of cancer cells.
We used the Millipore QCM 5 µm Chemotaxis Assay 24-well-colorimetric kit (Shanghai, China) which provides a quick and efficient system for quantitative determination of cell migration after cells are subjected to treatments. The Millipore 5 µm QCM Chemotaxis Assay 24-well-colorimetric is performed in a Migration Chamber, based on the Boyden chamber principle. The 5 µm pore size of this assay′s Boyden chambers is appropriate for studying a subset of fibroblast or cancer cell migration. In our experiments, cells were first subjected to drug treatment, trypsinized, counted and seeded in equal numbers on the provided 24-well plates. Assays were performed exactly as suggested by the Instructions. Colorimetric readings were recorded on a Shimadzu spectrophotometer (Japan).
Quantitative Rt-pcR. RNA was extracted from cells using Trizol (Thermo Fisher Scientific, Shanghai, China) and quantitative RT-PCR was performed as described earlier 19 . cDNA synthesis was carried out with SuperScript III reverse transcription (Thermo Fisher Scientific, Shanghai, China) and the quantitative real-time PCR performed using iTaq Universal SYBR Green Supermix (Bio-Rad, China). The mRNA expression levels of were quantified by measuring the threshold cycle (Ct). e-cadherin over-expression. E-cadherin, cloned in pCMV6, was obtained from Origene (Beijing, China).
Transfection of cloned E-cadherin into MCF-7TR cells was done using lipofectamine 2000 reagent (Thermo Fisher Scientific, Shanghai, China), following the procedure suggested by the company. promoter methylation assays. We studied the methylation of individual genes, using MethylDetector Bisulfite conversion kit (Active Motif, China) which involved heat denaturation in the presence of the bisulfite conversion reagent. All the steps were followed, as described in the Instructions, including PCR. Statistical analysis. Each experiment was performed a minimum of three times and Student t test (unpaired, one-tailed) was employed to compare individual test groups with the respective controls. Significant difference was set at 0.05.

Results
Characterization of resistant cells. MCF-7 cells chronically treated with tamoxifen eventually developed resistance against the drug. This resistance was confirmed by cell viability MTS assay. As shown in Fig. 1A, when MCF-7 cells were treated with increasing concentrations of tamoxifen for a time duration of 48 hours, a dose dependent decrease in cell viability was observed. Even the resistant cells showed a decreasing trend but the cells were significantly more viable than the parental MCF-7 cells at all the three doses -2 μM, 5 μM and 10 μM. Additionally, we prolonged the treatment time and now exposed the cells to tamoxifen for 72 hours and still found the same trend (Fig. 1B). There still was a dose-dependent decrease in viability of parental MCF-7 and tamoxifen resistant MCF-7-TAM cells, but the MCF-7-TAM cells were significantly (p < 0.01) more resistant than the parental cells. Thus, we found through dose-and time-dependent analyses that MCF-7-TAM cells were considerably more resistant to tamoxifen than the parental cells, and were thus an excellent model to study the mechanism of induced tamoxifen resistance.

tamoxifen-resistant cells exhibit reduced apoptosis and increased migration.
Chemotherapeutic drugs mostly function through induction of apoptosis and such action also results in reduced migratory abilities, therefore, we checked the induction of apoptosis by tamoxifen and the resulting affect on migration in our cell line models. As shown in Fig. 2A,B, tamoxifen doses induced a dose-dependent as well as time-dependent increase in apoptosis in both the cells. However, the induction of apoptosis in MCF-7-TAM cells was significantly (p < 0.01) less, compared to the parental MCF-7 cells. Even when tested for migration after treatment with a 10 μM dose for 48 hours, MCF-7-TAM cells still retained significantly (p < 0.01) more migratory potential that the parental MCF-7 cells. Thus, tamoxifen induced significantly more apoptosis in parental MCF-7 cells and also reduced the migratory potential of these cells, compared to the tamoxifen-resistant cells.
www.nature.com/scientificreports www.nature.com/scientificreports/ evaluation of eMt. EMT has been proposed to be a major factor affecting the response of cancer cells to therapy, therefore, we evaluated EMT in our cell line models. We tested three epithelial markers (Collagen IV alpha 1, E-cadherin and MUC1) and three mesenchymal markers (Slug, Snail and Twist). As shown in Fig. 3, we found decreased expression of all the three epithelial markers and the increased expression of all the three mesenchymal markers in MCF-7-TAM cells, as compared to the parental MCF-7 cells. In our results, we found E-cadherin to be the most affected epithelial marker and Twist to be the most affected mesenchymal marker. Thus, we found elevated mesenchymal markers and decreased epithelial markers in tamoxifen-resistant cells which means that EMT was induced in these cells. e-cadherin is epigenetically down-regulated. We evaluated methylation as a possible mechanism for the down-regulation of E-cadherin and the up-regulation of Twist, the two most differentially expressed EMT markers revealed in the last figure. We found significantly increased methylation of E-cadherin promoter (Fig. 4A) in tamoxifen-resistant MCF-7-TAM cells, compared to the parental MCF-7 cells. Further, even though the methylation of Twist was relatively low in MCF-7-TAM cells, compared to MCF-7 cells, the levels were not determined to be significantly different.
effect of e-cadherin overexpression. In our evaluation of EMT markers shown above, we found down-regulation of E-cadherin. Therefore, we overexpressed E-cadherin in tamoxifen-resistant MCF-7-TAM cells and tried to evaluate if restoration of E-cadherin could reverse the action of tamoxifen. As shown in Fig. 5A, we found that over-expression of E-cadherin significantly (p < 0.01) reduced the viability of MCF-7-TAM cells at all the three doses -2 μM, 5 μM and 10 μM, when cells were treated with tamoxifen for 72 hours. At the same time point, cells with over-expressed E-cadherin underwent significantly (p < 0.01) more apoptosis (Fig. 5B). Further, migration potential of cells was significantly decreased (p < 0.01) (Fig. 5C). Thus, we found that over-expression of E-cadherin significantly sensitized the estrogen receptor-positive breast cancer cells to tamoxifen and tamoxifen was found to reduce viability/migration and increase apoptosis in these cells.
Effect of treatment with 5-azacytidine. In the results described above, we observed increased methylation of E-cadherin in tamoxifen resistant cells. Therefore, we wanted to test if this was mechanistically linked to tamoxifen-mediated effects. We used 5-azacytidine as an agent that demethylates DNA. We chronically treated MCF-7-TAM cells with 1 μM 5-azacytidine for 6 cell passages and found that the expression of E-cadherin was

Discussion
Breast cancer remains a major killer among the females [20][21][22] . Treatment of breast cancer patients depends upon factors such as presence of estrogen receptor, progesterone receptor or HER2 overexpression. Presence of one or more of these receptors makes it convenient to enroll the patients for targeted therapy. Endocrine therapy, such as tamoxifen, is effective in breast cancer patients with overexpression of estrogen receptor. Tamoxifen functions via binding to estrogen receptor and competitively inhibiting the binding of estrogen to its receptor. It does not affect the levels of circulating estrogen 3 .  www.nature.com/scientificreports www.nature.com/scientificreports/ In our analyses, we found significantly higher methylation of E-cadherin promoter in tamoxifen-resistant cells. This can clearly be a reason for its observed down-regulation because methylation of gene promoters leads to their silencing 23 . Now, that does not mean that this is the only mechanism and certainly more mechanism(s) cannot be ruled out, particularly when we look at the relative expression of the most differentially expressed mesenchymal marker Twist. Twist was found to be highly elevated in tamoxifen resistant cells, however, even though the methylation of its promoter was slightly reduces, as would be expected in cells undergoing EMT, the degree of reduced methylation was not significant enough. This would mean that up-regulation of Twist is because of some other mechanism that does not necessarily involves methylation. On the other hand, the levels of methylation were found to be significantly more for E-cadherin, giving confidence in pointing out some epigenetic regulation. This was further confirmed by experiment using 5-azacytidine, which is a demethylating agent 24 and, by definition, reduces methylation by removing methyl groups. When we treated cells with 5-azacytidine, not only the suppression of E-cadherin inhibited in tamoxifen-resistant cells, but the resulting attenuation of tamoxifen effects, such as on cellular apoptosis and migration potential, was also observed which clearly gives weightage to the methylation of E-cadherin as a principal mechanism causing its down-regulation in the tamoxifen-resistant cells.
Epigenetic regulation of tumor suppressor genes has been realized for quite some time now [25][26][27] and the functional implications of these epigenetic events are coming to light now. The epigenetic changes within the tumor microenvironment assume further significance 28 where so many cell types mutually interact within a small space with many of them manipulated into serving the needs of cancer cells. Any changes in the expression of factors within the tumor microenvironment are thus bound to have profound effect on cellular signaling and the growth, proliferation, invasion, metastasis and angiogenesis of cancer cells because they happen inside or within a close vicinity of cancer cells. Further, E-cadherin has long been identified as a prominent biomarker with critical role in cancer progression. It has even been called a 'gatekeeper' to breast cancer tumorigenesis 29 and although there exists plethora of information on its up-or down-regulation under different conditions, the mechanisms remain almost unknown. Thus, keeping in view such gap in our information, our present study provides some novel information on the regulation of E-cadherin, particularly in breast cancers that have turned refractory to endocrine treatment. While we provide in vitro evidence in support of a novel mechanism of regulation of E-cadherin in tamoxifen-resistant breast cancer, we understand that further animal studies are needed to validate these findings so that the information could be of use in future clinical trials.