miR-200 affects tamoxifen resistance in breast cancer cells through regulation of MYB

Resistance to tamoxifen is a major clinical challenge. Research in recent years has identified epigenetic changes as mediated by dysregulated miRNAs that can possibly play a role in resistance to tamoxifen in breast cancer patients expressing estrogen receptor (ER). We report here elevated levels of EMT markers (vimentin and ZEB1/2) and reduced levels of EMT-regulating miR-200 (miR-200b and miR-200c) in ER-positive breast cancer cells, MCF-7, that were resistant to tamoxifen, in contrast with the naïve parental MCF-7 cells that were sensitive to tamoxifen. Further, we established regulation of c-MYB by miR-200 in our experimental model. C-MYB was up-regulated in tamoxifen resistant cells and its silencing significantly decreased resistance to tamoxifen and the EMT markers. Forced over-expression of miR-200b/c reduced c-MYB whereas reduced expression of miR-200b/c resulted in increased c-MYB We further confirmed the results in other ER-positive breast cancer cells T47D cells where forced over-expression of c-MYB resulted in induction of EMT and significantly increased resistance to tamoxifen. Thus, we identify a novel mechanism of tamoxifen resistance in breast tumor microenvironment that involves miR-200-MYB signaling.


Materials and Methods
Cell culture and other reagents. MCF-7 and T47D cells were obtained from Beijing Zhongyuan Limited (Beijing, China). Tamoxifen resistant MCF-7 cells, TAM-MCF7 were developed in our own laboratory by dose escalation method wherein we treated MCF-7 cells with increasing doses of tamoxifen for over 4 months. Cells were cultured in 5% CO 2 humidified atmosphere at 37 °C. Tamoxifen was obtained from Sigma Chemical Company (Shanghai, China). Real time RT-PCR Primers for c-MYB and GAPDH were same as reported elsewhere 14 . MTS assay. MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-t etrazolium) assay was employed to assess cell proliferation. Cell were first seeded overnight before the start of any experiment. Adequate replicates were taken and absorbance was measured on a spectrophotometer.
Quantitative RT-PCR. We used mirVana RNA isolation kit to extract RNA. Realtime RT-PCR was performed as per the conditions described earlier 14 on a BioRad Instrument. Reverse transcription was done using M-MuLV Reverse Transcriptase and Random Primer Mix (New England Biolabs Inc, USA). miRNA reverse transcription was performed using TaqMan miRNA reverse transcription reagents (Thermo Fisher Scientific, China) as per the instructions. Each sample was run in triplicate in 3 independent experiments. c-Myb transfection and silencing. Transfections were performed by transfecting cells with pcD-NA3-c-myb containing the full-length cDNA for human c-myb, using Lipofectamine 3000 (Thermo Fisher Scientific, Shanghai, China). Cells were transfected with c-MYB-targeting siRNA or scrambled controls (Dharmacon, China), following the exact conditions as specified by manufacturer.

Statistical analyses.
Statistical analyses were done with Student's t test. The level of significance was set at p-values less than <0.05 in all cases.

Results
Mesenchymal markers are up while miR-200s are down in tamoxifen resistant cells. With the intent to possibly find an EMT basis of tamoxifen resistance, and involvement of miR-200 family, we first evaluated the EMT markers such as vimentin, ZEB1 and ZEB2 for their mRNA expression, using real time RT-PCR. We used parental MCF-7 cells and compared the expression of markers in tamoxifen resistant TAM-MCF7 cells. As shown in Fig. 1A, the levels of all three markers, vimentin, ZEB1 and ZEB2 were elevated in TAM-MCF7 cells, as compared to the parental MCF-7 cells. Thus, we could verify that EMT is induced in our model of tamoxifen resistance.
We also evaluated the IC-50 values of tamoxifen in resistant cells vs. the parental ones and performed a dose-escalation experiment wherein cells were exposed to increasing concentrations of tamoxifen for different durations of time, and IC-50 of tamoxifen was calculated for the normal vs resistant cells. As shown in Table 1, whereas the IC-50 value decreased for tamoxifen, as would be expected for increasing time duration, the IC-50  www.nature.com/scientificreports www.nature.com/scientificreports/ values for tamoxifen resistant TAM-MCF7 cells were always higher than the parental MCF-7 cells at all the time points tested. In fact, the fold-changes in IC-50 values of TAM-MCF7 over IC-50 values of MCF-7 cells increased with the passage of time, which verified a robust resistance to tamoxifen in TAM-MCF7 cells (Table 1).
With     www.nature.com/scientificreports www.nature.com/scientificreports/ over-expression led to significantly increased resistance against tamoxifen, compared to the vector transfected T47D cells. This is evident from the values provided in Table 2. The IC-50 value of c-MYB overexpressing cells increased to 19.14 μM from a value of 12.81 μM in control cells. Further, all the mesenchymal markers (vimentin, ZEB1 and ZEB2) were significantly over-expressed in T47D cells transfected with c-MYB, compared to the control cells (Fig. 5B).

Discussion
ER-positive breast cancers are blessed to have a targeted therapy but as seen with most targeted therapies, continued administration over a time period results in development of resistance. While human samples are precious and often difficult to obtain, cell line models are acceptable alternatives. In this study, we generated tamoxifen resistant breast cancer cells, MCF-7 and tried to understand the mechanisms of resistance, with the focus on EMT and the regulation by miRNAs, particularly those belonging to miR-200 family. The focus on miR-200s in this study was not without a logic. This family has a history of being implicated in resistance against multiple therapies in different cancers [15][16][17] ; for example, EGFR targeting therapy in bladder cancer 18 and lung cancer 19 , sorafenib and imatinib resistance in renal cancer 20 , oxaliplatin 21 and 5-FU 22 resistance in colon cancer, paclitaxel and carboplatin resistance in ovarian cancer 23 and nintedanib 24 and paclitaxel 25 resistance in lung cancer. Despite the reports in several different cancers, there is very little evidence supporting such role of this miR family in tamoxifen resistance of breast cancers. Thus, our study fills that gap in the literature.
Having validated the down-regulation of miR-200s in our model system thus verifying a role of miR-200s in tamoxifen resistance, our primary task was to find a novel target gene that is regulated by miR-200s. We turned to bioinformatics which predicted c-MYB to be a potential target gene. c-MYB was predicted to be targeted by both miR-200b and miR-200c, and not by miR-200a. This is interesting because we found miR-200b and miR-200c to be significantly down-regulated in our model, and not the miR-200a. This, in itself, was a great evidence but we went ahead and validated the prediction our model system. Since miR-200b/c were downregulated, we expected c-MYB to be overexpressed. This is because miRNAs negatively regulate their target genes. We found this to be case as c-MYB was significantly upregulated in the tamoxifen resistant cells, compared to the parental cells.
Further to this evaluation of c-MYB in paired cell line model, we tested a direct regulation of c-MYB by miR-200b/c. Among the two cell lines in the paired cell lines, miR-200s were significantly downregulated in TAM-MCF-7 cells and therefore we tried to upregulate miR-200s in these cells by transfections with pre-miRs. When this was accomplished, we saw a significant downregulation of c-MYB as would be expected because of a negative regulatory relationship between miRNAs and its target. Furthermore, in the parental MCF-7 cells, where miR-200s were relatively at a higher endogenous levels, we tried to down-regulate miR-200s. This was accomplished by transfections with anti-miRs and resulted in significant upregulation of c-MYB as the negative regulation by miR-200s was now relieved. Thus, through multiple experiments, we clearly established a novel regulatory relationship between miR-200s and c-MYB.
Another important finding of this study is the role of miR-200-c-MYB in possibly regulating EMT which leads to tamoxifen resistance. The functional connection between c-MYB and EMT has been explored in several previous studies [26][27][28][29] and moreover, the non-coding RNAs connection with EMT that involves c-MYB has also studied 30 . In an interesting earlier study 31 , it was reported that miR-200s such as miR-200b and miR-200c downregulate c-MYB in ER-positive cells and that EMT-inducing TGF-β stabilizes c-MYB. These results support our observation and may provide intricate details of miR-200 regulation of c-MYB. In our study, we did not find any effect of c-MYB on miR-200s (results not shown). We only report the regulation of c-MYB by miR-200s. However, there is evidence in the literature suggesting an activation of miR-200s by c-MYB 32 which suggests that the relationship between miR-200s, c-MYB and EMT is complex and maybe even cell type or the cancer microenvironment dependent. More studies are needed to fully understand the mechanism.

Data availability
All data from this study is reported within the article.