Human breast cancer MCF-7 and MDA-MB-231 cells obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-activated fetal bovine serum (FBS; Biological Industries, Northern Kibbutz Beit Haemek, Israel), 100 U/ml penicillin and 100 µg/ml streptomycin at 37°C with 5% CO₂.

pGC-LV-GV308-PRMT2 (NM_001535) with the wild-type PRMT2 (LV-Tet-on-PRMT2) gene were constructed by the GeneChem Co., Ltd. (Shanghai, China). The pGC-LV-GV308 vector was used as a negative control. The packaging plasmid pHelper 1.0 and pHelper 2.0 were purchased from GeneChem Co. Ltd. We co-transfected the pGC-LV-GV308-PRMT2 vectors with the pHelper 1.0 and pHelper 2.0 packaging plasmid into 293T cells to generate recombinant lentiviruses. Culture medium was collected 72 h post-transfection, and MDA-MB-231 cells were then infected with the aforementioned lentiviruses. A total of 5×105 MDA-MB-231 cells were seeded into a 6-well cell plate and further incubated for 12 h to reach 30% confluency, and then infected with LV-Tet-on-PRMT2 (PRMT2 overexpression group) for 48 h in the presence of 8 µg/ml of Polybrene (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany). A stable cell line with the PRMT2-3Flag was obtained after infection with Lv-tet-on-PRMT2 cells which were selected by puromycin for 2 weeks. Western blot analysis was performed to verify the expression of PRMT2-3Flag induced by 5 µg/ml of Dox in the infected MDA-MB-231 cells.

The lentivirus-based PRMT2 shRNA expression plasmids, pYr-Lvsh-PRMT2, were purchased from Yingrun Biotechnology Co. (Changsha, China). The pYr-Lvsh vector was used as a negative control. The viral particles were produced and purified by cotransfecting pYr-Lvsh-PRMT2 and the lentivirus packaging plasmids (pVSVG, pLP1, pLP2) into 293FT cells. The stable clones with PRMT2 knockdown generated from the MCF-7 cells were selected in culture medium containing 0.5 µg/ml puromycine. The selected stable clones were verified by western blotting using the negative control cells as control.

Total cell lysates were lysed on ice for 30 min. Soluble proteins (40 µg) were probed with the anti-PRMT2 antibody (1:500; cat. no. 3667; Abcam, Cambridge, MA, USA), anti-ER-α36 antibody (1:1,000; cat. no. CY1109; Cell Applications, San Diego, CA, USA) and anti-p-AKT (cat. no. 4060), anti-AKT (cat. no. 4691), anti-p-ERK1/2 (cat. no. 4370) and anti-ERK1/2 (cat. no. 4695) (1:1,000; Cell Signaling Technology, Inc., Danvers MA, USA). Loading variations were normalized against β-actin, which was identified by the anti-β-actin antibody (1:1,000; cat. no. 4970; Cell Signaling Technology).

Cells were harvested, washed with phosphate-buffered saline (PBS), and fixed with 66% ethanol overnight at −20°C. Cells were centrifuged and washed with PBS, and then stained with propidium iodide (PI; BD Biosciences, Franklin Lakes, NJ, USA) in PBS solution for 1 h in the dark. Cell cycle distribution was evaluated by use of a FACSCalibur flow cytometer equipped with CellQuestPro software (BD Biosciences) (27).

Apoptosis was measured with use of an Annexin V-fluoroisothiocyanate (Annexin V-FITC) or Annexin V-PE apoptosis detection kit according to the manufacturer's instructions (BD Biosciences) and analyzed with use of a FACSCalibur flow cytometer and BD CellQuest Pro software (BD Biosciences) as previously described (28). Briefly, cells were harvested and washed, and incubated in binding buffer with Annexin V-FTIC/PI or Annexin V-PE/7AAD for 15 min at room temperature. The cells were washed and resuspended in binding buffer before flow cytometric analysis.

Transfections were performed using Lipofectamine 2000 transfection reagent (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions. Plasmid constructs expressing PRMT2 fused to an N-terminal GFP tag and pcDNA3.1-Myc-His-ERα36 were generated by Yingrun Biotechnologies Inc. (Hunan, China), and used to transiently co-transfect the MCF-7 or MDA-MB-231 cells grown on glass coverslips. At 48 h post-transfection, the cells were fixed in paraformaldehyde, permeabilized with °Triton X-100 and incubated with Cy3-conjugated secondary antibody. The cells were then stained with 4,6-diamidino-2-phenylindole (DAPI) and viewed under a Zeiss LSM 510 confocal microscope (Carl Zeiss, Oberkochem, Germany); images were acquired from typical cells using a ×63 oil-immersion lens.

GST and GST fusion proteins were expressed in Escherichia coli BL-21 cells by induction with a final concentration of 0.8 mM isopropyl-b-D-thiogalactopyranoside. Cells were lysed by sonication in 10 ml of 1X PBS (NaCl/Pi) supplemented with complete protease inhibitor tablets (Roche Applied Science, Basel, Switzerland). GST fusion proteins with ER-α36 and ER-α66 were purified using glutathione-agarose beads (Sigma-Aldrich; Merck KGaA). The recombinant human His-tag-PRMT2 protein (obtained from Yingrun Biotechnologies Inc.) were mixed with 10 mg of GST derivatives bound to glutathione-agarose beads in 0.5 ml of binding buffer (50 mM Tris/HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.3 mM dithiothreitol, 0.1% NP-40 and protease inhibitor tablets from Roche Applied Science). The binding reaction was performed at 4°C for 3 h and the beads were subsequently washed 4 times with the washing buffer (the same as the binding buffer), 30 min each time. The beads were eluted by boiling in SDS sample buffer and analyzed by SDS/PAGE, and PRMT2, ER-α36 and ER-α66 were analyzed by immunoblotting.

MCF-7 cells or MDA-MB-231 cells were transfected with the indicated plasmids. At 48 h, cells were lysed (50 mM Tris at pH 8.0, 500 mM NaCl, 0.5% NP-40, 1 mM dithiothreitol and protease inhibitor tablets from Roche Applied Science). Five hundred nanograms of lysate were precleared with 50 µl protein A-Sepharose beads (Sigma-Aldrich; Merck KGaA) for 2 h at 4°C. The PRMT2 antibody (Abcam) or ER-α36 antibody (Cell Applications) was then added and incubated overnight at 4°C. One hundred microliters of protein A agarose were then added to the antibody/lysate mixture for another 2 h at 4°C, and the beads were pelleted and washed thrice with lysis buffer. Bound proteins were eluted in SDS sample buffer, subjected to SDS/PAGE and analyzed using an ER-α36 antibody (Cell Applications) or PRMT2 antibody (Abcam).

A tissue microarray (BR1921; US Biomax, Inc., Rockville, MD, USA) consisting of 160 breast cancer samples was used. These samples were histologically interpretable and were analyzed for the correlation between PRMT2 and ER-α36. Immunohistochemical staining was performed as detailed in our previous study (24). The tissues were blocked with 1% bovine serum albumin (Roche Diagnostics, Basel, Switzerland) at room temperature for 1 h. Antibodies against PRMT2 (1:50; Abcam) and ER-α36 (1:50; Cell Applications) were used. The correlation analysis was performed using SPSS software (version 18.0; SPSS, Inc., Chicago, IL, USA). Spearman's rank correlation coefficients were generated to determine the degree of the correlation.

All experiments were performed ≥3 times, and the results are expressed as the mean ± standard deviation unless otherwise stated. GraphPad Prism software (version 5.0; GraphPad Software, Inc., La Jolla, CA, USA) was used for statistical analysis. Comparisons between two groups were performed using the two-tailed Student's t-test. Comparisons among multiple groups were performed using one-way analysis of variance with post hoc intergroup comparisons using the Tukey test. P<0.05 was considered to indicate a statistically significant difference.

In order to clarify the antitumor effect of anti-estrogen drug TAM in different breast cancer cell lines, flow cytometry after dual staining of Annexin V-FITC/PI was used to evaluate the cell death rate of MCF-7 and MDA-MB-231 cells after drug treatment. The results showed that the cell death of MCF-7 and MDA-MB-231 cells were both induced by TAM in a dose-dependent manner. However, the cell death rate of MDA-MB-231 cells was significantly lower than that of MCF-7 cells for the same drug concentration (Fig. 1), indicating that MCF-7 cells were more sensitive to TAM but MDA-MB-231 cells were relatively resistant to TAM.

To explore the mechanism of TAM resistance, we first examined the expression of resistance-related potential proteins using western blot analysis in MCF-7 and MDA-MB-231 cells. The results revealed no obvious change in expression of PRMT2 and ER-α36 in the MCF-7 cells, but a substantial decrease in PRMT2 and an increase in ER-α36 level after treatment with TAM in the MDA-MB-231 cells, and both proteins changed in a time- and dose-dependent manner (Fig. 2A). ER-α36 binds to estrogen and quickly activates estrogen non-genomic signaling pathways, such as PI3K/Akt signaling and MAPK/ERK signaling, and regulate anti-estrogen drug resistance (13–15). Therefore, we also evaluated the expression of phosphorylated Akt, ERK1/2 and total Akt, ERK1/2. The results demonstrated that the levels of phosphorylated Akt and ERK1/2 were decreased in the MCF-7 cells but were increased in the MDA-MB-231 cells after exposure to TAM, whereas no effects on total proteins were noted in the MCF-7 and MDA-MB-231 cells (Fig. 2A).

To confirm the relationship of PRMT2 and ER-α36, a tetracycline (doxycycline hyclate; Dox)-inducible lentiviral system was established to overexpress PRMT2, and the pGC-LV-GV308 vector was used as a negative control. Stable cell lines with the PRMT2-3Flag were obtained after selection with puromycin for 2 weeks (MDA-MB-231-PRMT2 cells). Western blot analysis verified the overexpression of PRMT2-3Flag induced by Dox in the infected MDA-MB-231 cells. Moreover, as shown in Fig. 2B, with treatment of 5 µg/ml of Dox, MDA-MB-231 cells carrying lentivirus PRMT2 expression exhibited markedly decreased ER-α36 and phosphorylated Akt, phosphorylated ERK1/2 compared to negative control cells. In turn, the lentivirus-based shRNA expression plasmids, pYr-Lvsh-PRMT2, was used to knockdown PRMT2. The pYr-Lvsh vector was used as a negative control. The stable clones of PRMT2 knockdown generated from MCF-7 cells were selected in the culture medium containing puromycine (MCF-7-shPRMT2 cells). The selected stable clones were verified by immunoblotting using the negative control cells as control. Knockdown of PRMT2 in MCF-7 cells increased the expression of ER-α36 and phosphorylated Akt and phosphorylated ERK1/2 compared to the negative control cells (Fig. 2B). The results in these two cell lines demonstrated that PRMT2 inhibited ER-α36 and its estrogen non-genomic signaling pathways, PI3K/Akt and MAPK/ERK.

To further study the interaction of PRMT2 and ER-α36, we first examined the subcellular localization of PRMT2 and ER-α36 by confocal microscopy. As shown in Fig. 3A, PRMT2 appeared to be largely localized to the nucleus excluding the nucleoli and a weak fluorescence was detected in the cytosol as described previously (29–31), whereas ER-α36 was predominantly localized to the cell membrane and cytoplasm. In addition, the fluorescence signals related to ER-α36 and PRMT2 overlapped, indicating co-localization of PRMT2 and ER-α36 inside MCF-7 cells. Similar results were observed in the MDA-MB-231 cells (Fig. 3A). PRMT2 is able to bind to ER-α66 both in the presence and absence of estrogen but can enhance ER-α66 activity only with estrogen as previously reported (25). To understand the molecular mechanisms of the ER-α36 dependence of PRMT2, we performed a protein-protein interaction study in vitro and inside the cells. As shown in Fig. 3B, PRMT2 displayed similar binding affinity to ER-α36 and ER-α66 in vitro. This binding activity of PRMT2 to ER-α36 was also confirmed inside MCF-7 and MDA-MB-231 cells. PRMT2 was co-immunoprecipitated by the PRMT2 antibody, and ER-α36 could be captured by the ER-α36 antibody, and vice versa (Fig. 3C). These results showed that PRMT2 directly associates with ER-α36.

Given that PRMT2 was significantly decreased after TAM treatment in MDA-MB-231 cells and was closely related to ER-α36, we reasoned that PRMT2 may be a critical mediator for regulating the response of breast cancer cells to TAM. To test this hypothesis, we evaluated the impact of silencing PRMT2 on the sensitivity of MCF-7 cells to TAM and the impact of overexpression of PRMT2 on the resistance of MDA-MB-231 cells to TAM. The stable MCF-7-PRMT2 cells exhibited sufficient upregulation of PRMT2 and the MDA-MB-231-shPRMT2 cells displayed marked downregulation of PRMT2 by western blot analysis (Fig. 2B). PRMT2 knockdown in MCF-7 cells obviously attenuated TAM-induced cell death when compared with the MCF-7-NC cells, as analyzed by flow cytometry (Fig. 4A). In turn, PRMT2 overexpression in MDA-MB-231 cells notably increased TAM-induced cell death when compared with the MDA-MB-231-NC cells (Fig. 4B). Moreover, MDA-MB-231-PRMT2 cells exhibited increased G1 arrest and decreased S phase (to reflect cell proliferation status) when compared with the MDA-MB-231-NC cells, whether or not with TAM treatment (Fig. 4C). These results suggest that PRMT2 is a critical mediator for regulating the response of breast cancer cells to TAM, and PRMT2 could reverse TAM resistance in breast cancer cells.

To further identify the association between PRMT2 and ER-α36 expression in breast cancer, a tissue microarray (BR1921), consisting of 160 breast cancer cases, was used. No significant correlation (r=−0.024; P=0.765; Fig. 5) was identified between the expression of PRMT2 and ER-α36 upon analysis of the human breast cancer tissue microarray, regardless of breast cancer types. The data suggest that there was no significant correlation between the expression of PRMT2 and ER-α36 in breast cancer tissues, which was not consistent with the results observed in the cells.