Estrogen Receptor-α36 Is Involved in Pterostilbene-Induced Apoptosis and Anti-Proliferation in In Vitro and In Vivo Breast Cancer

Pterostilbene (trans-3,5-dimethoxy-4′-hudroxystilbene) is an antioxidant primarily found in blueberries. It also inhibits breast cancer regardless of conventional estrogen receptor (ER-α66) status by inducing both caspase-dependent and caspase-independent apoptosis. However, the pterostilbene-induced apoptosis rate in ER-α66-negative breast cancer cells is much higher than that in ER-α66-positive breast cancer cells. ER-α36, a variant of ER-α66, is widely expressed in ER-α66-negative breast cancer, and its high expression mediates the resistance of ER-α66-positive breast cancer patients to tamoxifen therapy. The aim of the present study is to determine the relationship between the antiproliferation activity of pterostilbene and ER-α36 expression in breast cancer cells. Methyl-thiazolyl-tetrazolium (MTT) assay, apoptosis analysis, and an orthotropic xenograft mouse model were used to examine the effects of pterostilbene on breast cancer cells. The expressions of ER-α36 and caspase 3, the activation of ERK and Akt were also studied through RT-PCR, western blot analysis, and immunohistochemical (IHC) staining. ER-α36 knockdown was found to desensitize ER-α66-negative breast cancer cells to pterostilbene treatment both in vitro and in vivo, and high ER-α36 expression promotes pterostilbene-induced apoptosis in breast cancer cells. Western blot analysis data indicate that MAPK/ERK and PI3K/Akt signaling in breast cancer cells with high ER-α36 expression are mediated by ER-α36, and are inhibited by pterostilbene. These results suggest that ER-α36 is a therapeutic target in ER-α36-positive breast cancer, and pterostilbene is an inhibitor that targets ER-α36 in the personalized therapy against ER-α36-positive breast cancer.


Introduction
Breast cancer is the most common malignant tumor in women, and its incidence in the world is persistently rising [1]. It is a hormone-related systemic disease, and endocrine therapy effectively blocks its estrogen receptor (ER-a66, the classic estrogen receptor) pathway to inhibit tumor progression. ER-a66 expression is an important indicator of breast cancer, and previous studies show that patients with ER-a66-negative tumors have shorter disease-free intervals and worse overall survival than patients with ER-a66-positive tumors [2,3]. Tamoxifen (TAM, the selective estrogen receptor modulator) is the most effective drug commonly used for the endocrine therapy of ER-a66-positive breast cancer patients [4]. Patients with ER-negative-tumors are therefore not supposed to respond to TAM therapy. However, clinical studies show that most ER-positive tumors eventually resist TAM therapy despite initial responsiveness to TAM [5,6]. Finding new therapeutic strategies is an urgent topic in breast cancer research nowadays.
Wang et al. have recently identified and cloned a 36-kDa variant of ER-a66, named ER-a36, which is expressed in ERnegative tumor tissues and ER-negative breast cancer cell lines [19,20,21]. ER-a36 and ER-a66 are splicing variants from the same gene, but the previously unidentified promoter of ER-a36 is located in the first intron of ER-a66, indicating that ER-a36 expression is subject to a transcriptional regulation different from ER-a66 [22]. ER-a36 is also predominantly localized in both the plasma membrane and cytoplasm, and it may be associated with both genomic and non-genomic signaling networks [19,23,24]. It therefore is a compound of interest, and possibly is a new therapeutic target in breast cancer.
Alosi et al. showed that the pterostilbene-induced apoptosis in ER-a66-negative breast cancer cells MDA-MB-231 is more obvious than in ER-a66-positive breast cancer cells MCF-7 [14]. ER-a36-mediated mitogenic estrogen signaling in ER-negative breast cancer cells, such as MDA-MB-231, was previously found to lack ER-a66 expression, but highly express ER-a36 [25]. However, the exact mechanism of pterostilbene in ER-a66negative breast cancer cell apoptosis is still under investigation.
In this study, ER-a36 is reported to play an important role in mediating the in vivo and in vitro pterostilbene-induced apoptosis of breast cancer cells. ER-a36 was also found to mediate reducing the activation of the Akt and Erk1/2 pathway using pterostilbene.

Ethical statement
All of the animal experiments described in this study were approved by the Institutional Animal Care and USE Committee (IACUC) at Zhejiang University. All animals were maintained in accordance with the IACUC guideline.

RNA purification and RT-PCR
Total RNA was prepared using the ''TRIzol'' RNA purification reagent. The cDNA was synthesized through the reverse transcription of mRNA using oligo(dT) 20 primer and SuperScript III Reverse Transcriptase (Invitrogen). The RT-PCR analysis of ER-a36 and b-actin was performed using gene specific primers, described before as the following [25]. ER-a36: forward primer: 59-CAAGTGGTTTCCTCGTGTCTAAAG-39; reverse primer: 59-TGTTGAGTGTTGGTTGCCAGG-39; b-actin: forward primer: 59-CCTGGCACCCAGCACAAT-39; reverse primer: 59-GCTGATCCACATCTGCTGGAA-39. The PCR was performed using the PCR Master Mix kit (Beyotime, Nantong, China), according to the manufacturer protocol. PCR products were analyzed by electrophoresis in a 2.5% agarose gel, and were visualized by ethidium bromide staining under UV illumination. The relative mRNA expression was determined from the optical density (OD) ratio of the corresponding mRNA bands, determined using BandScan 5.0 software (Glyko inc., Upper Heyford, UK).

Methyl-thiazolyl-tetrazolium (MTT) assay
Mb231/Si36, Mb231, MCF-7/ER36, and MCF-7 cells were seeded at a density of 3610 3 /well in 96-well plates, and were incubated at 37uC overnight. The cells were then treated with various concentrations of pterostilbene for 72 h. Cell growth was measured by adding 20 ml of 5 mg/ml 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT) to each well, and the plates were incubated at 37uC for 4 h. The supernatant was removed afterwards, and the formazan crystals were dissolved in 200 ml dimethylsulfoxide (DMSO) for 15 min. Absorbance was then measured at 570 nm wavelength using a multiwell spectrophotometer (Bio-Rad). Cell viability is expressed as the percentage of surviving pterostilbene-treated cells vs. control cells (whose viability was considered 100%). The experiments were independently performed in triplicate.

Apoptosis analysis
Apoptosis was detected using an AnnexinV-FITC Apoptosis Detection Kit (Sigma-Aldrich, St. Louis, MO). Mb231/Si36, Mb231, MCF-7/ER36, and MCF-7 cells were treated with 30 mM pterostilbene for 72 h before apoptosis analysis. Cells were harvested by trypsinization, washed twice with PBS, and incubated in 500 mL of binding buffer and 10 mL of Annexin V-FITC at room temperature for 30 min. Subsequently, 5 mL of PI was added, and the cells were incubated for 5 min. Apoptosis data were collected using flow cytometry (BD FACSCanto II, BD Biosciences, San Jose, CA, USA), and were analyzed using the CellQuest software (Becton Dickinson, Franklin Lakes, NJ, USA). The experiments were independently performed three times.

Western blot analysis
Whole cell extracts were obtained by treating cells with RIPA buffer (50 nM Tris pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.5% Sodium deoxycholate, 0.1% SDS, 2 mM EDTA, and 5% Glycerol) containing a protease and phosphatase inhibitors cocktail (Sigma-Aldrich, St. Louis, MO, USA) for 30 min on ice. Appropriate protein extracts of cell lysates were fractionated through SDS-PAGE and were electro-transferred to polyvinyli-dene difluoride (PVDF) membranes (Millipore, Billerica, MA, USA). The membranes were probed with various primary antibodies and HRP-conjugated secondary antibodies, and were visualized using enhanced chemiluminescence (ECL) detection reagents (Millipore, Billerica, MA, USA). The molecular weights of the immunoreactive proteins were estimated based on the PageRuler Prestained Protein ladder (MBI Fermentas, USA). Experiments were repeated at least three times.

Orthotopic xenograft assay
Five to six weeks old female nude mice (01B74-Athymic NCrnu/nu) were ordered from the Experimental Animal Center of Zhejiang Chinese Medical University, and were housed in cages with wood chip beddings in a temperature-controlled room (68uF to 72uF) with a 12-h light-dark cycle and 45% to 55% relative humidity, and were permitted free access to diet and drinking water. Mice were injected with 1.6610 6 Mb231/Si36 and Mb231 cells suspended in 100 ml of L-15 medium containing 50% Matrigel (BD Bioscience) into one their left and right breast pads, respectively. The mice were randomly assigned to experimental and control groups (n = 3) when the tumors reached the size of ,50 mm 3 . The mice were fasted overnight, and were administered with 56 mg/kg pterostilbene or physiological saline (vehicle) by oral gavage once every four days for 3 weeks, as the methods described before [27]. Tumors were measured using calipers once every two days, and the tumor volumes were calculated (Volume (mm 3 ) = p6length 6width 2 /6). At the end of study, tumors were harvested, fixed, and embedded in paraffin. Tumor sections were subjected to standard H&E staining and immunohistochemical (IHC) staining using anti-ER-a36 antibody. The experiment was conducted with replication.

Immunohistochenical analysis
Immunohistochemical (IHC) analysis of ER-a36 expression was performed using 5 mm-thick paraffin-embedded tumor sections. The tissue sections were dewaxed and heated for 20 min with EDTA (pH 9.0) for antigen retrieval, endogenous peroxidase activity was quenched with 3% H 2 O 2 for 10 min at 37uC, and the slides were incubated with normal goat blood serum to block nonspecific binding sites. The slides were rinsed in phosphatebuffered saline, and were incubated for 2 h at 37uC with anti-ER-a36 antibody at 1:100 dilution. The slides were then washed with PBS and were stained with horseradish peroxidase (HRP)conjugated rabbit anti-mouse antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h. The slides were exposed to DAB after washing with PBS. The slides were counterstained using hematoxylin, and were coverslipped with neutrogum.

Statistical analysis
Data are expressed as mean 6 standard error (SE). Two-tailed Student's t-tests were used for analyzing statistical differences between two groups using IBM SPSS software (SPSS, Chicago, IL, USA), and the significance was set at p,0.05. Graphs were generated using GraphPad InStat software program (GraphPad Software, La Jolla, CA, USA).

Results
Breast cancer cells with high ER-a36 expression were sensitive to in vitro pterostilbene treatment RT-PCR was performed to determine ER-a36 gene expression. The ER-a66 and ER-a36 protein expressions in these breast cancer cells were analyzed through Western blotting. The PCR amplicon obtained was the same size as that described before [25].
The ER-a36 gene expression in Mb231/Si36 with ER-a36 knocked down was dramatically decreased compared to parental Mb231 cells. MCF-7/ER36 overexpressed ER-a36 compared to MCF-7 cells. b-actin gene expression was used as the internal control (Fig. 1A). The relative ER-a36 mRNA expressions were determined using the ratio of the OD of mRNA bands compared and the OD of the corresponding b-actin bands (Fig. 1B). Data from RT-PCR are consistent with the protein levels determined through western blot analysis (Fig. 1C). The ER-a66 protein expressions in Mb231 and Mb231/Si36 were undetectable, whereas it decreased in MCF-7/ER36 cells, compared with that in parental MCF-7 cells (Fig. 1C).
Whether the sensitivity of breast cancer cells to pterostilbene dependd on ER-a36 expression was determined. MTT assay was performed and Mb231/Si36 with negative ER-a36 expression were found to exhibit dramatically decreased sensitivity to pterostilbene compared to the parental Mb231 cells. MCF-7/ ER36 cells with ER-a36 overexpression were also more sensitive to pterostilbene than MCF-7 cells (Fig. 1D).
These data indicate that higher ER-a36 expression increases the sensitivity of breast cancer cells to pterostilbene.
Pterostilbene deactivates ER-a36-mediated MAPK/ERK and PI3K/Akt signaling in breast cancer cells    7.5 mM, 15 mM, and 30 mM) for 72 h before Western blot analysis. ERK1/2 and Akt phosphorylation was then found to decrease in a dose-depended manner in ER-a36-positive Mb231 and MCF-7/ER36 with ER-a36 overexperssion, but not in MCF-7 cells (Fig. 3A). The cells were treated with 30 mM pterostilbene for 24 h, 48 h, and 72 h to determine if the inhibition of ERK1/2 and Akt activation is in a time-dependent manner. Phosphor-ERK1/2 and phosphor-Akt expressions were found to be inhibited in a time-depended manner in Mb231 and MCF-7/ ER36 cells, but not in MCF-7 cells (Fig. 3B). The p-ERK1/2 and p-Akt expressions in Mb231/Si36 were nearly undetected with or without pterostilbene treatment. These results indicate that ER-a36 mediates pterostilbene to inhibit MAPK/ERK and PI3K/Akt phosphorylation in breast cancer cells, and ER-a36 knockdown totally deactivates ERK1/2 and Akt phosphorylation in Mb231/ Si36 cells.
Silencing ER-a36 reduces the sensitivity of the xenograft tumors of ER-a36-positive breast cancer to pterostilbene in vivo The in vitro finding that silencing ER-a36 expression reduces the sensitivity of breast cancer cells to pterostilbene led to supposing that ER-a36 is a therapy target for pterostilbene. A xenograft mouse model was further used to study the effect on xenograft tumor growth in nude mice in vivo. Mb231/Si36 and Mb231 cell suspensions were injected into the left and right breast pad, respectively, of each mouse. These mice were randomized into two groups, vehicle control (Veh) and pterostilbene treatment (+Pter), when the tumor size reached approximately 50 mm 3 . The ER-a36-depleted Mb231/Si36 tumors in the left breast pad of the control group mice show significant reduction in growth compared with ER-a36-positive Mb231 tumors in the right breast pad (p, 0.001, Fig. 4A (a)). The tumor weight significantly reduced in the left breast pad injected with ER-a36-depleted Mb231/Si36 cells (p,0.001, Fig. 4B). Pterostibene treatment was observed to significantly reduce the growth rate of the Mb231 tumors compared with physiological saline treatment (vehicle control group) (p,0.05, Fig. 4A(b)). The growth rate of Mb231/Si36 tumors did not significantly reduce after pterostilbene treatment compared to that of the vehicle control group (Fig. 4A(c)). The similar phenomena in tumor weight were also observed (Fig. 4B). Figure 4C shows the representative images of tumors in each group.
The tumors were dissected and processed for standard H&E staining and immunohistochemical staining of ER-a36 protein expression to investigate whether the reduction of tumor growth is associated with the downregulation of ER-a36 expression. A large area of necrosis was found in Mb231 xenograft tumors after pterostilbene treatment. However, the areas of necrosis in Mb231/ Si36 tumors after pterostilbene treatment are small and scattered, which is similar to that of the vehicle (Fig. 4D). Pterostilbene was also found to inhibit ER-a36 expression to some extent in Mb231 xenograft tumors (Fig. 4D). The ER-a36 protein expression of Mb231/Si36 tumors with knockdown ER-a36 is negative (Fig. 4D). These results are consistent with previous in vitro studies, showing that silencing ER-a36 inhibits ER-negative breast cancer proliferation [24,28]. The observations indicate that ER-a36 is a critical therapeutic target for breast cancer, and pterostilbene might be an ER-a36 inhibitor for ER-a36-positive breast cancer therapy.

Discussion
Breast cancer is the most common malignant tumor, and is the major cause of cancer-related death in women in the world [30]. TAM treatment has substantially reduced the recurrence rates and mortality rates of breast cancer patients with ER-positive tumors [31]. However, the initial responsiveness to TAM therapy is limited because most advanced breast tumors recur with acquired resistance [5,6]. Approximately 30% of breast cancers are ERnegative, and do not respond to endocrine therapies. Identifying a novel target is therefore an urgent topic of research. Previous studies report that ER-a36, a variant of conventional ER, is highly expressed in ER-negative tumors, and poorly expressed in ERpositive tumors [25,32]. The acquired TAM resistance could nevertheless be mediated by ER-a36. If patients with ER-a66positive breast cancer tumors have high ER-a36 expressions, they would be more resistant to TAM therapy than those with ER-a66positive/ER-a36-nagetive tumors [21,33]. The published literature suggests that ER-a36 is a critical therapeutic target for personalized therapies for ER-a66-negative/ER-a36-positive breast cancer and ER-a66-positive/ER-a36-high positive tumors with acquired TAM resistance. Finding agents which inhibit ER-a36 is therefore also our central issue. Pterostilbene is an antioxidant primarily found in blueberries, and is also considered to inhibit breast cancer, regardless of conventional estrogen receptor (ER-a66) status, by inducing apoptosis.
Two pairs of breast cancer cells, Mb231 and Mb231/Si36 cells and MCF-7 and MCF-7/ER36 cells, were used in this study. The ER status of Mb231 cells is ER-a66-negative/ER-a36-positive ( Fig. 1A-C), therefore its ER-a36 was knocked down to investigate the role of ER-a36 in pterostilbene-induced antiproliferation on ER-a66-negative/ER-a36-positive breast cancer cells. ER-a36 knockdown was found to desensitize Mb231/Si36 cells to pterostilbene treatment (Fig. 1D). Pterostilbene treatment was also found to drastically reduce the in vivo proliferation capacity of Mb231 tumors with ER-a36-positive expression (Fig. 4A(b)). Clinical studies additionally demonstrated that approximately 40% of patients with ER-a66-positive tumors have high ER-a36 expression, and ER-a36 overexpression is associated with poor disease-free survival and disease-specific survival in patients with ER-a66-positive breast cancer [21]. Moreover, preclinical studies reveal that the pterostilbene-induced apoptosis in Mb231 is more obvious than in MCF-7 [14]. MCF-7/ER36 cells with ER-a36 overexpression were therefore used to investigate whether breast cancer cells with ER-a36 overexpression obtain more benefit from pterostilbene treatment. Overexpressing ER-a36 in MCF-7/ER36 Figure 4. ER-a36 knockdown reduces the sensitivity of the xenograft tumors of ER-a36-positive breast cancer to pterostilbene. Mb231/Si36 and Mb231 cells were injected into the left and right breast pad, respectively of 5 to 6 weeks old nude mice. When the tumors reached about 50 mm 3 size, the mice were randomly assigned into two groups (n = 3): vehicle control (Veh) and pterostilbene treatment (+Pter). a. Tumor volumes were measured once every two days. Red arrows indicate the timed of pterostilbene administration. b. Tumor weights were calculated and are shown as a plot with the median and whiskers from minimum to maximum. c. The representative images of tumors in each group are shown. d. Paraffin-embedded tissue sections of the above tumors were subjected to H&E staining (X 40, left panels) and immunohistochemical (IHC) staining using the antibody against ER-a36 (X 100, right panels). (**p,0.01, NS: no significance). doi:10.1371/journal.pone.0104459.g004 ER-a36 Mediates Pterostilbene-Induced Apoptosis in Breast Cancer PLOS ONE | www.plosone.org cells increased the sensitivity of cells to pterostilbene (Fig. 1D) and enhanced pterostilbene-induced apoptosis ( Fig. 2A and B), as expected.
The MAPK/ERK pathway is a major intracellular communication in breast cancer [34], and PI3K/Akt pathway is also very important in cell proliferation and survival by inhibiting apoptosis [35]. Pterostilbene has been revealed to inhibit PI3K/Akt activation in breast cancer to suppress the heregulin-b1/HER-2modulated invasive and aggressive phenotype of breast cancer cells [16,36,37], and reduced matrix metalloproteinase 9 (MMP) expression, which is an enzyme implicated in micrometastasis [16]. Previous studies have additionally reported that the ERK1/2 tumorigenic pathway in cancer cells could also be inhibited by pterostilbene treatment [17,36,38]. Pterostilbene was found to inhibit ER-a36 expression in ER-a36-positive Mb231 cells and ER-a36-overexpressing MCF-7/ER36 cells (Fig. 2C), and ERK1/ 2 and Akt activation in Mb231 cells and MCF-7/ER36 cells in dose-and time-dependent manners, but not in parental MCF-7 cells. ERK1/2 and Akt phosphorylation were found to be abolished in Mb231/Si36 cells with ER-a36 knockdown (Fig. 3). ER-a36 may therefore be associated with MAPK/ERK and PI3K/Akt pathways, and ER-a36 knockdown directly inhibits ERK1/2 and Akt activation, which is consistent with previous studies which report that both MAPK/ERK and PI3K/Akt signaling activation are regulated by ER-a36 in breast cancer [19,23,25,29,39,40]. Ohshiro et al. also demonstrated that estradiol and anti-estrogenic agent treatment activated ERK1/2 only with the presence of ER-a36, but not ER-a66 [29]. ER-a36 expression was also found to be decreased in tumor tissues after pterostilbene treatment (Fig. 4D). These studies suggested that the possible mechanism of the antiproliferative effect of pterostilbene in ER-a66 negative breast cancer might inhibit MAPK/ERK and PI3K/Akt signal pathways via ER-a36. Further investigations are needed to elucidate the exact mechanisms.
Pterostilbene treatment has been suggested to increase caspase 3 activity and expression, a critical mediator of mitochondrial apoptosis, and pterostilbene also could induce both caspasedependent apoptosis and caspase-independent apoptosis in breast cancer cells [12,14,15]. In our study, caspase 3 expression was also found to be increased after pterostilbene treatment in ER-a36positive Mb231 cells, but not in ER-a36-negative Mb231/Si36 cells (Fig. 2C). The pterostilbene-induced caspase-dependent apoptosis might be dependent on ER-a36. However, Mena et al. recently demonstrated that caspase-independent apoptosis is also induced by pterostilbene, the mechanism of which involves lysosomal membrane permeabilization, and caspase 3 activity failed to increase after pterostilbene treatment in MCF-7 cells [15]. The caspase 3 expression was consistently shown in this tudy to be unchanged under pterostilbene treatment in MCF-7 cells with or without ER-a36 overexpression (Fig. 2C). The pathways of pterostilben-induced-apoptosis in MCF-7 and MCF-7/ER36 cells are therefore different from that in Mb231. Possibly, the underlying mechanism may involve lysosomal membrane permeabilization [15]. This speculation must be further investigated and verified.
Pterostilbene failed to inhibit ER-a36 expression, ERK1/2 and Akt activation in ER-a66-positive MCF-7 cells, though their ER-a36 expression was positive. However, the inhibition could be restored by ER-a36 overexpression (Fig. 3), probably because ER-a36 overexpression competes with ER-a66 for DNA-binding elements in estrogen-responsive genes, inhibiting the estrogendependent and estrogen-independent transactivation activities of ER-a66 [19]. ER-a66 has also been previously reported to suppress ER-a36 promoter activity in an estrogen-independent manner, which could be released by ER-a36 [22]. The ER-a66 protein expression in MCF-7/ER36 cells decreased after transfection with ER-a36 overexpression compared with parental MCF-7 cells (Fig. 1C), which is consistent with a previous study [33]. The proliferation of MCF-7/ER36 cells might thus be dependent on ER-a36-activated MAPK/ERK and PI3K/Akt signaling, which could be blocked by pterostilbene treatment. This warrants future studies on the exact mechanisms of pterostilbene therapy against ER-a66-positive/ER-a36-positive breast cancer.
In summary, strong evidences for a key role of ER-a36 in pterostilbene treatment against ER-a66-negative/ER-a36-positive breast cancer cells both in vitro and in vivo are provided in this study. A possible role for ER-a36 in mediating pterostilbene sensitivity in ER-a66-positive/ER-a36-positive breast cancer cells is demonstrated in vitro. High ER-a36 expression is demonstrated for the first time to promote pterostilbene-induced apoptosis in breast cancer cells and necrosis in xenograft tumors. Furthermore, ER-a36 expression knockdown in Mb231/Si36 cells reduced the proliferation rate of tumors in vivo, together with decreased p-ERK1/2 and p-Akt expression. These finding suggested that ER-a36 is a therapeutic target in ER-a36-positive breast cancer tumors which have resistance to TAM. Pterostilbene is a selective inhibitor that targets ER-a36 in the future therapy against ER-a36-positive breast cancer.