Luminal Breast Cancer Cell Lines Overexpressing ZNF703 Are Resistant to Tamoxifen through Activation of Akt/mTOR Signaling

Background Selective estrogen receptor modulators, such as tamoxifen, play a pivotal role in the treatment of luminal-type breast cancer. However, in clinical applications, nearly half of breast cancer patients are insensitive to tamoxifen, a small number of whom have early recurrence or disease progression when receiving tamoxifen. The underlying mechanism of this resistance has not been determined. ZNF703 is a novel oncogene in the 15% of breast cancers that harbor 8p12 amplifications. Therefore, the goal of our study was to explore the role of ZNF703 in tamoxifen resistance. Methodology/Principal Findings We used immunohistochemistry techniques to examine ZNF703 expression in stage I-III primary breast cancer specimens and found a positive expression rate of 91.3%. All patients were divided into either high or low ZNF703 expression groups. We found that high ZNF703 expression mainly occurred in ER+ and PR+ breast cancers. Furthermore, 4-hydroxytamoxifen had different modes of action in breast cancer cell lines with high or low ZNF703 expression. ZNF703 overexpression in MCF-7 breast cancer cells activated the Akt/mTOR signaling pathway, downregulated ERα, and reduced the antitumor effect of tamoxifen. Low-dose tamoxifen did not suppress, but rather, stimulated the growth of cells overexpressing ZNF703. ZNF703 knockdown in MDA-MB-134 and HCC1500 luminal B-type breast cancer cell lines by siRNA significantly decreased survival rates when cells were treated with tamoxifen. Furthermore, targeting ZNF703 with a mTOR inhibitor increased the inhibitory effects of tamoxifen in ZNF703-overexpressing cells. Conclusion/Significance Our study suggests that ZNF703 expression levels may predict tamoxifen sensitivity. Tamoxifen should be administered with caution to those patients bearing tumors with ZNF703 overexpression. However, large clinical trials and prospective clinical studies are needed to verify these results.


Introduction
The molecular typing of breast cancer provides a basis for the prognosis and treatment of breast cancer. There are four major molecular subtypes of breast cancers: triple negative/basal-like, human epidermal growth receptor (HER)-2 positive, Luminal A, and Luminal B [1]. Most breast cancers are luminal tumors. Luminal A and B tumors tend to be estrogen receptor-positive (ER+) and/or progesterone receptor-positive (PR+). A variety of endocrine therapies act through different mechanisms to antagonize the growth of tumors stimulated by estrogen. Selective estrogen receptor modulators (SERM), such as tamoxifen, can antagonize ERa activity, and have been used in breast cancer therapy. Tamoxifen is very effective for the treatment of luminal breast cancer; 5 years of tamoxifen therapy can reduce the risk of recurrence and death by 41% and 33%, respectively [2].
However, of those patients who receive adjuvant tamoxifen therapy for 5 years, 8% have early recurrence (within less than 2.5 years), and another 8% have recurrence within 2.5 years to 5 years [3]. In addition, one-third of women treated for 5 years will eventually relapse within 15 years, as their tumors often become endocrine-resistant [4]. In luminal metastatic breast cancer, the objective response rate for tamoxifen treatment is only 30%, and 20% show stable disease. In regard to neo-adjuvant tamoxifen therapy, the clinical objective response rate is 30% to 60%, with 30% to 50% having stable disease, and about 3% having disease progression during treatment [5,6].
Tamoxifen resistance may be primary or acquired. Insensitive molecular subtypes include luminal B, HER2+, and triple negative [7,8]. With the exception of those subtypes, predictors of tamoxifen resistance are poorly defined, making it difficult to identify patients who are less likely to benefit from tamoxifen treatment. Some clinical and pathological factors that can predict early recurrence include cancers that are lymph node-positive or low ER-expressing as well as a multigene score called EndoPredict Index [3,[9][10][11][12][13]. The failure of tamoxifen to prevent many early relapses highlights the need for more effective therapies to improve clinical outcomes.
The ER signaling pathway plays a key role in the development of estrogen dominant breast cancer. However, this pathway is not the only survival pathway of tumors; thus, when the ER signaling pathway is blocked, the escape pathways function [14]. Importantly, these proliferative pathways can cross-talk with the ER pathway and regulate ER to affect endocrine therapy [15][16][17][18]. Activation of these pathways leads to the formation of ERindependent tumors. These pathways can be activated by amplification or overexpression of oncogenes, or by loss of function of downstream signaling molecules [19]. Studies have revealed that phosphorylation of protein kinase B (PKB/Akt) or mammalian target of rapamycin (mTOR) can directly alter sensitivity to tamoxifen, leading to tamoxifen resistance [20,21]. However, the factors that induce ERa loss and activate Akt signaling remain unknown. Many clinical trials have been designed that block these escape pathways to increase the survival benefit in luminal breast cancer patients who have endocrine resistance.
Zinc finger 703 (ZNF703) is an oncogenic transcription factor that regulates many genes involved in multiple aspects of the cancer phenotype, including proliferation, increased self-renewal, and invasion [22][23][24]. ZNF703 was recently identified as a novel breast cancer oncogene in the 15% of breast cancers that harbor 8p12 amplifications, amplified second only to the well-known oncogenes, ERBB2 and cyclin D1 (CCND1) [25,26]. Nevertheless, ZNF703 is still largely unknown in the breast cancer field; for example, it remains to be determined how the expression or function of the encoded protein can be regulated, and the major downstream effectors of its oncogenic functions have not been identified. In the present study, we hypothesized that ZNF703 might contribute to endocrine resistance. We assessed ZNF703 expression in clinical breast cancer specimens, and used breast cancer cell lines to investigate the role of and underlying mechanisms of ZNF703 in endocrine therapy.

ZNF703 is Highly Expressed in Luminal Breast Tumors
We examined ZNF703 expression in breast cancer tissue microarray. In 127 breast cancer patients, the positive expression rate of ZNF703 protein was 91.3%. ZNF703 was mainly localized in the nucleus, although it was also expressed in the cytoplasm of some cells. All patients were divided into high and low ZNF703 expression groups ( Figure 1A, B). In the high expression group, 49.3% of patients were ER+, whereas only 23.1% of the low expression group was ER+. This difference was statistically significant ( Figure 1C, P = 0.003). Similarly, 58.7% of patients were PR+ in the high expression group compared to only 26.9% in the low expression group. This difference was also statistically significant ( Figure 1C, P = 0.001). In addition, high ZNF703 expression was associated with an intermediate grade (Table S1, P = 0.04).

Inhibitory Effects of TAM in Breast Cancer Cell Lines with High or Low ZNF703 Expression
Given the involvement of ZNF703 in luminal-type breast cancer, we assessed ZNF703 mRNA expression levels in breast cancer cell lines by real time and reverse transcription polymerase chain reaction (RT-PCR) (Figure 2A). The MCF-7 luminal A-type breast cancer cell line had low ZNF703 expression levels, while MDA-MB-134 and HCC1500, luminal B-type breast cancer cell lines, had high expression levels, similar to observations in previous studies [22,23]. The normal mammary epithelial cell line MCF-10A had almost no expression of ZNF703. Next, we performed experiments to determine if ZNF703 impacted 4hydroxytamoxifen (TAM) sensitivity in breast cancer cell lines. The results showed that the inhibitory effects of TAM differed between breast cancer cell lines. In MCF-7 cells, increasing concentrations of TAM caused inhibitory rates to gradually increase ( Figure 2B). At a concentration of 50 mM, the inhibitory rate reached 100%. In contrast, in the hormone receptor negative BT-549 cell line, the inhibitory rates were close to 0 within the range of 1 to 1610 23 mM TAM, indicating that BT-549 cells were insensitive to tamoxifen ( Figure 2C). Notably, although HCC1500 and MDA-MB-134 cell lines were much more sensitive to tamoxifen than BT-549, they exhibited particular modes of tamoxifen resistance. In HCC1500 cells, although the inhibitory rate reached 100% at 10 mM TAM, inhibitory rates of less than 30% were achieved in the range of 1 to 1610 23 mM ( Figure 2D). After treatment with 100 mM TAM, the inhibitory rate in MDA-MB-134 cells was still less than 100%. Notably, 10 22 mM and 10 23 mM tamoxifen promoted the growth of MDA-MB-134 cells ( Figure 2E).

ZNF703 Overexpression Activates the Akt Signaling Pathway and Reduces the Antitumor Effects of TAM
To determine the effect of ZNF703 on endocrine therapy, we overexpressed ZNF703 in MCF-7 cells (MCF-7-ZNF703 cells). These cells were treated with several concentrations of TAM, and the expression levels of several signaling proteins were measured. We detected considerable upregulation of phosphorylated Akt (p-Akt; P,0.01) and down-regulation of both ERa (P,0.05) and Ecadherin (P,0.05) in MCF-7-ZNF703 cells; total Akt protein levels remained unchanged (P.0.05). After treatment with increasing concentrations of TAM, p-Akt (Ser473) and ERa levels decreased in a dose-dependent manner in MCF-7-vector cells, whereas E-cadherin was upregulated ( Figure 3A). In MCF-7-ZNF703 cells, increasing concentrations of TAM abrogated these effects, causing down-regulation of p-Akt (P,0.01) and upregulation of ERa (P,0.001).
We next examined the effect of TAM on the survival of MCF-7-ZNF703 cells, and found that MCF-7-ZNF703 had significantly reduced inhibitory rates after treatment with various concentrations of TAM compared to MCF-7-vector cells ( Figure 3B, P,0.01), indicating the occurrence of resistance. Surprisingly, lowdose TAM (10 23 to 10 21 mM) resulted in a 20% to 40% increase in the growth of MCF-7-ZNF703 cells, which is consistent with what we previously observed in MDA-MB-134 cells ( Figure 2E). We performed a colony formation assay to verify the reduced inhibitory rate of MCF-7-ZNF703 cells upon treatment with TAM compared to MCF-7-vector cells ( Figure 3C). We also observed that low-dose TAM (10 23 to 10 21 mM) stimulated the growth of MCF-7-ZNF703 cells in a time-dependent manner ( Figure 4A), as 6 days of low-dose TAM treatment resulted in a 2-fold increase in cell growth compared to control cells.

siRNA-mediated Depletion of ZNF703 Counteracts the Effects of Anti-TAM
To determine whether targeting ZNF703 could reverse tamoxifen resistance in ZNF703-overexpressing cell lines, small interfering RNAs (siRNAs) were used to knockdown ZNF703. siRNA depletion of ZNF703 in MDA-MB-134 cells significantly increased the inhibitory effect of TAM compared to control cells ( Figure 4B, P,0.05). The stimulatory effect on cell growth by lowdose TAM was also eliminated. There was no difference between MDA-MB-134 siRNA-1 and MDA-MB-134 siRNA-2-treated cells (P.0.05).
Similar results were obtained in the HCC1500 cell line; siRNA depletion of ZNF703 revealed significantly increased inhibitory effects of TAM compared to control cells ( Figure 4C, P,0.05). Treatment with siRNA-2 led to better knockdown efficiency of ZNF703 compared to treatment with siRNA-1; therefore, slightly lower survival rates were observed in HCC1500 cells treated with siRNA-2 (P,0.05).

Rapamycin Increases the Effect of TAM on Cell Viability in ZNF703-overexpressing Cell Lines
To further elucidate whether the Akt signaling pathway was responsible for tamoxifen resistance in ZNF703-overexpressing cell lines, we treated cells with the mTOR inhibitor, rapamycin. Rapamycin significantly blocked p-mTOR (S2448) in MCF-7vector cells in a dose dependent manner. p-mTOR (S2448) was upregulated in MCF-7-ZNF703 cells, and treatment with rapamycin significantly inhibited its expression ( Figure 5A). In addition, rapamycin inhibited phosphorylation of eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP1; T37/ 46 and T70), and increased phosphorylation of Akt in MCF-7 cells ( Figure 5B), most likely due to negative feedback mechanisms [27].
To determine the effect of blocking mTOR on cell viability, we treated MCF-7-ZNF703 cells with either TAM alone or TAM in combination with rapamycin at a concentration selected based on its predetermined half maximal inhibitory concentration (IC50). TAM significantly decreased the number of viable cells when combined with rapamycin in MCF-7-ZNF703 cells ( Figure 5C).

Discussion
The ER signaling pathway plays a key role in the development of breast cancer. Tamoxifen has been widely used in the treatment of luminal breast cancer for over 30 years. However, nearly half of breast cancer patients are insensitive to tamoxifen therapy, some of whom relapse or have progressive disease when receiving tamoxifen, and thus have tamoxifen resistance. Although previous studies have reported various mechanisms of tamoxifen resistance, there are many ''unknowns'' that need further clarification.
ZNF703 is the most significant luminal potential oncogene [28]. ZNF703 is overexpressed in some breast cancer cell lines, but has expression was mainly present in ER+ and PR+ breast cancers. Left: proportion of patients with ER+ or ER2 breast cancer in the high or low ZNF703 expression groups, respectively. Right: proportion of patients with PR+ or PR2 breast cancer in the high or low ZNF703 expression groups, respectively. Also see Table S1. doi:10.1371/journal.pone.0072053.g001 low expression in normal mammary epithelial cells. Overexpression of ZNF703 increases genome instability and contributes to tumor aggressiveness. Moreover, one single-nucleotide polymorphism near ZNF703 is significantly associated with breast size [29]. IHC was used to evaluate ZNF703 expression in the breast cancer specimens, and high ZNF703 was mainly found in luminal breast cancers. In addition, overexpression of ZNF703 correlated with an intermediate grade, suggesting a moderate prognosis in breast cancer [30].
The modes of tamoxifen action were distinct between breast cancer cell lines with high or low ZNF703 expression. ZNF703 overexpression reduced the antitumor effect of tamoxifen. Both MDA-MB-134 and HCC1500 are luminal cell lines [31][32][33], with high ZNF703 levels. siRNA depletion of ZNF703 combined with tamoxifen increased its antitumor effects. Compared to luminal A- type, luminal B-type breast tumors are not sensitive to tamoxifen [7,8]. Several studies use MDA-MB-134 and HCC1500 as representatives of luminal B-type breast cancer cell lines [22,23,34]. Our findings provide a basis for potential cellular mechanisms of tamoxifen resistance in luminal B breast cancer.
Preclinical experiments have shown that the phosphatidylinositol 3-kinase (PI3K)/Akt/mTOR pathway is a potential escape pathway that functions after the ER signaling pathway is blocked [20,21]. In this regard, studies have shown that ER+ breast cancer patients with excessive activation of p-Akt (Ser473) who receive tamoxifen treatment have an increased recurrence rate and a reduced overall survival rate [35]. Levels of p-Akt (S473), p-mTOR (s2448), p-ER (S167), and p-ER (S305) expression can predict tamoxifen efficacy [36]. Early studies have shown that the mTOR signaling pathway is essential for the growth of estrogendependent luminal breast cancer cells. mTOR inhibitors with endocrine therapy synergizes to induce cell death [37]. Bolero-2 Phase III clinical trials showed that endocrine therapy combined with Everolimus can improve progression-free survival in luminal breast cancer patients with disease progression [20,38,39]. Overexpression of ZNF703 in MCF-7 cells can activate the Akt/mTOR signaling pathway. A mTOR inhibitor combined with siRNA knockdown of ZNF703 increased the antitumor effect of tamoxifen. Our results reveal that ZNF703 is a new molecule that can activate this pathway. Previous studies have found that mutations or loss of ERa led to primary tamoxifen resistance. However, the factors that induce ERa loss remain unclear. Varying degrees of ERa loss can be detected in 15% to 20% of tamoxifen-resistant breast tumors [40]. ZNF703 overexpression decreased ERa levels, leading to less tamoxifen targets, and high-dose TAM resulted in the recovery of ERa. The phenomenon that ZNF703 repressed ERa expression is not contradictory with our results showing that ZNF703 was highly expressed in luminal breast cancer. However, further studies are needed to verify if ZNF703 is negatively correlated with ER expression in breast cancer patients.
The ER signaling pathway is closely related to the growth factor receptor pathway. First, the ER signaling pathway can be modulated by membrane receptor tyrosine kinases (RTKs), such as epidermal growth factor receptor (EGFR), HER-2, and insulinlike growth factor-1 receptor (IGF1-R) [15]. These RTKs can eventually phosphorylate ER and its co-activators or co-repressors, thereby affecting their function [16,41,42]. Second, estrogen increases the expression of growth factors, such as transforming growth factor (TGF)-a and IGF1, and activates downstream signaling [43]. In turn, activated PI3K/Akt and p42/44 mitogenactivated protein kinase (MAPK) pathways can reduce the  expression of ER and PR [44][45][46][47][48][49]. In addition, ER which functions in non-nuclear transcription, can alter the amount of proteins regulated by growth factor receptors [17,18]. Excessive activation of growth factor receptor pathways can increase the number of non-nuclear ERs and non-transcriptional activity, while non-nuclear ER activity can also be activated by estrogen or tamoxifen, resulting in tamoxifen resistance. Overexpression of RTKs causes tamoxifen to increase the function of non-nuclear ER and activate RTK downstream pathways, leading to tamoxifen resistance. We found that fibroblast growth factor receptor 4 was upregulated in MCF-7-ZNF703 cells (data not shown). Interestingly, mimicking endogenous estrogen levels, lowdose tamoxifen stimulated the growth of tumor cells upon ZNF703 overexpression. Therefore, ZNF703 is likely to increase the expression of RTKs as well as non-nuclear ER activity, which further activates the PI3K/Akt pathway, leading to cell proliferation and antagonism of tamoxifen.
Many regulatory proteins related to the cell cycle have been found. For example, cyclin-dependent kinase inhibitors (CDKI) negatively regulate the cell cycle, and RB is a negative regulator of the transition from the G1 to S phase. Studies have shown that low p27 [50][51][52], CCND1 expression abnormalities [53,54], and RB inactivation are associated with tamoxifen resistance [55]. We found that overexpression of ZNF703 caused down-regulation of p27 and RB1, as well as upregulation of CCND1 (data not shown). These results suggest that ZNF703 can participate in tamoxifen resistance through intrinsic cell cycle mechanisms.
In summary, our data demonstrates the important role of ZNF703 as a novel molecule in tamoxifen resistance induced by activation of the Akt/mTOR signaling pathway and downregulation of ERa. Our findings provide a potential mechanism of early-or late-recurring tamoxifen-resistant luminal breast cancers. It can be predicted that targeting ZNF703 and its downstream signaling pathway can increase the efficacy of endocrine therapy in patients with ZNF703-overexpressing luminal breast cancer. In addition, tamoxifen should be used with caution in patients with ZNF703 overexpression.

Ethics Statement
Written informed consent was obtained from all patients. The study protocol was designed according to the principles of the Helsinki guidelines, and approved by the Institutional Review Board of Ruijin hospital (China).

Cell Culture
All breast cancer cell lines were purchased from American Type Culture Collection (Manassas, VA, USA) and cultured in complete growth medium as described by the company. Cell lines were monitored by performing polymerase chain reaction (PCR) using the Mycoplasma Detection Kit (Invitrogen, Carlsbad, CA, USA) on the Applied Biosystems Real-Time PCR System. Briefly, MDA-MB-134 cells were grown in L-15 medium (Gibco, Carlsbad, CA, USA) with 20% fetal bovine serum (FBS; Gibco); HCC1500 and BT-549 cells were cultured in RPMI 1640 medium (Gibco) supplemented with 2 mM L-glutamine with 10% FBS; MCF-7 cells were cultured in Minimum Essential Media (Gibco) supplemented with 10% FBS.

Immunoblot Analysis
The indicated cell lines were grown in 6-well plates, treated as indicated, and lysed in lysis buffer (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% SDS, 1% Triton X-100, 1% sodium deoxycholate) containing protease and phosphatase inhibitors (Beyotime, Shanghai, China). Lysates were subjected to SDS-PAGE and western blotting with the indicated antibodies as previously described [62]. Briefly, cell lysates were resolved by SDS-PAGE and electrotransferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). After blocking at room temperature in 5% nonfat milk in TBST (20 mM Tris-HCl (pH 7.4), 150 mM NaCl, and 0.1% (w/v) Tween 20) buffer for 1 h, the membranes were incubated overnight at 4uC with the appropriate primary antibodies. The next day, the membranes were washed three times with TBST, and then incubated with peroxidase-conjugated secondary antibodies for 1 h at room temperature. After three more washes with TBST, proteins were visualized using a Chemiluminescence Detection Kit (Millipore). GAPDH was used as a loading control.

siRNA Transfection and Cell Line Drug Sensitivity
Target sequences for ZNF703 siRNA using RNAiMAX (30 nM final concentration, Invitrogen) were as follows: siRNA-1 (sense strand-59 CCACACACUUUGGGCCUAA dTdT 39; antisense strand-39 dTdT GGUGUGUGAAACCCGGAUU 59), siRNA-2 (sense strand-59 GGACCCUACUAUUCGCCAU dTdT 39; antisense strand-39 dTdT CCUGGGAUGAUAAGCGGUA 59). Non-targeting control (NC) siRNA was designed and synthesized by Guangzhou RuiBoBio (Guangzhou, China). For assessment of endocrine therapy, cells were maintained in phenol red-free media, supplemented with 10% charcoal/DCC stripped serum (Tissue Culture Biologicals, Long Beach, CA, USA) and 1 nM estradiol, and then plated in 96-well plates and treated for 3 to 6 days with a range of TAM concentrations. Cell proliferation was determined using the sulforhodamine B (SRB) (Sigma) assay. Relative growth was calculated as the value relative to dimethyl sulfoxide (DMSO) treated cells (control).

Colony-forming Assay
Cells were seeded into 6-well plates (1000 cells per well). Standard growth medium, with TAM or DMSO, was changed every other day. On day 15, colonies were fixed in 10% methanol, 10% acetic acid, and 80% ddH 2 O, and then stained with crystal violet (0.5% w/v).

Analysis of ZNF703 Expression in Tissue Microarray
Tissue specimens were obtained from archived tissue samples of patients with primary breast cancer who underwent surgical treatment at Ruijin Hospital (China) between January 2001 and December 2003. The selection criteria were as follows: (1) subjects had a diagnosis of stage I-III primary breast cancer and no history of other tumors; (2) subjects had complete clinical data, such as age, gender, stage, lymph node status, tumor size (as according to pathology reports), and pathological type. Patients receiving chemotherapy or radiotherapy prior to surgery were excluded. IHC analysis of ZNF703 protein was performed on tissue microarray sections with antibodies against ZNF703, as previously described [64]. ZNF703 expression was independently evaluated by two pathologists, both of whom were blinded to the clinicopathologic data. Specimens were assigned scores according to the intensity of nuclear staining (no staining = 0; weak staining = 1, moderate staining = 2, strong staining = 3), and to the percentage of stained cells (0% = 0, 1-20% = 1, 21-40% = 2, 41-60% = 3, 61-80% = 4, 81-100% = 5). The final immunoreactive score was determined by multiplying the intensity score by the score representing the percentage of stained cells, and ranged from 0 (minimum score) to 15 (maximum score). Cases were further classified into negative/low and high expression groups according to the median score of nucleic acid staining. Evaluation of hormone receptor expression was based on the Allred scoring method [65].

Statistics
Each experiment was repeated at least three times. Data were presented as mean 6 the standard deviation (SD) of each experiment. Data analysis was performed using SPSS 17.0. Student's t-test was used to evaluate the numerical data. Chisquare test was used for comparisons of categorical data. Statistical tests were two-sided, and P values less than 0.05 were considered statistically significant.