ENVIRONMEN TAL HEALTH PER SPEC T I VES ehp Parabens and Human Epidermal Growth Factor Receptor Ligands Cross-Talk in Breast Cancer Cells

Background: Xenoestrogens are synthetic compounds that mimic endogenous estrogens by binding to and activating estrogen receptors. Exposure to estrogens and some xenoestrogens has been associated with cell proliferation and increased risk of breast cancer. Despite evidence of estrogenicity, parabens are among the most widely used xenoestrogens in cosmetics and personal care products, and generally considered safe. However, previous cell based studies with parabens do not take into account the signaling cross-talk between estrogen receptor (ER α ) and the human epidermal growth factor receptor (HER) family. Objectives: We investigated the hypothesis that the potency of parabens can be increased with HER ligands, such as heregulin (HRG). Methods: The effects of HER ligands on paraben activation of c-Myc expression and cell proliferation were determined by real-time PCR, western blots, flow cytometry and chromatin immunoprecipitation assays in ER α - and HER2-positive human BT-474 breast cancer cells. Results: Butylparaben (BP) and HRG produced a synergistic increase in c-Myc mRNA and protein levels in BT-474 cells. Estrogen receptor antagonists blocked the synergistic increase in c-Myc protein levels. The combination of BP and HRG also stimulated proliferation of BT-474 cells compared to BP alone. HRG decreased the dose required for BP-mediated stimulation of c-Myc mRNA expression and cell proliferation. HRG caused the phosphorylation of serine 167 in ER α . BP and HRG produced a synergistic increase in ER α recruitment to the c-Myc gene. Conclusion: Our studies demonstrate that HER ligands enhance the potency of BP to stimulate oncogene expression and breast cancer cell proliferation in vitro via ER α , suggesting that parabens might be active at exposure levels not previously considered toxicologically relevant from studies testing their effects in isolation.


Introduction
Xenoestrogens are a class of synthetic estrogens known as endocrine disrupting chemicals that bind to estrogen receptors in cells to mimic or antagonize the action of endogenous estrogens, such as 17β-estradiol (E2) (Zoeller et al. 2012). Numerous xenoestrogens are found in common household products, including plastics, food and soda cans, and personal care products. One class of xenoestrogens that is an increasing public health concern is esters of parahydroxybenzoic acid, commonly known as parabens (Nohynek et al. 2013;Karpuzoglu et al. 2013). They are common ingredients in cosmetics, shampoos, body lotions and sunscreens, where they are used to prevent microbial growth and prolong shelf life (Guo and Kannan 2013;Dodson et al. 2012). Detectable levels of multiple parabens are present in human urine (Calafat et al. 2010;Den Hond et al. 2013;Mortensen et al. 2014) and breast tissue (Darbre et al. 2004; Barr et al. 2012;Darbre and Harvey 2014).
While endocrine disrupting chemicals have been linked to a variety of medical conditions, one of the most troubling is their association with breast cancer (Zoeller et al. 2012;Vandenberg et al. 2012;Darbre and Harvey 2008). Endogenous estrogens promote breast cancer by binding to estrogen receptor α (ERα) (Burns and Korach 2012;Sommer and Fuqua 2001), which causes the activation of oncogenes, such as c-Myc and cyclin D1 (Liao and Dickson 2000;Leygue et al. 1995). Cyclin D1 and c-Myc cause cell proliferation by facilitating a G1 to S-phase transition (Foster et al. 2001). Approximately two-thirds of breast tumors express ERα, and therapeutic strategies aimed at preventing and treating ER positive breast tumors are directed at blocking the action of ERα. Parabens are known to bind to ERα (Routledge et al. 1998), promote a G 1 to S-phase cell cycle progression, stimulate the proliferation of MCF-7 breast Environ Health Perspect DOI: 10.1289/ehp.1409200 Advance Publication: Not Copyedited 4 cancer cells (Darbre et al. 2003;Wrobel and Gregoraszczuk 2013;Okubo et al. 2001) and activate transcription of cell cycle (Wrobel and Gregoraszczuk 2014) and reporter genes (Darbre et al. 2003;Gomez et al. 2005). These findings indicate that paraben exposure might increase the risk of breast cancer by activating ERα to promote the activation of proliferative genes.
However, parabens are considered to be safe due to their weak estrogenic binding affinity, transcriptional activation, and stimulation of cell proliferation, and the dose required for ERα activation often exceeds those found in the body (Lemini et al. 2003;Pugazhendhi et al. 2005).
The most estrogenic paraben, butylparaben, was found to be 10,000-fold less potent than E2 (Routledge et al. 1998). However, studies involving xenoestrogens have tested them in the absence of activators of the human epidermal growth factor receptor (HER) family of receptor tyrosine kinases (Wrobel and Gregoraszczuk 2013;Wrobel and Gregoraszczuk 2014), a second signaling pathway implicated in breast cancer (Liu et al. 2009).
The HER family comprises four receptors: EGFR/HER1, ErbB2/HER2, ErbB3/HER3 and ErbB4/HER4 (Davoli et al. 2010). HER2 is a transmembrane protein that is overexpressed in approximately 25% of breast tumors (Davoli et al. 2010). Its presence in human tumors is a negative prognostic indicator, associated with malignant transformation, fast growth, and more aggressive tumors (Barros et al. 2010;Davoli et al. 2010). The association between HER2 expression and breast cancer led to the development of the drug Herceptin (trastuzumab), a recombinant humanized monoclonal antibody against HER2 to treat HER2 positive tumors (Hudis 2007). At least eleven proteins, known as HER ligands, can bind to HER family members to cause dimerization, leading to the activation of the phosphoinositide 3-kinase/protein kinase B (PI3K/AKT) and other signal transduction pathways (Mosesson and Yarden 2004). Aberrant activation of the PI3K/AKT signaling pathway may increase the risk of cancer by inhibiting 5 apoptosis and stimulating cell proliferation (Liu et al. 2009). HER and ERα signaling pathways can cross-talk, as demonstrated by the observation that HER ligands stimulate phosphorylation of the serine 167 (ser167) residue in ERα (Al-Dhaheri and Rowan 2006;Lannigan 2003;Murphy et al. 2011). Eliminating the main source of endogenous estrogens by ovariectomy delays the formation of mammary tumors and increases the lifespan of transgenic mice that overexpress HER2 in the mammary gland (Anisimov et al. 2003). Furthermore, when HER2 transgenic mice are mated to ERα knockout mice, tumor onset is delayed compared to control HER2 transgenic mice (Hewitt et al. 2002). Based on these findings, we hypothesize that activators of the HER2 pathway might cause parabens to be stimulatory of ERα at lower doses than suspected based on studies that examined their effects in isolation. In the present study, we determined the potency of parabens in the presence of the HER ligand, HRG in BT-474 breast cancer cells that express both ERα and HER2.

Cell culture
Human BT-474, MCF-7, and SKBR3 breast cancer cell lines were obtained from ATCC and used in these studies because of their differences in expression of HER2 and ERα. BT-474 cells are HER2 and ERα positive, MCF-7 cells are ERα positive and HER2 negative and SKBR3 cells are HER2 positive and ERα negative (Neve et al. 2006). Cells were grown in phenol red-free Dulbecco's modified Eagle's medium/F12 (DMEM/F12) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 units/ml penicillin, 10 µg/ml streptomycin (Life Technologies) under 5% CO 2 at 37 o C. Three days prior to treatment, the cells were Technologies and used at a final concentration of 20 ng/ml to activate the HER2 signaling pathway. Estradiol, raloxifene, tamoxifen, methylparaben (MP), ethylparaben (EP), propylparaben (PP) and butylparaben (BP) were purchased from Sigma-Aldrich. The estrogen receptor antagonist, ICI 182,780 was purchased from Tocris. The compounds were dissolved in ethanol. The final concentration of ethanol was 0.1%, which had no effect on the cells. An ethanol vehicle was used for the control cells.

Real-time RT-PCR
BT-474 cells (passage numbers 86-95) were grown in 6-well tissue culture dishes to reach 80% confluence and then maintained in DMEM/F12 supplemented with 10% charcoaldextran stripped FBS for 3 days. The cells were treated with 0.01 µM E2 or 10 µM MP, EP, PP or BP in the absence or presence of 20 ng/ml HRG for 2 h. The 10 µM concentration of parabens was selected by performing preliminary dose-response studies. Total RNA was isolated and purified using an Aurum Total RNA Mini Kit (Bio-Rad Laboratories, Inc). RNA purity and concentration were determined using a NanoDrop ND-1000 spectrophotometer. Reverse transcription of total RNA was carried out using iScript (Bio-Rad Laboratories, Inc) as previously described (Paruthiyil et al. 2009  The expression levels c-Myc and GAPDH were determined by the comparative Ct method as previously described (Paruthiyil et al. 2009).

Western blot
Human BT-474, MCF-7, and SKBR3 cells were grown in 6-well tissue culture dishes in phenol red-free DMEM/F12 supplemented with 10% FBS, 2 mM L-glutamine, 100 units/ml containing 20 mM Tris-HCl, pH 7.5, 500 mM NaCl, and 0.1% tween 20 and probed overnight with rabbit anti-c-Myc IgG (sc-764, Santa Cruz Biotechnology, Inc.) at 0.5 µg/ml in 1% milk-TBST at 4ºC. After washing with TBST for 5 min for three times at room temperature, the membrane was incubated with goat anti-rabbit IgG conjugated to horseradish peroxidase (sc- Detection Reagent (GE Healthcare Life Sciences). The western blot for ERα phosphorylation was performed as described for c-Myc except that the cell lysis buffer contained a phosphatase inhibitor cocktail (PhosSTOP, Roche) and the primary antibody was anti-phospho-ERα S167 (Bethyl Laboratories).

Chromatin immunoprecipitation (ChIP)
Confluent BT-474 cells were treated with butylparaben in the absence and presence of HRG (20 ng/ml) for 1-3 h. The cells were harvested for ChIP assay as previously described (Cvoro et al. 2006) with some modifications. Briefly, to cross-link proteins to DNA, the cells were fixed by adding formaldehyde to the cell culture medium and incubated at 37ºC for 10 min, followed by the addition of glycine for 5 min at room temperature to quench the cross-linking reaction. The cell monolayer was then washed with Phosphate-Buffered Saline (PBS) containing cOmplete Protease Inhibitor Cocktail and collected by scraping. The cells were concentrated by centrifugation and lysed with buffer containing 0.5% of Triton X-100, 50 mM of Tris-HCl, pH 7.4, 150 mM of NaCl, 10 mM of EDTA and cOmplete Protease Inhibitor Cocktail Tablets as previously described (Vivar et al. 2010 PCR with specific primers for c-Myc enhancer region as previously described (Wang et al.).

Cell Cycle Analysis using Flow Cytometry
The effects of different treatments on cell cycle phase was analyzed by flow cytometry based on the method previously described (Wiepz et al. 2006 for 24 h. The cells were then trypsinized and collected by centrifugation at 1700 rpm for 5 minutes at room temperature. The cell pellets were washed once with ice cold PBS and centrifuged at 1700 rpm for 10 minutes at room temperature followed by resuspension in 500 µl propidium iodide solution (PBS containing 0.1% Triton 100, 0.1% sodium citrate, 10 µg/ml RNase and 0.05 mg/ml propidium iodide) to stain the cells. The cell suspensions were assayed with a Cytomics FC-500 flow cytometer (Beckman Coulter) using CXP software in the flow cytometry core facility at University of California, Berkeley and the data were then analyzed using FlowJo 7.6.5 (FlowJo).

Cell Proliferation Assay
BT-474 cells were plated in 6-well tissue culture dishes in phenol red-free DMEM/F12 supplemented with 10% charcoal-dextran stripped FBS. The next day, the cells were treated with

Statistical Analysis
Data are presented as the mean ± SD or mean ± SEM as indicated in the figure legend.
The statistical significance of differences was examined by one-way analysis of variance or twoway analysis of variance (ANOVA) tests as specified in the figure legend. All ANOVA tests are followed by Tukey's multiple comparisons post hoc tests to analyze the difference between different time periods or doses within groups treated with same reagents (BP, HRG or BP plus HRG). Bonferroni's multiple comparisons post hoc test was applied to analyze the difference between groups with and without HRG within the same paraben treatment or the same time period. Data analysis was performed by using GraphPad Prism (version 6.01; GraphPad Software Inc.; La Jolla, CA, USA).

Combined effects of parabens and heregulin on c-Myc transcript levels in BT-474 breast cancer cells.
Since BT-474 cells express both ERα and HER2 (Lazaro et al. 2013)  parabens at increasing c-Myc mRNA levels in the absence of HRG ( Figure 1A). HRG alone produced about a 3-fold increase in c-Myc mRNA levels, but a synergistic increase that was greater than additive was observed with PP and BP ( Figure 1A). BP was the most effective stimulator of the c-Myc mRNA levels in the absence and presence of HRG, and was selected for further studies. The maximal increase of c-Myc expression level by BP was observed at 10 µM ( Figure 1B). The synergistic effect of HRG was observed when BT-474 cells were treated with BP for 1 hour ( Figure 1C). These results demonstrate that HRG decreases the dose required for the BP-mediated increase in c-Myc mRNA levels and enhances the magnitude of the BP response.

Combined effects of heregulin and BP on c-Myc protein levels in ERα positive cell lines.
To  Figure 4C). The shift in BP potency was more pronounced after treatment for 5 days. In the absence of HRG, 1 µM BP was required to produce a significant increase in cell number ( Figure 4D), whereas in the presence of HRG, 0.01 µM BP significantly increased cell number ( Figure 4E). These findings indicate that HRG lowers the dose of BP required to stimulate BT-474 cell proliferation.

Effect of heregulin on serine 167 phosphorylation of ERα and the recruitment of ERα to the c-Myc enhancer by the heregulin and BP combination.
One potential mechanism whereby HRG and BP could cooperate to produce a synergistic   (Nagashima et al. 2008;Joel et al. 1998 A major rationale promulgated in favor of the safety of xenoestrogens in consumer products is that, at biologically relevant concentrations, they bind to estrogen receptors with too low affinity to produce significant biological effects in humans (Golden et al. 2005). For example, BP was found to bind to ERα with about 10,000 fold lower affinity than E2 (Bolger et al. 1998). Similarly, functional assays of ERα such as reporter assay activation and MCF-7 cell proliferation found that physiologically implausible concentrations of parabens are needed for ERα activation (Golden et al. 2005 While it is clear that endogenous estrogens increase the risk of breast cancer, the role of parabens in breast cancer is controversial, in part due to uncertainty about whether concentrations of parabens present in the body are sufficient to mimic the effects of endogenous estrogens on breast cells (Karpuzoglu et al. 2013;vom Saal et al. 2007;Harvey 2003). Our studies demonstrated that even in the presence of HRG, higher concentrations of parabens than The presence of multiple HER receptors and ligands in breast tissues may affect the activity of parabens. The known eleven endogenous HER ligands can bind to one or more of the HER receptors (Mosesson and Yarden 2004) with the notable exception of HER2, for which there is no known ligand (Harari and Yarden 2000). However, the binding of ligands to HER1,

16
HER3 or HER4 leads to a preferential dimerization and activation of HER2 (Rubin and Yarden 2001). Further work will be needed to determine if other HER ligands potentiate the effect of parabens, and their relative potency compared to HRG. Interestingly, breast cancer cells are autocrine producers of HER ligands. In a study of 363 breast tumors it was found that 80%-96% of the tumors expressed at least one of ten tested HER ligands (Revillion et al. 2008). Similarly, another study found that 48% of breast tumors express HRG (Esteva et al. 2001). Breast tumors may therefore potentiate their own response to estrogenic compounds by producing HER ligands.

Conclusion
Our data showing that lower doses of butylparaben are required to stimulate breast cancer cell proliferation in the presence of HRG together with the observations that breast tumors are exposed in vivo to both HER ligands (Revillion et al. 2008) and parabens (Darbre et al. 2004) indicate a potential synergy relevant to proliferation of tumor cells in humans. Further work is needed to assess if indeed HER ligands enhance the potency of parabens in the normal human breast cells and breast tumors. We suggest that reevaluation of the potency of other xenoestrogens in the presence of HER ligands is warranted in the light of our findings.