Bisphenol-A induces expression of HOXC6, an estrogen-regulated homeobox-containing gene associated with breast cancer

Abstract

HOXC6 is a homeobox-containing gene associated with mammary gland development and is overexpressed in variety of cancers including breast and prostate cancers. Here, we have examined the expression of HOXC6 in breast cancer tissue, investigated its transcriptional regulation via estradiol (E2) and bisphenol-A (BPA, an estrogenic endocrine disruptor) in vitro and in vivo. We observed that HOXC6 is differentially over-expressed in breast cancer tissue. E2 induces HOXC6 expression in cultured breast cancer cells and in mammary glands of Sprague Dawley rats. HOXC6 expression is also induced upon exposure to BPA both in vitro and in vivo. Estrogen-receptor-alpha (ERα) and ER-coregulators such as MLL-histone methylases are bound to the HOXC6 promoter upon exposure to E2 or BPA and that resulted in increased histone H3K4-trimethylation, histone acetylation, and recruitment of RNA polymerase II at the HOXC6 promoter. HOXC6 overexpression induces expression of tumor growth factors and facilitates growth 3D-colony formation, indicating its potential roles in tumor growth. Our studies demonstrate that HOXC6, which is a critical player in mammary gland development, is upregulated in multiple cases of breast cancer, and is transcriptionally regulated by E2 and BPA, in vitro and in vivo.

Keywords: Bisphenol A; Endocrine disruption; Epigenetics; Estrogen-receptors; HOXC6 expression; Mixed lineage leukemia (MLL).

Figure 1

HOXC6 expression in breast cancer cells and tissues. (Panel A-C) Immunohistological analyses of HOXC6 expression in breast cancer tissues: Human breast cancer tissue microarray (6 cases of breast cancer along with their matched adjacent normal breast tissues) were immunostained (DAB staining) with HOXC6 antibody (panel A) and the relative quantification of HOXC6 expression is shown in panel B. (C) RNA from different breast cancer cells (MCF7, T47D, and MDA-MB-231) was reverse transcribed and analyzed by qPCR using primers specific to HOXC6. HOXC6 expression relative to GAPDH is plotted. Each experiment was repeated at least thrice with three parallel replicates (n=3). Bars indicate standard errors.

Figure 2

Effects of E2 and BPA on HOXC6 expression in MCF7 cells. MCF7 cells grown in phenol red free DMEM-F-12 were treated with varying concentrations (or time) of E2 or BPA. RNA from the control and E2 or BPA-treated cells were analyzed by RT-PCR and qPCR using HOXC6 specific primers. GAPDH was used as the loading control. Panel A-B: RT-PCR and qPCR analysis of HOXC6 expression upon exposure varying concentrations of E2, respectively. Panel C-D: RT-PCR and qPCR analysis of HOXC6 expression upon exposure to 1 nM E2 for varying time periods. Panel E-F: RT-PCR and qPCR analysis of HOXC6 expression upon exposure varying concentrations of BPA, respectively. Panel G-H: RT-PCR and qPCR analysis of HOXC6 expression upon exposure to 10 nM BPA for varying time periods. Panel I-J; RT-PCR and qPCR analyses of HOXC6 expression in T47D cells upon exposure to 1nM E2 and 10 nM BPA. Each qPCR experiment was repeated for thrice times with three parallel replicates (n = 3). Bars indicate standard errors (p < 0.05).

Figure 3

Effects of E2 and BPA on HOXC6 gene expression, in ovariectomized (OVX) rats, in vivo. (Panel A-C): OVX female rats were administered with acute doses of E2 (5 µg) and BPA (25 µg/kg), for 24 h, either separately or in combination. Control animals were administered with the vehicle (peanut oil/saline). RNA and protein were isolated from the mammary glands of the control, E2 and BPA treated animals. RNA was analyzed by regular RT-PCR (Panel A); GAPDH was used as a loading control and qPCR (Panel B) using rat specific HOXC6 primers (relative to GAPDH). Each experiment was repeated thrice with three parallel replicates. Bars indicate standard errors, (p < 0.05)). Western blot analysis of the HOXC6 protein levels from the protein samples obtained from control, E2 and BPA treated mammary gland tissues of animals are shown in panel C. β-actin was used as the loading control. Each experiment was repeated at least thrice (n = 3).

Figure 4

HOXC6 promoter ERE is responsive to E2 and BPA treatment (luciferase assay). Panel A-B: (A) HOXC6 promoter showing the ERE that was cloned into luciferase construct, pGL3. (B) Luciferase assay: ERE-pGL3 construct was transfected into MCF7 cells for 30 h. Control cells were treated with empty pGL3 vector and no transfection controls were done in parallel. Cells were also co-transfected with renilla luciferase expression construct as an internal transfection control. Cells were then treated with 1 nM E2 or 10 nM BPA and then subjected to luciferase assay by using Dual-Glo Luciferase Assay System (Promega). The luciferase activities in presence of E2 and BPA (over untreated controls and normalized against renilla luciferase expression) were plotted. The experiment with four replicate treatments was repeated at least thrice (n = 3). Bars indicate standard errors (p < 0.05).

Figure 5

Enrichment of ERα, MLLs (MLL1–4), CBP/p300 and N-CoR at the HOXC6 promoter in presence of E2 and BPA. (Panel A-D): MCF7 cells were treated with 1nM E2 or 10 nM BPA for 4 h and analyzed by ChIP assay using ERα, MLL1, MLL2, MLL3, MLL4, CBP, p300, N-CoR and β-actin (negative control) antibodies. The ChIP DNA was PCR-amplified using primers specific to ERE region of HOXC6 promoter. The recruitment of ERα, MLL1, MLL2, MLL3, MLL4 and β-actin in the HOXC6 promoter are shown in the panel A (qPCR) and panel B (RT-PCR). Panel C-D shows the recruitment of CBP, p300, N-CoR and β-actin upon treatment with E2 or BPA. Bars indicate standard errors (p < 0.05).

Figure 6

Enrichment of RNA polymerase II (RNAPII), histone acetylation and H3K4-trimethylation at the HOXC6 promoter in presence of E2 or BPA. (Panel A-B): MCF7 cells were treated with 1 nM E2 or 10 nM BPA for 4 h and analyzed by ChIP assay using RNAPII, H3K4-trimethyl, histone acetylation and β-actin (negative control) antibodies. The ChIP DNA was PCR amplified using primers specific to ERE region of HOXC6 promoter. Panel A-B (qPCR and regular RT-PCR data respectively) shows the levels of RNAPII, H3K4-trimethyl and histone acetylation at the HOXC6 promoter. Bars indicate standard errors (p < 0.05).

Figure 7

ERα knockdown downregulates E2 and BPA induced expression of HOXC6. (Panels A-B): MCF7 cells were transfected with antisense specific for ERα or scramble antisense for 48 h and then treated with 1 nM E2 and 10 nM BPA separately for additional 4 h. RNA was subjected to RT-PCR (Panel A) and qPCR analyses (Panel B) using primers specific to ERα, HOXC6 and GAPDH. Bars indicate standard error (n = 3; P < 0.05).

Figure 8

HOXC6 over expression induces carcinogenesis. (A) Western blots of HOXC6 overexpression analysis. Proteins were isolated from HEK293 cells and stable cells expressing Flag-HOXC6 and subjected to Western blotting using β-actin (control) and Flag antibodies. (B) 3D-colony formation ability of HOXC6 overexpressed cells. HEK293 and Flag-HOXC6 stable cells were incubated in soft agar for 4–5 weeks stained with 0.005% crystal violet and analyzed under microscope. Magnified view of one HOXC6 overexpressed colony (C) Number of colonies grown in soft agar assays were counted using light microscope and plotted. (D) Analysis of expression of growth factors. The RNA from HEK293 and Flag-HOXC6 overexpressed cells were subjected to RT-qPCR with primers specific to HOXC6, bFGF, VEGF, PDGFRA, and GAPDH (control). Bars indicate standard error ((n = 3, p<0.05)

Figure 9

Models showing the roles of ERs, MLLs and other ER-coregulators during E2 and BPA mediated upregulation of HOXC6. During a typical estradiol mediated HOXC6 gene activation, ERs upon bindinig to E2 dimerize and then dimerized (activated) ERs bind to the EREs of the HOXC6 promoter. ER-coregulators such as MLL2, MLL3, CBP, p300, and other ER-coregulators are also recruited to the HOXC6 promoter. Promoter histones are methylated via MLL-methylases and acetylated via p300 activity, allowing access to RNA polymerase II (RNAP II) to the promoter, ultimately resulting in HOXC6 expression activation. Steroidogenic endocrine disrupting chemicals (EDCs) like BPA when enters into the cells, competes with esrtradiol and binds to ERs leading to activation of ERs and associated cell signaling and target gene activation, in a fashion very similar to E2. Thus, even in the absence of E2, BPA can induce ER-target genes such as HOXC6 leading to their misregulation.