Estrogen induces c-myc gene expression via an upstream enhancer activated by the estrogen receptor and the AP-1 transcription factor

Abstract

c-myc oncogene is implicated in tumorigenesis of many cancers, including breast cancer. Although c-myc is a well-known estrogen-induced gene, its promoter has no estrogen-response element, and the underlying mechanism by which estrogen induces its expression remains obscure. Recent genome-wide studies by us and others suggested that distant elements may mediate estrogen induction of gene expression. In this study, we investigated the molecular mechanism by which estrogen induces c-myc expression with a focus on these distal elements. Estrogen rapidly induced c-myc expression in estrogen receptor (ER)-positive breast cancer cells. Although estrogen had little effect on c-myc proximal promoter activity, it did stimulate the activity of a luciferase reporter containing a distal 67-kb enhancer. Estrogen induction of this luciferase reporter was dependent upon both a half-estrogen response element and an activator protein 1 (AP-1) site within this enhancer, which are conserved across 11 different mammalian species. Small interfering RNA experiments and chromatin immunoprecipitation assays demonstrated the necessity of ER and AP-1 cross talk for estrogen to induce c-myc expression. TAM67, the AP-1 dominant negative, partially inhibited estrogen induction of c-myc expression and suppressed estrogen-induced cell cycle progression. Together, these results demonstrate a novel pathway of estrogen regulation of gene expression by cooperation between ER and AP-1 at the distal enhancer element and that AP-1 is involved in estrogen induction of the c-myc oncogene. These results solve the long-standing question in the field of endocrinology of how estrogen induces c-myc expression.

Fig. 1.

Effect of estrogen on c-myc gene expression. A, Time course of E2 induction of c-myc expression in MCF-7 cells. Hormone-depleted cells were treated with vehicle or E2 for different time points as indicated, and c-myc mRNA was quantified by q-RT-PCR. c-Myc protein was analyzed by Western blot after E2 treatment for 6 h (inset). B, Effects of ActD and CHX on E2 induction of c-myc expression. Hormone-depleted MCF-7 cells were treated with ActD or CHX for 1 h before addition of vehicle or E2. RNA was harvested at 1 h after treatment with vehicle or E2. C, Effect of the pure ER antagonist ICI on E2 induction of c-myc expression. Hormone-depleted MCF-7 cells were treated with ICI for 1 h before addition of vehicle or E2. RNA was harvested at 1 h after treatment with vehicle or E2. D, Effect of estrogen on c-myc gene expression in different breast cancer cell lines. Hormone-depleted cells were treated with E2 for 2 h, and c-myc expression was quantified using q-RT-PCR.

Fig. 2.

Effect of AP-1 on estrogen-regulated c-myc expression. A, Effect of E2, TPA, or the combination of both on c-myc expression. RNA was harvested at 1 h after treatment. B, Effect of AP-1 blockade on E2 induction of c-myc expression. TAM67 Tet-Off MCF-7 cells were cultured in the presence or absence of DOX for 5 d and switched into CSS for 2 d before treatment with vehicle or E2. RNA was harvested at 0 h, 0.5 h, and 1 h after treatment. Asterisk denotes P < 0.01 between E2, DOX+ and E2, DOX-. C, Validation of TAM67 and ER expression in response to DOX withdrawal by Western blot.

Fig. 3.

Effects of AP-1 and c-myc on estrogen-regulated cell cycle progression. A, Effects of AP-1 blockade on estrogen-regulated cell cycle progression. TAM67 Tet-Off MCF-7 cells were cultured in the presence or absence of DOX for 5 d and switched into CSS for 2 d before vehicle/E2 treatment. Cells were harvested for cell cycle analysis after 16 h treatment with vehicle/E2. B, Effects of c-myc knockdown on estrogen-regulated cell cycle progression. Hormone-depleted MCF-7 cells were treated with vehicle/E2 for 16 h and then harvested for cell cycle analysis. C, Validation of the efficiency of siRNA knockdown of c-myc expression by Western blot.

Fig. 4.

Luciferase assay of c-myc enhancer/promoter luciferase constructs. A, Schematic representation of c-myc proximal promoter. B, Effect of E2 on the reporter activity of various length of c-myc promoter luciferase constructs. Hormone-depleted MCF-7 cells were transfected with c-myc promoter luciferase plasmid. Cells were treated 16 h later with vehicle/E2 for 6 h and lysed for luciferase analysis. C, More complete map of c-myc gene structure in the genomic context, showing the ER-binding sites that were previously identified by us and others (chr8:128749539-128750675, −67958 to −66822 from the transcriptional start site). D, Effect of E2 on the reporter activity of c-myc promoter luciferase construct with or without the distal 67-kb upstream enhancer.

Fig. 5.

Deletion and mutation analysis of c-myc enhancer-promoter luciferase constructs. A, Potential transcription factor-binding sites in the distal 67 kb upstream enhancer, showing the predicted half-EREs and AP-1 sites. B, Effect of E2 on the reporter activity of c-myc enhancer-promoter luciferase constructs with deletion and mutation of the half-ERE and AP-1 sites. Luciferase analysis was performed as described previously. Asterisk denotes P < 0.01 compared with control. Double asterisk denotes P < 0.001 compared with control. C, Bioinformatics analysis of the enhancer sequences in a variety of mammalian species. Shaded blocks refer to the elements with the same sequences as the elements in human. An evolutionarily conserved pattern was noticed only for the identified functional half-ERE and the AP-1 site, but not for the other nonfunctional half-EREs and AP-1 sites.

Fig. 6.

Effect of siRNA knockdown of ER or each AP-1 member on estrogen-induced c-myc expression. Hormone-depleted MCF-7 cells were transfected with siRNA against each gene as indicated. Cells were treated 36 h later with vehicle/E2 for 1 h and harvested for q-RT-PCR. A, Validation of the efficiency of siRNA knockdown of ER or each AP-1 member by Western blot. B, Estrogen-induced c-myc expression as measured by q-RT-PCR. Data from nonspecific siRNA-transfected and vehicle-treated samples were arbitrarily set as 1. Asterisk denotes P < 0.01 compared with siLuc control.

Fig. 7.

ChIP assay. A, Schematic representation of the primers used to amplify the c-myc enhancer, promoter, or exon 2 regions. B, Recruitment of ER, AP-1, p300/CBP, and Pol II to c-myc enhancer, promoter, or exon 2 regions. Hormone-depleted MCF-7 cells were treated with vehicle or E2 for 30 min and then fixed with formaldehyde. Data were presented as relative amount of immunoprecipitated DNA normalized to input as measured by q-PCR assay. Primers for the pS2 gene promoter were used as a positive control (data not shown). C, Recruitment of Pol II to the c-myc promoter region after knockdown of ER, JunD, or FosB. Hormone-depleted MCF-7 cells were transfected for 36 h with siRNA against ER, JunD, and FosB after which cells were treated with vehicle or E2 for 30 min and fixed with formaldehyde. Data were presented as relative amount of immunoprecipitated DNA normalized to input as measured by q-PCR assay. Asterisk denotes P < 0.05 compared with siLuc control.

Fig. 8.

Schematic representation of the cooperation of ER and AP-1 at the distal enhancer region in regulating estrogen-induced c-myc gene expression.