Estrogenic activation of ERα and ERβ by BPA, BPAF, and Zea. To evaluate ERE-mediated transcriptional activity of ERα and ERβ, we examined promoter activation in Ishikawa, HeLa, and HepG2 cells. The luciferase reporter assay system was used to determine differential treatment effects of BPA, BPAF, and Zea on ERα or ERβ in cells derived from different tissues. Reporter activation of ERE-luc in response to E2, BPA, BPAF, or Zea was observed only with both ERα and ERβ expression plasmids but not with pcDNA control plasmid.
For ERα activation in Ishikawa cells, E2 activated the ERE-luc reporter at a concentration of 1 nM (4-fold induction) (). At low concentrations (≤ 10 nM), BPAF and Zea had weak estrogenic activity compared with E2, but stronger activation was observed at higher concentrations (1,000 nM). In contrast, 100 nM BPA was required for activation in these cells. Compared with Ishikawa cells, HeLa cells were less active, with only a 2.5-fold increase in ERE-mediated ERα activation in response to E2 (10 nM) (). ERα activation in response to BPA resembled that for E2, whereas induction by BPAF and Zea did not occur except with higher concentrations (≥ 100 nM). HepG2 cells were highly responsive to E2, with 10- to 20-fold increases in ERE-mediated transactivation at 10–1,000 nM E2 (). Interestingly, BPA activity was not evident for concentrations ≤ 100 nM, whereas Zea showed strong activation at 1 and 10 nM but less activity at higher concentrations. All activations occurred in a concentration-dependent manner in the three cell lines tested, with the exception of reduced activity of ERα activation when Zea was used at higher concentrations (≥ 100 nM).
Figure 1 BPA, BPAF, and Zea act as agonists or antagonists for ERα and ERβ. (A–F) Dose–response curves for ERα (A–C) or ERβ (D–F) in Ishikawa (A,D), HeLa (B,E), and HepG2 (C,F) cells transfected with (more ...)
In Ishikawa and HepG2 cells, ERβ was weak in response to the three EDCs compared with E2
(). In HeLa cells, ERβ responses to 100 or 1,000 nM BPA or BPAF were similar to those of E2
, whereas the ERβ response was stronger for 1 nM Zea than for E2
(). In addition, the pure ER antagonist, ICI, blocked ERα and ERβ transcriptional activity in response to all three EDCs [see Supplemental Material, Figure 2
)]. Taken together, these results demonstrate that BPA, BPAF, and Zea can activate ERE-mediated transcription in different cell types via ERα and ERβ, and that the estrogenic activity of each compound is cell type and concentration dependent.
Activity of ERα and ERβ is antagonized by low doses of BPA, BPAF, and Zea. The EDCs showed weak activity in certain cell types at low doses. We therefore investigated antagonistic effects of BPA, BPAF, and Zea on ERα and ERβ using the ERE-mediated reporter assay system. Cells were transiently transfected with ERα or ERβ expression plasmids and then treated with 1 or 10 nM BPA, BPAF, or Zea with or without 10 nM E2 co-treatment. We observed antagonistic effects of BPA and Zea on ERα only in Ishikawa cells (). Both concentrations of BPA and Zea inhibited 70–80% of the E2-ERE–mediated reporter activity. At 1 and 10 nM, BPAF and Zea weakly activated ERα, but BPAF did not inhibit E2 activation. In addition, Zea (1 and 10 nM) induced ERα E2-ERE–mediated reporter activity, but E2-mediated activation was inhibited with E2 co-treatment (). We observed no antagonistic effects of these three EDCs on ERα in HeLa or HepG2 cells (data not shown).
For ERβ, E2-ERE–mediated reporter activity was reduced in HeLa cells treated with BPAF but not with BPA or Zea treatment (, ). None of the EDCs showed antagonistic effects on ERβ in Ishikawa or HepG2 cells (data not shown). Data demonstrate that low doses of BPA and BPAF (≤ 10 nM) antagonized ERα activity and ERβ activity, respectively; however, these effects were cell-type specific.
Effects of BPA, BPAF, and Zea on ERα functionality. To link BPA-, BPAF-, and Zea-mediated activation to a specific ERα functionality, we used four ERα mutants: H1-ERα (ERE-mediated activation, but no tethered-mediated activation), AA-ERα (tethered-mediated activation, but no ERE-mediated activation), E1-ERα (AF-1 inactive), and AF-2-ERα (AF-2 inactive). We observed estrogenic effects of BPA, BPAF, and Zea at 100 nM with WT-ERα in Ishikawa cells (). Therefore, we used this concentration in the mutant experiments. Cells were transiently transfected with an ERE-luc reporter plasmid and ERα mutant expression plasmids. Reporter activity was calculated for each mutant relative to activity in the vehicle control (). Responses to BPA, BPAF, and Zea were similar with H1-ERα and WT-ERα; however, none of the EDCs showed activation of AA-ERα; this is consistent with ERE-mediated activity, in which ERE reporter activity in response to BPA, BPAF, and Zea with E1-ERα (AF-1 inactive) was similar to activity with WT-ERα (). In addition, Zea activated the AF-2 mutant to an extent that was similar to ICI, which activates AF-2-ERα via the AF-1 function. Overall, these data indicate that ERE-mediated activation by BPA and BPAF was via the AF-2 function on ERα, whereas Zea activated ERα via both AF-1 and AF-2 functions.
Figure 2 Functional analysis of BPA, BPAF, and Zea on WT and ERα mutants and coactivation of ERα by SRC2 or p300 in Ishikawa cells. (A) For functional analysis, cells transfected with ERE-luc, pRL-TK, and pcDNA/WT ERα, pcDNA/H1 (more ...)
In determining a role for ERα and activation with coactivators, we found that both SRC2 and p300 coactivated ERα-ERE–mediated activity with all three EDCs (). These data indicated that SRC2 and p300 act as ERα coactivators to increase transactivation in the presence of BPA, BPAF, and Zea, consistent with their AF-2 functionality.
BPA and BPAF activate genes via p44/42 MAPK and
src tyrosine kinase pathways, but Zea activates only through the p44/42 MAPK signaling pathway.
To examine phosphorylation events, we investigated the involvement of rapid action responses by BPA-, BPAF-, and Zea-mediated signaling pathways in Ishikawa cells stably expressing the vector control or WT-ERα (Burns et al. 2011
). First, we confirmed ERα expression in Ishikawa/ERα cells by Western blot () and then confirmed ERE-mediated activation by E2
, BPA, BPAF, and Zea using the reporter assay system (). E2
, BPA, BPAF, and Zea induced phospho-p44/42 MAPK in Ishikawa/ERα cells but not in Ishikawa/vec cells (), suggesting ERα-dependent and ligand-dependent activation. In addition, phospho-GSK-3β expression was only weakly induced by E2
and the EDCs, and we observed no increase in phospho-Akt in response to E2
or the EDCs in this model system ().
Figure 3 BPA, BPAF, and Zea affect p44/42 MAPK and src tyrosine kinase pathways in Ishikawa/ERα–stable cells. (A) Detection of ERα protein expression by Western blot in whole cell lysates prepared from Ishikawa/vec or Ishkawa/ERα (more ...)
We next examined the effect of two specific kinase inhibitors—PD 98059 (MAPK inhibitor) and PP2 (src family tyrosine kinase inhibitor)—on BPA-, BPAF- and Zea-mediated expression of progesterone receptor (PR), a classic ER target gene, in Ishikawa/ERα cells. Both inhibitors blocked BPA- and BPAF-mediated endogenous PR gene expression, suggesting that BPA and BPAF are involved in the p44/42 MAPK and tyrosine kinase src pathways (). In contrast, induction of PR expression by E2 and Zea was inhibited by PD 98059 but not by PP2. Thus, while all three EDCs appeared to activate the p44/42 MAPK pathway in an ER-dependent manner, other kinase signaling pathways, such as the tyrosine kinase src, may also involve BPA and BPAF activation.
BPA, BPAF, and Zea induced expression of ER target genes.
We confirmed ERα-dependent responses to BPA, BPAF, and Zea by detecting endogenous gene expression of PR
(gene regulation by estrogen in breast cancer 1), MCM3
(minichromosome maintenance complex component 3), and SPUVE
(also known as PRSS23
; a member of the trypsin family of serine proteases) by real time-PCR (). In Ishikawa/ERα cells, E2
and all three EDCs induced endogenous PR
expression was induced only by E2
and BPA, whereas SPUVE
expression was weakly induced by BPA and Zea but not induced by E2
or BPAF. Low-dose BPA (1 or 10 nM) antagonizes E2
expression [see Supplemental Material, Figure 3
)]. In addition, we confirmed that expression of PR
, and SPUVE
was not induced in Ishikawa/vec cells treated with the three EDCs (data not shown). These data demonstrate that BPA, BPAF, and Zea show direct compound-specific gene regulation in an ER-dependent manner.
Figure 4 BPA, BPAF, and Zea regulate expression of the ER target genes PR, GREB, MCM3, and SPUVE in Ishikawa/ERα cells. Ishikawa/ERα cells were treated with vehicle (control), 10 nM E2, or 100 nM BPA, BPAF, or Zea for 18 hr and total RNA was extracted; (more ...)