This is the first report that BPA antagonizes chemotherapeutic agents in both ERα-positive and -negative breast cancer cells. Importantly, unlike some previous studies that have used micromolar concentrations of BPA, we obtained our data using low nanomolar concentrations, which are relevant to human exposure levels. BPA confers chemoresistance to the anticancer drugs doxorubicin, cisplatin, and vinblastine, which act by different mechanisms. As judged by specific ERα/ERβ antagonists, BPA does not appear to mediate its effects through either ERα or ERβ. Given that both cell lines express nonclassical ERs such as GPR30 and members of the ERR family, these could serve as putative BPA receptors. The ability of BPA to alter the expression of Bcl-2 and Bcl-xL suggests a potential mechanism by which it confers chemoresistance in the two breast cancer cell lines.
We postulated that BPA might play a role in chemoresistance based on reports that E2
antagonizes anti cancer drugs. For example, taxol-induced cytotoxicity in MCF-7 breast cancer cells was abrogated by 0.1 μM E2
(Huang et al. 1997
). This was confirmed in a later study implicating JNK activation in the modulation of apoptosis and E2
protection (Razandi et al. 2000
). In addition, E2
antagonizes doxorubicin-induced cytotoxicity in MCF-7 cells (Teixeira et al. 1995
). Our data show that BPA protects T47D cells from several anticancer drugs. More unexpected was the effect of BPA on the estrogen-unresponsive MDA-MB-468 cells, raising the prospect that ERα does not mediate the chemoresistant effects of BPA.
The few reports on the effects of BPA on mitogenesis have used the ER-positive MCF-7 cells. Olsen et al. (2003)
observed increased MCF-7 cell proliferation in response to BPA, with a relative proliferative potential 60,000 times lower than that of E2
. Samuelsen et al. (2001)
further confirmed such effects of BPA; their MCF-7 data are a prime example of an inverted U-shaped curve that is often observed when treating cells with increasing doses of BPA. In that study, cell proliferation was unchanged in the presence of 10 nM BPA, increased > 40% with 100 nM BPA, peaked with 1 μM BPA, and declined at higher doses. These studies are in agreement with our data, which show an approximate 25% increase in cell viability in T47D cells in response to BPA. Despite the lack of a mitogenic effect of BPA in MDA-MB-468 cells, we observed its ability to antagonize the anticancer drugs with as little as 0.01 nM BPA.
Of particular interest is the ability of BPA to antagonize the cytotoxic effects of three chemotherapeutic agents that induce cell death by different mechanisms. Doxorubicin causes DNA damage by chelating metal ions, generating free radicals, and inhibiting topo-isomerase, thereby blocking transcription (Aubel-Sadron and Londos-Gagliardi 1984
). Cisplatin, a platinum-based compound, causes DNA intrastrand cross-linking and inhibits replication (Stewart 2007
). Vinblastine acts by interfering with microtubule dynamics, resulting in mitotic arrest and cell death (Toso et al. 1993
). As mentioned above, E2
protects against microtubule-altering and DNA-damaging drugs (Huang et al. 1997
; Teixeira et al. 1995
). Thus, drugs with different intracellular targets may have a common mechanism for inducing cell death. Future studies should examine whether BPA protects cells from death ligands that induce apoptosis by binding to proapoptotic death receptors.
BPA weakly competes with 17β-E2
in binding to the ER. Using a cell-based transcription assay with a reporter gene, Hiroi et al. (1999)
reported that BPA exhibits agonistic activity when signaling through ERβ but has both agonistic and antagonistic activity when interacting with ERα. Whereas T47D cells express both ERα and ERβ, MDA-MB-468 cells have long been used as a model for ER-negative breast cancer. We show that MDA-MB-468 cells express ERβ protein, whose levels can be modulated by treatment with ICI or PHTPP. Like others (Fan et al. 2003
; Long and Nephew 2006
), we show that ICI rapidly and dramatically degraded the ERα protein, suggesting that the use of ICI is comparable with targeting the receptor with small interfering RNA (siRNA). The finding that BPA exerted its anticytotoxic effects when ERα or ERβ were inhibited suggests that BPA activates a nonclassical ER(s).
Nonclassical ERs include GPR30 and members of the ERR family: ERRα, ERRβ, and ERRγ. BPA binds to GPR30 with a 50% inhibitory concentration (IC50
) of 630 nM, compared with E2
with an IC50
of 17.8 nM (Thomas and Dong 2006
). Interestingly, ICI binds to GPR30 and acts as an agonist (Prossnitz et al. 2007
). Although 17β-E2
does not bind to members of the ERR family, ERRs can bind to functional estrogen-responsive elements in ER target genes (Huppunen and Aarnisalo 2004
). Among the ERRs, BPA binds strongly to ERRγ, with a dissociation constant (KD
) of 5.5 nM, a much more environmentally relevant dose than that needed to bind to ERα or ERβ (Matsushima et al. 2007
). This makes ERRγ the most likely candidate for mediating the protective effects of BPA. Importantly, ERRγ mRNA level was significantly elevated 3.9-fold in breast tumors relative to normal mammary epithelial cells (Ariazi and Jordan 2006
). We found that both T47D and MDA-MB-468 cells express GPR30, ERRα, and ERRγ, whereas ERRβ was undetectable. These data identified potential receptors that should be pursued using approaches such as siRNA to determine which receptor(s) mediates the chemoprotective effects of BPA.
The mechanisms underlying chemoresistance include altered expression of proapoptotic/antiapoptotic proteins, increased activity of membrane transporters such as P-glycoprotein, the status of tumor suppressors, and the efficiency of DNA repair processes. The antiapoptotic Bcl-2 and Bcl-xL proteins and survivin, a prosurvival inhibitor of apoptosis, are major players in tumor growth and resistance to cytotoxic insults. Estrogen increases Bcl-2 protein expression in MCF-7 cells, with cells transfected with Bcl-2 antisense twice as sensitive to doxo rubicin treatment in the presence of estrogen compared with controls (Teixeira et al. 1995
). Another study suggested that increased Bcl-2 in response to estrogen protects cells from taxol-induced cytotoxicity (Huang et al. 1997
). Our data indicate that up-regulation of Bcl-2 and Bcl-xL is a plausible mechanism by which BPA confers resistance to doxorubicin and possibly other anticancer drugs. The survivin data agree with another study that found increased expression of this protein after doxorubicin treatment (Tirro et al. 2006
). However, the contributions of survivin are less critical when proteins such as Bcl-2 and Bcl-xL, which are upstream of survivin, mediate survival.
In conclusion, we have shown that low doses of BPA confer chemoresistance to multiple anticancer drugs, possibly by increasing expression of antiapoptotic Bcl-2 proteins. Importantly, we observed the effects of BPA in a cell line lacking ERα, indicating that BPA acts via nonclassical receptors. These data highlight a previously unrecognized function of BPA in cancer management, thereby adding strong support to the growing recognition of the adverse effects of BPA on human health.