Bisphenol-A exposure in utero leads to epigenetic alterations in the developmental programming of uterine estrogen response

Abstract

Bisphenol-A (BPA) is a nonsteroidal estrogen that is ubiquitous in the environment. The homeobox gene Hoxa10 controls uterine organogenesis, and its expression is affected by in utero BPA exposure. We hypothesized that an epigenetic mechanism underlies BPA-mediated alterations in Hoxa10 expression. We analyzed the expression pattern and methylation profile of Hoxa10 after in utero BPA exposure. Pregnant CD-1 mice were treated with BPA (5 mg/kg IP) or vehicle control on d 9-16 of pregnancy. Hoxa10 mRNA and protein expression were increased by 25% in the reproductive tract of mice exposed in utero. Bisulfite sequencing revealed that cytosine-guanine dinucleotide methylation was decreased from 67 to 14% in the promoter and from 71 to 3% in the intron of Hoxa10 after in utero BPA exposure. Decreased DNA methylation led to an increase in binding of ER-alpha to the Hoxa10 ERE both in vitro as and in vivo as determined by EMSA and chromatin immunoprecipitation, respectively. Diminished methylation of the ERE-containing promoter sequence resulted in an increase in ERE-driven gene expression in reporter assays. We identify altered methylation as a novel mechanism of BPA-induced altered developmental programming. Permanent epigenetic alteration of ERE sensitivity to estrogen may be a general mechanism through which endocrine disruptors exert their action.

Figure 1

Targeted sequences of the HOXA10 gene. A) The 5′ promoter region, containing 20 CpG sites. B) Intron region, containing 9 sites.

Figure 2

Hoxa10 mRNA expression in mice that had been exposed to BPA in utero relative to controls, as demonstrated by quantitative real time RT-PCR. Hoxa10 mRNA expression is increased in the uteri of BPA-treated mice. *P = 0.02.

Figure 3

Immunohistochemical analysis shows that BPA exposure (5 mg/kg) alters Hoxa10 expression in 2-wk-old female mice exposed in utero. A) Normal Hoxa10 expression in the uterus of vehicle-exposed mice. B) BPA-treated mice showed increased expression of Hoxa10 in the uterus compared to controls. C) Negative control (no primary antibody).

Figure 4

A) Bisulfite conversion, cloning, and sequencing of the Hoxa10 promoter and intron. Average percentage methylation of each CpG site is presented. B) Methylation level of the Hoxa10 gene promoter and intron in BPA and vehicle control-treated mice. In the promoter, BPA treatment in utero led to decreased mean number of methylated sites compared to controls (67% control vs. 14% after BPA treatment, P=0.007). Similarly, BPA treatment significantly reduced intron methylation as well (71% control and 3% after BPA treatment, P=0.001).

Figure 5

Expression of DNMTs in 2-wk-old mice after in utero exposure to BPA. Real time RT-PCR showed no significant persistent changes in mRNA expression of DNMT1, DNMT3a, or DNMT3b after exposure to BPA relative to control.

Figure 6

Effect of DNA methylation on HOXA10-promoter transcription-factor binding. Left panel: ER-α-containing nuclear extract bound the HOXA10 ERE. Labeled DNA without nuclear extract fails to produce a shift (lane 1). Addition of ER-containing nuclear extract produces a robust shift (lane 2). This shifted complex is competed by unlabeled ERE (lane 3). Specificity of the DNA-protein interaction is suggested by the loss of binding seen when ER-α antibody was added to the binding reaction (lane 4), but not with the nonspecific IgG (lane 5). Right panel: Methylated oligonucleotide probe (M) failed to bind to ER. In contrast, the unmethylated oligonucleotide (UN) bound strongly to ER. NE, nuclear extract; Comp, unlabeled competitor ERE oligonucleotide; Ab, antibody; FP, free probe; NS, nonspecific shift.

Figure 7

ChIP reveals enhanced ER-α binding to the Hoxa10 ERE in vivo after BPA treatment in utero. Uterine tissue was used from BPA-treated and vehicle control-treated mice. After immunoprecipitation and amplification, PCR products were separated on a 3.0% gel; representative gel is shown. Lane 1, 100-bp ladder; lane 2, vehicle control; lane 3, BPA; lane 4, 5% input for vehicle control; lane 5, 5% input for BPA group; lane 6, normal mouse IgG as negative control; lane 7, anti-RNA polymerase as positive control; lane 8, water control.

Figure 8

Effect of DNA methylation in Hoxa10-promoter response to estradiol. Transfection of MCF-7 cells with a luciferase reporter construct containing the Hoxa10 ERE sequence showed the expected response to the addition of 10⁻⁶ M estradiol. However, methylation of the sequence prior to transfection showed a loss of response to the addition of estradiol (P<0.005). C, control plasmid without ERE; ERE, response after transfection of ERE-containing reporter plasmid in the absence of estradiol treatment; ERE+E, response after transfection accompanied by estradiol treatment.