Genetic deletion of the repressor of estrogen receptor activity (REA) enhances the response to estrogen in target tissues in vivo

Abstract

We previously identified a coregulator, repressor of estrogen receptor activity (REA), that directly interacts with estrogen receptor (ER) and represses ER transcriptional activity. Decreasing the intracellular level of REA by using small interfering RNA knockdown or antisense RNA approaches in cells in culture resulted in a significant increase in the level of up-regulation of estrogen-stimulated genes. To elucidate the functional activities of REA in vivo, we have used targeted disruption to delete the REA gene in mice. The targeting vector eliminated, by homologous recombination, the REA exon sequences encoding amino acids 12 to 201, which are required for REA repressive activity and for interaction with ER. The viability of heterozygous animals was similar to that of the wild type, whereas homozygous animals did not develop, suggesting a crucial role for REA in early development. Female, but not male, heterozygous animals had an increased body weight relative to age-matched wild-type animals beginning after puberty. REA mRNA and protein levels in uteri of heterozygous animals were half that of the wild type, and studies with heterozygous animals revealed a greater uterine weight gain and epithelial hyperproliferation in response to estradiol (E2) and a substantially greater stimulation by E2 of a number of estrogen up-regulated genes in the uterus. Even more dramatic in REA heterozygous animals was the loss of down regulation by E2 of genes in the uterus that are normally repressed by estrogen in wild-type animals. Mouse embryo fibroblasts derived from heterozygous embryos also displayed a greater transcriptional response to E2. These studies demonstrate that REA is a significant modulator of estrogen responsiveness in vivo: it normally restrains estrogen actions, moderating ER stimulation and enhancing ER repression of E2-regulated genes.

FIG. 1.

Design of targeting vector and targeting strategy for disruption of the REA gene. The targeting vector designed for disruption of the REA gene contained 5.2-kb (5′ targeting arm) and 6.9-kb (3′ targeting arm) mouse REA and neighboring genomic sequences. The neomycin resistance gene flanking loxP sequences (loxP-PGK-neo) was used as a positive selection marker, and the HSV-TK gene was used as a negative drug selection marker. The pBluescript II SK (+) plasmid (Stratagene) was used as the backbone vector. After homologous recombination on the REA genomic locus, the XhoI restriction enzyme sites within the REA gene were deleted, causing a different fragmentation pattern with XhoI treatment. 3′ and 5′ probes for Southern hybridization can detect the difference in fragmentation. Primer set P1 and P2 and primer set P3 and P4 were designed for genotyping analysis.

FIG. 2.

Isolation of ES cell clones and genotyping analysis by Southern hybridization (A) and PCR (B). The ES cells were selected by resistance to G418 (positive selection) and by growth on FIAU-containing media (negative selection) that excluded the random integration of the plasmid DNA. Isolated ES cell clones were analyzed by Southern hybridization to establish that a homologous recombination event that interrupted the REA gene had occurred. Mouse genomic DNA extracted from tail samples of all pups was genotyped by both PCR and Southern blotting and digested with XhoI. Two primer sets were designed: one is for detecting the deleted part of REA and another set is for detecting the 5′ arm and neo gene. 3′ and 5′ probes are identical to probes used for ES cell Southern screening. WT, wild type; KO, knockout.

FIG. 3.

Analysis of growth in wild-type and REA heterozygous mice. (A) Growth curve of females. REA^+/− female mice grow significantly faster than the wild-type female mice. Values are means ± SEM of at least 13 mice each. *, significantly different from age-matched wild-type females (P < 0.05). (B) Growth curve of males. The growth of REA^+/− male mice is similar to that of wild-type males. These mice have an outbred genetic background. Values are means ± SEM of at least 13 mice each.

FIG. 4.

E2 responsiveness of the uterus (A) and REA mRNA (B) and REA protein levels (C) in wild type (+/+) and REA heterozygous (+/−) mice. Immature (day 21) mice were injected s.c. daily with control vehicle or 0.5 μg of E2 for 4 days. Uterine weight (A) was measured 24 h after the last injection and adjusted to each animal's body weight. *, P < 0.05. RNA was extracted from the uteri with Trizol, and the REA expression level was measured by real-time PCR. (B) The REA mRNA level was calculated relative to that of the vehicle-treated +/+ control, which was set at 1. (C) REA protein level in uteri from four mice of each genotype. Each lane contains uterine protein from a separate mouse of the indicated genotype. REA was detected by Western immunoblotting with REA polyclonal antibody. The numbers indicate the relative intensities of bands analyzed with ImageQuant software. β-Actin was used as a loading control.

FIG. 5.

Histological analysis of control and E2-treated uteri of REA heterozygous (+/−) and wild type (+/+) mice. Immature REA +/+ and +/− mice (21 days old) were injected s.c. daily with control vehicle or 0.5 μg of E2 for 4 days. At 24 h after the last hormone injection, tissue samples were collected, fixed, and stained with hematoxylin and eosin as described in Materials and Methods. (A) Control vehicle-treated uteri of wild type (+/+) mice; (B) control vehicle-treated uteri of REA heterozygous (+/−) mice; (C) E2-treated uteri of wild-type (+/+) mice; (D) E2-treated uteri of REA heterozygous (+/−) mice. Magnification, ×40.

FIG. 6.

Modulation of E2-up-regulated and E2-downregulated gene expression in uteri of REA wild-type (+/+) and heterozygous (+/−) mice. RNA was isolated from uteri of animals that received control vehicle or E2 treatments as described in the legend to Fig. 5. Levels of complement component C3, lactoferrin, G6PDH, and PTα were examined by real-time PCR, and changes were calculated relative to the vehicle-treated controls. Levels of the downregulated genes PR, AR, and ERα were also examined by real-time PCR, and changes were calculated relative to the vehicle-treated controls.

FIG. 7.

Enhancement of ER-driven transcriptional activation in REA heterozygous MEFs and examination of REA mRNA and protein levels in wild-type and heterozygous MEFs. (A) MEFs of each genotype were transfected with ERα (10 ng), (ERE)₂-pS2-Luc reporter (1 μg), and β-galactosidase (300 ng) and treated with control vehicle (C) or the indicated concentration of E2 for 24 h, and luciferase activities were measured. *, significantly different from wild-type (+/+) MEFs at the same concentration of E2 (P < 0.05); †, P = 0.2. All values are means ± SEM and have been normalized for β-galactosidase activity. (B) MEFs were derived from either wild-type or REA heterozygous embryos as described in Materials and Methods. Total mRNA was isolated by using Trizol, and the REA mRNA level was measured by real-time PCR. (C) REA protein level was monitored with REA polyclonal antibody. β-Actin, used as a loading control, was monitored by using β-actin antibody. Numbers indicate the relative intensities of bands analyzed with ImageQuant software.