Uterine epithelial cell estrogen receptor alpha-dependent and -independent genomic profiles that underlie estrogen responses in mice

Abstract

Estrogens exert their activity through estrogen receptor alpha (ERalpha) to stimulate hypertrophy and hyperplasia in the uterus. A uterine epithelial ERalpha conditional knockout mouse model (Wnt7a(Cre+);Esr1(f/f) or cKO) demonstrated that ERalpha in the epithelial cells was dispensable for an initial uterine proliferative response to 17beta-estradiol (E2) but required for subsequent uterine biological responses. This study aimed to characterize the differential gene expression patterns induced by E2 in the presence or absence of epithelial ERalpha. RNA microarray analysis revealed that approximately 20% of the genes differentially expressed at 2 h were epithelial ERalpha independent, as they were preserved in the cKO uteri. This indicates that early uterine transcripts mediated by stromal ERalpha are sufficient to promote initial proliferative responses. However, more than 90% of the differentially expressed transcripts at 24 h were not regulated in the cKO, indicating that the majority of later transcriptional regulation required epithelial ERalpha, especially those involved in mitosis. This shows that loss of regulation of these later transcripts results in blunted subsequent uterine growth after 3 days of E2 treatment. Additionally, progesterone's ability to inhibit E2-induced epithelial cell proliferation was impaired, consistent with a uterine receptivity defect that contributes to cKO infertility. These transcriptional profiles correlate with our previously observed biological responses, in which the initial proliferative response is independent of epithelial ERalpha and thus dependent on stromal ERalpha, yet epithelial ERalpha is essential for subsequent tissue responsiveness.

Keywords: epithelium; estradiol/estradiol receptor; proliferation; stroma; uterus.

FIG. 1

Mice lacking ERα in uterine epithelial cells showed no alteration in uterine ERβ expression. ERα (A) and ERβ (B) immunohistochemistry (IHC) of WT control littermate (WT) and cKO animals. Representative images of uterine cross section (A) and ovarian and uterine cross section (B) are shown. Both WT and cKO ovarian sections, ERβ protein was detected in the granulosa cells. GE, glandular epithelium; LE, luminal epithelium; GC, granulosa cells. Bar = 100 μm.

FIG. 2

Loss of uterine epithelial ERα led to a blunted late-phase uterine proliferative response after E₂ treatment. A) H&E staining of uterine sections from WT and cKO mice treated with vehicle or E₂ for 24 or 72 h. Adult females were ovariectomized and treated with E₂ (0.25 μg/mouse) for 24 h or every day consecutively for 3 days (72 h). Uterine sections were stained with H&E, and epithelial cell heights were measured using ImagePro software as indicated in Materials and Methods. Yellow lines indicate examples of epithelial cell height measured. B) Uterine epithelial cell height after E₂ treatment of WT and cKO animals. Fold increase of uterine epithelial cell heights from each group was calculated relative to WT vehicle-treated group (arbitrary unit). Bar graphs represent mean ± SEM, n = 4–7 mice/group. ** and ***P < 0.01 and 0.001, respectively; significant difference when compared to vehicle-treated group within genotype. ns, no significant difference when compared between WT E₂ 24 h and cKO E₂ 24 h. ^#P < 0.0001, significant difference when compared between WT E₂ 72 h and cKO E₂ 72 h.

FIG. 3

Microarray analysis of uterine transcripts from WT versus cKO treated with vehicle, E₂ 2 h, or E₂ 24 h. A) Unsupervised hierarchical clustering of all replicates from all genotypes and treatment groups using Partek Genomic Suites with the following criteria: raw intensity of at least one sample of each probe ≥ 100, gene is considered differentially expressed when P < 0.01, |fold change| ≥ 2, n = 3 animals/group. Green or red represent normalized expression that is less or greater than the mean of all conditions, respectively. B and C) Venn diagram of transcripts differentially expressed relative to vehicle in WT and cKO at (B) 2 h and (C) 24 h after E₂ treatment using similar statistical criteria as A. Numbers in overlapping region are epithelial ERα-independent transcripts. Transcripts differentially expressed only in WT group are epithelial ERα-dependent transcripts.

FIG. 4

Uterine epithelial ERα-independent transcripts 2 (A) and 24 (B) h after E₂ treatment. Validated gene expression using real-time PCR analysis of uterine samples from WT and cKO treated with E₂ and collected 2 (A) or 24 (B) h after the treatment, n = 4–7 animals/group. Bar graphs represent mean ± SEM. *, **, and ***P < 0.05, 0.01, and 0.001, respectively; significant difference when compared to the vehicle-treated group within genotype.

FIG. 5

Uterine epithelial ERα-dependent transcripts 2 (A) and 24 (B) h after E₂ treatment. Validated gene expression using real-time PCR analysis of uterine samples from WT and cKO treated with E₂ and collected 2 (A) or 24 (B) h after the treatment, n = 4–7 animals/group. Bar graphs represent mean ± SEM. *, **, and ***P < 0.05, 0.01, and 0.001, respectively; significant difference when compared to the vehicle-treated group within genotype.

FIG. 6

Epithelial ERα mediates P₄ inhibition of E₂-induced epithelial cell proliferation. A) Uterine weight increased in WT and cKO ovariectomized animals after a series of E₂ and P₄ with nidatory E₂ treatments (called “E+Pe”), as described in Materials and Methods, compared to vehicle control, n = 5–8 animals/group. Bar graphs represent mean ± SEM. ***P < 0.001; significant difference when compared to vehicle control within genotype. ^#P < 0.0001, significant difference when compared between WT E+Pe and cKO E+Pe. B–D) Immunohistochemical analyses and quantification of Ki-67 (B and C) and phosphohistone H3 (D and E) in WT and cKO uterine cross sections. Representative images are shown (n = 5). Bar = 100 μm. C and E) Cells with positive staining of Ki-67 and phosphohistone H3 in the E+Pe-treated groups were counted (detail described in Materials and Methods) and presented as percentage of total cell count (luminal epithelial vs. stroma). Ki-67-positive cells were quantified from the whole uterine cross section. Phosphohistone H3-positive cells were quantified from the luminal and subepithelial stromal cells area. * and ***P < 0.05 and 0.001, significant difference between designated groups. ns, no significant difference.

FIG. 7

Epithelial ERα mediates A) Lif and Ihh induction in WT and cKO ovariectomized animals after E+Pe treatments. Bar graphs represent mean ± SEM, n = 5–8 animals/group. * and ***P < 0.05 and 0.001, respectively; significant difference when compared to vehicle control within genotype. Loss of epithelial ERα does not alter expression pattern of PR (B) and HAND2 (C) when treated with E+Pe. Representative images are shown (n = 5).

FIG. 8

Working model illustrating epithelial ERα function during uterine proliferation in response to E₂. A) Epithelial ERα-independent, early transcriptional phase (2 h) of E₂-induced initial epithelial cell proliferation (24 h). E₂ activates stromal ERα to produce and secrete mitogenic signals (such as growth factors and potentially other molecules shown in Supplemental Table S1) and stimulate uterine epithelial cell proliferation. This initial proliferative event is not regulated by epithelial ERα. B) Epithelial ERα-dependent, late transcriptional phase (24 h) of E₂-induced subsequent epithelial cell growth and proliferation (72 h). Both epithelial and stromal ERα are pivotal for a maximal uterine biological response regulated by E₂. At this stage, epithelial ERα may induce transcripts needed to generate autocrine signals or provide positive feedback to the stromal cells to induce gene products involved in mitosis and cell cycle progression as well as inhibition of apoptosis that led to an ultimate physiological end point induced by E₂. C) Inhibition of epithelial cell proliferation and induction of implantation markers during the uterine receptivity window require the expression of epithelial ERα. Stromal cell proliferation during this window appears to be independent from epithelial ERα expression.