COUP-TFII mediates progesterone regulation of uterine implantation by controlling ER activity

Abstract

Progesterone and estrogen are critical regulators of uterine receptivity. To facilitate uterine remodeling for embryo attachment, estrogen activity in the uterine epithelia is attenuated by progesterone; however, the molecular mechanism by which this occurs is poorly defined. COUP-TFII (chicken ovalbumin upstream promoter transcription factor II; also known as NR2F2), a member of the nuclear receptor superfamily, is highly expressed in the uterine stroma and its expression is regulated by the progesterone-Indian hedgehog-Patched signaling axis that emanates from the epithelium. To further assess COUP-TFII uterine function, a conditional COUP-TFII knockout mouse was generated. This mutant mouse is infertile due to implantation failure, in which both embryo attachment and uterine decidualization are impaired. Using this animal model, we have identified a novel genetic pathway in which BMP2 lies downstream of COUP-TFII. Epithelial progesterone-induced Indian hedgehog regulates stromal COUP-TFII, which in turn controls BMP2 to allow decidualization to manifest in vivo. Interestingly, enhanced epithelial estrogen activity, which impedes maturation of the receptive uterus, was clearly observed in the absence of stromal-derived COUP-TFII. This finding is consistent with the notion that progesterone exerts its control of implantation through uterine epithelial-stromal cross-talk and reveals that stromal-derived COUP-TFII is an essential mediator of this complex cross-communication pathway. This finding also provides a new signaling paradigm for steroid hormone regulation in female reproductive biology, with attendant implications for furthering our understanding of the molecular mechanisms that underlie dysregulation of hormonal signaling in such human reproductive disorders as endometriosis and endometrial cancer.

Figure 1. COUP-TFII Is Efficiently Deleted by PR-Cre, and Resulting Mutants Are Infertile

However, COUP-TFII is not deleted in the ovary and, thus, mutants exhibit no ovarian defects. (A–B) Immunohistological detection of COUP-TFII in the uterus. (A) Control, 4.5 dpc. COUP-TFII is highly expressed in the endometrial stroma, but is undetectable in the epithelial compartment. (B) Mutant, 4.5 dpc. COUP-TFII expression is efficiently ablated in the mutant uterus. (C–D) Immunohistological detection of COUP-TFII in the ovary. (C) Control, 4.5dpc. COUP-TFII is highly expressed in the theca cell layer. (D) Mutant, 4.5dpc. COUP-TFII expression is maintained in the theca cell layer. (E) The gene expression of COUP-TFII, assayed by quantitative real-time RT-PCR. COUP-TFII expression is efficiently ablated in the uterus but not affected in the ovary. White bar: control; black bar: mutant. *p < 0.001 (t-test, n = 9). (F) Summary of breeding studies. PR^Cre/+ COUP-TFII^flox/flox mutants are infertile, while both controls are normal in fertility.

Figure 2. COUP-TFII Mutants Are Defective in Embryo Attachment

PR^Cre/+ COUP-TFII^flox/flox mutants have implantation failure. (A) Control, 4.5 dpc. Implantation sites are visualized as blue spots. (B) Mutant, 4.5 dpc. Mutant uterine horns have no implantation sites and appear small in size. (C) Summary of implantation study. (D) Control, hematoxylin and eosin staining (HE), 4.5 dpc. Embryo is attached to the uterine lumen and the surrounding stroma is permeated as an initial decidual response. (E) Mutant, HE, 4.5 dpc. No embryos are attached to the mutant uterine lumen. A total of 20 embryos were examined in each group.

Figure 3. COUP-TFII Mutants Are Defective in Decidualization

(A–E) PR^Cre/+ COUP-TFII^flox/flox mutants have decidualization failure. (A) Control, 48 h after stimuli. The right horn (R) was stimulated, and the left horn (L) was unstimulated. Only right horn is decidualized. (B) Mutant, 48 h after stimuli. Neither right (stimulated) nor left (unstimulated) horn is decidualized. (C) Control, right horn. Alkaline phosphatase activity is detected in the decidualized stroma. (D) Mutant, right horn. No alkaline phosphatase activity is detected in the mutant uterus. (E) Ratio of right to left horn in weight. Mutant uterine horns fail to decidualize. White bar: control; black bar: mutant. *p < 0.001 (t-test, n = 6). k–o, cell proliferation is altered in PR^Cre/+ COUP-TFII^flox/flox mutants. (F) Control, phospho-H3, 3.5 dpc. (G) Mutant, phospho-H3, 3.5 dpc. Some positive cells in the epithelium of the mutant are marked by an arrowhead. (H–I) Percentage of phospho-H3–positive cells in the stroma (H) and in the epithelia (I). The number of cells were counted on multiple sections and averaged in each mouse. Stromal cell proliferation is decreased, while epithelial proliferation is enhanced in the mutant uterus. White bar: control; black bar: mutant. *p = 0.007; **p = 0.001 (t-test, n = 7). (J–K) PR^Cre/+ COUP-TFII^flox/flox mutants have an angiogenesis defect. (J) Control, 3.5 dpc. Lectin was intravenously administered and adhered to the inner lumen of vessels. (K) Mutant, 3.5 dpc. Vascular density is apparently reduced in the mutant uterus.

Figure 4. BMP2 Mediates COUP-TFII in Decidualization

(A–C) BMP2 induction is diminished in the mutant uterus. (A) The gene expression of Bmp2 in the stimulated horn (D) and the unstimulated horn (ND), assayed by quantitative real-time RT-PCR at 48 h after stimuli. Bmp2 expression is induced by stimuli for decidualization in the control, but is not sufficiently induced in the mutant uterus. White bar: control; black bar: mutant. *p < 0.001, (t-test, n = 6). (B) Control, right horn, 72 h after stimuli. BMP2 is immunohistologically detected in the secondary decidual zone. (C) Mutant, right horn, 72 h after stimuli. BMP2 is not detected in the mutant stroma. (D–I) Decidualization is rescued by the administration of recombinant human BMP2. (D) Control, 48 h after stimuli. The right horn was treated with BMP2, and the left horn was treated with BSA (vehicle). Both horns are decidualized. (E) Control, right horn. (F) Control, left horn. Alkaline phosphatase activity is detected in both uterine horns. (G) Mutant, 48 h after stimuli. Only the right horn is decidualized. (H) Mutant, right horn. (I) Mutant, left horn. Alkaline phosphatase activity is detectable in the BMP2-treated horn, but not in the BSA-treated horn of mutant uterus.

Figure 5. COUP-TFII Regulates ER Activity in the Epithelia

Estrogen-responsive genes are upregulated in the mutant uterine epithelia. (A) The gene expression of Ltf, assayed by qRT-PCR at 3.5 dpc. Ltf expression is high in the mutant uterus. White bar: control; black bar: mutant. *p = 0.012 (t-test, n = 6). (B) Ltf, qRT-PCR, 30 hPe. Ltf expression is consistently high in the mutant uterus. White bar: control; black bar: mutant. *p < 0.001 (t-test, n = 9). (C) Control, immunohistological detection of lactoferrin, 30 hPe. (D) Mutant, lactoferrin. (E) Control, DAPI. (F) Mutant, DAPI. Upregulated expression of lactoferrin is observed in the epithelial compartment. (G) The gene expression of C3, assayed by qRT-PCR at 30 hPe. C3 expression is high in the mutant uterus. White bar: control; black bar: mutant. *p = 0.015 (t-test, n = 9). (H) Clca3, qRT-PCR, 30 hPe. Clca3 expression is high in the mutant uterus. White bar: control; black bar: mutant. *p = 0.002 (t-test, n = 9).

Figure 6. Uterine Receptivity Is Disrupted by High Estrogen Activity in the Mutant Uterine Epithelia

(A) The gene expression of Muc1, assayed by qRT-PCR at 30 hPe. Muc1 expression is high in the mutant uterus. White bar: control; black bar: mutant. *p = 0.041 (t-test, n = 9). (B) Control, immunohistological detection of MUC1, 30 hPe. (C) Mutant, MUC1, 30 hPe. MUC1 is detected in the apical surface of uterine epithelia, and its expression level is high in the mutants. (D) Control, EM, 30 hPe. Membrane flattening is observed in the surface of uterine epithelia. (E) Mutant, EM, 30 hPe. Mutant uterine epithelia fail to undergo membrane flattening. (F) Control, EM, 30 hPe. (G) Mutant, EM, 30 hPe. Mutant microvilli are excessively coated with glycocalyx (arrow). (H) The gene expression of Galnt1 and Galnt7, assayed by qRT-PCR at 30 hPe. The expression of both enzymes is high in the mutant uterus. White bar: control; black bar: mutant. **p = 0.008, **p < 0.001 (t-test, n = 9). (I) Control, EM, 30 hPe. (J) Mutant, EM, 30 hPe. Mutant uterine epithelia exhibit the persistent presence of desmosome (arrow). (K) The gene expression of Dsc2, assayed by qRT-PCR at 30 hPe. Dsc2 expression is high in the mutant uterus. White bar: control; black bar: mutant. *p = 0.001 (t-test, n = 9).

Figure 7. COUP-TFII Regulates the Expression of ERα and SRC-1 in the Uterine Epithelia

(A–B) Immunohistological detection of PR at 30 hPe. (A) Control (PR^Cre/+ COUP-TFII^+/+) (B) Mutant (PR^Cre/+ COUP-TFII^flox/flox). The expression of PR is significantly reduced in the stroma of COUP-TFII mutant uterus. The same result is observed under another comparison between COUP-TFII^flox/flox and PR^Cre/+ COUP-TFII^flox/flox). (C) The gene expression of Esr1, assayed by qRT-PCR at 30 hPe. ERα expression is increased in the whole mutant uterus. White bar: control; black bar: mutant. *p = 0.012 (t-test, n = 9). (D) Immunohistological detection of ERα at 30 hPe of control uterus. (E) ERα expression is increased in the mutant uterus. (F) Western blot analysis of ERα, phospho-ERα, and SRC-1 in isolated luminal epithelial cells. Intensity of signals was measured by National Institutes of Health image software (http://rsb.info.nih.gov/nih-image) and normalized by cyclophilin-A. ERα, phospho-ERα, and SRC-1 were increased 2.5-fold, 2.4-fold, and 1.6-fold in the isolated luminal epithelia (LE) of the mutant uterus, respectively. *Nonspecific band. Arrow indicates the pERα, which is confirmed by reprobing with anti-ERα antibody. (G) Immunohistological detection of phospho-ERα at 30 hPe of control uterus. (H) Level of pERα is increased in the mutant uterine epithelia. (I) Immunohistological detection of SRC-1 at 30 hPe of control uterus. (J) Increased SRC-1 is detected in the epithelial compartment of mutant uterus.

Figure 8. Working Model of COUP-TFII in Mediating Progesterone Function in the Uterus

Progesterone activates Ihh–Ptch signaling to induce COUP-TFII expression in the stroma compartment. COUP-TFII, in turn, regulates stromal cell differentiation (decidualization) through the induction of BMP2. COUP-TFII also mediates the suppression of epithelial estrogen activity through inhibiting the expression of SRC-1 and ERα as well as ERα activation, which allows the uterine epithelia to undergo transformation and gain receptivity for embryo attachment.