MicroRNA-200a Locally Attenuates Progesterone Signaling in the Cervix, Preventing Embryo Implantation

Abstract

Although cervical pregnancy and placenta previa, in which the embryo and placenta embed in or adjacent to the cervix, are life-threatening complications that result in massive bleeding and poor pregnancy outcomes in women, the incidence of these aberrant conditions is uncommon. We hypothesized that a local molecular mechanism is normally in place to prevent embryo implantation in the cervix. The ovarian hormones progesterone (P₄) and estrogen differentially direct differentiation and proliferation of endometrial cells, which confers the receptive state for implantation: P₄ dominance causes differentiation of the luminal epithelium but increases stromal cell proliferation in preparation of the uterus for implantation. In search for the cause of cervical nonresponsiveness to implantation, we found that the statuses of cell proliferation and differentiation between the uterus and cervix during early pregnancy are remarkably disparate under identical endocrine milieu in both mice and humans. We also found that cervical levels of progesterone receptor (PR) protein are low compared with uterine levels during this period, and the low PR protein levels are attributed to elevated levels of microRNA(miR)-200a in the cervix. These changes were associated with up-regulation of the P₄-metabolizing enzyme 20α-hydroxysteroid dehydrogenase (200α-HSD) and down-regulation of its transcriptional repressor signal transducer and activator of transcription 5 in the cervix. The results provide evidence that elevated levels of miR-200a lead to down-regulation of P₄-PR signaling and up-regulation of (200α-HSD) in the cervix, rendering it nonresponsive to implantation. These findings may point toward not only the physiological but also the pathological basis of the cervical milieu in embryo implantation.

Successful embryo implantation requires an intimate physiological and molecular interaction between the receptive uterus and the implantation-competent blastocyst. This interaction is only initiated when embryonic development to the blastocyst stage is synchronized with the preparation of the endometrium to become receptive for implantation (window of receptivity) (1). Ovarian steroid hormones progesterone (P₄) and estrogen are key regulators that guide the uterus to the receptive state (1). In mice, preovulatory ovarian estrogen secretion induces proliferation of luminal and glandular epithelial cells during the first 2 days of pregnancy (day 1 = vaginal plug). On day 3, the newly formed corpora lutea begin to synthesize P₄. By day 4, the rising P₄ levels superimposed with a low level of preimplantation estrogen secretion result in extensive stromal cell proliferation with differentiation of epithelial cells, conferring the receptive phase. These timely changes in cell-specific proliferation and differentiation accompany the expression of several genes critical for uterine receptivity, followed by blastocyst activation and attachment to the luminal epithelium on the evening of day 4. Thus, P₄ dominance with a small amount of estrogen is critical for proper endometrial preparation and blastocyst implantation (1, 2). P₄, known as the hormone of pregnancy, acts via the nuclear progesterone receptor (PR) for transcriptional activation of genes involved in all stages of pregnancy, including ovulation, fertilization, endometrial receptivity, implantation, decidualization, placentation, and pregnancy maintenance until parturition is initiated (3).

Epidemiological evidence indicates that embryo implantation can occur in or near the cervix under rare circumstances (4–6), suggesting that normally the cervical milieu is not conducive to embryo implantation. In cervical pregnancy, the embryo implants in the lining of the endocervical canal. This event accounts for <1% of ectopic pregnancies (4, 5). Another disorder is placenta previa, a critical perinatal complication that occurs when the embryo implants in the lower segment of the uterus very close to the cervix. Its incidence is approximately 0.4% of births (6). Although these pathological states are relatively rare, the complications are life-threatening, with a high incidence of both maternal and fetal mortality due to massive hemorrhage (7–9). The underlying causes for these high-risk pregnancy complications are unclear and warrant further investigation.

Although the causes of these disorders remain unknown, the underlying mechanism that precludes embryos from implanting in the cervix despite the proximity and the continuity of the reproductive tract also remains elusive. To address this issue, we compared the steroid hormonal responsiveness between the uterus and cervix in both mice and humans in the context of blastocyst implantation. We found that P₄-PR signaling with respect to proliferation, differentiation, and gene expression for receptivity is incompetent in the cervix compared with that in the uterus. Intriguingly, the expression levels of Pgr transcript-encoding PR were similar between the cervix and uterus. To explore the mechanisms underpinning these tissue-specific contributions despite similar ovarian levels of 17β-estradiol (E₂) and P₄ and similar expression of Pgr, we found that the cervical P₄-PR signaling deficiency was directed by microRNA (miR)-200a through translational repression of PR. We also observed increased cervical expression of the P₄-metabolizing enzyme 20α-hydroxysteroid dehydrogenase (20α-HSD) with the decreased expression of its transcriptional mediator signal transducer and activator of transcription (Stat) 5, suggesting translational regulation of cervical PR in addition to heightened local P₄ degradation. Collectively, the results reveal a novel regulatory paradigm: that the deficiency of P₄-PR signaling perhaps renders the cervix incompatible for attaining receptivity for embryo implantation.

Materials and Methods

Mice

C57BL/6 wild-type (WT), CD1 WT, and Fkbp52^−/− mice were used in this study. Fkbp52^−/− mice were originally established by David Smith (Mayo Clinic, Scottsdale, Arizona). WT and Fkbp52^−/− females (2- to 5-month-old) were mated with WT fertile males to induce pregnancy (day 1 = vaginal plug). For RU486 treatment, mice were given a subcutaneous injection of RU486 (8 mg/kg of body weight) on day 3 of pregnancy (9:00 am). To supplement Fkbp52^−/− females with exogenous P₄, they were given a subcutaneous injection of P₄ (2 mg/mouse/d) on days 2 and 3 of pregnancy. Pregnant females were killed on day 3 or 4 (9:00 am) for sample collections and on day 5 (9:00 am) for the evaluation of embryo implantation. All mice used in this investigation were housed in the University of Tokyo Animal Care Facility according to the institutional guidelines for the use of laboratory animals. The experimental procedures were approved by the institutional animal experiment committee.

Human tissues

Endometrial tissues were obtained from women who had shown regular menstrual cycles without any hormonal treatment for at least 3 months. Endometrial samples were dated according to the women's menstrual history and standard histological criteria (10). The experimental procedures were approved by the institutional review board of the University of Tokyo, and signed informed consent for the use of tissues was obtained from each woman.

In situ hybridization

Paraformaldehyde-fixed frozen sections (12 μm) were hybridized with ³⁵S-labeled cRNA probes as described previously (11). Mouse-specific cRNA probes for Areg, Hdc, Ihh, and Ltf were used for hybridization.

Immunohistochemistry

Immunostaining was performed on formalin-fixed paraffin-embedded sections. Antibodies to Ki67 (Thermo Scientific), proliferating cell nuclear antigen (DAKO), PR (Abcam), estrogen receptor-α (ERα) (Abcam), phosphorylated Stat5 (Abcam), and total Stat5 (Bioworld Technology) were used.

Western blotting

Protein extraction and Western blotting were performed as described previously (12). Antibodies to actin (Santa Cruz Biotechnology), PR (Cell Signaling Technology), ERα, 20α-HSD (a gift from Geula Gibori, University of Illinois at Chicago, Illinois), phosphorylated Stat5, and total Stat5 were used. All antibodies were used at a 1:500 dilution. Bands were visualized by using an ECL Prime detection system (GE Healthcare). Actin served as a loading control.

Laser capture microdissection (LCM)

Frozen sections (20 μm) were mounted on PEN (polyethylene naphthalate) slides (Leica Microsystems), fixed in cold acetone, stained in 0.05% toluidine blue, and dehydrated in ethanol. The target regions in the sections were microdissected with a fully motorized LMD7000 system (Leica Microsystems).

RT-quantitative PCR (qPCR)

For conventional RNA analyses, total RNA extraction was performed as described previously (13). The extracted RNA was amplified using an Ovation PicoSL WTA system V2 (NuGEN Technology), and qPCR was performed as described previously (14). A housekeeping gene (β-actin [Actb]) was used as an internal standard for normalizing the relative mRNA expression. For miRNA analyses, miRNA extraction was performed with a NucleoSpin miRNA kit (Macherey-Nagel). RT-qPCR was performed using a Mir-X miRNA qRT-PCR SYBR Kit (Clontech). U6 was used as an internal standard for normalizing the relative miRNA expression.

Cell transfection with miRNA mimics

MCF-7 cells were pretreated with E₂ (10 nM) for 48 hours and then were transfected with mirVana miRNA mimics of miR-200a and controls (Invitrogen) by Lipofectamine RNAiMax transfection regent (Invitrogen) for another 48 hours. All analyses were performed in triplicate.

Luciferase reporter assay

The miRanda database (http://www.microrna.org) was used to identify a putative miR-200a binding site in human PGR. The PGR 3′-untranslated region (UTR) containing a potential miR-200a binding site (8871–8895 nucleotides) was amplified using the following primers: forward, 5′-TTAACTCGAGTTGTGTTGAAGTTGATGCAATCTTC-3′; and reverse, 5′-TTCTAGAATGAGAAAGATTCCCTGCATCTCTT-3′. The amplified fragment was then inserted into the pmirGLO Dual-Luciferase miRNA Target Expression Vector for the 3′-UTR luciferase reporter assay (Promega). For a mutation analysis, a KOD-Plus-Mutagenesis Kit (Toyobo, Osaka, Japan) was used to mutate 3 nucleotides in the putative miR-200a binding site (AGUGUU to AGUCCC). COS-7 cells were cotransfected with the reporter plasmid and the miRNA mimic of miR-200a or a negative control, and then cells were incubated for 24 hours. The extracts were used for the luciferase assay. Transfections and reporter assays were repeated 3 times.

Statistical analysis

Statistical analyses were performed using a two-tailed Student t test. A value of P < .05 was considered statistically significant.

Results

Proliferation and differentiation are differentially regulated in the uterus and cervix

Cell proliferation and differentiation are regulated in a highly coordinated manner during development and tissue homeostasis (15–18). Positive staining of cells for Ki67, a known marker of cell proliferation, indicates cell cycle progression, whereas cellular differentiation is often associated with its absence. As reported previously in mice (1), the uterine luminal epithelium ceases proliferation with the loss of Ki67 staining with the acquisition of epithelial cell differentiation as opposed to stromal cell proliferation on day 4 (the day of uterine receptivity and blastocyst attachment). We examined Ki67 immunostaining in WT cervices and uteri on days 3 and 4 of pregnancy and found that the day 3 uteri exhibited marked epithelial cell proliferation with less stromal cell proliferation, in contrast to the day 4 uteri which showed a transition of epithelial cell proliferation to differentiation with increased stromal cell proliferation (Figure 1A and Supplemental Figure 1A). In contrast, epithelial cell proliferation persisted in the cervix with attenuated stromal proliferation on days 3 and 4 of pregnancy (Figure 1A and Supplemental Figure 1A). These findings suggest that the cervix is unable to acquire receptivity despite exposure to an endocrine milieu appropriate for uterine receptivity. These results were further confirmed by immunostaining of proliferating cell nuclear antigen, another marker for cell proliferation status (Supplemental Figure 2). Taken together, these findings suggest that the appropriate cell proliferation and differentiation status in the endometrium during the periimplantation period is an indicator for receptivity and embryo implantation.

Figure 1.

The endometrial proliferation-differentiation switching is observed in the receptive uterus, not the cervix, and this switching is controlled by P₄-PR signaling. A, Representative photographs of Ki67 immunostaining in the uterus and cervix of WT mice on days 3 and 4 of pregnancy. The dynamic changes in proliferation status were observed only in the uterus. Ten independent samples from different mice were examined in each group. B, Representative photographs of Ki67 immunostaining in the uterus and cervix in the day 4 WT mice treated with vehicle or the PR antagonist RU486. RU486 stopped the appropriate changes in endometrial proliferation in the uterus and did not affect the proliferation status of the cervix. Three independent samples from different mice were examined in each group. C, Ki67 immunostaining of the day 4 uterus and cervix in WT mice and Fkbp52^−/− mice with or without P₄ supplementation. The Fkbp52^−/− uterus did not show the stromal proliferation and the epithelial differentiation, which was restored by P₄. Three independent samples from different mice were examined in each group. Scale bars correspond to 100 μm. le, luminal epithelium; s, stroma.

PR activity is compromised in the cervix

Because the uterine receptivity occurs on day 4 of pregnancy when postovulatory corpora lutea markedly secrete P₄ to execute its function via PR, we examined the role of PR signaling in mouse uterine and cervical cell-specific proliferation status by an injection of a PR antagonist, RU486 (mifepristone), and use of a P₄-resistant mouse model, FKBP52-null (Fkbp52^−/−) mice. With a subcutaneous injection of 8 mg/kg body weight of RU486 on day 3, at which dose WT females show complete implantation failure (n = 3/3, 100%), epithelial cell proliferation persisted without stromal cell proliferation on day 4 (Figure 1B and Supplemental Figure 1B). FKBP52, an immunophilin cochaperone and a part of the steroid receptor complex, is required for optimal PR activity (19). Fkbp52^−/− mice exhibit impaired uterine P₄ responsiveness with exacerbated estrogenic influence, leading to implantation failure (20). However, implantation and full-term pregnancy can be restored with excess P₄ supplementation in these mice on the CD1 background (20). Therefore, the Fkbp52^−/− females are considered to be a suitable model to study P₄ resistance. Again, we found persistent epithelial cell proliferation without much stromal cell proliferation on day 4 in Fkbp52^−/− females (Figure 1C and Supplemental Figure 1C). In contrast, no such changes in the proliferation pattern were noted in the cervix of either WT females after RU486 treatment or in Fkbp52^−/− females with or without P₄ supplementation (Figure 1C and Supplemental Figure 1C). These findings show that P₄-PR signaling regulates uterine, but not cervical, proliferation programming in day 4 pregnant mice.

Human endometrial proliferation and differentiation patterns have signatures similar to those in mouse

In many aspects, the patterns of ovarian hormones during the menstrual cycle in humans is similar to those in mice (21). During the preovulatory proliferative phase, the dominant ovarian hormone is E₂. After an E₂ surge and ovulation, P₄ from the newly formed corpus luteum gradually primes the human reproductive tract toward the secretory phase to support embryo implantation. We examined the status of cell proliferation in the human uterus and cervix by Ki67 immunostaining and found that the luminal epithelial cells had Ki67-positive staining during the proliferative phase in the uterus, cervix, and the transitional zone between these tissues called the isthmus (Figure 2A). During the secretory phase, Ki67 staining in the epithelium in the cervix and the isthmus was still present but was absent in the uterus (Figure 2B). On the other hand, Ki67-positive stromal cells in the uterus increased during the secretory phase (Figure 2B). In this context, the cervix is composed of more than one layer of epithelial cells; however, the isthmus zone has a single layer of epithelial cells similar to that of the uterus, yet still proliferates analogously with the cervix proper and does not normally accommodate embryo implantation. These findings show that the patterns of cell proliferation seen in the human endometrium and cervix are similar to those in the mouse.

Figure 2.

The endometrial proliferation-differentiation switching is observed in the human endometrium. A and B, Representative photographs of Ki67 immunostaining in the human endometrium during the proliferative phase (A) and the secretory phase (B). As for the uterine luminal epithelium, the intensity of uterine Ki67 staining was high in the proliferative phase and low in the secretory phase. In contrast, the intensity of the luminal epithelium in the isthmus and cervix was continuously high. Samples from 10 different individuals were examined. Scale bars correspond to 100 μm. le, luminal epithelium; s, stroma.

PR- and P₄-responsive genes are differentially expressed in the uterus and cervix

To further evaluate the mechanism underlying cervical nonresponsiveness to P_4, we evaluated PR protein levels in the mouse uterus and cervix. Interestingly, our Western blotting results showed that levels of 2 PR isoforms, PR-A and PR-B, were both reduced in the cervix compared with those in the uterus (Figure 3, A and B). Immunohistochemistry results also showed that PR was strongly expressed in the uterus but weakly in the cervix (Figure 3C). In contrast, ERα expression levels were comparable between the cervix and the uterus (Supplemental Figure 3, A–C). With these data in hand, we examined the expression of P₄-regulated genes in the uterine and cervical luminal epithelia isolated by LCM. We performed qPCR of RNA extracted from isolated tissues and found that the P₄-responsive genes amphiregulin (Areg), histidine decarboxylase (Hdc), and Indian hedgehog (Ihh), which are all expressed in the uterine luminal epithelium, were substantially down-regulated in the cervix (Figure 3D), suggesting down-regulation of P₄ responsiveness in the cervix. Our in situ hybridization results were also consistent with Western blotting and qPCR results (Figure 3F). We have previously shown that reduced P₄ signaling enhances estrogenic responses and induces expression of an estrogen-responsive gene, lactotransferrin (Ltf), in the mouse uterine epithelium (22). We also found that the Ltf transcript was up-regulated in the LCM-dissected cervical luminal epithelia (Figure 3, E and G). These results suggest exaggerated estrogenic effects in the cervix due to reduced P₄-PR signaling. We then sought to understand the mechanism underlying lower PR protein levels in the cervix.

Figure 3.

Lower expression of PR protein and P₄-responsive genes in the cervix than in the uterus. A and B, Lower protein levels of the PR isoforms PR-A and PR-B in the cervix than in the WT uterus on day 4 of pregnancy (means ± SEM; *, P < .05), which were evaluated by Western blotting. Band intensities of PR-A and PR-B were normalized against actin. Five independent samples from different mice were used in each group. C, Immunostaining of PR in the WT uterus and cervix on day 4 of pregnancy. Five independent samples from different mice were examined in each group. Scale bar corresponds to 100 μm. le, luminal epithelium; s, stroma. D, mRNA levels of P₄-responsive genes Areg, Hdc, and Ihh in the day 4 luminal epithelium of the WT uterus and cervix, as determined by qPCR. P₄-responsive genes were low in the cervix compared with those in the uterus (means ± SEM; *, P < .05). Luminal epithelia were dissected out by LCM. Ltf mRNA was normalized against Actb. All results are compared between the cervix and uterus and expressed as fold changes. Three independent samples from different mice were used in each group. F, In situ hybridization of P₄-responsive genes Areg, Hdc, and Ihh in the WT uterus and cervix on day 4 of pregnancy. Scale bars corresponds to 500 μm. G, In situ hybridization of Ltf in the WT uterus and cervix on day 4 of pregnancy. Scale bars correspond to 500 μm.

PR is differentially regulated in the uterus and cervix by miR-200a

To evaluate whether the differential expression of PR protein levels between the uterus and cervix is under transcriptional regulation, we quantified mRNA levels of Pgr in the luminal epithelium and stroma in LCM-dissected tissues. Intriguingly, Pgr mRNA levels were comparable in the luminal epithelia of the cervix and uterus in contrast to PR protein levels (Figure 4A). Stromal Pgr mRNA levels were also similar between the uterus and cervix (Figure 4A). Therefore, we postulated that reduced epithelial PR protein levels in the cervix are posttranscriptionally regulated. miRNAs play key roles in the translation from mRNAs to proteins by sequence-specific posttranscriptional gene regulation for protein synthesis (23). Recently, miR-200a has been shown to repress Stat5 and induce the P₄-metabolizing enzyme 20α-HSD in the pregnant myometrium during labor (24), suggesting that miR-200a is an important regulator of P₄-PR signaling in the uterus.

Figure 4.

Heightened miR-200a reduces PR protein levels in the cervix. A, mRNA levels of PR (Pgr) in the luminal epithelium and the stroma of the WT uterus and cervix on day 4 of pregnancy, as determined by qPCR. Epithelial and stromal Pgr mRNA levels were comparable between the uterus and cervix (means ± SEM; P > .05). Pgr mRNA was normalized against Actb and compared between the cervix and uterus as fold changes. Three independent samples from different mice were used in each group. B, Expression levels of miR-200a were high in the day 4 WT cervix compared with those in the uterus (means ± SEM; *, P < .05). The levels of miR-200a were normalized against U6 and compared between the cervix and uterus as fold changes. Seven independent samples from different mice were used in each group. C, MCF-7 cells were pretreated with E₂ (10 nM) for 48 hours and transfected with the miRNA mimics of miR-200a and the control for another 48 hours. Transfection of miR-200a mimics repressed PR protein expression in MCF-7 cells, which was evaluated by Western blotting. Actin served as an internal control. Scrambled miRNA (control mimics) had no effect. Experiments were performed 3 times, and the representative result was demonstrated. D, A luciferase-human PGR 3′-UTR reporter construct with a putative miR-200a binding site (8871–8895 nucleotides) and the sequences of WT and mutated binding sites. E, Transfection of miR-200a mimics diminished PGR 3′-UTR reporter activity in COS-7 cells, which was eliminated by the mutation of the putative miR-200a binding site. Control mimics had no effect on reporter activity. COS-7 cells were transfected with the miRNA mimics of miR-200a and control for 24 hours and then were used for the luciferase reporter assay. The experiments were repeated 3 times (means ± SEM; *, P < .05).

Interestingly, we found elevated miR-200a levels in the cervix compared with those in the uterus (Figure 4B). To determine whether miR-200a regulates PR expression, we assessed the effects of miR-200a on the PR protein expression in vitro. The transfection of MCF-7 cells with mimics of miR-200a significantly decreased the protein levels of both PR isoforms, PR-A and PR-B (Figure 4C and Supplemental Figure 4). To assess whether miR-200a directly targets PGR, we transfected COS-7 cells with miR-200a mimics and a luciferase reporter plasmid comprising a portion of the PGR 3′-UTR, which contained a putative miR-200a binding site subcloned downstream of the luciferase gene (Figure 4D). The heightened expression of miR-200a significantly repressed luciferase reporter activity, but the transfection with control mimics did not show such effects (Figure 4E). In cells transfected with a luciferase reporter construct in which the miR-200a binding site was mutated, this repression disappeared (Figure 4E). These findings show that PR is a direct target of miR-200a.

P₄ is locally metabolized in the cervix

In the myometrium, miR-200a represses Stat5b, and this reduction of Stat5 transcriptionally induces the P₄-metabolizing enzyme 20α-HSD (24). Because our results showed increased miR-200a levels in the cervix, we speculated that Stat5b would be down-regulated with up-regulation of 20α-HSD in the cervix. Indeed we found that Stat5b mRNA levels were lower in the cervix than those in the uterus on day 4 of pregnancy (Figure 5A). In addition, 20α-HSD mRNA levels were significantly increased in the cervix (Figure 5B). The protein levels of both total and phosphorylated Stat5 were decreased, and 20α-HSD levels were inversely increased in the cervix compared with those in the uterus, similar to their transcript expression levels (Figure 5, C and D, and Supplemental Figure 5). These findings provide evidence that increased cervical miR-200a attenuates P₄ signaling not only by down-regulating PR expression but also by metabolizing local P₄. Because the protein levels of PR and Stat5 are low in both the epithelium and stroma of the cervix compared with those in the uterus, we think that miR-200a in both tissue types contributed to reduced expression of PR in the cervix.

Figure 5.

Reduced expression of Stat5 and heightened expression of 20α-HSD in the day 4 cervix compared with those in the uterus. A, Stat5b mRNA levels were low in the day 4 luminal epithelium of the WT cervix compared with those in the uterus (means ± SEM; *, P < .05), as determined by qPCR. Luminal epithelia were dissected out by LCM. Stat5b mRNA was normalized against Actb. Three independent samples from different mice were used in each group. B, 20αHSD mRNA levels were high in the day 4 luminal epithelium of the WT cervix compared with those in the uterus (means ± SEM; *, P < .05), as determined by qPCR. 20α-HSD mRNA was normalized against Actb. Three independent samples from different mice were used in each group. All results were expressed and compared between the cervix and uterus as fold changes. C, Protein levels of phosphorylated and total Stat5 were low and those of 20α-HSD were high in the WT cervix compared with those in the uterus, as determined by Western blotting. Three independent samples from different mice were examined in each group. D, Representative photographs of Stat5 immunostaining in the WT uterus and cervix on day 4 of pregnancy. The immunoreactivity of phosphorylated and total Stat5 was high in the WT cervix compared with those in the uterus. Three independent samples from different mice were examined in each group. Scale bar corresponds to 200 μm. le, luminal epithelium; s, stroma.

Discussion

The present study reveals that the failure of P₄-mediated signaling is a major difference in the cervix compared with the uterus under similar hormonal conditions for receptivity. This difference may contribute to the inability of the cervix to permit embryo implantation under normal pregnancy conditions in both mice and humans. This was reflected in reduced PR signaling activity as evident from reduced PR protein levels and P₄-responsive gene expression in comparison to that seen in the uterus during the receptive phase. These observations are consistent with the fact that P₄-PR signaling is an absolute requirement for embryo implantation in the uterus. Differential P₄-PR signaling in the uterus vs the cervix is intriguing and raises an important question as to the mechanism of tissue-specific regulation of PR activity within the same reproductive tract under similar circulating ovarian hormone levels. Our results of higher levels of miR-200a in the cervix suggest that this miRNA regulates the differential response of the cervix vs the uterus toward P₄-PR signaling by regulating PR protein levels and local P₄ metabolism. This is evident from decreased PR levels by miR-200a with increases in the expression of 20α-HSD and reduction of its transcriptional repressor Stat5 in the cervix. These findings suggest that the miRNA-PR-P₄ signaling is a critical regulator of cervical nonresponsiveness to embryo implantation (Figure 6). We found persistent epithelial proliferation even in the isthmus with a single layer of epithelial cells, whereas the uterine epithelium ceased proliferation with induction of stromal proliferation during the receptive phase.

Figure 6.

A schematic depiction of the potential pathways that support uterine proliferation-differentiation switching and receptivity.

P₄ responsiveness is directed via PR whose activity is modulated by PR isoforms, several steroid receptor coregulators, and a steroid receptor chaperone complex. FKBP52 is required for optimal PR activity, as evident from studies of Fkbp52^−/− mice in which PR activity is compromised with exacerbated estrogenic influence despite normal levels of PR and P₄, resulting in implantation failure (19, 22). However, implantation and full-term pregnancy can be restored in Fkbp52^−/− mice by excess P₄ supplementation on the CD1 background, but not on C57BL6/129 background (22). Interestingly, epithelial cell differentiation and stromal cell proliferation were restored in uteri of CD1 Fkbp52^−/− mice after supplementation with excess P₄. However, this differential response was not present in the cervix of WT or Fkbp52^−/− females even after P₄ supplementation. These findings reinforce the observation that the cervix has consistent P₄ resistance, presumably due to down-regulation of PR and up-regulation of 20α-HSD.

One function of P₄ is to modulate estrogen's actions for the preparation of the uterine lining for implantation. This study showing the reduction in the protein levels of PR and the expression of P₄-responsive genes in the cervix compared with those in the uterus perhaps contributed to higher estrogenic effects in the cervix (22, 25), which is reflected in the higher expression of the E₂-responsive gene Ltf in the cervix on the day of uterine receptivity. Taken together, our results provide for the first time molecular evidence for the nonresponsiveness of the cervix to P₄-PR signaling, which may contribute its inability to allow implantation.

Although embryo implantation in or around the cervix is rare in humans, pathological issues associated with cervical pregnancy and placenta previa are serious and life-threatening due to massive bleeding. Our findings suggest a molecular basis for these disorders involving P₄-PR signaling as a determinant to specify the site of implantation along the reproductive tract. A prior history of curettage or caesarean section is one risk factor for these diseases (8, 9, 13). It is possible that local pathological changes related to previous cervical or uterine surgery endow the cervix with higher P₄-PR signaling and/or reduced levels of miR-200a. It would be interesting to determine whether the deficiency in P₄-PR signaling is reversed in cervical pregnancy and placenta previa. Further investigation is needed to clarify this issue.

Direct degradation and inhibition of the translation of target mRNAs are 2 major biological functions of miRNA. Whereas both PR and Stat5 are targets of miR-200a, our findings and those of others (24) suggest that miR-200a transcriptionally represses Stat5 expression, but not Pgr expression. Thus, miR-200a eventually is likely to suppress P₄-PR signaling through different targets using different approaches. It is possible that miRNAs are expressed in a cell- or tissue-specific manner to affect various target mRNAs. Therefore, further investigation is warranted to elucidate the epithelium and stroma-specific roles of miR-200a in the cervix and uterus.

Most mammalian miRNA is spatiotemporally regulated and influences a variety of biological processes (23). A previous study showed that the expression of cyclooxygenase-2, a gene critical for implantation, is posttranscriptionally regulated by miR-101a and miR-199a, whose expressions were colocalized with cyclooxygenase-2 in the mouse uterus during implantation (26). The present study indicated that miR-200a regulates P₄-PR signaling through the expressions of PR and 20α-HSD in the uterus and cervix during implantation. Our findings suggest that miRNAs spatiotemporally regulate endometrial receptivity and embryo implantation to possibly prevent cervical pregnancy. These results may help to better understand the pathogenesis of cervical pregnancy and placenta previa in humans.

Additional material

Supplementary data supplied by authors.