Ovarian regulation of endometrial gland morphogenesis and activin-follistatin system in the neonatal ovine uterus

Postnatal development of the ovine uterus between birth and Postnatal Day (PND) 56 involves differentiation of the endometrial glandular epithelium from the luminal epithelium followed by tubulogenesis and branching morphogenesis. Previous results indicated that ovariectomy of ewes at birth did not affect uterine growth or initial stages of endometrial gland genesis on PND 14 but did affect uterine growth after PND 28. Available evidence from a number of species supports the hypothesis that the ovary does not affect endometrial gland morphogenesis in the postnatal uterus. To test this hypothesis in our sheep model, ewes were assigned at birth to a sham surgery as a control or bilateral ovariectomy (OVX) on PND 7. Uteri were removed and weighed on PND 56. Ovariectomy did not affect circulating levels of estradiol-17beta. Uterine weight was 52% lower in OVX ewes. Histomorphological analyses indicated that the thickness of the endometrium and myometrium, total number of endometrial glands, and endometrial gland density in the stratum spongiosum stroma was reduced in uteri of OVX ewes. In contrast, the number of superficial ductal gland invaginations and gland density in the stratum compactum stroma was not affected by ovariectomy. The uteri of OVX ewes contained lower levels of betaA subunit, activin receptor (ActR) type IA, ActRIB, and follistatin protein expression but higher levels of betaB subunit. In the neonatal ovary, follistatin, inhibin alpha subunit, betaA subunit, and betaB subunit were expressed in antral follicles between PNDs 0 and 56. These results led to rejection of the hypothesis that the ovary does not influence endometrial adenogenesis. Rather, the ovary and, thus, an ovarian-derived factor regulates, in part, the coiling and branching morphogenetic stage of endometrial gland development after PND 14 and expression of specific components of the activin-follistatin system in the neonatal ovine uterus that appear to be important for that critical process.     

Mechanisms regulating norgestomet inhibition of endometrial gland morphogenesis in the neonatal ovine uterus

In many species, endometrial gland adenogenesis occurs neonatally in an ovary- and steroid-independent manner. Chronic exposure of the developing neonatal ovine uterus to norgestomet (NOR) from birth permanently ablates endometrial gland morphogenesis or adenogenesis, creating an adult ovine uterine gland knockout (UGKO) phenotype. This study was conducted to determine the mechanism(s) whereby NOR inhibits adenogenesis in the neonatal ewe. Ewe lambs received no implant or a NOR implant at birth and on postnatal day (PND) 14, and they were necropsied on PND28. Histological analyses of the tracts indicated NOR exposure specifically inhibited endometrial adenogenesis, but no histoarchitectural differences were observed in the oviduct, cervix, or vagina. No effect of NOR treatment was detected on proliferating cell nuclear antigen (PCNA) expression in the endometrial luminal epithelium (LE), stroma, or myometrium. In control (CX) ewes, estrogen receptor alpha (ER-alpha) and progesterone receptor (PR) mRNA and protein were expressed strongly in nascent and proliferating glandular epithelium (GE) but were undetected in epithelium of NOR uteri. Expression of c-met and fibroblast growth factor receptor 2IIIb (FGFR2IIIb) mRNA was detected in the LE and GE of CX uteri. In NOR uteri, c-met was expressed in the LE similar to CX uteri, but FGFR2IIIb mRNA levels were lower than in the LE of CX uteri. Uterine hepatocyte growth factor (HGF), the ligand for c-met, and FGFR2IIIb mRNA expression was substantially lower in NOR ewes, but expression of FGF-7 and FGF-10 mRNAs, ligands for FGFR2IIIb, was unaffected. These results indicate that NOR disrupts endometrial adenogenesis by ablating epithelial ER-alpha expression and altering expression of paracrine growth factors and/or receptors involved in epitheliomesenchymal interactions. Likewise, these mechanisms are proposed to be important regulators of normal uterine gland morphogenesis in the neonate.     

The activin-follistatin system in the neonatal ovine uterus

Uterine gland development or adenogenesis in the neonatal ovine uterus involves budding and tubulogenesis followed by coiling and branching morphogenesis of the glandular epithelium (GE) from the luminal epithelium (LE) between birth (Postnatal Day [PND] 0) and PND 56. Activins, which are members of the transforming growth factor beta superfamily, and follistatin, an inhibitor of activins, regulate epithelial branching morphogenesis in other organs. The objective of the present study was to determine effects of postnatal age on expression of follistatin, inhibin alpha subunit, betaA subunit, betaB subunit, activin receptor (ActR) type IA, ActRIB, and ActRII in the developing ovine uterus. Ewes were ovariohysterectomized on PND 0, 7, 14, 21, 28, 35, 42, 49, or 56. The uterus was analyzed by in situ hybridization and immunohistochemistry. Neither inhibin alpha subunit mRNA or protein was detected in the neonatal uterus. Expression of betaA and betaB subunits was detected predominantly in the endometrial LE and GE and myometrium between PND 0 and PND 56. In all uterine cell types, ActRIA, ActRIB, and ActRII were expressed, with the highest levels observed in the endometrial LE and GE and myometrium. Between PND 0 and PND 14, follistatin was detected in all uterine cell types. However, between PND 21 and PND 56, follistatin was only detected in the stroma and myometrium and not in the developing GE. Collectively, the present results indicate that components of the activin-follistatin system are expressed in the developing neonatal ovine uterus and are potential regulators of endometrial gland morphogenesis.     

WNT pathways in the neonatal ovine uterus: potential specification of endometrial gland morphogenesis by SFRP2

Endometrial glands are critical for uterine function and develop between birth (Postnatal Day [P] 0) and P56 in the neonatal ewe. Endometrial gland morphogenesis or adenogenesis involves the site-specific budding differentiation of the glandular epithelium from the luminal epithelium followed by their coiling/branching development within the stroma of the intercaruncular areas of the endometrium. To determine whether WNT signaling regulates endometrial adenogenesis, the WNT signaling system was studied in the neonatal ovine uterus. WNT5A, WNT7A, and WNT11 were expressed in the uterine epithelia, whereas WNT2B was in the stroma. The WNT receptors FZD2 and FZD6 and coreceptor LRP6 were detected in all uterine cells, and FZD6 was particularly abundant in the endometrial epithelia. Secreted FZD-related protein-2 (SFRP2), a WNT antagonist, was not detected in the P0 uterus, but was abundant in the aglandular caruncular areas of the endometrium between P7 and P56. Exposure of ewes to estrogens during critical developmental periods inhibits or retards endometrial adenogenesis. Estrogen-induced disruption of endometrial adenogenesis was associated with reduction or ablation of WNT2B, WNT7A, and WNT11, and with an increase in WNT2 and SFRP2 mRNA, depending on exposure period. Collectively, results implicate the canonical and noncanonical WNT pathways in regulation of postnatal ovine uterine development and endometrial adenogenesis. Expression of SFRP2 in aglandular caruncular areas may inhibit the WNT signaling pathway, thereby concentrating WNT signaling and restricting endometrial adenogenesis in the intercaruncular areas of the uterus. Further, estrogen-induced inhibition of adenogenesis may be mediated by a reduction in WNT signaling caused by aberrant induction of SFRP2 and loss of several critical WNTs.     

Endometrial glands are required for preimplantation conceptus elongation and survival

Endometrial glands secrete molecules hypothesized to support conceptus growth and development. In sheep, endometrial gland morphogenesis occurs postnatally and can be epigenetically ablated by neonatal progestin exposure. The resulting stable adult uterine gland knockout (UGKO) phenotype was used here to test the hypothesis that endometrial glands are required for successful pregnancy. Mature UGKO ewes were bred repeatedly to fertile rams, but no pregnancies were detected by ultrasound on Day 25. Day 7 blastocysts from normal superovulated ewes were then transferred synchronously into Day 7 control or UGKO ewes. Ultrasonography on Days 25-65 postmating indicated that pregnancy was established in control, but not in UGKO ewes. To examine early uterine-embryo interactions, four control and eight UGKO ewes were bred to fertile rams. On Day 14, their uteri were flushed. The uterus of each control ewe contained two filamentous conceptuses of normal length. Uteri from four UGKO ewes contained no conceptus. Uteri of three UGKO ewes contained a single severely growth-retarded tubular conceptus, whereas the remaining ewe contained a single filamentous conceptus. Histological analyses of these uteri revealed that endometrial gland density was directly related to conceptus survival and developmental state. Day 14 UGKO uteri that were devoid of endometrial glands did not support normal conceptus development and contained either no conceptuses or growth-retarded tubular conceptuses. The Day 14 UGKO uterus with moderate gland development contained a filamentous conceptus. Collectively, these results demonstrate that endometrial glands and, by inference, their secretions are required for periimplantation conceptus survival and development.     

Sex steroid receptor expression in the oviduct and uterus of sheep with estrus synchronized with progestagen or prostaglandin analogues

The objective of this study was to investigate differences in the expression of estrogen receptor-alpha (ERalpha), progesterone receptor (PR) and the proliferative indexes (Ki-67), in the uterus and oviduct of sheep with estrus synchronized either by prostaglandin analogues (Group PA, n=27) or by treatment with progestagens (Group P, n=29) on days 4 and 7 (day 0=estrus), when the embryos were collected. Immunohistochemical methods were used to quantify ERalpha, PR and Ki-67 in six superficial and deep compartments in the uterus and oviduct. The expression of ERalpha was significantly (P<0.01) lower in progestagen treated ewes than in prostaglandin analogues treated group in the luminal epithelium, superficial glands and superficial stroma in the uterus on day 4. The expression of PR was significantly lower in progesterone treated ewes than in the PA Group in the superficial gland (P<0.05) in both days studied. The lowest expression of PR was observed in the luminal caruncular epithelium and superficial glands in both treatments, obtaining the lowest levels on day 4 (P<0.05). There were significant differences between days 4 and 7 in the Ki-67 immunostaining in the luminal epithelium (P<0.01) and superficial glands (P<0.05). A higher cell proliferation was observed in the uterine epithelium (P<0.05) on day 4 in the animals treated with progestagens. Results indicate that sheep with synchronization of estrus with progestagens showed a reduction of ERalpha and PR protein expression in most of oviductal and uterine cells.     

Endometrial development and adenogenesis in the neonatal pig: effects of estradiol valerate and the antiestrogen ICI 182,780.

In the pig, appearance of endometrial glands between birth (postnatal day [PND] 0) and PND 14 involves development of estrogen receptor-alpha-positive (ER+) phenotype by, and increased DNA synthesis in, nascent glandular epithelium (GE). To determine whether ER activation is required for this process, gilts were treated daily with either vehicle, the antiestrogen ICI 182,780 (ICI), estradiol-17beta valerate (EV), or both ICI and EV. Treatments began on PND 0, before onset of adenogenesis, or on PND 7, after onset of gland proliferation. Uteri obtained on PNDs 7 and 14 (study one) or on PND 14 (study two) were weighed; uterine histology was evaluated; DNA synthesis in luminal epithelium and GE was characterized by determining 5-bromo-2'-deoxyuridine (BrdU) labeling index; and patterns of ER mRNA expression were evaluated in situ (study one). Gland genesis was inhibited by ICI, which decreased gland penetration depth by PND 14 in study one, both endometrial thickness and BrdU-labeling index in GE in study two, and increased stromal cell compaction in both studies. Uterotropic effects of EV included increased gland development and epithelial BrdU labeling and decreased stromal compaction. These effects were inhibited by coadministration of ICI. Treatments did not alter ER mRNA expression, which remained limited to stroma and GE. Data indicate that endometrial maturation and adenogenesis in the neonatal pig require expression and activation of a functional ER system.     

Actions of progesterone on uterine immunosuppression and endometrial gland development in the uterine gland knockout (UGKO) ewe

In ewes, the uterine gland knockout (UGKO) phenotype is caused by neonatal exposure to norgestomet to arrest uterine gland development and produce an adult which has a uterus characterized by the lack of endometrial glands. Since endometrial glands in the sheep produce the lymphocyte-inhibitory protein, ovine uterine serpin (OvUS), an experiment was conducted with ewes of the UGKO phenotype to evaluate whether the inhibitory actions of progesterone on tissue rejection responses in utero are dependent upon the presence of endometrial glands. Control and UGKO ewes were ovariectomized and subsequently treated with either 100 mg/day progesterone or corn oil vehicle for 30 days. An autograft and allograft of skin were then placed in each uterine lumen and treatments were continued for an additional 30 days before grafts were examined for survival. All autografts survived and had a healthy appearance after histological analysis. Allografts were generally rejected in ewes treated with vehicle but were present for hormone-treated ewes, regardless of uterine phenotype. Analysis of the histoarchitecture and protein synthetic capacity of the uterus revealed that progesterone induced differentiation of endometrial glands and synthesis and secretion of OvUS in UGKO ewes. The UGKO ewes had reduced density of CD45R+ lymphocytes in the endometrial epithelium and there was a tendency for progesterone to reduce this effect in luminal epithelium. Taken together, results confirm the actions of progesterone to inhibit graft rejection response in utero. Responses of UGKO ewes to progesterone indicate that the hormone can induce de novo development and differentiation of endometrial glands, at least when skin grafts are in the uterus.     

Ovine uterine gland knock-out model: effects of gland ablation on the estrous cycle

Ovine endometrial gland development is a postnatal event that can be inhibited epigenetically by chronic exposure of ewe lambs to a synthetic progestin from birth to puberty. As adults, these neonatally progestin-treated ewes lack endometrial glands and display a uterine gland knockout (UGKO) phenotype that is useful as a model for study of endometrial function. Here, objectives were to determine: 1) length of progestin exposure necessary from birth to produce the UGKO phenotype in ewes; 2) if UGKO ewes display normal estrous cycles; and 3) if UGKO ewes could establish and/or maintain pregnancy. Ewe lambs (n = 22) received a Norgestomet (Nor) implant at birth and every two weeks thereafter for 8 (Group I), 16 (Group II), or 32 (Groups III and IV) weeks. Control ewe lambs (n = 13) received no Nor treatment (Groups V and VI). Ewes in Groups I, II, III, and VI were hemihysterectomized (Hhx) at 16 weeks of age. After puberty, the remaining uterine horn in Hhx ewes was removed on either Day 9 or 15 of the estrous cycle (Day 0 = estrus). Histological analyses of uteri indicated that progestin exposure for 8, 16, or 32 weeks prevented endometrial adenogenesis and produced the UGKO phenotype in adult ewes. Three endometrial phenotypes were consistently observed in Nor-treated ewes: 1) no glands, 2) slight glandular invaginations into the stroma, and 3) limited numbers of cyst- or gland-like structures in the stroma. Overall patterns of uterine progesterone, estrogen, and oxytocin receptor expression were not different in uteri from adult cyclic control and UGKO ewes. However, receptor expression was variegated in the ruffled luminal epithelium of uteri from UGKO ewes. Intact UGKO ewes displayed altered estrous cycles with interestrous intervals of 17 to 43 days, and they responded to exogenous prostaglandin F(2 approximately ) (PGF) with luteolysis and behavioral estrus. During the estrous cycle, plasma concentrations of progesterone in intact control and UGKO ewes were not different during metestrus and diestrus, but levels did not decline in many UGKO ewes during late diestrus. Peak peripheral plasma concentrations of PGF metabolite, in response to an oxytocin challenge on Day 15, were threefold lower in UGKO compared to control ewes. Intact UGKO ewes bred repeatedly to intact rams did not display evidence of pregnancy based on results of ultrasound. Collectively, results indicate that 1) transient, progestin-induced disruption of ovine uterine development from birth alters both structural and functional integrity of the adult endometrium; 2) normal adult endometrial integrity, including uterine glands, is required to insure a luteolytic pattern of PGF production; and 3) the UGKO phenotype, characterized by the absence of endometrial glands and a compact, disorganized endometrial stroma, limits or inhibits the capacity of uterine tissues to support the establishment and/or maintenance of pregnancy.     

Matrix metalloproteinases and their tissue inhibitors in the developing neonatal mouse uterus

Postnatal development of the mouse uterus involves differentiation and development of the endometrial glands as well as the myometrium. Matrix metalloproteinases (MMPs) and their tissue inhibitors (TIMPs) are involved in extracellular matrix breakdown and morphogenesis of many epitheliomesenchymal organs. As a first step to understanding their roles in postnatal mouse uterine development, MMPs and TIMPs found to be expressed in the neonatal mouse uterus by microarray analysis were localized by in situ hybridization. The MMP-2 mRNA was detected only in the uterine stroma, whereas the MMP-10 mRNA was present only in the uterine epithelium from Postnatal Day (PND) 3 to PND 9. All other MMPs (MMP-11, MMP-14, and MMP-23) as well as TIMP-1, TIMP-2, and TIMP-3 were detected in both epithelial and stromal cells of the endometrium, but not in the myometrium. Uterine extracts were then analyzed by gelatin and casein gel zymography to detect active gelatinases and stromelysins, respectively. Five major gelatinase bands of activity were detected and inhibited by the MMP inhibitors, EDTA or 1,10-phenanthroline, but not by PMSF, a serine protease inhibitor. Western blot analysis confirmed the presence of MMP-2 and MMP-9 proteins in the uterus. Immunoreactive MMP-9 protein was detected only in the endometrial stroma, whereas immunoreactive MMP-2 protein was detected in both the stroma and epithelium of the uterus. Casein zymography detected three major bands of activity ( approximately 54, 63, and 80 kDa) that were inhibited by the serine protease inhibitor, PMSF, but not by the MMP inhibitors, EDTA or 1,10-phenanthroline, suggesting that they were serine proteases. These results support the hypothesis that MMPs and TIMPs regulate postnatal development of the mouse uterus.     

Developmental biology of uterine glands.

All mammalian uteri contain endometrial glands that synthesize or transport and secrete substances essential for survival and development of the conceptus (embryo/fetus and associated extraembryonic membranes). In rodents, uterine secretory products of the endometrial glands are unequivocally required for establishment of uterine receptivity and conceptus implantation. Analyses of the ovine uterine gland knockout model support a primary role for endometrial glands and, by default, their secretions in peri-implantation conceptus survival and development. Uterine adenogenesis is the process whereby endometrial glands develop. In humans, this process begins in the fetus, continues postnatally, and is completed during puberty. In contrast, endometrial adenogenesis is primarily a postnatal event in sheep, pigs, and rodents. Typically, endometrial adenogenesis involves differentiation and budding of glandular epithelium from luminal epithelium, followed by invagination and extensive tubular coiling and branching morphogenesis throughout the uterine stroma to the myometrium. This process requires site-specific alterations in cell proliferation and extracellular matrix (ECM) remodeling as well as paracrine cell-cell and cell-ECM interactions that support the actions of specific hormones and growth factors. Studies of uterine development in neonatal ungulates implicate prolactin, estradiol-17 beta, and their receptors in mechanisms regulating endometrial adenogenesis. These same hormones appear to regulate endometrial gland morphogenesis in menstruating primates and humans during reconstruction of the functionalis from the basalis endometrium after menses. In sheep and pigs, extensive endometrial gland hyperplasia and hypertrophy occur during gestation, presumably to provide increasing histotrophic support for conceptus growth and development. In the rabbit, sheep, and pig, a servomechanism is proposed to regulate endometrial gland development and differentiated function during pregnancy that involves sequential actions of ovarian steroid hormones, pregnancy recognition signals, and lactogenic hormones from the pituitary or placenta. That disruption of uterine development during critical organizational periods can alter the functional capacity and embryotrophic potential of the adult uterus reinforces the importance of understanding the developmental biology of uterine glands. Unexplained high rates of peri-implantation embryonic loss in humans and livestock may reflect defects in endometrial gland morphogenesis due to genetic errors, epigenetic influences of endocrine disruptors, and pathological lesions.     

Disruption of the TIMP-1 gene product is associated with accelerated endometrial gland formation during early postnatal uterine development

Postnatal uterine development is marked by periods of tissue remodeling. The objective of the present study was to examine the role of tissue inhibitor of metalloproteinase-1 (TIMP-1), a regulator of tissue remodeling events, during postnatal uterine development and to assess the phenotypic consequences of disruption of the TIMP-1 gene product during this time period. To accomplish this goal, wild-type and TIMP-1 null mice were sacrificed at Postnatal Days (PNDs) 5, 10, 15, 20, and 25 and uterine morphology, TIMP expression and matrix metalloproteinase (MMP) activity were assessed. In wild-type mice, TIMP-1 mRNA steady-state levels were highest at PND 5, after which expression decreased. TIMP-2 and TIMP-3 expression in wild-type mice showed no significant changes from PND 5 to 25. In TIMP-1 null mice, TIMP-2 and TIMP-3 expression patterns were similar to those in wild-type counterparts with the exception that, at PND 10, TIMP-2 and TIMP-3 expression was significantly lower in the null mice. Endometrial gland number and uterine histology were similar between genotypes at PNDs 5 and 10, but at PNDs 15 and 20, endometrial glands were more abundant in TIMP-1 null mice. Associated with the increased gland density in the null mice was an increase in total MMP activity above the levels expressed in wild-type mice. In summary, disruption of the TIMP-1 gene product is associated with reduced TIMP-2 and TIMP-3 steady-state mRNA levels, elevated MMP activity, and accelerated endometrial gland formation. We conclude that, during early postnatal uterine development, TIMP-1 may be critical for proper endometrial gland development.     

Progesterone inhibits uterine gland development in the neonatal mouse uterus

Uterine glands and their secretions are required for conceptus (embryo/fetus and associated placenta) survival and development. In most mammals, uterine gland morphogenesis or adenogenesis is a uniquely postnatal event; however, little is known about the mechanisms governing the developmental event. In sheep, progestin treatment of neonatal ewes permanently ablated differentiation of the endometrial glands. Similarly, progesterone (P4) inhibits adenogenesis in neonatal mouse uterus. Thus, P4 can be used as a tool to discover mechanisms regulating endometrial adenogenesis. Female pups were treated with sesame vehicle alone as a control or P4 from Postnatal Day 2 (PD 2) to PD 10, and reproductive tracts were examined on PD 5, 10, or 20. Endometrial glands were fully developed in control mice by PD 20 but not in P4-treated mice. All other uterine cell types appeared normal. Treatment with P4 stimulated proliferation of the stroma but suppressed proliferation of the luminal epithelium. Microarray analysis revealed that expression of genes were reduced (Car2, Fgf7, Fgfr2, Foxa2, Fzd10, Met, Mmp7, Msx1, Msx2, Wnt4, Wnt7a, Wnt16) and increased (Hgf, Ihh, Wnt11) by P4 in the neonatal uterus. These results support the idea that P4 inhibits endometrial adenogenesis in the developing neonatal uterus by altering expression of morphoregulatory genes and consequently disrupting normal patterns of cell proliferation and development.     

Postnatal deletion of Wnt7a inhibits uterine gland morphogenesis and compromises adult fertility in mice

The success of postnatal uterine morphogenesis dictates, in part, the embryotrophic potential and functional capacity of the adult uterus. The definitive role of Wnt7a in postnatal uterine development and adult function requires a conditional knockout, because global deletion disrupts müllerian duct patterning, specification, and cell fate in the fetus. The Wnt7a-null uterus appears to be posteriorized because of developmental defects in the embryo, as evidenced by the stratified luminal epithelium that is normally found in the vagina and the presence of short and uncoiled oviducts. To understand the biological role of WNT7A after birth and allow tissue-selective deletion of Wnt7a, we generated loxP-flanked exon 2 mice and conditionally deleted Wnt7a after birth in the uterus by crossing them with Pgr(Cre) mice. Morphological examination revealed no obvious differences in the vagina, cervix, oviduct, or ovary. The uteri of Wnt7a mutant mice contained no endometrial glands, whereas all other uterine cell types appeared to be normal. Postnatal differentiation of endometrial glands was observed in control mice, but not in mutant mice, between Postnatal Days 3 and 12. Expression of morphoregulatory genes, particularly Foxa2, Hoxa10, Hoxa11, Msx1, and Wnt16, was disrupted in the Wnt7a mutant uteri. Conditional Wnt7a mutant mice were not fertile. Although embryos were present in uteri of mutant mice on Day 3.5 of pregnancy, blastocyst implantation was not observed on Day 5.5. Furthermore, expression of several genes (Foxa2, Lif, Msx1, and Wnt16) was reduced or absent in adult Wnt7a-deleted uteri on Day 3.5 postmating. These results indicate that WNT7A plays a critical role in postnatal uterine gland morphogenesis and function, which are important for blastocyst implantation and fertility in the adult uterus.