| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Minireview |
a Center for Animal Biotechnology and Genomics, Department of Animal Science, Texas A&M University, College Station, Texas 77843
b Department of Animal and Dairy Sciences, Program in Cell and Molecular Biosciences, Auburn University, Auburn, Alabama 36849
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
developmental biology, female reproductive tract, mechanisms of hormone action, steroid hormone receptors, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
Uterine secretions are thought to be particularly important for conceptus survival and development in sheep, cattle, pigs, and horses, in which a prolonged period of preimplantation conceptus development precedes superficial attachment and placentation. Consequences of steroid-induced disruption of neonatal uterine development, when evaluated in terms of adult endometrial structure and function as well as conceptus survival, illustrate the importance of both uterine developmental success and functional endometrial integrity in domestic ungulates. Exposure of neonatal ewes to a progestin ablated endometrial gland differentiation and produced adult ewes that displayed the uterine gland knockout (UGKO) phenotype, which is characterized by the absence of endometrial glands [15–21]. Partial to complete UGKO phenotypes were also produced in adult cows exposed from birth to a combination of progesterone plus estradiol benzoate (P+E) [16, 22]. Studies of embryo development in the ovine UGKO uterus showed that a normal glandular endometrium is essential for peri-implantation conceptus survival and growth [18, 20, 21]. Consistently, pregnancy rates were also reduced in adult heifers with reduced endometrial gland numbers that were exposed neonatally to P+E [16, 23]. In mares, infertility and subfertility are associated with fibrotic lesions in the endometrium that are thought to compromise the integrity and functionality of endometrial glands [24]. In sheep [25, 26], cattle [27], and pigs [28, 29], endometrial glands undergo extensive hyperplasia and hypertrophy during pregnancy, presumably in response to increasing demands by the developing conceptus for uterine histotroph [3, 26, 30]. The precise fate and role of endometrial glands in human pregnancy are not known. Nonetheless, functional endometrial glands appear to be required during the peri-implantation period and through the first 8 wk after implantation as the placenta develops [5, 7].
The genetic potential for uterine function during pregnancy is defined at conception, but the success of developmental events regulating endometrial gland morphogenesis ultimately determines the functional capacity and embryotrophic potential of the adult uterus [16, 20, 21]. Therefore, high and unexplained rates of peri-implantation embryonic losses in humans and livestock may reflect, in part, unrecognized defects in endometrial adenogenesis or function induced during critical organizational periods in the neonate or adult [16, 18, 21]. In women and menstruating primates, the long pre- and peripubertal period, during which endometrial adenogenesis occurs, and the cyclical nature of adult endometrial regeneration provide significant and repeated opportunities for endometrial dysgenesis and development of pathological lesions that could contribute to infertility. The purpose of this review is to summarize current understanding of the developmental biology of uterine glands, with an emphasis on hormonal, cellular, and molecular mechanisms that regulate endometrial gland morphogenesis and differentiated function.
|
COMPARATIVE DEVELOPMENTAL BIOLOGY OF UTERINE GLAND MORPHOGENESIS |
|---|
|
TOPABSTRACTINTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
In all mammals, the uterus develops as a specialization of the paramesonephric or müllerian ducts, which gives rise to the infundibula, oviducts, uterus, cervix, and anterior vagina [31]. The mature uterine wall is comprised of two functional compartments, the endometrium and the myometrium. The endometrium is the inner mucosal lining of the uterus and is derived from the inner layer of ductal mesenchyme. Histologically, the endometrium consists of two epithelial cell types, luminal epithelium (LE) and glandular epithelium (GE), which are two stratified stromal compartments including a densely organized stromal zone (stratum compactum) and a more loosely organized stromal zone (stratum spongiosum), blood vessels, and immune cells. The myometrium is the smooth muscle component of the uterine wall and includes an inner circular layer derived from the intermediate layer of ductal mesenchymal cells and an outer longitudinal layer derived from subperimetrial mesenchyme.
Morphogenetic events common to development of all mammalian uteri include 1) organization and stratification of endometrial stroma, 2) differentiation and growth of the myometrium, and 3) coordinated development of the endometrial glands [16, 32]. As illustrated in Figure 1, genesis of uterine glands involves differentiation and budding of GE from LE, penetration of uterine stroma by tubes of GE, and extensive coiling and branching of GE. In humans and livestock, endometrial adenogenesis is completed postnatally and involves extensive coiling and branching morphogenesis (Fig. 2). Consequently, neonatal ungulates (e.g., sheep, cattle, and pigs) provide attractive models for the study of mechanisms regulating these processes [33]. Although adenogenesis is also a postnatal event in rodents, the adult rodent uterus does not contain the tightly coiled, branched glands that are characteristic of endometria in most other mammals. Thus, ungulate and primate species may provide more relevant models for the study of this process and related uterine organizational events affecting the embryotrophic and functional capacity of the adult uterus in women and many other mammals.
|
|
Sheep and Cattle
The endometrium in adult sheep and cattle consists of a large number of raised aglandular caruncles, dense stromal protuberances covered by a simple LE, and intensely glandular intercaruncular areas [25, 34]. Caruncular areas are the sites of superficial implantation and placentation [25, 31]. In synepitheliochorial placentation, which is characteristic of these ruminants, fusion of placental cotyledons with endometrial caruncles forms placentomes, which serve a primary role in fetal-maternal gas exchange and derivation of micronutrients by the placenta [25, 31, 35]. Intercaruncular endometrial areas contain large numbers of branched, coiled uterine glands that synthesize and secrete or transport a variety of enzymes, growth factors, cytokines, lymphokines, hormones, transport proteins, and other substances (i.e., histotroph) [2, 3]. The dichotomous nature of the adult ruminant endometrium, consisting of both aglandular caruncular areas and glandular intercaruncular areas, makes it an excellent model for the study of mechanisms underlying the establishment of divergent structural and functional areas within a single, mesodermally derived organ [36].
Fetus Uterine morphogenesis has been described in sheep [15, 19, 36–42] and, to a lesser extent, in cattle [34, 43]. The ewe and cow have gestational lengths of approximately 147 and 284 days, respectively. Paramesonephric duct fusion occurs between Gestational Days (GD) 34 and 55 in sheep, is partial, and produces a bicornuate uterus [36–38]. By GD 90, raised aglandular nodules are present that are destined to become caruncles [16, 36]. Endometrial gland development is first observed as shallow invaginations of LE in internodular clefts between GDs 135 and 150 in sheep [36] and at GD 250 in cattle [34, 43]. These shallow invaginations in the LE represent primordial bud formation of GE.
Neonate Postnatal uterine morphogenesis in sheep and cattle involves the emergence and proliferation of endometrial glands, development of endometrial folds, and, to a lesser extent, growth of endometrial caruncular areas and myometrium [16, 34, 36, 39, 40, 41, 43]. The progressive development of endometrial GE from the LE to the inner circular layer of myometrium is a coordinated event that involves bud formation and tubulogenesis and is completed with coiling and branching morphogenesis. In sheep, endometrial gland genesis is initiated between birth (or Postnatal Day [PND] 0) and PND 7, when shallow epithelial invaginations appear along the LE in presumptive intercaruncular areas [15, 41]. Between PNDs 7 and 14, nascent GE buds proliferate and invaginate into the stroma, forming tubular structures that coil and branch by PND 21 [41]. After PND 21, the majority of glandular morphogenetic activity involves branching morphogenesis of tubular and coiled endometrial glands to form terminal-end, bud-like structures in deeper stroma. By PND 56, the caruncular and intercaruncular endometrial areas are histoarchitecturally similar to those of the adult uterus [41]. In UGKO ewes, endometrial glands do not penetrate the intercaruncular stroma regularly, nor is a characteristic stratum spongiosum recognizable within the intercaruncular stroma [17–20]. Thus, in sheep, development of GE appears to direct or permit differentiation of uterine stroma into subluminal stratum compactum and stratum spongiosum in intercaruncular areas of the endometrium.
Adult Although the ovine uterine wall is histoarchitecturally mature by 8 wk after birth, final maturation and growth may not occur until puberty [39] or even first pregnancy [25, 26]. Extensive hyperplasia and hypertrophy of endometrial glands in the stratum compactum occur during each pregnancy [25, 26], presumably in response to increasing demands for histotrophic support by the growing fetoplacental unit [3, 26]. After parturition in the ewe, intercaruncular endometrial LE remains intact, but degeneration and apoptosis of many GE cells have been observed on the day after parturition [44]. In that study, glandular regeneration commenced by Day 8 postpartum, and the glands were substantially regenerated by Day 15. In caruncles, regeneration of LE commenced after Day 8 and was not complete until Days 28–31 postpartum [45, 46]. Caruncular LE appeared to emanate from epithelia in the intercaruncular areas of the endometrium [45, 46]. Overall, patterns of endometrial gland genesis and development in the neonatal and adult ovine uterus are very similar to the GE morphogenesis that is characteristic of the stages of mammogenesis, lactogenesis, and involution of the mammary gland [47].
Pig
Transformation of the porcine uterine wall from histoarchitectural infancy to maturity occurs within 120 days of birth [32, 48–54]. Uterine wall development in the neonatal pig is dramatic and includes the appearance and proliferation of endometrial glands, organization of the stroma, development of endometrial folds, and growth of the myometrium. Overall patterns of endometrial morphogenesis in the gilt are similar to those in the ewe. Endometrial glands are absent at birth, but they develop in a rapid, synchronous manner during early postnatal life [53]. Shallow, epithelial depressions can be observed at PND 0 and are the presumed precursors for the coiled, branched uterine glands that are characteristic of the adult porcine uterus. Endometrial adenogenesis is initiated when GE develops into simple epithelial tubes that extend radially from the luminal surface into the endometrial stroma. Eventually, tubular glands undergo coiling and branching within the stroma until they reach the adluminal border of the myometrium [49, 53–55]. Mature uterine histoarchitecture is observed by PND 120 in crossbred gilts [32, 48–54].
Neonate At birth, the porcine uterus consists of a simple, slightly corrugated columnar epithelium supported by unorganized stromal mesenchyme, encircled by a rudimentary myometrium [32, 53, 54, 56]. By PND 7, stromal zones, including a shallow stratum compactum and a deep stratum spongiosum, are evident, and distinct, simple, coiled tubular glands are present throughout shallow stroma. By PND 14, many coiled tubular glands are apparent that extend approximately a third of the distance from the LE to the myometrium, which has differentiated into inner circular and outer longitudinal layers. By PND 28, many of the coiled glands have obvious branches, and GE is present throughout the endometrial stroma. Additionally, well-developed endometrial folds are apparent by PND 28, increasing the uterine luminal surface area. By PND 56, endometrial glandularity is dense and extensive [53, 54, 56]. The porcine uterus is capable of supporting pregnancy by PND 120, indicating that it is functionally mature [32].
Adult Gestation in the pig lasts for approximately 114 days. On GD 30, porcine uterine glands appear as simple, coiled, tubular structures with a narrow lumen [28, 29]. The simple columnar GE includes both ciliated and secretory cells. At midpregnancy, endometrial glands are highly dilated and filled with substantial amounts of granular, acid phosphatase-positive material, which is indicative of high secretory activity. Glandular secretory activity remains high during the last third of pregnancy. After parturition, uterine glands undergo rapid involution [28].
Horse
Gestational length in the horse is 335–340 days. Detailed descriptions of fetal organogenesis and neonatal morphogenesis of the equine uterus, however, are few.
Fetus and neonate Histology of the equine uterus at GDs 100, 150–160, and 180–200 was reported by Ginther [57]. On GD 100, the equine uterine wall was immature and composed of a simple LE supported by undifferentiated mesenchyme that was surrounded by a single, longitudinal layer of smooth muscle cells. By GDs 150–160, mitotic figures were described in the simple columnar LE, and endometrial folding was evident. Pronounced folding of the mucosa was apparent by GDs 180–200. In some areas, the uterine epithelium was corrugated and contained slight invaginations into the stroma, but definitive GE was not present in the endometrium [57]. Equine endometrial gland development from birth to sexual maturity was described recently by Gerstenberg and Allen [58]. In marked contrast to observations in other large domestic animals, including sheep, cattle, and pigs, endometrial glands in the neonatal mare are simple tubular structures that remain in a comparatively nonproliferative, juvenile state of development through the first pubertal estrus but that proliferate rapidly during the first diestrus. Observations suggest a unique role for progesterone in equine endometrial adenogenesis [58, 59].
Adult The equine conceptus has a unique, early association with the maternal endometrium, because it remains spherical and virtually unattached to the endometrium until Day 35 postovulation [60]. Therefore, endometrial glands and their secretions are thought to be necessary to support of conceptus survival during this prolonged preimplantation period [30]. Older mares often display a decreased ability to produce healthy foals as a result of degenerative changes in the endometrium, termed endometrosis, that affect the number and morphology of endometrial glands [61]. These age-related, degenerative changes cause deposition of fibrous tissue in the stroma and grouping of endometrial glands into "gland nests" [62]. Degenerated uterine glands display functional abnormalities, including an increase in ciliated cells, atypical mucus production, reduced expression of epidermal growth factor (EGF) and P19 lipocalin-like protein, and abnormal patterns of proliferation [24, 63, 64]. Abnormal endometrial glands in uteri of mares exhibiting endometrosis have been suggested to be responsible for the observed decreases in fertility [24, 65]. Postpartum changes in the equine endometrium include rapid degeneration of microcaruncles and uterine glands [66]. By Day 7 postpartum, endometrial histology is similar to that observed during normal proestrus, with cuboidal LE and edematous stroma. By Days 9 and 10 postpartum, endometrial histology is similar to that observed in the mare during estrus.
Rodents (Rats and Mice)
Gestational length is 20 days in mice and 21 days in rats. Paramesonephric duct fusion occurs at GDs 15–16, is partial, and produces a bicornuate uterus. At birth, uteri of mice and rats lack endometrial glands, and the uterus consists of a simple epithelium supported by undifferentiated mesenchyme. On PND 5, epithelial invaginations appear, which represent the formation of GE buds [67]. Genesis of endometrial glands is not observed until PND 7 in mice and PND 9 in rats [68]. In the rat uterus, adenogenesis proceeds from PND 9 through PND 15 [68] and results in the development of simple tubular glands that, unlike ungulate endometrial glands, are neither tightly coiled nor extensively branched.
During the preimplantation period of early pregnancy in these rodents, endometrial glands synthesize and secrete several proteins required for the establishment of uterine receptivity and embryo implantation [11, 12]. If a successful pregnancy occurs, endometrial glands are ablated by stromal decidualization in response to conceptus implantation. Information is not currently available regarding the histoarchitecture of endometrial gland regeneration after parturition during uterine involution in rodent species.
Humans and Menstruating Primates
Humans have a simplex uterus consisting of a single uterine body or corpus lacking the uterine horns characteristic of species that possess a bicornuate uterus. Histologically, the adult human and primate endometrium is stratified into two zones, including the stratum functionalis and the stratum basalis [69]. The endometrial functionalis, which is lost during menses, is further subdivided into two parts. Zone I, which is lost almost entirely during menses, consists of LE and subadjacent stroma. Zone II consists of dense stroma surrounding the straight portions of endometrial glands [69]. The endometrial basalis is a dynamic, but structurally stable, compartment of the primate uterus that is not eroded during menstruation or at the end of gestation. This tissue functions as the germinal compartment of the endometrium in these species, and it provides the stem cells from which the functionalis regenerates with each cycle or after gestation. Histologically, the basalis includes endometrial zone III, which contains loose stroma and the bodies of uterine glands, and endometrial zone IV, where endometrial glands terminate and endometrial progenitor and stem cells are thought to reside [69]. The lower two-thirds of the endometrial glands are retained within the stratum basalis [70].
Fetus and Neonate As in other mammals, the prenatal uterus in humans is formed by fusion of the paramesonephric ducts, which occurs before 8 wk of gestation [71, 72]. As seen in rodents and ungulate species, the simple columnar epithelium of the undifferentiated uterine body gives rise to numerous invaginations, which represent primordial GE buds [71–73]. By 20–22 wk of gestation, the myometrium is well-defined, but endometrial gland development is very superficial [73–75]. Endometrial histoarchitecture at birth resembles that of the adult, though less developed. Neonatal endometrial LE is low columnar or cuboidal, and GE is sparse and limited to the adluminal stroma [71, 76]. From birth to the onset of puberty, uterine glands develop slowly. By 6 yr of age, endometrial glands extend from one-third to one-half of the distance to the myometrium. Mature uterine histoarchitecture is observed at puberty, with endometrial glands extending to the inner circular layer of the myometrium [76]. Although initiated during fetal life, endometrial gland proliferation in the human uterus is completed postnatally, in a manner similar to that observed for domestic ungulates [16, 32]. Thus, genesis of endometrial glands in the human fetus and neonate involves differentiation of GE from LE and development of GE through endometrial stroma to the myometrium. This pattern of endometrial development is the opposite of that observed for gland genesis in postmenstrual uteri of adult women and primates, in which endometrial glands develop adluminally from the basalis during the proliferative phase [77].
Adult In contrast to nonmenstruating primates, rodents, and livestock, the premenopausal endometrium in adult humans and primates undergoes programmed, phasic changes that include menses, a proliferative phase, and a secretory phase [78–80]. With each menstrual cycle, the primate functionalis is eroded away during menstruation and regenerates from the basalis during the proliferative period [81]. This process is primarily regulated by ovarian steroids [82]; however, responses to these hormones are cell and endometrial zone specific [70, 83]. As indicated above, endometrial regeneration and growth of uterine glands during the proliferative period involves stromal and epithelial cells of the deep stratum basalis, which proliferate and organize adluminally to complete formation of the stratum functionalis [77]. Interestingly, compared to those of livestock and rodents, the stromal layers in the primate endometrium are inverted. In menstruating primates, the more densely cellular or compact stroma (stratum compactum) is located in the stratum basalis, adjacent to the myometrium, whereas the more loosely organized stroma (stratum spongiosum) is characteristic of the adluminal primate functionalis. In both adult ungulates and primates, available evidence supports the hypothesis that compact stroma (e.g., stratum compactum or basalis) supports endometrial gland genesis and tubule formation, whereas loose stroma (e.g., stratum spongiosum or functionalis) supports coiling and branching morphogenesis of proliferating endometrial glands.
In the human uterus during the proliferative phase, endometrial glands undergo extensive branching morphogenesis as the functionalis is reconstructed [77, 82]. The early proliferative endometrium is thin, and it contains regenerating superficial epithelium of the functionalis as well as endometrial glands that are narrow, short, straight, and partially collapsed [82]. During the midproliferative phase, the glands elongate and become more tortuous. Proliferation is primarily observed in GE on Day 3, but by Day 5, proliferation of the endometrium is occurring in GE of the functionalis as well as in the surrounding stroma [77]. By the late proliferative phase, the glands are coiled tightly and are obviously branched. Tortuosity and branching of endometrial glands reach a maximum during the secretory phase by Day 8 postovulation. During this phase, glands increase in diameter and tortuosity, but they do not proliferate [77]. If pregnancy does not occur, endometrial glands regress late in the secretory phase, commensurate with luteolysis, and break down during menses. Similar changes in endometrial histoarchitecture occur in menstruating primates [77, 84, 85]. To our knowledge, the fate of and changes in the uterine glands during pregnancy in the nondecidualizing area of the uterus have not been described for humans and other primates.
|
MECHANISMS REGULATING ENDOMETRIAL GLAND MORPHOGENESIS |
|---|
|
TOPABSTRACTINTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
|
Site-Specific Alterations in Cell Proliferation
Differentiation and budding of endometrial GE from LE does not appear to require cell proliferation [53]. In developing lung epithelium, bud outgrowth and invagination does not involve localized cell proliferation [87] but does involve changes in ECM biochemistry in the basal lamina [88]. These data are supported by the results of histochemical studies regarding the stromal-epithelial interface in neonatal ovine [40] and porcine uteri [53]. Progestin-induced ablation of endometrial gland genesis in the neonatal ovine uterus does not appear to involve specific suppression of epithelial cell proliferation [19]. However, evidence of focused proliferation of GE in tips of developing glands, documented in neonatal ovine and porcine uteri as well as in the adult primate uterus, supports the idea that local microenvironmental conditions are important for gland proliferation [15, 41, 53, 77, 82].
Epithelial-Mesenchymal Interactions
Uterine development and function depend on epithelial-mesenchymal interactions [86, 89]. These interactions provide local control and coordination of morphogenetically important cell behaviors, including movement, adhesion, differentiation, and proliferation [90, 91]. Tissue recombination studies in rodents clearly indicate that uterine mesenchyme directs and specifies patterns of epithelial development, whereas epithelium is required to support the organization of endometrial stroma and myometrial differentiation [86, 89, 92]. Results from studies of neonatal ovine [19, 40, 41], porcine [32, 53], and rodent [93] uteri, as well as results from studies of regenerating postmenstrual primate uterus [77, 82, 94], are consistent with the idea that uterine gland morphogenesis is supported and regulated through interactions between the epithelium and stroma. Developmentally critical tissue microenvironments, which are necessary to support and maintain spatially focused changes in cell behaviors associated with gland genesis, are thought to evolve through such interactions [16, 32]. The concept that interactions between GE and stroma are required for endometrial morphogenesis and establishment of normal uterine histoarchitecture is supported by the results of elegant tissue recombination studies involving the mouse uterus [89, 95]. In adult UGKO ewes [16–18, 20] the endometrium not only is devoid of GE but is generally thinner, with a more compact stroma that lacks the characteristic zonation necessary for histological delineation of the stratum compactum and stratum spongiosum. This extreme endometrial phenotype is proposed to reflect disintegration of stromal-epithelial interactions that normally support epithelial morphogenesis and cytodifferentiation as well as stromal organization and development [16, 32].
Epithelial-mesenchymal interactions are mediated, in part, by changes in the composition and distribution of ECM components. Glycosaminoglycans (GAGs), which are oligosaccharide components of the ECM, can affect cell function both directly and indirectly (e.g., by mediating access of growth factors and other molecules to their receptors or target cells). During adenogenesis in many tissues, including salivary glands, prostate, and uterus, the sulfated GAGs, including chondroitins and heparans, become localized to morphogenetically inactive sites, such as the necks of glands, whereas the nonsulfated GAGs, such as hyaluronic acid, accumulate in morphogenetically active sites, such as the tips of proliferating glands [32, 40, 53, 95]. Although not investigated in the developing uterus, metalloproteinases and other factors that alter the biochemical nature of the basal lamina affect both physical and chemical interactions between epithelium and underlying stroma. In this context, the ECM can affect patterns of branching morphogenesis through control of the cell cycle, apoptosis, and related changes in stromal and epithelial gene expression that define such developmental programs [96, 97]. Thus, elements of cooperative signaling pathways mediating uterine organization likely include cell-cell and cell-ECM interactions that facilitate the actions of insoluble signals from the ECM as well as the actions of soluble hormones and growth factors [97].
Growth Factors
Communication between the epithelium and stroma is facilitated by paracrine and autocrine pathways within uterine tissues that involve peptide growth factors and their receptors [98, 99]. Stromal-derived growth factors play important roles in epithelial proliferation, differentiation, and branching morphogenesis in developing epitheliomesenchymal organs, including the uterus [19, 42, 98, 100]. Interactions between growth factors and their receptors can involve elements of the ECM, which not only affect patterns of growth factor presentation to target cells but also may participate as elements of cell-surface receptor complexes, as demonstrated for the fibroblast growth factor (FGF) family [101]. Peptide growth factors implicated in uterine development include FGF-7, FGF-10, hepatocyte growth factor (HGF), and insulin-like growth factor (IGF)-I and IGF-II [42, 100, 102].
Stromal growth factors An established paracrine growth factor, FGF-7 stimulates epithelial cell proliferation and differentiation [103–105]. Isolated originally from rat lung mesenchyme, FGF-10 was determined to be essential for the patterning of early events in branching morphogenesis [106–108]. The HGF functions as a paracrine mediator of mesenchymal-epithelial interactions that govern mitogenic, motogenic, and morphogenic behaviors of epithelia in developing lung and mammary tissues [109–111]. In the developing neonatal ovine uterus, FGF-7, FGF-10, HGF, and their epithelial receptors were identified as being organizationally important growth factor systems associated with endometrial morphogenesis [19, 42]. Although FGF-7 mRNA was constitutively expressed in uteri from PND 1 to PND 56, FGF-10 and HGF mRNA levels increased markedly after PND 21, which is a period characterized by coiling and branching development of endometrial glands in the neonatal ovine uterine wall [42]. Furthermore, progestin-induced inhibition of endometrial adenogenesis in the neonatal ewe altered expression patterns of these paracrine-acting growth factors and/or their receptors [19].
Insulin-like growth factors In addition to their roles as mitogens, IGF-I and IGF-II induce cellular differentiation and promote the expression of differentiated functions in cells and tissues [112–114]. These IGFs are, therefore, multifunctional regulators of cell proliferation, differentiation, and function that act through autocrine and/or paracrine mechanisms in many organ systems, including the uterus [114–116]. In such tissues, IGFs can also regulate responses to steroid hormones. Examples include complex responses of the immature rodent uterus to estrogens [102, 117–119] and human endometrial proliferative growth responses to ovarian estradiol [115]. Null mutation of the IGF-I gene in mice was employed to demonstrate the critical role of this growth factor in normal development of the female reproductive tract [120] as well as its requirement for estrogen-induced uterine growth in cyclic female rodents [121]. Mice lacking the type 2 IGF receptor displayed delayed lung development and poor alveolar differentiation, indicating a functional role for this system in lung morphogenesis [122]. In neonatal rodent and ovine uteri, the IGF system is involved in postnatal uterine morphogenesis and growth [42, 102, 120]. Gu et al. [102] observed that IGF-I mRNA expression in the neonatal rat uterus was confined to the stroma and myometrium and increased during the developmental period associated with uterine gland genesis. Expression of IGF-II was not detected, although it is present in endometrium from both nonpregnant and pregnant women [123].
In the neonatal ovine uterus, IGF-I and IGF-II are expressed in the stroma surrounding nascent and proliferating endometrial GE, which is both IGF-I receptor (IGF-1R) and estrogen receptor 
(ER) positive [41, 42]. Cross-talk between ER and IGF-1R signaling pathways results in synergistic growth stimulation in a number of systems [124, 125]. Activation of ER by growth factors such as IGF-I involves the mitogen-activated protein kinase (MAPK) pathway via direct serine phosphorylation. In addition, estrogen increases IGF-1R protein in the immature rat uterus [117] and modulates IGF-1R function by inducing tyrosine phosphorylation of IGF-I and insulin receptor substrate-1, which is followed by enhanced MAPK activation [126, 127]. Collectively, the available data can be interpreted to suggest that stromal IGF-I and IGF-II may stimulate proliferation of GE and support uterine gland genesis in the neonatal ovine uterine wall through such signaling pathways.
Steroids and Their Receptors
Jost [128], working with rabbits, established the concept that prenatal urogenital tract development in female mammals is an ovary-independent process. Since then, results of numerous studies have indicated that uterine development and endometrial adenogenesis can proceed normally in the absence of ovarian support for varying periods of time during early postnatal life as well. Circulating estrogens increase between PNDs 9 and 11 in the rat, in association with postnatal endometrial remodeling events [129]. However, early postnatal events in rat uterine development and endometrial adenogenesis are both ovary [130] and adrenal independent [93, 131]. In the neonatal pig, Tarleton et al. [54] determined that ovariectomy at birth, though antiuterotrophic after PND 56, did not affect the genesis of uterine glands or related endometrial morphogenetic events before PND 120. Similarly, ovariectomy of ewe lambs at birth did not affect the patterns of uterine gland genesis at PND 14 [15] but was antiuterotrophic after PND 28 [39]. This finding suggests a role for endogenous estrogens in branching morphogenesis, but not in differentiation or budding of GE from LE.
In the neonatal ewe lamb, numbers of growing and vesicular ovarian follicles, which are substantial by PND 28, decline significantly by PND 84 [39]. Consistently, serum estradiol-17ß was detected at relatively high levels in ewe lambs at PND 1, increased between PNDs 14 and 28, and then declined between PNDs 42 and 56 [41]. This pattern of circulating estrogen correlates with the ontogeny of endometrial gland development in the ewe lamb. However, the precise role of ovarian estrogens in ovine neonatal uterine development remains to be determined. In cyclic women and primates, estradiol-17ß of ovarian origin is found at high levels during the proliferative phase of the menstrual cycle and has been proposed to regulate endometrial morphogenesis [77, 82, 94, 115].
Although endometrial adenogenesis is ovary-independent in neonatal rodents and pigs, genesis of endometrial glands in porcine [54] and rodent [132–135] uteri involves coordinated changes in epithelial phenotype that are marked by ER expression in nascent and proliferating endometrial GE. Homozygous ER-null mice (ERKO) have hypoplastic uteri that contain all characteristic cell types, though in reduced proportions [136], including reduced uterine gland numbers [137]. Thus, ER expression is not essential for fetal murine uterine organogenesis, but it is essential for normal postnatal uterine growth and development [136].
In the neonatal rat uterus, gland genesis occurs between PNDs 9 and 15, and treatment with estradiol-17ß from PND 10 to PND 14 delayed the onset of gland genesis [68]. Interestingly, treatment with estradiol-17ß from PND 1 to PND 5 induced slightly premature gland genesis but, ultimately, reduced gland numbers between PNDs 15 and 26 [68]. Mechanisms mediating these age-specific effects of estrogen on uterine gland genesis are not known, but the effects could be due to estrogen-induced, negative regulation of ER expression. Treatment of rats with the antiestrogen tamoxifen, a mixed ER agonist/antagonist, from PNDs 1 to 5 or PNDs 10 to 14 elicited a dose-dependent inhibition of uterine gland genesis [138]. Interestingly, cotreatment of neonatal rats with ICI 182,780, a type II antiestrogen that is antiuterotrophic but not antiadenogenic in neonates of this species, inhibited the characteristic tamoxifen-induced effects on endometrial gland development [139]. Observations were interpreted to suggest that tamoxifen acts as an ER agonist in this physiological context. Compounds such as tamoxifen and estradiol-17ß, both of which cause pronounced uterine hypertrophy in the rat when administered as uterine glands are developing postnatally, disrupt the normal process of gland genesis. Given that the first step in uterine gland development involves differentiation of GE and invagination of nascent glands into the underlying stroma, it is conceivable that the extreme hypertrophic state induced by ER agonists prevents invagination of GE physically, as a consequence of alterations in cell shape and associated changes in cell-cell and cell-ECM relationships that would otherwise support this process [96, 97]. Under such conditions, epithelial cells could be unable to recognize, integrate, and respond as usual to the cooperative signals that normally drive gland genesis [96, 97, 139]. This does not explain how endometrial gland genesis is initiated, but it does explain how the process might be disrupted through disorganization of local control mechanisms at the tissue level.
In the neonatal pig, administration of ICI 182,780 from birth inhibited endometrial adenogenesis at PND 14 [55]. These data are generally consistent with the idea that uterine ER expression and changes in the state of uterine ER activation, which may be species specific, are important elements of the organizational program that determines patterns of uterine growth and endometrial morphogenesis. In this regard, results of elegant tissue recombination studies involving mouse uterine stroma and epithelium indicate that epithelial ER is neither necessary nor sufficient to mediate the mitogenic actions of estrogen [98, 140]. In addition to direct ligand-dependent activation of epithelial ER, proliferative effects of estrogen on epithelium appear to be mediated primarily by stromal ER via production of paracrine-acting, stromal-derived growth factors, such as EGF and IGF-I [98].
In neonatal ewes, all uterine cell types are ER-positive at PND 1 [41]. Endometrial gland morphogenesis is accompanied by ER expression in emerging, proliferating, and developing GE as well as in the surrounding stroma. A requirement for ER in ovine uterine adenogenesis is supported by the finding that progestin-induced ablation of endometrial gland genesis in neonatal ewes involves suppression of epithelial ER expression [19]. Ablation of endometrial gland genesis in neonatal ewes treated with norgestomet from birth may reflect the loss or attenuation of ER-dependent signaling [19]. Ovarian estradiol-17ß and growth factors such as IGF-I, IGF-II, and EGF also likely are involved in mediation of the endometrial adenogenesis characteristic of the proliferative phase of the menstrual cycle [115, 123]. Critical experiments remain to be conducted, but gland morphogenesis in the neonatal ovine endometrium is most likely an ER-dependent phenomenon, not unlike that described for the neonatal pig [16, 54, 55] and rat [68].
Estrogen receptor can be activated by estrogens in a ligand-dependent manner or by growth factor-coupled pathways in a ligand-independent manner [125, 141]. Results of transient transfection experiments indicate that ligand-independent ER activation can be induced by many factors, including dopamine, EGF, transforming growth factor , heregulin, and IGF-I [124, 141–144]. The EGF-like growth factor, heregulin, stimulated proliferation and progesterone receptor (PR) gene expression in an ER-dependent manner in MCF-7 cells [143]. In the neonatal ovine uterus and proliferative-phase human endometrium, high levels of PR expression were detected in uterine epithelia and attributed to ER activation [41, 145]. Progesterone levels in serum of neonatal ewes and follicular-phase women are low or undetectable, and these levels likely are not sufficient to inhibit endometrial gland development [41]. The precise roles and significance of ligand-dependent and ligand-independent actions of ER in endometrial gland morphogenesis, however, remain to be determined.
Prolactin
In other epitheliomesenchymal organs, prolactin (PRL) stimulates epithelial differentiation and development in a manner that may be facilitated by cooperative ECM signaling [97, 146, 147]. Using mammary glands from PRL receptor (PRL-R)-null mice, Brisken et al. [148] demonstrated that PRL affects lobuloalveolar development directly, but that it does not affect alveolar bud formation or ductal side branching. In neonatal ewes, serum PRL concentrations are high at birth, increase between PNDs 1 and 14, and then decline to PND 56, which is a pattern that correlates positively with the onset of endometrial gland proliferation in the developing uterine wall [41, 149]. Indeed, mRNAs for both the short and long PRL-R proteins were expressed in nascent and proliferating endometrial GE, and their relative levels of expression increased sevenfold between PNDs 7 and 56 [42]. In the adult endometrium of sheep, humans, and primates [26, 150, 151], the PRL-R gene is expressed exclusively by endometrial glands, and in ewes, increased PRL-R expression during pregnancy correlates with hyperplasia and hypertrophy of endometrial glands [26]. Expression of the PRL-R gene is associated with GE in the deeper coiled and branched regions of endometrial glands of neonatal and adult ovine endometrium. Whether PRL and the PRL-R system play a role in terminal gland development, bud formation, or branching morphogenesis in the neonatal ovine uterus, however, remains to be determined.
In the adult mouse, rabbit, and pig [152–154], hyperprolactinemia stimulates uterine glandular hypertrophy. Intrauterine administration of placental lactogen, a member of the PRL/growth hormone (GH) family, stimulated proliferation of endometrial glands, particularly the terminal ends of coiled, branched glands found in the deeper stroma of adult ewes [155]. Given the central role of PRL-R in mammary gland morphogenesis and function [97, 148], expression of PRL-R in nascent GE of the developing neonatal ovine uterus may play a similar role.
Several lines of evidence support the idea that proliferation of uterine GE and genesis of uterine glands could involve PRL-R-dependent, estrogen-independent activation of ER in developing endometrial GE. Prolactin can stimulate an increase in ER expression in the rat [156]. Activation of both the short and long forms of PRL-R stimulates MAPK signaling [157, 158]. Thus, PRL stimulation of PRL-R in developing uterine GE could activate the MAPK-signaling cascade, resulting in serine phosphorylation, ligand-independent activation of ER, up-regulation of ER, and IGF-1R gene expression [41, 42]. Other complementary pathways involving the PRL-R system could also be involved in regulation of the development and proliferation of uterine GE. In human endometrial stroma, PRL-R mRNA expression is up-regulated by both IGF-I and estrogen [159]. Moreover, stimulation of human secretory-phase endometrium with PRL increases expression of interferon regulatory factor 1 by GE [160], which may increase cell proliferation [161]. These data, taken together with the fact that PRL-R gene expression has been documented in adult endometrium of sheep, humans, mice, rats, and primates [26, 150, 151, 160] in association with periods of proliferation and hypertrophy of GE, support the idea that this receptor system and its cognate ligands are important mechanistic components of the developmental program that regulates endometrial morphogenesis and uterine gland development during both neonatal and adult life.
|
REGULATION OF ADULT UTERINE GLAND MORPHOGENESIS AND FUNCTION BY A SERVOMECHANISM |
|---|
|
TOPABSTRACTINTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
(IFN), PL, and GH, which may activate and maintain endometrial remodeling, secretory function, and uterine growth [155]. Interferon- serves as the signal for maternal recognition of pregnancy in ruminants, is produced by mononuclear cells of the conceptus trophectoderm between Days 8 and 21 in sheep (maximally on Days 15–16), and acts in a paracrine manner on the adult endometrium [162]. It also maintains pregnancy by preventing development of the endometrial luteolytic mechanism that, in turn, maintains the corpus luteum and maternal progesterone production [163]. Progesterone sequentially down-regulates PR expression in LE and GE, which is necessary for the onset of progesterone-induced secretory gene expression as well as glandular remodeling and differentiation [155]. Secretion of PL by the binucleate cells of the conceptus trophectoderm begins on Day 16 of pregnancy, with peak secretion occurring from Days 120 to 130 of gestation [164, 165]. Placental lactogen can bind to the long form of the PRL-R [166] as well as to a heterodimer of a PRL-R and GH receptor (GH-R) [167], which is expressed exclusively by GE and increases throughout gestation [26]. Growth hormone is secreted by the ovine placenta on Days 35–70 of gestation and binds to endometrial GH-R [168]. Sequential intrauterine administration of ovine IFN to ewes from Days 11 to 15 postestrus, followed by ovine PL from Days 21 to 25, increased proliferation of GE in the deep stratum spongiosum. Furthermore, infusion of ovine GH into the uterine lumen from Days 21 to 25 increased uterine gland density in the deep stratum spongiosum and increased the size of endometrial glands in the shallow stratum spongiosum [155]. Results of these studies indicate that a developmentally programmed sequence of events, mediated by specific paracrine-acting factors at the conceptus-endometrial interface, ultimately supports both endometrial remodeling and up-regulation of uterine secretory activity (i.e., increased expression of uterine milk protein and osteopontin genes) during ovine gestation. Whether similar gestational servomechanisms regulate uterine gland development and function in other species remains to be determined. However, strategic manipulation of such mechanisms may offer therapeutic schemes designed to improve uterine capacity, conceptus survival, and reproductive health.
|
|
SUMMARY AND FUTURE DIRECTIONS |
|---|
|
TOPABSTRACTINTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
Studies of conceptus development in the ovine UGKO model [15, 16, 21], in which endometrial gland development was prevented by strategic endocrine disruption of early postnatal uterine organizational events, provided definitive evidence that peri-implantation conceptus survival and growth can be related directly to the presence and state of development of endometrial glands in the adult uterus. These observations reinforce the idea that endometrial organizational mechanisms are critical determinants of functional uterine capacity and, therefore, must be defined. In humans and menstruating primates, regular, cyclical recrudescence of the endometrial functionalis offers repeated opportunities for developmental disruption. Such organizationally induced alterations in human endometrial gland formation and function may lead to infertility and early pregnancy loss. Consistent with observations in other species, endometrial glands and their secretions are critical for embryo development in humans. Decreased expression of cell-surface and secretory proteins within the human uterine environment correlate with abnormal uterine gland morphology that adversely affects uterine receptivity during the peri-implantation and early stages of placentation [166–171]. Future investigations into mechanisms regulating the process of uterine gland development and endometrial morphogenesis will provide insight regarding factors affecting early embryonic survival and development in both humans and livestock.
|
FOOTNOTES |
|---|
1 Supported in part by NIH grant R01-HD38274 to T.E.S., NRI Competitive Grants Program/USDA grant 98-35203-6322 to T.E.S. and grant 95-37203-1995 to F.F.B., and NIH grant P30 ES09106. 
2 Correspondence: Thomas E. Spencer, Center for Animal Biotechnology and Genomics, 442 Kleberg Center, 2471 TAMU, Texas A&M University, College Station, TX 77843-2471. FAX: 979 862 2662; tspencer{at}ansc.tamu.edu
3 Current address: Department of Animal and Veterinary Science, Center for Reproductive Biology, University of Idaho, Moscow, ID 83844-2330.
Accepted: June 4, 2001.
Received: March 8, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONCOMPARATIVE DEVELOPMENTAL...MECHANISMS REGULATING...REGULATION OF ADULT UTERINE...SUMMARY AND FUTURE DIRECTIONSREFERENCES |
|---|
