| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Minireview |
Center for Animal Biotechnology and Genomics,3
Department of Veterinary Anatomy and Public Health,4 College of Veterinary Medicine,
Department of Animal Science,5 College of Agriculture and Life Sciences, Texas A&M University, College Station, Texas 77843
 |
ABSTRACT |
|---|
conceptus, implantation, placenta, pregnancy, uterus
|
INTRODUCTION |
|---|
|
OSTEOPONTIN STRUCTURE AND FUNCTION |
|---|
In general, OPN is a monomer ranging in length from 264 to 301 amino acids that undergoes extensive posttranslational modification, including phosphorylation, glycosylation, and cleavage, resulting in molecular mass variants ranging from 25 to 75 kDa (see Fig. 1). OPN contains a hydrophobic leader sequence characteristic of a secreted protein, a potential calcium phosphate apatite binding region of consecutive asparagine residues, a cell attachment GRGDS sequence, a thrombin cleavage site, and two glutamines that are recognized substrates for transglutaminase-supported multimer formation [18]. Genes encoding OPN have been cloned from eight species, including rat [15], mouse [10], human [20, 21], cow [22], chicken [23], rabbit [24], and sheep [25]. The cDNA sequences from these species reveal only moderate conservation [25], except in the NH2-terminal region, around the Arg-Gly-Asp (RGD) integrin-binding sequence, and the COOH-terminus [18]. Sequences that are highly conserved in all these species include 1) 4 of the 9 or 10 residues in the poly-Asp region, 2) the GRGDS sequence, 3) 15 serine residues that include an SSEEK sequence (residues 26–30), 4) three threonines (residues 131, 140, and 145), 5) an NES sequence (residues 79–81), and 6) glutamines at positions 50 and 52. Two features shared by ovine and bovine OPN but not other species are deletion of 22 amino acids that would otherwise be inserted between residues 196 and 197 and replacement of the RS with a KS thrombin cleavage site [18].
|
OPN undergoes extensive posttranslational modifications believed to be important to its function. These modifications include proteolytic cleavage [26–28], phosphorylation on serine and threonine residues [13, 28], glycosylation with primarily O-linked oligosaccharides [29, 30], cross-linking with self and other macromolecules through transglutaminase [31, 32], and sulfation [33]. Originally isolated from bone [11], OPN has been found on epithelial cells and in secretions of the gastrointestinal tract, kidneys, thyroid, breast, uterus, placenta, and testes [25, 34–40]. OPN is also expressed by leukocytes, smooth muscle cells, and highly metastatic cancer cells [41–43]. OPN has been reported to 1) stimulate cell-cell adhesion, 2) increase cell-ECM communication, 3) promote migration of immune cells, osteocytes, and tumor cells, 4) decrease cell death by reducing reactive oxygen species and nitric oxide production by injured tissues, 5) stimulate immunoglobulin production by B cells, 6) induce changes in the phosphorylation state of focal adhesion kinase and paxillin, 7) stimulate phosphotidylinositol 3'-kinase activity, 8) alter intracellular calcium levels, and 9) affect tissue mineralization and promote calcium phosphate deposition in bone [44–54]. Increases in transcription of the OPN gene are induced by interleukin (IL) 1
and ß, transforming growth factor ß-1, fibroblast growth factor, tumor necrosis factor , interferon (IFN) 
, estrogen, progesterone, glucocorticoids, and 1,25-dihydroxy vitamin D3 [55–61].
The RGD amino acid sequence of OPN interacts with cell surface integrin receptors to mediate cell adhesion, migration, differentiation, survival, and immune function [44, 62–69]. Integrins are dominant glycoproteins in adhesion cascades. They belong to a ubiquitous family of cation-dependent, heterodimeric intrinsic transmembrane glycoprotein receptors composed of noncovalently bound and ß subunits that participate in cell-cell and cell-ECM adhesion, cause cytoskeletal reorganization to stabilize adhesion, and transduce signals through numerous signaling intermediates [70–72]. Although generally it has been accepted that OPN binds primarily to vß3 integrin heterodimers via its RGD sequence, recent identification of other receptors and alternative binding sequences has expanded this signaling repertoire (Fig. 1). The vß1, vß5, and 8ß1 integrins have affinities for the RGD motif of OPN that are similar to that of vß3 [73, 74], and OPN has binding affinity for the leukocyte integrins mß2 and xß2 [75]. Evidence for non-RGD-mediated integrin binding to OPN is illustrated by the discovery of dual 4ß1 binding sites through which OPN promotes leukocyte adhesion [76, 77]. There is also a cryptic binding sequence recognized by 9ß1 [78]. OPN is also an extracellular ligand for certain splice variants of the cell surface proteoglycan CD44 (CD44v3-v6), where it serves as a proadhesion cytokine immobilized on the luminal surface of endothelium to recruit leukocytes and trigger conversion of integrin expression to an active adherence configuration [27, 49, 79]. CD44 variants may require cooperation with ß1 to bind multiple yet unidentified domains in OPN and stimulate cell motility and chemotaxis [80]. Additionally, intracellular OPN is part of a hyaluronan-CD44-ERM (ezrin/radixin/moesin) attachment complex involved in fibroblast, macrophage, and tumor cell migration, suggesting that OPN stimulates or participates in motogenic activity of cells [81, 82].
The different signals that OPN elicits are attributed to its multiple receptors and binding sites and to its various molecular/structural forms. Upon freezing and thawing or treatment with serine proteases such as thrombin, the 70-kDa native protein gives rise to 24-kDa and 45-kDa fragments that were initially shown to bind antigen and to nonspecifically suppress T-helper lymphocytes, respectively [26, 27]. The 45-kDa amino-terminal OPN fragment retaining the RGD sequence was subsequently shown to improve cell attachment and spreading properties through increased accessibility of the RGD to integrin receptors [83]. Only the 45-kDa OPN supports 9ß1 and 4ß1-mediated cell migration and adhesion [76, 84] and attachment through several other non-RGD sites [85]. OPN is also a substrate for cleavage by matrix metalloproteinases (MMPs) through one novel site cleaved by both MMP-3 (stromelysin 1) and MMP-7 (matrilysin) and another site cleaved by MMP-3 alone (Fig. 1). Fragments formed by these cleavage events bind cell surface integrins to initiate adhesion and migration [28]. OPN exists in a polymerized form cross-linked by tissue transglutaminase (Fig. 1), which alters OPN conformation to increase binding to collagen. In bone, polymerized OPN is hypothesized to participate in cell adhesion, matrix assembly and maturation, and calcification by simultaneously binding multiple receptors on different cells [32]. Homotypic OPN-OPN bonds have high tensile strength in vitro, supporting the idea that self-assembly is involved in cell-cell and cell-matrix adhesion [86].
|
OPN IS A MAJOR COMPONENT OF HISTOTROPH |
|---|
The original observation that OPN is expressed by uterine epithelia arose from studies to determine possible roles of 2ar/OPN in the cell cycle and in cell proliferation during mouse embryogenesis [104]. In situ hybridization indicated high levels of OPN mRNA in a number of epithelial tissues, including kidney tubules, sensory epithelium of the embryonic ear, and uterine endometrium [104]. Although uterine LE and GE adjacent to deciduoma expressed OPN whereas epithelia in the contralateral horn were negative, the authors did not focus on OPN expression [104]. Subsequent immunocytochemical studies by Young and colleagues [35] localized OPN protein to secretory phase endometrial GE of cyclic women. Because OPN was absent from GE during the proliferative phase of the menstrual cycle, it was suggested that changes in expression in GE of normal cycling endometrium were the result of hormonal regulation and that the function(s) of OPN in the endometrium was likely associated with its ability to enhance cell attachment [35]. A significant advance in thinking about the function of epithelium-derived OPN was made in 1992, when OPN mRNA and protein were localized to the epithelia of a variety of organs, including the hypersecretory endometrial GE associated with pregnancy in women [36]. The investigators hypothesized that OPN secreted by epithelium binds integrins on luminal surfaces to effect communication between LE and the external environment [36].
A substantial body of work in the 1990s established that transient endometrial expression of vß3 and 4ß1 integrins is dependent on the stage of the sexual cycle and defines the window of implantation in women [4, 105, 106] and that altered expression of these integrins is correlated with several conditions associated with human infertility [3, 5]. In mice, null mutations of v, 5, ß1, or ß5 integrin genes result in peri-implantation lethality and failure of chorioallantoic membrane fusion [107], and a functional blockade of v and ß3 integrins in mice and rabbits reduces the number of implantation sites [108, 109]. Noting that the vß3 and 4ß1 integrin heterodimers present during the implantation window bind the RGD amino acid sequence in OPN, Lessey and coworkers [4] suggested that involvement of OPN and integrins in trophoblast-endometrial interactions during the initial attachment phase of implantation is plausible and merits closer examination. This suggestion was reinforced by observations that CD44, another OPN receptor, is present through the blastocyst stage of human preimplantation development [110] and that uterine epithelium expresses CD44 and its larger variant CD44 isoforms [111].
A comprehensive examination of temporal and spatial expression of uterine OPN mRNA and protein in sheep provided the first clear evidence that OPN is a secretory product of endometrial glands (histotroph) that binds integrins on endometrial LE and Tr. Comparison of expression patterns between in situ hybridization and immunofluorescence analyses of cyclic and pregnant ovine uterine sections revealed a significant divergence. OPN mRNA increased in the endometrial GE during the peri-implantation period; however, it was not present in LE or conceptus Tr [25]. In contrast, immunoreactive OPN protein was present at the apical surfaces of endometrial LE and GE and on Tr [38]. A 45-kDa proteolytic fragment of OPN, known to be more stimulatory to cell attachment and migration than native OPN [83], was detected exclusively on LE and Tr (Fig. 2). A wide array of OPN proteins representing the native 70-kDa form, 45- and 25-kDa cleavage fragments, and probable high-molecular-weight transglutaminase-stabilized multimers (Fig. 2) were detected by Western blot analysis of endometrium from pregnant ewes [38]. In addition, the integrin subunits v, 4, 5, ß1, ß3, and ß5, which could contribute to the assembly of several OPN receptors including vß3, vß1, vß5, 4ß1, and 5ß1 heterodimers, were constitutively expressed on Tr and the apical surface of endometrial LE [38, 103]. A key finding was an increase in the amount of 45-kDa OPN in uterine flushings from pregnant ewes during the attachment phase of implantation [38]. Pregnant Day 14 UGKO ewes, which lack uterine glands, exhibit an absence of OPN in uterine flushings compared with normal bred ewes [102]. These results indicated that OPN is a component of histotroph secreted from GE into the uterine lumen of pregnant ewes and that OPN binds integrin receptors expressed on endometrial LE and conceptus Tr, where it can stimulate changes in morphology of Tr and mediate adhesion between LE and Tr essential for implantation and placentation [38].
|
Uterine OPN expression has been detected during the peri-implantation period in humans [112, 113], pigs [114], and rabbits [115], confirming that OPN is an important epithelial protein involved in implantation in several mammalian species with diverse types of implantation and placentation. Results from two studies in the human endometrium were highly similar to results from sheep regarding the temporal/spatial deposition of OPN. In both studies, OPN mRNA and protein increased in the GE during the mid to late secretory phase [112, 113], and Apparao et al. [112] localized OPN protein, but not mRNA, to the apical LE surface. OPN protein was also present in GE cells showing evidence of secretory activity [112], and increasing levels of OPN were detected in uterine secretions from the secretory phase in women [113]. Hence, it is probable that in humans OPN is synthesized in the GE and secreted into the uterine lumen and then binds to integrins on the LE. In rabbits, OPN mRNA and protein are also expressed by the endometrial GE during the peri-implantation period. OPN expression in GE increases at the time of initial conceptus attachment but is not detectable in the nonpregnant endometrium [115]. Pigs have a novel expression pattern for OPN mRNA. Unlike in other species, in pigs OPN is transcribed directly in endometrial LE, and GE mRNA expression does not begin until Day 35 of pregnancy [114]. This unique expression of OPN mRNA in LE begins in discrete regions of the endometrial LE associated with the presence of the conceptus prior to conceptus attachment and expands to the entire LE by Day 20, when firm adhesion of conceptus to uterus has occurred [114]. The pig has an epitheliochorial type of placentation in which the LE remains intact throughout pregnancy, whereas other mammals undergo differing degrees of LE degeneration and fusion with Tr during placentation. Therefore, the LE of pigs provides an uninterrupted supply of OPN protein at the interface between LE and conceptus during implantation, whereas other species use OPN synthesized and secreted from GE to ensure continued expression at these sites. The expression of OPN mRNA in porcine LE offers strong evidence that OPN at the maternal-fetal interface is critical and emphasizes that functional properties of LE and interactions between LE and Tr are essential to the processes of implantation and placentation. In support of this hypothesis, the SPP1 gene (OPN) has recently been reported as one of several putative quantitative trait loci on porcine chromosome 8 associated with reproductive performance [116]. Specifically, SPP1 is a candidate for the litter size and prenatal survival traits based on positional data and growing physiological evidence for the involvement of OPN in implantation and placentation.
|
OPN IS A PROGESTERONE-REGULATED ENDOMETRIAL SECRETORY PROTEIN |
|---|
, the pregnancy recognition signal in sheep [118]. Although intrauterine infusion of IFN had no effect on OPN gene expression, exogenous progesterone induced OPN mRNA in GE and increased secretion of the 45-kDa OPN cleavage product. This upregulation of OPN in GE is correlated with silencing of expression of progesterone receptors (PRs) in uterine LE and superficial GE, although PR continues to be expressed in stroma and myometrium. It has been suggested that effects of progesterone on OPN expression in ovine GE are to downregulate epithelial PR to allow for upregulation of OPN and/or to regulate a growth factor(s) produced by the stroma or myometrium (progestamedin) that mediates progesterone effects on GE [118]. Administration of ZK137.316, a PR antagonist, ablated progesterone effects in these ewes [118]. In the second study, injection of progesterone alone or in combination with ZK137.316 convincingly created a stimulatory effect of progesterone on endometrial OPN mRNA expression in rabbits [115]. Progesterone induction of OPN mRNA in unmated rabbits was prevented by addition of the PR antagonist, and estrogen priming did not influence effects of progesterone. Although there is no direct evidence for progesterone regulation of OPN in the endometrium of women, in a third study Apparao and coworkers [112] demonstrated effects of progesterone on OPN expression in transformed human endometrial epithelial (Ishikawa) cells. Expression of OPN was increased by in vitro treatment of cells with estrogen plus progesterone, whereas cotreatment with progesterone and RU-486 reduced the stimulatory effect of progesterone on OPN [112]. In women, where PR-A and -B isoforms are differentially regulated in the endometrium during implantation, it is likely that endometrial OPN is modulated by PR-B alone. Progesterone treatment of cells transfected with PR-A or -B resulted in an increase in OPN in cells containing PR-B but not in those containing PR-A [119]. Global gene profiling using high-density microarray technology has identified genes that either increase or decrease in endometrium as it transitions to an implantation receptive state. About 20% of the changes are attributed to genes encoding cell surface receptors, adhesion and ECM proteins, and growth factors [120]. Two studies of human endometrium have confirmed that OPN expression is upregulated during the receptive phase for implantation. A comparison of gene expression in endometrium from Days 8–10 of the cycle and that from Days 8–10 after midluteal LH surge showed an 8.1-fold increase in OPN mRNA [121]. Similarly, microarray comparison of endometrium from Days 2–4 and Days 7–9 post-LH surge showed a 12.3-fold upregulation [120]. OPN is the most highly up-regulated ECM-adhesion molecule in the human uterus as it becomes receptive to implantation. Previous reports indicated that this upregulation occurs in the endometrial GE and results in increased secretion of OPN into the uterine lumen [35, 36, 112], suggesting a role in uterine-conceptus interactions during early adhesion of the conceptus to LE.
Recently, another microarray study extended uterine gene expression profiling throughout pregnancy in rats. OPN gene expression increased 60-fold between Day 0 of the estrous cycle and Day 20 of pregnancy, the second greatest increase of all genes measured [122]. It is unclear what cell type is responsible for the increase in endometrial OPN expression in rats, but both endometrial OPN mRNA and protein increase in pigs and sheep throughout gestation. In pigs, total endometrial OPN mRNA increases 20-fold between Day 25 and Day 85 of pregnancy. Although OPN mRNA continues to be present in endometrial LE, GE is responsible for the significant increase in OPN expression during gestation. OPN mRNA in GE was first detected on Day 35 and increased markedly between Days 40 and 85 [114] of gestation. Progesterone is likely responsible for this expression, because it regulates endometrial secretory activity in the pig, and secretory activity increases after Day 35 of pregnancy to maximal levels between Day 60 and Day 75 [92], a temporal pattern that mirrors transcriptional activity of the OPN gene in GE [114].
In sheep, the increase in OPN mRNA during pregnancy is even greater than that in pigs. Steady-state levels of OPN mRNA in total ovine endometrium remain constant between Day 20 and Day 40, increase 40-fold between Day 40 and Day 100, and remain maximal thereafter [123]. The major source of this OPN is GE, which undergoes hyperplasia through Day 50, followed by hypertrophy and maximal production of histotroph after Day 60 [124]. Endometrial gland morphogenesis and increases in the OPN from GE are likely influenced by uterine exposure to estrogen, progesterone, IFN, placental lactogen, and placental growth hormone. Sequential treatment of ovariectomized ewes with progesterone, IFN, placental lactogen, and growth hormone results in GE development similar to that observed during normal pregnancy. Administration of progesterone alone in these experiments induced expression of OPN in GE, and intrauterine infusion of IFN and placental lactogen to progesterone-treated ovariectomized ewes increased OPN mRNA levels above those for ewes treated with progesterone alone. Intrauterine infusion of recombinant ovine placental growth hormone did not affect OPN expression, although it did increase levels of uterine milk protein (serine protease inhibitor), another GE secretory product in sheep [125, 126]. Immunofluorescence microscopy revealed that secreted 45-kDa OPN cleavage fragment is exclusively, continuously, and abundantly present along the apical surface of LE, on Tr, and along the entire uterine-placental interface of both interplacentomal and placentomal regions through Day 120 of pregnancy [123] (Fig. 2). These findings are significant because they definitively localize a secretory product of the GE to regions of intimate contact between conceptus and uterus where OPN may influence fetal/placental development and growth and mediate communication between placental and uterine tissues to support pregnancy.
|
OPN BINDS AND ACTIVATES INTEGRINS IN ENDOMETRIAL LE AND TR |
|---|
vß3 integrin heterodimer by intrauterine injection of neutralizing antibodies against the respective integrin subunits, by masking ligand binding sites via intrauterine injection of RGD containing peptides, and by intrauterine injection of the disintegrin echistatin and reported that each procedure reduced numbers of implantation sites in mice. A similar study in rabbits demonstrated reduction in the number of implantation sites after intrauterine injection of neutralizing antibody or GRGDSP hexapeptides, whereas GRGESP hexapeptides had no effect [109]. Although these experiments do not address the role of OPN directly, they illustrate a critical role for vß3 and, by inference, one or more of its ligands, including OPN, during implantation. Immortalized and primary ovine and porcine endometrial LE and conceptus Tr cells, which abundantly express v and ß3 heterodimers at focal adhesion sites of cell anchorage to the basal substrate [103], show evidence of integrin activation via OPN binding. Transmembrane accumulation of talin at the apical surface of these cells in contact with OPN-coated beads revealed functional integrin activation and cytoskeletal reorganization to form in vitro focal adhesions in response to OPN binding. However, OPN-coated beads in which the RGD sequence was mutated to RGE or RAD failed to induce focal adhesions [103, 114].
Functional evidence for integrin-mediated signal transduction in vivo at ovine implantation sites is presented in Figure 3. When serial frozen sections of implantation sites from Day 45 of pregnancy were stained with antibodies to v, ß3, ß5, the ECM proteins OPN and fibronectin, and antibodies to detect focal adhesion formation to cytoskeletal proteins talin, vinculin, or -actinin, nearly identical intense punctate staining for v, ß5, and -actinin was observed. Colocalization of the vß5 integrin receptor with -actinin indicates the presence of specific ligand-activated integrins at implantation sites to form focal adhesions in sheep. In vivo focal adhesions have rarely been described [130]. OPN protein is present along the entire maternal-placental interface in sheep through Day 120 of pregnancy (Fig. 2) and may be an ECM ligand bound to these integrin receptors that is responsible for induction of the focal adhesions observed in vivo.
|
|
OPN IS A MARKER FOR STROMAL DECIDUALIZATION |
|---|
-smooth muscle actin [154], and ECM components such as OPN [155]. The end result is the formation of a morphologically and functionally distinct tissue that is a source of hormones, promotes conceptus nutrition, prevents fetal allograft rejection, and regulates placentation by limiting Tr invasion through generation of a local cytokine environment that promotes Tr attachment over invasion [156–158].
The earliest reports of OPN mRNA in uteri of mammals noted high expression in a population of periodic acid-Schiff-positive cells corresponding in size and distribution to granulated metrial gland cells in mice [104]. These OPN-positive cells were localized in the central zone and metrial triangle, extending to more antimesometrial regions adjacent to trophoblast giant cells. Both the pseudopregnant deciduoma resulting from intraluminal injection of oil and the endometrium in the uninjected horn of these mice contained OPN-positive cells. A similar pattern of OPN mRNA expression was reported by Waterhouse and coworkers [155] in 1992, who defined OPN as a marker gene for decidualization. In humans, OPN mRNA and protein have been localized in human decidual cells by reverse transcription polymerase chain reaction (RT-PCR), Northern blot, and in situ hybridization procedures. These decidual cells also exhibited intense cytoplasmic immunoreactivity to antiserum against human bone OPN [35]. Eleven years later, Apparao et al. [112] reported identical immunostaining in the decidua of human pregnancy and stressed the striking similarity in the temporal and spatial patterns of expression of the ligand OPN and its integrin receptor vß3. These results confirmed previous results showing coexpression of OPN and vß3 integrins in decidualizing stromal cells of baboons. Fazleabas et al. [117] concluded that increased stromal expression of specific integrins during pregnancy, along with changes in ECM such as OPN, may be essential for stromal cell proliferation and differentiation. Much of the 60-fold increase in OPN gene expression in the rat uterus during pregnancy probably occurs in decidua, because there is a paucity of endometrial GE in rats [122].
OPN mRNA and protein have also been detected in the stroma of pregnant sheep [159]. Although Mossman [160] and Kellas [161] described decidual cell characteristics in syncytial tissues of placentomal crypts of sheep and antelope, these reports have been largely ignored, and decidualization is not thought to occur in species with central and noninvasive implantation, including domestic animals. However, endometrial stromal cells increase in size and become polyhedral in shape in pregnant ewes following conceptus attachment. The classical decidualization markers desmin and -smooth muscle actin are expressed in these cells, suggesting that OPN expression in this stromal compartment is part of a decidualization-like differentiation during ovine pregnancy [159]. In contrast, no mophological changes or induction of OPN mRNA or protein have been detected in uterine stromal cells during porcine pregnancy [159]. One of the primary roles of decidua in invasive implanting species is to restrain conceptus trophoblast invasion to a circumscribed region of the endometrium. Both pigs and sheep have noninvasive implantation, but the extent of Tr invasion into the endometrium differs between these species. Pig conceptuses undergo true epitheliochorial placentation in which LE remains morphologically intact throughout pregnancy and the conceptus Tr simply attaches to the apical LE surface without contacting uterine stromal cells [162, 163]. Synepitheliochorial placentation in sheep involves extensive erosion of the LE due to syncytium formation with Tr binucleate cells. After Day 19 of pregnancy, conceptus tissue is apposed to but does not penetrate ovine uterine stroma [164]. Although speculative, differences in stromal expression of OPN between these species suggest that the extent of decidualization is correlated positively with degree of conceptus invasiveness. OPN expression increases in ovine stratum compactum stroma that has undergone considerable loss of ECM by Day 35 of pregnancy [159]. A possible explanation for this concomitant increase in OPN and decrease in overall ECM might be that stromal OPN is intracellular. OPN has generally been characterized as a protein secreted into biological fluids and luminal spaces, such as the uterine lumen, or into the ECM where it binds receptors to mediate various cell-cell and cell-ECM interactions. However, an intracellular form of OPN associated with a hyaluronan-CD44-ERM attachment complex has been implicated in migration of embryonic fibroblasts, activated macrophages, and invasive cancer cells. As such, intracellular OPN may modulate CD44-mediated changes in cytoskeletal architecture involved in cell differentiation and migration [81, 82]. The nature of the potential intracellular and extracellular OPN interactions with stromal cells will be critical to understanding the function of OPN during decidualization and placentation.
|
CONCEPTUS EXPRESSION OF OPN |
|---|
v integrin [60]. In hydatidiform mole, the most common type of gestational trophoblastic disease, there is a downregulation of OPN mRNA and protein, compared with normal placentae, concomitant with decreased levels of progesterone and PR [167]. In contrast to mice and humans, OPN mRNA is not expressed by sheep, pig, or rabbit conceptus tissues. However, in sheep and pigs very high levels of immunoreactive OPN protein, presumably supplied in histotroph and bound to placental integrin receptors, are detected on Tr beginning just prior to implantation and maintained throughout gestation [38, 112, 114, 123]. Although OPN is not synthesized by Tr in these species, Tr exhibits greater immunostaining for OPN than does any other cell type in the pregnant uterine environment. This high affinity may indicate essential roles for OPN in mediating embryonic placental changes in morphology and adhesion to the maternal uterus during implantation and placentation. |
OPN EXPRESSION BY ENDOMETRIAL AND PLACENTAL IMMUNE CELLS |
|---|
Johnson and coworkers [25] first described OPN mRNA in a small population of cells scattered within the stratum compactum stroma immediately beneath the endometiral LE during the early stages of the estrous cycle and pregnancy in sheep. Unlike OPN mRNA in endometrial GE, expression of OPN by these cells and/or migration of this cell population through the endometrium was not regulated by administration of exogenous progesterone to ovariectomized ewes [118]. Similar results were observed in pigs in which both OPN mRNA and protein were detected in a small population of cells in the endometrium on Day 9 of the estrous cycle and early pregnancy exclusively [114]. Immunostaining for OPN and various immune cell surface antigens in ovine serial uterine cross sections indicated that OPN-producing cells in the uterus may be CD8-bearing T cytotoxic or natural killer cells or CD172a-positive macrophages [168]. Both macrophages and natural killer cells are believed to influence the pregnant endometrium of humans and mice [176, 177]. In primates, OPN mRNA was detected in a population of CD45-positive leukocytes isolated from mid and late secretory phase human endometrium [113] and has been detected in scattered endometrial stromal cells, believed to be of immune origin, during the menstrual phase of cynomolgus monkeys [178]. OPN-producing immune cells are also present in the ovine placenta beginning on Day 25 of pregnancy, and the number of these cells increases as gestation progresses (Fig. 4). In contrast, OPN mRNA was never detected in the placenta of pigs at any stage of gestation studied, and differences in placental immune cell expression of OPN between sheep and pigs may relate to type of placentation [168].
|
Immune cell OPN is of particular interest because of its role in maintenance and remodeling of tissue during inflammation, where it mediates cell-cell and cell-matrix communication resulting in targeted chemotaxis of immune cells [27]. In vitro, OPN can induce migration of both T cells and macrophages [179, 180], and in vivo macrophages migrate toward OPN injected into rat dermis [181]. OPN induces chemotactic migration of dendritic cells from the skin to draining lymph nodes, a process required for many inflammatory responses [182]. Movement of leukocytes into tissues involves a cascade in which adhesion to endothelium is critical. OPN becomes immobilized on endothelium by CD44 to recruit leukocytes to the endothelial surface and triggers activation of integrins to an active adherence configuration [49, 79]. An intracellular form of OPN associates with the CD44 complex in cell processes and is believed to function in the migration of osteoclasts, which are specialized bone cells of macrophage lineage [183].
OPN is a key component of type 1 (Th-1) immunity characterized by increases in IL-12 and decreases in IL-10 expression by macrophages and monocytes [41, 83, 184]. Macrophage production of IL-12 promotes a cell-mediated or Th-1 immune response that effectively protects against the growth of viral and bacterial pathogens and transformed cells, whereas IL-10 inhibits this response and favors humoral (Th-2) immunity [185, 186]. Although conclusions of recent studies are being debated, OPN appears to regulate Th-1-mediated demyelinating disease in rats paralyzed by experimental autoimmune encephalomyelitis (EAE), a model for multiple sclerosis. OPN-deficient rats are resistant to progressive EAE and have higher levels of IL-10 than do OPN-positive mice [187–189]. A similar model in mice revealed that OPN-deficient mice have less severe clinical symptoms and faster recovery without spontaneous relapses than do OPN-positive mice. Deficient mice display decreased inflammatory cell infiltration and demyelination of spinal cords, suggesting that OPN sustains autoimmune responses by increasing and/or maintaining Th-1 immunity [190].
Trophoblasts and many tumors use OPN to selectively escape immune surveillance [191]. As a key component of innate immunity, the alternative complement system is directly activated by foreign epitopes on cell surfaces. Molecules such as the plasma protein Factor-H bind and inactivate inappropriately high levels of complement pathway components in serum [192]. OPN and several related SIBLINGS sequester Factor-H to the cell surface and inhibit complement-mediated cell lysis [191, 193]. Human decay-accelerating factor appears to play a similar protective role for endometrial LE during the secretory phase of the menstrual cycle [194].
|
SUMMARY |
|---|
Figure 5 illustrates a working model for the role of OPN in the pregnant ovine uterus, although data from other species indicate remarkably similar patterns of OPN expression. OPN functions within the uterus appear to be complex and omnifarious, encompassing 1) a role as a component of histotroph involved in adhesion and signal transduction at the uterine-placental interface, 2) expression in uterine stroma, suggesting a decidualizationlike transformation that is correlated with the degree of conceptus invasiveness, and 3) expression by resident uterine and placental immune cells that may regulate immune cell behavior and cytokine production. The regulation and immediate functional implications of these expression events, although independent, are temporally and spatially orchestrated to maintain successful pregnancy. Insights have been gained into the potential hormonal/cytokine, cellular, and molecular mechanisms underlying OPN-mediated adhesion, remodeling, and cell-cell/cell-matrix communication within the uterine-placental environment through comparisons among humans, monkeys, sheep, pigs, and rodents. Each animal model offers unique advantages, and investigations must continue to exploit the comparative aspects of OPN expression and function among species. Abundant and continuous immunoreactive OPN protein is present along all areas of intimate contact between maternal and fetal components of the placenta throughout pregnancy [38, 112, 114, 115, 123].
|
Future investigations must focus on the mechanistic nature of the functions of this multifaceted component of the uterine/placental environment. For example, studies are underway to identify specific signals that are transduced into LE and Tr, respectively, following binding of OPN to integrin and/or CD44 receptors. The specific signals elicited by OPN result not only from its multiple receptors and binding sites but also from the various molecular forms that arise from glycosylation, phosphorylation, and protease cleavage. Therefore investigators are challenged to focus on understanding the regulation and distribution of the various OPN forms within the uterus and placenta. Investigations should include a determination of whether intracellular OPN-CD44-ERM complex is involved in uterine and placental function. The identity of the molecular forms and functional significance of the induction of ovine stromal OPN accompanying the decidualizationlike differentiation of this tissue during implantation should also be informative. Similarly, the identity of OPN-producing immune cells and the determination of intracellular or secreted molecular forms generated is an important question. Specifically, the production of secreted OPN by immune cells would suggest cell-cell or cell-matrix communication resulting in upregulation of Th-1 inflammatory cytokines and/or chemotaxis of immune cells, whereas intracellular OPN would suggest the presence of a population of immune cells that are themselves migrating through the tissue. Other obvious functions to be addressed include potential roles of OPN in placental angiogenesis and in uterine tissue remodeling that occurs during and following pregnancy. Because nitric oxide and polyamines (products of arginine catabolism) are essential for placental angiogenesis and placental and fetal growth [195, 196], the role and mechanisms for OPN in regulating these biosynthetic processes should be defined. Collectively, clear understanding of these and other functions involving uterine OPN should be useful for the rational design of therapies for the prevention of pregnancy failure related to abnormalities in uterine/placental factors as opposed to current empirical therapies that have low efficiency.
|
FOOTNOTES |
|---|
2 Correspondence: Greg A. Johnson, Department of Veterinary Anatomy and Public Health, College of Veterinary Medicine, Texas A&M University, College Station, TX 77843-4458. FAX: 979 847 8981; gjohnson{at}cvm.tamu.edu
Received: 25 June 2003.
First decision: 8 July 2003.
Accepted: 16 July 2003.
|
REFERENCES |
|---|
TOP
ABSTRACT
INTRODUCTION
