| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Center for Animal Biotechnology and Genomics,
b Department of Veterinary Anatomy & Public Health, College of Veterinary Medicine
c Department of Animal Science, Texas A&M University, College Station, Texas 77843-2471
d Department of Veterinary Anatomy, Histology and Embryology, Justus-Liebig-University, 35392 Giessen, Germany
ABSTRACT
The extracellular matrix protein osteopontin (OPN) is a component of histotroph that increases in uterine flushings from pregnant ewes during the peri-implantation period and is localized on the apical surfaces of the uterine luminal epithelium (LE) and conceptus trophectoderm (Tr). The potential involvement of OPN in the implantation adhesion cascade in sheep was investigated by examining temporal, spatial, and potential functional relationships between OPN, Muc-1, and integrin subunits during the estrous cycle and early pregnancy. Immunoreactive Muc-1 was highly expressed at the apical surfaces of uterine luminal (LE) and glandular epithelium (GE) in both cycling and pregnant ewes but was decreased dramatically on LE by Day 9 and was nearly undetectable by Day 17 of pregnancy when intimate contact between LE and Tr begins. In contrast, integrin subunits
v, 4, 5, ß1, ß3, and ß5 were constitutively expressed on conceptus Tr and at the apical surface of uterine LE and GE in both cyclic and early pregnant ewes. The apical expression of these subunits could contribute to the apical assembly of several OPN receptors including the vß3, vß1, vß5, 4ß1, and 5ß1 heterodimers on endometrial LE and GE, and conceptus Tr in sheep. Functional analysis of potential OPN interactions with conceptus and endometrial integrins was performed on LE and Tr cells in vitro using beads coated with OPN, poly-L-lysine, or recombinant OPN in which the Arg-Gly-Asp sequence was replaced with RGE or RAD. Transmembrane accumulation of talin or -actinin at the apical surface of uterine LE and conceptus Tr cells in contact with OPN-coated beads revealed functional integrin activation and cytoskeletal reorganization in response to OPN binding. These results provide a physiological framework for the role of OPN, a potential mediator of implantation in sheep, as a bridge between integrin heterodimers expressed by Tr and uterine LE responsible for adhesion for initial conceptus attachment.
implantation, pregnancy, signal transduction, trophoblast, uterus
INTRODUCTION
Implantation in mammals is a highly coordinated process that begins with apposition, attachment, and adhesion of uterine lumenal epithelium (LE) and conceptus trophectoderm (Tr). Remodeling of the glycoprotein adhesion molecules of the apical surfaces of these cells precedes implantation [1]. In ruminants implantation is also influenced by secretions from both the embryo that signals its presence through interferon (IFN)
[2] and the endometrial glands that produce histotroph to nourish the conceptus [3]. In contrast to primates and rodents with invasive implantation, sheep have superficial/central implantation in which a prolonged preattachment period (
15 days) is followed by incremental apposition and attachment until zones of Tr invade and fuse with uterine LE (40 days) [4].
The LE is a simple, polarized cell layer that mediates cell-cell and cell-extracellular matrix (ECM) interactions. The apical LE domain is normally nonadhesive; however, this character is lost during development of receptivity [5, 6] when apical adhesion between LE and Tr defines the onset of implantation. The nonadhesive properties of LE are partially due to apical expression of mucins with extensive glycosylation and extended structure that sterically inhibits cell-cell and cell-ECM adhesion [7, 8]. These properties of the mucin Muc-1 presumably impair interactions between adhesive glycoproteins at the surfaces of LE and Tr [9]. The relationship between the mode of implantation and the spatial and temporal expression of Muc-1 is not obvious. In rodents and pigs, Muc-1 expression on uterine LE decreases just prior to implantation [10, 11], whereas it increases during the receptive phase in rabbits and humans but may be reduced locally at sites of conceptus apposition [12, 13]. Muc-1 expression in sheep uterus has not been reported.
Integrins are dominant glycoproteins in adhesion cascades. They comprise a ubiquitous family of cation-dependent, heterodimeric intrinsic transmembrane glycoprotein receptors that mediate cellular differentiation, motility, and adhesion [14–16]. The central role of integrins in the implantation adhesion cascade stems from their ability to bind ECM ligand(s) to mediate adhesion, cause cytoskeletal reorganization to stabilize adhesion, and transduce cellular signals through numerous signaling intermediates [17–19]. Lessey and coworkers have established that transient endometrial expression of vß3 and 4ß1 integrins is cycle-dependent and defines the implantation window in women [20, 21]. Altered expression of these integrins is correlated with several causes of infertility [22, 23]. Null mutations of v, 5, ß1, or ß5 integrin genes in mice lead to peri-implantation lethality, and failure of chorion-allantois fusion [24], while functional blockade of v and ß3 integrins reduces the number of implantation sites [25]. Endometrial integrin expression of several species that exhibit noninvasive implantation has also been reported including pig, sheep, goat, and cow [11, 26–28]. In the pig, integrin subunits expressed by uterine LE and conceptus Tr potentially form vß3, 4ß1, vß1, and 5ß1 heterodimers at implantation sites [11].
The vß3, 4ß1, and vß1 heterodimers present during the implantation window in humans and pigs bind the Arg-Gly-Asp (RGD) amino acid sequence found in osteopontin (OPN), an ECM ligand that binds integrins to promote cell-cell attachment and spreading [29–31]. Osteopontin is expressed at high levels by epithelium and decidualizing stroma of human uterus [32–34] and cytotrophoblast of the chorionic villus [35, 36]. It is also expressed in invading cytotrophoblast, glandular epithelium (GE), and decidualizing stromal cells in the baboon [37]. Osteopontin is produced by mouse trophoblast and metrial gland cells of decidua and placenta [38], while mRNA and protein are present in LE and GE of pregnant pigs during the peri-implantation period [39]. In sheep, OPN is a component of histotroph that increases in uterine flushings from pregnant ewes between Days 11 and 17 [26]. Although exposure to progesterone induces OPN mRNA expression only in the endometrial GE of pregnant ewes [40, 41], OPN protein is localized on the apical aspect of the endometrial LE, GE, and conceptus Tr [26, 40]. Therefore, it is hypothesized that OPN binds integrin heterodimers expressed by Tr and LE to 1) stimulate changes in morphology of conceptus extraembryonic membranes; and 2) induce adhesion between LE and Tr essential for implantation and placentation [26, 41]. Objectives of this study were to 1) examine relationships between Muc-1, selected integrins, and OPN in the implantation cascade in sheep; and, 2) perform functional analysis of potential OPN interactions with conceptus and endometrial integrins.
MATERIALS AND METHODS
Animals
Mature western-range ewes of primarily Rambouillet breeding were observed daily for estrous behavior in the presence of vasectomized rams. After experiencing at least two estrous cycles of normal duration (16–18 days), ewes were assigned randomly on Day 0 (estrus/mating) to cyclic or pregnant status. Ewes assigned to pregnant status were mated to intact rams three times at 12-h intervals beginning at estrus. Fifty-two ewes were hysterectomized (n = 4 ewes/day) on Day 1, 3, 5, 7, 9, 11, 13, or 15 of the estrous cycle or Day 11, 13, 15, 17, or 19 of gestation. At hysterectomy uteri were flushed with 0.9% NaCl, and pregnancy was verified by recovery of an apparently normal conceptus in uterine flushes. Several sections (1–1.5 cm) of uterine wall from the middle of each uterine horn were snap frozen in Tissue-Tek OCT compound (Miles, Oneonta, NY). The remaining endometrium was dissected from myometrium, frozen in liquid nitrogen, and stored at -80°C. All experimental and surgical procedures involving animals were approved by the Agricultural Animal Care and Use Committee of Texas A&M University (Animal Use Protocol AG-239AG).
Immunocytochemical Analyses
Antibodies used for immunocytochemistry included rabbit anti-Muc-1 (generously provided by Dr. Dan Carson, University of Delaware); rabbit anti-v (AB1930), 4 (AB1924), 5 (AB1928), ß1 (AB1952), ß3 (AB1932), and ß5 (AB1926) from Chemicon (Temecula, CA); mouse anti-talin clone 8d4 (7-3287), rabbit anti--actinin (A-2543), normal rabbit IgG (15006), and normal mouse IgG (15381) from Sigma (St. Louis, MO); and, fluorescein-conjugated goat anti-rabbit IgG (65-611) and fluorescein-conjugated goat anti-mouse IgG (65-6411) from Zymed (San Francisco, CA).
Proteins were localized in frozen uterine tissue sections (8–10 µm) and in cell lines grown on LabTek four-well chamber slides (Nunc, Naperville, IL) by immunofluorescence staining as previously described [26, 42]. Frozen sections or monolayer cell cultures were fixed in -20°C methanol, permeabilized with 0.3% Tween 20 in 0.02 M PBS, blocked in 5% normal goat serum, and incubated overnight at 4°C with 2 µg/ml primary antibody. Immunoreactive protein was then detected using a fluorescein-conjugated secondary antibody for 1 h at room temperature. Slides were overlaid with a coverglass and Prolong antifade mounting reagent (Molecular Probes, Eugene, OR). For each antibody, representative fluorescence images of cross sections for each day were recorded using a Zeiss Axioplan microscope (Carl Zeiss, Thornwood, NY) equipped with a Hamamatsu chilled 3CCD color camera (Hamamatsu, Japan) using Photoshop 5.0 (Adobe Systems, Seattle, WA) image capture software.
Reverse Transcription-Polymerase Chain Reaction
The presence of mRNAs for IFN and the integrin receptors v, 4, 5, ß1, ß3, and ß5 in cell lines (Tr and LE) and endometrium (Day 15 of the estrous cycle and pregnancy) was examined by reverse transcription-polymerase chain reaction (RT-PCR). Total cellular RNA was extracted using Trizol reagent (Gibco BRL, Grand Island, NY). Total RNA (5 µg) was reverse transcribed to obtain cDNAs using Superscript II reverse transcriptase (Gibco BRL), acid-ethanol precipitated, resuspended in 20 µl water, and stored at -20°C. The cDNA templates were then diluted (1:10) with sterile water and amplified by PCR using AmpliTaq DNA polymerase (Perkin Elmer, Foster City, CA) and the specific primers listed in Table 1. The PCR products were separated on 2% agarose gels, visualized by ethidium bromide staining, and analyzed using an Alpha Innotech (San Leandro, CA) imaging system. The identity of each amplified PCR product was verified by sequence analysis after cloning into the pCRII vector (In Vitrogen, San Diego, CA).
|
Functional Analysis of OPN-Induced Integrin Activation
A ligand-coated bead assay, described by Miyamoto et al. [43] and utilized to identify the transmembrane aggregation of cytoskeletal and signaling molecules induced by integrin activation, was employed to detect functional activation of integrins by OPN in immortalized LE and primary cultures of Tr from ewes. The differentiated properties of the immortalized ovine LE cells that include estrogen and progesterone receptors, STAT proteins, and several IFN-inducible genes expressed by ovine LE in vivo have been described [42]. Ovine LE cells were maintained in Dulbecco modified Eagle medium with F-12 salts (DMEM-F12; Sigma) supplemented with 10% fetal bovine serum and antibiotics. Primary ovine Tr cells were isolated using mechanical dispersion from conceptuses harvested on Day 15 of gestation [44–46] and were maintained in DMEM-F12 supplemented with 5% fetal bovine serum, antibiotics and 0.1 U/ml bovine insulin (Sigma). Both cell types were seeded onto two-well Lab-Tek coverglass chambered slides (Nunc) and cultured for 48 h prior to addition of ligand-coated beads.
Polystyrene beads (6.0 µm; Polysciences Inc., Warrington, PA) were washed in sterile high phosphate PBS (hPBS; 0.1 M NaCl, 2.7 mM KCl, 5 mM Na2PO4, 0.85 mM KH2PO4, 50 mM NaH2PO4), pH 7.4, and centrifuged at 7000 x g for 5 min five times. Beads were coated with poly-L-lysine, wild-type OPN or with mutated OPN in which the RGD sequence was replaced with RGE or RAD (100 µg/ml in hPBS; [47]) and incubated at room temperature with constant agitation for 24 h. The beads were washed five times in hPBS as described above and then incubated overnight with sterile BSA (1 mg/ml in hPBS) at room temperature with constant agitation and then stored at 4°C until used [48].
Cells were washed once in DMEM-F12 + 1 mg/ml BSA (810013, crystalline bovine albumin; ICN Pharmaceuticals) prior to initiation of the assay. Beads were diluted in DMEM-F12 + 1 mg/ml BSA, to a final concentration of 2 x 106 beads/2-ml/chamber, and incubated with cells at 37°C for 1 h in a humidified, 5% CO2/air incubator. Cells were then fixed with 4% paraformaldehyde in PBS (10.14 mM Na2PO4, 1.76 mM KH2PO4, 136.9 mM NaCl, 2.68 mM KCl) for 10 min at room temperature, rinsed twice with PBS, and once with PBS containing 2% BSA. Cells were incubated at room temperature for 30 min in 10% normal serum followed by overnight incubation at 4°C with anti-talin or anti--actinin primary antibody. Following removal of primary antibody, cells were rinsed once in PBS/BSA and three times in PBS for 5 min each. Secondary antibody conjugated to fluorescein isothiocyanate was diluted 1:200 and incubated with cells at room temperature for 1 h in the dark. Cells were rinsed twice with PBS and twice with Hepes buffer (10 mM Hepes, 150 mM NaCl, 0.08% NaN3). After removing the last rinse, 400 µl mounting medium (0.2 M PBS, 2 mg/ml p-phenylenediamine, 90% glycerol) was added, and slides were stored at 4°C in the dark prior to viewing.
Fluorescence imaging of ligand-coated beads was performed with a digital fluorescence imaging system consisting of a charge-coupled device camera and image-capturing software (CELLscan; Scanalytics Inc., Bedford, MA) integrated with a Zeiss Axiovert inverted fluorescence microscope. Sequential optical slices from the basal-to-apical surface of the cells were collected with a high numerical aperture objective lens. Image collection was initiated approximately 1 µm below the basal surface of the cell, and optical slices were collected at 0.5-µm steps up through the apical cell surface and attached beads.
The percentage of ligand-coated beads with apical surface-induced focal adhesions was determined by combined phase-contrast/fluorescence imaging. Phase-contrast imaging was used to locate cells and count adherent beads. Fluorescence microscopy was then used to identify transmembrane accumulation of immunoreactive talin below the ligand-coated beads. For each treatment group, the percentage of ligand-coated beads in contact with cells that exhibited apical focal adhesions was obtained by multiplying the ratio of focal adhesion-positive beads to the total number of beads in contact with the cell by 100. Three to five separate experiments on different days were performed for each of the ligands tested. Data collected from each experiment were subjected to arcsine transformation. General linear models analysis was conducted to test for significance followed by Tukey multiple comparison test to identify significance between groups. A P value of <0.05 was considered significant. Data are presented as mean percentage of beads ± SD inducing apical focal adhesions.
RESULTS
Localization of Muc-1 in Ovine Endometrium
Immunoreactive Muc-1 was restricted to the apical surface of uterine LE and GE (Fig. 1). Expression of Muc-1 on uterine GE remained elevated and constant throughout the estrous cycle and early pregnancy. In contrast, intense apical staining of Muc-1 on uterine LE was observed from Days 1 to 7 followed by a continuous reduction of immunostaining through Day 13 of the estrous cycle. Staining remained low for the remainder of the cycle and early pregnancy and was barely detectable on LE by Day 17 of pregnancy, when intimate adhesion between Tr and LE begins in the ewe [4]. No immunoreactive Muc-1 was detected on Day 15 conceptus flushed from the uterus or on Day 15–17 conceptus tissues examined in situ (data not shown).
|
Uterine/Conceptus Expression of Integrin Subunits
Primers were designed to amplify mRNAs for the integrin subunits v, 4, 5, ß1, ß3, and ß5 from Day 15 cyclic and Day 15 pregnant total ovine endometrial RNA (Table 1). Reverse transcription-PCR confirmed the presence of mRNAs for v, 4, 5, ß1, ß3, and ß5 integrins in endometrium from both cyclic and pregnant ewes (Fig. 2A). The identity of each specific PCR product was confirmed by sequence analyses (data not shown).
|
Immunoreactive v, 4, 5, ß1, ß3, and ß5 integrin subunits were detected at the apical surface of uterine LE and GE, and on conceptus Tr (Fig. 3). The 5 integrin subunit was also detected in uterine stroma. Each of these integrins was expressed by endometrium of both cyclic and pregnant ewes, and the apical expression patterns of each integrin subunit did not change during pregnancy. The apical expression of these subunits could potentially contribute to the apical assembly of several OPN receptors including the vß3, vß1, vß5, and 5ß1 heterodimers on endometrial LE and GE, and conceptus Tr in sheep. Attempts to identify other known OPN receptors including 9ß1 and CD44 using commercially available antibodies were unsuccessful.
|
Characterization of LE and Tr Cell Lines
Immortalized uterine LE [42] and primary Tr cells were developed to investigate integrin-mediated signaling at the apical surfaces of these cells. The Tr cells exhibited an epithelial morphology and were evaluated for the expression of IFN mRNA, the most definitive marker of this cell type. A specific IFN PCR product was detected in total RNA from Tr cells but was not present in total RNA from immortalized LE cells (Fig. 2B). Sequence analyses identified the PCR product as ovine IFN (data not shown).
Using RT-PCR (primers described in Table 1), mRNAs for the integrin subunits v, 5, ß1, and ß3 were detected in total RNA samples from immortalized ovine LE and primary conceptus Tr cells (Fig. 2A). Although 4 and ß5 integrin mRNAs were present in Tr cells, they were not detected in the immortalized LE cell line.
The v and ß3 integrin subunits were abundantly expressed at the basal surface of immortalized ovine LE and primary Tr cells (Fig. 4). The similar punctate subunit staining patterns at sites of cell anchorage to the substrate suggest both the presence of vß3 heterodimers aggregated at focal adhesion sites as well as the abundance of this OPN receptor in both cell types. Although immunoreactive ß1, ß5, and 5 integrin subunits were also detected in LE and Tr cell lines, the aggregation of these proteins at focal adhesions was not as well defined as those shown for v and ß3 integrins.
|
Functional Activation of LE and Tr Integrins by OPN
Figure 5 illustrates the cytoskeletal reorganization (outside-in signaling) observed at the cytoplasmic side of the interface between OPN-coated beads and the apical cell surface of immortalized ovine LE using immunocytochemical localization of the cytoskeletal protein talin to detect the response. Talin was selected as the response variable for this assay because of its central role in binding cytoplasmic domains of ß1 and ß3 integrins, cytoskeletal proteins, and focal adhesion kinase [49]. Within 1 h of adding OPN-coated beads to cultured ovine LE, approximately 20% of the beads in contact with cells exhibited intense immunostained talin aggregates. Similar aggregates were detected at the surface of conceptus Tr although about 50% of the cells associated with beads were talin positive. Similar staining patterns were detected in cells stained with antibody directed against -actinin, but the results were not quantified. Table 2 summarizes the talin results from multiple in vitro microbead-cell adhesion assays.
|
|
DISCUSSION
It is generally accepted that the process of implantation that begins with intimate association of the conceptus and uterine endometrium and ends with the formation of a placenta involves an adhesion cascade. Adhesive LE ligands, normally masked by mucins, become exposed during the receptive period, and various adhesion molecules then function sequentially, or in parallel, to stabilize adhesion of Tr to LE [50]. Temporal and/or spatial changes in expression of mucins, ECMs, and their integrin receptors influence the establishment of endometrial receptivity in rodents, primates, cattle, goats, and pigs [10, 11, 13, 20, 26, 37, 40, 51, 52]. The present and previous studies identified key structural and functional elements of the implantation adhesion cascade in sheep that include 1) elevated expression of antiadhesive Muc-1 at the apical aspect of uterine LE that declines prior to the attachment phase of implantation; 2) the integrin subunits v, 4, 5, ß1, ß3, and ß5 that could form known OPN receptors including vß3, vß1, vß5, and 4ß1, and are constitutively and highly expressed on the apical surface of uterine LE and on conceptus Tr; 3) OPN is secreted from the endometrial GE into the uterine lumen immediately before and during conceptus attachment [26, 41]; 4) OPN protein is present at the apical surfaces of both uterine LE and conceptus Tr [26]; and 5) OPN binds integrins expressed on LE and Tr cells in vitro to initiate integrin activation and outside-in signaling.
The lumen of the uterus is covered by mucins that are heavily glycosylated and project above the apical surface of LE cells. Human blastocysts also express Muc-1 [53]. Membrane mucins such as Muc-1 and Muc-4 serve to lubricate and protect the uterus against microbial infection; however, these proteins also sterically inhibit cell-cell and cell-ECM adhesion [7, 8], and presumably impair Tr access to the uterine LE [9]. Overall Muc-1 expression increases during the receptive phase in rabbits and humans but is locally reduced at sites of conceptus attachment [12, 13]. It is hypothesized that Muc-1 in human endometrium undergoes a reduction of repulsive charge density through the loss of keratin sulfate to allow conceptus apposition. Recent in vitro evidence suggests that paracrine signals from the blastocyst lead to down-regulation of Muc-1 from both LE and trophectoderm at sites of implantation [13, 53]. In contrast, Muc-1 expression at the apical LE surface decreases prior to implantation in rodents and pigs [10, 11, 52].
The present studies clearly show that in sheep, similar to rodents and pigs, the ovine adhesion cascade initiates through down-regulation of Muc-1. Although speculative, it is reasonable to infer that loss of Muc-1 on ovine LE is influenced by progesterone. Both implantation and Muc-1 down-regulation are blocked by administration of RU-486 in rodents [10, 52], and injection of ovariectomized gilts with progesterone decreases Muc-1 expression [11]. Further, Muc-1 decreases dramatically between Days 7 and 17 of pregnancy in sheep (see Fig. 1). This pattern of Muc-1 expression is temporally similar to that of progesterone receptor gene expression in pregnant ovine endometrium. Progesterone receptors are expressed in LE on Days 9 and 11 of pregnancy but are progressively lost between Days 13 and 19 [54]. It is noteworthy that expression of Muc-1 on LE declines to barely detectable levels by Day 17 of pregnancy because adhesion of conceptus Tr to a still-intact uterine LE occurs between Days 16 and 18 in sheep [4]. Loss of Muc-1 prior to ovine implantation may remove a barrier on LE that sterically hinders the ability of ECMs to interact with apically expressed integrins.
Cell surface integrin receptors connect the ECM to the intracellular cytoskeleton and transmit biochemical signals across the plasma membrane. Integrins are the primary mediators of matrix effects including cell-cell and cell-ECM adhesion, influence on cell shape, and regulation of gene expression [19, 55]. Hormone-dependent temporal and spatial distribution of integrins in the human uterus and developmental regulation on the invading mouse blastocyst provide strong circumstantial evidence for integrin involvement in the events of implantation [22, 56]. In sheep, v, 4, 5, ß1, ß3, and ß5 integrin subunits are expressed by endometrium of both cyclic and pregnant ewes and by conceptus Tr. All subunits were identified by RT-PCR in Day 15 cyclic and pregnant endometrium and in a conceptus Tr cell line (see Fig. 2). Expression of v, 5, ß1, and ß3 integrin subunits in LE was confirmed by detection of mRNAs in an immortal LE cell line (Fig. 2). Immunostaining showed that v, 4, 5, ß1, ß3, and ß5 are apically expressed on LE and GE and on conceptus Tr in vivo (Fig. 3). Apical expression of these integrins was constitutive and, in general, did not increase during the peri-implantation period. Although 4 and ß5 subunit proteins were present on endometrial epithelia, mRNA was minimal in endometrium and Tr cells and was not detected in LE cells. The primers used to amplify ovine 4 and ß5 were designed using the cDNA sequence for human 4 and ß5. Perhaps detection of these integrin subunits was limited due to an RT-PCR reaction that was less efficient than those for the other integrins.
The ewe does not appear to limit receptivity to implantation by modifying temporal and spatial patterns of integrin expression as integrins are expressed apically and constitutively. The constitutive apical expression of integrins on uterine epithelia in sheep contrasts with studies from other species which suggest that alterations in integrin expression may frame the putative window of implantation. In women, vß3 and 4ß1 increase on LE prior to embryo attachment [20, 21]. While vß3 and vß5 are found on the apical surface of LE [21, 57], 6ß4, 2ß1, and 3ß1 are distributed primarily along the lateral and basal cell surfaces [58]. In pigs, subunits v, 3, 4, 5, ß1, and ß3 are present on the apical LE surface, with v and ß3 integrins showing intense and constitutive expression. However 4, 5, and ß1 are spatially and temporally regulated throughout the estrous cycle and early pregnancy, reaching maximal levels during early implantation [11]. Baboons express vß3 integrin in both GE and decidualizing stromal cells of pregnant endometrium [37], whereas in mice, endometrial vß3 integrin is limited to stromal cells during implantation [25]. Integrin expression in the LE of sheep also appears to contrast with other ruminants. In cows, vß3, 3, 6, and ß1 are localized to the basolateral surfaces of LE, shallow GE, and within the sublumenal stroma. The pattern of expression for vß3 changes over the estrous cycle [28]. Similarly, ß1 integrin has been detected along the basolateral membranes of LE and GE in goats but is decreased in LE at sites of conceptus adhesion [27]. These variations of integrin expression among species may reflect differences in types of apposition, attachment, and invasion during implantation. However, the presence of integrin receptors on endometrial and conceptus surfaces suggests their involvement as mediators of conceptus attachment to uterine LE common to all mammalian species.
Integrin-mediated attachment may involve bifunctional bridging ligands that interact with receptors expressed on these apposing surfaces to adhere fetal and maternal membranes. OPN is a likely candidate integrin bridging ligand in sheep because the 45-kDa fragment is abundant within the uterine lumen during the peri-implantation period. Moreover, OPN protein is present at apical surfaces of LE and Tr cells that do not express OPN mRNA, providing circumstantial evidence that OPN accumulates at these surfaces through interaction with integrin receptors [26, 40]. Ovine uterine OPN is an acidic 70-kDa matrix glycoprotein that contains an RGD integrin receptor binding sequence [59]. Upon freezing and thawing or treatment with proteases, the 70-kDa protein gives rise to 24- and 45-kDa fragments [60]. The 45-kDa fragment has greater binding affinity for integrins than the native 70-kDa form, possibly because protease cleavage makes the adjacent RGD sequence more accessible for interaction with specific integrins [61].
Binding of integrins to ECM proteins promotes the aggregation of integrins and triggers a hierarchical response leading to transmembrane accumulation of cytoskeletal proteins and over 20 signal transduction molecules to the ß-integrin subunit cytoplasmic domain [43]. The result is assembly of well-developed aggregates composed of ECM proteins, integrins, and cytoskeletal proteins that are known as focal adhesions [19, 43, 62]. Attachment of the c-Src substrates, tensin, and focal adhesion kinase can result from integrin aggregation alone, but aggregation of cytoskeletal proteins including talin, -actinin, vinculin, and F-actin requires ligand occupancy and integrin aggregation [43]. Therefore, immunodetection of aggregated integrins, talin, or -actinin at focal adhesions can provide a sensitive functional index of integrin activation and outside-in signaling. The in vitro studies reported here exploit the ability to induce focal adhesions by integrin-ECM interactions to show, for the first time, functional integrin activation and cytoskeletal reorganization in uterine LE and conceptus Tr cells in response to OPN binding. Accumulation of talin and -actinin was detected at the interface between OPN-coated polystyrene beads and the apical membranes of LE and Tr cells. The focal adhesions are the result of RGD-integrin interactions because mutation of the OPN RGD sequence eliminated cytoskeletal aggregation (Table 2), although the identity of activated integrins remains unknown. Interestingly, v and ß3 integrin subunits that form the vß3 receptor, which is capable of binding multiple matrix proteins including OPN, vitronectin, and fibronectin [18], aggregate at sites of cell anchorage to the substrate in both LE and Tr cells, suggesting the presence of this versatile receptor at focal adhesion sites (see Fig. 4). OPN-coated beads also induce integrin activation in Tr cells. The number of beads associated with Tr cells was at least double the number formed at the apical surface of immortalized LE, and this was consistent with the greater abundance of v and ß3 containing aggregates at the basal surface of cultured Tr cells. This is not unexpected in light of the highly adhesive nature of the conceptus. Therefore, it is reasonable to predict that in the pregnant ovine uterus, OPN binding to integrin heterodimers induces similar focal adhesion sites that promote and stabilize attachment of Tr to LE for implantation.
The present study suggests that a decline in Muc-1 on uterine LE exposes apically oriented v, 4, 5, ß1, ß3, and ß5 integrins to interaction with OPN during the peri-implantation period. The coincident presence of these integrins that could form known OPN receptors including vß3, vß1, vß5, and 4ß1, on uterine LE and conceptus Tr, along with evidence that OPN can functionally interact with integrins expressed on LE and Tr cells, provide a physiological framework for OPN to act as a mediator of implantation in sheep that bridges integrin receptors expressed by Tr and LE to induce adhesion essential for initial conceptus attachment.
ACKNOWLEDGMENTS
The authors thank Dr. Dan Carson at the University of Delaware for rabbit anti-Muc-1 IgG; Dr. Magnus Höök at the Institute of Biosciences and Technology, Texas A&M University System Health Science Center, for the recombinant rat OPN containing the RGD sequence and mutants in which the RGD sequence was replaced with RAD or RGE; and Dr. Shawn Ramsey and Mr. Todd Taylor of the Texas A&M University Sheep and Goat Center for care and management of ewes.
FOOTNOTES
1 Research supported by USDA-NRICGP grants 95-37203-2185 and 98-35203-6223 to F.W.B. and R.C.B. and by NIH 1-F32-HD08501-01A1 to G.A.J. Use of microscopy and imaging facilities in the College of Veterinary Medicine Image Analysis Laboratory, which is supported, in part, by NIH grant P30 ES09106, is acknowledged. 
2 Correspondence: Robert C. Burghardt, Department of Veterinary Anatomy
& Public Health, College of Veterinary Medicine, Texas A&M University,
College Station, TX 77843. FAX: 979 847 8981; rburghardt{at}cvm.tamu.edu
3 Current address: Department of Animal and Veterinary Science, Center
for Reproductive Biology, Agricultural Sciences Building, University of
Idaho, Moscow, ID 83844-2330.
Accepted: April 20, 2001.
Received: March 7, 2001.
REFERENCES
vß3 integrin adversely affects implantation in the mouse. Biol Reprod 2000; 62:1285-1290vß3 integrin expression in the uterus and conceptus during the peri-implantation period. Biol Reprod 1999; 61:892-899vß3 in smooth muscle cell migration to osteopontin in vitro. J Clin Invest 1995; 95:713-724vß1 and vß5. J Biol Chem 1995; 270:26232-262384ß1 integrin. J Cell Science 1998; 111:1165-1174[Abstract]v in human and mouse endometrium: expression in stromal and glandular cells. Mol Human Reprod 1996; 2:527-534This article has been cited by other articles:
|
|
 |
|
  M. C. Satterfield, K. Hayashi, G. Song, S. G. Black, F. W. Bazer, and T. E. Spencer Progesterone Regulates FGF10, MET, IGFBP1, and IGFBP3 in the Endometrium of the Ovine Uterus Biol Reprod, December 1, 2008; 79(6): 1226 - 1236. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Dunlap, D. W. Erikson, R. C. Burghardt, F. J. White, K. M. Reed, J. L. Farmer, T. E. Spencer, R. R. Magness, F. W. Bazer, K. J. Bayless, et al. Progesterone and Placentation Increase Secreted Phosphoprotein One (SPP1 or Osteopontin) in Uterine Glands and Stroma for Histotrophic and Hematotrophic Support of Ovine Pregnancy Biol Reprod, November 1, 2008; 79(5): 983 - 990. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-C. Su, E. A. Mendoza, H.-i. Kwak, and K. J. Bayless Molecular profile of endothelial invasion of three-dimensional collagen matrices: insights into angiogenic sprout induction in wound healing Am J Physiol Cell Physiol, November 1, 2008; 295(5): C1215 - C1229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. S. Bridger, S. Haupt, R. Leiser, G. A. Johnson, R. C. Burghardt, H.-R. Tinneberg, and C. Pfarrer Integrin Activation in Bovine Placentomes and in Caruncular Epithelial Cells Isolated from Pregnant Cows Biol Reprod, August 1, 2008; 79(2): 274 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Farmer, R. C. Burghardt, F. D. Jousan, P. J. Hansen, F. W. Bazer, and T. E. Spencer Galectin 15 (LGALS15) functions in trophectoderm migration and attachment FASEB J, February 1, 2008; 22(2): 548 - 560. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. K Lewis, J. L Farmer, R. C Burghardt, G. R Newton, G. A Johnson, D. L Adelson, F. W Bazer, and T. E Spencer Galectin 15 (LGALS15): A Gene Uniquely Expressed in the Uteri of Sheep and Goats that Functions in Trophoblast Attachment Biol Reprod, December 1, 2007; 77(6): 1027 - 1036. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A Desmarais, F. L Lopes, H. Zhang, S. K Das, and B. D Murphy The Peroxisome Proliferator-Activated Receptor Gamma Regulates Trophoblast Cell Differentiation in Mink (Mustela vison) Biol Reprod, November 1, 2007; 77(5): 829 - 839. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Satterfield, K. A. Dunlap, K. Hayashi, R. C. Burghardt, T. E. Spencer, and F. W. Bazer Tight and Adherens Junctions in the Ovine Uterus: Differential Regulation by Pregnancy and Progesterone Endocrinology, August 1, 2007; 148(8): 3922 - 3931. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Hayashi, R. C. Burghardt, F. W. Bazer, and T. E. Spencer WNTs in the Ovine Uterus: Potential Regulation of Periimplantation Ovine Conceptus Development Endocrinology, July 1, 2007; 148(7): 3496 - 3506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. J White, R. C Burghardt, J. Hu, M. M Joyce, T. E Spencer, and G. A Johnson Secreted phosphoprotein 1 (osteopontin) is expressed by stromal macrophages in cyclic and pregnant endometrium of mice, but is induced by estrogen in luminal epithelium during conceptus attachment for implantation. Reproduction, December 1, 2006; 132(6): 919 - 929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Hao, N. Mathialagan, E. Walters, J. Mao, L. Lai, D. Becker, W. Li, J. Critser, and R. S. Prather Osteopontin Reduces Polyspermy During In Vitro Fertilization of Porcine Oocytes Biol Reprod, November 1, 2006; 75(5): 726 - 733. [Abstract] [Full Text] [PDF]
|
|
|
|
