| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center for Animal Biotechnology and Genomics,3 Department of Veterinary Integrative Bioscience,4 Department of Large Animal Clinical Sciences,5 College of Veterinary Medicine and Biomedical Sciences, and Department of Animal Science,6 College of Agriculture and Life Sciences, Texas A&M University, College Station, Texas 77843-4458
ABSTRACT
Pig conceptuses secrete estrogen for pregnancy recognition, and they secrete interferons (IFNs) gamma and delta during the peri-implantation period. The uterine effects of pig IFNs are not known, although ruminant conceptuses secrete IFN tau for pregnancy recognition, and this increases the expression of IFN-stimulated genes (ISGs) in the endometrium. In sheep, the transcriptional repressor interferon-regulatory factor 2 (IRF2) is expressed in the endometrial luminal epithelium (LE) and appears to restrict IFN tau induction of most ISGs, including IRF1, to the stroma and glands. Interestingly, MX1, which is an ISG in sheep, is also expressed in the endometrial stroma of pregnant pigs. The objective of the present study was to determine if estrogen and/or conceptus secretory proteins (CSPs) that contain IFNs regulate IRF1 and IRF2 in pig endometria. The endometrial levels of IRF1 and IRF2 were low throughout the estrus cycle. After Day 12 of pregnancy, the levels of the classical ISGs, which include IRF1, STAT2, MIC, and B2M, increased in the overall endometrium, with expression of IRF1 and STAT2 being specifically localized to the stroma. IRF2 increased in the LE after Day 12. To determine the effects of estrogen, pigs were treated with 17 beta-estradiol benzoate (E2). To determine the CSP effects, pigs were treated with E2 and implanted with mini-osmotic pumps that delivered control serum proteins (CX) to one ligated uterine horn and CSP to the other horn. Estrogen increased the level of IRF2 in the endometrial LE. The administration of E2 and infusion of CSP increased the level of IRF1 in the stroma. These results suggest that conceptus estrogen induces IRF2 in the LE and limits the induction of IRF1 by conceptus IFNs to the stroma. The cell-specific expression of IRF1 and IRF2 in the pig endometrium highlights the complex and overlapping events that are associated with gene expression during the peri-implantation period, when pregnancy recognition signaling and uterine remodeling for implantation and placentation are necessary for successful pregnancy.
conceptus,, estradiol,, estrogen,, gene regulation,, interferons,, pregnancy,, trophoblast,, uterus
The successful establishment and maintenance of pregnancy requires orchestrated communication between the conceptus (embryo/fetus and associated extraembryonic membranes) and the uterus, which includes: (i) secretions from the conceptus to signal pregnancy recognition [1]; (ii) secretions from the uterine luminal epithelium (LE) and glandular epithelium (GE), i.e., the histotroph, to support the attachment, development, and growth of the conceptus [2–4]; (iii) remodeling at the endometrial LE surface to allow intimate association between the conceptus trophectoderm and endometrium for implantation [5–7]; and (iv) remodeling of the endometrial stroma to generate a cytokine-rich environment that directly promotes angiogenesis, to provide hematotrophic support for the developing conceptus [8, 9].
In pigs, pregnancy recognition is the result of conceptus secretion of estrogens on Days 11 and 12 of pregnancy, which redirects PGF secretion from the uterine vasculature to the uterine lumen, where it is sequestered away from the CL [10–13]. In contrast to pigs, sheep conceptuses secrete interferon tau (IFNT) to signal maternal recognition of pregnancy [1, 14, 15]. In addition to its antiluteolytic actions on the endometrium, IFNT increases the expression of a number of IFN-stimulated genes (ISGs) in the stroma of the ruminant uterus [16–26], including MX1 [16], interferon regulatory factor 1 (IRF1) [17], signal transducer and activator of transcription 2 (STAT2) [16], major histocompatibility complex (MHC) class I polypeptide-related alpha chain (MIC), and beta-2-microglobulin (B2M) [18].
Peri-implantation pig conceptuses also secrete IFNs. Cultured conceptuses from Day 11 of pregnancy have been shown to secrete proteins that cross-react with antiserum against IFN alpha [27], although peak antiviral activity was not measured in the uterine flushings or conceptus culture media until Days 14 and 15 of pregnancy [28]. Both type I IFN and type II IFN are produced. The major species, which comprises 75% of the antiviral activity of pig conceptus secretory proteins (CSPs), is the type II IFN gamma (IFNG) and the minor species (25%) is the novel type I IFN delta (IFND) [29, 30]. These IFNs do not appear to have antiluteolytic activities during pregnancy, as intrauterine infusion of CSPs on Days 12 to 15 of the estrus cycle had no effect on the interestrus interval or temporal changes in plasma progesterone concentrations [31, 32]. However, paracrine effects for IFNs are suggested by the localization of IFN receptors on endometrial epithelial cells [30], increased secretion of prostaglandin E2 [31], and MX1 expression in the stratum compactum stroma of pigs on Day 18 of pregnancy [33]. The effects of these IFNs on pig endometrium have not been determined. Although it has been noted that high peri-implantation levels of IFNG coincide with the presence of uterine transforming growth factor ß, interleukin 6, and MHC class II antigens in pigs [34, 35], increased uterine expression of classical ISGs has not been detected [36]. Indeed, treatment of Madin-Darby bovine kidney cells and bovine endometrial explant cultures with pig CSPs increased ISG expression, whereas a similar treatment had no effect on ISG expression in pig endometrial explants [36].
Our working hypothesis is that pig conceptus IFNs increase uterine endometrial expression of ISGs during pregnancy, and that these genes have biological roles in uterine receptivity and conceptus implantation and development. IRF1 is a key intermediate in the induction cascades of many classical ISGs through its abilities to bind and transactivate IFN-stimulated response elements (ISRE) at their promoters [37–39]. Both type I and type II IFNs induce IRF1 [37], which plays a role in placental development in the murine reproductive tract [38]. In sheep, IRF1 expression increases in the stroma and GE, but not in the LE, during early pregnancy, presumably due to the expression of IRF2, which is a potent transcriptional repressor of ISGs that is constitutively expressed in the LE and increases during early pregnancy [16]. Therefore, the objectives of the present studies were to determine whether IRF1 and IRF2 are expressed in the pig endometrium during pregnancy, and if so, whether the expression of these genes is regulated by conceptus estrogen and/or CSPs that contain IFNG and IFND.
All the experimental and surgical procedures complied with the Guide for Care and Use of Laboratory Animals and were approved by the Texas A&M University Laboratory Animal Care and Use Committee. Pigs were observed daily for estrus (Day 0) and exhibited at least two estrus cycles of normal duration (18–21 days) before use in these studies.
Study 1. To evaluate the effect of pregnancy status on endometrial gene expression, sexually mature pigs were assigned randomly at estrus to either cyclic or pregnant status. The pigs in the pregnant group were bred upon detection of estrus and 12 h and 24 h thereafter. Pigs were ovariohysterectomized on Day 5, 9, 12, or 15 of the estrus cycle or on Day 9, 10, 12, 13, 14, 15, 20, 25, 30, 35, 40, 60, or 85 of pregnancy (n = 3 pigs/day/status). Pregnancy was confirmed by the presence of normal conceptuses in the uterine flushings (Days 9–15) or at hysterectomy (Days 20–85).
Study 2. To evaluate the effect of estrogen-induced pseudopregnancy on uterine gene expression, pigs were assigned randomly at estrus to receive daily i.m. injections of either 5 ml corn oil (CO) vehicle or 5 mg 17ß-estradiol benzoate (E2; Sigma Chemical Company, St. Louis, MO) in 5 ml CO on Days 11, 12, 13, and 14 postestrus (n = 5 pigs/treatment). All pigs were ovariohysterectomized on Day 15 postestrus.
Study 3. To evaluate the effect of pig CSPs on uterine gene expression, pigs (n = 3) were injected i.m. with 5 mg E2 5 ml of CO on Days 11, 12, 13, 14, and 15 postestrus, to induce pseudopregnancy. On Day 12 postestrus (coincident with the onset and prior to the peak of secretion of IFNs by the pig conceptuses [27, 28, 29]), each pig was surgically implanted with two indwelling ALZET osmotic pumps (Durect Corp., Cupertino, CA) with constant delivery rates of 10 µl/h. Briefly, each uterine horn was isolated via midline celiotomy, clamped, and severed from the uterine body at approximately 12.7 cm from the utero-tubal junction, while preserving the mesometrium and vascular supply to the uterine horn. The transected ends of each uterine horn and uterine body were closed using an inverting suture pattern of absorbable suture, and the serosa of the antimesometrial borders of each uterine horn and the uterine body were sutured together to prevent twisting of the horn. For each pump, a catheter was attached and inserted approximately 2 cm into the lumen of one uterine horn. Prior to surgery, the pumps were filled and equilibrated according to the manufacturer's instructions. For each pig, one uterine horn was infused by a pump that was filled with 35 mg of porcine serum albumin (Sigma), while the other uterine horn was infused by a pump that was filled with 35 mg of porcine CSPs. Thus, a 12.7-cm isolated section of the uterine horn, which retained full vascular supply, was completely exposed to the infusate; the uterine tissue samples were taken from these sections. Pilot studies were conducted with infusion of India ink to confirm coverage of the uterus by the infusate. All gilts were ovariohysterectomized on Day 16 postestrus (coincident with maximal antiviral activity in the pig uterine flushings [28]).
At hysterectomy, several sections (
0.5-cm thickness) from the middle of each uterine horn or from the isolated pouch of an infused uterus were fixed in fresh 4% paraformaldehyde in PBS (pH 7.2) and embedded in Paraplast-Plus (Oxford Laboratory, St. Louis, MO). Several sections from each uterine horn were also embedded in Tissue-Tek Optimal Cutting Temperature (OCT) Compound (Miles, Oneonta, NY), snap frozen in liquid nitrogen, and stored at –80°C before sectioning. The remaining endometrium was physically dissected from the myometrium, frozen in liquid nitrogen, and stored at –80°C for RNA extraction.
Preparation of porcine CSPs. Using procedures described previously [31, 40], the conceptuses from Day 15–17 pregnant pigs (coincident with maximal production of IFNs by the conceptuses [27, 28, 29]) were recovered by flushing each uterine horn with 20 ml of minimal essential medium (MEM). The conceptuses were then cultured in MEM for 30 h at 37°C with rocking in a 50% O2, 45% N2, 5% CO2 atmosphere. The culture medium was collected after centrifugation and protease inhibitors (Complete EDTA-free Protease Inhibitor Cocktail; Roche Diagnostics, Indianapolis, IN) were added. The culture supernatant was dialyzed (MWCO 3500; Spectrum Laboratories, Inc., Rancho Dominguez, CA) four times using 4 L of 10 mM Tris (pH 8.2) each time, and concentrated (MWCO 5000; Millipore Corp., Bedford, MA). The sample was then dialyzed (MWCO 1000; Spectrum Laboratories) against Dulbecco PBS (Sigma), protease inhibitors were added, and the sample was filter sterilized, assayed for protein concentration, and stored at 4°C until use.
RNA isolation. Total cellular RNA was isolated from endometrial tissue samples using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's recommendations.
Northern blot analysis.
Total endometrial RNA (8 µg) was loaded onto a 1.2% agarose gel, electrophoresed, and transferred to a 0.2-µm nylon membrane, as described previously [41]. The blot was hybridized with radiolabeled antisense cRNA probes that were generated from linearized ovine IRF1 [16], ovine IRF2 [16], human STAT2 [42], ovine MIC [17], and ovine B2M [17] plasmid templates. Radiolabeled riboprobes were generated by in vitro transcription with [
-32P]uridine 5-triphosphate (Perkin-Elmer Life Sciences, Inc., Boston, MA) and the MAXIscript kit (Ambion, Austin, TX). After washing, nonspecific hybridization was eliminated by RNase A digestion. Hybridization signals were detected by exposing the blot to a PhosphoImager screen and visualized using a Typhoon 8600 variable mode imager (Molecular Dynamics, Piscataway, NJ).
Slot blot analysis. The steady-state mRNA levels were assessed in endometrial total RNA samples (20 µg) by slot blot hybridization with radiolabeled antisense ovine IRF1 [16], ovine IRF2 [16], human STAT2 [42], ovine MIC [17], and ovine B2M [17] cRNA probes using methods described previously [41]. To correct for variability in total RNA loading, a duplicate RNA slot membrane was hybridized with radiolabeled antisense 18S rRNA (pT718S; Ambion) cRNA probe. The radiolabeled riboprobes were generated as described above. The membranes were washed, digested, and hybridization signals were detected as described above.
In situ hybridization analysis.
IRF1, IRF2, and STAT2 mRNAs were localized in paraffin-embedded pig uterine tissues by in situ hybridization, as previously described [43]. Briefly, deparaffinized, rehydrated, and deproteinated uterine cross-sections (5-µm thickness) were hybridized with radiolabeled antisense or sense ovine IRF1 [16], ovine IRF2 [16], and human STAT2 [42] cRNA probes, which were synthesized by in vitro transcription with [-35S]uridine 5-triphosphate (Perkin-Elmer Life Sciences). After hybridization, washing, and RNase A digestion, autoradiography was performed using NTB liquid photographic emulsion (Eastman Kodak, Rochester, NY). Slides were exposed at 4°C, developed in Kodak D-19 developer, counterstained with Harris modified hematoxylin (Fisher Scientific, Fairlawn, NJ), dehydrated, and protected with coverslips.
Immunoreactive IRF1 protein was localized in frozen porcine uterine cross-sections (8–10-µm thickness) by immunofluorescence staining using methods described previously [44]. Briefly, tissues were fixed in methanol at –20°C, washed in PBS that contained 0.3% (vol/vol) Tween-20, blocked in 10% normal goat serum, incubated overnight at 4°C with 1 µg/ml of rabbit antihuman IRF1 (sc-497; Santa Cruz Biotechnology, Santa Cruz, CA) or rabbit IgG (negative control; Sigma), and detected with fluorescein-conjugated goat anti-rabbit IgG (Chemicon International, Temecula, CA). The slides were overlaid with Prolong antifade mounting reagent (Molecular Probes, Eugene, OR) and a coverslip.
Digital photomicrographs of in situ hybridization and immunofluorescence staining were evaluated using an Axioplan 2 microscope (Carl Zeiss, Thornwood, NY), which was interfaced with an Axioplan HR digital camera and the Axiovision 4.3 software. Photographic plates were assembled using the Adobe Photoshop ver. 6.0 software (Adobe Systems Inc., San Jose, CA).
The data were subjected to least-squares ANOVA using the general linear models procedures of the Statistical Analysis System (SAS, Cary, NC). The slot blot hybridization data were analyzed using 18S rRNA as a covariate to correct for differences in RNA loading. The data from study 1 were analyzed for the effects of day and status and their interaction where appropriate. For all other studies, the effects of treatment were determined by preplanned orthogonal contrasts. All tests of significance were performed using the appropriate error terms according to the expectation of the mean squares for error, and P < 0.05 was considered statistically significant. Data are presented as least-squares means with standard errors (SEMs). The SEM represents the pool of the mean derived using the root mean square error term generated by the SAS software.
Effects of Pregnancy (Study 1)
Steady-state levels of IRF1, STAT2, MIC, B2M, and IRF2 mRNAs in the pig endometrium.
The ovine cDNAs for IRF1, MIC, and B2M revealed mRNAs of 2.1 kb, 1.7 kb, and 1.0 kb, respectively, and the human cDNA for STAT2 revealed mRNAs of 4.5 kb and 4.8 kb in Northern blot analysis of pig total endometrial RNA (data not shown). These mRNAs were similar in size to those detected using the same cDNAs with sheep total mRNA. The steady-state levels of IRF1, MIC, and B2M mRNAs in the porcine endometrium did not change (P > 0.10), whereas the STAT2 mRNA levels increased between Day 5 and Day 9 during the estrus cycle, and decreased thereafter (P < 0.005, cubic effect of day) during the estrus cycle. During pregnancy, the IRF1 mRNA levels were low on Day 9 to Day 12, increased almost 3-fold between Days 12 and 15, declined between Days 15 and 40, and remained low thereafter (P < 0.001, quartic effect of day) (Fig. 1A). The STAT2 mRNA levels increased more than 2-fold between Days 12 and 14 (P < 0.01, linear effect of day) (Fig. 1A). The MIC mRNA levels increased nearly 3-fold between Days 10 and 14 of pregnancy, remained high through Day 20, declined between Days 20 and 40, and remained low thereafter (P < 0.05, cubic effect of day) (Fig. 1A). The B2M mRNA levels gradually decreased between Day 9 and Day 13, increased 2-fold between Days 13 and 14, gradually declined through Day 25, and remained low thereafter (P = 0.1, quadratic effect of day) (Fig. 1A). Therefore, the levels of the mRNAs for the four classical ISGs, i.e., IRF1, STAT2, MIC, and B2M, are increased in the pig endometrium during the peri-implantation period.
|
The ovine cDNA for IRF2 revealed a 2.4-kb mRNA in both the pig and sheep endometrial total RNA. The steady-state levels of IRF2 mRNA did not change (P > 0.10) during the estrus cycle, whereas the IRF2 mRNA levels increased from Day 9 to Day 13 in pregnant pigs, were maximal on Days 13–15, and decreased thereafter (P < 0.001, quadratic effect of day) (Fig. 1B).
In situ hybridization for IRF1, STAT2, and IRF2 mRNAs in the pig endometrium. The levels of IRF1 (Fig. 2) and STAT2 (Fig. 3) mRNAs were low in all endometrial cell types during the estrus cycle. During pregnancy, the IRF1 and STAT2 mRNAs were noticeably upregulated in the stratum compactum stroma of the endometrium between Day 12 and Day 15. The IRF1 and STAT2 mRNAs remained in the stratum compactum stroma through Day 25 of pregnancy, and then decreased to very low levels through Day 85. The IRF1 (Fig. 2) and STAT2 (Fig. 3) mRNAs were not observed in the uterine LE of either the cyclic or pregnant pigs. Therefore, the mRNAs for two classical ISGs, i.e., IRF1 and STAT2, are increased specifically in pig endometrial stroma during the peri-implantation period.
|
|
As illustrated in Figure 4, the level of IRF2 mRNA was low during the estrus cycle. However, IRF2 mRNA appeared in the LE of pregnant pigs on Day 12 and remained in the LE through Day 30. IRF2 mRNA was not observed in the stroma or glands of the uteri isolated from both cyclic and pregnant pigs.
|
Immunoreactive IRF1 protein. Consistent with the in situ hybridization results, the level of IRF1 protein was low in the endometrium on Day 15 of the estrus cycle, but was present in the endometrial stroma on Day 15 of pregnancy (Fig. 5). IRF1 protein was not observed in the LE of pregnant endometrium (Fig. 5).
|
Collectively, these data document two expression patterns during the peri-implantation period of pigs: 1) IRF2 increases in LE cells on Day 12, at which time-point the elongated pig conceptuses secrete estrogen for pregnancy recognition [11]; and 2) the levels of classical ISGs increase in the endometrial stroma between Day 12 and Day 15, which temporally correlates with increased antiviral activity measured in uterine flushes exposed to conceptus secretion of IFNG and IFND [27–29].
Exogenous Estrogen Induces IRF2 But Not IRF1 in Porcine Endometrium (Study 2)
Intramuscular injections of E2 did not alter the steady-state levels of IRF1 mRNA in the pig endometrium compared to CO injection (for CO vs. E2, 146 634 vs. 115 756 ± 11 505 relative units of radioactivity; P > 0.10). Consistent with the slot blot hybridization results, the levels of immunoreactive IRF1 protein were similar in the endometria (Day 15) of cyclic pigs injected with E2 and those injected with CO vehicle (Fig. 6A).
|
In contrast, the levels of IRF2 mRNA were increased (P < 0.05) in the endometria of pigs injected with E2 as compared to those injected with CO vehicle (for CO vs. E2, 245 844 vs 343 684 ± 19 604 relative units of radioactivity). The in situ hybridization analyses revealed that IRF2 mRNA was increased specifically by E2 in the uterine LE (Fig. 6B).
CSPs Regulate IRF1 But Not IRF2 (Study 3)
Steady-state levels of IRF1 mRNA in the pig endometrium. Intrauterine infusion of CSPs into the uterine horn of pigs treated with exogenous estrogen increased (P < 0.1) the steady-state levels of endometrial IRF1 mRNA about 2-fold as compared to the uterine horns infused with control serum proteins. Consistent with the slot blot hybridization results, immunoreactive IRF1 protein abundance was noticeably greater in the stratum compactum stroma of the uterine horn infused with CSPs as compared to the uterine horn infused with control serum proteins (Fig. 7A).
|
In contrast, in situ hybridization for IRF2 revealed that intrauterine infusion of CSP into Day-16 pigs treated with exogenous estrogen did not increase IRF2 mRNA expression over the levels detected after intrauterine infusion of control serum proteins (Fig. 7B).
Our results demonstrate that the levels of IRF1, STAT2, MIC, and B2M increase in endometria in the peri-implantation period, during which elongated pig conceptuses secrete IFND and IFNG [29, 30]. Furthermore, IRF1 and STAT2 are expressed in the endometrial stroma. It seems likely that IFND and/or IFNG pass through altered tight junctions between the uterine LE cells [45], to act in a paracrine manner to induce these genes in the endometrial stroma, since IRF1, STAT2, MIC, and B2M are known to be induced by both type I and type II IFNs [16, 37]. Indeed, the lower magnitude of increased expression of these genes in the endometria of pigs as compared to sheep correlates well with the differences noted between these species in the antiviral activities measured in the uterine flushes [29, 46]. Increases in IRF1 and STAT2 gene transcription in several cell lines require the activation of STAT1, formation of a homodimer (termed 
-activated factor or GAF), translocation to the nucleus, and transactivation through GAF binding to a -activated sequence (GAS) in the IRF1 promoter [37, 47–49]. However, IRF1 can also co-ordinate with and maintain the induction cascade of other classical ISGs through binding and transactivating IFN-stimulated response elements (ISRE) at their promoters [37–39]. It is significant that the stromal distributions of IRF1 and STAT2 in the pig uterus are similar to the increases observed for IRF1 and STAT2 in pregnant sheep endometria exposed to IFNT [16], and this temporal/spatial pattern of expression is shared by several ISGs in sheep [16–19, 21, 22, 24, 26, 50]. Cattle, mice, and primates express ISGs in the endometrial stroma or decidua of pregnancy [20, 25, 51, 52]. This is the first report of temporal changes in the gene expression levels of IRF1, STAT2, MIC, and B2M and in the spatial distributions of IFR1 and STAT2 in the pig uterus, and the first direct linkage of CSPs that contain IFNG and IFND to endometrial ISG expression in pigs.
Although it has been suggested that IRF1 induces and/or maintains the transcription of selected ISGs in the endometrial stroma of sheep [36], the pregnancy-specific roles of uterine ISGs remain conjectural. Hess and coworkers [53] treated decidualized human endometrial stromal cells with conditioned media from human trophoblasts, in studies similar to the in vivo intrauterine infusion experiments of the present study, and found that many ISGs, including IRF1, were upregulated. The upregulation of ISGs from the secretory products of human trophoblasts is likely due to the production of type I IFNs [54]. Mouse trophoblast giant cells have also been shown to produce a type I IFN-like molecule, which induces ISG expression in endometrial stromal cells [55]. Thus, emerging evidence suggests that the induction and increase in ISGs in the endometrium or decidua by conceptus IFNs are phenomena of early pregnancy in many mammals [51, 53, 56], including pigs (present study). Interestingly, a decidual-like transformation has been reported in the pregnant endometrial stroma of sheep, which suggests that the endometrium of noninvasive implanting species undergoes remodeling that is somewhat similar to the uterine decidua of species with invasive implantation [57]. Therefore, it is likely that ISGs facilitate remodeling within the stromal compartment of the uterus for uterine receptivity to conceptus implantation and placentation across disparate mammalian species.
As players in decidual/stromal remodeling, individual ISGs may be involved in protecting the fetal semiallograft from immune rejection, limiting conceptus invasion through the uterine wall, and/or establishing a vascular supply to the conceptus. Since IFNG, a protein secreted by pig conceptuses, is involved in endometrial vascular development in mice [58], it is reasonable to hypothesize that conceptus-derived IFNs upregulate ISGs, such as IRF1, to facilitate the vascular changes that are needed to provide hematotrophic support to the developing conceptus. Whether or not this is true, it is becoming increasingly clear that IFN induction of genes within the uterine stroma of mammals is a universal response to or component of mechanisms for the establishment and maintenance of pregnancy.
IRF2 is a potent repressor and attenuator of ISG expression and inhibits ISRE-containing genes through direct ISRE binding and coactivator repulsion [59, 60]. As such, IRF2 is an important regulator in the gene networks of the IFN system [61, 62]. A previous study by Choi et al. [16] described the expression of IRF2 in the endometrial LE of early pregnant sheep. In addition, the transcriptional activity of a promoter-reporter construct that contained five consensus ISRE-binding sites was strongly repressed by transient transfection of immortalized sheep stromal cells [63] with vectors that overexpressed ovine IRF2 [16]. These data, along with the constitutive presence of IRF2 and lack of IRF1 and many other classical ISGs in the LE, have led to the hypothesis that IRF2 restricts the expression of ISGs in the LE by directly repressing their transcription and rendering IFNT unable to activate the classical JAK-STAT-IRF1 pathway [14, 16]. The present studies are the first to localize IRF2 in the pig endometrium, and the similar temporal and spatial patterns of expression of IRF2 and IRF1 in pigs and sheep supports the idea that IRF2 represses the expression of ISGs in the LE of pigs, and perhaps in mammals in general.
In the present study, the conceptus and injections of estrogen induced IRF2 expression, specifically in the endometrial LE. Estrogen receptor (ESR1) is present in the LE on Day 12 of pregnancy [64], at which time-point the conceptuses secrete estrogens. Furthermore, estrogen is capable of regulating gene transcription through ESR1/Sp1 interactions [15], and the human IRF2 promoter contains four Sp1 sites [62]. Estrogens are the maternal recognition signals that prevent CL regression [11]. In addition, conceptus estrogens modulate uterine gene expression to support the controlled inflammatory-like events that characterize changes in conceptus morphology and uterine remodeling for implantation in pigs [65]. Indeed, secreted phosphoprotein 1 (or osteopontin) is induced by estrogen in the LE [66]. Furthermore, conceptus secretion of estrogens correlates with conceptus secretion of interleukin 1ß, which may in turn modulate the uterine response to this cytokine [67]. The importance of estrogen to early survival of pig conceptuses is underscored by the fact that premature exposure of the pregnant uterus to estrogen on Day 9 and Day 10 results in the degeneration of all pig conceptuses by Day 15 [68]. The present results strongly suggest that conceptus estrogens induce IRF2 in the endometrial LE, thereby indirectly inhibiting conceptus IFNs from inducing IRF1, STAT2, and presumably other ISGs at the sites of conceptus attachment for implantation. The role that ISG repression plays in the establishment of pregnancy remains to be determined. However, in sheep, major MHC class I and ß2-microglobulin are silenced in the LE, presumably by IRF2 [17]. It has been hypothesized that the ablation of these key molecules, which are involved in host defense and immune histocompatibility of transplanted tissues at the maternal-placental interface, ensures acceptance of the conceptus semiallograft [17]. It is reasonable to predict similar mechanisms for the pig.
Interestingly, the placentas and offspring of Irf1–/– mice are smaller than their wild-type counterparts [35], a phenotype similar to that of several mouse strains that lack uterine natural killer (uNK) cells [69]. Uterine NK cells are associated with modification of the decidual spiral arteries that supply the conceptus with hematotrophic support [69]. In Irf1–/– mice, the uNK cells are fewer, smaller, and hypogranular [35]. Both Irf1 and Irf2 are involved in the development and/or function of peripheral NK cells. In Irf1–/– mice, the numbers of NK cells in the spleen and liver are reduced and cytolytic activity is absent [70, 71]. Irf1 also transcriptionally regulates interleukin 15 [72], which is involved in NK cell maturation [73]. In Irf2–/– mice, the number of NK cells and the NK cell cytotoxic activities of splenocytes are reduced [74]. While both Irf1 and Irf2 are important for peripheral NK cells, only Irf1 is involved in uNK cell development [35].
An attractive hypothesis for the pig is that conceptus secretion of IFNG and/or IFND increases IRF1 expression in the uterine stroma, which plays a role in increasing uNK cell cytolytic activity to expand maternal vascular support for developing conceptuses. Uterine NK cells are present and increase in the pig endometrium during early pregnancy due to the presence of the conceptus [75, 76]. These uNK cells may also transform into larger more granulated forms with increased cell cytolytic activity due to the uterine microenvironment [77]. This increase in functional activity of uNK cells in response to the conceptus does not occur in pseudopregnant pigs, which indicates an affect of the conceptus that is independent of conceptus estrogens [77]. Similar to the mouse, uNK cells may be involved in vascular changes that are important for embryo survival in pigs. Indeed, endometrial lymphocytes isolated near healthy conceptuses, but not those from sites of fetal arrest, are more numerous and express genes that are linked to angiogenesis [78]. These changes are associated with development of the subepithelial capillary bed, which is necessary for conceptus survival.
In conclusion, insights into the complex and overlapping events of pregnancy recognition and endometrial remodeling for implantation and placentation have been gained through examining the uterine expression levels of IRF1, STAT2, MIC, B2M, and IRF2 in terms of stage of estrus cycle, day of pregnancy, treatment with E2, and intrauterine infusion of CSP in pigs. The results suggest that pig conceptuses orchestrate precise temporal and spatial changes in uterine gene expression through initial secretion of estrogen, followed later by the expression of proteins, such as IFND and IFNG. Estrogens from pig conceptuses or injected E2 increase IRF2 expression in the LE and limit the expression of selected ISGs, including IRF1, to the underlying stroma. It is likely that many other uterine genes critical for pregnancy success are regulated by a similar interplay between conceptus steroids and proteins. Since the trophoblasts of ruminants, rodents, primates, and pigs share the characteristic of secretion of multiple paracrine factors that profoundly affect uterine gene expression and uterine remodeling, insights from the present studies advance our understanding of early pregnancy across mammalian species. While the key players at the uterine-placental interface require further definition, the interactions of estrogen, IFNs, and ISGs, including IRF1 and IRF2, described here highlight the complex and precisely orchestrated interplay between the endometrium and conceptus that influences conceptus survival, implantation, and development.
FOOTNOTES
1Supported in part by NIH grant P30ES0910607 and by NRI grant 2006-35203-17199 from the USDA Cooperative State Research, Education, and Extension Service. 
Correspondence: 2Greg A. Johnson, Department of Veterinary Integrative Biosciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX 77843-4458; FAX: 979-847-8981; e-mail: gjohnson{at}cvm.tamu.edu
Received: 17 February 2007.
First decision: 14 March 2007.
Accepted: 26 April 2007.
REFERENCES
by the uterine endometrium. Prostaglandins 1977; 14:397–400[CrossRef][Medline]
stimulates the Janus kinase-signal transducer and activator of transcription pathway in bovine endometrial epithelial cells. Biol Reprod 2001; 64:654–665This article has been cited by other articles:
|
|
 |
|
  M. M. Joyce, J. R. Burghardt, R. C. Burghardt, R. N. Hooper, F. W. Bazer, and G. A. Johnson Uterine MHC Class I Molecules and {beta}2-Microglobulin Are Regulated by Progesterone and Conceptus Interferons during Pig Pregnancy J. Immunol., August 15, 2008; 181(4): 2494 - 2505. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
