| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Animal Science (J.W.R., M.D.A., U.D., R.D.G.) and Biochemistry and Molecular Biology (P.J.A.), Oklahoma State University, Stillwater, Oklahoma 74078; Department of Agriculture (F.J.W.), Cameron University, Lawton, Oklahoma 73505; Department of Veterinary Integrative Biosciences (G.A.J.), Texas A&M University, College Station, Texas 77843; and Department of Animal Sciences (K.M.W., R.S.P.), University of Missouri-Columbia, Columbia, Missouri 65203
Address all correspondence and requests for reprints to: Rodney D. Geisert, S108 Animal Science Research Center, Columbia, Missouri 65211. E-mail: GeisertR{at}missouri.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Critical parameters exist with regards to the specific spatiotemporal release of estrogen during this period of pregnancy in the pig. Insufficient distribution of estrogen, as seen in litters with less than two piglets per uterine horn at the time of trophoblast elongation, results in the failure to prevent luteolysis and subsequent termination of pregnancy (4, 5). On the contrary, adverse timing of estrogen exposure to the dam on d-9 and -10 gestation results in conceptus degeneration during the period of placental attachment to the uterine surface (6, 7, 8, 9). However, the same dosage of estrogen given on d 11 and 12, in synchrony with conceptus synthesis and release of estrogen, has no adverse effect on conceptus development, pregnancy recognition, and implantation (10, 11). In the commercial swine industry, it is well known that ingestion of moldy corn before d-11 pregnancy causes total litter loss due to the presence of naturally occurring estrogenic aflatoxins, such as zearalenone (12).
Although estrogen is required as a maternal recognition of pregnancy signal in the pig, timing and extent of estrogen exposure dramatically affect conceptus development and survival. The diffuse type of porcine placental attachment occurs between d 13 and 18 of gestation, and is associated with a thickening of the uterine glycocalyx (2, 13, 14). Disruption of the uterine glycocalyx is closely associated with embryonic mortality that occurs in gilts treated with estrogen on d-9 and -10 gestation (14). The ability of estrogen to influence implantation success or failure also occurs in mice, in which high concentrations of endogenous estrogen refine the implantation window, and alter temporal and spatial gene expression in the uterine endometrium during blastocyst attachment (15). We propose that administration of exogenous estrogen on d-9 and -10 gestation prematurely initiates the cascade of molecular events normally induced by conceptus release of estrogen into the uterine lumen. The result is alteration of key endometrial secretory events during implantation that mistimes synchrony between uterine receptivity and conceptus development.
Therefore, the aim of this study was to identify and analyze endometrial molecular markers after endocrine disruption of pregnancy in the pig. Identification of endometrial changes in gene expression in response to premature exposure to exogenous estrogen will assist in determining factors regulated by conceptus estrogen that are critical to the initial stages of attachment during pregnancy in the pig.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Tissue collection
Pregnant gilts (four animals per treatment per day) were randomly assigned to one of the following treatment groups: (a) control [corn oil (CO)], im injection of CO (2.5 ml) on d-9 and -10 gestation; or (b) estrogen [estradiol cypionate (EC)], 5 mg im injection of EC (Legere Pharmaceutical, Inc., Scottsdale, AZ) in CO on d-9 and -10 gestation. Gilts were hysterectomized (four gilts per treatment per day) through midventral laparotomy, as previously described (7), on either d-10, -12, -13, -15, or -17 gestation. After induction of anesthesia with 1.8 ml im administration of a cocktail consisting of 2.5 ml xylazine (100 mg/ml; Phoenix Scientific, Inc., St. Joseph, MO) and 2.5 ml ketamine (ketamine HCl 100 mg/ml; Phoenix Scientific, Inc.) in 500 mg Telazol (tiletamine HCl and zolazepam HCl; Fort Dodge, Syracuse, NE), anesthesia was maintained with a closed circuit system of halothane (5%) and oxygen (1.5 liters/min). Immediately after removal, each uterine horn was flushed with 20 ml of a physiological saline, and conceptuses were removed. After conceptus removal, one uterine horn was cut along its antimesometrial border, and endometrium (5–10 g) was removed with sterile scissors. Endometrium was snap frozen in liquid nitrogen and stored at –80 C until analyzed. In addition, several sections (
0.5 cm) from each uterine horn were fixed in fresh 4% paraformaldehyde in PBS (pH 7.2) and embedded in Paraplast-Plus (Oxford Laboratory, St. Louis, MO) to be used for in situ hybridization.
RNA isolation
Total RNA was extracted from uterine endometrium tissue using the RNAwiz reagent (Ambion, Inc., Austin, TX) according to manufacturer’s recommendations. Approximately 500 mg endometrium was homogenized in 5 ml RNAwiz reagent using a Virtishear homogenizer (Virtis Co. Inc., Gardiner, NY). RNA pellets were rehydrated in nuclease-free H2O and stored at –80 C. RNA content was estimated spectrophotometrically and purity determined by the 260:280 ratio. RNA quality was assessed through agarose gel electrophoresis.
Microarray analysis
Microarray analysis was conducted using a spotted cDNA array that represents mRNA transcripts from pig brain, oviduct, uterine endometrium, oocytes, early embryos, peri-implantation conceptuses, fetal tissues, and ovarian tissues. Developed at the University of Missouri, the array contains 19,968 features, representing 14,970 PCR-amplified cDNA transcripts, of which 4,108 have not been identified in GenBank, meaning approximately 10,800 genes with known identity can be interrogated. Specific procedures used in the development of the array have been described (16).
cDNA synthesis and hybridization
Total RNA (20 µg) from endometrial tissue collected on d-10, -13, and -15 pregnancy representing each day x treatment combination was reverse transcribed into cDNA and labeled using the 3DNA Array 50 Expression Array Detection Kit (Genisphere Inc., Hatfield, PA). This method uses a dendrimer ridden with fluor molecules that hybridizes to a "capture" sequence specific to an oligonucleotide used in the RT reaction during cDNA synthesis. Hybridizations were conducted per the manufacturer’s recommendations. Four hybridizations were conducted for each day to compare mRNA abundance differences between CO and EC-treated gilts on d 10, 13, and 15. Day-12 tissue samples were not included in the microarray analyses due to morphological variation of conceptus development between the litters, and d 17 was not included in the analysis because conceptus mortality and degradation were conclusive. For each technical replication (n = 4), the total volume from the cDNA synthesis reaction was combined with the cDNA created for the opposite treatment to compare differences using dual-channel analysis. After combining the cDNA synthesis reactions for each treatment (260 µl), the cDNA was concentrated to 3–10 µl volume using Microcon YM-30 centrifugal devices (Millipore, Billerica, MA). Nuclease-free water was added to bring the total volume of the concentrated cDNA to 10 µl. Hybridizations were performed according to the manufacturer’s recommendations. For the primary hybridization, the hybridization solution (10 µl concentrated cDNA, 25 µl 2x formamide hybridization buffer, 2 µl LNA dT blocker, and 13 µl nuclease-free water) was denatured at 80 C for 10 min, applied to the array slide using a 22- x 40-mm LifterSlip (Erie Scientific Co., Portsmouth, NH), then incubated at 53 C for 16 h in a humidified hybridization cassette. After hybridization, coverslips were washed off in 2x standard saline solution (SSC), 0.2% sodium dodecyl sulfate at room temperature (RT), followed by a series of washes (2x SSC, 0.2% sodium dodecyl sulfate, 65 C, 15 min; 2x SSC, RT, 15 min; and 0.2x SSC, RT, 15 min). Slides were rinsed in 95% EtOH for 2 min at RT to fix cDNA molecules and then dried on a slide centrifuge. Secondary hybridizations, the attachment of the fluorescent dendrimer ridden capture sequence to probes bound during the primary hybridization, were conducted at 50 C for 3 h in a humidified hybridization cassette. The wash series after secondary hybridization was identical to the wash series after the primary hybridization. Slides were dried on a slide centrifuge and stored in a light-protected box.
Imaging and microarray data acquisition
Each microarray slide was scanned with both the Cy3 and Cy5 channels using the ScanArray Express (PerkinElmer, Wellesley, MA). Laser power and photomultiplier tube (PMT) gain were adjusted for each slide to minimize variation between wavelengths. Raw data files (.gal and .gpr files) are available (http://animalsciences.missouri.edu/research/supplemental_data/).
Analysis using GenePix Auto Processor (GPAP)
GPAP 3.0 software (http://darwin.biochem.okstate.edu/gpap3; Weng, H., and P. J. Ayoubi, in preparation) was used for data preprocessing, background correction, Local Loess pin by pin intensity normalization, and microarray statistical analysis. Genes that were determined to change at least 1.8-fold in response to EC with a P value < 0.1 were included in further analysis using the Database for Annotation, Visualization and Integrated Discovery (DAVID) (2.1; http://niaid.abcc.ncifcrf.gov/).
Analysis by DAVID
DAVID is a program that enables the use of microarray gene lists to generate specific functional annotations of the biological processes affected by the treatment as determined through microarray experiments (17). A gene list containing the GenBank accession numbers for all genes affected by the treatment for all days was compiled to determine the underlying biological themes that were altered due to EC treatment. The accession numbers used were those assigned to each gene during annotation as previously described (16). Based on the gene ontology (GO) assessment of the biological process, molecular function, and cellular component of each gene altered by the treatment, clusters of functional annotation terms were generated. Not all differentially expressed genes were used in the functional annotation because a number of genes differentially expressed were unique or lacked sufficient biological annotation to be useful for functional annotation.
Quantitative one-step RT-PCR
Quantitative analysis of secreted phosphoprotein 1 (SPP1) (also known as osteopontin), aldose reductase (AKR1B1), neuromedin B (NMB), and CD24 antigen (CD24) mRNA were assayed using quantitative real-time RT-PCR and a fluorescent reporter. AKR1B1, CD24, SPP1, and NMB were specifically chosen to validate the array data because they represent genes that, based on current literature and/or the DAVID annotation clustering, may pose a significant impact on pregnancy through their potential ability to affect conceptus growth and implantation to the uterine endometrium. The PCR amplification was conducted using the ABI PRISM 7500 Sequence Detection System (PE Applied Biosystems, Foster City, CA). The samples were evaluated for SPP1 expression differences using a dual-labeled probe designed to have a 5' reporter dye (6-FAM) and a 3' quenching dye (TAMRA), and to anneal between the forward and reverse primers. Thermal cycling conditions were 50 C for 30 min and 95 C for 15 min, followed by 40 repetitive cycles of 95 C for 15 sec and 60 C for 1 min. Amplification differences for AKR1B1, CD24, and NMB mRNA were detected using the intercalating dye, SYBR green. Primer and probe sequences are presented in Table 1
. Cycling conditions for SYBR green detection were 50 C for 30 min and 95 C for 15 min, followed by 40 repetitive cycles of 95 C for 15 sec, and variable annealing temperature for 30 sec, 72 C for 33 sec, and a variable temperature during fluorescent detection for 33 sec. Fluorescence detection temperature was determined by evaluating melting curve analysis for the samples and the no template control amplification plot. One hundred nanograms of total RNA were assayed for each sample in duplicate for each target template. 18S ribosomal RNA was assayed as a normalization control to correct for loading discrepancies for all samples assayed. To confirm that genomic DNA contamination was not contributing to the amplification, pooled samples were assayed in the presence and absence of reverse transcriptase to ensure only synthesized cDNA contributed to the amplification of the target. Relative mRNA abundance was determined using a previously established method (9). Relative mRNA expression units between sample means was calculated by assuming an amplification efficiency of two and applying the equation 2
Ct for each sample mean CT.
|
5 µm) were deparaffinized, rehydrated, and deproteinated, then hybridized with radiolabeled antisense or sense porcine cRNA probes (5.0 x 106 counts per min/slide) synthesized by in vitro transcription with [
-35S]uridine 5-triphosphate (MP Biomedicals, Irvine, CA). For in vitro transcription, appropriate RNA polymerases (T3, T7, SP6) were used with the linearized plasmid carrying the transcript printed on the array (16). After hybridization washes, and RNase A digestion, hybridized slides were exposed to Biomax maximum resolution film (Eastman Kodak, New Haven, CT) overnight to determine signal strength. Autoradiography was performed using NTB liquid photographic emulsion (Eastman Kodak). Slides were dipped in emulsion and exposed at 4 C for a period of time relative to signal strength, developed in Kodak D-19 developer, counterstained with Harris modified hematoxylin (Fisher Scientific, Fairlawn, NJ), dehydrated, and protected with coverslips.
Photomicrography
Digital photomicrographs of in situ hybridization, bright-field and dark-field images of liquid emulsion autoradiography, were collected using a Nikon Eclipse E6000 microscope (Nikon Corp., Tokyo, Japan) interfaced with the CoolSNAPcf digital camera equipped with a cooled charge-coupled device (Photometrics, Tucson, AZ) and imaging software (MetaVue; Molecular Devices, Downingtown, PA). Photographic plates were assembled using Adobe Photoshop (version 6.0; Adobe Systems Inc., San Jose, CA).
Statistical analysis
Quantitative RT-PCR CT values were analyzed using PROC MIXED of the Statistical Analysis System (SAS Institute Inc., Cary, NC). Analysis of endometrial gene expression tested for the effect of treatment, day, and day x treatment interaction. Significance (P < 0.05) was determined by probability differences of least squares means. Figures representing fold differences in gene expression have superscripts above bars depicting significant differences as determined by the CT values (P < 0.05).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Exogenous estrogen induces altered expression patterns for numerous genes
Microarray hybridizations were conducted to identify differential gene expression between EC and CO-treated pigs on d-10, -13, and -15 gestation. The specific days for microarray analyses represent three critical periods in the establishment of pregnancy: pre-conceptus estrogen release (d 10), post-conceptus estrogen release (d 13), and initiation of conceptus attachment to the uterine surface (d 15). After hybridization, scanning, raw data acquisition, and normalization, eight, 32, and five endometrial genes were identified up-regulated at least 1.8-fold (P < 0.1) in the EC gilts on d 10, 13, and 15, respectively. A total of one, 39, and 16 endometrial genes expressed in EC were down-regulated at least 1.8-fold (P < 0.1) on d 10, 13, and 15, respectively (Tables 2–4). Of the total 77 differentially expressed genes, 21 represented novel sequences. The remaining 56 differentially expressed genes were assigned a putative annotation based on their associated GenBank accession number representing factors involved with attachment, immunology, transcription regulation, protein regulation, and metabolism.
|
|
|
). Collectively, regulation of cellular/physiological processes, biosynthesis, immune response, multiple aspects of metabolism, apoptosis, transport, and calcium and metal ion binding are among the general functional themes affected. Other functional processes identified by DAVID based on the expression differences that were not included in a cluster include response to chemical stimuli, cytoskeleton protein binding, and signal transduction.
|
A day x treatment interaction (P < 0.03) was detected for endometrial AKR1B1 gene expression. In CO gilts, AKR1B1 expression was relatively low on d 10, 15, and 17 in contrast to a 25- and 45-fold greater expression that occurred on d 12 and 13, respectively (Fig. 1A). Although no significant difference in AKR1B1 abundance was detected between CO and EC treatments on d 10, 12, 15, or 17, EC treatment caused a 40-fold reduction in AKR1B1 expression (P < 0.001) on d 13.
|
). In comparison to d 10, endometrial CD24 expression increased 8, 53, 276 and 353-fold in CO gilts on d 12, 13, 15, and 17, respectively. EC treatment increased CD24 abundance on d 12 (6.7-fold) and 13 (6.5-fold) compared with CO, whereas expression was similar between the two treatments on d 10, 15, and 17 (Fig. 1B).
Quantitative analysis of SPP1 mRNA abundance was affected by day (P < 0.001). Gene expression increased from d 10–17 with d-15 and -17 endometrial SPP1 expression 61 and 115-fold greater, respectively, compared with d 10 (Fig. 1C). Although not statistically significant, SPP1 expression was numerically greater (7-fold) in EC-treated gilts compared with CO-treated gilts on d-13 pregnancy. No day x treatment interaction was detected (P = 0.19).
A significant day x treatment interaction (P < 0.001) was detected for endometrial NMB mRNA expression (Fig. 1D). In comparison to CO gilts, expression of NMB was greater in the EC-treated gilts on d 10 (P < 0.002) and 12 (P < 0.04) by 6.7- and 2.8-fold, respectively. Alternately, expression tended to be reduced (2.7-fold) in EC-treated gilts on d 13 (P < 0.07) and significantly reduced 4.4-fold on d 15 (P < 0.01) when compared with CO gilts.
Disrupted gene expression patterns of AKR1B1, CD24, SPP1, and NMB are spatially associated primarily with the uterine luminal epithelium (LE)
Endometrial expression of AKR1B1 was localized to the uterine LE of both CO and EC gilts. Expression of AKR1B1 was detectable in the LE on d 12 and 13 in CO-treated gilts, and on d-12 EC-treated gilts (Fig. 2). Administration of estrogen on d-9 and -10 pregnancy completely ablated the expression of AKR1B1 in the LE of gilts on d-13 pregnancy. Expression of AKR1B1 mRNA in the LE was low by d-15 and -17 gestation in both CO and EC gilts.
|
). Expression of SPP1 mRNA was apparent in the LE on d-12 and -13 EC-treated gilts with a strong signal detectable in the uterine LE on d 15.
|
). Slight CD24 expression was detected in the glandular epithelium (GE) on d 13–17 in both CO and EC gilts. Interestingly, the GE CD24 expression appeared to be elevated on d 10 in EC compared with d-10 CO gilts. Expression of CD24 is also present in the conceptus during the peri-implantation stage of development and appears to be expressed at much greater levels in the LE adjoining to a conceptus compared with LE distal from the conceptus within the same cross-section (Fig. 5).
|
|
). Although expression of NMB in the LE of EC gilts on d-13 pregnancy was greater compared with CO gilts, NMB expression was not visibly different on d 10, 12, 15, and 17.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The relatively few mRNAs whose expression was altered by EC treatment relative to the number of mRNAs spotted on the array is intriguing, however, conceptus loss due to the EC treatments was consistent, indicating that the alteration of just a few select transcripts is sufficient to negatively affect implantation and pregnancy establishment. Based on our microarray analysis, the greatest number of genes altered from the EC treatment occurred on d-13 gestation, 5 d after the initial estrogen injection. It may be possible that the disruptive response is a result of the interaction between the EC treatment and endogenous secretions of the elongating conceptus. Not only does the elongating conceptus produce estrogen that alters uterine secretion during normal pregnancy (1), but also there is elevated conceptus production of IL-1ß and a concomitant increase in endometrial expression of IL-1 receptors (20, 21). Interestingly, the EC injection on d-9 and -10 gestation in gilts does not alter the pattern of IL-1ß secretion by the conceptus (9), suggesting that synergistic effects that occur from simultaneous conceptus secretion of estrogen and IL-1ß may be abrogated by the premature estrogen exposure to the uterine endometrium. This may lead to the alteration or inhibition of the initial attachment of the trophectoderm to the uterine endometrium, which occurs on d-13 gestation (2, 13).
Based on evaluation through the DAVID, EC treatment altered genes that could be clustered into biological themes, suggesting effects on specific biological processes (Table 5). It is not surprising that the alterations in gene expression represent processes such as immune, inflammatory, wound response (annotation cluster 3; Table 5), and apoptosis-related events (annotation cluster 6) because the phenotypic outcome is pregnancy loss resulting from conceptus degeneration (9) and sloughing of the extracellular matrix (ECM) (14). Disruption of cation (calcium or zinc) ion binding (annotation cluster 9) has previously been due to exogenous estrogen administration on d-9 and -10 gestation (7).
Many of the genes affected by the EC treatment were previously differentially regulated in the porcine endometrium during peri-implantation, such as retinol binding protein 4 (22, 23), spermidine/spermine N1-acetyltransferase (24, 25), and SPP1 (26, 27). Numerous novel sequences for which a quality annotation could not be performed were also identified to be regulated by estrogen during the period of endometrial receptivity and conceptus attachment during pregnancy (Tables 2–4).
Several genes, such as CD24, NMB, AKR1B1, and allograft inflammatory factor 1, have good annotation but have not been significantly associated with the uterine endometrium during implantation in the pig. Because of their biological effects in protein modification/metabolism, immune/inflammatory response, and Ca2+ transport that are relative to critical events associated with conceptus development and endometrial reorganization during peri-implantation pregnancy, we further investigated endometrial expression of AKR1B1, NMB, SPP1, and CD24 using tissues from our endocrine disruption model through quantitative RT-PCR and in situ hybridization. The quantitative RT-PCR data for AKR1B1, SPP1, NMB, and CD24 was consistent with the DNA microarray data with respect to the pattern of gene expression in that the day with the largest difference between CO and EC gilts was consistent with the day for which microarray hybridizations identified a difference. However, numerically the gene expression differences were larger, as detected by quantitative PCR. The pattern of gene expression for AKR1B1, SPP1, and NMB determined by quantitative PCR correlates well with the temporal changes of expression in the uterine LE, as detected via in situ hybridization. The expression pattern of CD24 between quantitative PCR and in situ hybridization is less noticeable. This may in part be due to the spatial differences occurring as a result of conceptus location (Fig. 5) that complicate the effort to associate the data between the different techniques.
AKR1B1 is the rate-limiting enzyme of the polyol pathway responsible for the reduction of glucose to sorbitol, and is also involved in the reduction of toxic aldehydes, created by reactive oxygen species, to inactive alcohols (28). After the conversion of glucose to sorbitol via AKR1B1, sorbitol can be used to produce fructose. During normal pregnancy in the pig, the expression of AKR1B1 mRNA is transient and localized to the endometrial LE (Figs. 1A and 3). Peri-implantation expression of AKR1B1 in sheep appears to be regulated by the trophectoderm, for which expression is elevated between d-12 and -17 gestation (29). In human endometrial stromal (ST) cells, IL-1ß significantly up-regulates the expression of AKR1B1 (30). The induction of AKR1B1 gene expression in the pig may be regulated through conceptus IL-1ß because the peak AKR1B1 expression is concurrent to previously reported peak IL-1ß gene and protein production (21). The critical nature of AKR1B1 expression in the LE during implantation may be associated with both glucose and toxic aldehyde reduction. It is likely that AKR1B1 plays a critical role in production of sorbitol from glucose to be used in fructose production. Although glucose itself is a vital energy substrate, conversion of glucose to fructose may also provide a critical carbon source for DNA and RNA synthesis, whereas conceptuses undergo the dramatic increase in transcriptional activity and cellular mitosis after the initiation of trophoblastic elongation, occurring concomitant with LE AKR1B1 expression. Indirectly, the alteration in AKR1B1 expression leading to the potential shift in available fructose in the uterine lumen may affect the fructation and, subsequently, the activity and biological function of specific proteins during peri-implantation development in the pig. The ablated AKR1B1 expression preceding conceptus death on d-13 gestation demonstrated in this study is temporally associated with increased production of glucose and fructose in the uterine lumen during normal pig pregnancy (31). This suggests that the inability to use glucose through disrupted AKR1B1 function could be a mechanism leading to conceptus mortality.
SPP1 has been previously expressed in the uterine LE and GE of pigs during early pregnancy, and is sustained through at least to d-85 gestation (26). When given on d 11–15 of the estrous cycle, estrogen induces the expression of SPP1 in the LE, but not in GE, on d 15 (32). Our data indicate that estrogen stimulates LE expression of SPP1 in a progesterone-primed uterus. In the CO pregnant gilts, conceptus estrogen synthesis and release stimulated SPP1 expression on d 15, whereas early administration of estrogen in EC gilts prematurely induced LE expression.
SPP1 has numerous potential functions in the uterus during the establishment of pregnancy in species forming an epitheliochorial type of placentation, such as the pig (27). These include contributions to ECM formation, migration of immune cells, alterations in intracellular calcium levels, and activation of phosphatidylinositol 3'-kinase activity, and SPP1 has also been hypothesized to form as a bridging ligand between the uterine LE and the conceptus trophectoderm (27). The premature induction of SPP1 after EC treatment may affect several of these physiological roles to disrupt the implantation process. Blair et al. (14) documented the sloughing off of the ECM of gilts by d-15 gestation after injections of estradiol valerate (EV) on d 9 and 10, also resulting in total conceptus mortality. A single estradiol injection of EV on d 11 of the estrous cycle results in an approximate 4.5-fold increase in calcium concentrations in uterine flushings 12 and 24 h post injection (33). However, there is no increase in uterine luminal calcium content 24 h after treatment of gilts with EV on d 9, which also dramatically (>12-fold) reduces the surge release of calcium in uterine flushing of pregnant gilts on d 12 (7). Although EC alteration of SPP1 expression in LE cannot be directly linked to overall disruption of endometrial ECM formation and calcium transport, its patterns in our endocrine disruption model are similar. Although SPP1 expression increase occurs during normal peri-implantation pregnancy in the pig, the premature expression in the LE may be sufficient to affect ECM integrity and function before and during trophectoderm attachment, alter calcium concentrations in the uterine lumen, and alter immune cell trafficking, all of which, collectively or individually, could impact the developmental potential of the conceptuses.
Similar to SPP1, the expression of CD24 was also localized to the uterine LE and prematurely elevated in EC-treated gilts. CD24 is a membrane-bound molecule that lacks a cytoplasmic component and has specific binding ability for P selectin, phenotypically suggesting that CD24 may serve as a mucin-like adhesion molecule (34). However, the interaction of P selectin and CD24 may also induce TH1 lymphocyte trafficking because overexpression of P selectin is associated with increased TH1 lymphocytes in human patients suffering from spontaneous miscarriage (35). Interestingly, treatment of abortion prone mice (CBA/J x DBA/2J) with anti-P selectin monoclonal antibody before implantation significantly reduced the occurrence of abortion and the production of interferon-
and TNF- production by decidual lymphocytes (36). Because of the multiple roles that CD24 has within the immune response, premature CD24 expression in EC-treated gilts may elicit an inappropriately timed endometrial inflammatory response that disrupts conceptus attachment to the uterine surface. Although P selectin expression is not well characterized in the pig, an mRNA sequence has been isolated from uterine endometrium during early pregnancy, sequenced, and annotated as P selectin (GenBank accession no. DQ097865), suggesting that CD24 may serve as a bridging ligand during the attachment and adhesion of the conceptus trophectoderm to the uterine endometrium. It is clear that as early as d-12 gestation, both the conceptus and endometrium are producing copious amounts of CD24 in a spatiotemporal manner. If expression of P selectin by the endometrium is required for CD24 bridging during attachment, it is possible that increased CD24 expression by the uterine endometrium in response to the EC treatment could result in excessive endometrial P selectin/CD24 binding before the conceptus ability to express CD24 and bind P selectin on the LE and, thus, contributing to attachment failure in EC gilts.
NMB is a bombesin-like peptide first discovered in porcine spinal cord (37). Immunologically, NMB has been expressed by multiple cancer cell lines, and, like other bombesin-like peptides, can negatively influence IL-12 production and the maturation, antigen presentation, and the endocytotic capability of dendritic cells (38). Interestingly, the greatest endometrial NMB gene expression occurred in CO animals on d-13 gestation. Treatment of pregnant gilts with estrogen caused a significant increase, or advancement, in NMB expression on d 12 in the LE. Although NMB has stimulated smooth muscle contractions in a variety of tissues, including the uterus (39), there are no reports of endometrial NMB expression during establishment of pregnancy in pigs or other species. Although the expression of NMB and its receptor, NMBR, has affected maternal and emotional behavior (40, 41), NMB has also increased the release of arachidonic acid and promoted cell growth (42) and elevated cellular Ca2+ concentrations (43) in C6 glioma cancer cells through bombesin receptors. The promoted cell growth is consistent with NMB and NMBR, functioning as mitogens for epithelial cells lining the colon, whereas elevated NMBR was associated with those epithelial cells that had differentiated into tumor cells (44). The expression of bombesin-like peptides does appear to be a significant factor during pregnancy because neuromedin U was also shown to be differentially expressed in response to EC on d-13 gestation. Bombesin-like peptide receptors, specifically gastrin-releasing peptide receptor and NMBR, are expressed in the developing mouse embryo throughout gestation (45). The expression of NMB and neuromedin U in the uterine endometrium and their ability for bombesin-like peptides to function as mitogens may be critical for pregnancy establishment in pigs. The significant reduction of NMB expression on d 13 demonstrated in this study may alter tissue growth and reorganization in both the conceptus and uterine endometrium, negatively affecting implantation and conceptus survival.
Premature estrogen exposure advances the expression profile of multiple endometrial genes approximately 48 h compared with expression during normal pregnancy. These alterations in gene expression may cause asynchrony between the endometrium and conceptus, similar to that described by Polge (46), in which embryos transferred greater than 48 h out of synchrony with the uterine endometrium fail to establish pregnancy. Geisert et al. (47) demonstrated that d-6 embryos transferred into a more advanced uterine environment (48 h) are degenerate within 36 h after transfer. Uniquely, conceptus loss as a result of estrogen exposure on d 9 and 10 does not occur before d 12 because conceptuses continue to elongate and do not visibly degenerate until approximately d-15 gestation. These data suggest that exogenous estrogen administration causes significant alterations in endometrial gene expression on d-12 and -13 gestation, and may prevent attachment and promote rejection of the conceptus, resulting in its deterioration by d-15 gestation. It is difficult to pinpoint specifically the mechanism underlying the delayed conceptus death, and there is likely more than one specific contributor to this phenomenon. Based on the present data, we suggest that delayed conceptus degeneration is a result of aberrant gene expression pattern in the uterine LE caused by EC treatment on d 9 and 10, preventing attachment of the conceptus trophectoderm to the uterine epithelium, a process vitally important for continued development beyond d-15 gestation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Present address for J.W.R. and R.D.G.: University of Missouri-Columbia, College of Agriculture, Food and Natural Resources, Animal Science Department, Columbia, Missouri 65203.
Disclosure Statement: None of the authors have anything to disclose.
First Published Online July 19, 2007
Abbreviations: AKR1B1, Aldose reductase; CD24, CD24 antigen; CO, corn oil; DAVID, Database for Annotation, Visualization and Integrated Discovery; EC, estradiol cypionate; ECM, extracellular matrix; EV, estradiol valerate; GE, glandular epithelium; GO, gene ontology; GPAP, GenePix Auto Processor; LE, luminal epithelium; NMB, neuromedin B; NMBR, neuromedin B receptor; RT, room temperature; SPP1, secreted phosphoprotein 1; SSC, standard saline solution; ST, stromal.
Received May 7, 2007.
Accepted for publication July 12, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
by the uterine endometrium. Prostaglandins 14:397–401[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  F. W Bazer, T. E Spencer, G. A Johnson, R. C Burghardt, and G. Wu Comparative aspects of implantation Reproduction, August 1, 2009; 138(2): 195 - 209. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Song, K. A. Dunlap, J. Kim, D. W. Bailey, T. E. Spencer, R. C. Burghardt, G. F. Wagner, G. A. Johnson, and F. W. Bazer Stanniocalcin 1 Is a Luminal Epithelial Marker for Implantation in Pigs Regulated by Progesterone and Estradiol Endocrinology, February 1, 2009; 150(2): 936 - 945. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
