| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Departments of Animal Science,
b Surgery, and
c Immunology, University of Alberta, Edmonton, Alberta, Canada T6G 2P5
d AFRC Institute of Animal Health Compton, Newbury, Berkshire, RG16 0NN, United Kingdom
e Department of Veterinary Anatomy and Public Health, College of Veterinary Medicine, Texas A&M University, College Station, Texas 77843
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although MHC antigen expression within the placentas of farm animals such as sheep [3] cows [4, 5], and horses [6, 7] resembles that of the better studied mouse and human models (see review [8]), the development and morphology of placentas in the two groups are quite different [9]. Especially notable is the epithelio-chorial placentation of the pig [9].
The length of gestation in the pig is approximately 114 days. Between Days 14 and 22, the entire outer surface of the chorion (trophoblast) becomes attach to the endometrial epithelium of the uterus [9]. Around this time, the outer endodermal surface of the developing allantois begins fusing with the inner endodermal layer of the chorion, starting at the embryo and progressing to both ends of the elongated blastocysts. The mesoderm then develops between the two endodermal layers to innervate the allanto-chorionic sac. Practically the entire surface of the allanto-chorion forms the placenta, hence the name placenta diffusa. The trophoblast remains a noninvasive single layer in the pig [9]. Beginning around midgestation, the capillary plexuses at the tips of the chorionic villi penetrate between the trophoblast cells to about 2 µm from maternal epithelium at term [9].
The present study is the first to localize class I MHC expression on periimplantation blastocysts and extra-fetal membranes at term. We have characterized pig class I MHC expression both at the pre- and posttranslation levels, the former by Northern and in situ (mRNA) analyses with cross-reactive bovine and homologous porcine cDNA probes and the latter immunohistochemically with a monoclonal antibody that recognizes a common determinant on all polymorphic class I MHC antigens. We found that MHC class I antigens could not be detected on the surfaces of extra-embryonic/fetal membranes during gestation but that specific mRNA could be detected at low levels.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tissue collection and preparation Outbred pigs of the Yorkshire breed maintained at the University of Alberta Swine Facility were used in this study. The recommendations of the Canadian Council on Animal Care were followed in the treatment and care of the animals. Gilts were bred at the onset of their second estrous cycle (the first day of standing heat = Day 0), and blastocysts were collected at slaughter on Days 14, 16, and 22 after mating. Uterine tracts were excised from the animals within 15 min of slaughter, chilled on ice, and taken to the laboratory. Blood was also collected from each pregnant animal at slaughter for isolation of total RNA from peripheral blood leukocytes (PBL). PBL isolation was as previously described [10]. Animals of the same gestational age were not all slaughtered on the same day. In the laboratory, blastocysts from individual animals were flushed from both uterine horns with cold PBS and pooled. The total number of blastocysts flushed from pigs of the same gestational age were as follows: Day 14 (group 1), n = 5 pigs, 62 blastocysts as indicated by the number of embryonic discs found, 73 corpora lutea counted on the ovaries; Day 16 (group 2), n = 5 pigs, 58 blastocysts flushed, 70 corpora lutea counted on the ovaries; Day 22 (group 3), n = 3 pigs, a total of 32 attached (implanted) blastocysts mechanically removed.
Samples of tissues from different regions of the blastocysts were dissected for immunohistochemistry and samples from Day 16 blastocysts, for in situ hybridization (described below). The remaining blastocyst tissues from each pig were then aliquoted into 50-ml polypropylene centrifuge tubes (Corning Inc., Corning, NY; ~1 g of tissue/tube) containing guanidine thiocyanate salt (GTC; Sigma, St. Louis, MO) solution (7 ml) and immediately homogenized with an electronic homogenizer (Ultra Turrax T25-Sl; Janke&Kunkel KG, Staufen, Germany). The homogenates were then flash-frozen in liquid nitrogen and stored at -70°C. After flushing, the uterine horns were ligated at both ends and infused with a 0.2% collagenase solution and incubated for 30 min at 38°C. The dissociated endometrial cells from each reproductive tract were then pelleted and washed. Cell smears were prepared for immunocytochemistry, and the rest of the cells were used for isolation of total RNA. Approximately 108 cells were resuspended in GTC (7 ml) and frozen and stored as mentioned above.
In group 4, placental tissues at term were collected from two multiparous sows that farrowed on different days. The amniotic and chorioallantoic membranes were separated immediately after extrusion of the placentas (~20 min after birth of the last member of each litter). Samples from 10 different conceptuses from pig number 1 and 11 from pig number 2 were collected, minced, and immediately homogenized in 50-ml tubes (~1 g of tissue/tube) containing GTC and then frozen in liquid nitrogen to stop the action of placental ribonucleases. Frozen samples were taken to the laboratory and stored at -70°C for the extraction of total RNA. Also, samples of the amniotic and chorioallantoic membranes from randomly selected conceptuses were dissected and directly spread onto microscope slides with either the fetal- or maternal-face surface exposed for immunohistochemical staining.
Immunohistochemistry The samples for immunohistochemistry were treated as follows: the pieces of extra-embryonic tissues, embryos, and uterine endometrium were spread (embryos were squashed) onto microscope slides, fixed with cold acetone for 20 min at -18°C, and then air-dried and stored at -18°C until use. Term amniotic and chorioallantoic membrane pieces, uterine endometrial cells, and PBL were also air-dried onto slides and similarly fixed with cold acetone and stored. In addition, Day 14, 16, and 22 embryos, pieces of extra-embryonic tissues, and uterine endometrium randomly selected, were embedded in OCT compound (Miles Inc., Elkhart, IN) and frozen in liquid nitrogen. Frozen sections (6 µm) were cut and collected onto slides, air-dried, and fixed with cold acetone. Immunohistochemistry was carried out on at least three slides randomly selected from each of the above-mentioned tissues and cell types. The immunohistochemistry procedure was repeated three times in each experiment.
The indirect immunoperoxidase labeling procedure was as previously described [10]. A mouse monoclonal anti-porcine class-1 antibody (mAb, PT85, 1 mg/ml stock solution; VMRD Inc., Pullman, WA) was used as the primary antibody in the detection of class I MHC antigen expression. This mAb specifically recognizes a monomorphic determinant on all polymorphic class I swine leukocyte antigens (SLA; [11]). It was titrated from 1:25 to 1:1600 in 2-fold serial dilutions. The 1:200 dilution showed optimal staining and was selected for use in this study. A mouse monoclonal anti-porcine vimentin antibody (Dakopats, Glostrup, Denmark) supplied as a tissue culture supernatant was titrated in 2-fold serial dilutions. The 1:20 dilution was found optimal for indirectly labeling term-placental tissues. Peripheral blood leukocytes and endometrial cells in experiment I and endometrial tissue in experiment II were used as positive controls. In addition to normal mouse serum, mouse IgG1 and IgG2 isotype antibodies (Zymed Laboratories, San Francisco, CA) were used as reagent controls for the anti-vimentin and anti-class I MHC antibodies, respectively.
Preparation of radiolabeled probes
The bovine class I MHC probe used for Northern blot analysis was previously described [12]. Briefly, two cDNA clones 5.1 and 2.1, which code for the Aw10 and KN104 class I MHC molecules, respectively, were inserted into the plasmid pUC19 to obtain separate clones. The inserts, 1378 base pairs (bp) for Aw10 and 1396 bp for KN104, were isolated by digestion with EcoRI from these clones. A mixture of the Aw10 and KN104 cDNA inserts were labeled with 
-[32P]dCTP by nick translation and used for hybridization.
The BL3–7 bovine class I MHC cDNA clone [13] was used to prepare riboprobes for in situ hybridization. Polymerase chain reaction (PCR) was used to amplify a fragment of BL3–7 between exons 2 and the junction of 5 and 6, approximately 770 bp in length. This fragment was further amplified using oligonucleotide primers designed to generate an EcoRI site at the 5' end and a blunt 3' end. The transcription vector pcDNA 3 (Invitrogen, San Diego, CA) was cut with EcoRI and EcoRV, and the PCR-amplified fragment was ligated between the T7 and Sp6 promoters. The pcDNA plasmid was linearized by digestion with BamHI and NotI restriction enzymes for the generation of the antisense and sense riboprobes, respectively. The production of riboprobes was performed with an in vitro transcription kit (Ambion, Austin, TX). Linearized plasmid (1 µg) was incubated with 0.5 mCi/ml [35S]CTP (1200 Ci/mmol; NEN, Du Pont Canada Inc., ON, Canada), 0.5 mM each of nucleotides ATP, GTP, and UTP, RNasin, 10 mM dithiothreitol, 0.2 U/ml RNA polymerase Sp6 (anti-sense), and T7 (sense) for 1 h at 37°C. Approximately 107 cpm were incorporated into RNA probes per microgram of template DNA.
Northern blot analysis Blastocysts from the same pig were pooled and then divided into approximately 1-g aliquots of tissues and frozen in 50-ml polypropylene tubes containing 7 ml GTC. Because of the small amount of blastocyst tissues per animal at Days 14 (group 1) and 16 (group 2) of gestation, two pools were made of the tissues per group. In group 3, blastocyst tissues from each of the two pigs at Day 22 of gestation were treated as a separate pool. Total RNA was isolated at different times from each of the two pools per gestational age group. Samples of PBL and endometrial cells from all three groups were pooled before RNA isolation. Total RNA was isolated from the following tissues and cells: Day 14, 16, and 22 blastocysts, uterine endometrial cells, PBL, amniotic and chorioallantoic membranes from term placentas (group 4), and the Jag-1 trophoblast cell line [10]. Samples of cells (~1 x 108) and tissues (~1 g), previously homogenized in GTC and frozen upon collection, were thawed in a 37°C water bath. The DNA in the samples was sheared using 20-ml syringes fitted with 18-gauge needles. The total RNA from the tissues and cell homogenates was extracted by the GTC/CsCl method of Chirgwin et al. [14]. Approximately 50 µg of total RNA from each sample (except for that of PBL, in which 20 µg, 10 µg, and 5 µg of RNA were separated in lanes 8, 9, and 10 of Fig. 3, respectively) was separated by electrophoresis on 1% agarose gels containing 2.2 M formaldehyde and blotted onto nitrocellulose filters (NEN Research Products, Boston, MA).
|
Filters were baked for 2 h at 80°C under a vacuum and prehybridized for 3 h at 42°C in a solution of 50% formamide, 20 mM NaH2PO4, 4-strength SSC (single-strength = 0.15 M sodium chloride, 0.015 M sodium citrate), 2 mM EDTA, 4-strength Denhardt's solution (single-strength = 0.02% BSA, Ficoll, and polyvinylpyrrolidone), 1% SDS, and 100 µg/ml sonicated denatured salmon sperm DNA. Hybridization was performed in the same solution at 42°C for 16–20 h with a probe prepared from a mixture of two bovine class I cDNAs. The filters were washed at room temperature with double-strength SSC and 0.1% SDS. Autoradiography was performed by exposing Kodak X-Omat film (Rochester, NY) to the nitrocellulose filters at -70°C in the presence of an enhancing screen. This experiment was repeated at least three times with different durations of autoradiographic exposure. The filters were stripped and rehybridized to a ß-actin cDNA probe.
In situ hybridization The procedure used for in situ hybridization was previously described [15]. Briefly, samples of extra-embryonic tissues and embryos from Day 16 blastocysts and strips of Day 16 endometrial epithelium were fixed in a 4% paraformaldehyde/PBS solution overnight at 4°C. The tissues were cryoprotected by immersion for 2 h in each of a graded series of sucrose/PBS solutions (12%, 16%, 18%), embedded in OCT compound, and frozen in liquid nitrogen. Frozen sections (6 µm) were placed on slides, fixed with 4% paraformaldehyde/PBS for 20 min, and then dehydrated in a graded series of ethanol dilutions.
For in situ hybridization, 1 ml of prehybridization mixture was added to each slide, and the slides were incubated for 1 h at 37°C. Hybridization mixture (100 µl) containing the labeled RNA probes was added to each slide, and hybridization was performed at 45°C for 15 h in a humidified chamber. The slides were then washed, and autoradiography was performed at -20°C for 10 days. Thereafter, the slides were developed, fixed, washed, and counterstained with hematoxylin and photographed under brightfield illumination. This experiment was carried out twice on different tissue samples for a total of at least four slides per tissue type.
Experiment II
Tissue collection and preparation Gilts were bred on the first day of their second estrous cycle and hysterectomized on Day 25 of pregnancy or allowed to go to term (~114 days postcoitum). The reproductive tracts from Day 25 gilts were immediately placed on ice, and five conceptuses were mechanically removed from randomly selected attachment (implantation) sites from each tract. At each site the embryo, extra-embryonic membranes, and a section of the uterine endometrium were collected into separate tubes, flash-frozen in liquid nitrogen, and stored at -70°C until total RNA extraction. Additional samples of embryos, extra-embryonic membranes, and uterine endometrium from different implantation sites were collected for immunostaining (described above). Samples were collected from a total of three Day 25 pregnant gilts representing three replicates.
At term, placental tissues were collected from three gilts that farrowed on different days. Sections of the chorioallantoic and amniotic membranes belonging to five different piglets from each pig were flash-frozen in liquid nitrogen and stored at -70°C for later isolation of total RNA. Immediately after slaughter, sections of the heart, lungs, kidney, spleen, and liver from three mature gilts were flash-frozen in liquid nitrogen and stored at -70°C for later isolation of RNA. Additional samples of these tissues were embedded in OCT compound, frozen in liquid nitrogen, and sectioned as described above for immunohistochemistry.
Immunohistochemistry Samples of embryos proper, extra-embryonic chorioallantoic membranes, and the uterine endometrial epithelium from Day 25 pregnant animals were spread onto microscope slides and prepared for immunostaining as described in experiment I. Frozen sections (6 µm) of these tissues were also cut and fixed onto slides, as well as samples from the heart, lungs, kidney, liver, and spleen from the three mature gilts to be used as positive controls. The immunostaining procedure was as described in experiment I.
Northern analysis Total RNA was isolated from individual samples of embryos, extra-embryonic membranes, and uterine endometrium at each implantation site per pig. Tissues were homogenized manually under liquid nitrogen using a mortar and pestle on a bed of dry ice. Total RNA was extracted using Trizol reagent (Life Technologies, Grand Island, NY) according to the manufacturer's specifications. Total RNA (50 µg) from the three tissue types at each implantation site was electrophoresed in adjacent lanes (representing 5 implantation sites per replicate) and blotted onto nitrocellulose filters. Total RNA was also isolated from chorioallantoic and amniotic membranes from each piglet at term (representing extra-embryonic membranes from 5 different piglets for each of the three gilts farrowed). As positive controls, total RNA was isolated from heart, liver, lungs, kidney, and spleen from three different mature gilts. Total RNA (50 µg) from these tissues were electrophoresed in adjacent lanes and blotted onto nitrocellulose filters, and Northern hybridization was performed as described in experiment I.
Preparation of radiolabeled probes A homologous porcine class I cDNA probe was used in experiment II for Northern hybridization. This probe was prepared from four class I SLA genes (Pd-1, Pd-6, Pd-7, and Pd-14) kindly supplied by Dr. Douglas M. Smith, Oklahoma University Health Sciences Center, Oklahoma City). The cDNAs were cloned into the vector pBK-CMV. The Pd-1, Pd-7, and Pd-14 clones were digested with EcoRI and XhoI, and the Pd-6 clone with EcoRI and XbaI; the inserts were separated from the vector sequence by agarose gel purification. Each insert had an approximate size of 1.5 kilobases (kb). A mixture of equal concentrations of the four class I cDNAs were radioactively labeled and used as a probe for Northern hybridization to total RNA isolated from pig tissues on Day 25 of pregnancy. RNA blots were stripped after hybridization to the class I MHC probe and reprobed with the ß-actin probe as described in experiment I.
Statistics In experiment II, data obtained by Northern blot hybridization was analyzed on a Macintosh computer using the public domain NIH Image program (developed at the U.S. National Institute of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/). The data for class I MHC hybridization was normalized to that of the ß-actin control, and a ratio was obtained for each respective lane and analyzed by ANOVA. Mean separation was accomplished by Fisher's protected least-significant-difference method.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Class I MHC antigens were not expressed in periimplantation blastocysts In experiment I, extra-embryonic membranes of Day 14, 16, and 22 periimplantation blastocysts, whether prepared by microsectioning or whole mounts spread and fixed directly onto slides, did not stain immunohistochemically for class I MHC antigens. Figure 1a shows that neither the outer trophectoderm (closed arrow) nor the inner endoderm (open arrow) of a whole-mounted sample of Day 16 extra-embryonic tissue stained for class I MHC (data for Days 14 and 22 were similar and are not shown). Whole mounts or microsections of the embryonic portion of Day 14, 16, and 22 blastocysts also did not stain immunohistochemically (data not shown).
|
Class I MHC were not found on surfaces of the chorioallantoic membranes at term Exposure of the fetal face of term chorioallantoic membranes to the anti-class I mAb did not stain the parietal endoderm covering this surface; only hematoxylin-counterstained nuclei were evident (Fig. 1b, whole mount, closed arrows). On the other hand, the outer surface of the intricate branches of blood vessels underlying the endoderm were intensely labeled (Fig. 1b, open arrow). When the maternal face of term chorioallantoic membranes was exposed to anti-vimentin antibodies, only the inter-epithelial capillary plexuses at the tips and sides of the chorionic folds were intensely stained (Fig. 1c, whole mount). Vimentin, a cytoskeletal intermediate filament, is found in tissues of mesodermal origin (see [16]) such as blood vessels. The origin of the trophoblast epithelium covering the maternal-face surface of the chorioallantoic membrane is ectodermal. A pattern of labeling similar to that found with the anti-vimentin antibody was evident with the anti-class I mAb on this outer surface. The chorionic epithelium (trophoblast) did not stain (Fig. 1d, whole mount, closed arrows). However, the inter-epithelial capillary plexuses at the tips and sides of the chorionic folds were intensely stained (Fig. 1d, open arrows). The close proximity of these blood vessels to the surface (~1 µm, [9]) facilitated their reaction with the anti-class I MHC antibody.
Class I MHC antigens were not found on surfaces of the amniotic membranes at term
The epithelial surface of the amnion that faces the fetus did not express class I MHC (Fig. 2a, whole mount, closed arrow). Instead, prominently labeled foci of rounded cells were evident (Fig. 2a, open arrows). These class I-positive cell clusters were numerous and in some areas were confluent over this surface of the amnion (not shown). These cells were not further characterized in this study. The abembryonic surface of the amnion also did not stain for class I MHC (Fig. 2b, whole mount, closed arrow) except for the blood vessels found on this side (Fig. 2b, open arrows). Both uterine endometrial cells (Fig. 2c) and PBL (data not shown; see [10]) smeared onto slides were stained intensely for class I MHC and served as positive controls. Replacement of the mouse anti-class I MHC antibody with normal mouse serum or a mouse IgG2 isotype-matched antibody served as reagent controls and did not stain cells or tissues (reagent controls looked similar to negatively stained tissues exposed to the class I MHC antibody shown in Fig. 2; data not shown). Figure 2d (fetal-face surface) shows an example of vimentin-positive, round and stellate cells that were commonly detected on both surfaces of the amnion. However, these cells did not stain positive for class I MHC antigens and were not characterized further in this study.
|
Messenger RNA for class I MHC antigens was expressed at very low levels in periimplantation blastocysts and extra-fetal tissues at term Figure 3 shows a Northern blot hybridization analysis using the class I MHC-specific cDNA probe, which detected a single band with an approximate size of 1.5 kb. This band is similar in size to that found in cattle with the same probe [12].
It was evident that embryo/fetus-derived tissues and cells expressed considerably lower levels of class I MHC mRNA than did maternal tissues. Even though 2.5–10 times less total RNA from PBL (20 µg, 10 µg, and 5 µg, lanes 8, 9, and 10, respectively) than that from Jag-1 (50 µg, Fig. 3 lane 6) was loaded, the intensity of the MHC bands for PBL was much greater than that of Jag-1 RNA. The relative intensities of MHC bands (Fig. 3, center panel), rRNA (Fig. 3, upper panel), and ß-actin (Fig. 3, lower panel) control bands confirm this tendency. In addition, although similar amounts of RNA from chorioallantoic membrane (partially degraded, Fig. 3, lane 1), term amnion (lane 2), Day 22 (lane 3), Day 16 (lane 4), and Day 14 (lane 5) blastocysts were loaded, a clear and reproducible band was seen only for the tissues at term and Day 22 blastocysts, probably because of the vascularization of the latter tissues. Expression of class I MHC mRNA from endometrial cells (Fig. 3, lane 7) was similar to that of PBL (Fig. 3, lane 8). The large discrepancy in class I mRNA expression between embryo/fetus-derived tissues and those of maternal origin is emphasized in Figure 3, which shows results from titration of total RNA from PBL: 20 µg (lane 8), 10 µg (lane 9), and 5 µg (lane 10). Titration showed that 5 µg of total RNA from PBL contained many more class I mRNA transcripts than 50 µg of RNA from any embryo/fetus-derived tissues or cells.
Messenger RNA for class I MHC antigens was ubiquitously and uniformly localized at low levels in all tissues of Day 16 blastocysts Specific hybridization to the antisense riboprobe indicated that class I MHC mRNA was ubiquitously expressed in Day 16 blastocysts. In Figure 4, hybridization to the antisense probe is depicted on the left (a, c, e, and g) and that to the sense (control) probe on the right (b, d, f, and h). Silver grains were localized at low densities in both the extra-embryonic (Fig. 4, a and c) and embryonic (Fig. 4e) tissues when compared to the background densities of the sense controls (Fig. 4, b, d, and f) and to the endometrial epithelium tissue controls (g and h). The endodermal layer of extra-embryonic chorioallantoic tissue (Fig. 4c, arrow) showed a level of class I mRNA expression similar to that of the trophectoderm. The profound reduction in MHC class I mRNA in extra-embryonic and embryonic tissues was obvious when the uniformly low levels of hybridization in these tissues were compared to the considerably higher level of signal detected for Day 16 maternal endometrial epithelial strips (Fig. 4g, antisense and 4h, sense) as positive controls.
|
Experiment II
Class I MHC antigens were not found on Day 25 embryos and extra-embryonic membranes In experiment II, tissue samples taken from each implantation site were kept separately for immunohistochemistry. Representative tissue samples from the embryo, extra-embryonic membranes, and uterine endometrium taken from the same implantation site are shown in Figure 5. Class I MHC antigens were not detected by immunostaining on Day 25 embryos proper, whether compressed onto microscope slides (data not shown) or microsectioned for staining (Fig. 5a, hematoxylin-counterstained). Figure 5b represents a cross section of the chorioallantoic membrane adjacent to the point of attachment of the embryo. There was no staining for class I MHC antigens in this tissue except for that seen for large blood vessels in some of the microsections (Fig. 5b, inset). This pattern of class I MHC antigen expression was more distinct in whole mounts of tissues taken from this region, in which a larger surface area was exposed for immunostaining (Fig. 5d). Strong reactivity of the class I MHC antibody was evident along the larger, and lacking in the smaller, blood vessels. The chorionic epithelium (trophoblast) did not react with this antibody as only hematoxylin-counterstained nuclei were evident (Fig. 5d).
|
Tissues sections (Fig. 5c) or whole mounts (Fig. 5c, inset) taken from the ends of the chorioallantoic sac (where there are fewer blood vessels and the trophoblast cells are more squamous than columnar in shape) also did not stain for class I MHC antigens as only hematoxylin counterstaining was evident. Uterine endometrial samples at the site of implantation were used as positive controls. In whole mounts, the surface epithelial cells stained strongly for class I MHC antigens; so too did the underlying blood vessels (Fig. 5e, arrow). A cross section through the endometrium shows staining for class I MHC antigens in all tissues (Fig. 5f). Normal mouse serum and a mouse IgG2 isotype-matched antibody were used as negative controls and did not stain tissues (reagent controls looked similar to negatively stained tissues exposed to the class I MHC antibody shown in Fig. 5, data not shown). Tissues from the other parts of the body such as the heart (Fig. 5g), liver (Fig. 5h), kidney, lungs, and spleen (data not shown) expressed high levels of class I MHC antigens comparable to those seen for the uterine endometrium (Fig. 5, e and f). Figure 5, i and j, represents antibody isotype controls for the heart and liver tissues, respectively, showing no staining.
Messenger RNA for class I MHC antigens was expressed at very low levels in embryos and extra-embryonic tissues at Day 25 of pregnancy and at term Figure 6A1 shows the ethidium bromide-stained total RNA for the chorioallantoic (lane 1, ch1) and amniotic (lane 3, am1) membranes from the same piglet after farrowing, and the chorioallantoic membrane from a different piglet (ch2), together with samples from the heart, lungs, and liver from a mature gilt. The levels of hybridization to the class 1 MHC probe (Fig. 6A2) and to the ß-actin probe (Fig. 6A3) are shown. The mean intensities of the signal seen for the chorioallantoic (0.56 ± 0.1) and amniotic (0.63 ± 0.1) membranes were significantly lower than those of the heart (3.21 ± 0.6), lung (4.34 ± 0.4), liver (7.31 ± 0. 7), kidney (3.35 ± 0.5), and spleen (6.46 ± 0.43) (p < 0.0001). There was no significant difference between the chorioallantoic and amniotic membrane signal intensities.
|
Total RNA was isolated from individual tissue samples taken from 5 randomly selected implantation sites from each pig in this experiment. Approximately 50 µg of total RNA was separated by electrophoresis from the extra-embryonic membrane (m), embryo/fetus (f), and uterine endometrium (ue). Figure 6B1 (upper panel) shows the ethidium bromide-stained total RNA for one such implantation site. The integrity of the total RNA was not compromised, and the intensities of the 28S and 18S rRNA bands indicate relatively equal loading. This was confirmed by hybridization to a porcine ß-actin cDNA probe (Fig. 6B3, lower panel). The middle panel (Fig. 6B2) shows the hybridization pattern of the RNA blot to a homologous class I cDNA probe. The intensities of the signal seen for extra-embryonic membranes (1.67 ± 0.5, mean ± SD) and the fetuses (1.63 ± 0.18) were significantly lower than that observed for the maternal uterine endometrial samples (4.19 ± 0.18; p < 0.0001). The difference in class I MHC signal between the extra-embryonic membranes and that of the fetus was not significant. After the RNA blot was stripped and hybridized to the ß-actin cDNA probe, residual radioactive signal for class I MHC mRNA was still evident and was stronger for the endometrial than for the extra-embryonic and fetal samples (Fig. 6B3, lower panel arrow).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Immunohistochemistry on whole mounts of tissues was undertaken because with this technique morphology can be maintained and at the same time show discriminate staining among the different tissue layers. In addition, a greater surface area is available for staining. There was no staining for class I MHC on trophectoderm or endoderm of the extra-embryonic membranes of periimplantation blastocysts nor on the chorioallantoic membrane of term placentas. Placentas from the intermediate stages of pregnancy were not examined in this study. The absence of detectable cell surface expression of class I antigens on pig trophoblast at any stage of development is in accord with similar observations of its absence in sheep placentas [3].
Since the PT85 mAb is directed against pig thymocytes, cells that very likely do not express monomorphic, pregnancy-associated class I MHC antigens, it may not recognize these unique forms; hence, the possibility that monomorphic forms are expressed on pig trophoblast cannot be formally excluded. Human extravillous trophoblast subpopulations express the HLA-G antigen [18, 19], while the basal trophoblast of the rat expresses the Pa antigen [20, 21]; both are unique monomorphic class I antigens. Recently, HLA-G was shown to be restricted to differentiated cytotrophoblast [19], to be co-dominantly expressed in first-trimester trophoblast cells [22], and to present peptides in a manner similar to that of polymorphic class I HLA molecules [23, 24]. The W6/32 mAb directed against a monomorphic determinant of class I MHC detects both polymorphic and monomorphic MHC antigens in humans [25]. Therefore, it is possible that the PT85 mAb would also detect putative monomorphic forms of class I MHC antigens in the pig.
The lack of polymorphic MHC molecules on pig trophoblast follows that of both the sheep [3] and the horse, in which, except for the expression of class I MHC on the invasive trophoblast of the transient chorionic girdle cells, the noninvasive trophoblasts of the chorioallantoic membranes are class I-negative [6, 7, 26]. In contrast, in the cow placenta, which is structurally similar to that of the sheep (syndesmochorial), the noninvasive trophoblast of the interplacentomal allanto-chorion have been found to express class I in some instances [5]. This is especially perplexing since the same monoclonal anti-sheep class I antibody (SBU-1) was used in both the cow and sheep studies [3, 5].
The ontogeny of class I MHC antigen expression on the embryonic portion of periimplantation blastocysts detectable by immunoperoxidase staining was not addressed in this study. However, it should be noted that Day 14, 16, 22, and 25 embryos did not stain for class I MHC antigens. Furthermore, a very low level of class I mRNA was detected by Northern blotting in Day 25 embryos using a mixture of homologous cDNAs as a probe. A correspondingly low level of class I mRNA was localized by in situ hybridization in Day 16 embryos and extra-embryonic tissues. Day 16 represents the stage immediately following the onset of maternal recognition and attachment of the blastocysts [9], and periimplantation tissues at Day 16 of development were chosen for in situ hybridization. On the other hand, whether the homologous or heterologous class I MHC probes used in this study can detect all swine class I MHC mRNAs is not known. Mouse class I MHC antigens are barely detectable in the embryo proper as late as Day 13.5 of gestation [27]. Although preimplantation mouse blastocysts express class I MHC mRNA [28] and surface antigens [29], these appear to be absent during implantation and are detected again around Day 10 postimplantation, after which levels increase progressively to term [30].
In experiment I, although there was generally more degradation in RNA samples from tissues than from isolated cells, RNA from the chorioallantoic membranes at term appeared to have been the most degraded, probably because of the ~20-min retention time of the placentas after birth of the last member of the two litters in this study. However, in no instance did this degradation preclude some degree of hybridization to the class I MHC and ß-actin-specific cDNA probes. In experiment II, total RNA from the chorioallantoic and amniotic membranes of term placentas, and total RNA from the heart, liver, lungs, spleen, and kidneys of mature gilts were analyzed to confirm results found in experiment I because of the lower quality of RNA obtained for analysis in experiment I and to expand the panel of positive controls to confirm the difference seen between fetal and maternal MHC class I RNA expression. The patterns of expression of mRNA among the different tissue types were similar whether tissues from different implantation sites were pooled (experiment I) or kept separate (experiment II). The low levels of class I MHC mRNA expressed in extra-embryonic tissues correlates well with the inability to detect class I antigen expression in the pig placenta. In general, the amount of class I mRNA found in tissues and cells of fetal origin was considerably less than that found in maternal tissues. Results were similar with the homologous pig and heterologous cattle MHC class I cDNAs used as probes. Whether these low mRNA levels are reflected by correspondingly low levels of protein expression that are difficult to detect or whether the antigen is translationally regulated is not clear. Although the Jag-1 pig trophoblast cell line expressed slightly higher levels of class I mRNA than Day 14–22 blastocysts, MHC antigen was not detected [10], suggesting translational or posttranslational regulation. Posttranslational regulation of class I expression in embryo-derived cell lines of the mouse has been suggested [31]. On the other hand, transcriptional regulation of MHC expression has been suggested in mouse [32] and human [33] trophoblast subpopulations and may involve DNA hypermethylation [34].
Trophoblast subpopulations in the human and the rat appear to employ multiple levels of control, allowing the expression of monomorphic class I while simultaneously repressing polymorphic forms (see Introduction). Different regulatory mechanisms appear to control MHC expression in trophoblast and the embryo during development [35]. In addition, special mechanisms appear to have evolved in trophoblast to control MHC expression [36]. For instance, most trophoblast is refractory to cytokines, such as interferon 
, that are known to up-regulate MHC expression on other somatic cell types [37]. The expression of MHC class I antigens on trophoblast thus appears to be regulated temporally and spatially during development in a complex but highly efficient manner that operates at multiple levels. It appears to be independent of the degree of trophoblast invasiveness and disruption of blood vessels.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence and current address: J.J. Ramsoondar, Alexion Pharmaceuticals, ;co Columbus Farming Corp., Sherburne, NY 13460. FAX: 607 674 2410; jag;cacolumbusfarming.com
Accepted: September 18, 1998.
Received: June 1, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This 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]
|
|
|
|
 |
|
 Y. Choi, G. A. Johnson, T. E. Spencer, and F. W. Bazer Pregnancy and Interferon Tau Regulate Major Histocompatibility Complex Class I and {beta}2-Microglobulin Expression in the Ovine Uterus Biol Reprod, May 1, 2003; 68(5): 1703 - 1710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. Hill, D. H. Schlafer, P. J. Fisher, and C. J. Davies Abnormal Expression of Trophoblast Major Histocompatibility Complex Class I Antigens in Cloned Bovine Pregnancies Is Associated with a Pronounced Endometrial Lymphocytic Response Biol Reprod, July 1, 2002; 67(1): 55 - 63. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Mao, B. K. Treacy, F. R.C.L. Almeida, S. Novak, W. T. Dixon, and G. R. Foxcroft Feed Restriction and Insulin Treatment Affect Subsequent Luteal Function in the Immediate Postovulatory Period in Pigs: Progesterone Production In Vitro and Messenger Ribonucleic Acid Expression for Key Steroidogenic Enzymes Biol Reprod, January 1, 2001; 64(1): 359 - 367. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
