The rat placenta expresses paternal class I major histocompatibility antigens

The immunological problem of pregnancy: 50 years with the hope of progress. A tribute to Peter Medawar

This article is a tribute to the late Sir Peter Brian Medawar for his pivotal contribution in 1953 to the establishment of the field of Reproductive Immunology, with a brief and selective assessment of the progress made during the ensuing 50 years towards elucidation of the mechanisms responsible for the paradoxical survival of the conceptus as an intra-uterine allograft within the immunologically competent genetically alien female host. Medawar's succinct and stimulating theories have been central throughout the whole of this time and his basic conclusion, that the single most important factor ensuring the success of gestation is the anatomical separation of the fetus from its mother, remains substantially valid to this day. The extent to which other factors are of significance, particularly those relating to the relative roles of maternal adaptive and innate immune responses to the developing feto-placental unit, has yet to be fully defined.

Effect of rat placental culture supernatants on cellular and humoral immune responses

PROBLEM

To evaluate the effect of rat placental culture supernatants (PS) on spontaneous, mitogen- and alloantigen-induced lymphoproliferation, antibody synthesis regulation, and symmetric/asymmetric antibody ratio.

METHOD OF STUDY

The effect of PS was determined: (a) on cell proliferation of murine hybridoma cells and on spontaneous or ConA-induced proliferation of murine and rat splenocytes by thymidine incorporation; (b) on rat or mouse cell-mediated cytotoxicity (CMC) by 51Cr release; and (c) on antibody synthesis by enzyme-linked immunoadsorbent assay (ELISA).

RESULTS

With 20% PS, hybridoma cell inhibition was 37% and that of splenocytes up to 60%, whereas it was 75 and 43%, respectively, in the presence of ConA. Despite marked cell death, hybridoma proliferation index increased significantly. There was a drop in total antidinitrophenylated (DNP) immunoglobulin G1 (IgG1) antibody production and an increase in asymmetric antibody percentage, correlating with placental supernatant concentration.

CONCLUSIONS

Rat placental culture supernatants inhibit cell proliferation in all cases, diminish total antibody production, and increase the percentage of asymmetric antibodies by the hybridoma, and they increase antibody production by rat splenocytes.

Analysis of rat placental plasma membrane proteins by two-dimensional gel electrophoresis

The placenta plays an essential role in fetal growth and the maintenance of pregnancy and its functions are strictly controlled in a stage-specific manner. To gain an insight into placental functions and their regulation, we analyzed the plasma membrane proteins of rat placenta by two-dimensional polyacrylamide gel electrophoresis (2D/E). Plasma membrane fractions of the placenta obtained on days 12, 14, 16, 18 and 20 of pregnancy were purified by Percoll gradient centrifugation, and subjected to 2D/E analysis. After the proteins on the 2D/E gels had been visualized by silver staining, the patterns on the gels at different stages of pregnancy were compared using image analysis software. Proteins within an isoelectric point (pI) range of 4.0 to 7.0 and a molecular weight (Mw) range of 20-100 kDa were analyzed in detail, and about 800 proteins on average were recognized on each gel. Of these, the expression of 150 proteins was found to change dramatically according to the stage of pregnancy. According to their expression patterns, these proteins were categorized into two groups, Group I and Group II. The proteins belonging to Group I showed a higher intensity of expression on day 12 and disappeared on day 20. They included 119 plasma membrane proteins and were divided into five subgroups. Group II, which consisted of three subgroups, included 31 proteins showing a low or negligible expression on day 12 and higher expression on day 20. Most of the other membrane proteins (about 600) were expressed constantly during pregnancy. On the basis of our data, we constructed a database for plasma membrane proteins of the rat placenta.

Alloimmunization against well defined polymorphic major histocompatibility or class I MHC transfected L cells antigens can prevent poly IC induced fetal death in mice

METHOD

It is possible to induce increased fetal resorption in a number of inbred murine matings by injecting Poly (I) Poly (C12U) 3.5 days postconception, a maneuver associated with natural killer-mediated damage to the feto placental unit such as occurs in spontaneous fetal resorptions.

RESULTS

We show here that alloimmunization can block this effect. In addition, maternal immune responses induced by alloimmunization against isolated mutant class I or class II, as well as by immunization with class I MHC alloantigens (Kd) transfected L cells are sufficient to restore normal fetal viability. It is not necessary that the maternal immune response be specifically directed against paternal alloantigens fr the fetal protection to ensue, since the effect occurs in inbred matings when the mother is immunized against unrelated class I or class II alloantigens. As in previous studies conducted in the murine species, not all MHC alloimmunizations are protective. In addition, as control, immunization with a monomorphic class I MHC molecular (37), transfected L cells, sheep red blood cells or hen egg lysozyme is without effect.

CONCLUSION

These results indicate that defined MHC antigens can mediate fetal protection from induced fetal resorption, and suggest that one driving force in promoting MHC antigen polymorphism in mammals is their capacity to confer protection from NK mediated fetal demise.

Analysis of mRNA for class I HLA on human gametogenic cells

We have studied mRNA expression for Class I HLA (human leukocyte antigen) on male germ cells by amplification of gene fragments in PCR technique and by Northern hybridization. RNA was extracted from fractionated gametogenic cells (isolated from testis) and reversely transcribed. Then, cDNA was amplified for specific HLA sequence (HLA, -A, -B, -C). The specificity of this product was confirmed in "nested" PCR of 400 bp gene fragment coding for alpha 2 domain, alpha 3 domain, and the transmembrane portion of Class I HLA. The results indicate minimal expression of classical Class I HLA on gametogenic cells. Northern hybridization with 669 bp cDNA fragment (spanning for alpha 3 domain, transmembrane, cytoplasmic, and 3' untranslated region) resulted in a low intensity signal from gametogenic cell fractions and confirmed our findings obtained by PCR. The minimal expression of classical HLA antigens may create a neutral cover for the male reproductive system, thereby preventing an immunological response during germ cell differentiation.

Analysis of the expression and immunostimulatory capacity of class I major histocompatibility antigens on rat trophoblast cell lines

In rat strains expressing the a and other major histocompatibility complex (MHC) haplotypes, subpopulations of placental trophoblast cells synthesize the nonclassical class I Pa antigen in preference to the classical RT1.Aa antigen. In this study, a rat trophoblast cell line, R8RP.3, which was derived from midgestation placentas of PVG.R8 (RT1.Aa) rats, was shown to express class I antigens similarly to those of trophoblast cells in situ. Both unstimulated and IFN-gamma-exposed metabolically labeled R8RP.3 cells synthesized more Pa than RT1.Aa antigen. The reverse was true for labeled spleen cells from PVG.R8 rats. The R8RP.3 cells failed to stimulate allogeneic lymphocyte proliferation even when high levels of both classical and nonclassical class I MHC antigens were expressed on their membranes after incubation with IFN-gamma. These experiments thus supply the first evidence that the inductive phase of the immune response is not promoted by trophoblast cell class I MHC antigens, which could explain the failure of mothers to mount immune responses to class I MHC positive trophoblast cells.

Placental MHC class I antigen expression is induced in mice following in vivo treatment with recombinant interferon-gamma

Allogeneically pregnant mice (NFR/N (Swiss-derived) H2q x 57/B1 H2b) were injected daily from day 11 to 18 of gestation with recombinant rat interferon-gamma (IFN-gamma) and the effects on placental MHC antigen expression were evaluated by immunohistochemistry. The results demonstrated that a daily dose of 200,000 U/mouse induces a significant increase in placental MHC class I expression in the decidua and the basal zone spongiotrophoblast as well as the fetal mesenchyme, while the labyrinthine trophoblast was almost completely MHC negative. No significant induction of MHC class II antigen expression was observed in the placental tissue, with the exception of a few scattered cells. The MHC class II-inducing efficiency of the treatment protocol used was ascertained by staining maternal skin (ear), which proved to contain considerably elevated numbers of MHC class II-positive cells (mainly keratinocytes). The IFN-gamma treated group showed a higher incidence of fetal resorptions than the corresponding controls, but this was probably due to a non-specific toxic effect of the treatment as none of the sera analysed contained detectable amounts of anti-paternal antibodies, no infiltrating lymphocytes were detectable on histological examination of placental sections, and most of the IFN-gamma treated animals successfully completed their pregnancy.

Expression and analysis of the rat placental class I cDNA clone encoding the Pa antigen

The previously sequenced cDNA clone pARI.5 was recloned into the mammalian expression vector pcEXV3, and transient and permanent transfectants were prepared in COS7 green monkey kidney fibroblasts. The transfectants were analyzed by indirect immunofluorescence using monoclonal and polyclonal antibodies raised in specifically selected rat strain combinations. These studies showed that pARI.5 encodes the Pa antigen and that the Pa molecule is distinct from the Aa molecule. Probes were derived from the pARI.5 clone and used to study the genomic DNA from Pa-positive and Pa-negative strains. Two probes derived from the 3' untranslated region (3'apARI.5 and 3'bpARI.5) and one probe derived from the 5' region (5'pARI.5) hybridized nonspecifically in all strains under moderate stringency conditions. By contrast, an Xba I restriction fragment unique to the Pa gene was detected with the 5'pARI.5 probe under high stringency conditions. This probe hybridized with a 1.8 kilobase (kb) fragment in the Pa-positive strains and with a 1.7 kb band in the Pa-negative strains. These studies suggest that the gene encoding the Pa antigen, or a fragment thereof, is present in both Pa-positive and Pa-negative strains but may not be expressed in the latter.

Major histocompatibility complex class I antigens expressed on rat trophoblast cells

There is controversy about the size of the major histocompatibility complex antigens of trophoblast cells from placenta. There are some reports that the heavy chains of these molecules are smaller (39-43 kd) than those of the classical class I antigens (45-46 kd), while there are others which show that both the light and the heavy forms of class I antigens occur in the trophoblast cells. In order to investigate this problem, we studied the classical class I antigen (RT1.Aa) and the pregnancy-associated class I antigen (Pa) of the rat from 125I-labeled basal trophoblast cells, isolated from the placenta of WF females pregnant by DA males, using very mild conditions. These antigens were compared with those of the syngeneic (DA x DA) trophoblast cells or paternal (DA) lymphocytes by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Both the Aa and Pa antigens, precipitated from the two trophoblast preparations, showed a heavy chain of 46 kd associated with a 12 kd beta 2-microglobulin component, as did the same molecules precipitated from the lymphocytes. Heavy chains in the range of 39-43 kd could not be detected in any of the samples. The results suggest that the smaller molecular weight heavy chains are methodological artifacts and could arise from loss of a glycan(s) during isolation.

Altered placental morphology associated with murine trisomy 16 and murine trisomy 19

The morphology of placentas from trisomy 16 and trisomy 19 mouse conceptuses aged 12 to 18 gestational days was studied at the light microscopic level. Comparisons were made with placentas from normal littermate animals. Trisomy 16 placentas showed marked changes from normal: 1) the junctional zone showed little indication of normal morphologic differentiation throughout gestation; 2) clusters of germinal trophoblast cells persisted in the labyrinth throughout gestation, whereas these cells disappeared by gestational day 16 in the normal littermate placentas; 3) the labyrinth was reduced in size in the trisomic placentas, and the differentiation of the interhemal membranes was delayed. The size of the labyrinths from trisomy 19 placentas appeared to be decreased, but otherwise the placentas appeared to have normal morphology. These observations and others from the literature show that placental development is affected by the presence of a trisomic genome, and that different trisomies influence the development of the placenta differently. For trisomy 16, we propose that the striking changes of the junctional zone may be associated with the trisomy 16-related gene dosage effect for alpha- and beta-interferon cell surface receptors. Because of the homology for this and other genes on mouse chromosome 16 with genes on human chromosome 21, findings related to the altered development of the trisomy 16 mouse may be relevant to understanding some of the phenotypic variations associated with human trisomy 21, the Down syndrome.

Expression of histocompatibility antigens, transferrin receptors, intermediate filaments, and alkaline phosphatase by in vitro cultured rat placental cells and rat placental cells in situ

Rat placental cells (RPCs) derived from the chorioallantoic placenta of day-12 Holtzman rats were tested for the expression of class I and class II RT I histocompatibility antigens, transferrin receptors, intermediate filaments, and alkaline phosphatase. The binding of mouse monoclonal antibodies to those antigens by RPCs was compared with the binding of the same reagents to rat placental cells in situ. RPCs expressed low levels of class I antigens and failed to express detectable levels of class II antigens. RPCs resisted up-regulation of expression of class I antigens by interferon-gamma, and did not express class II antigens following exposure to medium containing interferon. Transferrin receptors; cytokeratin intermediate filaments, and alkaline phosphatase were universally expressed by RPCs. Taken together with the patterns of expression of the same antigens by rat placental cells in situ, the results suggest that RPCs comprise labyrinthine trophoblast cells. Those cells may provide a valuable new approach for studying the structures and functions of trophoblast cells in vitro.

Electron microscopic localization of the Pa and RT1.Aa antigens on the placenta of the rat

A major factor in the ability of the placenta to avoid allograft rejection is the differential expression of MHC class I antigens on its surface. Using monoclonal antibodies and the electron microscopic immunogold technique, we have demonstrated that only the pregnancy-associated (Pa) antigen, which carries a broadly shared antigenic determinant, is expressed on the placental surface in the rat, whereas the allele-specific classical transplantation antigens are not. Both types of antigens are, however, present in the cytoplasm of the basal trophoblast but completely absent from the labyrinthine trophoblast.

Major histocompatibility antigens on trophoblast and their regulation: implications in the maternal-fetal relationship

Recent technological advances have provided methods of detecting antigens encoded by the major histocompatibility complex with greater precision, allowing the expression of such antigens on the components of the placenta to be clarified. Of specific interest is the expression of these antigens on trophoblast cells, the fetal-derived epithelial cells that confront maternal blood and tissues at the maternal-fetal interface. It is now clear that the different trophoblast subpopulations differentially express class I antigens, although none appear to express class II antigens. Class I antigens can be induced by exposure to interferons on some populations but apparently not others, suggesting that the regulation of their expression differs for subpopulations of trophoblast cells, depending on gestational stage and location. This restricted expression has important implications for maternal-fetal immune interactions during the different phases of pregnancy and perhaps also bears on physiological functions of the feto-placental unit, such as growth and differentiation.

Class I major histocompatibility complex antigen expression on early murine trophoblast and its induction by lymphokines in vitro. II. The role of gamma interferon in the responses of primary and secondary giant cells

The induction of paternal Class I and II MHC antigens by crude lymphokine preparations or purified recombinant gamma interferon was investigated on (C57BL/6J X CBA/H)F1 primary and secondary trophoblast giant cell outgrowths from 3.5-day post-coital (pc) blastocyst and 7.5-day pc ectoplacental cone preparations, respectively, using sensitive immunogold labelling techniques and electron microscopy. Class I MHC (but not Class II) antigens could readily be induced on secondary trophoblast giant cells, by incubation in vitro with gamma interferon for 40 h. However, repeated attempts to induce detectable MHC antigens on primary trophoblast giant cells failed. Mock-treated (C57BL/6J X CBA/H)F1 secondary trophoblast giant cell control preparations failed to express detectable MHC antigens. These findings suggest that, at the time of implantation, there is a time window during which MHC antigens are neither expressed constitutively nor are inducible by soluble factors which normally modulate cell surface MHC antigen concentration.

Placental immunoregulatory factors

Local immune suppression appears to be a key feature in the success of the fetal allograft. Various placental factors have been described endowed with immunoregulatory activity. Much of the available evidence concerning the biological function and biochemical characteristics of these molecules is reviewed here, as well as the controversial issues surrounding them. In the final section, some outstanding questions relating to these immunoregulators are considered.