The involvement of interleukin (IL)-15 in regulating the differentiation of granulated metrial gland cells in mouse pregnant uterus.

Previous studies have suggested that granulated metrial gland (GMG) cells are bone marrow-derived lymphoid cells, which differentiate in situ in the mouse pregnant uterus into natural killer (NK)-like cells. Similar to NK cells, GMG cells express an abundant level of cytolytic mediators such as perforin. The factor(s) regulating the differentiation of GMG cells remain(s) to be identified, although cytokines previously implicated in the stimulation/activation of NK cells (e.g., IL-2, IL-6, IL-7, and IL-12) can be considered as potential candidates. Recently, IL-15, a novel cytokine, which displays biological activities similar to IL-2, has also been shown to be capable of activating NK cells. Using reverse transcription-polymerase chain reaction (RT-PCR) analysis, we have demonstrated in the present study that IL-15 and its cognate receptor, but not the other cytokines, are expressed in the mouse pregnant uterus, with a time course concomitant with those of cytolytic mediators in differentiated GMG cells. Moreover, IL-15, though not IL-2, is capable of inducing the expression of perforin and granzymes in pregnant uterine tissues explanted in vitro. Data obtained from in situ hybridization study have suggested that the macrophages present in the pregnant uterus may be responsible for the production of IL-15. These results suggest that IL-15 is involved in regulating the differentiation of GMG cells during mouse pregnancy.

Uterine natural killer cells in perforin and beta(2)-microglobulin deficient mice.

Uterine natural killer (uNK) cells have roles for immune responses at the feto-maternal interface in mice. We studied the effects of beta(2)-microglobulin (beta(2)m) and perforin on proliferation and differentiation of uNK cells in pregnancy, using beta(2)-microglobulin-deficient (beta(2)m(-/-)) mice and perforin-deficient (P(-/-)) mice. The cell population of uNK cells in the metrial gland of P(-/-) mice was tended to be higher than the control B6 mice. The cell population of uNK cells in the metrial gland of beta(2)m(-/-) mice was significantly increased at Day 12 of pregnancy comparing to B6 and P(-/-) mice. On the other hand, the cell population of uNK cells in the decidua basalis of beta(2)m(-/-) mice was tended to be lower than B6 and P(-/-) mice. These results indicate that beta(2)m may be involved in proliferation of uNK cells in the metrial gland, and that beta(2)m may affect the maturation of uNK cells in the decidua basalis.

[Immunosuppressor activity of rat endometrial granulated cells and their differentiation].

Natural killer and natural suppressor activities of the rat endometrial granulated cells were assayed on day 13 of pregnancy or pseudopregnancy. Metrial gland granulated cells were used as endometrial granulated cells. The natural killer activities of metrial gland granulated cells and other cells were determined by means of Hashimoto-Sudo test with K562 cells as targets. The estimation of natural killer activity included removal of the cells sticking to glass from a suspension of material gland granulated cells. Cytochemically, metrial gland granulated cells were identified by the presence of PAS-positive granules in the cytoplasm after treatment of the cells with diastase and identification of a specific antigen with the help of specific antisera. The natural killer activity of metrial gland granulated cells was twice weaker than that of splenocytes from the same pregnant or pseudopregnant females. The level of natural killer activity was proportional to the content of metrial gland granulated cells in a cell system. These data suggest that the natural killer activity of metrial gland granulated cells is realized via their contact with cell targets. Natural killer and suppressor activities were determined simultaneously for metrial gland granulated cells and splenocytes of the same rat with common cell targets. When estimating the nuclear suppressor activity of metrial gland granulated cells, the splenocytes of the same rat were used as an effector in a natural killer test. Various amounts of metrial gland granulated cells were added to the effector : target system at a ratio of 50:1. The natural suppressor activity of metrial gland granulated cells did not depend on the amount of metrial gland granulated cells present in a natural killer system. After fractionation in a Percoll gradient, the highest natural killer activity was recorded in a 30% Percoll fraction. The highest and lowest natural suppressor activities were recorded in 30% and 60% Percoll fractions, respectively. The culture medium was characterized by natural suppressor activity as well. The differences in mean areas of metrial gland granulated cells in 30 and 60% Percoll fractions between the pregnant (144.7 +/- 13.4 and 75.0 +/- 12.5 microm2, respectively) and pseudopregnant (97.5 +/- 4.9 and 69.2 +/- 3.5 microm2) females were reliable. The natural killer activity was estimated in all studied 23 samples of metrial gland granulated cells, among which 18 (79.6 +/- 7.8%) displayed the natural suppressor activity as well. The absence of natural suppressor activity in five samples was combined with the absence of this activity in their culture medium and with a reduction in the mean area of metrial gland granulated cells in 30% Percoll fraction to 109.1 +/- 5.2 microm2. The data obtained confirm the known data on a low activity of metrial gland granulated cells and demonstrated for the first time the natural suppressor activity of these cells. It was concluded that the natural suppressor activity of metrial gland granulated cells is due to their differentiation from metrial gland granulated cells with natural killer activity.

Monoclonal antibodies directed against surface and intracellular antigens of mouse granulated metrial gland cells.

Rat monoclonal antibodies with reactivity directed against mouse GMG cells have been produced. One of the antibodies (GMG-1) reacts with a surface antigen of GMG cells and cross-reacts with T lymphocytes. Another (GMG-2) reacts with an intracellular antigen in GMG cells and with asialo-GM1 positive cells in the spleen. Three antibodies (GMG-3, -4, -5) bind to intracellular antigens in GMG cells. The cross-reactivity of these antibodies is discussed with reference to the lineage relationship of GMG cells to NK cells and T cells and the recent suggestion that NK cells and T cells have a common progenitor cell. It is proposed that GMG cells share this common progenitor cell but are otherwise independent of the NK or T cell lineages.

A unique antigen, RT11, in the placenta of the rat.

In the course of exploring the antibody response in the unsensitized WF (u) female pregnant by a DA (a) male, we prepared a hybridoma that secreted an antibody (mAb 213) that was specific to the a haplotype but identified an antigen different from Pa. This antigen was designated RT11. It is present from the twelfth day of gestation on the collagen fibers of the placenta and of all organs in fetal and adult rats. It is particularly prominent on red blood cells; in the yolk sac epithelium; in the walls of the endodermal sinus, blood vessels and bronchioles; and in capsules and trabeculae. A very small amount is present on DA lymphocytes, since 17-20% of them react with mAb 213 by cytofluorimetry. The RT11 antigen is absent from the basal and labyrinthine trophoblast cells, from the parenchymal cells of all organs, and from T and B cells. This distribution pattern is completely different from that of the Aa and Pa antigens. Inhibition and absorption studies showed that RT11 is not an integral part of the collagen molecule. The SDS-PAGE analysis of the immunoprecipitates of RT11 from radioiodinated whole-membrane extracts of red blood cells and from the glycoprotein fraction thereof showed that it is an unglycosylated protein of molecular weight 29,000. The evidence to date suggests that RT11 is a blood group antigen. Studies on the genetic control of the expression of RT11 were undertaken to determine whether a gene linked to the MHC was involved and whether the control mechanism was unigenic or polygenic. Backcrosses generated using inbred strains--(DA x BN)F1 x DA-- and using complementary congenic strains--(DA.1N x BN.1A) F1 x BN.1A--showed that the expression of RT11 was under polygenic control, and that both an MHC-linked gene (1.2 cM from RT1.Aa) and genes not linked to the MHC are involved. By contrast, the expression of the Pa antigen is under the control of an MHC gene only.

The killing of rat placental cells by rat and mouse granulated metrial gland cells in vitro.

Small round cells which migrated from explant cultures of rat metrial gland were identified as granulated metrial gland (GMG) cells. They contained large amounts of glycoprotein and displayed the leucocyte common antigen. Other cells which migrated from the explants were probably derived from the fibroblast-like stromal cells of the metrial gland. The asialo-GM1 antigen was found on rat GMG cells in culture and in cryostat sections of rat metrial gland. The rat GMG cells in culture exhibited locomotion and, when co-cultured with placental cells, made numerous contacts with the placental cells. A small number of these contacts (less than 1 per cent) were followed rapidly by the death of the placental cell. Mouse GMG cells which had migrated from explant cultures of mouse metrial gland were also co-cultured with rat placental cells. The migratory activity of the mouse GMG cells also involved numerous contacts being made with rat placental cells and a small number (less than 1 per cent) of these contacts were cytotoxic for the rat placental cells. The observations support previous suggestions that GMG cells are a type of killer cell. The cytotoxic activity of rat and mouse GMG cells against co-cultured rat placental cells is discussed in relation to the nature of the target molecule involved.

Granulated metrial gland cells--not part of the natural killer cell lineage?

The relationship of rodent granulated metrial gland (GMG) cells to the natural killer (NK) cell lineage is reviewed. The antigenic profile of GMG cells provides insufficient evidence to indicate that these cells form part of the NK cell lineage. No good evidence is found that GMG cells are cytotoxic to the NK cytotoxicity target, Yac-1. In Beige mice, which have a defect in the lytic pathway of NK cells, GMG cells were found to kill labyrinthine cytotrophoblast. It is concluded that GMG cells are not a type of NK cell but a novel member of the leucocyte population.

Immunoglobulin G in the decidua basalis and metrial gland of the pregnant mouse uterus.

The distribution of immunoglobulin G (IgG) in the decidua basalis and metrial gland of the mouse uterus at days 10 and 14 of pregnancy has been investigated using immunohistochemical methods. IgG was found in the intercellular spaces but none was found in decidual cells, stromal cells of the metrial gland or granulated metrial gland cells. These results differ from those of other studies which have localised IgG in the cytoplasm of rat granulated metrial gland cells.

Granulated metrial gland cells: hypotheses concerning possible functions during murine gestation.

Granulated metrial gland (GMG) cells are morphologically distinctive lymphoid cells found in the murine uterus only during gestation. The life history of GMG cells suggests that they have important roles during mammalian gestation but these roles have been difficult to define. Genetic and immunologic data suggest that GMG cells may be a specialized subset of natural killer (NK) lymphocytes. This has directed research on GMG cell functions towards questions of effector cell-target cell interactions. A broader range of potential functions is discussed and shifts in functional roles played by GMG cells are proposed over the course of gestation.

Localisation of immunoglobulin (IgG) within the rat metrial gland.

An immunohistological method was used to assess the IgG content of the rat metrial gland at different stages of pregnancy. The result apparently varied according to the type of fixative used. Saturated alcoholic mercuric chloride was found to produce the most consistent demonstration, with the IgG located in cells which also contained diastase-fast, PAS-positive granules. Using single cell suspensions prepared from metrial glands, significantly more cells were shown to contain cytoplasmic IgG at day 13, compared to day 14 of pregnancy. However, there were no other significant differences at other stages of pregnancy examined. Surface IgG was detected on a small proportion of the cells and the findings are discussed in relation to the hypothesis of a lymphocytic origin for the granulated metrial gland cells.

Localisation of immunoglobulin (IgG) in granulated metrial gland cells of deciduomata of the pseudopregnant rat.

Using an immunoperoxidase technique, after fixation in saturated alcoholic mercuric chloride, it was possible to detect cytoplasmic immunoglobulin (IgG) in granulated metrial gland cells from deciduomata of pseudopregnancy in the rat. The extent of the reaction for IgG was variable but did not appear to be related to the day of pseudopregnancy or to the extent of the decidual reaction. Examination of single cell suspension enabled quantification of IgG-containing cells, but no significant differences were detected in the numbers of positive cells at the days of pseudopregnancy which were examined. Surface IgG was detected on a small proportion of the cells, and they were distinguished from the Fcgamma receptor-bearing cells in the metrial gland of deciduomata of pseudopregnancy.

Immunohistochemical analysis of beta 1-integrin receptors displayed by murine uterine natural killer cells over the course of successful pregnancy.

Granulated metrial gland (GMG) cells are a natural killer (NK) cell-like population present in large numbers in the pregnant rodent uterus. By day 8 of gestation GMG cells are large and granulated and localized to the mesometrial side of each implantation site. GMG cells appear to be highly migratory both in vivo and in vitro; however, little is known regarding their functions. Using indirect fluorescence immunohistochemistry, murine uteri and implantation sites were studied on successive days of gestation to characterize the extracellular matrix receptors of the VLA-integrin family displayed by GMG cells. On days 3 and 6 of gestation, double immunostaining using the monoclonal antibody LGL-1 was employed to recognize GMG cells because their morphology early in pregnancy resembles that of other lymphocytes. Between days 8-15 of gestation, GMG cells can be recognized by their unique morphology. The day 3 and day 6 LGL-1+ cells were positive for all antigens examined; that is, beta 1 plus alpha 1, alpha 3, alpha 4, alpha 5 and alpha 6. From days 8-15 of gestation, GMG cells were beta 1+, alpha 4+, alpha 5+ but alpha 1-, alpha 3-, alpha 6-. Thus, between days 6-8 of gestation, major changes occur in the uterine NK/GMG cell population which include the loss of the surface molecules VLA alpha 1, alpha 3 and alpha 6 or the rapid expansion of NK cells not expressing these proteins. It is postulated that major changes in the functions of uterine NK cells are likely to be associated with these alterations in cell surface phenotype and that functional studies of uterine NK cells should focus upon this relatively early gestational time point.

Immunological characterisation of surface receptors on rat metrial gland cells.

Cells from metrial glands of pregnant rats were examined for surface receptors. No E, EA mu or EAC receptors were demonstrated. Fc gamma receptors were detected by EA gamma rosette formation, using sheep red blood cells sensitised with the IgG fraction of rabbit or rat antisera. Significantly more of the metrial gland cells, and of the rat peritoneal exudate and spleen cells examined as controls, formed rosettes with red cells sensitised with rabbit IgG than with those sensitised with rat IgG. The proportion of metrial gland cells forming EA gamma rosettes decreased significantly between day 12 and day 15 of pregnancy but increased by day 19. Metrial gland cells from deciduomata formed EA gamma rosettes, and the proportions varied during pseudopregnancy. At day 13 of pregnancy a greater proportion of metrial gland cells displayed Fc gamma receptors in multiparous rats than in primigravid rats. The binding affinity of the Fc gamma receptors was characterised by inhibition studies with homologous and heterologous IgGs. Maximal inhibition occurred when the inhibitory IgG was homologous to the IgG used to sensitise the red cells. EA gamma rosette formation by cells from the metrial gland was inhibited by both monomeric and heat-aggregated IgGs.

Receptors for IgG2b on cells of the mouse metrial gland.

Single cells prepared from metrial glands of mice killed at days 10, 13 and 17 of pregnancy were assayed for the expression of Fc gamma receptors in a standard rosetting assay using sheep red blood cells sensitised with a mouse monoclonal IgG2b antibody. Rosettes, indicating Fc gamma receptors, were found on both granulated metrial gland (GMG) cells and non-GMG cells, comprising mainly stromal cells, from each stage of pregnancy. Some animals were given an intravenous injection of horseradish peroxidase 2 h before they were killed in order to identify endocytic cells. No GMG cells were found to have endocytosed the horseradish peroxidase. Non-GMG cells which showed endocytic activity all expressed Fc gamma receptors but these receptors were also found on some of the non-GMG cells which had not exhibited endocytosis. The finding of Fc gamma receptors on GMG cells provides further evidence that these cells may be related to NK cells.

Evaluation of the murine metrial gland for immunological function.

The metrial gland (MG) is a transient uterine structure associated with rodent pregnancy. The gland is a complex structure consisting of stromal and vascular elements, as well as a population of histologically distinctive, large, granulated metrial gland (GMG) cells. The functions of the MG and of the GMG cells, as well as their relationship to the success of pregnancy, are unknown. Based upon morphological and morphometric studies it has been proposed that the MG might be involved in the immunology of pregnancy and that GMG cells could be immunocompetent. Explant cultures of MG have therefore been evaluated for immunological function. Lytic activity against the NK sensitive target cell line YAC and mitogen responsiveness could not be detected. MG tissue and medium conditioned by overnight culture of MG tissue (MG-CM) suppressed the response of murine spleen cells to Con A. MG-CM also reduced the lytic activity of splenic NK cells against YAC target cells. However, uptake of [3H]thymidine was elevated when YAC cells were cultured in MG-CM. The response of embryonic and uterine cells to growth in MG-CM was complex. MG-CM had little effect on isotope incorporation by decidual cells recovered at 6.5 days or by embryonic cells recovered from 12.5 day embryos. However, thymidine incorporation was less in MG-CM than in control medium for 12.5 day placental cells, 6.5 day embryonic sac, 6.5 day ectoplacental cone and 3.5 day blastocysts. Cytotoxicity and cytostasis accounted for reduced uptake of isotope in cultures of 3.5 day blastocysts and 6.5 day embryonic tissues. Loss of viability could not be detected in any other assays. Both YAC cells and unstimulated splenocytes showed altered morphology and improved viability when cultured in MG-CM. This study suggests that the only immunological role the MG might have during normal pregnancy is that of non-specific intra-uterine suppression. Alternatively, differential regulation of cell proliferation might be a function of the MG, within the pregnant uterus. The latter mechanism could also account for the apparent observation of non-specific immunosuppression.

Mouse metrial gland cells do not kill Yac-1 myeloma cells.

Mouse metrial gland cell suspensions, which included granulated metrial gland cells, were assessed for their ability to lyse NK cell target Yac-1 myeloma cells in a 51chromium release cytotoxicity assay. Metrial gland cells did not kill Yac-1 cells even after in vivo stimulation of NK cytotoxicity activity by polyinosilic-cytidilic acid. The precise relationship of granulated metrial gland cells to the NK cell lineage remains to be clarified.

Leucocyte membrane antigens on mouse granulated metrial gland cells.

An indirect immunofluorescence technique has been used to study the expression of leucocyte membrane antigens on mouse granulated metrial gland (GMG) cells. GMG cells isolated from cultured metrial gland explants and GMG cells in cryostat sections of uterine implantation sites were examined. GMG cells were found to express both the asialo-GM1 antigen and the Thy-1 antigen. GMG cells did not express the Lyt-1, Lyt-2, Mac-1, L3T4 or IgM antigens. These results provide new evidence that GMG cells are a type of NK cell.

Characterization of the cells that migrate from metrial glands of the pregnant mouse uterus during explant culture.

Granulated metrial gland (GMG) cells are estrogen-receptor and Interleukin 2 (IL-2) receptor positive lymphocytes of the Natural Killer cell lineage found in the murine uterus during pregnancy. Functional studies of these cells, which are now more frequently called uterine NK (uNK) cells, have been limited due to technical difficulties. The cells are difficult to isolate and their proliferation and differentiation have not been achieved in culture. In 1988, Mukhtar and Stewart (Cell Tiss. Res., 253, 413-417) reported a method for explant culture of metrial glands isolated from pregnant rodents that yielded an almost pure population of uNK cells. This major technical advance has supported most of the subsequent functional and molecular studies of rodent uNK cells. However, the quality of the cells isolated by the explant culture procedure has not been established. A cytochemical approach was used to identify and quantify the cells migrating from metrial glands. At midpregnancy, almost all (> 90%) migrating nucleated cells were NK cells. Earlier in gestation, a significant proportion (25%) of cells having lymphoid morphology could not be assigned to the lineage. The viability of cells migrating from explants was assessed by DNA isolation and electrophoresis on days 6-16 of gestation. At all times evidence for apoptosis was found, even after culture intervals as brief as 4 h. Parallel analyses of histological sections of the metrial gland, using terminal deoxytransferase labelling to detect nuclear fragmentation, did not support significant levels of uNK cell death in situ prior to day 12 of gestation. Supplementation of the explant culture medium with estrogen, IL-2, various extracellular matrices, decidual cells or combinations of these did not lead to in vitro proliferation of uNK cells and usually did not extend the short term viability of these cells in serum supplemented or serum free media. Thus, the optimal culture conditions for uNK cells remain undefined.

Preliminary characterization of lymphoid hybridoma cell lines derived from the pregnant mouse uterus.

Seven independent cell lines were derived from the fusion of migratory cells recovered from explant cultures of metrial glands to SP 2/0, a non-Ig secreting B cell myeloma. The migrating cells came from a pool of metrial glands from day 6-8 pregnant random bred CD1 mice and were assumed to be cells early in the differentiation pathway to granulated metrial gland (GMG) cells. The fused cells were cloned twice at the limiting dilution. Hybridization was confirmed by quantitation of cellular DNA using propidium iodide staining and by karyotyping. Electron microscopy revealed that each of the hybrid cell lines was composed of cells which were lymphoid in appearance, but lacked the granules found in mature GMG cells. The surface phenotype of all lines is CD45+, LGL-1-, asialo GM-1-, IgG-, IgM-, CD3- and CD25- (p55 of IL-2 receptor). Although the hybridomas lack those phenotypic markers which were used to show that GMG cells are related to the natural killer (NK) cell lineage (ie LGL-1, asialo GM-1), they do express the pan-leukocyte marker CD45 as well as the lytic protein, perforin, at levels intermediate to those of SP 2/0 cells and GMG cells. In addition, the hybridomas were observed to preferentially bind the NK target cell YAC and to be capable of lytic activity at temperatures below 30 degrees C. Because these hybridomas may represent fusion to an early progenitor cell of the NK/GMG cell lineage, their continued characterization is of merit.