In-vitro development and degeneration of stromal cells from the uterus of the rat at three stages of pregnancy

On Days 6, 11 or 16 of pregnancy, endometrial tissue (Day 6) or decidual tissue (Days 11 and 16) were removed from rat uteri, dissociated into single cells and grown in culture. At intervals during the culture period, the cells were examined and the cell density was calculated. The cells from Days 11 and 16 of pregnancy were similar in appearance, being very large and flattened. Initially, cells from Day 6 were much smaller, but showed an increase in size and came to resemble cells from the later stages of pregnancy. Growth curves for cells from each of the three stages of pregnancy over the first 16 days of culture differed. Cells from Day 6 of pregnancy increased exponentially in number. In cultures of cells from Day 11, cell numbers were stable, and in cultures of cells from Day 16, an exponential decline was seen. Cells from Days 6 and 11 of pregnancy were cultured for 42 days, by which time only a few cells remained viable. Cultures of cells from Day 16 degenerated within 3 weeks. These results indicate that, in vivo, decidual cells are not controlled solely by a 'programmed lifespan'. Changes occurring during pregnancy, however, limit the potential of the cells for division and survival in culture.     

Characterization of immune effector cells present in early murine decidua

It has been suggested that murine decidual cells act as an important immunoregulatory population localized to the pregnant uterus. We have examined early murine decidua to determine if immune effector cells occur in the decidual environment in proximity to the conceptus. High levels of natural killer (NK) cell activity were found consistently in decidual cell suspensions with peak activity occurring on Day 6.5 of gestation. NK activity declined as pregnancy proceeded and was not significant by Day 12.5 of gestation. Decidual cell suspensions did not appear to contain significant numbers of functional B or T effector cells. No antipaternal T-cell response could be demonstrated even in the decidua of immune mice. Lack of T-cell responses was attributed to the absence of T cells from decidua rather than to their inactivation because precursors of cytotoxic T lymphocytes (pCTL) could not be detected in decidual cell suspensions. Furthermore, the levels of pCTL detectable in spleen cell suspensions could not be reduced by mixing spleen cells with 7.5-day decidual cells. These results suggest that B cells and T cells may not occur in early decidua while NK cells are present and regulated independently.     

Suppression of lymphocyte alloreactivity by early gestational human decidua. I. Characterization of suppressor cells and suppressor molecules

We examined the immunosuppressor role of the first trimester human decidua on lymphocyte alloreactivity in vitro in order to identify (1) the major cell classes in the decidua mediating the suppressor effect; (2) the stages in the lymphocyte alloreactive responses susceptible to the suppressor effects of the decidua; and (3) the precise nature of the suppressor molecules. Irradiated (2800 R), Ficoll-Paque-separated nucleated cells of the collagenase-dispersed early gestational (6.5-9.5 weeks menstrual age) decidua containing 70-94% typical decidual cells (identified on the basis of distinctive morphology and numerous cytoplasmic or surface markers) or their plastic-nonadherent fractions further enriched for decidual cells (approximately 96% pure) caused a strong dose-dependent suppression of the one way mixed lymphocyte reaction (MLR, i.e., proliferative response measured on Days 3, 4, or 5), when added at the onset of the mixed lymphocyte cultures (MLC). As few as 10(3) decidual cells caused a detectable inhibition of the MLR exhibited by 10(5)-1.5 X 10(5) responder lymphocytes. A smaller degree of suppression was noted with the plastic-adherent fractions of the early decidua (which retained all macrophages and granulocytes, but still included many decidual cells) or unfractionated cells of later gestational (10-13 weeks) decidua containing a higher incidence of leukocytes, granulocytes, and macrophages in particular, or the plastic-adherent fraction thereof, enriched for macrophages. Thus, decidual cells seem to represent an important suppressor cell class in the early gestational human decidua; however, suppression by decidual leukocytes, macrophages in particular, was also evident. The suppressor effect was unrelated to the major histocompatibility phenotype of the responder or the stimulator cells. It was not caused by cell crowding, since an equivalent number of irradiated K562 erythroleukemia cells had little effect on the MLR. The effect was exerted during both the initiation and the progression of the MLR. A delay in the addition of regulator cells progressively minimized the effect on the Day 4 MLR, but did not abolish it completely even when added as late as on Day 3. The major class of mediator molecules was identified as prostaglandins, primarily PGE2, on the basis of the following results: (1) the presence of indomethacin (10(-5) M) or varying dilutions of an anti-PGE2 antibody abrogated this suppression substantially or completely. (2) Addition of pure PGE2 (3 X 10(-7) to 1.1 X 10(-5) M), but not PGF2 alpha, reproduced a dose-dependent suppressor effect.(ABSTRACT TRUNCATED AT 400 WORDS)     

Collagen concentrations in dissected tissue compartments of rat uterus on days 6, 7 and 8 of pregnancy.

Implantation and non-implantation sites were dissected into myometrial and stromal components; a decidual/embryonic region was obtained on Days 7 and 8 of pregnancy. The concentration of collagen (as a percentage of the dry weight of tissue), measured by hydroxyproline analysis, was significantly lower in the implantation regions than in the non-implant regions in all areas studied. The concentrations in the antimesometrial myometrium and stroma of the implantation region remained the same over the days studied. In contrast, the mesometrial collagen concentration in the implantation region declined from Day 6 to Day 8 of pregnancy. Collagen concentration was low within the decidual/embryonic tissue on Days 7 and 8 of pregnancy. Remodelling of collagen within the embryonic area appears to be an important feature of the uterine response to implantation in rats.     

Swine leukocyte antigen-DQ expression and its regulation by interferon-gamma at the maternal-fetal interface in pigs

Successful pregnancy requires an appropriate intrauterine immune response to the conceptus, which is a semiallograft within the uterus. We reported that swine leukocyte antigen-DQA (SLA-DQA), a major histocompatibility complex (MHC) class II gene, is expressed in the uterine endometrium at the time of conceptus implantation in pigs. Because MHC molecules play critical roles in the immune system, SLA-DQ was hypothesized to be involved in immune regulation during pregnancy. Therefore, we examined expression of SLA-DQ in uterine endometrial tissues obtained during the estrous cycle and pregnancy. SLA-DQA and SLA-DQB mRNAs were detected as 1.3-kb and 1.2-kb bands, respectively. Real-time RT-PCR analysis indicated that SLA-DQA and SLA-DQB mRNA expression was affected by day and pregnancy status, with the highest expression on Day 15 of pregnancy. SLA-DQ was localized primarily to subepithelial stromal cells and endothelial cells of the uterus. Using endometrial explant cultures from Day 12 of the estrous cycle, we determined that expression of SLA-DQA and SLA-DQB mRNAs increased in response to interferon-gamma (IFNG), which is produced by pig conceptus trophectoderm between Days 14 and 18 of pregnancy. The abundance of SLA-DQ protein was less in endometria from gilts with conceptuses resulting from somatic cell nuclear transfer compared with endometria from gilts with conceptuses resulting from natural mating. These results support our hypothesis that SLA-DQ is expressed in response to IFNG from the conceptus, and likely regulates immune response at the maternal-fetal interface to support the maintenance of pregnancy in pigs.     

Evidence for a selective migration of fetus-specific CD4+CD25bright regulatory T cells from the peripheral blood to the decidua in human pregnancy

During pregnancy, the maternal immune system has to tolerate the persistence of fetal alloantigens. Many mechanisms contribute to the prevention of a destructive immune response mediated by maternal alloreactive lymphocytes directed against the allogeneic fetus. Murine studies suggest that CD4(+)CD25(+) T cells provide mechanisms of specific immune tolerance to fetal alloantigens during pregnancy. Previous studies by our group demonstrate that a significantly higher percentage of activated T cells and CD4(+)CD25(bright) T cells are present in decidual tissue in comparison with maternal peripheral blood in human pregnancy. In this study, we examined the phenotypic and functional properties of CD4(+)CD25(bright) T cells derived from maternal peripheral blood and decidual tissue. Depletion of CD4(+)CD25(bright) T cells from maternal peripheral blood demonstrates regulation to third party umbilical cord blood cells comparable to nonpregnant controls, whereas the suppressive capacity to umbilical cord blood cells of her own child is absent. Furthermore, maternal peripheral blood shows a reduced percentage of CD4(+)CD25(bright)FOXP3(+) and CD4(+)CD25(bright)HLA-DR(+) cells compared with peripheral blood of nonpregnant controls. In contrast, decidual lymphocyte isolates contain high percentages of CD4(+)CD25(bright) T cells with a regulatory phenotype that is able to down-regulate fetus-specific and fetus-nonspecific immune responses. These data suggest a preferential recruitment of fetus-specific regulatory T cells from maternal peripheral blood to the fetal-maternal interface, where they may contribute to the local regulation of fetus-specific responses.     

Influence of the endometrium, protease inhibitors and freezing on antiviral activity of proteins secreted by pig conceptuses

In Exp. 1, only medium from cultures containing conceptus tissue had antiviral activity (P less than 0.05). Addition of Day-15 pregnant endometrium or Day-14 cyclic uterine flush proteins to cultures containing 200 mg conceptus tissue decreased antiviral activity (conceptus x endometrial protein interaction, P less than 0.06). Effects of endometrium (-54%) and uterine flush proteins (-40%) on antiviral activity of conceptus cultures did not differ from each other (P greater than 0.10). In Exp. 2, antiviral activity was only detected in cultures containing conceptus tissue (P less than 0.06). The amount of antiviral activity in cultures of Day-15 conceptus tissue was not influenced differently (P greater than 0.10) by culture in medium conditioned by endometrium from Day 10 or Day 12 of pregnancy. However, antiviral activity was undetectable in medium conditioned by endometrium from one of the Day-12 gilts. In Exp. 3, antiviral activity was present in medium from only 1 of 3 cultures from Day-12 gilts when assayed unfrozen. Antiviral activity was lower (P less than 0.01) in cultures of conceptuses from Day 12 than Day 14 of pregnancy; however, antiviral activity increased quadratically (P less than 0.05) when cultures contained 0, 0.01, 0.1 and 1.0 units/ml aprotinin, respectively. Freezing and thawing culture medium did not reduce (P greater than 0.10) antiviral activity compared to medium assayed unfrozen (1438 vs 1354 units/ml, respectively). These results suggest a regulatory influence of the endometrium on secretion of antiviral proteins by pig conceptuses in vitro.     

Relationship between luteinizing hormone and decidual luteotropin in the maintenance of luteal steroidogenesis

Between Days 6-11 of pregnancy or pseudopregnancy, the decidual tissue of the rat produces a prolactin-like hormone, decidual luteotropin, which can sustain luteal progesterone production when prolactin is suppressed. However, this effect is dependent upon the presence of the pituitary. The present investigation was undertaken to determine whether decidual luteotropin and luteinizing hormone (LH) act together to sustain luteal steroidogenesis and if so, to find out whether the need for LH is due to the inability of the decidual tissue to produce LH-like material and/or whether LH affects decidual luteotropin production. Pseudopregnant rats with or without decidual tissue were hypophysectomized on Day 8 and treated with either 1.5 IU human chorionic gonadotropin (hCG)/day or with vehicle. Within 24 h, serum progesterone dropped in both vehicle-treated groups and decidual luteotropin levels declined by 80% in the decidual tissue. Human CG administration had no effect on progesterone production in the control group. Yet in rats with decidual tissue, hCG stimulated progesterone production for at least 48 h and maintained the decidual tissue content of decidual luteotropin. Progesterone, but not hCG treatment, maintained decidual luteotropin concentrations in ovariectomized rats. To compare the luteotropic activity of the decidual tissue with that of the placenta, pregnant or pseudopregnant rats with decidual tissue were hypophysectomized on Day 8 and treated with 1.5 IU hCG. Control groups had decidual tissue or placentas removed and were similarly treated. Human CG stimulated progesterone production only in rats with placental or decidual tissue.(ABSTRACT TRUNCATED AT 250 WORDS)     

In vitro prostaglandin release from and platelet-activating factor accumulation in isolated endometrial cells from pregnant and pseudopregnant rabbits.

Prostaglandin (PG) release from and platelet-activating factor (PAF) accumulation by enzymatically isolated endometrial epithelial and stromal cells from Day 6 pregnant and Day 6 pseudopregnant rabbits were studied in vitro, using RIA for PG measurement and a platelet aggregation assay for PAF measurement. On the first day of culture in serum-free media, PGF release into the medium was significantly higher from epithelial cells from Day 6 of pregnancy than from stromal cells from Day 6 of pregnancy or pseudopregnancy. PGE release did not differ significantly among these cell types. The addition of indomethacin (10(-5) M) to similar cultures inhibited release of both PGs from both cell types, but to a much greater extent from stromal than from epithelial cells. Significant stimulation of PG release by A23187 was achieved under all conditions on the fifth day of culture; PGE release was significantly greater than PGF release from stromal cells from Day 6 of pregnancy and pseudopregnancy, and release of both PGs from stromal cells was significantly greater from Day 6 of pregnancy than from Day 6 of pseudopregnancy. PG release from similar cells, cultured in medium containing 10% calf serum, was highest on the first or second day of culture and then, especially for PGF, declined with continued culture. PGE release was significantly higher than PGF release from stromal cells on the third and fourth days of culture. The ratios of PGF/PGE release from epithelial cells were significantly higher than those from stromal cells over the 5-day culture period for both reproductive stages. These ratios indicate the differential release of PGE and PGF from rabbit endometrial cell subpopulations and indicate a preferential release of PGE from stromal and of PGF from epithelial cells. Under basal conditions, PAF was not detected in epithelial or stromal cells cultured for 2 or 4 days, or in the associated culture media. If PAF had been released into the medium, it would have rapidly metabolized. Short exposure to calcium ionophore A23187 (10(-5) M) was able to stimulate PAF accumulation in epithelial and stroma cells in serum-free media, probably via the remodeling pathway. PAF was not detected in the medium. Intracellular PAF accumulation after exposure to A23187 (10(-5) M) for 5 min was significantly greater on the second day of culture than on the fourth day in epithelial and stromal cells from Day 6 of pregnancy.(ABSTRACT TRUNCATED AT 400 WORDS)     

Preeclamptic women are deficient of interleukin-10 as assessed by cytokine release of trophoblast cells in vitro

BACKGROUND: It is well known that the acceptance of the fetoplacental unit in human pregnancy requires maternal immune tolerance, which is thought to be regulated locally by the placenta. Therefore an anti-inflammatory cytokine such as IL-10 plays a critical role in different pregnancy disorders including preeclampsia. In the present study, we examined the expression of both proinflammatory (TNF-alpha, IL-1beta, IL-2) and immunoregulatory (IL-6, IL-10) cytokines from normal term and preeclamptic patients in human trophoblast cultures. METHODS: Eleven patients with preeclampsia and 11 patients with a normal pregnancy at term were included in the study. Trophoblast cells isolated from placentas were cultured up to 48 h under standard tissue culture conditions and cytokine release was determined by ELISA. IL-10 synthesis was significantly decreased in the third trimester in preeclamptic patients in comparison with the control group. RESULTS: There were no significant differences in IL-1beta, IL-2, IL-6 or TNF-alpha expression but a significant alteration in IL-10 release in trophoblast cultures in vitro in term placentas from preeclamptic patients compared with normal pregnancy. CONCLUSIONS: Because IL-10 is a potent regulator of anti-inflammatory immune response these abnormalities may be associated with the inadequate placental development in preeclampsia.     

Trophoblast regulation of maternal-paternal lymphocyte interactions

The modification of immunological responses by murine embryonic trophoblast cells was investigated using the mixed-lymphocyte reaction (MLR) and cell-mediated lympholysis (CML) test. In MLR containing Balb/c (responder) and C57BL/6 (stimulator) splenocytes DNA synthesis was markedly reduced in the presence of ectoplacental cone or placental trophoblast cells. These same trophoblast cell populations inhibited the in vitro generation of cytotoxic lymphocytes while cultures containing nontrophoblast regulator cells expressed normal cytotoxicity. DNA synthesis in MLR and cytotoxic activity in CML were not suppressed in cultures containing $\text{3}\text{1}\text{2}$ day blastocyst outgrowths. The potent immunosuppressive properties of the trophoblast may be important during pregnancy by protecting the genetically dissimilar fetus from potentially harmful maternal immune effector mechanisms.     

Differential temporal and spatial expression of mRNA encoding extracellular matrix components in decidua during the peri-implantation period.

Changes in the temporal and spatial patterns of expression of mRNA encoding uterine extracellular matrix (ECM) proteins were determined during the peri-implantation period. Northern blot hybridization of cDNAs corresponding to laminin (LM) B1, LM B2, entactin, fibronectin, collagen (CL) type IV alpha 1, and CL IV alpha 2 was performed on RNA extracted from either whole mouse uteri or endometrial explants between Day 4, i.e., the day of implantation, and Day 7 of pregnancy, when the decidual response is well established. These analyses revealed a dramatic increase in LM B2, CL IV alpha 1, and CL IV alpha 2 mRNA expression by Day 7 of pregnancy. Relative levels of the mRNA encoding other ECM components, including LM B1, were not altered when compared to changes in the relative level of expression of glyceraldehyde-3-phosphate dehydrogenase mRNA. The differential expression of the B chains of LM appeared to be limited to the stromal cells of the endometrium. In situ hybridization of uterine sections with cRNA probes corresponding to LM B1, LM B2, and CL IV alpha 1 demonstrated that LM B1 was expressed temporally in high amounts in the primary decidual zones (PDZ) and persisted throughout PDZ degeneration. LM B2 mRNA was expressed in both primary and secondary decidual zones and persisted through Day 8 of pregnancy. CL IV alpha 1 mRNA expression mimicked that of LM B2. Oviduct ligation on Day 2 of pregnancy was used to prevent embryo transport to one uterine horn, whereas decidualization and embryo implantation were permitted in the contralateral horn. This experiment demonstrated that the increases in uterine ECM mRNA expression were not due solely to the changing hormonal milieu of the uterus. ECM components, including CL IV, have been shown to bind growth factors such as transforming growth factor-beta (TGF-beta) in an insoluble but biologically active form. The remarkable similarity between the pattern of CL IV and LM B2 expression and previously reported TGF-beta deposition (Tamada et al., Mol Endocrinol 1990; 4:965-972) prompted examination of the effects of this growth factor on blastocyst development in vitro. TGF-beta 1 was tested for its ability to alter embryo outgrowth on LM-coated tissue culture surfaces; however, significant differences in the rate or extent of outgrowth in the presence of TGF-beta were not detected.(ABSTRACT TRUNCATED AT 250 WORDS)     

Suppressor cell activity of ovine caruncular and intercaruncular tissues during the placentation period

We assessed the suppression of lymphocyte proliferation ovine endometrial cells recovered during each trimester (Days 45, 90, and 135) of pregnancy. We evaluated fractionated and unfractionated caruncular (C) and intercaruncular (IC) cells for suppression of cocultured peripheral blood lymphocytes (PBL) in phytohemagglutinin (PHA)-treated cultures. We also evaluated cells for the release of the suppressor factor into the culture medium and tested the factor for transforming growth factor-beta (TGF-beta) activity. Suppression of PHA-induced proliferation of PBL was evident for C and IC cells recovered from each day of pregnancy, and the activity was predominantly attributed to the population(s) of low-density (1.006-1.054 g/ml) cells. The activity was greater for unfractionated C than for IC cells on Day 45, whereas the pattern was reversed by Day 135 of pregnancy. For the C cells, the activity was greatest on Day 45 and least by Day 135. Although suppressor factor was released into the medium from cultured C and IC cells, its activity was not apparently mediated by TGF-beta. In conclusion, we observed a temporal pattern in suppressor activity for unfractionated endometrial cells during pregnancy. Suppression was predominately mediated by a population(s) of low-density cells. In addition, the cells released a soluble suppressor factor that seems to be unrelated to TGF-beta. The suppressor cells may provide immunological protection to the fetoplacental unit by suppressing specific lymphocyte responses directed toward conceptus tissues.     

Suppression of cytolytic T-cell activity by staphylococcal enterotoxin B-induced suppressor cells: Role of interleukin 2

We have previously shown that staphylococcal enterotoxin B (SEB) is a potent inducer of suppressor T cells which function to inhibit antibody production in vitro. In the present paper we extend these studies and show that the SEB-induced suppressor cells also inhibit the development of cytotoxic lymphocytes in mixed-lymphocyte reaction (MLR) cultures. Since further analysis also showed that the level of interleukin 2 (IL-2) in cultures of SEB-primed cells was significantly reduced, experiments were carried out to determine the role of IL-2 in the inhibition of cytotoxic cell activity. Attempts to neutralize the suppression by supplementing MLR cocultures with delectinated supernatants from concanavalin A (Con A)-stimulated rat splenocytes were not successful. In addition, MLR cocultures supplemented on Day 0 with 50 units of affinity-purified IL-2 were also suppressed. Further analysis showed that the IL-2 activity in the supplemented MLR cocultures containing suppressor cells were significantly reduced by Day 3. However, repeated supplementation of the MLR cocultures with exogenous IL-2 was successful in achieving (and maintaining) “normal” levels of IL-2. The cytotoxic cell activity in these MLR cocultures remained suppressed. These results suggest that the inhibition of cytotoxic cell activity by SEB-induced suppressor cells is independent of IL-2 levels in the MLR coculture.     

Cellular origin of cytotoxic effectors and secondary educated lymphocytes in human mixed leukocyte reaction

Human lymphocytes sensitized in vitro during a mixed leucocyte reaction (MLR) against an allogeneic-stimulating cell respond by blast transformation and generation of specific cytotoxic effector cells. Both proliferation and cytotoxicity are maximum on Days 6 and 7 of culture. On Day 14, no more dividing cells or cytotoxic cells are detected in such primary cultures. Restimulation by the specific priming cell triggers a secondary proliferative response and rapid reappearance of specific cytotoxic effector cells. The velocity sedimentation cell separation method which separates cells according to their size was applied to human lymphocytes sensitized in vitro during an MLR on Day 7 of culture. Blast cells were separated from nondividing small lymphocytes. It was shown that: (1) cytotoxic effectors generated at the peak of a primary response are exclusively present in the isolated blast population; (2) highly cytotoxic secondary effector cells are induced to reappear mainly from the blast-derived population upon restimulation; and (3) secondary educated proliferative cells mainly derive from the blast population. Conversely, the blast-depleted small lymphocyte population is operationally depleted of cells able to respond by proliferation to the priming cell while responding normally against third party control cells. HLA-D region specificity of the secondary proliferative response is suggested.     

Suppression of lymphocyte alloreactivity by early gestational human decidua. II. Characterization of the suppressor mechanisms

We have earlier shown that first trimester human decidual cells (typical decidual cells and decidual macrophages) suppress lymphocyte alloreactivity (MLR and CTL generation) in vitro in an MHC-unrestricted manner and that this suppression is mediated by PGE2. The present study explored the mechanisms underlying this suppression by noting the effects of decidual cells (+/- indomethacin or anti-PGE2 antibody) or chemically pure PGE2 on numerous T lymphocyte activation events following allogeneic stimulation in mixed lymphocyte culture (MLC) or polyclonal activation with Con A. The results revealed that this suppression was the net result of an action of PGE2 on at least two events during lymphocyte activation: (i) a down-regulation of IL-2 receptor development on lymphocytes, quantitated with a radioimmunoassay and radioautography; this was noted in MLC or Con A-stimulated lymphocyte cultures in the presence of decidual cells (reversible in the presence of indomethacin or anti-PGE2 antibody), or PGE2, but not PGF2 alpha; (ii) an inhibition of IL-2 production in the MLC, measured with a bioassay using an IL-2-dependent T cell line (CTLL-2) and a recombinant IL-2 standard. These effects blocked cell proliferation and eventual generation of killer cells in the MLC. Decidual cells or PGE2 did not interfere with IL-2-dependent proliferation of CTLL-2 cells, which require an interaction between IL-2 receptors on these cells and IL-2. Finally, neither agent interfered with the lytic function of CTL, once generated. These results indicate that the PGE2-mediated immunosuppressor function of early gestational human decidual cells is accomplished by an afferent blockade of the early events in T lymphocyte activation.     

睾丸酮对离体培养人蜕膜细胞形态的影响

实验选用早孕人工流产蜕膜组织进行体外培养,观察睾丸酮对蜕膜细胞形态的影响并与RU 486加以比较。研究结果提示:(1)睾丸酮(6.9×10~(-5)mol/L)能抑制离体培养人蜕膜细胞的生长发育,但这种抑制作用是暂时和可恢复的且与用药剂量及持续时间有关。(2)睾丸酮对蜕膜细胞形态的影响与RU 486(4.7×10~(-4)mol/L)的作用效果相似。     

Growth characteristics of human first trimester decidual cells cultured in serum-free medium: production of prolactin, prostaglandins and fibronectin

A procedure for the establishment of pure human first trimester decidual cells in primary cultures has been developed. The high rate of success in obtaining such cultures resulted from the elimination of fibroblasts by appropriate enzyme dissociation and filtration of the initial tissue sample, and subsequent maintenance of the cells in a serum-free medium supplemented with insulin, fibroblast growth factor (FGF), estradiol, progesterone, hydrocortisone, transferrin and sodium selenite. Under these culture conditions, we obtained pure and actively dividing decidual cells forming tightly packed and nonoverlapping epithelioid cell monolayers covering more than 75% of the culture area. The cultured decidual cells retained their in vivo capacity to produce prolactin and various prostaglandins (PGs), primarily PGE2. There was a marked reduction in hormone production after 20 days in culture. A massive network of fibrillar surface fibronectin was detected by indirect immunofluorescence staining of the cultured cells. The production of prolactin and PGs together with the secretion of fibronectin may play a role in the implantation and subsequent growth of the embryo. The described procedure of obtaining fibroblast-free decidual cell monolayers will promote studies on the hormonal regulation of this tissue at the time of early intrauterine life.