P.M. Starkey

Interleukin-6, turnour necrosis factor and soluble turnour necrosis factor receptors in women with pre-eclampsia

Objectives Generalised maternal endothelial cell dysfunction appears to be an underlying problem in pre‐eclampsia presumed to be caused, directly or indirectly, by one or more circulating factors derived from the placenta. Recently it has been suggested that tumour necrosis factor (TNF) may play an important role in pre‐eclampsia and contribute to endothelial activation. This study was designed to investigate this proposal.

The effect of placental syncytiotrophoblast microvillous membranes from normal and pre‐eclamptic women on the growth of endothelial cells in vitro

Objectives To determine if placental syncytiotrophoblast microvillous (STBM) membranes contain factors which could cause the maternal endothelial cell disturbance thought to be central to the pathophysiology of the maternal syndrome of pre‐eclampsia.

Flow cytometric characterisation of cell populations in human pregnancy decidua and isolation of decidual macrophages.

Methods have been developed for isolating human tissue macrophages from first trimester or term pregnancy decidua. After a two stage enzymic digestion, viable cells were separated from cellular debris by velocity sedimentation at unit gravity or by Percoll centrifugation. Cell populations were analysed by flow cytometry after labelling with monoclonal antibodies. In term decidua, 47% of the cells were of bone marrow origin, comprising 18% macrophages, 3% large granular lymphocytes and 8% T cells. The remaining cells, the proportion of which varied between individuals, were CD16-positive granulocytes. Macrophages were isolated flow cytometrically from both first trimester and term decidual cell dispersions after labelling with an antibody to MHC class II. Yields of up to 4 X 10(6) macrophages, greater than 95% pure, were routinely obtained.

Localization of tumour necrosis factor production in cells at the materno/fetal interface in human pregnancy

Biologically active tumour necrosis factor (TNF) was detected in medium conditioned by incubation with cxplants of human pregnancy decidua or fetal chorionic villous tissue, taken in the first trimester and at term. Addition of endotoxin increased TNF release in most cases. ELISA assays gave similar results for TNF–α and also demonstrated low levels of TNF–β. Using cell populations purified by flow cytometry, secretion of biologically active TNF was shown to be localized to the macrophages. Cytotrophoblast purified from term amniochorion produced no TNF. Both decidual and chorionic villous tissue at term contained mRNA for TNF–α and TNF–β. TNF–α mRNA was confined to decidual macrophages in first trimester tissue, and was not present in chorionic cytolrophoblast. TNF–β mRNA, in contrast, was detected in both macrophage and no‐acrophage populations in term dccidua.

Cytotoxic activity against trophoblast and choriocarcinoma cells of large granular lymphocytes from human early pregnancy decidua.

Large granular lymphocytes (LGL) are the most abundant cell type in first trimester human pregnancy decidua. We have shown previously that CD56-positive decidual LGL have cytotoxic activity against the natural killer (NK) target K562, and that this cytotoxicity is augmented by pretreatment with interleukin-2 (IL-2). We now report that flow cytometrically purified populations of CD56-positive decidual LGL have no cytotoxic activity against either the BeWo choriocarcinoma cell line or freshly isolated term trophoblast. Incubation of unfractionated decidual cells with IL-2 induced cytotoxicity against BeWo, but term trophoblast remained resistant to lysis. Both BeWo and trophoblast showed much lower binding frequencies to decidual or peripheral blood cells than K56 targets, and excess trophoblast did not inhibit cytotoxic activity against K562. This suggests that the resistance of trophoblast to lysis by either decidual or peripheral blood LGL is due to the lack of accessible NK target structures on the surface of trophoblast.

Peritoneal fluid cytokines and the relationship with endometriosis and pain

It is generally accepted that the current scoring system for endometriosis has little correlation with clinical symptoms such as pain, and therefore we may deduce that either endometriosis does not cause pain, or that the current scoring system does not indicate the biological activity of the disease. Pain may occur because the presence of endometriosis produces an intraperitoneal inflammatory response, and several studies have shown that the cytokine content of peritoneal fluid differs between women with and without endometriosis. We studied the relationship between tumour necrosis factor alpha (TNF alpha), platelet-derived growth factor (PDGF), interleukin (IL)-6, IL-4 and TNF (alpha and beta) activity in peritoneal fluid and the clinical history of pain and infertility. TNF alpha concentrations were increased in peritoneal fluid of women with endometriosis and of infertile women; PDGF concentrations were increased in peritoneal fluid of parous women; IL-6 was increased in peritoneal fluid of women with adhesions; IL-4 was absent from peritoneal fluid. PDGF and IL-6 concentrations were cycle related, with the highest amounts in the menstrual and proliferative phases respectively. We failed to demonstrate any association between concentrations of cytokines in vitro and pain symptoms or severity of endometriosis.

Antigenic heterogeneity of human cytotrophoblast and evidence for the transient expression of MHC class I antigens distinct from HLA-G

Expression of MHC class I antigens on trophoblast populations in first trimester human chorionic villous tissue was assessed by immunohistology. Antibodies used were W6/32 which recognizes a non-polymorphic framework determinant of HLA- A, -B, -C, MHM5 specific for HLA-B, C and 4E and B23.1 which are specific for HLA-B. Syncytiotrophoblast and villous cytotrophoblast were negative with all the anti (HLA class I) antibodies tested. Interstitial trophoblast cells within the maternal decidua were identified with a new antibody, NDOG5, which is specific for extravillous cytotrophoblast. Double labelling showed that they bind W6/32 but not 4E, MHM5 or B23.1; consistent with the expression of the monomorphic HLA-G. In contrast the cytotrophoblast cells of the cell islands and cytotrophoblast shell, which also express the NDOG5 antigen, were positive with W6/32, 4E, MHM5 and B23.1. Cell column cytotrophoblast cells were negative with all four MHC class I antibodies. These results suggest that differentiation of cytotrophoblast from noninvasive to invasive forms is associated with transient expression of class I antigens other than HLA-G on cytotrophoblast shell and cell island cytotrophoblast.

Tumor necrosis factor-α selectively stimulates prostaglandin F2α production by macrophages in human term decidua

OBJECTIVE: Our objective was to investigate the effect of tumor necrosis factor-α on prostaglandin production by human term decidual cell subtypes in vitro. STUDY DESIGN: We measured the effect of tumor necrosis factor-α on prostaglandins F 2α , E 2 , D 2 , F metabolite, and E metabolite production by decidual cells ( n = 4) with radioimmunoassay. We used flow cytometry after labeling with an antibody to histocompatibility antigen DR, L243, which is specific for macrophages in this tissue, to prepare pure populations of decidual macrophages ( n = 3). Differences in prostaglandin output were analyzed by Wilcoxon and Kruskal-Wallis tests. RESULTS: Tumor necrosis factor-α stimulated prostaglandin F 2α , output by unfractionated decidual cells, without altering the output of any other prostaglandins. Tumor necrosis factor-α (10 nmol/L) increased prostaglandin F 2α production from decidual macrophages (HLA-DR-positive cells) by about threefold: from a median of 727 (range 423 to 1226) to a median of 1974 (range of 1550 to 2201), fmol/10 6 cells per 18 hours, as compared with a 1.4-fold increase from nonmacrophages from a median of 247 fmol/10 6 cells per 18 hours (range 125 to 611) to a median of 340 fmol/10 6 cells per 18 hours (range 201 to 505). CONCLUSION: Stimulation of decidual prostaglandin F 2α production by tumor necrosis factor-α may be important in the etiology of spontaneous labor at term or preterm labor associated with infection.

Plasminogen activators in ectopic and uterine endometrium

OBJECTIVE To investigate the expression of the plasminogen activator (PA)-plasmin system components in ectopic endometrium and in uterine endometrium from women with and without endometriosis. DESIGN Plasminogen, PAs (urokinase and tissue plasminogen activator), and PA inhibitors (1 and 2) were detected by immunohistochemistry using a alkaline phosphatase staining method. RESULTS No differences in staining were found between uterine endometrium of women with endometriosis and women without endometriosis with any of the antibodies used. However, we did find differences between uterine and ectopic endometrium. Although the expression of the components of the PA-plasmin system reflected the cyclic changes in the hormonal levels in uterine endometrium, ectopic endometrium maintained a very high level of plasminogen and urokinase in every sample. We were unable to detect the presence of PA inhibitors in either uterine or ectopic endometrium. CONCLUSIONS There is no evidence that uterine endometrium from women with endometriosis is originally more able to implant than that of women without the disease because of an increase in their PA expression. The high levels of urokinase and plasminogen in ectopic endometrium may reflect a more invasive nature of the endometriotic implants in the peritoneal cavity.

Localization of anti-endometrial antibody binding in women with endometriosis using a double-labelling immunohistochemical method.

Summary. Anti‐endometrial antibody binding was localized using a double‐labelling immunohistochemical method on frozen sections of endometrium taken from a woman without pelvic disease. Serum from 40 women with endometriosis was tested and, as controls, serum samples from 20 adult males and 20 umbilical cords. The method allowed compensation for endogenous immunoglobulins in endometrium and accurate localization of anti‐endometrial antibody binding in the cytoplasm of the glandular epithelium. Significantly more women with endometriosis (14/40) were found to have anti‐endometrial antibodies than controls (1/40) (P< 0.001; χ2). There was no correlation between disease severity and the presence of anti‐endometrial antibodies or the intensity of staining.