Maternal cells in chorionic villi from placentae of normal and abnormal human pregnancies

PROBLEM

We asked if activated macrophages and CD4 positive T lymphocytes in placental chorionic villi with villitis were of maternal or fetal origin.

METHOD

We employed a double antibody immunocytochemical technique on placental sections from three normal and four abnormal pregnancies with small-for-gestational-age infants. All studied placentae were mismatched for the maternal-fetal HLA-DRw 52 antigen. Areas of immunopathology were identified by using a monoclonal antibody to a monomorphic determination on HLA-DR, and the origin of immunological cells in areas of immunopathology was identified by using a monoclonal antibody to a polymorphic determinant on HLA-DRw 52.

RESULTS

We used a double antibody technique that employed monoclonal antibodies to HLA-DR and HLA-DRw 52 antigens and placentae that were mismatched for the maternal-fetal HLA-DRw 52 antigen. We found that the vast majority of immunological cells within villi with inflammation were of maternal origin. Quantitative studies showed that between 75 and 100% of the cells in normal as well as in abnormal pregnancies were of maternal origin, and that abnormal pregnancies had a significantly higher percentage of villi with maternal cellular infiltrates.

CONCLUSION

Our data show unequivocally that cells in areas of placental immunopathology are predominantly of maternal origin, and that abnormal pregnancies are associated with significantly more villi containing immunological cells of maternal origin.

Intercellular adhesion molecule-1 (ICAM-1) and HLA-DR antigens are expressed on endovascular cytotrophoblasts in abnormal pregnancies

PROBLEM

We asked if the lack of normal trophoblastic invasion of spiral arteries in the basal plate of abnormal pregnancies was associated with the expression of HLA-DR antigens and intercellular adhesion molecules (ICAM-1) on endovascular cytotrophoblasts.

METHOD

The basal plates of placentae from 15 normal and 55 abnormal pregnancies, including preeclampsia, small-for-gestational age infants, and mothers with history of secondary recurrent spontaneous abortion, were studied immunocytochemically by using monoclonal antibodies to HLA-DR and ICAM-1. Spiral and uteroplacental arteries were identified by using a triple antibody technique with antibodies to cytokeratin, alpha-smooth muscle actin, and von Willebrand factor to detect cytotrophoblasts, arterial smooth muscle cells, and endothelium, respectively.

RESULTS

Placentae with normal placentation showed the presence of uteroplacental arteries that contained endovascular cytotrophoblasts that were negative for HLA-DR and ICAM-1 antigens. Placentae from abnormal pregnancies showed the presence of spiral arteries without trophoblastic invasion and uteroplacental arteries that were surrounded by numerous macrophages and T lymphocytes. Endovascular cytotrophoblasts in uteroplacental arteries of placentae from abnormal pregnancies reacted with antibodies to HLA-DR and ICAM-1 antigens.

CONCLUSION

Placentae from normal pregnancies show uteroplacental arteries that contain endovascular cytotrophoblasts that do not react with antibodies to ICAM-1 and HLA-DR antigens, and placentae from abnormal pregnancies with uteroplacental arteries that are associated with arteries that do not show physiological changes contain endovascular cytotrophoblasts that react with antibodies to ICAM-1 an HLA-DR antigens. Normal uteroplacental arteries were found to be not surrounded by round cell infiltrates, but uteroplacental arteries associated with arteries that lack physiological changes were surrounded by round cell infiltrates, indicating that round cell infiltrates and endovascular cytotrophoblasts which react with antibodies to ICAM-1 and HLA-DR antigens are associated with abnormal pregnancies. These findings suggest that the cellular infiltrates are associated with endovascular cytotrophoblasts that react with ICAM-1 and HLA-DR antigens.

Natural anticoagulant and fibrinolytic pathways in renal allograft failure

This is an immunocytochemical study of the relationship between depletion of natural anticoagulant and fibrinolytic pathways and allograft survival following renal transplantation. Patients (n = 44) were classified in three groups according to the length of time between transplantation and allograft failure: group 1 (n = 14) failed within a month of transplantation; group 2 (n = 14) failed between one month and one year after transplantation; and group 3 (n = 16) failed after one year of transplantation. Control biopsies were from donor kidneys (n = 16) prior to transplantation. There were no statistically significant differences in recipient age, gender, donor kidney type (living-related versus cadaver), histocompatibility, and plasma cholesterol, triglycerides, or creatinine concentrations between groups. However, group 1 allografts had a greater depletion of the vascular heparan sulfate proteoglycan-antithrombin III natural anticoagulant pathway than allografts in group 2 or 3 (P < or = 0.05), and this depletion was associated with significantly greater fibrin deposition in group 1 than in either group 2 or 3 (P < or = 0.05). All three groups demonstrated severe depletion of tissue plasminogen activator from arteriolar smooth muscle cells and depressed fibrinolysis as evidenced by increased fibrin/plasmin ratios. However, no significant differences were found for either endothelial thrombomodulin or T cell, neutrophil, or macrophage infiltration between the groups. These data indicate that differences in graft outcome may be determined more by compromised vascular function than by the presence of cellular infiltrates.

Antigenic identification of cells in spiral artery trophoblastic invasion: validation of histologic studies by triple-antibody immunocytochemistry

OBJECTIVE

Trophoblastic invasion of spiral arteries in the placental bed is essential for normal human placentation. Because the absence of these physiologic changes is associated with abnormal pregnancies, we developed a simple immunocytochemical approach that allows simultaneous study of three cell types involved in spiral artery changes (i.e., trophoblastic, endothelial, and smooth muscle cells).

STUDY DESIGN

Spiral and uteroplacental arteries in the basal plate of 70 term placentas were studied simultaneously with antibodies to keratin to identify cytotrophoblasts, alpha-smooth muscle actin to identify vascular smooth muscle cells, and von Willebrand factor to identify endothelial cells in 15 normal placentas and 55 placentas from abnormal pregnancies.

RESULTS

Normal term placentas showed keratin-positive, alpha-smooth muscle actin-negative and von Willebrand factor-negative uteroplacental arteries. Most (37/55) placentas from abnormal pregnancies had keratin-negative, alpha-smooth muscle actin and von Willebrand factor-positive spiral arteries. Spiral artery atherosclerosis was keratin negative, alpha-smooth muscle actin positive, and von Willebrand factor positive and was identified in 22 abnormal and 0 normal placentas.

CONCLUSION

Simultaneous identification of three principal cell types involved in human placentation allowed qualitative and quantitative studies of placentas from normal and abnormal pregnancies that validated previous histologic investigations and provided a new approach to detect spiral arterial changes.

Tissue plasminogen activator, plasminogen activator inhibitor-1, and fibrin as indexes of clinical course in cardiac allograft recipients. An immunocytochemical study

BACKGROUND

Tissue-type plasminogen activator (TPA) is the principal activator of plasminogen. Since hemostasis in the microcirculation of allografts is a well-recognized complication of transplantation, we asked (1) whether the distribution and amount of cellular TPA in biopsies of transplanted human hearts are associated with fibrin deposits in and around the microcirculation, (2) whether such changes involve the physiological inhibitors of TPA and plasmin, and (3) whether the presence of these activators and inhibitors of fibrinolysis in tissue is correlated with clinical outcome.

METHODS AND RESULTS

We immunocytochemically quantified the presence of fibrin, plasmin, TPA, and the TPA inhibitor PAI-1 in 938 biopsies from 68 consecutive cardiac allografts over a 54-month period. The localization, distribution, and quantification of TPA in arteriolar smooth muscle cells revealed that 35 of the 68 allografts maintained vascular TPA reactivity consistent with time-zero biopsies of autologous donor hearts: this was designated as the normal TPA group. In contrast, 33 of the 68 allografts significantly lost vascular TPA reactivity compared with time-zero biopsies of autologous donor hearts: this was designated as the depleted TPA group. Analysis of sequential biopsies from both groups during 54 months revealed that the mean cumulative quantitative TPA value for the normal TPA group was 1.0 +/- 0.01, whereas the depleted TPA group value was 1.9 +/- 0.02 (P = .0001), and the mean cumulative quantitative fibrin value for the normal TPA group was 1.0 +/- 0.01, whereas the depleted TPA group value was 1.5 +/- 0.05 (P = .0001). Biopsies of allografts in the depleted TPA group contained endothelial reactivity for TPA-PAI-1 complexes, whereas biopsies from the normal TPA group did not. Plasmin-associated molecules were rarely identified in biopsies of the normal TPA group but were present in the depleted TPA group, and the fibrin-to-plasmin ratio in the normal TPA group always was less than the fibrin-to-plasmin ratio in biopsies from the depleted TPA group. Analysis of demographic and risk factors revealed no significant differences between patients in the normal and depleted TPA groups, but none of the 35 patients in the normal TPA group died or were retransplanted, and 13 of the 33 patients in the depleted TPA group died or required retransplantation (P = .0001).

CONCLUSIONS

Time-zero hearts (n = 68) and 34 of 38 stable allografts contained immunocytochemically detectable TPA only in vascular smooth muscle cells. Twenty-nine of 30 patients with normal TPA in their time-zero biopsies who subsequently developed a poor clinical outcome were found to have depleted TPA in biopsies evaluated during their first postoperative month and remained depleted throughout the study. Of 33 patients with depleted TPA, 39% died or required retransplantation. Depleted arteriolar TPA associated significantly with vascular and interstitial deposits of fibrin, plasmin, and endothelial TPA-PAI-1 complexes. These findings indicate that hemostatic and fibrinolytic pathways are activated in falling allografts, and they reveal evidence of depleted TPA before clinical or histopathological signs of failure. Patients with such allografts were found to be at high risk of death independently of other widely used clinical/laboratory parameters of prediction.

Predominant expression of the beta subunit of prolyl 4-hydroxylase (disulfide isomerase) in human extravillous trophoblasts

Prolyl 4-hydroxylase is a heterodimeric enzyme that is crucial in the biosynthesis of collagen. The beta subunit of this enzyme is a multifunctional protein which is also known as protein-disulfide isomerase. Immunofluorescence and monoclonal antibody (Mab) 5B5 were used to localize the beta subunit in human extraembryonic tissues. The strongest sites of 5B5 reactivity were extravillous cytotrophoblasts in the basal plate, uteroplacental arteries and amniochorion, syncytiotrophoblast displayed variable weaker reactivity. Only a small fraction of placental 5B5 antigen was detected as a component of prolyl-4-hydroxylase by affinity chromatography on immobilized polyproline. The results indicate a difference in the expression of an endoplasmic reticulum marker between villous and extravillous trophoblast. The predominance of 5B5 antigen in extravillous trophoblast could be associated with an increased ability to synthesize collagen or other enzymatic reactions associated with prolyl 4-hydroxylase beta subunit.

Immunoglobulin M antibodies in transplanted human hearts

We have studied 56 human hearts before and after transplantation for the presence of immunoglobulin M. None of the 56 time-zero biopsy specimens studied contained immunocytochemically detectable immunoglobulin M, but they all had immunoglobulin M deposits on vascular endothelial cells after transplantation. The vascular location of immunoglobulin M was confirmed in double-antibody experiments with antibodies to von Willebrand factor and immunoglobulin M. None of the immunoglobulin M antibody reactivity was shown to colocalize with complement. Biopsy specimens from 29 of the 56 patients contained abundant deposits of immunoglobulin M. These patients were clinically stable, and only one of them has died. Biopsy specimens from 27 of the 56 patients had scant deposits of immunoglobulin M. These patients were clinically unstable; eight of them have died, and one has undergone retransplantation. In addition, biopsy specimens from stable allografts contained fewer fibrin deposits than biopsy specimens from unstable grafts, suggesting a protective role for non-complement fixing immunoglobulin M. The observation that immunoglobulin deposits were found soon after transplantation prompts us to consider them to be natural antibodies. Efforts to identify the antigen for immunoglobulin M natural antibodies in heart transplant recipients are currently under study.

Tissue plasminogen activator in human cardiac allografts

The activation of hemostasis and fibrinolysis frequently is observed in allografted organs. Plasminogen is activated by urokinase and tissue plasminogen activator (tPA). We have studied human hearts before and after transplantation to determine if fibrin deposition within the microcirculation is associated with a depletion of myocardial tPA, and if such depletion of tPA is associated with decreased fibrinolysis. We found that tPA in pretransplanted hearts and in biopsies from hearts of most patients with a stable clinical course is confined to arterial and arteriolar smooth muscle cells. The depletion of smooth muscle cell reactivity was associated with microvascular fibrin deposition in unstable allografts, and the appearance of endothelial cell tPA reactivity heralded a bad prognosis. Successful medical management was signaled by a loss of endothelial tPA reactivity and a return of tPA reactivity in arterial and arteriolar smooth muscle cells. These findings indicate a central role for tPA in maintaining the integrity of the microcirculation in transplanted human hearts.

Vascular lesions in biopsy specimens devoid of cellular infiltrates: qualitative and quantitative immunocytochemical studies of human cardiac allografts

Biopsy specimens from 80 cardiac allografts were studied immunocytochemically with nine antibodies selected to identify cellular, immunologic, and vascular aspects of rejection. Results from these experiments were converted to a numeric base and used to calculate rejection indexes for each of these aspects of rejection. Pretransplantation biopsy specimens of donor hearts were studied to determine normal values, and clinical, catheterization, and conventional biopsy findings were used to classify patients as stable or unstable. The standard deviation of cellular and immunologic reaction indexes were overlapping for stable and unstable patients, but the vascular indexes for these patient groups were significantly different, suggesting that vascular damage defines clinical instability in cardiac allograft recipients. To study the vascular index in the absence of cellular infiltrates, biopsy specimens from 80 allografted hearts were selected to represent 40 patients with cellular infiltrates and 40 patients devoid of infiltrates. The results of this analysis revealed no difference in the vascular index between hearts that contained or did not contain infiltrates; the spectrum of vascular damage between the groups ranged from trivial to severe, suggesting that vascular damage was not dependent on cellular infiltrates. Concluding that these were independent events was not possible, however, because none of the hearts continuously were devoid of cellular infiltrates. Two groups of patients with significantly different vascular indexes then were studied for a systemic manifestation of coagulation by measurement of their thrombin times. The group with low vascular index was found to have significantly prolonged thrombin times compared to the group with high vascular index, and a trend to less prolongation of thrombin times was observed in another group with high cellular index. The relation of these results with cardiac allograft vasculopathy is not established, but the results suggest hitherto unexplored roles for quantitative expressions of hemostasis, fibrinolysis, and anticoagulation in the pathophysiology of vascular disease in allografts.

Modulation of vascular antithrombin III in human cardiac allografts

The natural anticoagulant pathway involving heparan sulfate proteoglycan and antithrombin III (ATIII) was studied in serial biopsies from 90 cardiac allograft recipients. The ATIII component of this pathway was identified immunocytochemically on venous endothelium and arterial smooth muscle cells and intima of normal donor hearts and stable allografts. Unstable grafts lacked vascular ATIII and contained fibrin deposits. Neither stable nor unstable grafts had ATIII-reactive capillary endothelium. Grafts with absent vascular ATIII could (1) result in death, (2) revert to an arterial/venous ATIII distribution or (3) develop ATIII-reactive capillary endothelium. The development of ATIII-reactive capillaries was associated with a survival advantage, and such reactivity seemed to be promoted by heparin.

Laboratory-clinical correlates of time-associated lesions in the vascular immunopathology of human cardiac allografts

Ninety cardiac allograft recipients were studied for clinical and functional parameters during a 40-month period. Baseline histologic and immunocytochemical data were obtained from donors' hearts before transplantation, and serial endomyocardial biopsy specimens were studied histologically for cellular infiltrates and immunocytochemically for complement and immunoglobulin deposits and for components of the hemostatic, fibrinolytic, and natural anticoagulant pathways. Results were grouped according to the time from transplantation: the first 3 months, 4 to 21 months, and 22 to 40 months. Each group was evaluated for most frequently obtained immunocytochemical findings and results relevant to clinical-laboratory cooperation in patient management. During the first 3 months, findings of biopsy specimens from allografts that subsequently were going to be problem cases revealed depleted tissue plasminogen activator in arteriolar smooth muscle cells, and recipient IgM that was deposited on donor endothelium of stable grafts was diminished or absent in unstable allografts. In addition, vascular deposits of activated complement components were identified in 50 of 70 allografts. From 4 through 21 months after transplantation, vascular deposits of complement rarely were identified (even in patients who previously had positive biopsy specimens), and the principal vascular lesion was fibrin deposits with impaired anticoagulant pathways and inadequate fibrinolysis, usually without associated cellular infiltrates. From 22 through 40 months after transplantation, the principal vascular lesion was graft-induced atherosclerosis. These immunocytochemically defined qualitative and quantitative changes in unstable or failing allografts form a time-related spectrum of lesions that encompasses the emerging diagnostic entity of vascular rejection.

Immunological studies of lactoferrin in human placentae

Lactoferrin (LF) and transferrin (Trf) are glycoproteins with strong affinities for ferric ions. Human syncytiotrophoblastic membranes analyzed by enzyme linked immunosorbent assay (ELISA) and immunoblotting were negative with monoclonal and polyclonal antibodies to LF. Immunohistological studies of 35 normal placentae showed that LF was absent from the trophoblast basement membranes, stroma and fetal stem vessel endothelium, but positive cells were occasionally noted in intervillous spaces and fetal stem vessels. In contrast, many LF-positive cells were identified within areas of immunopathology identified by the presence of T cells, HLA-DR-positive macrophages and platelets. Double-antibody experiments showed that the LF-positive cells in these areas reacted with CD15 and CD16 monoclonal antibodies (mAbs), indicating that the cells were polymorphonuclear neutrophils (PMN). PMN from peripheral blood analyzed by flow cytometry and immunocytology also showed reactivities with anti-LF, CD15 and CD16 and we consistently found that circulating PMN reacted better than placental PMN with antibodies to MHC class I antigens and gp 100, (CD67), which is a neutrophil activation marker. PMN adherent within placentae had no detectable MHC class I or CD67 antigens. These findings suggest PMN adherent to placental tissues down-regulate or alter plasma membrane markers. LF appears to play a role in placental inflammation, for LF-positive cells were significantly enriched in areas of immunopathology.

Expression of CD59, a human complement system regulatory protein, in extraembryonic membranes

CD59 (leukocyte cluster of differentiation antigen 59), is a phosphatidylinositol glycan-anchored membrane protein that inhibits lysis of cells by terminal complement system components. To further define complement regulatory proteins relevant to pregnancy, this study characterized the expression of CD59 in human extraembryonic membranes. Immunohistology with CD59 monoclonal antibody MEM-43 showed that this molecule was normally present on the apical surface of the syncytiotrophoblast, on extravillous cytotrophoblast, and amniotic epithelium. Immunoblotting confirmed that first and second trimester syncytiotrophoblast microvilli (STM) contained a glycoprotein similar in mass and glycosylation to CD59 from adult cells and tissues. Reactivities of STM with MEM-43 in ELISA were 2- to 6-fold higher than those of kidney, erythrocyte and platelet membranes. Term placental STM from recurrent spontaneous aborting patients after immunotherapy, reacted with MEM-43 in ELISA similarly to STM from normal individuals. Plasmas from pregnant women and umbilical cords had 50% or greater reactivity with MEM-43 than did normal plasmas. CD59 could help protect extraembryonic epithelia from damage by complement in maternal blood and amniotic fluid. The apical location of CD59 reflects the immunological roles and functional polarization of plasma membranes in the syncytiotrophoblast.

Vascular immunopathology and atheroma development in human allografted organs

Human atherosclerosis requires decades to develop spontaneously, and its development customarily is not monitored by serial biopsies. Atherosclerosis in allografts develops within months, and biopsy specimens are usually obtained from the grafts. We have used immunocytochemical techniques to study biopsy specimens of cardiac and renal allografts for parameters of vascular changes. The antibodies used in this investigation were specific for T lymphocytes, macrophages, IgM, and complement, and detailed studies were done with the use of antibodies specific for components of the hemostatic, fibrinolytic, and natural anticoagulant pathways. The results indicated a lack of association between the appearance of cellular infiltrates and measurable alterations in vascular endothelium and smooth-muscle cells. Although infiltrating macrophages and lymphocytes were identified in biopsy specimens with vascular change, such changes were also observed in biopsy specimens that were devoid of cellular infiltrates. The most prominent vascular changes in endothelium and smooth-muscle cells were accelerated hemostasis, depressed fibrinolysis, and deranged anticoagulant pathways. A negative association between the presence of IgM and fibrin on endothelial cells was also identified. Whether vascular changes are due to immunologically mediated reactions or to as yet undefined metabolic stresses of the graft-host relationship remains to be determined. This study provides a model for the study of vascular changes in atheroma development as viewed through the window of transplant-induced atherosclerosis.

NOVEL IMMUNOHISTOCHEMICAL MARKERS OF HUMAN RENAL ALLOGRAFT DYSFUNCTION—ANTITHROMBIN III, THY-1, UROKINASE, AND ALPHA–SMOOTH MUSCLE ACTIN

We have studied the expression of alpha-smooth muscle actin (alpha sm-1) by mesangial cells, and the expression of Thy-1 glycoprotein, antithrombin III (ATIII), and urokinase by tubular epithelial cells in normal kidneys and dysfunctional renal allografts. Kidney biopsies were studied immunocytochemically for changes in each of these markers and the findings were classified into two groups and compared with creatinine plasma levels at the time the biopsies were taken. In dysfunctional grafts, mesangial alpha sm-1 and tubular epithelial Thy-1 reactivities were greatly diminished, and urokinase and ATIII were missing from proximal renal tubular epithelial cells. Urokinase, which was absent from normal renal glomeruli, appeared in glomeruli of some dysfunctional allografts. The possible usefulness of these markers in patient evaluations was supported by our finding that the distribution of vinculin, fibronectin, myosin, actin B4, desmin, glomerular HLA-DR, and the tubular expression of CD15 remained unchanged. These data prompt us to suggest that the immunocytochemical localization and evaluation of alpha sm-1, Thy-1, ATIII, and urokinase in kidney allografts may be useful adjuncts in the assessment of function in renal allografts.

Vascular events in placentae and organ allografts

This investigation explores the transplantation analogy of placentae with allografted human organs. Biopsies of cardiac and renal allografts and placentae were studied immunocytochemically with antibodies to components of the immunological, coagulational, anticoagulational, and fibrinolytic systems. Cellular rejection of cardiac and renal allografts was identified by infiltrating lymphocytes and macrophages. This was accompanied by vascular damage characterized by loss of endothelial anticoagulant pathways, vascular deposits of fibrin, and depletion of arterial tissue plasminogen activator (tPA). Failing allografts, including placentae from abnormal pregnancies, demonstrated coagulation/fibrinolytic changes consistent with vascular rejection, regardless of the presence of cellular infiltrates. An IgM autoantibody to allogeneic endothelium was associated with vascular protection. Its presence in cardiac and renal transplant biopsies was associated with an absence of fibrin deposits, and its absence was associated with vascular damage. Atherosclerosis commonly was identified in allograft biopsies (including placentae from abnormal pregnancies). These changes were shown in serial biopsies of transplanted hearts to be preceded by or associated with diminished IgM autoantibody, impaired endothelial anticoagulant pathways, vascular fibrin deposits, and depleted arterial reactivity for tPA. These results indicate the transplantation analogy of pregnancy should be viewed more specifically as vascular smooth muscle cell and endothelial responses to as yet unidentified microenvironmental stimuli.

Tissue factor in normal and transplanted human kidneys

Tissue factor (TF) plays a central role in the initiation of blood coagulation that frequently is enhanced in renal allografts. The identification and localization of TF was studied immunocytochemically in biopsies from normal and transplanted human kidneys and classified according to its distribution. The clinical status of each allograft was then correlated with the TF classifications. From these correlations, four distributional types of TF were identified. In normal kidneys, TF was localized to glomerular epithelium and basement membranes. Glomerular TF expression did not colocalize with mesangial or endothelial HLA-DR reactivity as determined by double antibody techniques. Tissue factor in donor kidneys also was identified in the renal capsule and in the adventitia of large arteries. These structures were not reactive in long-term transplanted grafts. Some cadaver kidneys prepared for transplantation had depleted glomerular TF, and exhibited TF reactivity within stromal tissues. Long-term allografts with progressive loss of renal function and kidneys with advanced rejection exhibited diminished TF reactivity of glomerular epithelium and basement membranes. This was frequently associated with fibrin deposition within the glomeruli and in the intertubular microcirculation. These findings indicate that the evaluation of TF in transplanted kidneys is related to the prognosis of graft survival.

Glycosylation of membrane cofactor protein (CD46) in human trophoblast, kidney and platelets

Many cell surface glycoconjugates are differentiation markers and are involved in cell-cell and intermolecular interactions in development, immunity and cancer. Membrane cofactor protein (MCP) comprises structurally related 65 and 55 kDa glycoproteins that bear O- and N-linked glycans. MCP prevents amplification of autologous complement action on human cells. We used immunoblotting with MCP-specific monoclonal antibody TRA-2-10 to determine lectin-binding properties and glycosidase sensitivities of MCP in a study of cell-specific variation in glycosylation of this protein. The results showed that N-linked glycans on placental syncytiotrophoblast and cytotrophoblast, kidney and platelet MCP are similar in binding to concanavalin A and Lens culinaris lectins, but are not bound by leucophytohemagglutinin. Lectin binding prior to and after neuraminidase digestion indicates that MCP from these sources is highly sialylated. 65 kDa MCP was confirmed to contain more O-linked glycans than 55 kDa MCP. A fraction of platelet 65 kDa MCP is distinct, however, in bearing peripheral fucose residues. Syncytiotrophoblast is unique in containing a 110 kDa form of MCP in non-reducing SDS-PAGE that resembles 65 kDa MCP in glycosylation. Chorion laeve MCP in 4 of 8 preparations was unusually heterogeneous and differed from syncytiotrophoblast MCP after neuraminidase digestion in the forms bound to peanut agglutinin and WGA. The results indicated for the first time, differences in O-linked glycosylation of MCP in chorion laeve cytotrophoblast relative to syncytiotrophoblast, platelet and kidney MCP. We conclude that structures of MCP glycans can differ between trophoblasts and other cell types.

Microvascular perturbations in human allografts: analogies in preeclamptic placentae

The thromboresistance of endothelium is maintained as long as natural anticoagulant pathways are functionally present on endothelial plasma membranes. The principal anticoagulant pathways in human hearts and kidneys are thrombomodulin (TM) and heparan sulfate proteoglycan-antithrombin III (HSPG-ATIII). The downregulation of TM or the loss of ATIII is associated with fibrin deposition. This sequence of events occurs when stable allografts of hearts or kidneys become unstable or rejected. Human placentae do not contain the HSPG-ATIII natural anticoagulant pathway, but the TM system is uniformly represented on endothelium of normal chorionic villi. However, many villi in placentae from preeclamptic pregnancies contain thrombomodulin-negative endothelium, and these vessels contain fibrin thrombi. These thrombi compromise blood flow through the placental microcirculation and are associated with ischemic changes either with or without the presence of cellular infiltrates.

The complement system in human reproduction

Regulation of the complement system in reproduction is unique inasmuch as reproductive tissues represent the only condition where allogeneic interactions occur naturally. Both allogeneic extraembryonic membranes and semen that contact and interact with maternal cells and tissues must avert complement-mediated damage to ensure reproductive success. Several regulators of complement activation exist. Membrane cofactor protein (MCP) and decay accelerating factor (DAF) inactivate C3 and C5 convertases on cell surfaces. In addition, CD59 inhibits the membrane attack complex (MAC) of the complement cascade. Strong expression of these membrane glycoproteins by trophoblast and amniotic epithelium has been observed. MCP, DAF, and CD59 likely safeguard extraembryonic tissues from complement damage originating from maternal and fetal blood or amniotic fluid. Different reproductive tract fluids vary in complement levels. With the exception of ovarian follicular fluid, these levels are generally much less than those in blood. Endometrial and cervical content of C3 appear to be regulated by hormones. These observations suggest that the effects of complement activation may vary in reproductive tissues. MCP is absent from the surfaces of oocytes. Sperm express MCP and DAF in discrete areas that would not be associated with the known complement-regulatory functions of these proteins. Seminal plasma contains MCP and the MAC inhibitor SP-40,40 but not DAF.SP-40,40 may exemplify how complement-regulatory proteins perform alternative functions as it interacts with molecules other than complement components. We have reviewed aspects of the complement system that relate to allogeneic interactions in reproduction and that suggest fruitful areas for further research.

The endothelial heparan sulfate-antithrombin III natural anticoagulant pathway in normal and transplanted human kidneys

This is the first study of the antithrombin III-heparan sulfate natural anticoagulant pathway in human kidneys. Immunocytochemical experiments were done to demonstrate the pathway on normal renal endothelial cells. Enzymatic studies were done to show that the antithrombin III was anchored to endothelium by molecules of heparan sulfate. Displacement studies were done with glycosaminoglycans to show that the antithrombin III was bound to its glycosaminoglycan anchor via a heparinlike binding site, and replacement studies showed that antithrombin III could be returned to the same endothelial cells from which it was displaced. Immunocytochemical studies of biopsies showed that normally functioning renal allografts manifested the endothelial antithrombin III-heparan sulfate anticoagulant pathway. The pathway was compromised or absent from the microcirculation of biopsies from rejecting or rejected renal allografts, and the diminishment of endothelial ATIII was associated with the presence of fibrin deposition. It is concluded that compromise of the antithrombin III-heparan sulfate natural anticoagulant pathway results in compromised renal function in transplanted kidneys.