AMERICAN JOURNAL OF REPRODUCTIVE IMMUNOLOGY 27:97-100 © 1992 MUNKSGAARD

AJJ!!nancg Immunolog.l

Fetal Stem Vessel Endothelial Changes in Placentae From Normal and Abnormal Pregnancies C. A. LABARRERE AND W. PAGE FAULK Methodist Hospital of Indiana, Indianapolis, Indiana ABSTRACT: Areas of immunocytochemically defined immunopathology are common in placentae from secondary recurrent spontaneous aborters. Endothelial cells of affected villi were found to be negative for the thrombomodulin natural anticoagulant pathway, and these cells were reactive with monoclonal antibodies to tissue factor and fibrin. Endothelial cells of normal villi adjacent to affected villi also were negative for thrombomodulin and positive for tissue factor and fibrin, but endothelium of normal villi distant to affected villi were thrombomodulin-positive, tissue factor-negative, and did not contain deposits of fibrin. These findings indicate that a loss of endothelial anticoagulant activity precedes the onset of immunopathology. This observation, coupled with the findings that fetal stem vessels normally lack the heparan sulfate proteoglycanantithrombin III natural anticoagulant pathway and that normal cord blood contains a heparin-like anticoagulant, suggest that a failure of endothelial anticoagulation may be a primary event in the development of placental immunopathology. (Am J Reprod Immunol. 1992; 27:97-100.) Key words: Thrombomodulin, endothelium, placentae, secondary recurrent spontaneous abortion, natural anticoagulant pathways, fibrin, antithrombin III. INTRODUCTION

Spontaneous abortions during the first trimester of pregnancy are frequent complications of human pregnancies.' Some of these are associated with karyotypic anomalies in the embryo.2,3 The remaining are related with different harmful maternal factors such as uterine, endocrine, metabolic, infectious, or immunological disorders." Some of these patients have clinical histories of recurrent spontaneous abortions.v'' Laboratory and clinical studies of many of these patients have tended to define two groups of recurrent spontaneous aborters; these generally are referred to as primary and secondary aborters." It recently has been demonstrated that human placentae have immunocytochemical evidence ofimmunopathology characterized by the presence of many helper T l~mphocytes, activated m~crophages,platel~ts7and reaction products of coagulation and fibrinolysis. -13 These findings are more marked in placentae from secondary recurrent spontaneous aborters than in placentae from either normal deliveries or from primary recurrent spontaneous aborters.v'" We have proposed that these lesions

Submitted for publication December 24,1991: accepted February 24,1992. Address reprint requests to W. Page Faulk, M.D., FRC Path, Methodist Hospital of Indiana, Inc., 1701 N. Senate Boulevard, Indianapolis, IN 46202.

promote damage offetal stem vessel endothelial cells and result in fetal anoxia, hypoperfusion, and nutrient deprivation secondary to thrombus formation within placental stem vessels." To further characterize the impact of these lesions on the placental circulation, we have studied endothelial and coagulation markers in chorionic villi of placentae from secondary aborters. This article is a presentation of these findings. MATERIALS AND METHODS

The placentae from 25 normal term pregnancies and six term pregnancies from women with a previous history of secondary recurrent spontaneous abortion were collected by vaginal delivery at Methodist Hospital, Indianapolis, Indiana. Criteria used for the definition of secondary recurrent spontaneous abortion were those described by McIntyre and colleagues.P The placentae were placed in an ice bath and transported to the Center for Reproduction and Transplantation Immunology within 1 h of delivery. Tissue blocks (1 em x 1 em) of chorionic villi were taken from the central cotyledon. These were washed in 0.01 M, pH 7.2, phosphate buffered saline (PBS) and snap-frozen in liquid nitrogen. Tissues were kept at - 20°C until 4.0 urn sections were prepared from them with the use of a Tissue-Tek cryostat (Elkhart, Indiana). Tissue sections were removed from the cryostat blade by flash condensation onto glass microscope slides and air-dried with an electric fan for 1 h at room temperature. No chemical fixation was used. To determine the extent of villitis in our study population, a section of each block from each placenta was stained with hematoxylin and eosin and examined by conventional light microscopy. The sections were examined for villitis according to criteria used in other studies that employed sections stained with hematoxylin and eosin. 16,17 The percentage of villi with evidence of villitis was determined by studyinu four blocks per placenta according to Knox and Fox. 1 Each tissue section was washed for 30 min in PBS and incubated for 30 min in a moisture box with 50 fl.l of antibody. The antibodies used in this study are listed in Table I. After being incubated with antibody, each section was rinsed in PBS and taken through three 10-min washes in a glass reservoir (7.5 x 10 x 12.5 em) filled with PBS that was constantly mixed with a magnetic stirrer. After the third wash, each section was incubated for 30 min in a moisture box with 50 p.l of F(ab/)2 antibody-conjugates of fluorescein isothiocyanate (FITC) or rhodamine isothiocyanate (RITC). After the final wash, each section was mounted with a coverslip containing PBS-buffered glycerol (60%) adjusted to pH 8.0 with solid TRIS (hydroxymethyl) aminomethane (TRIS, Sigma). Each experiment was accompanied by PBS, antibody, and conjugate controls. These were done by exposing the cryostat sections to only PBS, antibody, or conjugate when

98

LABARRERE AND FAULK

TABLE I. Definition of Antibodies Used in the Present Study Antibody

Specificity

Source

Ig Class

B 183

bovine and human thrombomodulin humanATIII

C. Esmon Oklahoma City, OK Dakopatts Santa Barbara, CA American Diagnostics New York, NY Protos ImmunoResearch San Francisco, CA

Mouse, IgG2ak

1:10

Rabbit

1:300

Mouse, IgG1

1:20

Goat

1:50

Goat

1:40

Antithrombin III (ATIIIl Fibrin FITC goat F(ab'l2 antiRabbit IgG RITC goat F(ab'l2 antiMouse IgG (H&Ll

human fibrin rabbitIgG mouseIgG

Protos ImmunoResearch San Francisco, CA

they were taken through the above immunohistological procedures. Isotype-matched irrelevant primary antibodies also were used as negative controls. The dilution of each antibody used in this investigation is given in Table I. These dilutions were determined by checkerboard titrations against conjugates on tissue sections from different control tissues according to Faulk and Johnson. 19 The working dilution selected for each antibody was defined as two serial dilutions before endpoint titration. These procedures were performed at room temperature. Before application, each antibody was ultracentrifuged IOO,OOOg at 4°C for 1 h. Titration and ultracentrifugation were done for each antibody at least every 6 wk, and all newly received antibodies were ultracentrifuged and titrated. All sections were studied by epi-illumination and interference optics according to Faulk and Hijmans.f" A Lei tz microscope fitted with an HBO-IOO mercury-arc lamp was used. The Ploem epi-illuminator in this microscope contained an 1-2 type FITC filter complex that consisted of a 450-490-nm excitation filter, a 5IO-nm dichroic mirror, and a 515-nm barrier filter that allowed wavelengths greater than 515 nm to pass. The epi-illuminator also contained an N2.I-type RITC filter complex that consisted of a 515-560-nm excitation filter, a 580-nm dichroic mirror, and a 580-nm barrier filter that allowed 580 nm wavelengths to pass. The microscope was connected to a Leitz camera that contained ABA 200 daylight 35 mm Ektachrome film. After the immunofluorescence studies, the sections were stained with hematoxylin and eosin and examined for histopathological evidence of villitis.

Dilution

its importance from the well-established observation that thrombomodulin is one of the three known natural anticoagulant pathways.f" The other two natural anticoagulant pathways of endothelium are (1) dermatan sulfate-heparin cofactor 11,23 and (2) heparan sulfate proteoglycan-antithrombin 111. 24 ,25 We have found the dermatan sulfate-heparin cofactor II pathway to be represented on placental endothelium.f" In contrast, we have not been able to identify elements of the heparan sulfate proteoglycan-antithrombin III natural anticoagulant pathway on plasma membrane of fetal stem vessel endothelium (Fig. 2). A summary of the status of endothelial natural anticoagulant pathways in placentae from normal pregnancies and secondary recurrent spontaneous aborters is presented in Table II. The availability of tissue factor on endothelial cells associated with the disappearance or downregulation of thrombomodulin converts the normally thromboresistant endothelial plasma membranes to thrombogenic surfaces. 2 6-28 In order to test for the presence of this conversion from thromboresistance to thrombogenicity in placental stem vessel endothelium, we have studied fetal stem vessel endothelial cells for the presence of fibrin in placentae from normal pregnancies and compared this result with our findings of endothelial fibrin in placentae from secondary recurrent spontaneous aborters. Endothelial fibrin rarely was identified in placentae from

RESULTS

We previously have reported that the immunopathological areas of villitis of unestablished etiology are significantly increased in villi of placentae from mothers with clinical histories of secondary recurrent spontaneous abortions.P The percentage ofthese lesions in term placentae from patients with a history of secondary recurrent spontaneous abortions was 11% ± 7.1 compared with a value of 1.8% ± 0.8 from normal term placentae, giving a P value of 0.05. 1 4 When we studied normal chorionic villi around areas of villitis, we detected a significant increase in fetal stem vessels with endothelial cells that were reactive with monoclonal antibody to tissue factor," We now are able to report that these morphologically normal villi from secondary aborters' placentae that manifest tissue factor in fetal stem vessel endothelium have either lost or downregulated their endothelial thrombomodulin (Fig. 1). The significance of this finding derives

Fig. 1. Cryostat section of snap-frozen unfixed chorionic villi following incubation with monoclonal antibody to human thrombomodulin and rhodamine-labeled second antibody. Note fetal stem vessels in normal villi adjacent to an area of chronic villitis (V) are not reactive (arrows). x 250.

PLACENTAL ENDOTHELIAL CHANGES

99

TABLE III. Vessels with Fibrin in Normal Chorionic Villi Adjacent to Areas of Immunopathology Placentae Normal 2°RSN

Villi with endothelial fibrin

Vessels/villus with endothelial fibrin

3.3%a 50%

10%b 100%

'Number of villi with endothelial fibrin/lOO villi. "Vessels per villus with endothelial fibrin. "Secondary recurrent spontaneous aborters.

Fig. 2. Cryostat section of snap-frozen unfixed chorionic villi reacted with polyclonal antibody to antithrombin III and fluorescein-labeled second antibody. Note fetal stem vessels (F) are not reactive. Focal reactivity is observed in an area of fibrinoid necrosis (arrow). x 250.

normal pregnancies (Table III). In contrast, fetal stem vessel endothelium commonly contained fibrin in placentae from secondary aborters (Fig. 3). DISCUSSION

Areas of immunopathology defined by immunocytochemical criteria have been reported in normal human placentae.Y'" These same areas have been described for many years by light microscopists as areas of villitis of unestablished etiology. 16-18 Results of histological studies from many laboratories have confirmed the presence of villitis in normal term placentae, and results of quantitative studies have revealed less immunopathology in normal placentae than in placentae from abnormal pregnancies.f In our experience, immunocytochemical studies of these areas have shown the presence of activated macrophages and CD4-reactive T lymphocytes associated with components of the hemostatic and fibrinolytic systems. 6,7,13

of normal chorionic villi not associated with areas of immunopathology in the same placentae were thrombomodulin-positive and tissue factor-negative, and did not contain deposits of fibrin. This raises the possibility that thrombogenic changes in natural anticoagulant pathways are causally related to fibrin deposition and thrombus formation. How do these thrombogenic changes occur within chorionic villi without evidence of inflammation? The spiral (i.e., uteroplacental) arteries in recurrent or spontaneous abortions do not demonstrate the physiological change of trophoblast invasion; thus, they retain their musculoelastic coats and are capable of vasoconstrictive response that can cause placental ischemia. 29,3o Ischemia produces free radicals that damage endothelial cells and predispose them to thrombus formation." Such thrombi extend ischemic changes that release leukotactic molecules that attract leukocytes.f and these are the predominant cells in villitis areas. 6 •7.21 Among these cells are activated macrophages, which release cytokines such as interleukin-l (lL-l) and tissue necrosis factor that make tissue factor available on endothelial cells, downregulate thrombomodulin, and convert endothelial plasma membranes from thromboresistant to thrombogenic surfaces.ll.33.34 The downregulation of thrombomodulin is especially important in placentae when one considers that, unlike other organs, placental endothelial cells do not normally contain the heparan sulfate proteoglycan-antithrombin III natural anticoagulant pathway. We have suggested that

TABLE II. Endothelial Natural Anticoagulant Pathways in Human Placentae Placentae Normal 2° RSAb

TM a

ncn-

+

+

HSPG-ATIII a

?

'The natural anticoagulant pathways are thrombomodulin (TM), heparin cofactor II (Hem, and heparan sulfate proteoglycan-antithrombin III (HSPG-ATIIl). bSecondary recurrent spontaneous aborters.

Endothelial cells in these areas of immunopathology are thrombomodulin-negative, tissue factor-positive, and show the presence of fibrin thrombi.9-12 We have proposed that these inflammatory lesions are secondary to endothelial damage and thrombosis." However, in the present study we found that fetal stem vessels in normal chorionic villi adjacent to areas of villitis also were tissue factor-positive, thrombomodulin-negative, and contained deposits of fibrin (Table III). In addition, vessels

Fig. 3. Cryostat section of snap-frozen unfixed chorionic villi reacted with monoclonal antibody to fibrin and rhodamine-labeled second antibody. Note many fetal stem vessels are fibrin-reactive (arrows). These villi are adjacent to an area of villitis. Fetal stem vessels in normal chorionic villi (N) not adjacent to villitis areas are negative. x 250.

100

LABARRERE AND FAULK

the absence of this endothelial natural anticoagulant pathway is associated with the presence of heparin-like molecules in the fetal blood.i" and this suggestion is supported by a report of a heparin-like inhibitor of coagulation present in cord blood. 3 5 ,36 Although the possibility of thrombi formation in fetal stem vessels following damage by immune cells is not excluded, a deficiency of heparin-like molecules in fetal blood also could predispose to thrombosis in fetal stem vessels. Both of these mechanisms are currently being investigated in our laboratory. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

REFERENCES Edwards RG. Causes of early embryonic loss in human pregnancy. Hum Reprod. 1986; 1:185-198. Boue JG, Boue A. Chromosomal anomalies in early spontaneous abortion. Current Topics in Pathology. 1976; 62:193-208. Plachot M. Chromosome analysis of spontaneous abortion after IVF. A European survey. Hum Reprod. 1989; 4:425-429. Faulk WP, Coulam CB, McIntyre JA. Recurrent pregnancy loss. In: Seibel MM, ed. Infertility. A Comprehensive Text. Norwalk, Ct: Appleton and Lange, 1990; 273-284. McIntyre JA, Coulam CB, Faulk WP. Recurrent spontaneous abortion. Am J Reprod Immunol. 1989; 21:100-104. Labarrere CA. Allogeneic recognition and rejection reactions in the placenta. Am J Reprod Immunol. 1989; 21:94-99. Labarrere CA, McIntyre J A, Faulk WP. Immunohistologic evidence that villitis in human normal term placentas is an immunologic lesion. Am J Obstet Gynecol. 1990; 162:515-522. Labarrere CA, Faulk WP, McIntyre JA. Villitis in normal term human placentae: Frequency of the lesion determined by monoclonal antibody to HLA-DR antigen. J Reprod Immunol. 1989; 16:137 -150. Labarrere CA, Carson SD, Faulk WP. Tissue factor in chronic villitis of un established etiology. J Reprod Immunol. 1991; 19:225-235. Faulk WP, Labarrere CA, Carson SD. Tissue factor: Identification and characterization of cell types in human placentae. Blood. 1990; 76:86-96. Faulk WP.Placental fibrin. AmJ Reprod Immunol. 1989; 19:132-135. Labarrere CA, Esmon CT, Carson SD, Faulk WP. Concordant expression of tissue factor and class II MHC antigens in human placental endothelium. Placenta. 1990; 11:309-318. Labarrere CA, Faulk WP. Fibrinolyses in normal and abnormal chorionic villi. Am J Reprod Immunol. 1991; 25:66. Labarrere CA, Faulk WP. MHC class II reactivity of human villous trophoblast in chronic inflammation of unestablished etiology. Transplantation. 1990; 50:812-816. McIntyre JA, McConnachie PR, Taylor CG, Faulk WP. Clinical, immunologic and genetic definitions of primary and secondary recurrent spontaneous abortions. Fertil Steril. 1984; 42:849-855. Labarrere C, Althabe 0, Telenta M. Chronic villitis of unknown aetiology in placentae of idiopathic small for gestational age infants. Placenta. 1982; 3:309-318. Russell P. Inflammatory lesions of the human placenta: III. The

18. 19. 20. 21.

22. 23.

24.

25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

histopathology of villi tis of unknown aetiology. Placenta. 1980; 1:227-244. Knox WF, Fox H. Villitis of unknown aetiology: Its incidence and significance in placentae from a British population. Placenta. 1984; 5:395-402. Faulk WP, Johnson PM. Immunological studies of human placentae: Identification and distribution of proteins in mature chorionic villi. Clin Exp Immunol. 1977; 27:365-375. Faulk Wp, Hijmans W. Recent developments in immunofluorescence. Prog Allergy. 1972; 16:9-39. Labarrere CA, Faulk WP. Immunopathology of extraembryonic tissues in abnormal pregnancies. In: Andreani D, Bompiani GD, DiMario U, Faulk WP, Galluzzo A, eds. Immunobiology of Normal and Diabetic Pregnancies, John Wiley & Sons, 1990:165-186. Esmon CT. The regulation of natural anticoagulant pathways. Science. 1987; 235:1348-1352. Ofosu FA, Modi GJ, Smith LM, Cerskus AL, Hirsh J, Blajchman MA. Heparan sulfate and dermatan sulfate inhibit the generation of thrombin activity in plasma by complementary pathways. Blood. 1984; 64:742-747. Absher E, Labarrere CA, Carter C, Haag B, Faulk WP. The endothelial heparan sulfate-antithrombin III natural anticoagulant pathway in normal and transplanted human kidneys. Transplantation. 1992; 53:828-834. Faulk WP, Tollefsen DM, Labarrere CA. Anticoagulant pathways in human placentae. Am J Reprod Immunol. 1991; 25:66-67. Labarrere CA, Faulk WP. Anchoring villi in human placental basal plate: Lymphocytes, macrophages and coagulation. Placenta. 1991; 12:173-182. Faulk WP, Gargiulo P, McIntyre JA, Bang NU. Hemostasis and fibrinolysis in renal transplantation. Sem Thromb Hemost. 1989; 15:88-98. Rosenberg RD, Rosenberg JS. Natural anticoagulant mechanisms. J Clin Invest. 1984; 74:1-6. Khong TY, Liddell HS, Robertson WB. Defective haemochorial placentation as a cause of miscarriage. A preliminary study. BrJ Obstet Gynaecol. 1987; 94:649-655. Hustin J, Jauniaux E, Schaaps JP. Histological study of the matemoembryonic interface in spontaneous abortion. Placenta. 1990; 11:477-486. Zweier JL, Rayburn BK, Flaherty JT, Weisfeldt ML. The effect of superoxide dismutase on free radical concentrations in post-ischemic myocardium. Circulation. 1986; 74(suppl):371. Petrone WF, English DK, Wong K, McCord JM. Free radicals and inflammation: Superoxide dependent activation of a neutrophil chemotactic factor in plasma. Proc Nat! Acad Sci USA. 1980; 77:1159-1163. Cotran RS. New roles for the endothelium in inflammation and immunity. AmJ Patho1.1987; 129:407-413. Pober JS. Cytokine-mediated activation of vascular endothelium. Physiology and pathology. Am J Pathol. 1988; 133:426-433. Muller AD, van Doorm JM, Hemker HC. Heparin-like inhibitor of blood coagulation in normal newborn. Nature. 1977; 267:616-617. Faulk Wp, Labarrere CA, Xiao HY, Bang NU. Anti-coagulatory pathways in human feto-placental unit. Am J Reprod Immunol 1992; 27:43.