Cover Image


View/Hide Left Panel
Click for next page ( 216

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement

Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 215
19 Reproductive Immunology: Biologic Markers of Compromised Pregnancies Women experiencing repeated pregnancy losses have been reported in all popula- tions (Bloom, 1981~. Many of these losses are probably not due to exposures to toxic chemicals. Therefore, it is important to be able to distinguish women experienc- ing recurrent pregnancy losses because of toxic effects from those experiencing losses ~ or other reasons. This chapter discusses assessments that distinguish some mechanisms of spontaneous abortion. The chapter also discusses other immuno- logic assessments that might be used to characterize the changes that occur during normal pregnancies. These changes might be altered by toxicants, thereby causing a pregnancy loss. MATERNAL IMMUNOLOGIC RECOGNITION AND REACTION DURING NORMAL PREGNANCY Efforts to understand the immunology of human pregnancy have focused on extra- embryonic membranes, because the point of contact between maternal tissues and the conceptus is the trophoblast. Cells of the inner cell mass differentiate to become the embryo, and extraembryonic components form an interface with maternal blood and uterine cells (Faulk and McIn- tyre, 1983~. This materno-trophoblastic - 215 interface exists at all sites of contact, including placenta, point of contact of the amnion and chorion (the amniochorion), spiral (uteroplacental) arteries, basal plate, and interstitial tissues. The trophoblast is strategically impor- tant in potential maternal immune recogni- tion and rejection. Its plasma membranes are unique, inasmuch as none expresses the polymorphic form of class I or class II transplantation histocompatibility antigens (human leukocyte antigens or hLAs); however, some cytotrophoblasts are reactive with monoclonal antibodies thought to recognize class I hLA. The gen- eral lack of transplantation antigens has led to speculation of trophoblastic im- munologic neutrality; however, numerous investigators have shown that trophoblast membranes are not immunologically inert (Faulk and Hsi, 1983~. The immunogens that signal and maintain maternal recognition are extraembryonic structures at the ma- terno-trophoblastic interface, i.e., trophoblast antigens (Faulk and McIntyre, 1981~. The immune system in mammals consists of B lymphocytes, which are thought to arise from bone marrow and are responsible for antibody immunity, and T lymphocytes, which emerge from the thymus and are re- sponsible for cell-mediated immunity.

OCR for page 215
216 Both lymphocyte populations are activated during normal human pregnancy. B-lymphocyte activation by trophoblast antigens has been shown by elusion of ma- ternal antitrophoblast antibodies from homogenates of individual placentae (Faulk et al., 1974~. T-lymphocyte activa- tion by trophoblast antigens has been con- firmed by demonstrating that trophoblast antigens cause maternal lymphocytes to release a lymphokine (migratory inhibi- tion factor, or MIF), which is a quantita- tive measure of cell-mediated immunity (Rocklin et al., 1982~. An early step in the generation of cyto- toxic reactions in cell-mediated immunity is allogeneic recognition of the target cell. Blocking allogeneic recognition by B-cell-produced antibodies to tropho- blast can inhibit T-cell activation and the succeeding steps that result in cell death. Antibodies that impede such immune reactions are called blocking antibodies. Ample evidence from studies done in mice and human beings shows the presence of blocking antibodies in maternal blood and placental eluates during normal pregnancy (Faulk et al., 1974; Rocklin et al., 1982~. Some trophoblast antigens (McIntyre and Faulk, 1979a) and some antitrophoblast antibodies (McIntyre and Faulk, 1 979b) can modulate allogeneic recognition reac- tions. Although antitrophoblast antibod- ies have been identified in some normal and abnormal pregnancies (McIntyre et al., 1 984a), antitrophoblastic activity usual- ly cannot be identified in serum taken during a normal pregnancy (Davies and Browne, 1985~. Antitrophoblast antibodies are not always found in serum taken during a preg- nancy, for at least five reasons: Antibody-combining sites might be bound by trophoblast antigens in immune complexes. Trophoblast immunogens might stimu- late blocking or incomplete antibody pro- duction. Autoanti-idiotypic antibodies might prevent antitrophoblast immunity. Inhibitors of antigen-antibody reac- tions or of their products (such as comple- ment fixation) might mask the presence of antibodies. TOXICI7~YDURING PREGNANCY Antitrophoblast antibodies might not always be present in serum taken during a pregnancy. A proposed model for- immunologic re- sponse is presented in Figure 19-1. Stud- ies of human serum taken during pregnancy have revealed the presence of circulating immune complexes (i.e., antigen in com- bination with its antibody) comprising five biochemically identifiable tropho- blast and antitrophoblast components, two of which also can be identified at sig- nificantly lower concentrations in serum of nulliparous, nonpregnant women (Da- vies, 1985~. Increased quantities of trophoblast antigens have been reported in maternal serum as pregnancy progresses (Faulk and McIntyre, 1983), and this has been cited as further support for immune complexes containing trophoblast antigen. Immune complexes might account for the difficulty in demonstrating antitropho- blast immunity in some maternal serum, because the antitrophoblastic components are captured in these complexes (Davies, 1985~. Blocking or incomplete antibodies have been suspected of being important manifestations of immunity in reproduc- tion and cancer research (Gorer et al., 1959; Voisin et al., 1972~. Whether the antigens responsible for host responses to transplants, cancers, tolerance induc- tion, and pregnancy have chemical similar- ities is not known, but blocking or incom- plete antibodies have been described in all these conditions. In pregnancy research, human tropho- blast antigens and antibodies to them have been shown to block mixed-lymphocyte- culture (MLC) reactions (McIntyre and Faulk, 1 979a,b). Maternal antipaternal blocking immunity appears to be important in normal pregnancy, but is absent in some abnormal pregnancies (Rocklin et al., 1982; McIntyre et al., 1984a). Experiments in mice have shown that immunization with placental extracts and an additional, unrelated antigen promotes production of more blocking and less cytotoxic anti- bodies (Due et al., 1985~. Thus, blocking activity in maternal blood might explain the lack of antitrophoblast immunity in vitro.

OCR for page 215
REPRODUCTIVE IMMUNOLOGY ANT I - I D I OTY PE ( Ab2) **,? ANTI -TLX (Ab1 ) T LX ANT I GEN (Seminal plasma) (T rophoblast antigen) 217 CONTROL Cl RCUIT IN HUMAN PREGNANCY (TLX /Anti-TLX /Anti-anti-TLXJ ANTI-TLX-ANTI-IDIOTYPE (Abl-Ab2) - - ANTI-IDIOTYPIC CONTROL ~ / \ Ab 1)' (/b2 FIGI~E 19-1 Proposed model of immunologic response. Source: Faulk et al., 1987. Autoanti-idiotypic antibodies to anti-hLA antibodies in recipients of do- nor-specific transfusions have been found in serum of patients who lack sero- logic responses to hLA; that indicates that the presence or absence of detectable antibodies might reflect the amount of anti-idiotype (Reed et al., 1985~. Serum from people alloimmunized through preg- nancy, transfusion, or transplantation can react with autologous T lymphoblasts primed against the immunizing donor (Su- ciu-Foca et al., 1983~. Coupled with the finding that primed T cells display idiotypelike receptors for alloantigens, those observations have prompted the idea that the receptors induce formation of anti-idiotypic antibodies. Such autoanti - idiotypic antibodies to hLA can be found during and after pregnancy (Reed et al., 1983~. Maternal anti-idio- typic antibodies to trophoblast antigens have not been studied in human pregnancies, but they should help to explain why anti- trophoblast antibodies cannot always be found in maternal serum. Inhibitors of antigen-antibody reac- tions in complement-dependent assays (Torry et al., 1986) and complement-in- dependent assays (McIntyre and Faulk, 1985) have been described. Inhibitors of inhibitors have also been noted (Faulk 1 et al., 1 989a) that are heat-labile, sen- sitive to calcium concentrations, de- stroyed by Russell's viper venom, and absent from the plasma of patients with deficiency in clotting factor V. Some cytotrophoblasts in placental villi react with antibodies to factor V; factor V might play a role in modulating maternal antigen- antibody interactions within the placen- tal bed. Knowledge of the inhibitors and inhibitors of inhibitors of antigen- antibody reactions is only beginning to emerge; these inhibitors might be impor- tant factors when maternal antitropho- blast antibodies are not detected, par- ticularly if a sample is heated, an incor- rect anticoagulant is used, or the calcium concentration is erroneous. There seems to be a novel and unexplored link between blood clotting and immunologic reactions. Another possible explanation for the inability to detect maternal antitropho- blast antibodies in maternal serum taken during pregnancy is that the antibodies are not present. Antibodies were not thought to be a response to transplantation until Gorer et al. (1959) developed methods to reveal them; and antibodies were not thought to be an aspect of neonatal tol- erance until Voisin et al. (1972) demon- strated their presence. In many cases, what was thought to be a lack of serologic . .

OCR for page 215
218 reactivity to hLAs has been found to be an inhibition of reactivity by pregnancy- induced autoanti-idiotypic antibodies (Reed et al., 1983; Suciu-Foca et al., 1983~. That matter will be clarified by time and more research, but it is reason- able to assume that all mothers have an- tibodies to trophoblast antigens in their blood (Faulk et al., 1978~. TESTS TO DETERMINE MARKERS OF MECHANISMS OF RECURRENT PREGNANCY LOSS Infertility is classified as primary or secondary. Primary infertility occurs in women who have never conceived, and secondary infertility occurs in women who have conceived but have failed to conceive during 1 or more years of intercourse with- out contraception (Coulam, 1981~. Recur- rent spontaneous abortion comprises at least two groups of persons designated as primary and secondary spontaneous abor- ters (McIntyre et al., 1984a). Repeated pregnancy loss in primary aborters is thought to be associated with compatibil- ity of trophoblast-lymphocyte-cross- reactive (TLX) antigens between mating partners, which results in lack of neces- sary protective or blocking maternal re- sponses during pregnancy (McIntyre et al., 1986~. The trophoblast antigens that eli- cited antibodies to cross-reacting anti- gens on lymphocytes were designated TLX antigens (Faulk and McIntyre, 1981~. Mixed-Lymphocyte-Culture Reactions One of the best examples of cell- mediated immunity in pregnancy is the MLC reaction, in which lymphocytes from two persons are mixed and cultured under conditions that permit measurement of their DNA metabolism, which is an index of the intensity of reaction of one cell to the other. If the father's cells are sufficiently irradiated to prevent their immunologic response but their capacity to stimulate the mother's lymphocytes is retained, then the reaction is called a one-way MLC reaction. This test assesses the reaction of the mother's cells to the father's cells. TOXICI7~YDURING PREGNANCY Perhaps the most convincing support for an important role of blocking factors in pregnancy comes from clinical inves- tigations done in some conditions of abnor- mal pregnancies, particularly unexplained spontaneous abortions. Often, a primary spontaneous aborter does not produce a blood factor that blocks lymphocytes from reacting with her mate's cells in in vitro models of cell-mediated immunity. In some cases, that deficiency can be overcome by immunizing the woman with lymphocytes (Taylor et al., 1985~. Lymphocytotoxic Antibodies and Histocompatibility Typing The presence of lymphocytotoxic anti- bodies in maternal serum never has been explained adequately, because these an- tibodies are not always identified in ma- ternal serum (Kajino et al., 1988~. When they are detected, it usually is in serum collected during a second or later pregnan- cy, although such antibodies have been reported in samples from first pregnancies (Davies and Brown, 1985~; lymphocyto- toxins probably are not of central impor- tance in the immunobiology of human preg- nancy. Lymphocytotoxins of pregnancy often are of broad reactivity and might react with antigens common to several class I hLAs (Konoeda et al., 1986~. Lymphocytotoxins in maternal serum taken during pregnancy traditionally have been characterized as anti-hLA. That clearly is not the case in secondary spontaneous abortion, inasmuch as cyto- toxicity is removed from serum by absorp- tion with hLA-negative trophoblast or platelets unrelated to hLAs (Faulk and McIntyre, 1986~. During the last several years, inves- tigators have theorized that normal preg- nancy requires maternal immunologic rec- ognition of the TLX antigens inherited by the conceptus (Faulk and McIntyre, 1981 ) and that failure of recognition or inap- propriate recognition results in faulty blastocyst implantation and ultimately spontaneous abortion. The nature of the immunogen is speculative, and results of animal-model studies have prompted some investigators to suggest that maternal

OCR for page 215
REPRODUCTIVE LUMUNOLOGY recognition depends on incompatibility of major histocompatibility complex (MHC) antigens (Kiger et al., 1985~. That sug- gestion is supported by the comparative success of outbred matings (as opposed to inbred matings), by the benefit of al- logeneic, third-party leukocyte immuniza- tions to primary aborters (McIntyre et al., 1986), and by the demonstration of a beneficial effect of mating with an MHC- incompatible male on pregnancy outcome in mice (Clark et al., l 986~. The role of MHC antigens in defining incompatibility has been controversial (Palm, 1974; Komlos et al., 1977; Gerencer et al., 1979; Schacter et al., 1979; Gill, 1983; Thomas et al., 1985~. Reports have supported (Taylor and Faulk, 1981; Beer et al., 1981; McIntyre and Faulk, 1983; Unander and Olding, 1983) and refuted (Lauritsen et al., 1976; Caudle et al., 1983; Mowbray et al., 1983; MacQueen and Sanfilippo, 1984; Jeannet et al., 1985) an association between hLAs and reproduc- tive performance. The variation in results can be explained by small sample sizes and lack of homogeneity of the populations investigated. Controversy involving association of hLAs and reproductive per- formance could be resolved in part by pro- perly classifying recurrent spontaneous aborters and inexplicably infertile cou- ples. Immunologists often liken pregnancy to an allogeneic graft, inasmuch as the developing conceptus expresses paternal genes that are foreign to the mother. Men and women who share hLAs might be expected to be ideal reproducers, because chances for immunologic recognition and rejection by the mother would tee low. However, genet- ic identity between mother and conceptus does not appear to be necessary for the pregnancy to continue; surrogate mothers and in vitro fertilization techniques have shown that a fertilized egg can develop successfully if it is genetically differ- ent from the recipient. Furthermore, cou- ples that share hLAs often are not able to have successful pregnancies. Women suffering from primary recurrent spontaneous abortions often have hLA (TLX) profiles more similar to those of their mates than is the case in normal childbear- 219 i] ng couples (McIntyre et al., 1986~. Those women do not manifest antipaternal humoral immunity, and their cell-mediated re- sponses often are absent or suboptimal, as measured by mother-father MLC reac- tions. The cellular deficit appears to be intrinsic, in that it occurs regardless of the serum supplement used in the culture system. The deficit can be used to distin- guish primary from secondary recurrent spontaneous abortion, because secondary aborters have extrinsic or acquired MLC- inhibitory activity (McIntyre and Faulk, 1983~. Autoimmunity Tests Spontaneous abortion is common in pa- tients with some autoimmune diseases, such as systemic lupus erythematosus (Derue et al., 1985), although how antibod- ies associated with these diseases inter- rupt pregnancies is unknown. The lupus anticoagulant is an antibody that reacts with the phospholipid component of pro- thrombinase (Hougie, 1985) and affects many in vitro blood coagulation tests (Sha- piro and Thiagarajan, 1982~. Preeclamp- sia and fetal growth retardation, which . sometimes appear in pregnancies compli- cated by lupus anticoagulant, are associ- ated with decreased prostacyclin produc- tion by maternal and fetal vascular tissues (Bussolino et al., 1980; Remuzzi et al., 1980~. Identification of lupus anticoagu- lant in maternal blood taken during preg- nancy should signal a high risk of pregnan- cy loss. PROMISING MARKERS OF MATERNAL-FETAL INTERACTIONS Trophoblast Antigens and Classes of Couples with Recurrent Spontaneous Abortions Trophoblast antigen- 1 (TA 1) and tropho- blast antigen-2 (TA2) were among the first human trophoblast antigenic groups to be identified with the use of polyclonal an- tisera (Faulk et al., 1978). Other poly- clonal and monoclonal antibodies have been used to define several trophoblast anti- gens (Faulk and Hsi, 1983). The study of

OCR for page 215
220 such antibodies is essential for building an understanding of trophoblast antigen functions in mammalian pregnancies, be- cause trophoblast tissues account for the operational interfaces between maternal and extraembryonic cells in the allogeneic relationship of human pregnancy (Faulk, 1983). Antisera to TA 1 (anti-TA 1 ~ contain an- tibodies to syncytiotrophoblast and sub- populations of cytotrophoblast found in the amniochorion, basal plate (Wells et al., 1984a), spiral arteries (Wells et al., 1984b), and uterine interstitial tissues (Hsi et al., 1984a). In contrast with their lack of reactivity with normal somatic cells, anti-TA1 reacts with many human-transformed cell lines (Faulk et al., 1979~. Normal, abnormal, and extra- embryonic tissues from other species do not react with human anti-TA 1. Anti-TA 1 inhibits MLC reactions without affecting nonspecific mitogenic stimula- tion of lymphocytes (McIntyre and Faulk, 1979b). The mechanism of MLC inhibition seems to be recognition and stimulation, rather than T-cell proliferation (Mc- Intyre and Faulk, 1979c). All normal human extraembryonic tissues have TA 1 at the materno-trophoblastic interface, including the amniochorion (Hsi et al., 1982~. Two immunohistologic observations of the amniochorion have been made: the cytotrophoblast forms a barrier having TA 1 that is three to five cells thick (Faulk et al., 1982), and the amniot- ic epithelium does not have TA1 but ex- presses an antigen called amniotic anti- gen-3 (AA3) (Hsi et al., 1984b). In new- borns with epidermolysis bullosa letalis (EBL; also called polydysplastic epider- molysis bullosa), the barrier of TA 1 - bearing cytotrophoblast is so thin that maternal uterine cells sometimes come into contact with amniotic epithelium-derived tissues (Faulk et al., 1988b). In EBL, am- niotic epithelial cells react with TA 1 and AA3 antibodies. The only other cir- cumstance in which anti-TA1 reacts with amniotic epithelial cells is when they are transformed (Yeh et al., 1984~. Am- niotic epithelium is representative of somatic ectoderm (Faulk and McCrady, 1983~. Babies with EBL have defective TOXICI7~YDURING PREGNANCY ectodermal derivatives, such as junc- tional epidermolysis bullosa, and defec- tive ectodermal thymus. The absence of AA3 from skin biopsies is a useful pre- natal diagnostic criterion for EBL (Kennedy et al., l985~. TA2 was established on the basis of three observations: So lub ilized tro p ho b last me mb ranes were partitioned into two peaks (TAT and TA2) with chromatography. The antigens responsible for gener- ating lymphocytotoxic antibodies were present in the second (TA2) peak of solu- bilized, chromatographed trophoblastic microvilli. Lymphocytotoxic antibodies were re- moved from antitrophoblast serum by ab- sorption with lymphocytes (Faulk et al., 1978~. Absorption of anti-TLX antigens with trophoblast membranes from different placentas has indicated that these TLX antigens are allotypic (Faulk et al., 1980; McIntyre and Faulk, 1982~. At least three groups of TLX antigens can be found with the use of rabbit antibodies (McIntyre et al., 1984b). One TLX antigen in the fer- tilized egg must be incompatible with the mother and be capable of signaling allogeneic recognition and immunologic protection through the generation of ma- ternal blocking antibodies and suppressor cells (McIntyre and Faulk, 1985~. Maternal recognition of paternal TLX can be initiated by its immunogenicity in seminal plasma (Faulk and McIntyre, 1986~. Failure of recognition or inappro- priate recognition of TLX antigens could result in spontaneous abortion that occurs so early in the pregnancy that it goes undetected and the couple is thought to be infertile (Faulk and McIntyre, 1986~. Major Basic Protein Major basic protein (MBP) is a protein whose blood concentrations rise by the sixth week of gestation and return to nor- mal by 6 weeks after birth (Maddox et al., 1984~. Pregnancy-associated MBP can be purified from human placentas and is bio-

OCR for page 215
REPRODUCTIVE IMMUNOLOGY chemically indistinguishable from MBP found in eosinophil granules. MBP, which accounts for most of the granule protein (M.S. Peters et al., 1986), is toxic to mammalian cells in vitro (Gleich et al., 1979) and parasites in viva (Kephart et al., 1984) and mediates inflammation in asthma (Frigas and Gleich,1986~. In human pregnancy, increases in MBP in peripheral blood are independent of either eosino- phils or eosinophil proteins other than MBP(Maddoxetal.,1983~; immunohistolog- ic techniques show increases to be local- ized in the extravillus trophoblast (Maddox et al., 1983~. Quantitative studies have indicated that MBP concentrations plateau by 20 weeks of gestation at more than 10 times the nonpregnant value, and they rise sharp- ly in the third trimester in women who ex- perience a spontaneous onset of labor (Wasmoen et al., 1987~. The late increase, which accounts for 40% of the total in- crease in MBP during pregnancy, begins at least 3 weeks before onset of labor. Women who experience preterm labor have a similar increase; those with oxytocin- induced labor do not, nor do those with prolonged gestation (Coulam et al., 1987~. These observations suggest that MBP concentration is a marker of the onset of labor. Early Pregnancy Factor Early pregnancy factor (EPF) is an im- munosuppressive molecule that increases the ability of antilymphocyte antibodies to inhibit active, spontaneous rosette formation between lymphocytes and red cells (Rolfe et al., 1984~. EPF is produced by the mother within 24 hours of fertiliza- tion, and it wanes in midpregnancy, at which time its function is assumed by a placental form of EPF. The functions of maternal EPF and embryonic EPF seem to be indistinguishable. In the mother, the molecule is assembled from an oviduct com- ponent (EPF-A) and an ovarian component (EPF-B) and is synthesized under the in- fluence of prolactin and an ovum factor (Cavanaghetal., 1982~. Because EPF appears in maternal blood so soon after implantation, it provides a method to distinguish infertile couples 221 from those who are becoming pregnant but aborting very early. It might also be use- ful in judging success of in vitro fertili- zation/embryo transfer programs. During pregnancy, the molecule is pres- ent only when there is a viable embryo and could be used as an early marker of em- bryonic viability (Morton et al., 1982~. Detection of EPF in a nonpregnant patient suggests a tumor of germ-cell origin (Rolfe etal., 1983~. PROMISING TECHNIQUES THAT MIGHT YIELD BIOLOGIC MARKERS Fluorescence-Activated Cell Sorting In the past several years, molecular biology, genetics, and immunology have converged and produced mutually advanta- geous techniques, such as molecular probes, gene cloning, and hybridoma forma- tion. One of the most promising techniques for pregnancy immunology is fluorescence- activated cell sorting (FACS). FACS uses fluorochrome-labeled antibodies to iden- tify and quantify membrane markers on cells in a heterogeneous mixture (Parks et al., 1979~. For example, FACS has been used to measure the flux of fetal cells into mater- nal blood during pregnancy. The technique can be used to isolate immunologically marked cells from a complex mixture, such as blood, and it has been used to measure trophoblast membranes in the peripheral circulation of pregnant women (Kawata et al., 1984~. This approach might make it possible to harvest fetal cells for cyto- genetic investigations without resorting to the more invasive techniques of amnio- centesis or chorionic villus biopsy. Immunotherapy to Prevent Spontaneous Abortion Immunotherapy to prevent primary spon- taneous abortions was begun in 1979 (Taylor and Faulk, 1981~. Primary aborting women were given leukocyte transfusions from nonpaternal blood donors (Taylor et al., 1985~; the rationale was that non- paternal leukocytes would express TLX allotypes foreign to the mother and cause her to mount a protective anti-TLX response to the infused cells that cross-

OCR for page 215
222 reacted with TLX antigens on the blasto- cyst, thereby protecting the developing embryo from maternal immune rejection (Faulk and McIntyre, 1981, 1983; Beer et al., 1986~. More than 45 couples receiving this immunotherapy had a rate of successful pregnancy comparable with that of normal childbearing women (Mowbray, 1987~. No graft-versus-host reactions were observed in any of the offspring (Mowbray and Under- wood, 1985; McIntyre et al., 1986), and no evidence of intrauterine growth retar- dation was found. Immunopharmacology Immunopharmacology brings a promising new approach to the study of biologic mark- ers in pregnancy by joining two previously unconnected fields of medical investiga- tion, particularly in relation to toxicol- ogy and environmental pollutants. Immuno- pathologic reactions can be grouped into four types of causal mechanisms that are central to understanding diagnostic, therapeutic, and prognostic variables of immunologic disease. Careful clinical identification of immunologic reactions and failed pregnancies in association with environmental exposures might help to clarify the action of toxic chemicals. Type I lesions are mediated by immediate hypersensitivity reactions-such as ana- phylaxis, urticaria, and angioedema and usually can be attributed to a reaginic (IgE) antibody that fixes to receptors on tissue mast cells and blood basophils. When the reagin meets its antigen-which can be a drug, such as penicillin, or an environmental allergen, such as ragweed pollen-the mast cells or basophils degran- ulate and release mediators (such as his- tamine) that cause the signs and symptoms of immediate hypersensitivity reactions. Patients with preeclampsia have been found to have significant elevations of IgE in their blood (Alanen, 1984~. Type II immunopathologic reactions are caused by direct interaction of anti- body and complement to cause cell lysis, usually with the collaboration of blood complement. An example of such antibodies in pregnancy is the maternal antipaternal lymphocytotoxin seen in secondary spon- TOXICI7~YDURING PREGNANCY taneous aborters (McConnachie and Mc- Intyre, 1984~. Type II immune responses also can be caused by drugs, such as insu- lin, and they are represented by pathophys- iologic conditions in maternal isoimmuni- zation and erythroblastosis fetalis. Anti-D vaccine was developed for Rh-nega- tive mothers with Rh-positive fetuses, to prevent rhesus isoimmunization, a type II reaction (Whitfield, 1976~. Type III immunopathologic reactions are caused by immune complexes, usually with the participation of complement. These reactions include urticarial skin eruptions, arthralgia or arthritis, lymphadenopathy, and fever. The reactions generally last 6-12 daYs and subside when the offending ant~gen ~s el~m~nated (Salmon, 1982~. Environmental pollu- tants, such as mercury, can cause type III reactions (Roman-Franco et al., 1978) that are demonstrated by granular glomerular deposits of immunoglobulin and complement in renal biopsies. Similar immunohisto- logic findings have been reported in pre- eclampsia, but mercury intoxication was not evident in these cases (Matter and Faulk, 1980~. Type IV reactions are independent of antibody and complement and are mediated by lymphocytes. (That is commonly referred to as cell-mediated immunity.) Owing to the role of soluble lymphocyte products (e.g., lymphokines), there might be no such thing as a pure type IV reaction, but it is useful in thinking about diagnosis, treatment, and prognosis. Graft rejection and delayed-hypersensitivity reactions are generally thought to be type IV reac- tions (Turk, 1975), and these types of reactions are inhibited by TA1 and anti- TLX antibodies, as measured by MLC reaction (McIntyre and Faulk, 1979a). Type IV reac- tions are central in host defense reactions against some infectious diseases (Chandra and Newberne, 1977), and the lymphoid axis on which such immunity is based (i.e., T cells) is severely damaged by protein-cal- orie malnutrition. Protein-calorie mal- nutrition is a major factor in diminu- tion of reproductive capacity as demon- strated by decreased ovulation. T-cell function, reproduction, and diet also are closely linked.