JournalofReproductive Immunology, 8 (1985) 33-44 Elsevier
Identification of selectively solubilised syncytiotrophoblast plasma membrane proteins as potential antigenic targets during normal human pregnancy M. Davies and C.M. Browne Reproductive Immunology Research Group, Department of Pathology, The Medical School, University of Bristol, Bristol BS8 1 TD, U.K. (Accepted for publication 21 March 1985)
Syncytiotrophoblast plasma membranes prepared from term placentae were selectively solubilised in non-ionic detergents. The solubilised proteins and the insoluble residue were tested in an ELISA assay for their ability to function as antigenic targets for anti-trophoblast antibodies present in normal first trimester pregnancy sera. The soluble proteins were fractionated by gel filtration and four major antigen forms were identified. The antigens were reactive with affinity purified anti-trophoblast antibody isolated from maternal sera and hence were termed maternally-recognised trophoblast antigens (MRTA); these were designated MRTA-I (M r 400,000 D), MRTA-II (M r = 142,000), MRTA-III (M r 50,000) and MRTA-IV (M r = 13,000). The relationship between MRTA-I, II, III and IV and antigens identified in maternal sera in the form of immune complexes is discussed. =
Key words: human syncytiotrophoblast, membrane proteins, maternal antibodies
By means of a recently developed ELISA (Davies, 1985a), anti-trophoblast antibodies have been detected in sera from women during the course of a normal pregnancy (Davies and Browne, 1985). The IgG anti-trophoblast antibody levels were maximal during the first trimester and declined gradually as pregnancy progressed. On a population basis, the incidence of women with demonstrable anti-trophoblast antibodies decreased with increasing parity. The antibodies were directed against determinants present on plasma membrane vesicles prepared from the microvillous surface of the syncytiotrophoblast layer of term placentae and the antigens involved appeared to be organ-specific rather than alloantigenic since the response could be detected in sera using placentae obtained from unrelated females ( D a v i e s a n d B r o w n e , 1985). A n t i g e n i c a n a l y s i s o f t h e p l a s m a m e m b r a n e o f the s y n c y t i o t r o p h o b l a s t has d e t e r m i n e d t h a t it is d e v o i d o f i n t a c t H L A d e t e r m i n a n t s , b e t a - 2 - m i c r o ° g l o b u l i n , s e r u m p r o t e i n s a n d the A B O a n d R h e s u s b l o o d g r o u p a n t i g e n s ( S z u l m a n , 1973;
0165-0378/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
34 Goto et al., 1976; Davies et al., 1981, 1982). However, in recent years, studies using mouse monoclonal antibodies and heterologous antisera have demonstrated that unique trophoblast-specific antigenic systems are expressed on the syncytiotrophoblast and a number of antigens have been identified (Faulk et al., 1978, 1979; McIntyre and Faulk, 1979; Loke et al., 1980; Sunderland et al., 1981; Johnson et al., 1981; Brown et al., 1983). In the present study, syncytiotrophoblast plasma membranes were selectively solubilised using non-ionic detergents in order to identify those membrane proteins which are recognised as antigens by the maternal anti-trophoblast antibodies obtained from primiparous women during the 5th to 10th wk of gestation. Although only partially purified, these maternally-recognised trophoblast antigens have been compared in terms of their polypeptide profiles with the previously reported xenogeneically-recognised trophoblast antigens.
Materials and Methods
Preparation of placental plasma membrane vesicles Plasma membrane vesicles from the brush border of the syncytiotrophoblast layer of term placentae were prepared according to the method of Smith et al. (1974), and described by Davies et al. (1981). The membranes were suspended in phosphate buffered saline (PBS), pH 7.4, by means of a tight-fitting glass homogeniser.
Solubilisation of placental plasma membrane vesicles Solubilisation was effected either by Nonidet P40 (Alkylphenyl ethoxylate; BDH Ltd.) or Tween 20 (Polyoxyethylene sorbitane monolaurate; Sigma Chemicals Co. Ltd., U.S.A.). Briefly, membrane vesicles were suspended in 0.0038 M glycine-HCl buffer, pH 9.2, containing 1% (v/v) detergent, to a final concentration 3 mg protein m1-1. The suspension, agitated at 4°C for 60 min, was differentially centrifuged at 9000 g for 10 min to yield a high density pellet (HDP) followed by 100,000 g for 60 min to produce a low density pellet (LDP). The material remaining unsedimented was considered to be the solubilised membrane proteins. The complete solubilisation procedure was repeated and the fractions pooled. The fractions were dialysed against PBS and the protein content estimated by the Folin method (Lowry et al., 1951) using the modification described by Bonsall and Hunt (1971).
Enzyme-linked immunosorbent assay (ELISA) The anti-trophoblast antibody activity was estimated by an ELISA based on the use of the plasma membranes or their solubilised components as the antigenic target and utilising a urease-conjugated sheep anti-human IgG antiserum (Sera-Lab, Crawley Down, England). The assay has been fully described elsewhere (Davies, 1985a). Besides estimating the antibody titre, a 'reverse' assay was used to assess the antigenic potential of the various membrane fractions. Microtitre plates (Flow Laboratories, Irvine, Scotland) were coated with the membrane fractions diluted to 50 /tg protein, m1-1 and incubated with a 1/10 diluted suitable maternal serum
35 sample. The antigenic potential of the fractions was based on spectrophotometric analysis at 588 nm following treatment of the plates with the developing antiserum and the enzyme substrate.
P olyacrylamide gel electrophoresis (PAGE) The polypeptide profiles of both intact, solubilised and fractionated placental plasma membranes were determined by means of SDS polyacrylamide gradient gel electrophoresis. The method was based on the procedure described by Laemmli (1971) and is fully presented elsewhere (Davies et al., 1981).
Solubilised membrane protein fractionation The detergent solubilised plasma membrane proteins were fractionated on the basis of molecular size by means of column chromatography using Sephacryl S-300 a n d / o r Sepharose 4B (Pharmacia Fine Chemicals) and eluting with PBS, pH 7.4, at a flow rate of 15 ml. h -~. The protein content of the fractions was estimated by a comparison of the ratio of E280 : E260 nm and diluted in ELISA coating buffer to a concentration of 50 /~g-ml-i for use as potential antigenic targets in the ELISA. Since the fractionation procedure was used to identify, rather than purify, the antigens, a limited number of fractions were obtained from each column to ensure ease of handling in the ELISA. The molecular weights were estimated by comparison with the elution profiles obtained with standard proteins. Sera
Maternal sera were prepared from routine blood samples obtained from women in their 5th to 10th wk of a first pregnancy. Control sera were obtained from males or nulliparous non-pregnant females.
Affinity chromatography purification of the anti-trophoblast antibody Intact placental plasma membrane vesicles were bound to Sephadex G.150 using a concanavalin A (Con A) link by a modification of the method described by Sela and Edelman (1977). Five ml Sephadex G.150 (Pharmacia Fine Chemicals) were stirred at room temperature for 30 min with 5 mg Con A (Sigma Chemical Co.). The excess Con A was decanted and the Con A-Sephadex conjugate incubated for 15 min at room temperature with intact placental plasma membrane vesicles. The conjugate was poured into a column and stabilised by washing for 5 h with 3% glutaraldehyde in PBS and then allowed to stand in the glutaraldehyde overnight. The column was washed extensively with PBS prior to use. Two ml of Con A-Sephadex G.150 absorbed maternal serum were added to the column which was incubated at room temperature for 30 min and the unbound material eluted with PBS. The bound antibodies were eluted with 1 M glycine, pH 2.5, and the eluate neutralised prior to use.
Solubilisation of membrane proteins I n t a c t p l a c e n t a l p l a s m a m e m b r a n e vesicles were solubilised in either N o n i d e t P40 or Tween 20 a n d differentially centrifuged to yield a high d e n s i t y pellet ( H D P , s e d i m e n t e d at 9000 g), a low density pellet ( L D P , centrifuged at 100,000 g ) a n d the solubilised m e m b r a n e p r o t e i n s ( n o n - s e d i m e n t a b l e at 100,000 g). A s a p r o p o r t i o n of the initial p r o t e i n concentration, 50-70% was c o n t a i n e d in the H D P , 5 - 2 0 % in the L D P and 2 5 - 3 0 % in the solubilised fractions, regardless of the detergent used ( T a b l e 1). O n a p e r mg p r o t e i n basis, the m a j o r antigenic activity a p p e a r e d to reside in the L D P with c o n s i d e r a b l y less being d e t e c t a b l e in the H D P a n d solubilised proteins. However, strict c o m p a r i s o n s of the relative antigenic p o t e n t i a l s of the different fractions are not valid owing to the different m o l e c u l a r structures b e i n g c o m p a r e d , i.e., solubilised a n d m e m b r a n e - b o u n d proteins. The solubilised fraction was further e x a m i n e d since it was in a form a m e n a b l e to gel filtration studies.
Gel filtration of Nonidet P40 solubilised proteins The N o n i d e t P40 solubilised m e m b r a n e p r o t e i n s were f r a c t i o n a t e d on a Sephacryl S-300 c o l u m n a n d the p r o t e i n c o n t e n t a n d antigenic p o t e n t i a l of each fraction assessed. T h e m a j o r i t y of the p r o t e i n s eluted from the c o l u m n over the m o l e c u l a r weight range 4 8 0 - 1 0 kD, with two m a j o r p e a k s at 230 k D a n d 50 k D (Fig. 1). A n t i g e n i c analysis, b y m e a n s of reactivity against a specified m a t e r n a l serum ( M S 5119), i n d i c a t e d that there was some reactivity t h r o u g h o u t the c o m p l e t e range of eluted proteins, b u t three distinct p e a k s of activity were evident. T h e r e was a m a j o r peak, Mr400 a n d two m i n o r peaks, M r 142 a n d 50; the M r of the m a j o r p e a k was c o n f i r m e d b y r e c h r o m a t o g r a p h i n g the p r o t e i n fraction on Sepharose 4B. These three
TABLE 1 The distribution of antigenic material following detergent solubilisation a Membrane fraction
Intact vesicles High density pellet (HDP) Low density pellet (LDP) Non-sedimented material
% Total protein b
IgG titre c Maternal sera
100 63 11 28
8.0 5.5 8.5 6.0
6.0 3.0 6.0 4.0
% Total protein b
IgG titre ~ Maternal Normal sera sera
100 56 17 27
8.0 6.0 8.0 6.0
6.0 3.0 5.0 4.0
a Placental plasma membranes were solubilised in either Nonidet P40 or Tween 20 according to the Materials and Methods. b The protein content of each fraction was estimated and expressed as a percentage of the total protein present in the intact vesicles. c The fractions were diluted to 50/~g. ml-l and used as a target for maternal sera (MS5119, MS5089 and MS5098) obtained from women in their first pregnancy at the 9th wk of gestation or from nulliparous non-pregnant females. The figures represent the mean values obtained from 5 observations.
37 08 1-5
. ~ I.. 1"o 1"5"~~25 FRACTION NUMBER Fig. 1. The fractionation on Sephacryl S-300 of Nonidet P40-solubilised placental plasma membrane vesicles. Each fraction was estimated for protein content (mg.ml-l; • *) and the ability to function as an antigen in the ELISA (Esssnm; O---O) using maternal serum as the source of the anti-trophoblast antibodies.
antigen peaks on the basis of their reactivity with maternal serum were termed maternally-recognised trophoblast antigens and designated MRTA-I, II and III, respectively; they were recognised by all the maternal sera tested. A similar experiment to assess the antigenic potential of the fractions using normal control sera indicated that the reactivity of the fractions was extremely variable, depending on the serum sample. Some normal sera contained reactivity against all three antigens while others had activity against only MRTA-I and II. When the relative strengths of these reactions was measured, the maternal sera had greater antibody activity against all three peaks compared with the normal sera; the mean IgG antibody titre using MRTA-I, II and III as targets in the ELISA was 6.6, 7.7 and 5.3 respectively, when maternal sera were used as the antibody source, and 4.6, 6.3 and 3.0 respectively, using normal sera (Table 2). Analysis of the data demonstrated that the anti-MRTA I, II and III antibody levels in normal sera represented only 20, 26 and 17%, respectively, of the levels observed in the pregnancy sera. However, the differential between maternal and normal sera was not as great as had been previously observed with intact membranes or the LDP (Davies and Browne, 1985; Table 1). In addition, absorption of MRTA-I, II and III with immobilised protein A to eliminate any immunoglobulins demonstrated a decrease in the anti-MRTA II activity. Hence the majority of the activity assigned to MRTA-II was probably due to the non-specific binding of the developing antiserum to the Fc regions of immunoglobulins contained in that fraction. MRTA-II had an apparent M r by gel filtration of 142 (Fig. 1). However, once the immunoglobulins were removed a residual antigenic activity could still be detected in the presence of maternal sera but not with normal sera. This residual
38 TABLE 2 The use of the maternally-recognisedtrophoblast antigens as targets in the ELISA~ Tween 20-solubilised
Nonidet P40-solubilised MRTA-I
Pregnant, 10th wk of gestation Nulliparous, non-pregnant female
a The titre of the IgG anti-trophoblast antibody activity (log2) was estimated in pregnancy and normal sera using the antigen fractions (MRTA) obtained by fractionating the detergent-solubilisedproteins on Sephacryl S-300. The figures represent the means + SEM for 3-4 observations.
activity was not attributed to lgG3 subclass o n the basis of the polypeptide profile (Fig. 5).
Gel filtration of Tween 20 solubilised proteins The T w e e n 20 solubilised proteins were fractionated on Sephacryl S-300 a n d produced a similar elution profile to the N o n i d e t P40 solubilised proteins except that the peak of protein material had an a p p a r e n t M r of 142. W h e n the fractions were screened for antigenic activity using m a t e r n a l serum (MS5119), two regions were identified (Fig. 2). O n e eluted with a n a p p a r e n t M r of 400 a n d c o r r e s p o n d e d to M R T A - I , while the second eluted in the 13 region a n d was termed M R T A - I V . Both M R T A - I a n d M R T A - I V reacted with m a t e r n a l a n d n o r m a l sera; the m a t e r n a l sera had m e a n I g G a n t i b o d y titres of 7.0 a n d 6.0 for M R T A - I a n d IV, respectively,
Fig. 2. The fractionation on Sephacryl S-300 of Tween 20-solubilizedplacental plasma membrane vesicles. Each fraction was estimated for protein content (mg. ml- 1; • O) and the ability to function as an antigen in the ELISA (E588nm; O---O) using maternal serum as the source of the anti-trophoblast antibodies.
0, A ;
o" : "
Fig. 3. The affinity purification of anti-trophoblast antibodies. The fractions eluted with glycine buffer were estimated for pH( X x ), protein (#g per fraction; • •) and antibody activity (E588,m; D---D) using intact placental membrane vesicles(code number 2802) as the antigen source. Fig. 4. The ability of affinity purified antibody, eluted at pH 2.5, to bind to Sephacryl S-300 fractionated protein solubilised by Nonidet P40 (• •) and Tween 20 (D---D).The binding of the antibody to the protein fractions is measured by absorption at 588 nm. compared with the corresponding titres obtained with normal sea of 4.5 and 3.6. Both forms of antigenic activity were not absorbed by immobilised protein A.
Antigenic analysis using affinity purified antibody By means of immobilised intact membranes, anti-trophoblast antibody activity could be bound and subsequently eluted from columns. As the p H of the fractions fell the protein bound to the column eluted in two peaks, one at approximately p H 6.8 and the other p H 2.5. Using intact membranes as the target in the ELISA, both peaks of protein were found to contain anti-trophoblast antibody activity (Fig. 3). When the targets in the ELISA were M R T A I-IV, it was observed that the first peak of antibody activity (pH 6.8) reacted with all four M R T A (data not shown). However, the second antibody peak (pH 2.5) reacted with MRTA-I, II and IV but not with M R T A - I I I (Fig. 4). The purified antibody was used to confirm the nomenclature of the antigens as maternally-recognised.
Polypeptide analysis of the maternally-recognisedtrophoblast antigens The polypeptide components of the various fractions of the solubilised membranes were analysed by SDS P A G E following treatment with SDS and reduction with mercaptoethanol. The data, indicating the apparent molecular weights of the polypeptide bands, are presented in Fig. 5. Each of the four M R T A have distinct
Fig. 5. The polypeptideprofilesof the various membrane fractionson SDS PAGE followingprior sample reduction. Track 1 represents intact plasma membranepreparation; Track 2, HDP; Track 3, LPS; Track 4, Nonidet P40 detergent-solubilisedproteins; Track 5, MRTA-1;Track 6, MRTA-II; Track 7, MRTA-III; and Track g, MRTA-IV. The arrows represent the molecularweight markers. polypeptide profiles although several bands are common to some of the MRTA. MRTA-I has major bands of M r 105 and 60, while MRTA-II has major components of 68, 29 and 24 kDal. MRTA-III has approximately 10 polypeptide bands with the major components having apparent M r of 117, 68, 60, 50, 29, 24, 20 and 13. Finally, MRTA-IV consisted of a single polypeptide band of apparent M r of 13.
Discussion Plasma membrane vesicles isolated from the syncytiotrophoblast layer of term placentae were selectively solubilised by non-ionic detergents to yield soluble (27% of the membrane proteins) and insoluble (74% of the membrane proteins) fractions that were able to function as antigenic targets in an ELISA for anti-trophoblast antibodies reported to be present in first trimester pregnancy sera (Davies and Browne, 1985). By differential centrifugation two detergent-insoluble fractions, H D P and LDP, were obtained; while the latter represented 10-20% of the insoluble proteinaceous material, it did, however, on a per/~g protein basis, bind approximately eight times more antibody than the former (Table 1). This latter calculation is, however, unable to account for any difference in the coating efficiency in the ELISA of the two potential targets. The detergent soluble proteins were relatively poor targets and were comparable with the HDP. This pattern was achieved regardless of the non-ionic detergent involved. It has been demonstrated previously
that placental plasma membrane vesicles prepared by the method of Smith et al. (1974) can be separated into two subpopulations using sucrose gradient centrifugation (Davies et al., 1981). These two subpopulations can be distinguished in their binding pattern for IgG, IgM and transferrin, but not for insulin. The high density fraction (density > 1.08 g. cm -3) represented 84% of the membrane protein, while the lighter fraction (density 1.048-1.05 g. cm-3) represented the remaining 16%. The HDP and LDP were not suitable in their present forms for the protein separation techniques and undoubtedly retained a high level of membrane organisation since it has been reported that the 'leakiness' of the vesicles following treatment with Nonidet P40 was similar to that of untreated vesicles (Davies and Sutcliffe, 1982). From this it can be concluded that the Nonidet 40 (and probably the Tween 20) - soluble proteins are not necessary to maintain placental membrane integrity. Although Nonidet P40 and Tween 20 solubilised the membranes to a comparable degree they appeared to differ in their selective solubilisation of the various proteins. Following fractionation of the solubilised proteins on Sephacryl S-300, four regions of antibody-binding activity (antigen activity) could be identified with apparent M r of 400, 142, 50 and 13 (Fig. 1 and 2). Since these regions were reactive with the affinity-purified anti-trophoblast antibody (Fig. 4) they have been termed maternally-recognised trophoblast antigens (MRTA) and designated MRTA-I, II, III, and IV. However, the degree to which they were solubilised was dependent on the detergent used. MRTA-I was the major antigenic component in Nonidet P40-solubilised proteins with MRTA-II and III representing minor, but significant, components. With Tween 20, MRTA-I and IV were the major antigenic components with very little contribution from MRTA-II and III. By absorption of the four antigenic forms with immobilised protein A it was found that the majority of the antigen activity in MRTA-II, but not in the other fractions, could be explained on the basis of non-specific binding of the developing antiserum to the Fc regions of immunoglobulins present in the MRTA-II fraction; this fraction co-eluted from the Sephacryl S-300 with a commercial preparation of human IgG. There was, however, a residual activity remaining in the MRTA-II after the absorption. It is generally considered that for membrane solubilisation non-ionic detergents, such as Nonidet P40 and Tween 20, fulfill all the necessary criteria, namely that the solubilisation should apparently destroy the integrity of the membrane and release its constituent proteins into solution so that they retain their antigenicity. Ionic detergents such as SDS and sodium deoxycholate, although more efficient at solubilisation, are more denaturing (Hjrrten and Johansson, 1972). The disadvantage of non-ionic detergents is the selective solubilisation achieved, with the degree of solubilisation, under standard physical conditions, being dependent on the membrane. For example, it has been reported that Tween 20 is able to solubilise 30% of the proteins from erythrocyte membranes (Liljas et al., 1974) or 70% from Acholeplasma laidlawii (Hjrrten and Johansson, 1972); this compares with the 27% for Tween 20 and 28% for Nonidet P40 for the solubilisation of the placental membranes. It is thought that the insoluble proteins contain more hydrophobic side chains than solubilised proteins and hence are more resistant to a weak dissociating agent such as Tween 20.
Acknowledgements This work was supported by grants from the Medical Research Council and the Wellcome Trust. We wish to thank Dr. W.D. Billington for his assistance in the preparation of the manuscript.
References Bonsall, R.W. and Hunt, S. (1971) Characteristics of interactions between surfactants and the human erythrocyte membrane. Biochim. Biophys. Acta 193, 265-271. Brown, P.J., Molloy, C.M. and Johnson, P.M. (1983) Immunochemical identification of human trophoblast membrane antigens using monoclonal antibodies. J. Reprod. lmmunol. 5, 351-361. Brunda, M.J. and Minden, P. (1977) Antibodies to bacterial and tumor-derived antigens in sera from normal guinea pigs. J. Immunol. 119, 1374-1377. Davies, M. (1985a) An EL1SA for the detection of maternal anti-trophoblast antibodies in human pregnancy. J. Immunol. Methods 77, 109-118. Davies, M. (1985b) Antigenic analysis of immune complexes formed in normal human pregnancy. Clin. Exp. Immunol. (in press). Davies, M. and Browne, C.M. (1985) Anti-trophoblast antibody responses in normal human pregnancy. J. Reprod Immunol. 7, 285-297. Davies, M., McLaughlin, M.E.E. and Sutcliffe, R.G. (1982) Immune responsiveness against the human placenta. I. Generation of cellular and humoral activity in experimental animals. Immunology 47, 459-468. Davies, M., Parry, J.E. and Sutcliffe, R.G. (1981) Examination of different preparations of human placental plasma membrane for the binding of insulin, transferrin and immunoglobulins. J. Reprod. Fertil, 63, 315-324. Davies, M. and Sutcliffe, R.G. (1982) A quantitative assay for the measurement of immune responses directed against the human placenta. J. Immunol. Methods 48, 373-383. Faulk, W.P. and Mclntyre, J.A. (1983) Immunological studies of human trophoblast: markers, subsets and functions, Immunol. Rev. 75, 139-175. Faulk, W.P., Temple, A., Lovins, R.E. and Smith, N. (1978) Antigens of human trophoblasts: a working hypothesis for their role in normal and abnormal pregnancies. Proc. Natl. Acad. Sci. U.S.A. 75, 1947-1951. Faulk, W.P., Yeager, C., Mclntyre, J.A. and Ueda, M. (1979) Oncofetal antigens of human trophoblast. Proc. R. Soc. London B 206, 163-182. Goto, S., Hoshino, M., Tomoda, U. and lshizuka, N. (1976) Immunoelectron microscopy of the human chorionic villus in search of blood group A and B antigens. Lab. Invest. 35, 530-536. Hj6rten, S. and Johansson, K.E. (1972) Selective solubilisations with Tween 20 of membrane proteins from Acholeplasma laidlawii. Biochim. Biophys. Acta 288, 312-325. Johnson, P.M., Cheng, H.M., Molloy, C.M., Stem, C.M.M. and Slade, M.B. (1981) Human trophoblastspecific surface antigens identified using monoclonal antibodies. Am. J. Reprod. Immunol. 1, 246-254. Laemmli, U.K. (1971) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227, 680-685. Liljas, L., Lundahl, P. and Hj6rten, S. (1974) Selective solubilisation with Tween 20 of proteins from water-extracted human erythrocyte membranes. Analysis by gel electrophoresis in dodecylsulphate and in Tween 20. Biochim. Biophys. Acta 352, 327-337. Loke, Y.W., Whyte, A. and Davies, S.P. (1980) Differential expression of trophoblast specific membrane antigens by normal and abnormal human placentae and neoplasms of trophoblastic and non-trophoblastic origin. Int. J. Cancer 25, 459-461. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem. 193, 265-275.
44 Mclntyre, J.A. and Faulk, W.P. (1979) Antigens of human trophoblast. Effects of heterologous antitrophoblast sera on lymphocyte responses in vivo. J. Exp. Med. 149, 824-836. Minden, P., Sharpton, T.R. and McClatchy, J.K. (1976a) Shared antigens between human malignant melanoma cells and Mycobacterium bovis (BCG). J. Immunol. 116, 1407-1414. Minden, P., Jarrett, C. McClatchy, J.K., Gutterman, J.U. and Hersh, E.M. (1976b) Antibodies to melanoma cells and BCG antigens in sera from tumour-free individuals and from melanoma patients. Nature (London) 263, 774-776. O'Sullivan, M.J., McIntyre, J.A., Prior, M., Warriner, G. and Faulk, W.P. (1982) Identification of human trophoblast membrane antigen in maternal blood during pregnancy. Clin. Exp. Immunol. 48, 279-287. Smith, N.C., Brush, M.E. and Luckett, S. (1974) Preparation of human placental villous surface membrane. Nature (London) 252, 302-303. Sela, B.-M. and Edelman, G.M. (1977) Isolation by cell-column chromatography of immunoglobulin specific for cell surface carbohydrates. J. Exp. Med. 145, 339-443. Sunderland, C.A., Redman, C.W.G. and Stirrat, G.M. (1981) Monoclonal antibodies to human syncytiotrophoblast. Immunology 43, 541-546. Szulman, A.E. (1973) The A, B and H blood group antigens in human placenta. N. Engl. J. Med. 286, 1029-1031. Theofilopoulos, A.N. and Dixon, F.J. (1979) The biology and detection of immune complexes. Adv. Immunol. 28, 89-220.