Possible mechanisms of immunotherapy for maintaining pregnancy in recurrent spontaneous aborters: analysis of anti-idiotypic antibodies directed against autologous T-cell receptors

Koichi Ito1,4, Tadao Tanaka2,5, Norio Tsutsumi2, Fumiya Obata3 and Noboru Kashiwagi1

1 Department of Immunology, Kitasato University School of Medicine, 1–15–1 Kitasato, Sagamihara 228-8555, 2 Department of Obstetrics and Gynecology, National Okura Hospital, 2–10–1 Okura, Setagaya 157-0074 and 3 Department of Immunology, Kitasato University School of Allied Health Science, 1–15–1 Kitasato, Sagamihara 228-8555, Japan


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We examined whether immunotherapy for recurrent spontaneous abortion (RSA) using paternal lymphocytes induces anti-T-cell receptor (TCR) idiotypic antibodies in RSA patients. The sera of these patients were assessed for inhibitory activity against mixed lymphocyte reactions (MLR) between maternal responder cells and paternal stimulator cells. Sera of four of the five women who maintained pregnancy successfully after immunotherapy showed significant MLR inhibition, whereas none of the five women who had unsuccessful pregnancies showed significant MLR inhibition. These sera inhibited the MLR of autologous responder T-cells, when stimulated with lymphocytes having the same HLA-DR antigens as the patient's husband, but not when stimulated with lymphocytes having unrelated HLA-DR antigens. This MLR inhibitory activity was absorbed by autologous maternal T-lymphoblasts induced by stimulation with lymphocytes having the paternal HLA-DR type but not by those induced by stimulation with lymphocytes having other HLA-DR types. The maternal serum inhibited the proliferation of autologous T-cells, but not of non-autologous T-cells, stimulated with paternal lymphocytes. These results indicate that anti-TCR idiotypic antibodies were induced in RSA patients by immunotherapy. These antibodies may contribute to maintaining pregnancy by negatively regulating maternal T-cells directed against HLA-DR antigens of the fetus.

Key words: anti-idiotypic antibody/HLA/immunotherapy/recurrent spontaneous abortion/TCR


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Immunological regulation by which pregnancy is maintained is a central problem in the field of reproductive immunology. In particular, human transplantation antigens, HLA, may play important roles in the maintenance of the fetus, because the fetus could be regarded as an allograft which escapes maternal immunological rejection. In this sense, recurrent spontaneous abortion (RSA) is considered to be a model of immunological rejection of the fetus by the mother. To elucidate the aetiology of RSA, it is important to clarify the immunological regulatory mechanisms by which the fetus is protected from maternal attack in normal pregnancy.

The immunological regulation which maintains pregnancy is reported to involve both cellular and humoral regulation. It has been reported that suppressor cells found in the decidua suppress the maternal immune response to the fetus (Daya et al., 1985Go; Clark et al., 1986Go). HLA-G antigens expressed on trophoblast cells are reported to inhibit natural killer cell activity (Ellis et al., 1990Go; Kovats et al., 1990Go; Mandelboim et al., 1997Go; Rouas-Freiss et al., 1997Go). Anti-paternal cytotoxic antibodies have been detected with high frequency in the sera of normal pregnant women (Jensen, 1962Go; Payne, 1962Go; Ahrons, 1971Go; Doughty and Gelsthorpe, 1971Go; Regan et al., 1991Go). Furthermore, anti-T-cell receptor (TCR) idiotypic antibodies capable of specifically inhibiting maternal autologous immune responses have been demonstrated in normal pregnant women (Sucui-Foca et al., 1983; Singal et al., 1984Go).

As immunotherapy for RSA, women are immunized with their partner's lymphocytes (paternal lymphocytes) in an attempt to enhance the production of immune-suppressing antibodies. Although high success rates of maintaining pregnancy have been achieved using this therapy (Taylor and Faulk, 1981Go; Mowbray et al., 1985Go; Reznikoff-Etievant et al., 1988Go; Daya and Gunby, 1994Go; Agrawal et al., 1995Go; Carp et al. 1997Go), the mechanisms of this therapeutic effect have yet to be elucidated. Some investigators have suggested that the RSA patient's ability to produce anti-paternal cytotoxic (Mowbray et al., 1985Go; Reznikoff-Etievant et al., 1988Go; Lubinski et al., 1993Go; Agrawal et al., 1995Go) and anti-autologous activated T-cell antibodies (Sugi et al., 1991Go; Lubinski et al., 1993Go) correlates with successful immunotherapy. However, none of these studies have determined the specificity or identified the possible target molecules of such antibodies. In this study, we examined whether immunotherapy for RSA induces anti-TCR idiotypic antibodies against autologous TCR in the sera of RSA patients. The study group comprised RSA patients who successfully maintained pregnancy after receiving this treatment as well as patients who had an unsuccessful pregnancy after immunotherapy. The results indicate that such antibodies were produced in successful, but not in unsuccessful, immunotherapy cases.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
RSA patients
We selected nine RSA patients who had been immunized with their husband's lymphocytes at the Department of Obstetrics and Gynecology of the National Okura Hospital, Tokyo. Patients with any anatomical, hormonal, autoimmunological or infectious causes of abortion were excluded from the analysis. In addition, if either the husband or wife had an abnormal karyotype, the couple was excluded. The pretherapeutic abortion histories and reproductive outcomes after immunotherapy of these nine women are summarized in Table IGo. One patient (7) was treated twice by immunotherapy and is included in more than one group. Of the seven patients in the primary RSA group, four had experienced spontaneous abortion three times (patients 7, 9, 10 and 21) and three had experienced spontaneous abortion twice (patients 1, 8 and 7). Of the three patients in the secondary RSA group, one had experienced abortion seven times, one had experienced abortion three times and the other patient had two spontaneous abortions (patients 38, 26 and 50 respectively). Overall, pregnancy was maintained until delivery in five of the RSA patients, but not in the other five, after immunotherapy. The pregnancy of patient 7 resulted in abortion after the first course of treatment, but the pregnancy after a second course was maintained successfully.


View this table:
[in this window]
[in a new window]
 
Table I. Pretherapeutic medical histories and pregnancy outcomes after lymphocyte immunization of the patients in this study
 
Immunotherapy
The paternal lymphocytes were isolated from peripheral blood by the Hypaque density-gradient method (Pharmacia Biotech, Uppsala, Sweden) and were stored in liquid nitrogen until required for the treatment and experiments. RSA patients received 5x107 paternal lymphocytes intracutaneously in their upper arms, three times at 2-week intervals. After pregnancy had been confirmed, identical booster immunizations were given at 2-week intervals until the end of the first trimester.

Serum samples
Serum samples were collected from each RSA patient before and after immunotherapy, at various times during pregnancy, and at delivery. The sera were subjected to complement inactivation at 56°C for 30 min and were stored at –20°C until required for the experiments.

HLA typing
Twenty-seven different HLA-DRB1 genotypes were determined by DNA typing using the polymerase chain reaction and hybridization with sequence-specific oligonucleotides established in our previous study (Obata et al., 1991Go).

Mixed lymphocyte reaction (MLR) inhibition test
The presence of anti-TCR idiotypic antibodies was examined by determining the inhibitory activity of the maternal serum against the MLR of the paternal lymphocytes (stimulator cells) with the pre-immunotherapy autologous maternal lymphocytes (responder cells). In the MLR test, the responder cells (2.5x106) were treated with test serum (0.5 ml) at room temperature for 30 min, followed by incubation with rabbit complement (2.5 ml) at 37°C for 1 h. Then the responder cells were washed with culture medium three times and the proliferation assay was carried out in triplicate using the serum-treated responder cells and 5x104 irradiated (3000 rad) stimulator cells in 200 µl RPMI-1640 supplemented with 20% (v/v) pooled human male serum (complete medium), in 96-well round-bottomed microtitre plates at 37°C in a 5% CO2 incubator. On day 5, the cells were incubated with 1 µCi [3H]-thymidine (6.7 µCi/mM) for 18 h, were harvested and 3H incorporation was determined by liquid scintillation counting. As the control of the MLR experiments, the responder cells were treated with pooled human male serum and rabbit complement. Lymphocytes from other donors having HLA types either common or unrelated to those of the RSA patients and their husbands were also used as responder or stimulator cells to examine the specificity of the anti-TCR idiotypic antibodies.

Absorption test
Serum sample 7S-9, which was obtained at 9 weeks of gestation (Figure 2BGo), was used for this study. Serum 7S-9 (1 ml) was absorbed with patient 7 autologous T-lymphoblasts (1x107) that had been induced by MLR using various donor lymphocytes as stimulator cells (husband, HLA-DRB1*0901/1502; KT14, HLA-DRB1*0901/0901; Pitout, HLA-DRB1*0701/0701) in complete medium at 37°C in a 5% CO2 incubator for 9 days. After absorption, fresh pre-immunotherapy patient 7 maternal lymphocytes were treated with either absorbed or non-absorbed 7S-9 serum and rabbit complement, and used as responder cells in a second set of MLR, in which irradiated paternal lymphocytes were used as the stimulator cells.



View larger version (28K):
[in this window]
[in a new window]
 
Figure 2. Inhibition of the maternal mixed lymphocyte reactions (MLR) by autologous serum (patient 7). A and B indicate the results of MLR inhibition after the first and second courses of treatment respectively. The asterisks indicate significant inhibition (**P < 0.01). aThe antibody designated as 7S-9 was used for studies of specificity (Figure 3Go) and absorption (Figure 4Go). W = weeks.

 
Complement-dependent cytotoxicity (CDC) test against Tlymphoblasts
Irradiated paternal stimulator (1x107) and maternal responder (1x107) cells were cultured in 10 ml of complete medium in a 25-cm2 tissue culture flask at 37°C in a 5% CO2 incubator. After 9 days, the T-lymphoblasts (2x104) obtained were treated with autologous maternal test serum (0.2 ml) and rabbit complement (1 ml), and stained with Trypan Blue. The dead cells were counted under a microscope.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
MLR inhibition test
All results of MLR inhibition by the sera of these patients are summarized in Table IIGo. In this table, when at least one among all sera collected from each RSA patient showed MLR inhibition, we considered the sample to be anti-TCR idiotypic antibody-positive. The sera obtained from patient 1 during the 19th, 25th and 31st weeks of gestation inhibited the MLR significantly (P < 0.01 and P < 0.05) by 29.6, 38.2 and 24.6%, respectively (Figure 1Go).


View this table:
[in this window]
[in a new window]
 
Table II. Summary of anti-T-cell receptor (TCR) idiotypic antibodies detected in this study
 


View larger version (17K):
[in this window]
[in a new window]
 
Figure 1. Inhibition of the maternal mixed lymphocyte reactions (MLR) by autologous serum (patient 1). Anti-T-cell receptor idiotypic antibody was detected by inhibition of the MLR between the paternal lymphocytes and maternal lymphocytes treated with autologous serum obtained at different stages of pregnancy and complement. The asterisks indicate significant inhibition compared with that of the control experiment, in which maternal lymphocytes were treated with pooled human male serum (*P < 0.05; **P < 0.01). W = weeks.

 
After the first treatment with paternal lymphocytes, the pregnancy of patient 7 ended in abortion, and none of her serum samples after the first treatment showed MLR inhibition compared with that of the control (Figure 2AGo). Ten months after her abortion, patient 7 received a second course of immunotherapy which was maintained until delivery. All sera collected after the second treatment showed significant MLR inhibition (range of inhibition, 53.4–71.4%) (P < 0.01), except for the sample collected at 5 weeks of gestation (Figure 2BGo).

For the two other successful patients, 8 and 26, a serum sample collected at 20 weeks of gestation from patient 8 showed significant MLR inhibition of 29.6% (P < 0.01), but none of sera collected from patient 26 showed MLR inhibition (Table IIGo). The five women who aborted despite receiving immunotherapy (21, 10, 7, 50 and 38) showed no MLR inhibition.

Specificity of MLR inhibition
The specificity of MLR inhibition was investigated using serum 7S-9, which was obtained at 9 weeks of gestation (Figure 2BGo) from patient 7, which was considered to be a representative sample. Patient 7 maternal lymphocytes were treated with autologous serum 7S-9 and complement and used as responder cells in MLR, in which irradiated lymphocytes from her husband and various unrelated donors were used as stimulator cells. The MLR was inhibited significantly by 41.9, 76.3 and 44.6% when patient 7 maternal lymphocytes were stimulated with lymphocytes from her husband (DRB1*0901/1502), donor O.M. (DRB1*1502/1502), and donor KT14 (DRB1*0901/0901) respectively. Each of these donors had one HLA-class II antigen in common with the patient's husband (Figure 3Go). By contrast, no MLR inhibition was observed when patient 7 maternal lymphocytes were stimulated with lymphocytes from donor KT11 (DRB1*1302/1302) who shared no HLA-class II antigens with patient 7's husband. These results suggest that 7S-9 serum specifically inhibited autologous T-cells that recognized the husband's HLA-DR molecules (i.e. DRB1*0901 and 1502), but not those that recognized the other DR molecules, such as DRB1*1302. Furthermore, we investigated whether 7S-9 serum could inhibit MLR by non-autologous T-cells. T-cells from donor 8F (DRB1*0802/1201) were treated with 7S-9 serum and complement and were used as responder cells for MLR. As shown in Figure 3, Go7S-9 serum did not show any inhibition against 8F T-cells even when they were stimulated with cells from patient 7's husband. Thus, 7S-9 serum appears to contain an antibody that specifically reacts with an antigenic structure that is characteristic of the autologous T-cells, but not of non-autologous T-cells, recognizing the particular allogenic HLA-DR molecules. It is reasonable to suggest that this antigenic structure on the T-cells is a TCR idiotype and, therefore, that 7S-9 serum contains antibodies against the TCR idiotypes.



View larger version (15K):
[in this window]
[in a new window]
 
Figure 3. Specificity of mixed lymphocyte reaction (MLR) inhibition. The lymphocytes of patient 7 were treated with autologous serum 7S-9 (open bar) and complement, and were stimulated with lymphocytes from the patient's husband or from various unrelated donors. In an additional experiment, unrelated responder cells (8F) were treated with 7S-9 serum and complement. HLA-DRB1 types of stimulator and responder cells are also shown. The asterisks indicate significant inhibition compared with that of the control experiment, in which responder cells were treated with pooled human male serum (closed bar) (*P < 0.05; **P < 0.01).

 
Absorption test
The MLR inhibitory activity of 7S-9 was abolished by absorption with T-lymphoblasts induced by stimulation with lymphocytes from the patient's husband (DRB1*0901/1502) and from donor KT14 (DRB1*0901/0901), who had the same HLA-class II antigens as her husband, but not by those induced by stimulation with lymphocytes from Pitout (DRB1*0701/0701), which had HLA-class II antigens that differed from her husband's (Figure 4Go). These results indicate that MLR inhibition by 7S-9 serum was due to antibodies that specifically recognized idiotypes on TCR directed against allogenic HLA-DRB1*0901, but not on TCR directed against allogenic HLA-DRB1*0701.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 4. Absorption of mixed lymphocyte reaction (MLR) inhibitory activity. Serum 7S-9 from patient 7 was absorbed with autologous T-lymphoblasts generated by stimulation with various donor lymphocytes. Pre-immunotherapy maternal lymphocytes from patient 7 were treated with either absorbed or non-absorbed 7S-9 and complement, and used as responder cells for MLR with paternal lymphocytes as the stimulator cells. The HLA-DRB1 genotypes of the stimulator cells used to generate the maternal T-lymphoblasts used for absorption are shown in parentheses. The asterisks indicate a significant difference compared with that of the non-absorbed serum (*P < 0.05).

 
CDC test
Patient 9, who had experienced a successful outcome after immunotherapy, was examined for anti-TCR idiotypic antibodies using the CDC test. The number of dead cells increased significantly, by 26 and 23%, in sera taken at the 16th and 22nd weeks of gestation respectively, compared with that of the control (Figure 5Go). The remaining vital T-lymphoblasts exhibited significantly less proliferative reaction in the secondary MLR when they were restimulated with paternal lymphocytes, compared with those treated with control sera (24.9% and 26% inhibition by sera obtained at 16 and 22 weeks of gestation respectively) (P < 0.01). In summary, anti-TCR idiotypic antibodies were detected in four (80%) of five patients whose pregnancies after immunization were successful and in none of the five patients (0%) whose pregnancies after immunization aborted (Table IIGo).



View larger version (14K):
[in this window]
[in a new window]
 
Figure 5. CDC activity of maternal serum against autologous activated T-lymphocytes (patient 9). The maternal T-lymphoblasts generated by stimulation with the paternal lymphocytes were treated with autologous serum obtained at different stages of pregnancy and complement, and stained with Trypan Blue to determine the percentage of dead cells. The asterisks indicate a significant difference compared with that of the control experiment, in which maternal T-lymphoblasts were treated with pooled human male serum (**P < 0.01).

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In this study, we examined whether anti-idiotypic antibodies directed against autologous TCR were induced in RSA patients who received immunotherapy with paternal lymphocytes. The following three findings proved the existence of such antibodies in sera of these patients: (i) maternal serum inhibited the MLR proliferation of autologous responder T-cells that were stimulated by cells having the same HLA-DR antigens as the husband, but did not inhibit MLR proliferation when the stimulator cells had unrelated HLA-DR antigens; (ii) the MLR inhibitory activity was absorbed by autologous maternal T-lymphoblasts induced by stimulation with lymphocytes having the paternal HLA-DR type but not by those induced by stimulation with lymphocytes having other HLA-DR types; and (iii) the maternal serum inhibited the proliferation of autologous T-cells, but not of non-autologous T-cells, stimulated with paternal lymphocytes. These results indicate that the maternal sera contained antibodies directed against particular idiotypes expressed by autologous TCR that recognized paternal allogenic HLA-DR antigens. The complement-dependent killing of maternal T-lymphoblasts confirmed that the MLR inhibitory activity was due to immunoglobulin (Ig) molecules, possibly IgG, but not to other humoral factors or suppressing cells. Although Sugi et al. (1991) and Lubinski et al. (1993) detected antibodies to autologous activated T-cells in the sera of RSA patients who had received immunotherapy, the specificities and possible target molecules of these antibodies were not analysed. In this sense, this is first study to suggest that anti-idiotypic antibodies directed against autologous TCR are induced in RSA patients who receive immunotherapy with paternal lymphocytes.

Anti-TCR idiotypic antibodies were detected in four of the five patients whose pregnancies were maintained, but in none of the five patients whose pregnancies resulted in abortion. This result indicates that the presence of the anti-TCR idiotypic antibodies correlates strongly with successful pregnancy. Although in this study we focused our analysis on the antibodies against the maternal T-cells, the reactor cells against the fetus, the antibodies against the paternal HLA, the target of the maternal T-cells, may be another important humoral factor to be considered. In our previous study, in which we tested a large number of samples, anti-paternal HLA antibody was induced in a significant number of women who had successful pregnancies compared with that of women with unsuccessful pregnancies, after immunotherapy (K.Ito, unpublished). This finding indicates that the presence of anti-HLA antibody, together with the anti-TCR idiotypic antibody detected in this study, is another indicator for successful pregnancy in patients given immunotherapy.

Some reports have disputed the efficiency of immunotherapy (Cauchi et al., 1991Go; Ho et al., 1991Go; Gatenby et al., 1993Go; Perino et al., 1997Go). Christiansen (1996), in a review, stated that recurrent miscarriage is caused by some immunological or non-immunological factors, and therefore specific therapies are needed in order to prevent recurrent miscarriage. Although it is questionable whether all our RSA patients required this immunotherapy, it is interesting that paternal lymphocyte immunization induced anti-TCR idiotypic antibody in successful cases, but not in unsuccessful ones. In successful cases, paternal lymphocytes may act as an immunogen to enhance the maternal immunoresponse, and to induce antibodies as an immunological regulator for maintaining pregnancy. To clarify the correlation between the presence of anti-TCR idiotypic antibody and successful pregnancy, future studies of larger numbers of RSA patients are required.

The aetiology of RSA is still not clear. One possible mechanism is that the RSA patient has some kind of defect in the immunological regulation that protects the fetus from maternal attack. Immunotherapy with paternal lymphocytes has been performed in an attempt to modulate this immunological regulation in RSA patients, and a high success rate of maintenance of pregnancy has been obtained. Our previous results and those of other investigators have suggested that the anti-paternal HLA antibody produced by the immunotherapy is involved in the successful maintenance of pregnancy. These antibodies would mask the fetal HLA antigens and prevent them from being attacked by the maternal T-cells. The anti-TCR idiotypic antibody that has been reported to be present in the sera of normal pregnant women (Sucui-Foca et al., 1983; Singal et al., 1984Go) and which was detected in this study would provide another mechanism for the immunotherapy effect. After immunization with paternal lymphocytes, maternal T-cells recognizing paternal HLA antigens (one of the HLA antigens of the fetus) would expand and serve as immunogens to produce anti-TCR idiotypic antibodies. The anti-TCR idiotypic antibody would then bind specifically to the TCR and suppress the maternal immunoresponse against the fetus, allowing the fetus to escape the maternal immunological attack.


    Acknowledgments
 
We thank the midwives at the National Okura Hospital for their assistance with blood collection and Mr. K.Aoki and other staff at the Radioisotope Center of Kitasato University for their assistance with radioisotope handling.


    Notes
 
4 To whom correspondence should be addressed Back

5 Present address: Department of Obstetrics and Gynecology, Jikei University School of Medicine, 3–25–8 Nishishinbashi, Minato-ku 105-0003, Japan Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Agrawal, S., Kishore, R., Halder, A. et al. (1995) Outcome of pregnancy in women with recurrent spontaneous abortion following immunotherapy with allogeneic lymphocytes. Hum. Reprod., 10, 2280–2284.[Abstract]

Ahrons, S. (1971) Leukocyte antibodies: occurrence in primigravidae. Tissue Antigens, 1, 178–183.[Medline]

Carp, H.J.A., Toder, V., Torchinsky, A. et al. (1997) Allogenic leukocyte immunization after five or more miscarriages. Hum. Reprod., 12, 250–255.[Abstract]

Cauchi, M.N., Lim, D., Young, D.E. et al. (1991) Treatment of recurrent aborters by immunization with paternal cells – controlled trial. Am. J. Reprod. Immunol., 25, 16–17.[ISI][Medline]

Christiansen, O.B. (1996) A fresh look at the causes and treatments of recurrent miscarriage, especially its immunological aspects. Hum. Reprod., 2, 271–293.

Clark, D.A., Slapsys, R., Chaput, A. et al. (1986) Immunoregulatory molecules of trophoblast and decidual suppressor cell origin at the maternofetal interface. Am. J. Reprod. Immunol. Microbiol., 10, 100–104.[Medline]

Daya, S. and Gunby, J. (1994) The effectiveness of allogeneic leukocyte immunization in unexplained primary recurrent spontaneous abortion. Am. J. Reprod. Immunol., 32, 294–302.[ISI][Medline]

Daya, S., Clark, D.A., Devlin, C. et al. (1985) Suppressor cells in human decidua. Am. J. Obstet. Gynecol., 151, 267–270.[ISI][Medline]

Doughty, R.W. and Gelsthorpe, K. (1971) A scheme for screening blood samples to acquire lymphocyte cytotoxic antisera for tissue typing and results of an initial screening programme. Transplantation, 12, 429–433.[ISI][Medline]

Ellis, S.A., Palmer, M.S. and McMichael, A.J. (1990) Human trophoblast and the choriocarcinoma cell line BeWo express a truncated HLA class I molecule. J. Immunol., 144, 731–735.[Abstract/Free Full Text]

Gatenby, P.A., Cameron, K., Simes, R.J. et al. (1993) Treatment of recurrent spontaneous abortion by immunization with paternal lymphocytes: results of a controlled trial. Am. J. Reprod. Immunol., 29, 88–94.[ISI][Medline]

Ho, H.N., Gill, T.J., Hsieh, H.J. et al. (1991) Immunotherapy for recurrent spontaneous abortions in a Chinese population. Am. J. Reprod. Immunol., 25, 10–15.[ISI][Medline]

Jensen, K.G. (1962) Leucocyte antibodies in serums of pregnant women. Serology and clinic. Vox Sang (Basel), 7, 454–469.

Kovats, S., Main, E.K., Librach, C. et al. (1990) A class I antigen, HLA-G, expressed in human trophoblasts. Science, 248, 220–223.[ISI][Medline]

Lubinski, J., Vrdoljak, V.J., Beaman, K.D. et al. (1993) Characterization of antibodies induced by paternal lymphocyte immunization in couples with recurrent spontaneous abortion. J. Reprod. Immunol., 24, 81–96.[ISI][Medline]

Mandelboim, O., Pazmany, L., Davis, D.M. et al. (1997) Multiple receptors for HLA-G on human natural killer cells. Proc. Natl. Acad. Sci. USA, 94, 14666–14670.[Abstract/Free Full Text]

Mowbray, J.F., Gibbings, C., Liddell, H. et al. (1985) Controlled trial of treatment of recurrent spontaneous abortion by immunisation with paternal cells. Lancet, i, 941–943.

Obata, F., Ito, K., Kaneko, T. et al. (1991) HLA-DR gene frequencies in the Japanese population obtained by oligonucleotide genotyping. Tissue Antigens, 38, 124–132.[ISI][Medline]

Payne, R. (1962) The development and persistence of leukoagglutinins in parous women. Blood, 19, 411–424.[ISI]

Perino, A., Vassiliadis, A., Vucetich, A. et al. (1997) Short-term therapy for recurrent abortion using intravenous immunoglobulins: results of a double-blind placebo-controlled Italian study. Hum. Reprod., 12, 2388–2392.[Abstract]

Regan, L., Braude, P.R. and Hill, D.P. (1991) A prospective study of the incidence, time of appearance and significance of anti-paternal lymphocytotoxic antibodies in human pregnancy. Hum. Reprod., 6, 294–298.[ISI][Medline]

Reznikoff-Etievant, M.F., Durieux, I., Huchet, J. et al. (1988) Human MHC antigens and paternal leucocyte injections in recurrent spontaneous abortions. In Bear, R.W. and Sharp, F. (eds), Early Pregnancy Loss: Mechanisms and Treatment. RCOG, London, UK, pp. 375–384.

Rouas-Freiss, N., Marchal, R.E., Kirszenbaum, M. et al. (1997) The alpha 1 domain of HLA-G1 and HLA-G2 inhibits cytotoxicity induced by natural killer cells: is HLA-G the public ligand for natural killer cell inhibitory receptors ? Proc. Natl. Acad. Sci. USA, 94, 5249–5254.[Abstract/Free Full Text]

Singal, D.P., Butler, L., Liao, S.K. et al. (1984) The fetus as an allograft: evidence for antiidiotypic antibodies induced by pregnancy. Am. J. Reprod. Immunol., 6, 145–151.[ISI][Medline]

Suciu-Foca, N., Reed, E., Rohowsky, C. et al. (1983) Anti-idiotypic antibodies to anti-HLA receptors induced by pregnancy. Proc. Natl. Acad. Sci. USA, 80, 830–834.[Abstract]

Sugi, T., Makino, T., Maruyama, T. et al. (1991) A possible mechanism of immunotherapy for patients with recurrent spontaneous abortions. Am. J. Reprod. Immunol., 25, 185–189.[ISI][Medline]

Taylor, C. and Faulk, W.P. (1981) Prevention of recurrent abortion with leucocyte transfusions. Lancet, ii, 68–69.

Submitted on April 20, 1998; accepted on October 14, 1998.