Molecular Human Reproduction, Vol. 9, No. 11, 729-735,
November 2003
© 2003 European Society of Human Reproduction and Embryology
Article |
Immunogenicity of the soluble isoforms of HLA-G
Submitted on March 24, 2003; resubmitted on July 11, 2003. accepted on July 28, 2003
Departments of 1 Anatomy and Cell Biology, 2 Pathology and Laboratory Medicine and 3 Molecular and Integrative Physiology, University of Kansas Medical Center, Kansas City, KS 66160, USA and 4 Department of Human Genetics, The University of Chicago, Chicago, IL 60637, USA
5 To whom correspondence should be addressed at: Department of Anatomy and Cell Biology, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160-7400, USA. e-mail: jhunt{at}kumc.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Soluble class Ib HLA-G glycoproteins synthesized in the placenta are abundant in the pregnant uterus and circulate in maternal blood throughout pregnancy. To establish immunogenicity of these proteins, we tested sera from 64 women with at least one successful pregnancy (multigravid), 21 women who had never been pregnant, and 54 males for antibodies to epitopes present on recombinant sHLA-G isoforms (sHLA-G1, sHLA-G2) derived from HLA 6.0 cDNA (HLA-G*0101 allele). By indirect enzyme-linked immunosorbent assay, antibodies to sHLA-G isoforms were identified in six sera, all from multigravid women; all other sera were negative (P = 0.0083). Immunoblots showed that two of the positive sera reacted exclusively with sHLA-G1 and -G2 whereas four reacted to both sHLA-G and pooled HLA class I antigens. To establish potential relationships between anti-sHLA-G and exposure to foreign paternal alleles (*0101, *0103, *0104, *0106), all multigravid women and their partners were genotyped. No relationship between allelic disparity and antibody production was identified. Taken together, these results indicate that (i) tolerance to HLA-G is the usual condition as antibodies to HLA-G were not detected in 91% (58/64) multigravid women, and (ii) pregnancy stimulates loss of tolerance in 9% (6/64) of multigravid women. All six women delivered healthy babies, demonstrating that maternal antibodies to epitopes on sHLA-G do not abrogate pregnancy.
Key words: antibodies/HLA-G/immunity/pregnancy/tolerance
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
HLA-G is a class Ib gene that is expressed primarily during pregnancy in fetal tissues at the maternal-interface (reviewed by Hunt and Orr, 1992; Le Bouteiller and Mallet, 1997; Ober, 1998; Le Bouteiller et al., 2003). HLA-G is synthesized by trophoblast cells located in the placenta and decidua (Yelavarthi et al., 1991) as well as by fetal cells circulating in maternal blood (Van Wijk et al., 2001). The single transcript from this gene is alternatively spliced to produce messages that encode six to seven different isoforms (Ishitani and Geraghty, 1992; Fujii et al., 1994; Paul et al., 2000), many of which are present in placental lysates. Two isoforms are soluble due to the inclusion in the mature message of an intron 4 sequence containing a premature stop codon, which precludes translation of the transmembrane portion of the molecule (Fujii et al., 1994). One of the soluble isoforms, sHLA-G1 (also known as HLA-G5), contains the
1, 2 and 3 domains and associates with ß2-microglobulin (ß2m) light chains, whereas the second, sHLA-G2 (also known as HLA-G6), is missing the 2-domain and does not associate with ß2m (Tysoe-Calnon et al., 1991; P.J. Morales and J.S.Hunt, unpublished data). Soluble HLA-G is abundant in the maternal circulation during pregnancy (Rebmann et al., 1999; Hunt et al., 2000), and appears to be comprised mainly of sHLA-G2 and/or free heavy chains (Hunt et al., 2000).
Whether tolerance to HLA-G develops during fetal life remains uncertain. HLA-G has been identified in the fetal thymus (Mallet et al., 1999) and in amniotic fluid (McMaster et al., 1998; Rebmann et al., 1999). Thus, fetuses could develop tolerance through classical thymic selection or in the periphery by swallowing amniotic fluid throughout gestation. However, proof of tolerance to HLA-G has not been demonstrated. If tolerance is not developed during fetal life, mothers might respond immunologically by producing antibodies to the sHLA-G circulating in their blood during pregnancy (Rebmann et al., 1999; Hunt et al., 2000). Even if tolerance to self HLA-G were developed, mothers might synthesize antibodies to immunogenic epitopes on paternally derived allogeneic HLA-G. Three amino acid substitutions have been identified in HLA-G that result in differences from the most common allele (*0101) at the amino acid level: Thr
Ser at amino acid 31 (defining the HLA-G*0103), LeuIle at amino acid 110 (defining the HLA-G*0104), and ThrMet at amino acid 258 (defining the G*0106) (Hviid et al., 1997; Ober and Aldrich, 1997).
In this study we posed three questions: Are women tolerant to HLA-G? Can tolerance be broken during pregnancy? Are non-self, paternal HLA-G alleles immunogenic to mothers? To investigate these questions, we analysed sera from multigravid married and nulligravid single women and from male blood donors by enzyme-linked immunosorbent assay (ELISA) for antibodies that bind recombinant soluble sHLA-G1 and sHLA-G2. Positive sera were further analysed by immunoblotting for reactivity to HLA-G and pooled HLA class I antigens. Our results indicate that tolerance to HLA-G is the normal state but that production of antibodies to epitopes present on soluble isoforms may be stimulated during pregnancy.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Collection of sera and DNA
Sera were collected in March, 2001, from 64 married women who had at least one successful pregnancy (multigravid women) and from 21 women who had never been married or pregnant (nulligravid women). The married women had participated previously in studies on HLA and fertility (Ober et al., 1998b) and were selected to represent different combinations of husband–wife HLA-G genotypes. Their age and pregnancy history are shown in Table I. These women had been married only once and had only one sexual partner; their mean number of children was 5.6 (median 5, range 1–15) and their mean age was 44.5 years. Many were still in their reproductive years at the time of this study. Whole blood samples were collected in South Dakota and shipped on wet ice to Chicago, where the serum was separated within 48 h of collection and stored at –70°C until shipped on dry ice to the University of Kansas Medical Center. A second sample from five of six women with high values in the ELISA (see Results), which was collected 4–5 years earlier, was also tested. In addition, serum from a multigravid woman with five successful pregnancies who was homozygous for a null allele for the HLA-G1 isoform (Ober et al., 1998a) was included. All married women and their partners were genotyped for HLA-G alleles, as previously reported (Aldrich et al., 2001; Hviid et al., 2001). In addition, when possible, children of the 64 married women were genotyped (n = 234). The latter included nearly all children who were
6 years old at the time of our study. Male sera were collected from 54 blood donors at the Community Blood Center of Kansas City, as described (Hunt et al., 2000); these sera were stored continuously at –80°C.
|
Preparation of FLAG-sHLA-G1 and -sHLA-G2
The recombinant sHLA-G proteins were generated from cDNA sequences in HLA 6.0-transfected mouse fibroblasts and have been characterized (Morales et al., 2004). The 6.0 sequence exhibits the *0101 allele. The cDNA constructs for sHLA-G1 and -G2 were modified to express a FLAG® peptide sequence (APLADYKDDDDKLA) between the signal peptide and the N-terminus of the mature proteins. In brief, recombinant FLAG-sHLA-G1 and FLAG-sHLA-G2 proteins were generated in human embryonic kidney (HEK293) cells and were affinity-purified using the FLAG-M2 affinity resin (Sigma–Aldrich, USA). Both FLAG-sHLA-G1 and -G2 mRNAs terminate in intron 4 and their proteins are appropriately glycosylated. Recombinant FLAG-sHLA-G1 and -G2 are recognized in immunoblots by the mouse monoclonal antibody 16G1 (Figure 1A), which was generated to an intron 4-specific sequence (gift of D.Geraghty, Seattle, WA, USA) and mouse anti-FLAG monoclonal antibody, FLAG-M1 (Figure 1B) (Sigma–Aldrich) and in ELISA (Table II). FLAG-sHLA-G1 is associated with ß2m whereas FLAG-sHLA-G2 is not (Morales et al., 2004).
|
|
Preparation of GST-sHLA-G1 and -sHLA-G2
To confirm the patterns obtained using the FLAG-tagged proteins, all sera were tested against a second set of sHLA-G proteins using the same ELISA technique. The prokaryotic recombinant fusion proteins GST-sHLA-G1 and GST-sHLA-G2 were generated by cloning the blunt-ended cDNA encoding sHLA-G1 and -G2 into the SmaI polylinker site of the vector pGEX-5X-2 (Amersham Pharmacia Biotech, USA). The proteins were expressed in E. coli and purified by affinity chromatography using Glutathione-Sepharose 4B (Amersham Pharmacia Biotech). Briefly, each clone was grown in LB medium containing ampicillin (100 µg/ml) in an orbital shaker at 37°C overnight. This primary culture was used to inoculate a preparative culture, induced with 1 mmol/l isopropylthio-ß-D-galactoside (IPTG; Promega, USA) after the OD600 reached
0.5, and was then cultured for a further 3 h. The bacterial pellet was collected by centrifugation and the cells were resuspended in phosphate-buffered saline (PBS) containing 5 mmol/l EDTA and 0.15 mmol/l phenylmethylsulphonyl fluoride. Subsequently, the cells were disrupted by using a French-Press (SLM-Aminco, SLM Instruments, Inc., USA). The supernatant of E. coli harbouring the vector contained the GST protein. The GST-sHLA-G1 and GST-sHLA-G2 were localized in the insoluble fraction forming inclusion bodies. An extra step of denaturation–renaturation was required to perform the purification of the sHLA-G fusion proteins. Figure 1A shows that GST-sHLA-G1 and GST-sHLA-G2 (predicted molecular weight of 60.7 and 50.2 kDa respectively) were identified by 16G1, and Figure 1C demonstrates that both proteins were also detected with rabbit anti-GST (Santa Cruz Biotechnology, Inc., USA). The additional bands detected by anti-GST in Figure 1C are most likely due to degradation products containing the GST portion of GST-sHLA-G molecules but lacking the epitope recognized by 16G1. Degradation has been a consistent finding of the GST-tagged sHLA-G isoforms, while oligomerization has been a consistent finding for the FLAG-tagged isoforms. Due to their production in bacteria, the GST-sHLA-G proteins were not glycosylated (data not shown) and may or may not have been correctly folded.
ELISA assays
An indirect horseradish peroxidase (HRP)-based ELISA was used for detection of antibodies to HLA class I antigen, as previously described (Langat et al., 2002) with modifications. Purified preparations of sHLA-G1, sHLA-G2 or a mixture of pooled HLA proteins (Kao, 1987) were prepared at a concentration of 5 µg/ml in 0.01 mol/l PBS, pH 7.2 and used to coat wells of 96-well plates. The pooled HLA class I antigens were acquired from outdated platelet concentrates by affinity chromatography using the anti-pan HLA class I monoclonal antibody, W6/32, which detects all ß2m-associated HLA class I antigens (Brodsky and Parham, 1982). Wells treated with PBS or coated with GST fusion proteins were used as controls. After blocking with 5% milk (BioRad Laboratories, USA) in PBS containing 0.02% Tween-20 for 1 h at room temperature, the wells were washed and 50 µl of test serum, diluted 10-fold in blocking buffer, was added to each well. The plates were incubated for 30 min at 37°C and washed four times. Bound Ig was detected using HRP-conjugated goat anti-human IgGAM (Zymed, USA) followed by washing and the addition of 100 µl 3,3',5,5'-tetramethylbenzidine (TMB) substrate (Kirkgaard and Perry Laboratories, USA). The enzymatic reaction was stopped after 20 min incubation at room temperature by adding 100 µl of 1% sodium dodecyl sulphate (SDS)/well. The plates were then read at 630 nm using an ELx 808 (Bio-Tek Instruments, Inc., USA) microplate reader.
Identification of anti-sHLA-G1, anti-sHLA-G2 and anti-HLA class Ia by indirect ELISA
The indirect ELISA was used to identify antibodies to HLA-G in sera from multigravid women (n = 64), nulligravid women (n = 21) and men (n = 54). Each serum sample was assayed in duplicate on the same plate against PBS-treated or sHLA-G coated wells arranged in clusters to assure that incubation times and washing procedures were similar. Antigen-coated wells treated with blocking buffer in place of human serum were included on each plate to define assay background and confirm negligible binding of the secondary HRP-conjugate. A sample of pooled male serum was also included on each plate to evaluate reproducibility of the assay. A positive control which utilized the mouse monoclonal antibody which identifies a peptide derived from the intron 4 sequence (16G1, 1 µg/ml), was included to control for the possibility that differential reactivity to sHLA-G2 compared with HLA-G1 could be due to uneven coating of the microwells. The coefficients of variation of the OD values obtained for this sample among different plates for all antigens ranged from 5 to 14%.
Analysis by immunoblotting
One microgram of purified proteins (FLAG- and GST-tagged proteins, pooled HLA) was resuspended in Laemmli buffer under reducing conditions and separated by electrophoresis in acrylamide gels (10 or 15% SDS–polyacrylamide gel electrophoresis). The proteins were electro-transferred onto nitrocellulose membranes (Schleider and Schuell, USA) and sHLA-G was detected using 16G1 or rabbit anti-GST (Santa Cruz Biotechnology, Inc., USA). The membrane subsequently was incubated with a peroxidase-conjugated anti-mouse IgG antibody or anti-rabbit IgG (both from Jackson Immuno Research Inc., USA). The signal was developed using the chemiluminescent substrate SuperSignal West Pico (Pierce, USA) and detected by exposing to Hyperfilm ECLTM (Amersham Pharmacia Biotech, USA). The human serum samples to be tested for the presence of antibodies against sHLA-G antigens were diluted to either 1:500 or 1:1000 in blocking buffer (3% non-fat dry milk in 0.05 mol/l Tris-buffered saline, pH 8.0, containing 0.05% Tween-20) incubated at RT for 30 min in a nutator, and then used as a first antibody. The bound antibody was detected using anti-human IgG-HRP (Zymed) at a dilution of 1:10 000 and detected as described above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Identification of anti-HLA-G1 and anti-HLA-G2 in sera from multigravid but not nulligravid women by indirect ELISA
Table II shows that male sera tested against FLAG-tagged sHLA-G1 and -G2 produced very low ELISA readings of –0.01 ± 0.04 (mean and SD) and –0.02 ± 0.04 respectively. Values for male sera were also very low when tested against GST-tagged sHLA-G1 and -G2 (0.03 ± 0.04 and 0.05 ± 0.04 respectively). Female sera that yielded values >2 SD above the mean for the male sera and that reacted with at least one FLAG-tagged and one GST-tagged recombinant protein were designated as positive.
Sixty-four sera from multigravid women, 21 sera from nulligravid women and 54 sera from men were tested in the two ELISA assays. All sera from the nulligravid women and the 54 men were negative. Sera from six of the 64 multigravid women (9%) were positive with both FLAG-tagged and GST-tagged recombinant proteins. None of these six women had been previously transfused. Thus, there was a significant association between the presence of antibodies and exposure to pregnancy (6/64 versus 0/75; Fisher’s exact test, P = 0.0083). One additional sample from a married woman exhibited reactivity with only GST proteins and was not tested further. None reacted with FLAG-tagged but not GST proteins.
The results of the ELISA assays for the six positive women are shown in Table II. To evaluate the reproducibility of the ELISA, the test was repeated on a second serum sample obtained previously from five of these six women (HT1, HT2, HT31, HT32, HT33). At the time of the previous sampling (1996–1997) the number of pregnancies for each woman was 14, 3, 6, 4 and 3 respectively. The repeated assays yielded essentially the same results as in the initial screen; although the absolute values varied, all of the second samples were positive as defined above.
Specificity of ELISA for HLA-G and pooled HLA class I antigen
To investigate the specificity of the HLA-G ELISA, we tested 85 of the female sera in an ELISA where the microwells were coated with pooled HLA class I protein. Twenty-seven per cent (17 of 64) of the sera samples from multigravid women were positive with pooled HLA class I antigen; none of the sera from nulligravid women was positive. This was not unexpected as up to 40% of multigravid women synthesize antibody to HLA (Payne and Rolfs, 1958) and the 17 women with antibodies all were incompatible with their partners at the HLA-A and/or HLA-B loci (data not shown). Five of the six serum samples that were positive in the HLA-G ELISA were negative in the pooled HLA ELISA, where antigens bound to the microplates would comprise primarily the HLA class Ia antigens, HLA-A, -B, -C. The one serum sample (HT32) that was positive in both the HLA-G ELISA and pooled HLA class I antigen ELISA was shown by immunoblotting to contain both antibodies. Collectively, these assays demonstrated high specificity in the HLA-G ELISA; the anti-HLA-G ELISA did not identify anti-HLA class I antibodies as anti-HLA-G, and, conversely, the anti-HLA class I ELISA did not falsely identify anti-HLA-G as reactive with other HLA class I antigens.
Analysis by immunoblotting demonstrates that anti-sHLA-G antibodies stimulated in pregnancy may or may not be accompanied by anti-HLA class Ia antibody
Immunoblots where maternal antibodies were blotted against GST, GST-sHLA-G1, GST-sHLA-G2, FLAG-sHLA-G1, FLAG-sHLA-G2 and pooled HLA class I antigen were performed to determine the specificities of the antibodies that were detected in indirect ELISA in the six multigravid women. In each case, the patterns of reactivity were those expected from the known migration patterns of the recombinant proteins as shown in Figure 1.
One serum sample (from subject HT1) that exhibited exclusively anti-HLA-G in the ELISA showed no reactivity to pooled HLA or to GST (Figure 2A). The same pattern was seen in subject HT2 (data not shown). Sera from the other four subjects who were identified as positive for anti-sHLA-G in ELISA assays exhibited reactivity to both HLA-G and pooled HLA. HT32 is shown in Figure 2B. The same general pattern was observed in sera from subjects HT66, HT31 and HT33 (data not shown). These experiments demonstrate that the ELISA assays were less sensitive than the immunoblotting assays. More importantly, while some women produce a broad range of anti-class I antibodies, some produce antibodies exclusively to sHLA-G.
|
Sera that were negative in the ELISA assays were negative in immunoblots
To confirm the low-to-negative reactivity of most serum samples in the sHLA-G ELISA assays, we tested several sera by immunoblotting against the sHLA-G glycoproteins. The results were uniformly negative (data not shown). Positive and negative controls (not shown) were as expected.
In a previous study, we identified a woman with a homozygous deletion that abrogated production of the HLA-G1 isoform (Ober et al., 1998a). This woman had five successful pregnancies (including one twin pregnancy) with a partner who did not carry the null allele. No antibodies were detected in her serum by either ELISA or immunoblot analysis (data not shown), despite the fact that she was exposed to sHLA-G1 proteins for the first time during pregnancy.
Modelling of HLA-G alleles
Four alleles have been identified in the HLA-G gene, with *0101 being the most common (Ober and Aldrich, 1997; Hviid et al., 1997). In order to evaluate potential immunogenicity of the *0103, *0104 and *0106 alleles, we first used the SWISS-MODEL program (Guex and Peitsch, 1997) to predict the three-dimensional structure of the HLA-G1 heavy chain. The HLA-A*0201, HLA-Cw4 and HLA-E heavy chains were used as templates. Figure 3 shows that amino acids 31 and 110, which define the *0103 and *0104 alleles respectively, are located in the beta sheet structures forming the base of the antigen binding cleft. Both are unlikely binding entities for antigens; neither is in position to interact with peptide in the cleft (Ober and Aldrich, 1997). The *0104 allele in the 2-domain has an exterior position and could be more accessible and therefore more immunogenic than the *0103 allele in the 1-domain. The *0106 allele (amino acid 258 in the 3-domain) is located in the pleated sheet structure of the 3 domain where it is unlikely to be highly immunogenic but might affect recognition of CD8 (Purbhoo et al., 2001).
|
In the sHLA-G2 isoform, the *0103 allele in the
1 region remains potentially antigenic but the *0104 allele is unexpressed because amino acid 110 is located in the 2 domain, which is absent in HLA-G2. The *0106 allele would be expressed in this molecule and could be important to binding of CD4 (König et al., 1992) since HLA-G2 H chains are predicted to form homodimers (Ishitani and Geraghty, 1992).
Production of anti-sHLA-G is not invariably associated with exposure to non-self, paternal HLA-G
To determine whether non-self paternal HLA-G alleles stimulated maternal antibody production during pregnancy in the six women with positive results, we classified couples by their HLA-G genotypes. Table III shows that thirty-seven of the multigravid women were potentially exposed to foreign paternal alleles (incompatible) and 27 women were entirely compatible with their husbands with respect to HLA-G alleles.
|
Among the 37 incompatible couples, the following HLA-G alleles were present in the father but not in the mother: HLA-G*0101 (n = 5), HLA-G*0103 (n = 7), HLA-G*0104 (n = 12), HLA-G*0106 (n = 8), both HLA-G*0103 and *0104 alleles (n = 2), and both HLA-G*0104 and *0106 (n = 2). Some of the children of these couples were available for genotyping (234 of 359). Of the 37 mothers who could have been exposed to a foreign HLA-G allele during pregnancy, 25 had at least one genotyped child who inherited a foreign allele (Table III).
Of the six sera with positive results in the sHLA-G ELISA assays, two were incompatible (HT2, HT66) and four were compatible (HT1, HT31, HT32, HT33) with their partners for HLA-G alleles (Table III). Thus, antibodies were generated in mothers regardless of exposure to allogeneic HLA-G alleles in their fetuses (2/37 versus 4/27; Fisher’s exact test, P = 0.23). Importantly, none of the sera comprised exclusively allele-specific antibody; the rsHLA-G1 and -G2 used to detect anti-sHLA-G in ELISA assays were derived from the HLA 6.0 *0101 allele, which does not bear the amino acid sequences encoding the *0103 and *0104 alleles, and the allelic disparities of antibody-producing mothers were *0103 and *0104 (Table IV).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The results of this study indicate that tolerance to HLA-G is a normal and expected condition; antibody to sHLA-G was undetectable in the sera of 91% of women with one or more successful pregnancies (range 1–15). Further, exposure to a foreign paternal HLA-G allele only occasionally resulted in antibody production; 94% of women exposed to non-self HLA-G alleles during pregnancy failed to produce anti-HLA-G. Tolerance is presumably developed during fetal life when production of HLA-G membrane-bound and soluble isoforms is high in placentas (reviewed in Hunt and Orr, 1992; Le Bouteiller and Mallet, 1997; Ober, 1998). HLA-G is also present in amniotic fluid (McMaster et al., 1998) and has been identified in the fetal thymus (Mallet et al., 1999). Thus, fetuses could develop tolerance through classical thymic selection or in the periphery by the normal swallowing of amniotic fluid throughout gestation.
Nonetheless, this study demonstrated that tolerance to HLA-G can be overcome by exposure during pregnancy; six of the 64 multigravid women (9%) produced readily detectable antibodies to sHLA-G whereas antibody was not detected in 54 men and 21 nulligravid women (P = 0.0083). Breaking of tolerance could be related to a number of conditions. First, constant stimulation of the immune system could be important; subject HT1, who had the highest levels of antibody, had 14 prior pregnancies, and all of the antibody-producing mothers had multiple pregnancies (range 3–15). Second, allelic differences between the mother and fetus could contribute. However, our data failed to support this hypothesis. Even though the amino acids in the *0103 and *0104 alleles were spatially positioned for immunogenicity, we failed to show an association between exposure to allogenic HLA-G and antibody production. Four of 27 (15%) women who had entirely compatible pregnancies and two of 37 (5%) of women who had incompatible pregnancies had detectable antibodies (P = 0.230). Even if we consider only women who had confirmed exposure to a foreign HLA-G allele during pregnancy, the prevalence of antibodies between exposed and unexposed women is not significant (4/27 versus 2/25, P = 0.670). Further studies are in progress to determine whether mothers incompatible with their partners produce allele-specific antibody, but the data collected thus far suggest that this is not likely.
Third, ectopic expression of sHLA-G1 or -G2 in adult tissues could elicit antibody production. HLA-G protein has been identified in activated macrophages, dendritic cells and T cells in inflamed tissues (Yang et al., 1996; Khosrotehrani, 2001; Pangault et al., 2002). However, if this route of sensitization resulted in antibody production, we would expect antibodies to be present in males and nulligravid women at frequencies similar to those in multigravid women but this was not the case. Lastly, it is possible that cross-reactive autologous or allogeneic antigen could account for these results. We have not yet tested the sera of women producing anti-sHLA-G for antibody of other specificities such as might be found in autoimmune disorders. Nor do we know the histories of these women with regard to microorganisms that could bear antigens mimicking HLA-G. However, women in this study had no evidence of autoimmune disease, and organisms mimicking HLA-G have not been reported.
It is also notable that although antibodies were produced against both sHLA-G1 and -G2 by all six women, anti-sHLA-G2 was more abundant than anti-sHLA-G1 whether the ELISA plate was coated with eukaryotic or prokaryotic sHLA-G. This could result if the 1/3 domain bridge that is present only in the HLA-G2 isoforms is particularly immunogenic and/or if mothers are exposed to higher levels of sHLA-G2 than sHLA-G1. The latter is consistent with two observations. First, evidence has been collected suggesting that sHLA-G2 may be the predominant soluble isoform in the sera of pregnant women (Hunt et al., 2000). Second, we could not detect antibodies in a woman who is homozygous for an HLA-G1 null allele despite the fact that she was exposed to HLA-G1 proteins in five (successful) pregnancies (Ober et al., 1998a). This suggests that the HLA-G1 isoform may not be highly immunogenic. In support of these ideas is our finding that antibody associated with polymorphisms in the 2-domain (e.g. *0101/*0101 mothers with *0104+ fetuses or *0104/*0104 mothers with *0101+ fetuses) was not detected. However, it is also possible that antibodies generated against class I (or even class II) HLA are more cross-reactive to sHLA-G2 than to -G1 as all six women with anti-sHLA-G antibodies were exposed to foreign HLA-A and/or HLA-B alleles during pregnancy. In fact, analysis by immunoblotting demonstrated that only two women (HT1, HT2) synthesized antibody exclusively to sHLA-G whereas the others generated antibody to both sHLA-G and pooled HLA class I. Among the latter, it is possible that a polyclonal antibody response to HLA class I included antibodies that cross-react with HLA-G, as discussed above. If this were the case, antibodies entirely exclusive to sHLA-G may be quite uncommon, as these were present in only two (HT1 and HT2) of the six (3%) multigravid women producing these antibodies.
Regardless of the mechanism for production, the presence of antibody to sHLA-G did not preclude normal pregnancy; all six women with antibodies had many successful pregnancies. The most recent pregnancy in these women ranged from <1 to 33 years, indicating that antibodies can persist for many years. We also note that the miscarriage rate in these six women was low (4/38 pregnancies, 11%). Thus, the presence of antibodies is not associated with adverse pregnancy outcome.
In summary, these data demonstrate that multigravid women do not ordinarily synthesize antibodies against HLA-G, indicating that tolerance toward HLA-G is the normal state. Such tolerance is difficult to break as even a woman who is homozygous for a null allele that prevents synthesis of the HLA-G1 isoforms did not demonstrate antibody against the HLA-G1 proteins produced by all six of her fetuses. Of equal importance, we show that tolerance can be broken irrespective of exposure to non-self HLA-G alleles during pregnancy, although this occurs in a minority (<9%) of pregnancies. Lastly, our study demonstrates that the presence of HLA-G antibodies does not preclude successful pregnancy, as is also true for antibody to the HLA class I proteins.
|
Acknowledgements |
|---|
The authors appreciate the gift of the mouse monoclonal antibody to HLA-G intron 4 sequence (16G1) from D.Geraghty, Fred Hutchinson Cancer Research Institute and the gift of HLA class Ia proteins from K.J.Kao, University of Florida. This work was supported by grants from the National Institutes of Health to J.S.H. (HD26429, HD35859, HD39878), C.O. (HD21244), the University of Kansas Medical School, Center for Reproductive Sciences (U54 HD33994), and the Kansas Mental Retardation Center (HD02528).
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Aldrich, C.L., Stephenson, M.D., Karrison, T., Odem, R.R., Branch, D.W., Scott, J.R., Schreiber, J.R. and Ober, C. (2001) HLA-G genotypes and pregnancy outcome in couples with unexplained recurrent miscarriage. Mol. Hum. Reprod., 7, 1167–1172.
Brodsky, F.M. and Parham, P. (1982) Monomorphic anti-HLA-A,B,C monoclonal antibodies detecting molecular subunits and combinatorial determinants. J. Immunol., 128, 129–135.[Abstract]
Fujii, T., Ishitani, A. and Geraghty, D.E. (1994) A soluble form of the HLA-G antigen is encoded by a messenger ribonucleic acid containing intron 4. J. Immunol., 153, 5516–5524.[Abstract]
Guex, N. and Peitsch, M.C. (1997) SWISS-MODEL and the Swiss-PDBViewer: an environment for comparative protein modelling. Electrophoresis, 18, 2714–2723.[CrossRef][Web of Science][Medline]
Hunt, J.S. and Orr, H.T. (1992) HLA and maternal–fetal recognition. FASEB J., 6, 2344–2348.[Abstract]
Hunt, J.S., Jedhav, L., Chu, W., Geraghty, D.E. and Ober, C. (2000) Soluble HLA-G circulates in mothers during pregnancy. Am. J. Obstet. Gynecol., 183, 682–688.[CrossRef][Web of Science][Medline]
Hviid, T.V., Meldgaard, M., Sorensen, S. and Morling, N. (1997) Polymorphism of exon 3 of the HLA-G gene. J. Reprod. Immunol., 35, 31–42.[CrossRef][Web of Science][Medline]
Hviid, T.V., Christiansen, O.B., Johansen, J.K., Hviid, U.R., Lundegaard, C., Moller, C. and Morling, N. (2001) Characterization of a new HLA-G allele encoding a nonconservative amino acid substitution in the alpha3 domain (exon 4) and its relevance to certain complications in pregnancy. Immunogenetics, 53, 48–53.[CrossRef][Web of Science][Medline]
Ishitani, A., and Geraghty, D.E. (1992) Alternative splicing of HLA-G transcripts yields proteins with primary structures resembling both class I and class II antigens. Proc. Natl Acad. Sci. USA, 89, 3947–3951.
Kao, K.J. (1987) Plasma and platelet HLA in normal individuals: quantitation by competitive enzyme-linked immunoassay. Blood, 70, 282–286.
Khosrotehrani, K. (2001) HLA-G expression in atopic dermatitis. J. Invest. Dermatol., 117, 750–752.[CrossRef][Web of Science][Medline]
König, R., Huang, L.-Y and Germain, R.N. (1992) MHC class II interaction with CD4 mediated by a region analogous to the MHC class I binding site for CD8. Nature, 356, 796–798.[CrossRef][Medline]
Langat, D.K., Morales, P.J., Fazleabas, A.T., Mwenda, J.M and Hunt, J.S. (2002) Baboon placentas express soluble and membrane-bound Paan-AG proteins encoded by alternatively spliced transcripts of the class Ib major histocompatibility complex gene, Paan-AG. Immunogenetics, 54, 164–173.[CrossRef][Web of Science][Medline]
Le Bouteiller, P. and Mallet, V. (1997) HLA-G and pregnancy. Rev. Reprod., 2, 7–13.[Abstract]
Le Bouteiller, P., Legrand-Abravanel, F. and Solier, C. (2003) Soluble HLA-G1 at the materno-fetal interface. Trophoblast Res., 17, 510–515.
Mallet, V., Blaschitz, A., Crisa, L., Schmitt, C., Fournel, S., King, A., Loke, Y.W., Dohr, G. and Le Bouteiller, P. (1999) HLA-G in the human thymus: a subpopulation of medullary epithelial but not CD83+ dendritic cells expresses HLA-G as a membrane-bound and soluble protein. Int. Immunol., 11, 889–898.
McMaster, M., Zhou, Y., Shorter, S., Kapasi, K., Geraghty, D., Lim, K.H. and Fisher, S. (1998) HLA-G isoforms produced by placental cytotrophoblasts and found in amniotic fluid are due to unusual glycosylation. J. Immunol., 160, 5922–5928.
Morales, P.J., Pace, J.L., Platt, J.S., Phillips, T.A., Morgan, K., Fazleabas, A.T. and Hunt, J.S. (2004) Placental cell expression of HLA-G2 isoforms is limited to the invasive trophoblast phenotype. J. Immunol., submitted.
Ober, C. (1998) HLA and pregnancy: the paradox of the fetal allograft. Am. J. Hum. Genet., 62, 1–7.[CrossRef][Web of Science][Medline]
Ober, C. and Aldrich, A. (1997) HLA-G polymorphisms: neutral evolution or novel function? J. Reprod. Immunol., 36, 1–21.[CrossRef][Web of Science][Medline]
Ober, C, Aldrich, C., Rosinsky, B., Robertson, A., Walker, M.A., Willadsen, S., Verp, M.S., Geraghty, D.E. and Hunt, J.S. (1998a). HLA-G1 protein expression is not essential for fetal survival. Placenta, 19, 127–132.[CrossRef][Web of Science][Medline]
Ober, C., Hyslop, T., Elias, S., Weitkamp, L.R. and Hauck, W.W. (1998b) Human leukocyte antigen matching and fetal loss: results of a 10-year prospective study. Hum. Reprod., 13, 33–38.
Pangault, D., Le Friec, G., Caulet-Maugendre, S., Lena, H., Amiot, L., Guilloux, V., Onno, M. and Fauchet, R. (2002) Lung macrophages and dendritic cells express HLA-G molecules in pulmonary diseases. Hum. Immunol., 63, 83–90.[CrossRef][Web of Science][Medline]
Paul, P., Adrian Cabestre, F., Ibrahim, E.C., Lefebvre, S., Khalil-Daher, I., Vazeux, G., Moya Quiles, R.M., Bermond, F., Dausset, J. and Carosella, E.D. (2000) Identification of HLA-G7 as a new splice variant of the HLA-G mRNA and expression of soluble HLA-G5, -G6, and -G7 transcripts in human transfected cells. Hum. Immunol., 61, 1138–1149.[CrossRef][Web of Science][Medline]
Payne, R, and Rolfs, M.R. (1958) Fetomaternal leucocyte incompatibility. J. Clin. Invest., 37, 1756–1763.[Web of Science][Medline]
Purbhoo, M.A., Boulter, J.M., Price, D.A., Vuidepot, A.-L., Hourigant, C.S., Dunbar, P.R., Olson, K., Dawson, S.J., Phillips, R.E., Jakobsen, B.K. et al. (2001) The human CD8 coreceptor effects cytotoxic T cell activation and antigen sensitivity primarily by mediating complete phosphorylation of the T cell receptor 
chain. J. Biol. Chem., 276, 32786–32792.
Rebmann, V., Pfeiffer, K., PäBler, M., Ferrone, S., Maier, S., Weiss, E. and Grosse-Wilde, H. (1999) Detection of soluble HLA-G molecules in plasma and amniotic fluid. Tissue Antigens, 53, 14–22.[CrossRef][Web of Science][Medline]
Tysoe-Calnon, V.A., Grundy, J.E. and Perkins, S.J. (1991) Molecular comparisons of the ß2-microglobulin-binding site in class I major-histocompatibility-complex a-chains and proteins of related sequences. Biochem. J., 277, 359–369.
VanWijk, I. J., Griffioen, S., Tjoa, M.L., Mulders, M.A.M., van Vugt, J.M.G., Loke, Y.W. and Oudejans, C.B.M. (2001) HLA-G expression in trophoblast cells circulating in maternal peripheral blood during early pregnancy. Am. J. Obstet. Gynecol., 184, 991–997.[CrossRef][Web of Science][Medline]
Yang, Y., Chu, W., Geraghty, D.E. and Hunt, J.S. (1996) Expression of HLA-G in human mononuclear phagocytes and selective induction by interferon-
. J. Immunol., 156, 4224–4231.[Abstract]
Yelavarthi, K.K., Fishback, J.L. and Hunt, J.S. (1991) Analysis of HLA-G mRNA in human placental and extraplacental membrane cells by in situ hybridization. J. Immunol., 146, 2847–2854.[Abstract]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Y. Tan, J. F.V. Ho, Y. S. Chong, A. Loganath, Y. H. Chan, J. Ravichandran, C. G. Lee, and S. S. Chong Paternal contribution of HLA-G*0106 significantly increases risk for pre-eclampsia in multigravid pregnancies Mol. Hum. Reprod., May 1, 2008; 14(5): 317 - 324. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Hunt, M. G. Petroff, R. H. McIntire, and C. Ober HLA-G and immune tolerance in pregnancy FASEB J, May 1, 2005; 19(7): 681 - 693. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||