Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The HLA-G genotype is potentially associated with idiopathic recurrent spontaneous abortion

The HLA-G genotype is potentially associated with idiopathic recurrent spontaneous abortion Abstract The causes for recurrent spontaneous abortion (RSA) remain unknown in a large proportion of the cases. Human leukocyte antigen (HLA)-G and HLA-E are expressed on invasive trophoblast cells, and are supposed to confer to materno–fetal tolerance. A total of 14 different nucleotide sequences have been described for HLA-G, including one dysfunctional null allele (HLA-G*0105N), while five different sequences have been described for HLA-E. In this study, 78 RSA couples and 52 fertile controls were typed for HLA-G and HLA-E by direct sequencing or single strand conformational polymorphism (SSCP) respectively. The overall analysis showed no significant difference in allele frequencies for either HLA-G or HLA-E between the two groups. However, HLA-G allele frequencies in women who had suffered from five or more RSA differed significantly from fertile controls (P = 0.001), and from women who had undergone three or four RSA (P = 0.027). Detailed analysis demonstrated a significant increase in the proportion of the HLA-G alleles *01013 and *0105N in the whole group of RSA women compared with fertile controls (P = 0.007). When studying the prognostic value of HLA genotyping for pregnancy outcome (n = 41), 31 patients (76%) gave birth to a living child without performing immunotherapy. Seven out of 10 (70%) couples suffering from a further RSA carried the HLA-G*01013 or *0105N allele, compared with 10 out 31 (32%) couples giving birth (P = 0.06). This study suggests that the HLA-G genotype may be a contributing factor in RSA. genotyping, HLA-G, HLA-E, recurrent spontaneous abortion Introduction Although numerous studies have tried to determine the reasons for habitual abortion, definite causes for recurrent spontaneous abortion (RSA) remain unknown in a large proportion of the cases (Harger et al., 1983). Identified causes for RSA include uterine malformation, antiphospholipid antibodies, and cytogenetic anomalies, but these account for only 20–50% of the cases. Immunological causes and consecutive immunological therapies are proposed if no other reasons for the RSA are found. Nevertheless, in idiopathic cases, psychological support can achieve results comparable with those of immunotherapy (Stray-Pederson and Stray-Pederson, 1984). Human leukocyte antigens (HLA) as major determinants of allograft rejection have been studied intensively in the context of RSA. Conflicting data have been obtained on matters such as HLA sharing (Beer et al., 1985, Christiansen et al., 1989), HLA risk haplotypes (Ober et al., 1998a; Christiansen et al., 1999) and, more recently, on HLA-G (Karhukorpi et al., 1997) and HLA-E (Steffensen et al., 1998) allelic distribution. HLA-G is a non-classical HLA-Ib molecule which has gained a great deal of attention due to its selective and co-dominant expression on invasive trophoblast (Kovats et al., 1990) and in the thymus (Crisa et al., 1997). Some studies have demonstrated that invasive trophoblast lacks the polymorphic classical transplantation antigens, with the exception of HLA-C (King et al., 1996), but expresses the non-classical HLA molecules HLA-G and HLA-E (Wei and Orr, 1990; Le Bouteiller et al., 1999). Both HLA-G and HLA-E are less polymorphic than classical HLA-I molecules and their polymorphisms do not involve the peptide binding groove, as it is typical for classical HLA-I molecules in order to present different peptides (van der Ven et al., 1998a). HLA-E protein expression is stabilized by binding the leader peptide of classical HLA-I and HLA-G molecules (Llano et al., 1998). HLA-G binds a nonamer with leucine or isoleucine at position 2, proline at position 3, and leucin at position 9 (O'Callaghan and Bell, 1998). The lack of polymorphism in the peptide binding groove of HLA-E and HLA-G supports the hypothesis of a function that differs from classical HLA-I (which are extremely polymorphic), in order to present as many different peptides as possible. Nevertheless, HLA-G mismatches are able to induce an immune response in mice (Horuzko et al., 1997) and, therefore, the tolerance of a semiallogenic fetus during pregnancy remains a matter of discussion. The detailed function of HLA-E and HLA-G is still under investigation, but in-vitro studies have demonstrated that both interact with different inhibitory natural killer (NK) cell receptors (Braud et al., 1998), and that they may be able to protect cells which lack classical HLA-I molecules from NK cell attack (Rouas-Freiss et al., 1997). However, this question has not yet been finally clarified (King et al., 2000). For soluble HLA-G, induction of apoptosis in CD8+ T-cells has been demonstrated (Fournel et al., 2000). Hence, HLA-G and HLA-E are possible candidates for inhibiting the maternal anti-fetal immune response. HLA-G mRNA exists in at least six different alternative splicing forms, with two of them encoding for a soluble protein (soluble HLA-G) (Fujii et al., 1994). Patients with lower soluble HLA-G concentrations in the peripheral blood are at a greater risk of undergoing an RSA after IVF than patients with higher soluble HLA-G concentrations (Pfeiffer et al., 2000). An earlier study on non-pregnant controls has shown that the serum concentrations of soluble HLA-G are determined by the HLA-G genotype (Rebmann et al., 1999). For HLA-G, 14 different nucleotide sequences, encoding four different proteins (HLA-G *0101, *0103, *0104 and *0105) have been described. Among these, the HLA-G *0105N and *01013 alleles are associated with significantly lower serum concentrations of soluble HLA-G (Rebmann et al., 2001). HLA-G*0105N includes a one base deletion polymorphism in exon 3, which leads to a consecutive frame shift, and is therefore expected to encode a non-functional protein. The reason why the silent polymorphism, HLA-G*01013, is associated with lower serum HLA-G concentrations is still unknown, although it may be in linkage disequilibrium with polymorphisms affecting transcription, thus resulting in reduced soluble HLA-G. For HLA-E, five polymorphisms can be distinguished at the nucleotide level: two (HLA-E*0102 and *0104) are very rare; the other three (HLA-E*0101, *01031, *01032) encode two different variants of the HLA-E protein depending upon a non-synonymous substitution occurring in exon 3 at codon 107 (AGG rather than GGG). In-vitro studies have not yet demonstrated any functional implication of this polymorphism. The aim of this study was to investigate whether the HLA-G and HLA-E polymorphisms and their interaction could play a role in the pathogenesis and prognosis of patients with RSA compared with fertile controls. Materials and methods Patients and controls All patients were Caucasians attending the outpatient clinic of the Department of Obstetrics and Gynecology, University of Bonn, Germany. Inclusion criteria for the study were as follows: three or more consecutive RSA from the same partner, the absence of uterine or genetic anomalies, normal results for lupus anticoagulans and cardiolipin antibodies, and informed consent of the patients. Of 100 couples investigated, 78 couples fulfilled the inclusion criteria and were included in this study. The mean age of the patients was 33.9 years (range 22–42) and the average number of RSA was 3.7 (range 3–8). Out of 78 patients, 22 (28%) were secondary RSA patients, i.e. they had already had a child with the same partner. After completion of diagnostic procedures, patients were encouraged to attempt another pregnancy; preconceptional folic acid substitution was recommended and luteal phase support was offered during early pregnancy if progesterone concentrations were <20 ng/ml. During the first trimester a supportive therapy by means of weekly ultrasound and human chorionic gonadotrophin (HCG) determination was offered. No immunological therapy was performed. Controls consisted of 52 Caucasians (mean parity 1.9; mean age 31.6 years), who had achieved at least one successful pregnancy and without any history of RSA or infertility. This study was approved by the local ethics comittee (no. 113/94) of the University of Bonn. Methods DNA isolation Genomic DNA extraction from whole blood was carried out using the salting out procedure, as described previously (Miller et al., 1988). HLA-G typing For HLA-G typing, the polymorphic exons 2 and 3 which determine the allele definition were amplified and sequenced as described previously (van der Ven et al., 1994). Briefly exons 2 and 3 were amplified with intronic primer pairs: for exon 2, sense primer G2i5 5′-GAG GGT CGG GCG GGT CTC AAC-3′ and antisense primer GCS 5i3 5′-GCA TGG AGG TGG GGG TCG TGA-3′, for exon 3, sense primer GCS 6i5 5′-GAC CCT CTA CCT GGG AGA ACC CCA-3′ and antisense primer GCS 4i3 5′-CCT CCA CTC CCT CAG AGA CTT CAT C-3′ were used. For both exons, the same polymerase chain reaction (PCR) programme was used: after 94°C for 5 min, 35 cycles of 94°C for 1 min and 68°C for 2 min were followed by 72°C for 10 min. PCR products underwent direct sequencing of exon 2 (sense primer G2i5) and exon 3 (antisense primer G3i3 5′-TCT GTG GAG CCA CTC CAC GCA CGT-3′) on an automated sequencer (A.L.F. Express DNA Sequencer; Amersham Pharmacia) according to a cycle sequencing protocol (Cycle sequencing Kit; Amersham Pharmacia). The cycling programme was 95°C for 3 min, followed by 25 cycles of 95°C for 30 s, 69°C for 36 s and 72°C for 1.24 min, ending with 72°C for 5 min. Allele assignment was based on combinations of polymorphisms at codons 31, 35, and 57 (exon 2), and codons 93, 107, 110, and 130 (exon 3) as described previously (van der Ven et al., 1998b). These polymorphisms give rise to 14 different alleles (some of which are listed in Table I) and four different proteins (HLA-G *0101, *0103, *0104 and *0105). HLA-E typing For determination of HLA-E alleles, the polymorphism at codon 107 (exon 3) was investigated, because this is a non-synonymous polymorphism described for HLA-E (AGG versus GGG encoding arginine or glycine respectively). PCR was performed by amplifying exon 3, as described previously (Geraghty et al., 1992). Briefly, the sense primer E3i5 5′-CGG GAC TGA CTA AGG GGC-3′ and the antisense primer E3i3 5′-AGC CCT GTG GAC CCT CTT-3′ were used in the following PCR protocol: 94°C for 5 min followed by 30 cycles of 94°C for 1 min, 65°C for 1.15 min, 72°C for 2.15 min ending in 72°C for 7 min. 8 μl of PCR products were screened for sequence variations by single strand conformational polymorphism (SSCP) analysis (6 W for 14 h on Blue Seq gel apparatus; Serva, Heidelberg, Germany) and consecutive silver staining. SSCP led to three clearly distinguishable patterns (Figure 1), which were confirmed by direct sequencing in 25 cases. Statistical analysis Statistical analysis was performed with SAS system software applying either χ2 or Fisher's exact test. In general, the χ2 test was used (Tables I, II and II). Fisher's exact test was applied if the criteria (i.e. number of expected counts per cell >5) for the χ2 test were not fulfilled (Table IV). Estimated haplotypes and maximum likelihood ratio were analysed in a log linear model as previously described (Korner et al., 1994). Since the analysis was descriptive, a final correction for multiple testing was not necessary. Results A total of 78 idiopathic RSA couples and 52 controls were genotyped for HLA-G and HLA-E. Detailed HLA-G and HLA-E allele frequencies are shown in Tables I and II respectively. An overall analysis of RSA women compared with fertile controls showed no significant difference between the two groups (P = 0.075 for HLA-G and P = 0.17 for HLA-E). Polymorphisms of HLA-G and HLA-E were in Hardy–Weinberg equilibrium in all groups. A more detailed analysis (Table I) showed that the HLA-G allele frequencies of patients who had suffered five or more RSA differed significantly from fertile controls (P = 0.001), and from patients with three or four RSA (P = 0.027). This difference was mainly due to an increase of HLA-G*01013 alleles in the RSA group. As HLA-G*0105N encodes a non-functional protein and there are data which show that the HLA-G *01013 and *0105N alleles are associated with lower amounts of the soluble isoform of HLA-G in peripheral blood (Rebmann et al., 1999, 2001), we investigated whether there is a difference between these groups with respect to the two alleles. The results of this analysis are shown in Table III: there is a significant increase in the proportion of the alleles *01013 and *0105N in the group of RSA women compared with the fertile controls (P = 0.007). These allele frequencies (which are rare in the general population; van der Ven et al., 1998b), increase further to 29% in the group of RSA women having had five or more RSA compared with RSA women having had only three or four RSA (11%) (P = 0.015) or compared with fertile controls (4%) (P = 0.001). Patient analysis on the genotype basis confirmed a significantly higher proportion of HLA-G *01013 and *01015N carriers in the group of RSA patients (17 out of 78, i.e. 22%) versus the fertile controls (four out of 52, or 8%) (P = 0.032) (data not shown). In the RSA group, there were two individuals homozygous for HLA-G*0105N and three homozygous for HLA-G*01013, while in the fertile control group, no HLA-G *0105N or *01013 homozygous individuals were observed. Primary RSA couples with five or more RSA revealed a higher incidence (although not statistically significant) of HLA-G *01013 and *01015N, compared with secondary RSA couples. For HLA-E polymorphisms, analysis of the patients with five or more RSA compared with fertile controls revealed an increase of the alleles encoding arginine (E*0101 and E*0102), but this was not statistically significant (Table II). In order to evaluate whether there is an interaction between the HLA-G and HLA-E alleles, a comparison of estimated haplotypes was carried out using likelihood ratio tests. This analysis revealed no indication of a linkage disequilibrium, nor any additional influence of HLA-E genotype on the prognosis or history of RSA compared with the fertile controls. HLA-G genotype and prospective pregnancy outcome In order to evaluate the prognostic value of HLA-G genotyping for the next pregnancy, we investigated the correlation of HLA-G genotype with pregnancy outcome in idiopathic RSA. Only the first pregnancy after diagnostic procedures of each patient in the study population was evaluated. After a positive pregnancy test, patients received weekly ultrasounds, folic acid supplementation and, if necessary, luteal phase support. No immunotherapy was performed. During the follow-up, 43 out of the 78 study patients became pregnant. One pregnancy was interrupted due to the development of a non-immunological hydrops fetalis in the 22nd week of gestation, and another patient decided to perform immunotherapy, so both patients were excluded from further analysis. Ten patients (24%) suffered from another miscarriage, 31 patients (76%) gave birth to a living child. Seven out of 10 couples (70%) suffering from another RSA carried either the HLA-G*01013 or *0105N alleles, compared with 10 out 31 (32%) couples giving birth to a child (P = 0.06) (Table IV). The maternal and paternal contribution of HLA-G*01013 and *0105N alleles did not cause any significant difference in prognostic pregnancy outcome or in historical number of RSA. Discussion This study demonstrates a significant increase in the number of HLA-G*01013 and HLA-G *0105N alleles in patients with RSA compared with fertile controls. Earlier studies showed that the alleles HLA-G*01013 and *0105N are associated with lower plasma soluble HLA-G concentrations (Rebmann et al., 1999, 2001), and lower soluble HLA-G concentrations have been demonstrated to be correlated with adverse pregnancy outcome after IVF (Pfeiffer et al., 2000). Consecutively, the results of the current study indicate the possibility, that low concentrations of soluble HLA-G could adversely affect the pregnancy outcome after RSA. Other studies (Hunt et al., 2000) have indicated that the soluble HLA-G2 isoform might be especially relevant in pregnancy. In HLA-G2 isoforms, the HLA-G*0105N allele does not lead to a distorted protein, because exon 3, which contains the frame shift mutation, has been spliced out. The HLA-G*0105N allele can therefore only be responsible for a decreased amount of HLA-G1 isoforms. On the basis of our own results, we favour the hypothesis that the HLA-G isoforms retaining exon 3 are relevant in pregnancy. This includes soluble and membranous isoforms, because HLA-G1 isoforms are the only transmembrane isoforms reaching the cell surface (Bainbridge et al., 2000). Nevertheless, quantitative data on membranous expression of HLA-G1 and a possible allelic association are not yet available. Additional factors could also modify the influence of HLA-G genotype on pregnancy. HLA-G molecules can inhibit NK cells and CD8 T-cells (Riteau et al., 1999) and HLA-G expression on trophoblast can be induced by interleukin-10 (Moreau et al., 1999), a cytokine belonging to the pregnancy protecting TH2 group. The interaction between genetic and biochemical factors could, therefore, determine pregnancy outcome. Another group (Karhukorpi et al., 1997) performed HLA-G typing by means of restriction fragment length polymorphism (RFLP) in 38 couples with RSA and could not find a difference compared with controls. However, RFLP detects only four different alleles, and it cannot detect HLA-G*0105N, nor can it differentiate between HLA-G *01013 and other HLA-G *0101X alleles. A Japanese group (Yamashita et al., 1999) performed SSCP and sequencing of HLA-G in 20 couples with RSA and found an allele frequency of 13% for HLA-G *01013 in RSA and 6% in controls, but this was not statistically significant in the overall analysis of allelic distribution. The impact of HLA-G *01013 and HLA-G *0105N on RSA is especially relevant in the group of women who had experienced five or more RSA, and who are known to have a worse prognosis than patients who had had only three or four RSA. Recent studies have shown that sporadic causes for RSA (i.e. abnormal karyotypes) decrease with the number of abortions, implying that other systemic (but still unknown) causes increase (Ogasawara et al., 2000). If the HLA-G `low-secretor' allele is such a factor involved in the pathogenesis of RSA, one would expect an increase in the number of carriers with the number of RSA, and this is in agreement with the results of this study. Our data suggest that both the maternal and paternal HLA-G genotypes may contribute to pregnancy outcome. The question of whether a live birth or an RSA occurs might, therefore, depend not only on maternal genotype but also on the fetal genotype. Further studies are in progress to analyse the contribution of the fetal HLA-G genotype. The results of HLA-E typing did not show any significant differences between RSA couples and fertile controls. This is in agreement with a previous study (Steffensen et al., 1998), which found no difference in allelic distribution of HLA-E between RSA and fertile controls. In Caucasian populations, both polymorphisms are equally distributed (Geraghty et al., 1992; Grimsley et al., 1997). In our population, patients with five or more RSA showed a slight increase in the alleles encoding arginine (AGG at codon 107; HLA-E*0101 and *0102); but this was not statistically significant in comparison with the controls. Idiopathic RSA is a multicausative disease. Due to its relatively good prognosis even without immunotherapy, genetic factors can only have a modifying, not a deterministic, impact. This hypothesis is supported by data on homozygous individuals with normal fertility (Ober et al., 1998b; Castro et al., 2000). On the other hand, several HLA-G *0105N and *01013 homozygous individuals have been described (van der Ven et al., 2000), most of whom suffered from pregnancy-associated disease. This study demonstrates that certain HLA-G alleles are associated with a higher risk of RSA. In this study, no immunotherapy was performed and the birth rate reached 76%. In our sample, the number of HLA-G *01013 and *0105N genotype carriers was higher in the group aborting again, compared with couples giving birth. It is therefore tempting to speculate that the subgroup of HLA-G*01013 and *0105N might profit from immunotherapy, e.g. paternal immunization (Ober et al., 1999) or i.v. immunoglobulin therapy (Coulam et al., 1995; Perino et al., 1997). Nevertheless, clinical application of HLA-G genotyping cannot be recommended before more data are available. Table I. Distribution of allele frequencies for HLA-G. Values in parentheses are percentages HLA-G allele  RSA women (n = 78) (all patients)  RSA women (n = 64) (3–4 abortions)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  No significant difference between RSA women (total) and fertile controls (degrees of freedom = 4, χ2 = 8.500).  No significant difference between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 4, χ2 = 6.665).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 4, χ2 = 18.407).  Significant difference (P = 0.027) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 4, χ2 = 10.960).   *01011,2,e,f,g  119 (76)  104 (81)  15 (54)  86 (83)  *01013  17 (11)  10 (8)  7 (25)  4 (4)  *01031  3 (1.9)  2 (2)  1 (4)  2 (1.9)  *01041/2  12 (8)  8 (6)  4 (14)  12 (11.5)   *0105N  5 (3.2)  4 (3)  1 (4)  0 (0)  Total no. of alleles  156 (100)  128 (100)  28 (100)  104 (100)  HLA-G allele  RSA women (n = 78) (all patients)  RSA women (n = 64) (3–4 abortions)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  No significant difference between RSA women (total) and fertile controls (degrees of freedom = 4, χ2 = 8.500).  No significant difference between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 4, χ2 = 6.665).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 4, χ2 = 18.407).  Significant difference (P = 0.027) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 4, χ2 = 10.960).   *01011,2,e,f,g  119 (76)  104 (81)  15 (54)  86 (83)  *01013  17 (11)  10 (8)  7 (25)  4 (4)  *01031  3 (1.9)  2 (2)  1 (4)  2 (1.9)  *01041/2  12 (8)  8 (6)  4 (14)  12 (11.5)   *0105N  5 (3.2)  4 (3)  1 (4)  0 (0)  Total no. of alleles  156 (100)  128 (100)  28 (100)  104 (100)  View Large Table II. HLA-E polymorphism at codon 107. Values in parentheses are percentages HLA-E allele  RSA women (all patients)  RSA male (n = 75) (all patients)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 50)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  A = AGG at codon 107, encoding arginine (E*0101,E*0102).  G = GGG at codon 107, encoding glycine (E*01031,E*01032,E*0104).  No significant difference between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 1.923).  No significant difference between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 2.049).  A  88 (58)  70 (47)  18 (64)  49 (49)  G  64 (42)  80 (53)  10 (36)  51 (51)  Total number of alleles  152 (100)  150 (100)  28 (100)  100 (100)  HLA-E allele  RSA women (all patients)  RSA male (n = 75) (all patients)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 50)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  A = AGG at codon 107, encoding arginine (E*0101,E*0102).  G = GGG at codon 107, encoding glycine (E*01031,E*01032,E*0104).  No significant difference between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 1.923).  No significant difference between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 2.049).  A  88 (58)  70 (47)  18 (64)  49 (49)  G  64 (42)  80 (53)  10 (36)  51 (51)  Total number of alleles  152 (100)  150 (100)  28 (100)  100 (100)  View Large Table III. HLA-G alleles and number of abortions in RSA women compared with fertile controls. Values in parentheses are percentages HLA-G allele  RSA women (n = 78) (all 78 patients)  RSA women ≥5 abortions (n = 14)  RSA women 3–4 abortions (n = 64)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  Significant difference (P = 0.007) between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 7.293).  Significant difference (P = 0.045) between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 1, χ2 = 4.032).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 16.319).  Significant difference (P = 0.015) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 1, χ2 = 5.897).  G*01013 or G*0105N  22 (14)  8 (29)  14 (11)  4 (4)  Others  134 (86)  20 (71)  114 (89)  100 (96)  Total number of alleles  156 (100)  28 (100)  128 (100)  104 (100)  HLA-G allele  RSA women (n = 78) (all 78 patients)  RSA women ≥5 abortions (n = 14)  RSA women 3–4 abortions (n = 64)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  Significant difference (P = 0.007) between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 7.293).  Significant difference (P = 0.045) between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 1, χ2 = 4.032).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 16.319).  Significant difference (P = 0.015) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 1, χ2 = 5.897).  G*01013 or G*0105N  22 (14)  8 (29)  14 (11)  4 (4)  Others  134 (86)  20 (71)  114 (89)  100 (96)  Total number of alleles  156 (100)  28 (100)  128 (100)  104 (100)  View Large Table IV. HLA-G and pregnancy outcome in RSA couples. Values in parentheses are percentages HLA-G genotype  Female RSA patients (n = 41)  Male RSA patients (n = 41)  RSA couples (n = 41)  Pregnancy outcome  Abortion  Live birth  Abortion  Live birth  Abortion  Live birth  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  P = 0.06 (Fisher's exact test).  Carrier of HLA-G  2  6  5  6  7 (70)  10 (32)  *01013 or *0105N              No carrier of HLA-G  8  25  5  25  3 (30)  21 (68)  *01013 or *01015N              Total  10  31  10  31  10 (100)  31 (100)  HLA-G genotype  Female RSA patients (n = 41)  Male RSA patients (n = 41)  RSA couples (n = 41)  Pregnancy outcome  Abortion  Live birth  Abortion  Live birth  Abortion  Live birth  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  P = 0.06 (Fisher's exact test).  Carrier of HLA-G  2  6  5  6  7 (70)  10 (32)  *01013 or *0105N              No carrier of HLA-G  8  25  5  25  3 (30)  21 (68)  *01013 or *01015N              Total  10  31  10  31  10 (100)  31 (100)  View Large Figure 1. View largeDownload slide Single strand conformational polymorphism (SSCP) pattern of human leukocyte antigen (HLA)-E polymorphism at codon 107. Lane 1 = heterozygous A/G; lane 2 = homozygous A/A; lane 3 = heterozygous A/G; lane 4 = homozygous G/G; lane 5 = homozygous G/G. Figure 1. View largeDownload slide Single strand conformational polymorphism (SSCP) pattern of human leukocyte antigen (HLA)-E polymorphism at codon 107. Lane 1 = heterozygous A/G; lane 2 = homozygous A/A; lane 3 = heterozygous A/G; lane 4 = homozygous G/G; lane 5 = homozygous G/G. 3 To whom correspondence should be addressed. E-mail: [email protected] This work was supported by DFG VE174/1/1–1/5 (K.van der Ven) and BONFOR (K.Pfeiffer). References Bainbridge, D.R.J., Ellis, S.A. and Sargent, I.L. ( 2000) The short forms of HLA-G are unlikely to play a role in pregnancy because they are not expressed at the cell surface. J. Reprod. Immunol. , 47, 1–16. Google Scholar Beer, A.E., Semprini, A.E., Xiaoyu, Z. and Quebbeman, J.F. ( 1985) Pregnancy outcome in human couples with recurrent spontaneous abortions: HLA antigen profiles; HLA antigen sharing; female serum MLR blocking factors; and paternal leukocyte immunization. Expl. Clin. Immunogenet. , 2, 137–153. Google Scholar Braud, V.M., Alla, D.S., O'Callaghan, C.A. et al. ( 1998) HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature , 391, 795–799. Google Scholar Castro, M.J., Morales, P., Rojo-Amigo, R. et al. ( 2000) Homozygous HLA-G*0105N healthy individuals indicate that membrane-anchored HLA-G1 molecule is not necessary for survival. Tiss. Antigens , 56, 232–239. Google Scholar Crisa, L., McMaster, M.T., Ishii, J.K. et al. ( 1997) Identification of a thymic epithelial cell subset sharing expression of the class Ib HLA-G molecule with fetal trophoblasts. J. Exp. Med. , 186, 289–298. Google Scholar Christiansen, O.B., Riisom, K., Lauritsen, J.G. and Grunnet, N. ( 1989) No increased histocompatibility antigen-sharing in couples with idiopathic, habitual abortion. Hum. Reprod. , 4, 160–162. Google Scholar Christiansen, O.B., Ring, M., Roosgard, A. et al. ( 1999) Association between HLA-DR1 and -DR3 antigens and unexplained repeated miscarriage. Hum. Reprod. Update , 5, 249–255. Google Scholar Coulam, C.B., Krysa, L., Stern, J.J. and Bustillo, M. ( 1995) Intravenous immunoglobulin for treatment of recurrent pregnancy loss. Am. J. Reprod. Immunol. , 34, 333–337. Google Scholar Fournel, S., Aguerre-Girr, M., Hue, X. et al. ( 2000) Cutting edge: soluble HLA-G1 triggers CD95/CD95 ligand-mediated apoptosis in activated CD8+ cells by interacting with CD8. J. Immunol. , 164, 6100–6104. Google Scholar 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. Google Scholar Geraghty, D.E., Stockschleader, M., Ishitani, A. and Hansen, J.A. ( 1992) Polymorphism at the HLA-E locus predates most HLA-A and -B polymorphism. Hum. Immunol. , 33, 174–184. Google Scholar Grimsley, C. and Ober, C. ( 1997) Population genetic studies of HLA-E: evidence for selection. Hum. Immunol. , 52, 33–40. Google Scholar Harger, J.H., Archer, D.F., Marchese, S.G. et al. ( 1983) Etiology of recurrent pregnancy loss and outcome of subsequent pregnancies. Obstet. Gynecol. , 62, 574–581. Google Scholar Horuzko, A., Antoniou, J., Tomlinson, P. et al. ( 1997) HLA-G functions as a restriction element and a transplantation antigen in mice. Int. Immunol. , 9, 645–653. Google Scholar Hunt, J.S., Jadhav, L., Chu, W. et al. ( 2000) Soluble HLA-G circulates in maternal blood during pregnancy. Am. J. Obstet. Gynecol. , 183, 682–688. Google Scholar Karhukorpi, J., Laitinen, T. and Tiilikainen, A.S. ( 1997) HLA-G polymorphism in Finnish couples with recurrent spontaneous miscarriage. Br. J. Obstet. Gynaecol. , 104, 1212–1214. Google Scholar King, A., Boocock, C., Sharkey, A.M. et al. ( 1996) Evidence for the expression of HLA-C class I mRNA and protein by human first trimester trophoblast. J. Immunol. , 156, 2068–2076. Google Scholar King, A., Hiby, S.E., Gardner, L. et al. ( 2000) Recognition of trophoblast HLA class I molecules by decidual NK cell receptors – a review. Placenta , 21, S81–S85. Google Scholar Korner, J., Rietschel, M., Hunt, N. et al. ( 1994) Association and haplotype analysis at the tyrosine hydroxylase locus in a combined German-British sample of manic depressive patients and controls. Psychiatr. Genet. , 4, 167–175. Google Scholar Kovats, S., Main, E.K., Librach, C. et al. ( 1990) A class I antigen, HLA-G expressed in human trophoblasts. Science , 248, 220–223. Google Scholar Le Bouteiller, P., Solier, C., Pröll, J. et al. ( 1999) Placental HLA-G expression in vivo: where and what for? Hum. Reprod. Update , 5, 223–233. Google Scholar Llano, M., Lee, N., Navarro, F. et al. ( 1998) HLA-E- bound peptides influence recognition by inhibitory and triggering CD94/NKG2 receptors: preferential response to an HLA-G-derived nonamer. Eur. J. Immunol. , 28, 2854–2863. Google Scholar Miller, S.A., Dykes, D.D. and Polesky, H.F. ( 1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. , 16, 1215–1218. Google Scholar Moreau, P., Adrian-Cabestre, F., Menier, C. et al. ( 1999) IL-10 selectively induces HLA-G expression in human trophoblasts and monocytes. Int. Immunol. , 11, 803–811. Google Scholar Ober, C., Hyslop, T., Elias, S. et al. ( 1998) HLA matching and fetal loss: results of a 10-year prospective study. Hum. Reprod. , 13, 33–38. Google Scholar Ober, C., Aldrich, c., Rosinsky, A. et al. ( 1998) HLA-G1 protein expression is not essential for fetal survival. Placenta , 19, 127–132. Google Scholar Ober, C., Karrison, T., Odem, R.R. et al. ( 1999) Mononuclear-cell immunisation in prevention of recurrent miscarriages: a randomised trial. Lancet , 354, 365–369. Google Scholar O'Callaghan, C.A. and Bell J.I. ( 1998) Structure and function of the human MHC class Ib molecules HLA-E, HLA-F, and HLA-G. Immunol. Rev. , 163, 129–138. Google Scholar Ogasawara, M., Aoki, K., Okada, S. and Suzumori, K. ( 2000) Embryonic karyotype of abortuses in relation to the number of previous miscarriages. Fertil Steril , 73, 300–304. Google Scholar 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. Google Scholar Pfeiffer, K.A., Rebmann, V., van der Ven, K. et al. ( 2000) Soluble histocompatibility antigen levels in early pregnancy after IVF. Hum. Immunol. , 61, 559–564. Google Scholar Rebmann, V., van der Ven, K., Pfeiffer, K.A. et al. ( 1999) Detection of soluble HLA-G molecules and their association to HLA-G Allotypes. Hum. Immunol. , 60, S29. Google Scholar Rebmann, V., van der Ven, K., Päβler, M. et al. ( 2001) Association of soluble HLA-G plasma levels with HLA-G alleles. Tiss. Antigens , 57, 15–21. Google Scholar Riteau, B., Menier, C., Khalil-Daher, I. et al. ( 1999) HLA-G inhibits the allogeneic proliferative response. J. Reprod. Immunol. , 43, 203–211. Google Scholar 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. Google Scholar Steffensen, R., Christiansen, O.B., Bennett, E.P. and Jersild, C. ( 1998) HLA-E polymorphism in patients with recurrent spontaneous abortion. Tiss. Antigens , 52, 569–572. Google Scholar Stray-Pedersen, B and Stray-Pedersen, S. ( 1984) Etiologic factors and subsequent reproductive performance in 195 couples with a prior history of habitual abortion. Am. J. Obstet. Gynecol. , 148, 140–146. Google Scholar van der Ven, K. and Ober, C. ( 1994) HLA-G polymorphisms in African–Americans. J. Immunol. , 153, 5628–5633. Google Scholar van der Ven, K., Skrablin, S., Ober, C. and Krebs, D. ( 1998) HLA-G polymorphisms – ethnic differences and implication for potential molecule function. Am. J. Reprod. Immunol. , 40, 145–157. Google Scholar van der Ven, K., Skrablin, S., Engels. G. and Krebs, D. ( 1998) HLA-G polymorphisms and allele frequencies in caucasians. Hum. Immunol. , 59, 302–312. Google Scholar van der Ven, K., Pfeiffer, K. and Skrablin, S. ( 2000) HLA-G polymorphisms and molecule function – questions and more questions – a review. Placenta , 21, S86–S92. Google Scholar Wei, X. and Orr, H.T. ( 1990) Differential expression of HLA-E, HLA-F, and HLA-G transcripts in human tissue. Hum. Immunol. , 29, 131–142. Google Scholar Yamashita, T., Fujii, T., Tokunaga, K. et al. ( 1999) Analysis of human leukocte antigen-G polymorphism including intron 4 in Japanese couples with habitual abortion. Am. J. Reprod. Immunol. , 41, 159–163. Google Scholar © European Society of Human Reproduction and Embryology http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Molecular Human Reproduction Oxford University Press

The HLA-G genotype is potentially associated with idiopathic recurrent spontaneous abortion

Download PDF
6 pages

Loading next page...
 
/lp/oxford-university-press/the-hla-g-genotype-is-potentially-associated-with-idiopathic-recurrent-AJFYgRp62m

References (35)

Publisher
Oxford University Press
Copyright
© European Society of Human Reproduction and Embryology
ISSN
1360-9947
eISSN
1460-2407
DOI
10.1093/molehr/7.4.373
Publisher site
See Article on Publisher Site

Abstract

Abstract The causes for recurrent spontaneous abortion (RSA) remain unknown in a large proportion of the cases. Human leukocyte antigen (HLA)-G and HLA-E are expressed on invasive trophoblast cells, and are supposed to confer to materno–fetal tolerance. A total of 14 different nucleotide sequences have been described for HLA-G, including one dysfunctional null allele (HLA-G*0105N), while five different sequences have been described for HLA-E. In this study, 78 RSA couples and 52 fertile controls were typed for HLA-G and HLA-E by direct sequencing or single strand conformational polymorphism (SSCP) respectively. The overall analysis showed no significant difference in allele frequencies for either HLA-G or HLA-E between the two groups. However, HLA-G allele frequencies in women who had suffered from five or more RSA differed significantly from fertile controls (P = 0.001), and from women who had undergone three or four RSA (P = 0.027). Detailed analysis demonstrated a significant increase in the proportion of the HLA-G alleles *01013 and *0105N in the whole group of RSA women compared with fertile controls (P = 0.007). When studying the prognostic value of HLA genotyping for pregnancy outcome (n = 41), 31 patients (76%) gave birth to a living child without performing immunotherapy. Seven out of 10 (70%) couples suffering from a further RSA carried the HLA-G*01013 or *0105N allele, compared with 10 out 31 (32%) couples giving birth (P = 0.06). This study suggests that the HLA-G genotype may be a contributing factor in RSA. genotyping, HLA-G, HLA-E, recurrent spontaneous abortion Introduction Although numerous studies have tried to determine the reasons for habitual abortion, definite causes for recurrent spontaneous abortion (RSA) remain unknown in a large proportion of the cases (Harger et al., 1983). Identified causes for RSA include uterine malformation, antiphospholipid antibodies, and cytogenetic anomalies, but these account for only 20–50% of the cases. Immunological causes and consecutive immunological therapies are proposed if no other reasons for the RSA are found. Nevertheless, in idiopathic cases, psychological support can achieve results comparable with those of immunotherapy (Stray-Pederson and Stray-Pederson, 1984). Human leukocyte antigens (HLA) as major determinants of allograft rejection have been studied intensively in the context of RSA. Conflicting data have been obtained on matters such as HLA sharing (Beer et al., 1985, Christiansen et al., 1989), HLA risk haplotypes (Ober et al., 1998a; Christiansen et al., 1999) and, more recently, on HLA-G (Karhukorpi et al., 1997) and HLA-E (Steffensen et al., 1998) allelic distribution. HLA-G is a non-classical HLA-Ib molecule which has gained a great deal of attention due to its selective and co-dominant expression on invasive trophoblast (Kovats et al., 1990) and in the thymus (Crisa et al., 1997). Some studies have demonstrated that invasive trophoblast lacks the polymorphic classical transplantation antigens, with the exception of HLA-C (King et al., 1996), but expresses the non-classical HLA molecules HLA-G and HLA-E (Wei and Orr, 1990; Le Bouteiller et al., 1999). Both HLA-G and HLA-E are less polymorphic than classical HLA-I molecules and their polymorphisms do not involve the peptide binding groove, as it is typical for classical HLA-I molecules in order to present different peptides (van der Ven et al., 1998a). HLA-E protein expression is stabilized by binding the leader peptide of classical HLA-I and HLA-G molecules (Llano et al., 1998). HLA-G binds a nonamer with leucine or isoleucine at position 2, proline at position 3, and leucin at position 9 (O'Callaghan and Bell, 1998). The lack of polymorphism in the peptide binding groove of HLA-E and HLA-G supports the hypothesis of a function that differs from classical HLA-I (which are extremely polymorphic), in order to present as many different peptides as possible. Nevertheless, HLA-G mismatches are able to induce an immune response in mice (Horuzko et al., 1997) and, therefore, the tolerance of a semiallogenic fetus during pregnancy remains a matter of discussion. The detailed function of HLA-E and HLA-G is still under investigation, but in-vitro studies have demonstrated that both interact with different inhibitory natural killer (NK) cell receptors (Braud et al., 1998), and that they may be able to protect cells which lack classical HLA-I molecules from NK cell attack (Rouas-Freiss et al., 1997). However, this question has not yet been finally clarified (King et al., 2000). For soluble HLA-G, induction of apoptosis in CD8+ T-cells has been demonstrated (Fournel et al., 2000). Hence, HLA-G and HLA-E are possible candidates for inhibiting the maternal anti-fetal immune response. HLA-G mRNA exists in at least six different alternative splicing forms, with two of them encoding for a soluble protein (soluble HLA-G) (Fujii et al., 1994). Patients with lower soluble HLA-G concentrations in the peripheral blood are at a greater risk of undergoing an RSA after IVF than patients with higher soluble HLA-G concentrations (Pfeiffer et al., 2000). An earlier study on non-pregnant controls has shown that the serum concentrations of soluble HLA-G are determined by the HLA-G genotype (Rebmann et al., 1999). For HLA-G, 14 different nucleotide sequences, encoding four different proteins (HLA-G *0101, *0103, *0104 and *0105) have been described. Among these, the HLA-G *0105N and *01013 alleles are associated with significantly lower serum concentrations of soluble HLA-G (Rebmann et al., 2001). HLA-G*0105N includes a one base deletion polymorphism in exon 3, which leads to a consecutive frame shift, and is therefore expected to encode a non-functional protein. The reason why the silent polymorphism, HLA-G*01013, is associated with lower serum HLA-G concentrations is still unknown, although it may be in linkage disequilibrium with polymorphisms affecting transcription, thus resulting in reduced soluble HLA-G. For HLA-E, five polymorphisms can be distinguished at the nucleotide level: two (HLA-E*0102 and *0104) are very rare; the other three (HLA-E*0101, *01031, *01032) encode two different variants of the HLA-E protein depending upon a non-synonymous substitution occurring in exon 3 at codon 107 (AGG rather than GGG). In-vitro studies have not yet demonstrated any functional implication of this polymorphism. The aim of this study was to investigate whether the HLA-G and HLA-E polymorphisms and their interaction could play a role in the pathogenesis and prognosis of patients with RSA compared with fertile controls. Materials and methods Patients and controls All patients were Caucasians attending the outpatient clinic of the Department of Obstetrics and Gynecology, University of Bonn, Germany. Inclusion criteria for the study were as follows: three or more consecutive RSA from the same partner, the absence of uterine or genetic anomalies, normal results for lupus anticoagulans and cardiolipin antibodies, and informed consent of the patients. Of 100 couples investigated, 78 couples fulfilled the inclusion criteria and were included in this study. The mean age of the patients was 33.9 years (range 22–42) and the average number of RSA was 3.7 (range 3–8). Out of 78 patients, 22 (28%) were secondary RSA patients, i.e. they had already had a child with the same partner. After completion of diagnostic procedures, patients were encouraged to attempt another pregnancy; preconceptional folic acid substitution was recommended and luteal phase support was offered during early pregnancy if progesterone concentrations were <20 ng/ml. During the first trimester a supportive therapy by means of weekly ultrasound and human chorionic gonadotrophin (HCG) determination was offered. No immunological therapy was performed. Controls consisted of 52 Caucasians (mean parity 1.9; mean age 31.6 years), who had achieved at least one successful pregnancy and without any history of RSA or infertility. This study was approved by the local ethics comittee (no. 113/94) of the University of Bonn. Methods DNA isolation Genomic DNA extraction from whole blood was carried out using the salting out procedure, as described previously (Miller et al., 1988). HLA-G typing For HLA-G typing, the polymorphic exons 2 and 3 which determine the allele definition were amplified and sequenced as described previously (van der Ven et al., 1994). Briefly exons 2 and 3 were amplified with intronic primer pairs: for exon 2, sense primer G2i5 5′-GAG GGT CGG GCG GGT CTC AAC-3′ and antisense primer GCS 5i3 5′-GCA TGG AGG TGG GGG TCG TGA-3′, for exon 3, sense primer GCS 6i5 5′-GAC CCT CTA CCT GGG AGA ACC CCA-3′ and antisense primer GCS 4i3 5′-CCT CCA CTC CCT CAG AGA CTT CAT C-3′ were used. For both exons, the same polymerase chain reaction (PCR) programme was used: after 94°C for 5 min, 35 cycles of 94°C for 1 min and 68°C for 2 min were followed by 72°C for 10 min. PCR products underwent direct sequencing of exon 2 (sense primer G2i5) and exon 3 (antisense primer G3i3 5′-TCT GTG GAG CCA CTC CAC GCA CGT-3′) on an automated sequencer (A.L.F. Express DNA Sequencer; Amersham Pharmacia) according to a cycle sequencing protocol (Cycle sequencing Kit; Amersham Pharmacia). The cycling programme was 95°C for 3 min, followed by 25 cycles of 95°C for 30 s, 69°C for 36 s and 72°C for 1.24 min, ending with 72°C for 5 min. Allele assignment was based on combinations of polymorphisms at codons 31, 35, and 57 (exon 2), and codons 93, 107, 110, and 130 (exon 3) as described previously (van der Ven et al., 1998b). These polymorphisms give rise to 14 different alleles (some of which are listed in Table I) and four different proteins (HLA-G *0101, *0103, *0104 and *0105). HLA-E typing For determination of HLA-E alleles, the polymorphism at codon 107 (exon 3) was investigated, because this is a non-synonymous polymorphism described for HLA-E (AGG versus GGG encoding arginine or glycine respectively). PCR was performed by amplifying exon 3, as described previously (Geraghty et al., 1992). Briefly, the sense primer E3i5 5′-CGG GAC TGA CTA AGG GGC-3′ and the antisense primer E3i3 5′-AGC CCT GTG GAC CCT CTT-3′ were used in the following PCR protocol: 94°C for 5 min followed by 30 cycles of 94°C for 1 min, 65°C for 1.15 min, 72°C for 2.15 min ending in 72°C for 7 min. 8 μl of PCR products were screened for sequence variations by single strand conformational polymorphism (SSCP) analysis (6 W for 14 h on Blue Seq gel apparatus; Serva, Heidelberg, Germany) and consecutive silver staining. SSCP led to three clearly distinguishable patterns (Figure 1), which were confirmed by direct sequencing in 25 cases. Statistical analysis Statistical analysis was performed with SAS system software applying either χ2 or Fisher's exact test. In general, the χ2 test was used (Tables I, II and II). Fisher's exact test was applied if the criteria (i.e. number of expected counts per cell >5) for the χ2 test were not fulfilled (Table IV). Estimated haplotypes and maximum likelihood ratio were analysed in a log linear model as previously described (Korner et al., 1994). Since the analysis was descriptive, a final correction for multiple testing was not necessary. Results A total of 78 idiopathic RSA couples and 52 controls were genotyped for HLA-G and HLA-E. Detailed HLA-G and HLA-E allele frequencies are shown in Tables I and II respectively. An overall analysis of RSA women compared with fertile controls showed no significant difference between the two groups (P = 0.075 for HLA-G and P = 0.17 for HLA-E). Polymorphisms of HLA-G and HLA-E were in Hardy–Weinberg equilibrium in all groups. A more detailed analysis (Table I) showed that the HLA-G allele frequencies of patients who had suffered five or more RSA differed significantly from fertile controls (P = 0.001), and from patients with three or four RSA (P = 0.027). This difference was mainly due to an increase of HLA-G*01013 alleles in the RSA group. As HLA-G*0105N encodes a non-functional protein and there are data which show that the HLA-G *01013 and *0105N alleles are associated with lower amounts of the soluble isoform of HLA-G in peripheral blood (Rebmann et al., 1999, 2001), we investigated whether there is a difference between these groups with respect to the two alleles. The results of this analysis are shown in Table III: there is a significant increase in the proportion of the alleles *01013 and *0105N in the group of RSA women compared with the fertile controls (P = 0.007). These allele frequencies (which are rare in the general population; van der Ven et al., 1998b), increase further to 29% in the group of RSA women having had five or more RSA compared with RSA women having had only three or four RSA (11%) (P = 0.015) or compared with fertile controls (4%) (P = 0.001). Patient analysis on the genotype basis confirmed a significantly higher proportion of HLA-G *01013 and *01015N carriers in the group of RSA patients (17 out of 78, i.e. 22%) versus the fertile controls (four out of 52, or 8%) (P = 0.032) (data not shown). In the RSA group, there were two individuals homozygous for HLA-G*0105N and three homozygous for HLA-G*01013, while in the fertile control group, no HLA-G *0105N or *01013 homozygous individuals were observed. Primary RSA couples with five or more RSA revealed a higher incidence (although not statistically significant) of HLA-G *01013 and *01015N, compared with secondary RSA couples. For HLA-E polymorphisms, analysis of the patients with five or more RSA compared with fertile controls revealed an increase of the alleles encoding arginine (E*0101 and E*0102), but this was not statistically significant (Table II). In order to evaluate whether there is an interaction between the HLA-G and HLA-E alleles, a comparison of estimated haplotypes was carried out using likelihood ratio tests. This analysis revealed no indication of a linkage disequilibrium, nor any additional influence of HLA-E genotype on the prognosis or history of RSA compared with the fertile controls. HLA-G genotype and prospective pregnancy outcome In order to evaluate the prognostic value of HLA-G genotyping for the next pregnancy, we investigated the correlation of HLA-G genotype with pregnancy outcome in idiopathic RSA. Only the first pregnancy after diagnostic procedures of each patient in the study population was evaluated. After a positive pregnancy test, patients received weekly ultrasounds, folic acid supplementation and, if necessary, luteal phase support. No immunotherapy was performed. During the follow-up, 43 out of the 78 study patients became pregnant. One pregnancy was interrupted due to the development of a non-immunological hydrops fetalis in the 22nd week of gestation, and another patient decided to perform immunotherapy, so both patients were excluded from further analysis. Ten patients (24%) suffered from another miscarriage, 31 patients (76%) gave birth to a living child. Seven out of 10 couples (70%) suffering from another RSA carried either the HLA-G*01013 or *0105N alleles, compared with 10 out 31 (32%) couples giving birth to a child (P = 0.06) (Table IV). The maternal and paternal contribution of HLA-G*01013 and *0105N alleles did not cause any significant difference in prognostic pregnancy outcome or in historical number of RSA. Discussion This study demonstrates a significant increase in the number of HLA-G*01013 and HLA-G *0105N alleles in patients with RSA compared with fertile controls. Earlier studies showed that the alleles HLA-G*01013 and *0105N are associated with lower plasma soluble HLA-G concentrations (Rebmann et al., 1999, 2001), and lower soluble HLA-G concentrations have been demonstrated to be correlated with adverse pregnancy outcome after IVF (Pfeiffer et al., 2000). Consecutively, the results of the current study indicate the possibility, that low concentrations of soluble HLA-G could adversely affect the pregnancy outcome after RSA. Other studies (Hunt et al., 2000) have indicated that the soluble HLA-G2 isoform might be especially relevant in pregnancy. In HLA-G2 isoforms, the HLA-G*0105N allele does not lead to a distorted protein, because exon 3, which contains the frame shift mutation, has been spliced out. The HLA-G*0105N allele can therefore only be responsible for a decreased amount of HLA-G1 isoforms. On the basis of our own results, we favour the hypothesis that the HLA-G isoforms retaining exon 3 are relevant in pregnancy. This includes soluble and membranous isoforms, because HLA-G1 isoforms are the only transmembrane isoforms reaching the cell surface (Bainbridge et al., 2000). Nevertheless, quantitative data on membranous expression of HLA-G1 and a possible allelic association are not yet available. Additional factors could also modify the influence of HLA-G genotype on pregnancy. HLA-G molecules can inhibit NK cells and CD8 T-cells (Riteau et al., 1999) and HLA-G expression on trophoblast can be induced by interleukin-10 (Moreau et al., 1999), a cytokine belonging to the pregnancy protecting TH2 group. The interaction between genetic and biochemical factors could, therefore, determine pregnancy outcome. Another group (Karhukorpi et al., 1997) performed HLA-G typing by means of restriction fragment length polymorphism (RFLP) in 38 couples with RSA and could not find a difference compared with controls. However, RFLP detects only four different alleles, and it cannot detect HLA-G*0105N, nor can it differentiate between HLA-G *01013 and other HLA-G *0101X alleles. A Japanese group (Yamashita et al., 1999) performed SSCP and sequencing of HLA-G in 20 couples with RSA and found an allele frequency of 13% for HLA-G *01013 in RSA and 6% in controls, but this was not statistically significant in the overall analysis of allelic distribution. The impact of HLA-G *01013 and HLA-G *0105N on RSA is especially relevant in the group of women who had experienced five or more RSA, and who are known to have a worse prognosis than patients who had had only three or four RSA. Recent studies have shown that sporadic causes for RSA (i.e. abnormal karyotypes) decrease with the number of abortions, implying that other systemic (but still unknown) causes increase (Ogasawara et al., 2000). If the HLA-G `low-secretor' allele is such a factor involved in the pathogenesis of RSA, one would expect an increase in the number of carriers with the number of RSA, and this is in agreement with the results of this study. Our data suggest that both the maternal and paternal HLA-G genotypes may contribute to pregnancy outcome. The question of whether a live birth or an RSA occurs might, therefore, depend not only on maternal genotype but also on the fetal genotype. Further studies are in progress to analyse the contribution of the fetal HLA-G genotype. The results of HLA-E typing did not show any significant differences between RSA couples and fertile controls. This is in agreement with a previous study (Steffensen et al., 1998), which found no difference in allelic distribution of HLA-E between RSA and fertile controls. In Caucasian populations, both polymorphisms are equally distributed (Geraghty et al., 1992; Grimsley et al., 1997). In our population, patients with five or more RSA showed a slight increase in the alleles encoding arginine (AGG at codon 107; HLA-E*0101 and *0102); but this was not statistically significant in comparison with the controls. Idiopathic RSA is a multicausative disease. Due to its relatively good prognosis even without immunotherapy, genetic factors can only have a modifying, not a deterministic, impact. This hypothesis is supported by data on homozygous individuals with normal fertility (Ober et al., 1998b; Castro et al., 2000). On the other hand, several HLA-G *0105N and *01013 homozygous individuals have been described (van der Ven et al., 2000), most of whom suffered from pregnancy-associated disease. This study demonstrates that certain HLA-G alleles are associated with a higher risk of RSA. In this study, no immunotherapy was performed and the birth rate reached 76%. In our sample, the number of HLA-G *01013 and *0105N genotype carriers was higher in the group aborting again, compared with couples giving birth. It is therefore tempting to speculate that the subgroup of HLA-G*01013 and *0105N might profit from immunotherapy, e.g. paternal immunization (Ober et al., 1999) or i.v. immunoglobulin therapy (Coulam et al., 1995; Perino et al., 1997). Nevertheless, clinical application of HLA-G genotyping cannot be recommended before more data are available. Table I. Distribution of allele frequencies for HLA-G. Values in parentheses are percentages HLA-G allele  RSA women (n = 78) (all patients)  RSA women (n = 64) (3–4 abortions)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  No significant difference between RSA women (total) and fertile controls (degrees of freedom = 4, χ2 = 8.500).  No significant difference between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 4, χ2 = 6.665).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 4, χ2 = 18.407).  Significant difference (P = 0.027) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 4, χ2 = 10.960).   *01011,2,e,f,g  119 (76)  104 (81)  15 (54)  86 (83)  *01013  17 (11)  10 (8)  7 (25)  4 (4)  *01031  3 (1.9)  2 (2)  1 (4)  2 (1.9)  *01041/2  12 (8)  8 (6)  4 (14)  12 (11.5)   *0105N  5 (3.2)  4 (3)  1 (4)  0 (0)  Total no. of alleles  156 (100)  128 (100)  28 (100)  104 (100)  HLA-G allele  RSA women (n = 78) (all patients)  RSA women (n = 64) (3–4 abortions)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  No significant difference between RSA women (total) and fertile controls (degrees of freedom = 4, χ2 = 8.500).  No significant difference between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 4, χ2 = 6.665).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 4, χ2 = 18.407).  Significant difference (P = 0.027) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 4, χ2 = 10.960).   *01011,2,e,f,g  119 (76)  104 (81)  15 (54)  86 (83)  *01013  17 (11)  10 (8)  7 (25)  4 (4)  *01031  3 (1.9)  2 (2)  1 (4)  2 (1.9)  *01041/2  12 (8)  8 (6)  4 (14)  12 (11.5)   *0105N  5 (3.2)  4 (3)  1 (4)  0 (0)  Total no. of alleles  156 (100)  128 (100)  28 (100)  104 (100)  View Large Table II. HLA-E polymorphism at codon 107. Values in parentheses are percentages HLA-E allele  RSA women (all patients)  RSA male (n = 75) (all patients)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 50)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  A = AGG at codon 107, encoding arginine (E*0101,E*0102).  G = GGG at codon 107, encoding glycine (E*01031,E*01032,E*0104).  No significant difference between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 1.923).  No significant difference between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 2.049).  A  88 (58)  70 (47)  18 (64)  49 (49)  G  64 (42)  80 (53)  10 (36)  51 (51)  Total number of alleles  152 (100)  150 (100)  28 (100)  100 (100)  HLA-E allele  RSA women (all patients)  RSA male (n = 75) (all patients)  RSA women (n = 14) (≥5 abortions)  Fertile controls (n = 50)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  A = AGG at codon 107, encoding arginine (E*0101,E*0102).  G = GGG at codon 107, encoding glycine (E*01031,E*01032,E*0104).  No significant difference between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 1.923).  No significant difference between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 2.049).  A  88 (58)  70 (47)  18 (64)  49 (49)  G  64 (42)  80 (53)  10 (36)  51 (51)  Total number of alleles  152 (100)  150 (100)  28 (100)  100 (100)  View Large Table III. HLA-G alleles and number of abortions in RSA women compared with fertile controls. Values in parentheses are percentages HLA-G allele  RSA women (n = 78) (all 78 patients)  RSA women ≥5 abortions (n = 14)  RSA women 3–4 abortions (n = 64)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  Significant difference (P = 0.007) between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 7.293).  Significant difference (P = 0.045) between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 1, χ2 = 4.032).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 16.319).  Significant difference (P = 0.015) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 1, χ2 = 5.897).  G*01013 or G*0105N  22 (14)  8 (29)  14 (11)  4 (4)  Others  134 (86)  20 (71)  114 (89)  100 (96)  Total number of alleles  156 (100)  28 (100)  128 (100)  104 (100)  HLA-G allele  RSA women (n = 78) (all 78 patients)  RSA women ≥5 abortions (n = 14)  RSA women 3–4 abortions (n = 64)  Fertile controls (n = 52)  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  Significant difference (P = 0.007) between RSA women (all patients) and fertile controls (degrees of freedom = 1, χ2 = 7.293).  Significant difference (P = 0.045) between RSA women (3–4 abortions) and fertile controls (degrees of freedom = 1, χ2 = 4.032).  Significant difference (P = 0.001) between RSA women (≥5 abortions) and fertile controls (degrees of freedom = 1, χ2 = 16.319).  Significant difference (P = 0.015) between RSA women (3–4 abortions) and RSA women (≥5 abortions) (degrees of freedom = 1, χ2 = 5.897).  G*01013 or G*0105N  22 (14)  8 (29)  14 (11)  4 (4)  Others  134 (86)  20 (71)  114 (89)  100 (96)  Total number of alleles  156 (100)  28 (100)  128 (100)  104 (100)  View Large Table IV. HLA-G and pregnancy outcome in RSA couples. Values in parentheses are percentages HLA-G genotype  Female RSA patients (n = 41)  Male RSA patients (n = 41)  RSA couples (n = 41)  Pregnancy outcome  Abortion  Live birth  Abortion  Live birth  Abortion  Live birth  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  P = 0.06 (Fisher's exact test).  Carrier of HLA-G  2  6  5  6  7 (70)  10 (32)  *01013 or *0105N              No carrier of HLA-G  8  25  5  25  3 (30)  21 (68)  *01013 or *01015N              Total  10  31  10  31  10 (100)  31 (100)  HLA-G genotype  Female RSA patients (n = 41)  Male RSA patients (n = 41)  RSA couples (n = 41)  Pregnancy outcome  Abortion  Live birth  Abortion  Live birth  Abortion  Live birth  HLA-G = human leukocyte antigen-G; RSA = recurrent spontaneous abortion.  P = 0.06 (Fisher's exact test).  Carrier of HLA-G  2  6  5  6  7 (70)  10 (32)  *01013 or *0105N              No carrier of HLA-G  8  25  5  25  3 (30)  21 (68)  *01013 or *01015N              Total  10  31  10  31  10 (100)  31 (100)  View Large Figure 1. View largeDownload slide Single strand conformational polymorphism (SSCP) pattern of human leukocyte antigen (HLA)-E polymorphism at codon 107. Lane 1 = heterozygous A/G; lane 2 = homozygous A/A; lane 3 = heterozygous A/G; lane 4 = homozygous G/G; lane 5 = homozygous G/G. Figure 1. View largeDownload slide Single strand conformational polymorphism (SSCP) pattern of human leukocyte antigen (HLA)-E polymorphism at codon 107. Lane 1 = heterozygous A/G; lane 2 = homozygous A/A; lane 3 = heterozygous A/G; lane 4 = homozygous G/G; lane 5 = homozygous G/G. 3 To whom correspondence should be addressed. E-mail: [email protected] This work was supported by DFG VE174/1/1–1/5 (K.van der Ven) and BONFOR (K.Pfeiffer). References Bainbridge, D.R.J., Ellis, S.A. and Sargent, I.L. ( 2000) The short forms of HLA-G are unlikely to play a role in pregnancy because they are not expressed at the cell surface. J. Reprod. Immunol. , 47, 1–16. Google Scholar Beer, A.E., Semprini, A.E., Xiaoyu, Z. and Quebbeman, J.F. ( 1985) Pregnancy outcome in human couples with recurrent spontaneous abortions: HLA antigen profiles; HLA antigen sharing; female serum MLR blocking factors; and paternal leukocyte immunization. Expl. Clin. Immunogenet. , 2, 137–153. Google Scholar Braud, V.M., Alla, D.S., O'Callaghan, C.A. et al. ( 1998) HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature , 391, 795–799. Google Scholar Castro, M.J., Morales, P., Rojo-Amigo, R. et al. ( 2000) Homozygous HLA-G*0105N healthy individuals indicate that membrane-anchored HLA-G1 molecule is not necessary for survival. Tiss. Antigens , 56, 232–239. Google Scholar Crisa, L., McMaster, M.T., Ishii, J.K. et al. ( 1997) Identification of a thymic epithelial cell subset sharing expression of the class Ib HLA-G molecule with fetal trophoblasts. J. Exp. Med. , 186, 289–298. Google Scholar Christiansen, O.B., Riisom, K., Lauritsen, J.G. and Grunnet, N. ( 1989) No increased histocompatibility antigen-sharing in couples with idiopathic, habitual abortion. Hum. Reprod. , 4, 160–162. Google Scholar Christiansen, O.B., Ring, M., Roosgard, A. et al. ( 1999) Association between HLA-DR1 and -DR3 antigens and unexplained repeated miscarriage. Hum. Reprod. Update , 5, 249–255. Google Scholar Coulam, C.B., Krysa, L., Stern, J.J. and Bustillo, M. ( 1995) Intravenous immunoglobulin for treatment of recurrent pregnancy loss. Am. J. Reprod. Immunol. , 34, 333–337. Google Scholar Fournel, S., Aguerre-Girr, M., Hue, X. et al. ( 2000) Cutting edge: soluble HLA-G1 triggers CD95/CD95 ligand-mediated apoptosis in activated CD8+ cells by interacting with CD8. J. Immunol. , 164, 6100–6104. Google Scholar 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. Google Scholar Geraghty, D.E., Stockschleader, M., Ishitani, A. and Hansen, J.A. ( 1992) Polymorphism at the HLA-E locus predates most HLA-A and -B polymorphism. Hum. Immunol. , 33, 174–184. Google Scholar Grimsley, C. and Ober, C. ( 1997) Population genetic studies of HLA-E: evidence for selection. Hum. Immunol. , 52, 33–40. Google Scholar Harger, J.H., Archer, D.F., Marchese, S.G. et al. ( 1983) Etiology of recurrent pregnancy loss and outcome of subsequent pregnancies. Obstet. Gynecol. , 62, 574–581. Google Scholar Horuzko, A., Antoniou, J., Tomlinson, P. et al. ( 1997) HLA-G functions as a restriction element and a transplantation antigen in mice. Int. Immunol. , 9, 645–653. Google Scholar Hunt, J.S., Jadhav, L., Chu, W. et al. ( 2000) Soluble HLA-G circulates in maternal blood during pregnancy. Am. J. Obstet. Gynecol. , 183, 682–688. Google Scholar Karhukorpi, J., Laitinen, T. and Tiilikainen, A.S. ( 1997) HLA-G polymorphism in Finnish couples with recurrent spontaneous miscarriage. Br. J. Obstet. Gynaecol. , 104, 1212–1214. Google Scholar King, A., Boocock, C., Sharkey, A.M. et al. ( 1996) Evidence for the expression of HLA-C class I mRNA and protein by human first trimester trophoblast. J. Immunol. , 156, 2068–2076. Google Scholar King, A., Hiby, S.E., Gardner, L. et al. ( 2000) Recognition of trophoblast HLA class I molecules by decidual NK cell receptors – a review. Placenta , 21, S81–S85. Google Scholar Korner, J., Rietschel, M., Hunt, N. et al. ( 1994) Association and haplotype analysis at the tyrosine hydroxylase locus in a combined German-British sample of manic depressive patients and controls. Psychiatr. Genet. , 4, 167–175. Google Scholar Kovats, S., Main, E.K., Librach, C. et al. ( 1990) A class I antigen, HLA-G expressed in human trophoblasts. Science , 248, 220–223. Google Scholar Le Bouteiller, P., Solier, C., Pröll, J. et al. ( 1999) Placental HLA-G expression in vivo: where and what for? Hum. Reprod. Update , 5, 223–233. Google Scholar Llano, M., Lee, N., Navarro, F. et al. ( 1998) HLA-E- bound peptides influence recognition by inhibitory and triggering CD94/NKG2 receptors: preferential response to an HLA-G-derived nonamer. Eur. J. Immunol. , 28, 2854–2863. Google Scholar Miller, S.A., Dykes, D.D. and Polesky, H.F. ( 1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res. , 16, 1215–1218. Google Scholar Moreau, P., Adrian-Cabestre, F., Menier, C. et al. ( 1999) IL-10 selectively induces HLA-G expression in human trophoblasts and monocytes. Int. Immunol. , 11, 803–811. Google Scholar Ober, C., Hyslop, T., Elias, S. et al. ( 1998) HLA matching and fetal loss: results of a 10-year prospective study. Hum. Reprod. , 13, 33–38. Google Scholar Ober, C., Aldrich, c., Rosinsky, A. et al. ( 1998) HLA-G1 protein expression is not essential for fetal survival. Placenta , 19, 127–132. Google Scholar Ober, C., Karrison, T., Odem, R.R. et al. ( 1999) Mononuclear-cell immunisation in prevention of recurrent miscarriages: a randomised trial. Lancet , 354, 365–369. Google Scholar O'Callaghan, C.A. and Bell J.I. ( 1998) Structure and function of the human MHC class Ib molecules HLA-E, HLA-F, and HLA-G. Immunol. Rev. , 163, 129–138. Google Scholar Ogasawara, M., Aoki, K., Okada, S. and Suzumori, K. ( 2000) Embryonic karyotype of abortuses in relation to the number of previous miscarriages. Fertil Steril , 73, 300–304. Google Scholar 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. Google Scholar Pfeiffer, K.A., Rebmann, V., van der Ven, K. et al. ( 2000) Soluble histocompatibility antigen levels in early pregnancy after IVF. Hum. Immunol. , 61, 559–564. Google Scholar Rebmann, V., van der Ven, K., Pfeiffer, K.A. et al. ( 1999) Detection of soluble HLA-G molecules and their association to HLA-G Allotypes. Hum. Immunol. , 60, S29. Google Scholar Rebmann, V., van der Ven, K., Päβler, M. et al. ( 2001) Association of soluble HLA-G plasma levels with HLA-G alleles. Tiss. Antigens , 57, 15–21. Google Scholar Riteau, B., Menier, C., Khalil-Daher, I. et al. ( 1999) HLA-G inhibits the allogeneic proliferative response. J. Reprod. Immunol. , 43, 203–211. Google Scholar 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. Google Scholar Steffensen, R., Christiansen, O.B., Bennett, E.P. and Jersild, C. ( 1998) HLA-E polymorphism in patients with recurrent spontaneous abortion. Tiss. Antigens , 52, 569–572. Google Scholar Stray-Pedersen, B and Stray-Pedersen, S. ( 1984) Etiologic factors and subsequent reproductive performance in 195 couples with a prior history of habitual abortion. Am. J. Obstet. Gynecol. , 148, 140–146. Google Scholar van der Ven, K. and Ober, C. ( 1994) HLA-G polymorphisms in African–Americans. J. Immunol. , 153, 5628–5633. Google Scholar van der Ven, K., Skrablin, S., Ober, C. and Krebs, D. ( 1998) HLA-G polymorphisms – ethnic differences and implication for potential molecule function. Am. J. Reprod. Immunol. , 40, 145–157. Google Scholar van der Ven, K., Skrablin, S., Engels. G. and Krebs, D. ( 1998) HLA-G polymorphisms and allele frequencies in caucasians. Hum. Immunol. , 59, 302–312. Google Scholar van der Ven, K., Pfeiffer, K. and Skrablin, S. ( 2000) HLA-G polymorphisms and molecule function – questions and more questions – a review. Placenta , 21, S86–S92. Google Scholar Wei, X. and Orr, H.T. ( 1990) Differential expression of HLA-E, HLA-F, and HLA-G transcripts in human tissue. Hum. Immunol. , 29, 131–142. Google Scholar Yamashita, T., Fujii, T., Tokunaga, K. et al. ( 1999) Analysis of human leukocte antigen-G polymorphism including intron 4 in Japanese couples with habitual abortion. Am. J. Reprod. Immunol. , 41, 159–163. Google Scholar © European Society of Human Reproduction and Embryology

Journal

Molecular Human Reproduction – Oxford University Press

Published: Apr 1, 2001

There are no references for this article.