AU2636399A

AU2636399A - Hla linked pre-eclampsia and miscarriage susceptibility gene

Info

Publication number: AU2636399A
Application number: AU26363/99A
Authority: AU
Inventors: John Bermingham; David M. Jenkins; Tommie V. Mccarthy; Margaret O'brien; Kathleen A. Quane
Original assignee: University College Cork
Current assignee: University College Cork
Priority date: 1998-02-25
Filing date: 1999-02-25
Publication date: 1999-09-15
Also published as: AU2003231618A1; NO20004222D0; EP1056886A1; NO20004222L; CA2321223A1; WO1999043851A1; JP2003517267A

Description

WO 99/43851 1 PCT/IE99/00012 HLA linked pre-eclampsia and miscarriage susceptibility gene The present invention relates to a susceptibility gene for pre-eclampsia and eclampsia, and the use of such a gene in methods for diagnosing susceptibility to these diseases. The invention also relates to a test kit for diagnosis of susceptibility and to pharmaceutical compositions for the prevention or treatment of the 5 diseases. The invention can also be used in the diagnosis of susceptibility to miscarriage and/or miscarriage-related infertility and/or intrauterine growth retardation. Specifically, the present invention relates to methods and materials used to detect a HLA linked human pre-eclampsia and miscarriage predisposing gene (HLA-G), some alleles of which, or linked alleles of linked genes of which, cause susceptibility to pre-eclampsia and miscarriage. More specifically, the 10 invention relates to sequence variation in the HLA-G gene and linked genes and their use in the diagnosis of susceptibility to pre-eclampsia and miscarriage. The invention further relates to sequence variations in the HLA-G gene and their use in diagnosis and prognosis of pre-eclampsia and miscarriage. Additionally, the invention relates to the therapy for pre-eclampsia and miscarriage and for susceptibility to pre eclampsia and miscarriage including protein therapy, gene therapy and protein mimetics. The invention 15 also relates to screening for drugs for pre-eclampsia and miscarriage therapy and for susceptibility to pre eclampsia and miscarriage therapy. Finally, the invention relates to the screening of the HLA-G gene and linked genes for sequence variations which are useful for diagnosing susceptibility to pre-eclampsia and miscarriage. Pre-eclampsia is the major cause of foetal and maternal morbidity and mortality with probable long term 20 adverse effects on health due to the prolonged associated intrauterine hypoxia. Pre-eclampsia occurs in approximately five to ten percent of all population births and is uniquely a disease of pregnancy. Acute pathological changes begin to resolve soon after delivery. The pathologic mechanisms causing pre eclampsia are unclear and no marker predictive for the disease prior to clinical evidence of the disease has been identified. Furthermore an association has been observed between miscarriage and pre-eclampsia 25 (Cooper et al.. 1988). Epidemiological studies show the disease to be highly heritable, mainly confined to first pregnancies and largely prevented by normal first pregnancy by the same partner. Patients affected in first pregnancies have a 13.1% recurrence risk for their second, whereas with a normal first pregnancy, the incidence in the second is of the order of 1%. Thus, the first pregnancy appears to have a significant protective effect 30 against pre-eclampsia in a subsequent pregnancy. Therefore, it follows that pre-eclampsia is preventable in principle (Lie et al., 1998). SUBSTITUTE SHEET (RULE 26) WO 99/43851 PCT/IE99/00012 2 Several classification schemes have been proposed to aid clinical recognition of pre-eclampsia. The classification advocated by the US National Institutes of Health working group on hypertension in pregnancy, is a rise in blood pressure of >15mm Hg diastolic or >30mm Hg systolic from measurement in early pregnancy, or to >140/90 mm Hg in late pregnancy if no early reading is available; plus proteinuria 5 (>0.3g per 24 h) and/or odema. However, in practice, proteinuria measurements may not always be determined and symptoms additional to a rise in blood pressure such as headache, visual disturbance and/or epigastric pain indicate a deterioration in pregnancy consistent with pre-eclampsia and form a basis for clinical intervention of early delivery by caesarean section to resolve the condition. Spinillo et al. (1994) reported that women with pre-eclampsia had a significantly increased incidence of intrauterine 10 growth retardation (IUGR) small for gestational age (SGA) infants. Although the cause of pre-eclampsia is unknown, hypertension is observed in pre-eclampsia and has been the focus of a large amount of research on the disorder. However, the pathological and physiological changes of pre-eclampsia show that this syndrome is much more than pregnancy-induced hypertension. Evidence to date implicates the action of placental trophoblasts as the underlying cause. 15 In pre-eclampsia, cytotrophoblast invasion is shallow and spiral arteriol invasion is abnormal, resulting in reduced blood perfusion of the intervillous space. Moreover the characteristic pattern of integrin switching that takes place during normal trophoblast differentiation does not occur in pre-eclampsia. The outermost layer (trophoblasts) of the human placenta is devoid of classical class I human leukocyte antigens (HLA-A and HLA-B) and class II proteins (HLA-DR, HLA-DQ and HLA-DP). Although this 20 prevents recognition by maternal T lymphocytes, the lack of class I molecules leaves these cells susceptible to attack by natural killer (NK) cells. However, trophoblast cells directly in contact with maternal tissues selectively express a characteristic nonclassical class Ib molecule, HLA-G. HLA-E and limited HLA-C expression also occurs . Expression of HLA-G has been shown to be sufficient to protect otherwise susceptible target cells from NK cell mediated lysis. NK cells usually express several different 25 inhibitory receptors of various specificities at the same time. Cross linking of any single inhibitory receptor is sufficient to inactivate NK cell activity against all possible targets. It has been shown that membrane bound HLA-G molecules were able to inhibit alloreactive NK cells with NK inhibitory receptor 1 and inhibitory receptor 2 (NKI and NK2). It has been shown that CD94 / NKG2 is the predominant inhibitory receptor involved in recognition of HLA-G by decidual and peripheral NK cells. Thus, at a 30 functional level, HLA-G is able to protect target cells from destruction by NKl-,NK2-and NKG2 specific effector cells (Loke and King, 1997). More recently, HLA-G has been shown to modulate the ability of blood mononuclear cells to release cvtokines (Maejima et al. 1997) suggesting a role for HLA-G in triggering matemrnal-foetal immune interplay. Specifically, coculturing of HLA-G expressing cells with WO 99/43851 PCT/IE99/00012 3 peripheral blood mononuclear cells (PBMC) increased the amount of interleukin-3 (IL-3) and interleukin 1 beta (IL-I beta) and decreased the amount of tumour necrosis factor-alpha (TNF-alpha) release from the PBMC cells. HLA-G binds a diverse but limited array of peptides in a manner similar to that found for classical class I 5 molecules and it has been reported that HLA-G is expressed in the human thymus raising the possibility that maternal unresponsiveness to HLA-G expressing foetal tissues may be shaped in the thvmus by central presentation of this MHC molecule on the medullary epithelium (Crisa et al. 1997) HLA-G is known to be capable of stimulating a HLA-G restricted cytotoxic T lymphocyte response and HLA-G molecules can serve as target molecules in lytic reaction with cytotoxic T lymphocytes and HLA-G 10 expressed internally in vivo in transgenic animals is involved in education of the lymphocytic repertoire (Schmidt et al., 1997). Major histocompatibility (MHC) molecules bind a diverse array of peptides for presentation to T cells as part of a mechanism for recognition of self and non-self cells and pathologically altered cells. A detailed analysis of peptides bound to the soluble and membrane HLA-G proteins shows that, like MHC class 1 15 molecules, HLA-G also binds a diverse, although less complex array of peptides (Lee et al., 1995). Some of these peptides, which are derived from intracellular proteins, constitute minor histocompatibility antigens which in conjunction with MHC molecules provoke an immune reaction by blood mononuclear cells such as T cells. HLA bound peptides can readily be fractionated, fully or partially purified and sequenced and can be assayed for their capacity to promote an immune reaction by measurement of their 20 capacity to reconstitute lysis of target cells by cytotoxic T cells (den Haan et al., 1998). The entire gene sequence of HLA-G is known and DNA sequence analysis of HLA-G has shown that the HLA-G gene exhibits limited polymorphism, van der Van & Ober, 1995 examined the first six exons of HLA-G in 45 healthy African-Americans and observed variations in exons 2 and 3, which correspond to the alpha 1 and alpha 2 domains of the peptide binding grove. The most common polymorphism observed 25 was a C to T transition at position 1488, corresponding to codon 93. Another common polymorphism was identified by Harrison et al, 1993 and is a 14 bp deletion in exon 8 of the gene. These results indicate that HLA-G is a polymorphic gene potentially capable of presenting a wide variety of peptides. Patterns of variability in HLA-G are similar to those of other class I MHC genes, where amino acid substitutions are clustered in the alpha 1 and alpha 2 domains. 30 Three observations of altered expression of HLA-G in pre-eclampsia have been reported. Colbern et al.. 1994 showed that the level of HLA-G in placental tissue was reduced in pre-eclampsia and that the decreased expression appeared to be related to a reduced number of trophoblasts in pre-eclamptic placental tissue. Hara et al., 1996, showed that clusters of extravillous trophoblasts were devoid of HLA- WO 99/43851 PCT/IE99/00012 4 G in pre-eclamptic patients. Examination of human preimpantation blastocysts showed that only 40% of the blastocvsts expressed HLA-G (Jurisicova, et al. 1996). Inheritance Several bodies of evidence show that pre-eclampsia and eclampsia are largely under genetic control. 5 However the genetic mechanisms underlying susceptibility to pre-eclampsia remain unclear. This is largely due to confounding factors peculiar to its inheritance. First, the condition is specific to pregnancy and genetic studies to date have not been able to clarify whether the genes responsible are acting through the maternal or foetal genotype or through some interaction between the two. Secondly, pre-eclampsia is largely confined to primagravidas with a much lower incidence in subsequent pregnancies and thirdly, as 10 the condition is specific to pregnancy, the genetic contribution of males is difficult to assess. Diagnosis of true pre-eclampsia can be complicated by other hypertensive disorders such as essential hypertension and hypertension arising from renal disease. Such hypertensive disorders are distinct from true pre-eclampsia but nonetheless can confound diagnosis and thus pose problems for genetic studies. The classification of pre-eclampsia by some investigators as a disease of immune dysfunction has 15 prompted a number of studies on the role of the major histocompatibility complex in the genetics of pre eclampsia. There are numerous published studies on HLA associations with pre-eclampsia (Cooper et al., 1993). Besides the fact that the positive associations are, with one exemption, not reproduced in studies by others, these studies suffer from other difficulties. The number of individuals are generally small in 20 comparison to the large number of antigens at each of the HLA loci. There is a tendency for only significant associations to be reported and so there may be completed studies showing no association that have not been published besides those reported here. The first four associations reported are with antigen sharing or homozygosity. The number of antigens recognised has vastly increased with time. Antigens have been split as new sera become available, and the use of DNA techniques has split these further so 25 that there are over 100 HLA-A and 100 HLA-B alleles and over 25 HLA-DRB alleles (including five different, common sequences recognised as DR4 serologically). Thus what were typed as the same allele in homozygotes or shared antigens in the early studies cannot be relied on to be homogenous in sequence or function. Detecting homozygotes with sera in early studies suffers from the extra difficulty of distinguishing them from heterozygotes for another allele for which sera did not exist (blanks). 30 At least three studies have further investigated the association between pre-eclampsia and HLA-DR by linkage analysis (Winton et al.. 1990; Hayward et al., 1992, Harrison et al.. 1997). In these definitive studies no evidence was found for linkage of the HLA region to pre-eclampsia. Hayward et al. (1992) WO 99/43851 PCT/IE99/00012 5 also investigated several candidate genes and random DNA markers. Overall, no evidence was found for linkage to several candidate genes implicated in the pathogenesis of hypertension and their results excluded linkage to several markers. In these studies, an autosomal recessive model was assumed. Winton et al. (1990) also analysed their data for a HLA linkage using the affected sib pair method and the 5 affected pedigree-member method. Both of these methods make no assumption about the mode of inheritance and neither gave any indication of linkage. The majority of pre-eclampsia cases are considered sporadic. A familial pregnancy-induced hypertensive disorder has been described and two loci have been implicated in the familial form of the disorder, namely, a candidate region on chromosome 4 and the eNOS gene region on chromosome 7 (Harrison et al., 1997, Arngrimsson et al., 1997). The epidemiology 10 of PET is consistent with familial pregnancy-induced hypertensive disorder and sporadic PET being distinct entities. Humphrey et al., 1995, investigated the HLA-G deletion polymorphism for association with pre eclampsia. Specifically, pre-eclamptic patients, offspring of pre-eclamptic mothers, blood relatives of pre eclamptic patients, husbands of pre-eclamptic patients and a normal control group were genotyped for the 15 polymorphism. The was no detectable association between pre-eclampsia in mothers or in offspring of pre-eclamptic mothers and the HLA-G deletion polymorphisms. Karhukorpi et al., 1997 investigated HLA-G polymorphisms for association with recurrent spontaneous miscarriage. Specifically, they showed that there was no association between several HLA-G restriction fragment length polymorphisms and recurrent spontaneous miscarriage. 20 In the largest study of monozygotic twins, pre-eclampsia was reported in five first pregnancies, and all affected mothers were discordant with their twin. A second well documented report on an identical set of twins also showed clear discordance for pre-eclampsia in their first pregnancies. These observations argue against a recessive model and further support a role for the foetal paternal genotype in the disorder. Furthermore, although the subject of some controversy, pre-eclampsia occurs in mothers with mono-and 25 di-zygotic twins arguing against a recessive foetal genotype and in favour of a dominant paternal gene in the foetus. Some studies have considered the possibility of changing paternity as a contributing factor in the occurrence of pre-eclampsia in multiparae. Most notably, a strong association between pre-eclampsia and changing paternity has been observed (Lie et al., 1998). 30 Much of the work on pre-eclampsia has been based on the hypothesis of a major susceptibility locus in the affected mother and almost all of the genetic studies to date have focused on linkage or association between the genotype of the mother and pre-eclampsia. In order to test the hypothesis that foetal HLA-G WO 99/43851 PCT/IE99/00012 6 is the most likely candidate gene for the disorder, we have investigated HLA-G genotypes in pre eclamptic and control trios and have shown that HLA-G is linked to both normal and pre-eclampsia pregnancy outcome and associated with recurrent spontaneous abortion. We have also investigated HLA G genotypes in second pregnancies of control and pre-eclamptic trios and have shown that the presence of 5 specific HLA-G alleles in the foetus in first pregnancy permits the occurrence of different HLA-G alleles in second pregnancy showing that HLA-G can induce tolerance to antigens in the first pregnancy and/or can modify the maternal immune system to accept foetuses in the second pregnancy in the absence of pregnancy related disorders that are selected against and/or cause pregnancy related disorders in first pregnancy. 10 Early pregnancy loss is the most common complication of human gestation of women attempting pregnancy. The majority of these losses are clinically unrecognised. Using a highly sensitive assay, the total incidence of miscarriage was estimated to be 31%, including 22% of losses which occurred at the very early stages of pregnancy i.e. before the pregnancy was clinically recognised (Wilcox et al., 1988). Recurrent spontaneous abortion (RSA) or recurrent miscarriage, defined as the loss of three or more 15 spontaneous pregnancies before 20 weeks gestation, occurs in less than 1% of pregnant women. Studies suggest that the chance of a successful pregnancy in an untreated woman who has experienced two or more first trimester miscarriages and no live births is approximately 30% to 50%. It is generally accepted that RSA is a condition with many different causes, however, about 50% of all RSA cases are not explained by structural genetic, endocrine, infectious or anatomic factors. Within the past few years there 20 has been a growing recognition that recurrent pregnancy loss may have autoimmune (immunity against self) and alloimmune (immunity against another person) causes, even in women with no clinically diagnosed autoimmune diseases. This has lead to investigation of the role of the HLA system and RSA. in particular, much emphasis has been placed on the degree of sharing of HLA alleles and haplotypes between RSA couples. It has been suggested that fetuses whose HLA alleles do not differ from maternal 25 alleles (i.e. histocompatible fetuses) are more likely to be aborted than fetuses with HLA alleles that differ from maternal alleles (i.e. histoincompatible fetuses). It would follow then that couples who match for HLA alleles or haplotypes would produce histocompatible fetuses and hence be at risk of miscarriage. Ober et al. (1998) conducted a 10 year prospective study of HLA matching and pregnancy outcome. A significant increase in fetal loss was observed in couples who matched for a 16-locus haplotype 30 encompassing the entire HLA locus. Christiansen et al. (1997) examined HLA-C and HLA-Bw in unexplained RSA couples. They found no variation in HLA-C, but a significantly higher number of RSA couples have the HLA-Bw4 haplotype than control couples. Jin et al., (1995) examined the degree of sharing of HLA-A. HLA-B, HLA-DR and HLA-DQ haplotypes. They found a significant excess of WO 99/43851 PCT/IE99/00012 7 HLA-DR sharing in couples with RSA, and also a significant excess of HLA-DQ sharing in couples with unexplained infertility. Several groups have recorded conflicting results. Billingard et al., (1995) examined sharing of HLA-A. HLA-B. and HLA-DR alleles and found no higher degree of HLA sharing in couples with RSA than in 5 fertile couples. Caudle et al., (1983) reported similar findings. HLA-A, HLA-B and HLA-DR alleles were typed in a large population of unexplained RSA couples and in control couples (Sbracia et al.. 1996). No increased sharing in HLA alleles was observed. In addition, there was no difference in the frequency of HLA alleles between RSA couples and control couples. Saski et al., (1997) reported an increase in the frequency of the HLA-DR4 allele in women who suffered from RSA compared to control 10 women. The role of HLA sharing as a risk factor for RSA remains controversial, and no studies have reported any diagnostic or prognostic significance to HLA sharing in individual couples. In addition, reports of significant sharing of class II genes are difficult to explain as fetal cells in contact with the maternal immune system during pregnancy are devoid of HLA class II expression. 15 According to the present invention there is provided a method for diagnosing susceptibility to normal pregnancy, pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility comprising the steps of: a) obtaining a fluid and/or tissue sample from a female and/or male and/or foetus; and either b) determining the sequence of all or part of the HLA-G nucleic acid, and/or HLA-G linked nucleic 20 acid: or c) detecting variant forms of all or part of the HLA-G protein, and/or proteins encoded by HLA-G linked genes or: d) measuring the functional activity of all or part of the HLA-G encoding protein and/or proteins encoded by HLA-G linked genes or: 25 e) measuring the size and/or level of all or part of HLA-G mRNA or mRNA transcribed from HLA-G linked genes or: f) measuring the size and/or level of all or part of HLA-G protein and/or protein encoded by HLA-G linked genes or: g) quantifying cells or molecules whose concentration changes as a result of HLA-G action; and 30 h) comparing any of the parameters b) to g) with those of a female and/or male and/or foetus of a normal pregnancy and/or a pregnancy with pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related fertility outcome.

WO 99/43851 PCT/IE99/00012 8 Preferably the HLA-G nucleic acid is analysed by the presence of the C and/or T allele of codon 93 in exon 3 and/or the insertion and/or deletion allele of exon 8. Preferably the effect of one or more of the HLA-G sequence variants on the functional activity of HLA-G and/or on the size and/or the level of all or part of the HLA-G mRMA and/or its encoded peptide is 5 measured. In its simplest the present invention provides a method of diagnosing susceptibility to pre-eclampsia and/or eclampsia and/or intrauterine growth retardation a nd/or susceptibility to miscarriage and/or miscarriage-related infertility comprising the steps of: a) obtaining nucleic acid from a parent and/or a prospective parent and/or foetus; O10 b) establishing the HLA-G sequence variants present in the parent and/or foetus by analysing the nucleic acid isolated in step (a); and c) comparing the HLA-G sequence variants identified in step (b) with known HLA-G sequence variants. Preferably, the HLA-G sequence variants are established by characterising all or part of the DNA 15 sequence of the HLA-G gene by methods selected from DNA sequencing, PCR-restriction fragment length polymorphism analysis, glycosylase mediated polymorphism detection, oligonucleotide hybridisation, gel electrophoretic detection of polymorphisms and amplification based detection approaches. Suitably, a stratified approach is used whereby the C/T-93 in exon 3 and insertion/deletion polymorphism 20 in exon 8 are first genotyped, followed by genotyping of other variations in exon 3, exon 2. intron 2, followed by exon 1 and 4, followed by the remainder of the HLA-G gene. The invention also provides a test kit for the diagnosis of susceptibility to normal pregnancy, pre eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility comprising: 25 a) oligonucleotide primers for amplification of all or part of the HLA-G gene and/or HLA-G linked DNA: b) amplification reagents for amplification of genomic DNA and/or RNA segments, selected from a DNA / RNA polymerase. a reverse transcriptase, the deoxyribonucleotides dATP, dCTP, dGTP. dTTP and dUTP. and/or ribonucleotides ATP, CTP, GTP, TTP and UTP, and reaction buffer; 30 c) reagents for identifying sequence variants in DNA and/or RNA; d) control DNA and/or RNA. Preferably the primers of (a) allow specific amplification of all or part of the HLA-G gene using the polymerase chain reaction. Several polymorphisms are known to occur in the HLA-G gene. A C to T WO 99/43851 PCT/IE99/00012 9 polymorphism occurs at nucletide 1488 in the third position of codon 93 and is referred to as C/T-93 herein where C-93 is one allele of the polymorphism and T-93 is the other allele of the polymorphism. Suitably, the C/T-93 polymorphism is genotyped by PCR amplification of a section of intron 2 - exon 3 using the primers 5 '-TACTCCCGAGTCTCCGGGTCTG-3'(SEQ ID NO. 1) as the forward primer and 5 5'-AGGCGCCCCACTGCCCCTGGTAC-3'(SEQ ID NO. 2) as the reverse primer giving rise to the amplified C-93 allele (SEQ ID NO. 4) and amplified T-93 allele (SEQ ID NO. 5) followed by by semi nested PCR amplification using the forward primer 5'-GACCGAGGGGGTGGGGCCAGGTTCT 3'(SEQ ID NO. 3) and the reverse primer 5'-AGGCGCCCCACTGCCCCTGGTAC-3'(SEQ ID NO. 1). In the semi-nested amplification reaction dTTP is replaced by dUTP. The 3' end of the forward primer is 10 designed so that the first U incorporated downstream of the forward primer is at, or distal to, the polymorphic site in codon 93. Following amplification using end labelled forward primer, glycosylase mediated cleavage of the amplified product is performed. Cleavage products are resolved by denaturing gel electrophoresis (20% polyacrylamide) and visualised by autoradiography. The C-93 allele is detected as a 32 n fragment (SEQ ID NO. 8)and the T-93 allele as a 27 n fragment (SEQ ID NO. 9). 15 The common 14 base pair insertion / deletion polymorphism in exon 8 of the HLA-G gene is referred to as I/D-E8 herein where I-E8 is one allele of the polymorphism and D-E8 is the other allele of the polymorphism. Suitably, genotyping of the HLA-G exon 8 deletion polymorphism is performed by amplifying a short section flanking the deletion location in exon 8. This is achieved using the polymerase 20 chain reaction with primers designed to hybridise to known DNA sequence in exon 8. The forward primer is 5'-TGTGAAACAGCTGCCCTGTGT-3' (SEQ ID NO. 10) and the reverse primer is 5' AAGGAATGCAGTTCAGCATGA-3' (SEQ ID NO. 11). The I/D exon 8 polymorphism is genotyped by size separation of the PCR products on a 10% non denaturing polyacrylamide gel and visualised by staining with ethidium bromide, the I-E8 insertion allele giving rise to a 151 bp product (SEQ ID NO. 12) 25 and the D-E8 deletion allele giving rise to a 137 bp product (SEQ ID NO. 13). Suitably, allele specific genotyping is performed in cases where maternal and paternal C/T-93 and I/D E8 HLA-G haplotypes cannot be directly assigned. This is achieved using allele specific primers which allows selective amplification of the I-E8 or D-E8 allele. Following allele specific amplification, the C/T 30 93 polymorphism is then genotyped using the GMPD assay described above. Primers for amplification of the I-E8 allele are 5'-TACTCCCGAGTCTCCGGGTCTG-3' (SEQ ID NO. 1) as the forward primer and 5"-CAAAGGGAAGGCATGAACAAATCTTG-3' (SEQ ID NO. 14) as the reverse primer. Primers for amplification of the D-E8 allele are 5'-TACTCCCGAGTCTCCGGGTCTG-3' (SEQ ID NO. 1) as the forward primer and 5'-GTTCTTGAAGTCACAAAGGGACTTG -3' (SEQ ID NO. 15) as the reverse 35 primer. Such allele specific amplification gives rise to four possible haplotypes, namely I-E8 and C-93 WO 99/43851 10 PCT/IE99/00012 10 haplotype (SEQ ID NO. 16), I-E8 and T-93 haplotype (SEQ ID NO. 17), D-E8 and C-93 haplotype (SEQ ID NO. 18) and D-E8 and T-93 haplotype (SEQ ID NO. 19). Suitably, haplotypes are constructed and suitably, transmitted and non-transmitted alleles to offspring / 5 foetus are assigned. Preferably, the amplification reagents include a thermostable DNA polymerase, amplification buffer and DNA precursor nucleotides. All or part of any HLA-G sequence and/or HLA-G linked sequence may also be amplified, by a method or combination of methods selected from nucleic acid sequence based amplification, self-sustained 10 sequence replication, transcription-mediated amplification, strand displacement amplification and the ligase chain reaction. Preferably the comparison of one or more variants identified is performed by association and/or linkage analaysis and/or transmission analysis. Preferably all or part of the HLA-G sequence is cloned into a vector. 15 The invention may involve: a) obtaining nucleic acid or fluid or tissue sample from a parent and/or prospective parent and/or foetus: b) establishing the HLA genotype or serotype of the parent/prospective parent and/or foetus by analysing the nucleic acid or fluid or tissue sample isolated in step (a); 20 c) comparing the HLA genotypes or serotypes identified in step (b) with known HLA genotypes or serotypes respectively. Preferably the method involves the measuring of cellular and/or soluble HLA-G levels. Preferably HLA G levels are measured by immunoassay using an antibody for specific HLA-G protein. The invention may involve identifying the variant form of HLA-G protein and/or the levels thereof present 25 in the sample. Preferably HLA-G variant proteins and/or levels thereof are detected and/or quantified by immunoassay using specific antibodies which detect HLA-G variants, or HLA-G protein. Alternatively, antibody specific for HLA-G protein variants and/or electrophoretic separation methods and/or chromatographic separation methods may be used. Preferred methods for detecting HLA-G protein and variants thereof include, enzyme linked immunosorbent assays (ELISA), radioimmuno-assays (RIA), 30 immunoradiometric assays (IRMA) and immnunoenzymatic assays (IEMA), including sandwich assays using monoclonal and/or polyclonal antibodies.

WO 99/43851 PCT/IE99/00012 11 The invention may involve measuring of level of molecules whose concentration changes as a direct and/or indirect result of HLA-G action. Preferably the molecules are selected from IL-.1, IL-2, IL-3, IL-4, IL-6, IL-10 beta and tumour necrosis factor alpha. Preferably the levels of such molecules are measured by immunoassay using antibodies specific for the molecules. 5 Alternatively, the method may involve measuring of levels of trophoblast specific markers. Preferably the trophoblast markers are cytokeratins pregnancy specific glycoprotein 1, human chorionic gonadotrophin and human placental lactogen. Preferably the levels of such molecules are measured by immunoassay using antibodies specific for the molecules. In one embodiment the method may comprise the steps of: 10 a) incubating blood mononuclear cells and/or a subset of such cells with one or more HLA-G variants and/or any combination thereof and/or cells expressing all or part of one or more variants of the HLA-G gene and/or a combination of one or more variants thereof, wherein the blood mononuclear cells and/or HLA-G variant is from a female and/or male and/or foetus b) analysing the activity of the blood mononuclear cells and/or the HLA-G and/or cells expressing one 15 or more HLA-G variant. Preferably, the blood mononuclear cells are obtained as a blood sample and/or tissue sample from the female and/or are obtained through matching the females blood mononuclear cells with blood mononuclear cells from a donor and/or cell line panel. Preferably populations of T cells and/or NK cells are isolated from the blood sample by density centrifugation and/or immunoselection. Preferably, blood 20 mononuclear cells matching the females blood mononuclear cells are identified from a test panel by matching the HLA serotype and/or extended HLA genotype and/or HLA-G genotype of the female with the HLA serotype and/or extended HLA genotype and/or HLA-G genotype of blood mononuclear cells. Preferably, HLA-G matching the male and/or female HLA-G is identified from a test panel by matching the HLA-G type and/or HLA-G genotype of the male and/or female with the HLA-G type and/or HLA-G 25 genotype of HLA-G proteins and/or cells expressing one or more HLA-G gene variants in the test panel. Preferably, such a test panel is assembled by growing cells expressing one or more HLA-G variants. Such cells may be derived from natural tissue such as placenta and/or created artificially by the introduction of one or more vectors bearing HLA-G gene variants which are capable of promoting the expression of the HLA-G gene into a cell and/or by inducing the expression of native HLA-G in cells. Suitably, the vector 30 used is plasmid, phage, viral, and/or artificial chromosome based. Preferably HLA-G protein is used as a crude preparation and/or fully or partially purified from such cells. HLA-G protein may be loaded with binding peptides naturally or artificially.

WO 99/43851 12 PCT/IE99/00012 Preferably, the HLA-G - blood mononuclear cell interaction is measured by assessing blood mononuclear cell activation including assessment of one or more of the following; cell proliferation, transformation cytotoxic response, surface marker expression, cytokine production, conjugate formation and target specificity. 5 The method may comprise the steps of: a) cloning the HLA-G gene from a parent and/or prospective parent and/or foetus; b) expressing the HLA-G protein from the cloned gene in vitro and/or in vivo; c) measuring the levels of activity of the expressed HLA-G protein; d) comparing the levels of activity of the expressed HLA-G protein with the levels of activity observed 10 for the normal HLA-G protein. The method may also comprise: a) establishing all or part of the HLA-G sequence and/or HLA-G linked sequences present in a sample from a female and/or male and/or foetus by analysing the nucleic acid from said sample; b) determining whether one or more of any variants or any combination thereof, identified in step (a) are 15 indicative of susceptibility to normal pregnancy or pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility by comparative analysis and/or analysis of the effect of one or more of the variants on the functional activity of HLA-G and/or on HLA-G mRNA. Preferably, the HLA-G sequence variants are established by characterising all or part of the DNA 20 sequence of the HLA-G gene and/or closely linked DNA including HLA-A, HLA-B, HLA-C, HLA-E, HLA-F and HLA-H genes by amplifying all or parts HLA-G or closely linked DNAs and identifying the sequence variants present using one or more sequence variation detection methods. Suitably, one or more copies of all or parts of the HLA-G gene is amplified by any of several amplification approaches such as the polymerase chain reaction (PCR), nucleic acid sequence based 25 amplification (NASBA), self sustained sequence replication (3SR), transcription-mediated amplification (TMA) and strand displacement amplification. Amplification of a target nucleic acid molecule may also be carried out using a the ligase chain reaction (LCR) and a variation of the LCR which employs a short PCR step (PLCR). Suitably, DNA or mRNA is used as the amplification substrate. Suitably, mRNA is converted into DNA using reverse transcriptase. Suitably, the amplified molecules are analysed directly 30 and/or may be cloned into a vector to facilitate analysis. Suitably, DNA sequence variations are detected by any one or more of a variety of gene variation detection methods including DNA sequencing, glycosylase mediated polymorphism detection, restriction fragment length polymorphism analysis, enzymatic or chemical cleavage assays, hybridisation to DNA probe arrays, allele specific oligonucleotide hybridisation assays, allele specific amplification methods such as the amplification refractory method WO 99/43851 13 PCT/IE99/00012 (ARMS), electrophoretic detection of polymorphisms based on migration through a gel matrix, 5' nuclease assay and ligase chain reaction. Suitably, it can be determined if one or more variants identified are known variants associated with susceptibility to normal pregnancy and/ or susceptibility to pre-eclampsia and/or eclampsia and/or 5 intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility. Alternatively, comparative analysis is performed by gene association and/or gene linkage methods to determine whether HLA-G variants and/or HLA-G linked variants are associated with normal pregnancy and/ or susceptibility to pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility. 10 Alternatively, HLA-G variants associated with normal pregnancy and/ or susceptibility to pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility can be identified by the effect of HLA-G variants on HLA-G function. Suitably, HLA-G variants are functionally analysed by measuring the interaction of one or more of the HLA-G variants and/or any combination thereof, with blood mononuclear cells and/or measuring the size 15 and level of the HLA-G messenger RNA and/or the size and level of HLA-G gene product and/or peptide binding for one or more of the HLA-G variants and/or any combinations thereof. HLA-G - blood mononuclear cell activity is measured by assessing blood mononuclear cell activation including assessment of one or more of the following; cell proliferation, cytotoxic response, surface marker expression, cytokine production, conjugate formation and target specificity. 20 The invention also relates to a pharmaceutical composition comprising a pharmaceutically effective amount of HLA-G protein and/or cells expressing HLA-G and/or one or more peptides which binds to HLA-G and/or blood mononuclear cells from a donor and/or a cells from a test panel known to interact with HLA-G variants, cytokines and any combination thereof including IL-I beta, IL-2. IL-3, IL-4, IL-6. IL-10 and tumour necrosis factor-alpha and/or inhibitors of cytokines and/or tumour necrosis factor 5 alpha and/or derivatives of cytokines and/or tumour necrosis factor-alpha, optionally with pharmaceutically-acceptable carriers or excipients. The invention also provides a method for screening for agents which can potentially be used as diagnostic indicators and/or drug targets for pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation by: 0 a) measuring the expression level of one or more genes and / or proteins in HLA-G expressing cells and /or blood mononuclear cells and / or T cell and /or natural killer cell subsets thereof following interaction with HLA-G and / or HLA-G expressing cells; WO 99/43851 14 PCT/IE99/00012 b) comparing the expression level identified in step (a) with the expression level in HLA-G expressing cells and /or the blood mononuclear cells and / or T cell and /or natural killer cell subsets thereof following interaction with HLA-G and / or HLA-G expressing cells in normal pregnancy and/ or pre eclampsia pregnancy and/or intrauterine growth retardation pregnancy and/or miscarriage pregnancy 5 and/or miscarriage-related infertility. In a further aspect the invention provides a method for screening for potential pre-eclampsia and eclampsia and intrauterine growth retardation and miscarriage and miscarriage-related infertility therapeutic agents selected from: a) identifying agents which alter the expression of HLA-G; 10 b) identifying agents which alter the activity of HLA-G; c) identifying agents which mimic the action of HLA-G; d) identifying agents which bind to HLA-G; e) identifying peptides which bind to HLA-G; f) identifying agents which bind to HLA-G receptors; 15 g) identifying expressed genes using DNA probe arrays in a cellular background in HLA-G expressing cells and/or blood mononuclear cells interacting with HLA-G and/or cells expressing HLA-G interacting with blood mononuclear cells; h) identifying expressed genes using DNA probe arrays in a cellular background whose expression is altered in response to HLA-G expression in the cells and/or in response to interacting cells expressing 20 HLA-G; i) identifying expressed proteins using mass spectrometry methods in HLA-G expressing cells and/or blood mononuclear cells interacting with HLA-G and/or cells expressing HLA-G interacting with blood mononuclear cells. Preferably sperm and/or semen and/or female reproductive tissue are screened for agents: 25 a) which alter the expression of HLA-G in fertilised eggs and/or embryos; b) which alter the cell cleavage rate of fertilised eggs and/or embryos; c) which induce cellular factors in cell in culture and/or cell in vivo that alter the cell cleavage rate of fertilised eggs and/or embryos. The method may involve: 30 a) measuring the expression level of one or more genes and/or proteins in HLA-G expressing cells and/or blood mononuclear cells and/or T cell and/or natural killer cell subsets thereof following interaction with HLA-G and/or HLA-G expressing cells; b) comparing the expression level identified in step (a) with the expression level in HLA-G expressing cells and/or the blood mononuclear cells and/or T cell and/or natural killer cell subsets thereof following 35 interaction with HLA-G and/or HLA-G expressing cells associated with normal pregnancy and/ or pre- WO 99/43851 PCT/IE99/00012 15 eclampsia pregnancy and/or intrauterine growth retardation pregnancy and/or miscarriage pregnancy and/or miscarriage-related infertility. Preferably, blood mononuclear cells and/or HLA-G expressing cells are obtained from a female and/or male and/or foetus and/or test panel of blood mononuclear cells and/or HLA-G expressing cells. 5 Preferably, gene expression is measured by any one or combination of several methods including hybridisation between cDNA and/or RNA from the cells and DNA probes and/or RNA probes and/or DNA probe arrays, quantitative amplification approaches such as quantitative (reverse transcriptase polymerase chain reaction) RT-PCR, 5' nuclease assay, ribonuclease protection assay and S 1 nuclease assay. 10 Preferably, protein expression is measured by any one or combination of several methods including one dimensional and/or two dimensional gel electrophoresis and staining of proteins and/or detection of one or more proteins using, enzyme linked immunosorbent assays (ELISA), radioimmunoassays (RIA), immunoradiometric assays (IRMA) and immunoenzymatic assays (IEMA), including sandwich assays and Western blotting using monoclonal and/or polyclonal antibodies. 15 Alternatively, the method may involve: a) measuring the expression level of one or more genes and/or proteins in cells expressing HLA-G; and b) comparing the expression level identified in step (a) with the expression level in HLA-G non expressing cells. Preferably, the cells are fertilised animal eggs and/or animal embryos. 20 The invention also provides a method for the prevention of pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility selected from: a) treatment of a female with all or part of a pharmaceutically effective amount of an effective HLA-G protein and/or peptides which bind to HLA-G and/or cells expressing HLA-G; 25 b) treatment of a female with all or part of a pharmaceutically effective amount of molecules or inhibitors of molecules whose level or activity is directly or indirectly altered by HLA-G; c) treatment of a female with all or part of a pharmaceutically effective amount of an agent which alters HLA-G expression; d) treatment of a female with all or part of a pharmaceutically effective amount of an agent which 30 alters NK cell activity: e) treatment of a female with all or part of a pharmaceutically effective amount of an agent which mimics all or part of HLA-G action: WO 99/43851 PCT/IE99/00012 16 f) treatment of a female with all or part of a pharmaceutically effective amount of molecules which inhibit the interaction between HLA-G and one or more of its receptors; g) treatment of a female with all or part of a pharmaceutically effective amount of an agent which alters the size and/or level of HLA-G mRNA; 5 h) treatment of a female with all or part of a pharmaceutically effective amount of an agent which alters HLA-G related blood mononuclear cell activity; i) treatment of a female with blood mononuclear cells that recognise foetal and/or self HLA-G; j) treatment of a female with HLA-G protein and/or cells expressing HLA-G. The invention may comprise: 10 a) obtaining blood mononuclear cells and/or T cell and/or natural killer cell subsets thereof and/or HLA-G and/or HLA-G expressing cells from a female and/or male and/or foetus and/or test panel; b) measuring the expression level of one or more genes and/or proteins in the HLA-G expressing cells and/or blood mononuclear cells following interaction with HLA-G and/or HLA-G expressing cells; c) comparing the expression level identified in step (b) with the expression level in the blood 15 mononuclear cells and/or HLA-G expressing cells in normal pregnancy and/ or pre-eclampsia pregnancy and/or intrauterine growth retardation pregnancy and/or miscarriage pregnancy and/or miscarriage-related infertility. Preferably, the blood mononuclear cells and/or HLA-G expressing cells are obtained as a blood sample and/or tissue sample. Preferably populations of T cells and/or NK cells are isolated from the blood sample 20 by density centrifugation and/or immunoselection. Preferably HLA-G expressing cells are isolated by immunoselection. The invention also provides a method for improving fertility and pregnancy outcome wherein male and/or female partners and/or sperm and/or ova and/or recipients of fertilised eggs and/or zygotes / and/or embryos are selected on the basis of HLA-G so that their genotypes and/or serotypes are associated with 25 normal pregnancy outcomes and/or not associated with pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility. In particular there is provided a method for improving pregnancy success selected from: a) pre-treating the female with sperm and/or attenuated forms thereof, and/or semen and/or fractions thereof from a male with a known HLA-G genotype, prior to mating with a male of a different HLA-G 30 genotype, and/or in vitro fertilisation using sperm from a male of a different HLA-G genotype and/or embryo transfer where the male HLA-G is of a different HLA-G genotype; b) mixing sperm of a known HLA-G genotype with sperm and/or attenuated forms thereof, and/or semen and/or fractions thereof from a male with a different HLA-G genotype prior to in vitro fertilisation.

WO 99/43851 17 PCT/IE99/00012 Preferably fertility and/or pregnancy outcome are improved by selection of male and / or female partners and / or sperm and / or ova and / or recipients of fertilised eggs and / or zygotes / and / or embryos so that (a) their HLA-G and /or HLA genotypes and /or serotypes or (b) the activity of their HLA-G and / or blood mononuclear cells interacting with HLA-G are indicative of normal pregnancy outcomes and / or 5 not associated with pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility. Cloning of all or part of one or more HLA-G genes in any of the above methods may be achieved by amplification of all or part of one or more HLA-G genes and insertion of all or part of the amplified product into a vector capable of expressing the inserted gene. Expression of the HLA-G protein from the 10 cloned gene in any of the above methods may be achieved by introduction of the expression vector into a suitable host such as a bacterium or an eukaryotic cell in culture. The level of activity of the expressed HLA-G protein in any of the above methods may be achieved by a) directly and/or indirectly measuring the interaction of the HLA-G protein and/or cells expressing HLA-G protein with blood mononuclear cells and/ or b) detecting one or more molecules whose level is altered as a result of the interaction of the HLA 15 G protein and/or cells expressing HLA-G protein with blood mononuclear cells and/or c) measuring changes in cell cleavage rate due to direct and/or indirect action of the HLA-G protein and/or cells expressing HLA-G protein with blood mononuclear cells. HLA-G as defined herein refers to any form of HLA-G and / any complex involving HLA-G including different isoforms of HLA-G arising from alternative splicing pathways, combination of different HLA-G 20 isoforms, secreted HLA-G. membrane bound HLA-G HLA-G with peptides bound and HLA-G associated with beta -2-microglobulin. HLA-G protein refers to any crude, partially and/or fully purified form of HLA-G. The invention also provides use of a DNA sequence selected from any one of sequence I.D.s 1 to 21 for diagnosis of susceptibility to or in a test kit for the diagnosis of susceptibility to normal pregnancy, pre 25 eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility, for monitoring progress of pregnancy, for use in the manufacture of a medicament. in a method for screening potential therapeutic agents, in a method for screening for potential diagnostic indicators and/or drug targets, in a method for improving pregnancy success or in a method for the prevention of pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or 30 susceptibility to miscarriage and/or miscarriage-related infertility, for monitoring progress of pregnancy. The invention also provides a method for induction of tolerance in a host to a non-self tissue which comprises administering HLA-G and /or HLA-G loaded with peptides from the non-self tissue and /or WO 99/43851 18 PCT/IE99/00012 HLA-G expressing cells derived from or related to the non-self tissue, and/or a non-self tissue bearing an introduced HLA-G gene so that HLA-G is expressed in all or part of the tissue. In a further aspect the invention provides a method for the treatment of autoimmune disease which comprises administering HLA-G and /or HLA-G loaded with peptides from a self or non-self tissue and / 5 or with specific autoimmune antigen and /or HLA-G expressing cells from a self and/or non-self tissue and/or a self and/or non-self tissue bearing an introduced HLA-G gene so that HLA-G is expressed in all or part of the tissue. Methods Identification of Subjects 10 In the initial phase of sampling pre-eclamptic patients were identified as primagravidas who were delivered by caesarean section at or prior to 36 weeks gestation because of a deterioration in pregnancy indicative of pre-eclampsia. Diagnostic symptoms were a rise in blood pressure >15mm Hg diastolic or >30mm Hg systolic from measurement in early pregnancy or to >140/90mm Hg in late pregnancy, and one or more of the following: proteinuria, odema, headache, visual disturbance, epigastric pain. Control 15 patients were identified as primagravidas with normal delivery and normal blood pressure. 5-10mls of blood were taken from the offspring of primagravida pre-eclampsia and normal pregnancies with informed consent. In the second phase of sampling, blood samples and/or cheek swab sample for DNA extraction were collected from control trios following delivery. The appropriate informed consent was obtained from all 20 subjects. Control mothers were identified as primagravidas under the age of thirty three with normal delivery and normal blood pressure. All individuals were Irish and Caucasian by origin. Mothers were interviewed to ensure that they were primagravidas. Primagravida (first pregnancy) pre-eclampsia trios where the mothers suffered severe pre-eclampsia and a matching control group of normal primagravida trios were identified and sampled. Families (mother, father, first and second offspring) where the mother 25 had two or more successful normal pregnancies in the absence of pregnancy related disorders including pre-eclampsia and miscarriage were also identified and sampled. Families (mother, father, first and second offspring) where the mother had pre-eclampsia in the first pregnancy and a normal second pregnancy in the absence of pregnancy related disorders including pre-eclampsia and miscarriage were also identified and sampled. Couples with recurrent spontaneous abortion were also identified and sampled. To minimise 30 the possibility of misdiagnosis of PE, we applied stringent criteria to ascertainment of samples. Essentially pre-eclampsia cases were identified as primagravidas under the age of 35 who were delivered by caesarean section at, or prior to. 36 weeks gestation because of a deterioration in pregnancy indicative WO 99/43851 19 PCT/IE99/00012 of pre-eclampsia. Diagnostic symptoms were a rise in blood pressure of >15mm Hg diastolic or >30mm Hg systolic from measurement in early pregnancy or to >140/90mm Hg in late pregnancy, and one or more of the following: proteinuria, odema, headache, visual disturbance, epigastric pain. Diagnostic symptoms were completely resolved within 3 months after delivery. A preliminary survey of the sisters of 5 the pre-eclamptic women in this study did not reveal an increased incidence of the condition, indicating that pre-eclampsia in the cohort of mothers investigated here is sporadic. A cohort of couples where the mother had three or more consecutive miscarriages were identified (recurrent miscarriage). Genotyping of HLA-G polymorphism Genomic DNA was extracted from peripheral blood samples and/or cheek swab samples by standard 10 methods. DNA concentration was determined by absorbance at 260nm for samples where DNA was isolated from blood. The integrity and purity of the genomic DNA was determined by agarose gel electrophoresis and OD260:OD280 ratio respectively. The C-93T HLA-G polymorphism is also known as the C/T codon 93 polymorphism (and as HLA-G C 1488T) and referred to as C/T-93 herein where C-93 is one allele of the polymorphism and T-93 is the 15 other allele of the polymorphism. In order to genotype the C/T-93 polymorphism in the genomic DNA samples. exon 3 of the HLA-G gene was first amplified using the polymerase chain reaction with primers designed to hybridise to the known DNA sequence flanking exon 3. The forward primer was 5' TACTCCCGAGTCTCCGGGTCTG-3' (SEQ ID NO. 1) and the reverse primer was 5' GAGGCGCCCCACTGCCCCTGGT-3'. 20 The polymerase chain reaction was carried out in a total volume of 25 l, with 100 ng genomic DNA, 50Ong of each primer, 0.2mM of each deoxynucleoside triphosphate (dATP. dCTP, dGTP and dTTP) 50mM KCI, 10mM Tris-HCl, pH 9.0 at 25 0 C, 0.1% Triton X-100, 1.5mM MgCl and 0.5U of Taq Polymerase. Reaction mixtures were covered with an equal volume of mineral oil and amplification was carried out using the "hot start" technique in a thermal cycler. The conditions for amplification involved denaturation 25 at 94 OC for 5 min followed by addition of Taq Polymerase. Thirty cycles were then performed: 94 oC for 1 min,. 63oC for 1 min, 72 0 C for 1 min and finally a 10 min extension at 72 0 C. Genotyping of the C/T-93 HLA-G polymorphism was then performed using a semi nested amplification approach and the Glycosylase Mediated Polymorphism Detection method (Vaughan and McCarthy 1998). A 319bp section of the HLA-G gene encompassing the C/T-93 polymorphism location was amplified 30 using a semi nested polymerase chain reaction approach from the previously amplified exon 3 of the HLA-G gene using the exon 3 reverse primer and the internal forward primer 5' GACCGAGGGGGTGGGGCCAGGTTCT-3' (SEQ ID NO. 3). The forward primer was end labelled by WO 99/43851 PCT/IE99/00012 20 incubation with polynucleotide kinase in the manufacturers buffer (New England Biolabs) and 5ptCi 32P ATP (3000Ci/mmol) for 30 min at 37 0 C followed by ethanol precipitation to remove unused labelled nucleotide. The semi nested amplification reaction was carried out in a total volume of 10[ l, with 1 tl of a l in 500 dilution of the previously amplified exon 3 product, 3pmoles of forward and reverse primer, 5 0.2mM of each deoxynucleoside triphosphate (dATP, dCTP, dGTP and dUTP) 50mM KCI, 10mM Tris HCI, pH 9.0 at 25 0 C, 0.1% Triton X-100,1.5mM MgCI and 0.5U of Taq Polymerase. Reaction mixtures were covered with an equal volume of mineral oil and amplification was carried out using the "hot start" technique in a thermal cycler. The conditions for amplification involved denaturation at 94 0 C for 5 min followed by addition of Taq Polymerase. Thirty cycles were then performed: 94 0 C for 1 min,. 64 0 C for 1 10 min,. 72 0 C for 1 min and finally a 10 min extension at 72'C. The reaction mixture was then treated with exonuclease I to digest the primers not extended in the amplification step. This was achieved by incubating the PCR reaction mixture with 0.4 units of exonuclease I at 37 0 C for 30 min. The exonuclease was subsequently heat inactivated by incubating the reaction at 80 0 C for 15 min. Uracil DNA-glycosylase (0.5 units) was then added and the incubation continued at 37 0 C for min. 15 Following treatment with uracil DNA-glycosylase, the AP sites generated in the amplified product were cleaved to completion by adding NaOH to a final concentration of 0.05M and heating the mixture for 15 min at 95 0 C. Under these conditions, cleavage occurs on the 5' side of each AP site. The reaction was then neutralised by addition to Tris base to 30mM final concentration. Both Exol and UDG are diluted containing 0.07M Hepes KOH pH 8.0, ImM EDTA, ImM DTT and 50% glycerol. 20 An equal volume of formamide loading dye (90% formamide, 0.025% Bromophenol blue, 0.025% Xylene cylanol) was added to the sample which was then heated at 85 0 C for 5 min. The sample was then loaded onto a 20% denaturing (7M urea) polyacrylamide gel and electrophoresis was carried out for 3-4 hours at 60W for size analysis of the cleaved products in the sample. Following electrophoresis, autoradiography was carried out by exposing the gel directly to X-ray photographic film for 12 hrs at -70 0 C. During the 25 second phase of the genotyping, an improved protocol was used. Essentially, PCR amplification was carried out in 25 ml reactions, each of which contained 100 ng genomic DNA, PCR buffer (100 mMTris HCI pH 8.3 (20 0 C), 500 mM KCI, 15 mM MgCl2,), 200 mM of each dNTP, 300 nM of each primer and 0.5 U Taq polymerase (Boehringer). Conditions for amplification of exon 3 were 30 cycles at 940 C for 45 s. 6 1 0 C for 45 s, 7 2 0 C for 60 s using 5'-TACTCCCGAGTCTCCGGGTCTG-3'(SEQ ID NO. 1) as 30 the forward primer and 5'-AGGCGCCCCACTGCCCCTGGTAC-3'(SEQ ID NO. 2) as the reverse primer giving rise to the amplified C-93 allele (SEQ ID NO. 4) and amplified T-93 allele (SEQ ID NO. 5). All of the samples were then genotyped for the HLA-G C/T-93 polymorphism using the recently described glycosylase mediated polymorphism detection (GMPD) method (Vaughan & McCarthy, 1998).

WO 99/43851 21 PCT/IE99/00012 21 Essentially a 319 bp fragment was amplified by semi-nested PCR from exon 3 using the forward primer 5"-GACCGAGGGGGTGGGGCCAGGTI'CT-3'(SEQ ID NO. 3) and the reverse primer 5' AGGCGCCCCACTGCCCCTGGTAC-3'(SEQ ID NO. 1) giving rise to the amplified C-93 allele (SEQ ID NO. 6) and amplified T-93 allele (SEQ ID NO. 7). In the semi-nested amplification reaction dTTP 5 was replaced by dUTP. The 3' end of the forward primer was designed so that the first U incorporated downstream of the forward primer was at, or distal to, the polymorphic site in codon 93. Following amplification using p 3 2 end labelled forward primer, glycosylase mediated cleavage of the amplified product was performed. Cleavage products were resolved by denaturing gel electrophoresis (20% polyacrylamide) and visualised by autoradiography. The C-93 allele was detected as a 32 n fragment 10 (SEQ ID NO. 8)and the T-93 allele as a 27 n fragment (SEQ ID NO. 9). The common 14 base pair insertion / deletion polymorphism in exon 8 of the HLA-G gene is referred to as I/D-E8 herein (also known as where I-E8 is one allele of the polymorphism and D-E8 is the other allele of the polymorphism. Genotyping of the HLA-G exon 8 deletion polymorphism was performed by amplifying a short section flanking the deletion location in exon 8. This was achieved using the 5 polymerase chain reaction with primers designed to hybridise to known DNA sequence in exon 8. The forward primer was 5'-TGTGAAACAGCTGCCCTGTGT-3' (SEQ ID NO. 10) and the reverse primer was 5'-AAGGAATGCAGTTCAGCATGA-3' (SEQ ID NO. 11). The polymerase chain reaction was carried out in a total volume of 25RIl, with 100 ng genomic DNA, 50 Ong of each primer, 0.2mM of each deoxyribonucleoside triphosphate (dATP, dCTP, dGTP and dTI'P) 0 50mM KCI. 10mM Tris-HCl, pH 9.0 at 25 0 C, 0.1% Triton X-100, 0.5mM MgCl and 0.5U of Taq Polymerase. Reaction mixtures were covered with an equal volume of mineral oil and amplification was carried out using the "hot start" technique in a thermal cycler. The conditions for amplification involved denaturation at 94oC for 5 min followed by addition of Taq Polymerase. Thirty cycles were then performed: 94 0 C for 1 min, 54oC for 1 min, 72 0 C for 1 min and finally a 10 min extension at 72 0 C. The 5 I/D exon 8 polymorphism was genotyped by size separation of the PCR products on a 10% non denaturing polyacrylamide gel and visualised by staining with ethidium bromide, the I-E8 insertion allele giving rise to a 151 bp product (SEQ ID NO. 12) and the D-E8 deletion allele giving rise to a 137 bp product (SEQ ID NO. 13). Allele specific genotyping. In order to gain more information from the transmission of HLA-G S polymorphisms in the second phase of the work, allele specific genotyping was performed. In the majority of cases, maternal and paternal C/T-93 and I/D-E8 HLA-G haplotypes could be directly assigned. In cases where all members of a trio were heterozygous for either C/T-93 or I/D-E8 polymorphisms, allele specific amplification was performed in order to assign haplotypes. This was achieved using allele WO 99/43851 PCT/IE99/00012 22 specific primers which allowed selective amplification of the I-E8 or D-E8 allele. Following allele specific amplification, the C/T-93 polymorphism was then genotyped using the GMPD assay described above. Conditions for amplification of the I-E8 allele were 30 cycles at 94 0 C for 45s, 64oC for 45s, 72 0 C for 60s using 5'-TACTCCCGAGTCTCCGGGTCTG-3' (SEQ ID NO. 1) as the forward primer and 5' 5 CAAAGGGAAGGCATGAACAAATCTTG-3' (SEQ ID NO. 14) as the reverse primer. Conditions for amplification of the D-E8 allele were 30 cycles at 94oC for 45s, 56 0 C for 45 s, 72 0 C for 60s using 5' TACTCCCGAGTCTCCGGGTCTG-3' (SEQ ID NO. 1) as the forward primer and 5' GTTCTTGAAGTCACAAAGGGACTTG -3' (SEQ ID NO. 15) as the reverse primer. Such allele specific amplification gives rise to four possible haplotypes, namely I-E8 and C-93 haplotype (SEQ ID 10 NO. 16), I-E8 and T-93 haplotype (SEQ ID NO. 17), D-E8 and C-93 haplotype (SEQ ID NO. 18) and D-E8 and T-93 haplotype (SEQ ID NO. 19). All individuals in both sets of trios were genotyped for some or all of these polymorphisms (table 1 and 2), transmitted and non-transmitted alleles were assigned and haplotypes were constructed. All individuals were genotyped for the C/T-93 and I/D-E8 polymorphism. Comparative statistical analysis was 15 performed. In the second phase of the work more elaborate analysis was performed. Comparison of allele and haplotype frequencies and genotype distribution for the polymorphisms between and within the sample cohorts, and was performed using chi-squared contingency table analysis and/or log linear model analysis and/or transmission disequilibrium testing. In cases where all members of a trio were heterozygous for C/T-93 or I/D-E8 polymorphisms, allele 20 specific amplification was performed to determine haplotypes and thus assign transmitted and non transmitted alleles. This was achieved using primers which allowed specific amplification of a section of the HLA-G gene from either the insertion, or, deletion in exon 8 to a site 5' of codon 93. The C/T-93 polymorphism was then genotyped in the specifically amplified allele. Using this approach transmitted and nontransmitted alleles to offspring were assigned. 25 Results All of the genetic studies to date apart from the one concerning the HLA-G deletion polymorphism have examined the genotype of the pre-eclamptic mother. We took the view that foetal HLA-G is the most likely candidate gene for pre-eclampsia. Pre-eclampsia trios where the offspring was the offspring of the primagravida pre-eclampsia pregnancy and a control group of offspring of normal primagravida 30 pregnancies were studied. 54 pre-eclamptic offspring and 48 control offspring were included in the investigation.

WO 99/43851 23 PCT/IE99/00012 23 Genetic analysis of the HLA-G C/T-93 polymorphism in the pre-eclamptic offspring group revealed homozygosity for the C-93 allele in 7 cases (1 3 %),homozygosity for the T-93 allele in 3 cases (5.6%) and heterozygosity in 44 cases (81.4%). In comparison, control offspring showed homozygosity for the C-93 allele in 18 cases (37.5%), homozygosity for the T-93 allele in 8 cases (16.6%) and heterozygosity in 22 5 cases (45.8%). The frequencies of the C-93 and T-93 allele in the pre-eclamptic offspring were 0.537 and 0.463 respectively. In comparison, the frequencies in the control offspring were 0.604 and 0.396 respectively. The expected frequency distribution of the C-93 and T-93 alleles can be estimated with the formula p2 + 2pq + q2 = 1. With the allelic frequencies of p=0.537 and q=0.463 in the pre-eclamptic offspring group, 10 the expected distribution of genotypes should be C-93/C-93 = 0.288, C-93/T-93 = 0.502 T-93/T-93 = 0.214. With the allelic frequencies of p=0.604 and q=0.396 in the control offspring group, the expected distribution of genotypes should be C-93/C-93 - 0.36, C-93/T-93 = 0.478, T-93/T-93 = 0.156. In comparison with the control group the distribution of genotypes in the pre-eclamptic offspring group is significantly different (Chi-square - 11.01, p<0.001, table 1). 15 It has been reported that no significant association was observed between the HLA-G exon 8 deletion polymorphism and pre-eclamptic offspring (Humphrey et al., 1995). In this work we also genotyped the pre-eclamptic and control offspring for the HLA-G deletion polymorphism. Genetic analysis of the HLA G deletion polymorphism in the pre-eclamptic (n=51) and control offspring (n=55) groups revealed homozygosity for the normal allele in 12 cases (23.5%) and 13 cases (23.6%) respectively, homozygosity 20 for the deletion allele in 8 cases (15.7%) and 12 cases (21.8%) respectively, and heterozygosity in 31 cases (60.8%) and 30 cases (54.5%) respectively. The frequencies of the normal and deletion allele in the pre-eclamptic offspring were 0.539 and 0.46 respectively. In comparison, the frequencies in the control offspring were 0.509 and 0.491 respectively. In the pre-eclamptic offspring group, the expected distribution of genotypes should be normal 25 allele/normal allele = 0.29 1, normal allele/deletion allele = 0.497, deletion allele/deletion allele = 0.212. The expected distribution of genotypes in the control offspring group should be normal allele/normal allele = 0.259. normal allele/deletion allele = 0.500, deletion allele/deletion allele = 0.241. In comparison with the control group the distribution of genotypes in the pre-eclamptic offspring group is not significantly different (Chi-square = 0.69, p<0.30, table 1). 30 The parents of the pre-eclamptic and control offspring were genotyped for the HLA-G C/T-93 genotype and the genotypes were analysed in conjunction with the offspring genotypes. In this analysis we scored the number of cases where the offspring had inherited a paternal C/T-93 allele that was not present in the maternal genotype. For the pre-eclamptic offspring, 41% of cases had a paternal allele of the C/T-93 WO 99/43851 24 PCT/IE99/00012 genotype that was not present in the maternal genotype. By contrast, for the control offspring, 28% of cases had a paternal CFF/T-93 allele not represented in the maternal genotype. Discussion In this investigation, two common polymorphisms in the HLA-G gene in the offspring of pre-eclamptic 5 and normal mothers were examined for association with pre-eclampsia. The two offspring groups are from Southern Ireland and are from the same ethnic background. This is the first report determining association between pre-eclampsia in the mother and the foetal HLA-G genotype. Our results show a strong association between pre-eclampsia in the mother and heterozygosity for the C/T-93 polymorphism in offspring. This indicates that transmission of HLA-G alleles to offspring 10 is different in normal offspring than in pre-eclampsia offspring. The result indicates that screening for susceptibility to pre-eclampsia can be achieved by genotyping of HLA-G in the mother and partner. Furthermore. since pre-eclampsia is associated with intrauterine growth retardation and miscarriage, it is likely that screening for susceptibility to intrauterine growth retardation, miscarriage and miscarriage-related infertility may also be achieved by HLA-G genotyping in the potential parents. 15 Following these results and to lend further support to this key finding, we expanded the number of subjects analysed to include additional normal primagravida trios where the mother had no history of pregnancy related problems, pre-eclampsia primagravida trios, families (mother, father, first and second offspring) where the mother had two or more successful normal pregnancies in the absence of pregnancy related disorders including pre-eclampsia and miscarriage, families (mother, father, first and second 20 offspring) where the mother had pre-eclampsia in the first pregnancy and a normal second pregnancy in the absence of pregnancy related disorders including pre-eclampsia and miscarriage. As pre-eclampsia has been shown to be associated with miscarriage, we also included a cohort of recurrent miscarriage couples. The C/T-93 and I/D-E8 polymorphisms were genotyped in all individuals (Table 2 and Table 6), haplotypes were assigned and more elaborate statistics were applied to support our initial finding. 25 Linkage of HLA-G to pregnancy success in normal primagravidas For transmission analysis of individual polymorphisms, we applied the log linear model and the transmission disequilibrium test. We also examined other polymorphisms in the HLA-G gene. The frequency of the A/T-31, A/T-107 and C/A- 10 polymorphisms in the sample cohorts was 1.0/0.0), 0.95/0.05 and 0.94/0.06 respectively. For the analysis, we still utilised the commonly occurring C/T-93 30 and I/D-E8 polymorphisms. Using the allele specific amplification approach we assigned transmitted and nontransmitted alleles in the trios (Table 3). Since maternal genotype, foetal genotype (and specifically the paternal origin of foetal alleles) could potentially influence pregnancy outcome, a number of different WO 99/43851 25 PCT/IE99/00012 comparisons were made using this data. The log linear model of Weinberg et al. 1998 allows for causal scenarios in which the foetal genotype, parental genotypes or combinations thereof are directly relevant to risk. Maximum likelihood log-linear models of case-parent triad data were fitted. The log-linear model predicts the expected numbers of the 16 possible family types observed (for a bi-allelic marker), allowing 5 for several "genetic risk factors" i.e. (a) whether the offspring carried one or more copies of an allele, (b) whether the mother carried one or more copies of an allele (c) a maternal origin effect and (d) a paternal origin effect (Table 4). The analysis was stratified on parental mating type assuming Hardy-Weinberg equilibrium, this corrects for the number of a particular allele that is found among the four parental alleles in a particular mating type, which has an obvious but uninteresting effect on allele distributions among the 10 offspring. The advantage of this framework is that nested models of differing complexity are validly compared. starting with simple allelic effects and adding extra factors. We use a four-factor model, and then a stepwise reduced model which has removed the less significant factors automatically, thus providing a simpler model that accounts for important departures from expectations. Fitting a four-parameter model to the data (Table 4) was significant for both the C-93 allele (p=0.006) 15 and especially significant for the I-E8 allele (p=0.001). Stepwise elimination of parameters which are not significant revealed that most of this can be accounted for by two effects: an effect due to foetal HLA-G alleles and, most strongly, by a parental origin effect for each allele (Table 4). Thus, both the C-93 and I E8 alleles are significantly under-represented among offspring, and where they do occur, tend to be of paternal and maternal origin respectively. The maternal alleles in themselves are not significantly biased 20 once these other effects are allowed for. The I-E8 allele is over 4 times more likely to be of maternal origin than expected (95% confidence interval 2.2-9.8). When the transmission frequencies of maternal and paternal alleles to offspring were compared using chi squared contingency table analysis, highly significant differences were observed for the C/T-93 and I/D E8 alleles (p-l=0.009 and p-l=0.000001 respectively, Table 5, Table 3). This reflected a deficit in 25 transmission of maternal C-93, D-E8 and paternal T-93, I-E8 alleles (Table 3). A significant difference was also observed between maternal and paternal non-transmitted I/D-E8 alleles showing that the maternal genotype plays a role in pregnancy outcome. We verified that the highly significant findings from log-linear modelling by simpler comparisons. The transmission disequilibrium test transmission disequilibrium test assesses whether assess whether 30 transmission of maternal and paternal alleles from heterozygous parents to offspring differed from the null expectation (of 50:50) and is valid even when Hardy-Weinberg equilibrium is violated by unusual population structure. When the transmission disequilibrium test was applied, the transmission of the C/T 93 and I/D-E8 alleles to offspring did not differ from the null expectation (Table 5). However, significant WO 99/43851 PCT/IE99/00012 26 deviations from the null expectation were observed when maternal and paternal transmission frequencies were analysed independently of each other. Transmission of the maternal T-93 (chi-squared transmission disequilibrium test, p=0.032) and I-E8 (chi squared transmission disequilibrium test, p=0.0005) alleles and also the paternal D-E8 allele (chi-squared 5 transmission disequilibrium test, p=0.0 1) to offspring was markedly more frequent than expected (Table 5). There was an excess transmission of the maternal T-93 allele from heterozygous mothers to offspring. Specifically, thirty out of forty four offspring inherited the maternal T-93 allele from C/T-93 heterozygote mothers and forty out of fifty two offspring inherited the maternal I-E8 allele from I/D-E8 heterozygote mothers(Table 3). In addition, a contrasting excess transmission of the paternal D-E8 allele from 10 heterozygote fathers to offspring (thirty one out of forty three cases) was observed (Table 3). These findings closely match the findings of the log-linear model. We examined the data to determine if the transmission distortion could be accounted for by transmission to female or male offspring alone. However, there was no evidence that this was the case since of the forty out of fifty two offspring inheriting the maternal I-E8 allele from heterozygote mothers, 19 were females 15 and 21 were males. Similarly, of the thirty one out of forty three offspring inheriting the paternal D-E8 allele from heterozygote fathers, 15 were females and 16 were males. The primagravida mothers investigated here differed significantly from Hardy-Weinberg expectations (pl =0.006) for I/D-E8 genotype frequencies (observed genotype frequencies: I/I; 17, I/D; 58, D/D; 15, expected I/I; 24, I/D; 45, D/D; 21). The transmission disequilibrium test results shows a significant effect 20 without assuming Hardy-Weinberg equilibrium and thus support the log linear model results which was calculated assuming Hardy-Weinberg equilibrium. Assignment of alleles transmitted and non transmitted to offspring was determined for five independent HLA-G polymorphisms permitting haplotype construction, and comparison of transmitted and non transmitted haplotypes. Thirteen haplotypes were observed (Table 6). Four of these, a-a-a-a-b, a-b-a-a-a 25 a-a-a-a-a and a-b-a-a-b, were relatively common. Differences between the frequencies of maternal and paternal transmission were apparent for all four common haplotypes indicating that the distortion of HLA-G allele transmission to primagravida offspring could still be accounted for by biases in transmission of the C/T-93 and I/D-E8 polymorphisms. We then constructed haplotypes for the C/T 93 and I/D-E8 polymorphisms alone for comparison purposes. 30 The C-D and T-I haplotype were most common (Table 7). The strength of linkage disequilibrium between the two markers is indicated by the high frequency of the C-D haplotype, for which the disequilibrium, expressed as a proportion of the maximum disequilibrium (D/Dmax). is 0.344. Comparison of maternally WO 99/43851 PCT/IE99/00012 27 and paternally transmitted haplotypes to offspring using chi-squared contingency table analysis, revealed a highly significant difference between maternally and paternally transmitted haplotypes to offspring (p3=0.

0 00003) reflecting a deficit in transmission of both maternal haplotypes bearing the D-E8 allele (C D and T-D) and both paternal haplotypes bearing the I-E8 allele (C-I and T-I) (Table 7). A significant 5 difference was also observed between maternally and paternally non-transmitted haplotypes to offspring (P3=0.028) showing that the maternal HLA-G genotype plays a role in pregnancy outcome. Maternally and paternally transmitted haplotypes to individual offspring are shown in Table 8. The maternally transmitted T-I haplotype and the paternally transmitted C-D haplotype combination occurs in twenty one (34%) of the control offspring. By contrast the possible alternative combination (maternally 10 transmitted C-D and paternally transmitted T-I) does not occur in any of the control offspring even though twenty two of the matings have this possibility. In the primagravida offspring, homozygosity did not deviate from Hardy-Weinberg expectations. Comparison of the observed and expected number of homozygotes vs. heterozygotes within the offspring did not reveal any significant differences (p 1 = 0.256 for C/T-93 and p I = 0.82) showing that selection 15 against homozygotes does not occur in primagravidas. The significant distortion observed for transmission of HLA-G alleles to primagravida offspring provides evidence for maternal and paternal allele specific HLA-G based selection of foetuses in normal primagravidas. The selection observed was most pronounced for the HLA-G I/D-E8 polymorphism. The log linear model and transmission disequilibrium test analysis shows strong selection for the maternal I-E8 20 and paternal D-E8 allele in offspring. The selection effect is most dramatic for foetal combinations of the I/D-E8 alleles. In the total sample, thirty seven of the eighty four offspring have a maternal I-E8 paternal D-E8 allele combination. By contrast, there are only five offspring with the alternative maternal D-E8 paternal I-E8 allele combination (calculated from Table 8). These results show that maternal D-E8 paternal I-E8 foetuses are subject to significantly increased postzygotic prenatal loss and identify HLA-G 25 as a key gene influencing this process. The deficiency of maternal D-E8 paternal I-E8 offspring is approximately 29% and closely matches postzygotic prenatal loss in prospective mothers which has been estimated to occur at a frequency of about 31% and in more than 20% of these cases, such loss occurs very early in pregnancy and is clinically unrecognisable. The dichotomous effect whereby the maternal HLA-G I-E8 allele seemingly imparts a protective effect to 30 the foetus while the equivalent paternal allele is detrimental is somewhat suggestive of genomic imprinting. however reports to date indicates that imprinting does not occur at the HLA-G locus.

WO 99/43851 PCT/IE99/00012 28 It is not clear why there is an excess of heterozygote mother for the I/D-E8 genotype. The excess of heterozygote mother for the I/D-E8 genotype the population of mothers investigated have been selected on the basis of normal pregnancy outcome and as such represent a select group of the female population since the whole female population would include several other categories of females including infertile 5 women (10-15% incidence), women that had a miscarriage (10-15% incidence) or pre-eclampsia (5-10% incidence). In the primagravida offspring, homozygosity did not deviate from Hardy-Weinberg expectations. Comparison of the observed and expected number of homozygotes vs. heterozygotes within the offspring did not reveal any significant differences (p = 0.45 for C/T-93 and p I = 0.81 I/D-E8) indicating that 10 selection against homozygotes does not occur in primagravidas. The preferential transmission of a maternal I-E8 and paternal D-E8 HLA-G alleles to offspring might be expected to result in increased heterozygosity in the offspring. However, as the selection of offspring appears to be for maternal I-E8 and paternal D-E8 HLA-G allele combinations and against paternal I-E8 and maternal D-E8 allele combinations, the excess of the former heterozygote will be balanced by the deficiency of the latter. 15 As linkage disequilibrium occurs across the HLA locus and the results are proof that HLA-G and/or a HLA-G linked gene, cause the selection effects observed here. The C/T-93 polymorphism is a silent mutation while the IID-E8 polymorphism occurs in the 3' untranslated region (UTR) of the gene. These polymorphisms have been considered innocuous. However, the evidence indicates that the deletion polymorphism has a functional effect on the HLA-G gene. The 14bp sequence of I/D-E8 polymorphism is 20 largely conserved in primates and in the 3' UTR and/or in the last intron of HLA-B, C, J, A and E. 11 of the 14bp of the polymorphism is repeated in intron seven of the HLA-G gene. The core sequence "atttgt" is repeated one or more times in the 3' UTR of all class I genes but is absent in coding sequences. Examination of the secondary structure around the I/D-E8 polymorphism using the mfold programme (Zuker, 1994) shows that the 14n sequence is involved in a region of the 3'UTR having extensive 25 secondary structure and that the secondary structure is altered depending on the presence or absence of the 14n sequence (data not shown). Thus the presence or absence of the polymorphism may' affect the stability and/or alternative splicing of HLA-G mRNA through formation of alternative secondary structures. Examination of human preimpantation blastocysts showed that only 40% of the blastocvsts expressed 30 HLA-G and such expression was associated with an increased cleavage rate by comparison with embryos lacking the HLA-G transcript. Thus, polymorphisms affecting expression of HLA-G are likely to influence the rate of postzygotic prenatal loss by altering the cleavage rate in the embryo.

WO 99/43851 PCT/IE99/00012 29 Taken together, the results provide evidence that different HLA-G alleles and/or combinations thereof and/or variations in DNA in linkage disequilibrium with HLA-G in the foetus and/or one or both parents of the foetus are responsible for postzygotic prenatal loss which may manifest as miscarriage or undetectable early miscarriage which would manifest as unexplained infertility. 5 Identification of HLA-G as the Pre-eclampsia gene A cohort of pre-eclampsia primagravida trios (mother, father and first offspring) were identified in maternity hospitals, sampled and genotyped for the following polymorphisms in the HLA-G gene: C/T at codon 93 (C/T-93) (Table 9), AFT at codon 107 (A/T 107), C/A at codon 110 (C/A 110), and the insertion / deletion polymorphism in the non-translated region of the gene in exon 8 (I/D-E8) (Table 9). 10 Alleles transmitted and non transmitted to offspring were assigned (Table 10 and Table I11). HLA-G genotypes and haplotypes in pre-eclampsia trios were examined independently using transmission segregation analysis. pre-eclampsia trios were also compared to the cohort of control primagravida trios. A significant difference in C/T-93 allele frequency was observed between control and pre-eclampsia mothers (pl= 0.03, Table 12). A significant difference was also observed for the allele frequency of the 15 I/D-E8 polymorphism between control and pre-eclampsia fathers (p I= 0.02, Table 12). The frequency of the 93-E8 haplotypes differed significantly between control and pre-eclampsia mothers (p3=0.03), control and pre-eclampsia fathers (p3=0.008, Table 12) and also between control and pre-eclampsia offspring (p3=0.03, Table 12). The distribution of C/T-93 genotypes differed markedly between control and pre-eclampsia trios (Table 20 12), with a highly significant difference being observed between control and pre-eclampsia offspring (P2=0.001), between control and pre-eclampsia fathers (p2=0.02), and also between control and pre eclampsia mothers (p2= 0.05). These differences reflected a significant excess of C/T-93 heterozygotes over Hardy-Weinberg equilibrium expectations in both pre-eclampsia offspring (p l=0.0002) and pre eclampsia fathers (pl =0.021). A significant excess of I/D-E8 heterozygotes over Hardy-Weinberg 25 expectations was also observed in control mothers, pre-eclampsia offspring and in pre-eclampsia fathers (Table 12). Analysis of HLA-G haplotype sharing between offspring and mothers was also performed. No significant difference between pre-eclamptic cases and controls was observed for foetal-maternal sharing of HLA-G alleles or for sharing of the paternally transmitted HLA-G allele. There was no significant difference 30 observed for offspring sex between controls and pre-eclampsia cases.

WO 99/43851 PCT/IE99/00012 30 Comparison of the frequency of maternally and paternally transmitted alleles and haplotypes to control and pre-eclampsia offspring revealed significant differences (Table 12). In particular, these differences showed an excess of the maternally inherited T-93/I-E8 haplotype, and paternally inherited C-93/D-E8 haplotype and a deficiency of the maternally inherited C-93/D-E8 haplotype, and paternally inherited T 5 93/I-E8 haplotype in control offspring by comparison with pre-eclampsia offspring (Table 10). Furthermore, there was a significant difference between non-transmitted maternal I/D-E8 alleles and 93 E8 haplotypes showing that the maternal non-transmitted alleles are associated with pregnancy outcome. Only twelve out of fifty two control offspring inherited the maternal D-E8 allele from heterozygous (I/D E8) mothers (Table 3). By contrast, the maternal D-E8 allele was transmitted to twenty one of thirty six 10 pre-eclampsia offspring (Table 11). Taken together, these findings show a significant deficit of maternal D-E8 transmission to control offspring and a contrasting excess of maternal D-E8 transmission to pre eclampsia offspring. Further analysis within control and pre-eclampsia trios was achieved by comparison of maternally and paternally transmitted alleles and haplotypes to the offspring. In control offspring a highly significant 15 difference was observed between transmission of maternal and paternal C/T-93 alleles (Table 5. pl=0.009, calculated from Table 3). This reflected a deficit of maternal C-93 and paternal T-93 transmitted alleles (Table 3). A highly significant difference was also observed between transmission of maternal and paternal I/D-E8 alleles (Table 4, pl=0.000001, calculated from Table 3), showing a deficit of transmission of maternal D-E8 and paternal I-E8 alleles to the offspring (Table 3). A significant, but 20 contrasting difference between transmission of maternal and paternal alleles was present in pre-eclampsia offspring where an excess of transmission of maternal D-E8 and paternal I-E8 alleles, maternal C-93 and paternal T-93 was observed (Table 13, Table 10). A significant difference was present between maternally and paternally transmitted haplotypes in both control and pre-eclampsia offspring (P3=0.000003, Table 5 and P3= 0

.

00 5 respectively) showing a deficit in transmission of maternal C-93/D 25 E8 haplotypes and paternal T-93/I-E8 haplotypes to the control offspring and an excess in transmission of maternal C-93/D-E8 haplotypes and paternal T-93/I-E8 haplotypes to the pre-eclampsia offspring (Table 10). Maternally and paternally transmitted haplotypes to individual offspring are shown in Table 14. The maternally transmitted C-93/D-E8 haplotype and the paternally transmitted T-93/I-E8 haplotype 30 combination occurs in twenty two (35%) of the pre-eclampsia offspring but does not occur in any of the control offspring. By contrast, the paternal C-93/D-E8, maternal T-93/I-E8 haplotype combination occurs in both control and pre-eclampsia offspring but is in excess in the control offspring. This finding WO 99/43851 PCT/IE99/00012 31 provides evidence that combinations of HLA-G alleles / haplotypes in the foetus are causative of pre eclampsia in the mother. Taken together, the results provide evidence that different foetal HLA-G alleles and/or combinations thereof and/or variations in DNA in linkage disequilibrium with HLA-G in the foetus and/or one or both 5 parents of the foetus are responsible for pre-eclampsia Association of HLA-G with Recurrent Miscarriage A cohort of couples where the mother had three or more consecutive miscarriages were identified, sampled and genotyped for the C/T-93 and I/D-E8 polymorphisms in the HLA-G gene and 93-I-E8 haplotypes were assigned. RSA mothers and RSA fathers were compared to the cohort of control and pre 10 eclampsia primagravida trios for C/T-93 and I/D-E8 allele frequency, C/T-93 and I/D-E8 genotype distribution and 93-E8 haplotype frequency. The genotypes and haplotypes of the couples are shown in Table 16. A significant difference in C/T-93 allele frequency was observed between control and pre-eclampsia mothers (pl= 0.03) and control and RSA mothers (pl = 0.002) but not between pre-eclampsia and RSA 15 mothers. The frequency of the 93-E8 haplotypes differed significantly between control and pre-eclampsia mothers (p3=0.03), control and RSA mothers (p3=0.01), but not between pre-eclampsia and RSA mothers. The 93-E8 haplotypes of female and male mating partners were constructed. 50% of couples have the possibility of producing foetuses with the maternally transmitted C-93/D-E8 haplotype and the paternally 20 transmitted T-93/I-E8 haplotype combination. This compares with a 24% possibility in control couples and a 44% possibility in pre-eclampsia couples. The possibility of producing foetuses with the maternally transmitted D-E8 allele and the paternally transmitted I-E8 allele is 46% for control couples, 70% for pre-eclampsia couples and 85% for recurrent miscarriage couples. Maternally transmitted C-93/D-E8 haplotype and the paternally transmitted T-93/I-E8 haplotype combinations in foetuses are only found in 25 pre-eclampsia offspring. Recurrent miscarriage couple no. 15 must produce such a foetus. Taken together, the results provide evidence that foetal genotype associated with pre-eclampsia are also associated with miscarriage. Finally, mating couples where the female is homozygous for the T-93/I-E8 haplotype and the male has the C-93/D-E8 and T-93/I-E8 haplotypes were found in seven of sixty three pre-eclampsia cases and were absent in controls. Two of twenty of the recurrent miscarriage mating 30 couples (no. 10 and 13, Table 17) also had the same haplotype combinations. This provides evidence that partners where the female is homozygous for the T-93/I-E8 haplotype and the male has the C-93/D-E8 and T-93/I-E8 haplotypes are susceptible to pre-eclampsia and/or miscarriage. Taken together, the results WO 99/43851 PCT/IE99/00012 32 provide evidence that miscarriage and pre-eclampsia are closely related and that miscarriage is a severe expression of PE. One offspring of T-93/I-E8 and C-93/D-E8 haplotype combination in control trios was found where the mother was homozygous for T-93/I-E8 and father had the C-93/D-E8 haplotype and the C-93/I-E8 haplotype. One offspring of T-93/I-E8 and T-93/I-E8 haplotype combination in control trios 5 was also found where the mother was homozygous for T-93/I-E8 and father had the T-93/I-E8 haplotype and the C-93/I-E8 haplotype. This result shows that the non-transmitted male haplotype has a major influence on pregnancy outcome and indicates that sperm/semen contains a factor which influences susceptibility to pre-eclampsia and miscarriage. Taken together, the results provide evidence that different HLA-G alleles and/or combinations thereof 10 and/or variations in DNA in linkage disequilibrium with HLA-G in the foetus and/or one or both parents of the foetus are responsible for miscarriage. Induction of tolerance to HLA-G pre-eclampsialmiscarriage haplotypes in first pregnancy The results show that maternally transmitted D-E8 allele and the paternally transmitted I-E8 to offspring are linked to pregnancy outcome and that the maternally transmitted C-93/D-E8 haplotype and the 15 paternally transmitted T-93/I-E8 haplotype combination cause pre-eclampsia and miscarriage in primagravidas. We analysed HLA-G transmission in fifty three couples that have had two successful pregnancies without a history of miscarriage or PE. The possibility of producing foetuses with the maternally transmitted D-E8 allele and the paternally transmitted I-E8 allele was 15% for this cohort of normal couples. The results thus show a clear 20 correlation between successful pregnancy outcome and the probability of possibilities of producing foetuses with the maternally transmitted D-E8 and the paternally transmitted I-E8 HLA-G alleles. Recurrent miscarriage - 85%, pre-eclampsia - 70%, first pregnancy normal - 46%, first and second pregnancy normal 15%. The maternally transmitted C-93/D-E8 haplotype and the paternally transmitted T-93/I-E8 haplotype 25 combination cause pre-eclampsia and miscarriage in primagravidas. We determined if primagravida normal pregnancy induced tolerance to the foetal pre-eclampsia / miscarriage haplotype combinations when they occur in pregnancy two. There were no first offspring detected bearing the maternally transmitted C-93/D-E8 haplotype and the paternally transmitted T-93/I-E8 haplotype combination. There were five second offspring bearing the maternally transmitted C-93/D-E8 haplotype and the paternally 30 transmitted T-93/I-E8 haplotype combination (Table 18). We also determined if primagravida pre cclampsia pregnancy induced tolerance to the foetal pre-eclampsia / miscarriage haplotype combinations when they occur in pregnancy two. We analysed nine families where the mother suffered pre-eclampsia in WO 99/43851 PCT/IE99/00012 33 her first pregnancy and had a normal second pregnancy. There were three second offspring bearing the maternally transmitted C-93/D-E8 haplotype and the paternally transmitted T-93/I-E8 haplotype combination in the absence of pre-eclampsia even though the same combination caused pre-eclampsia in the first pregnancy (Table 19). 5 This proves that tolerance to the paternal antigens in the foetus is induced in the first pregnancy. More specifically, tolerance to the problematic maternally transmitted C-93/D-E8 haplotype and the paternally transmitted T-93/I-E8 haplotype was induced in first pregnancy. Thus exposure to HLA-G alleles and/ or combinations thereof and/or paternal antigens presented to the maternal immune system by HLA-G in the first pregnancy induces tolerance to the pre-eclampsia / miscarriage haplotype combination so that these 10 problematic haplotypes can occur in the second pregnancy without associated pre-eclampsia or miscarriage. The genetic linkage, association and correlation approaches used in the large number of subject cohorts provides proof that HLA-G is a susceptibility gene for normal pregnancy, pre-eclampsia and miscarriage. As pre-eclampsia is associated with intra uterine growth retardation, the HLA-G gene is also a 15 susceptibility gene for intra uterine growth retardation. As miscarriage frequently occurs so early that it is not detected, the HLA-G is also a susceptibility gene for miscarriage related unexplained infertility. Exposure to foetal antigens including HLA-G in the first pregnancy has been shown to induce tolerance to antigens that are problematic in first pregnancy and thus provide a means for potential treatment of pre eclampsia, miscarriage, intrauterine growth retardation, miscarriage related infertility and autoimmune 20 disease and provide a means to induce tolerance to foreign antigens for purposes such as transplantation of foreign tissue. The HLA-G I/D-E8 polymorphism has been investigated previously in pre-eclampsia and no detectable relationship was observed between susceptibility to pre-eclampsia and HLA-G (24). This result is consistent with the results reported here in that an HLA-G effect is not seen when the I/D-E8 25 polymorphism is independently analysed by association studies alone. The results presented here show that genetic screening of maternal and/or paternal and/or foetal DNA is of value for predictive testing of susceptibility to pre-eclampsia, eclampsia, intrauterine growth retardation, miscarriage and miscarriage-related infertility. Furthermore, transmission of HLA-G alleles to offspring in normal pregnancy differs from the normal expectation. Therefore, the results presented 30 here show that genetic screening of maternal and/or paternal foetal DNA is of value for predictive testing of susceptibility to normal pregnancy.

WO 99/43851 PCT/IE99/00012 34 Preferably, foetal nucleic acid is isolated from any material containing nucleic acid of foetal origin in the mother such as amniotic fluid, maternal blood or chorionic villus. Furthermore, the results show that genetic screening of parents will also be of value for predictive testing of susceptibility to pre-eclampsia. Although the function of HLA-G apart from its role in regulating NK cell activity and induction of IL-3 5 and IL-I beta in PBMCs is as yet poorly understood. HLA-G is an excellent candidate for pre-eclampsia since HLA-G is considered to play a key role in foetal-maternal immune interactions. The C/T-93 HLA G allele associated with pre-eclampsia is a silent polymorphism but its effect on HLA-G mRNA stability or splicing is unknown. It is likely that this polymorphism and/or variations linked to this polymorphism play a causative role in pre-eclampsia. 10 In this work, we have demonstrated a difference between pre-eclamptic and control offspring with respect to sharing of the paternal C/T-93 allele between the offspring and their mothers. This result indicates that pre-eclampsia may arise due to the presence of a HLA-G haplotype in the foetus that has not previously been encountered by the mother. Since HLA-G is in tight linkage disequilibrium with the HLA locus, it is likely that the paternal HLA-G itself and/or the presence of an extended paternal HLA haplotype in the 15 foetus that has not previously been encountered by the mother causes pre-eclampsia. Furthermore, since HLA-G is in tight linkage disequilibrium with the HLA locus, determination of the extended paternal HLA haplotype segregating in the foetus and comparison of the haplotype with the maternal HLA haplotypes will allow diagnosis of susceptibility to pre-eclampsia. While other associations have been reported between pre-eclampsia and the maternal genotype, the results 20 reported here are much more consistent with epidemiological studies on pre-eclampsia. In particular, the association between a foetal HLA-G genotype is consistent with the observation that a) pre-eclampsia is more common in sisters than in the normal population, b) pre-eclampsia is discordant in identical twin mothers and c) pre-eclampsia can occur with a change of male partner. Pre-eclampsia is rare in second or later pregnancies indicating that initial exposure to functional HLA-G prevents pre-eclampsia. In 25 addition, HLA-G is now known to induce synthesis IL-3 and IL-1 beta and down-regulate tumour necrosis factor-alpha production. These observations coupled with the results presented here indicates that HLA-G protein, IL-3 and/or IL-1 beta or inhibitors of tumour necrosis factor-alpha will be useful for treatment of intrauterine growth retardation, pre-eclampsia, miscarriage and miscarriage-related infertility. 30 The HLA-G genotype associated with pre-eclampsia and miscarriage is likely to have one of a small number of consequences: i) it could result in reduced expression of HLA-G which would be reflected as decreased levels of cellular and/or soluble HLA-G (the HLA-G primary transcript is alternatively spliced to yield several WO 99/43851 PCT/IE99/00012 35 different mRNAs. One of these alternatively spliced forms includes intron 4. The open reading frame in this mRNA continues into intron 4, terminating 21 amino acids after the alpha3 domain - encoded by exon 4. Thus, the transmembrane region encoded by exon 5 and the cytoplasmic tail of HLA-G is excluded. The resultant protein is hence soluble). Thus measuring of cellular and/or soluble HLA-G levels and 5 comparing these levels with the normal observed levels would allow one to diagnose susceptibility to pre eclampsia and miscarriage; ii) the HLA-G genotypes associated with pre-eclampsia may lead to variations in HLA-G mRNA and/or HLA-G protein which in turn could be detected by characterisation of HLA-G mRNAand/or protein. Thus. characterisation of HLA-G protein in pregnant females, foetuses and/or respective male mating 10 partner would allow one to diagnose susceptibility to pre-eclampsia and miscarriage; iii) expression of the HLA-G protein leads directly or indirectly to alterations in the levels of certain molecules such as IL-3, IL-I beta and/or tumour necrosis factor alpha. The HLA-G genotypes associated with pre-eclampsia may result in changed expression of such molecules. Thus measuring of the levels of such molecules and comparing these levels with the normal observed levels would allow one to diagnose 15 susceptibility to pre-eclampsia and miscarriage; iv) The HLA-G genotypes associated with pre-eclampsia may result in decreased expression of HLA-G. This in turn would lead to increased lysis of trophoblasts by NK cells. Thus measuring of the levels of trophoblast specific marker and comparing these levels with the normal observed levels would allow one to diagnose susceptibility to pre-eclampsia and miscarriage. 20 The HLA-G variants associated with pre-eclampsia and miscarriage and normal pregnancy are likely to have one of a small number of consequences: i) a variant could result in altered expression of HLA-G splice forms and levels thereof which would be reflected as altered levels of HLA-G splice forms including soluble HLA-G in the serum.. Thus measuring of size, levels and/or splice forms of HLA-G mRNA and/or protein including soluble HLA-G 25 levels and comparing these levels with the normal observed levels would allow one to diagnose susceptibility to pre-eclampsia and miscarriage; ii) the HLA-G variants associated with pre-eclampsia and miscarriage may result in variations in HLA G protein which in turn could be detected by protein characterisation of cellular and/or soluble HLA-G. Thus characterisation of HLA-G protein in pregnant females, foetuses and/or respective male mating 30 partner would allow one to diagnose susceptibility to pre-eclampsia and miscarriage; iii) expression of the HLA-G protein leads directly or indirectly to alterations in the levels of certain molecules such as IL-3, IL-1 beta and/or tumour necrosis factor alpha. The HLA-G variants associated with pre-eclampsia may result in changed expression of such molecules. Thus measuring of the levels of such molecules and comparing these levels with the normal observed levels would allow one to diagnose 35 susceptibility to pre-eclampsia and miscarriage; WO 99/43851 PCT/IE99/00012 36 iv) the HLA-G variants associated with pre-eclampsia and miscarriage may result in increased or decreased expression of paternal and/or maternal HLA-G. This in turn would lead to increased lysis of trophoblasts by NK cells and/or cytotoxic T cells. Thus measuring of the levels of trophoblast specific marker and comparing these levels with the normal observed levels would allow one to diagnose 5 susceptibility to pre-eclampsia. v) the HLA-G variants associated with pre-eclampsia and miscarriage may result in increased or decreased cell cleavage rates in the embryo. Thus measuring of cell cleavage rates in cells expressing one or more HLA-G variants and any combinations thereof would allow one to diagnose susceptibility to pre eclampsia and miscarriage. 10 The results show that HLA-G polymorphism plays a major role in predisposition to normal, pre eclampsia and miscarriage outcome in pregnancy and that haplotypic combinations and parent-of-origin effects mediate the influence ofHLA-G polymorphism on these outcomes. The results show a strong association between foetal and paternal HLA-G genotypes and PE, and analysis of heterozygote v. homozygote mating outcomes indicate that transmission of HLA-G alleles to the pre-eclampsia offspring, 15 but not to control offspring, is distorted. The results provide evidence for linkage of the maternal HLA-G I-E8 allele to normal pregnancy outcome in primagravidas and the observed deficit of maternal D-E8 allele and C-93/D-E8 haplotype transmission to control offspring indicates selection for foetuses on the basis of HLA-G genotype in primagravida normal pregnancies. The transmission distortion of the maternal D-E8 allele to the foetus indicates that the effect seen in normal primagravidas is mediated by 20 the maternal allele acting primarily in the foetus. Thus, the maternal HLA-G imparts a protective effect to the foetus which enhances normal pregnancy outcome. This finding indicates that maternal selection of the HLA-G I-E8 and other protective HLA-G alleles occurs in normal pregnancy. By contrast, the maternal D-E8 allele was prevalent in heterozygous pre-eclampsia offspring, indicating that susceptibility to pre eclampsia partly arises through the lack of a protective maternal HLA-G allele in the foetus. The chi 25 squared contingency table analysis agreed with the log linear model analysis in that the C-93 allele was over-represented in pre-eclampsia offspring and a bias towards maternal inheritance of I-E8 was present in controls. Furthermore, the log linear model showed that the foetal C-93 allele is under-represented in control offspring with a strong bias towards paternal inheritance of the allele. This indicates that the paternal C-93 allele also imparts a protective or alternatively does not introduce a problematic effect to 30 the foetus which improves the prospect of a normal pregnancy outcome. These results are in good agreement with the findings observed when maternal and paternal haplotype combinations were constructed for individual control and pre-eclampsia offspring where more than one third of the pre cclampsia cases had a maternal C-93/D-E8 paternal T-93/I-E8 haplotype combination that was absent in the controls. Taken together. the data indicate a strong association between both maternal and paternal 35 HLA-G alleles acting through the foetus and normal pregnancy outcome and indicate that pre-eclampsia WO 99/43851 37 PCT/IE99/00012 arises through the absence of protective maternal and protective or problematic paternal HLA-G alleles in the foetus. Furthermore, considering that there are likely to be several HLA-G alleles with functional differences, and as more than one third of pre-eclampsia cases can be accounted for by a particular maternal / paternal haplotype combination, the results show that the magnitude of the effect of HLA-G in 5 normal and pre-eclampsia pregnancies is large. Alternatively, a protective foetal-maternal HLA-G allele is likely to arise through the transmission of a dominant maternal allele to the foetus which is recognised as self by the maternal immune system. A protective foetal-paternal allele is likely to arise through cross recognition of the paternal allele as self by the maternal immune system. A problematic foetal-paternal allele is likely to arise through cross 10 recognition of the paternal allele as non- self by the maternal immune system. The results indicate maternal education of the lymphocyte repertoire for maternal HLA-G during and/or prior to pregnancy and for paternal HLA-G during pregnancy. The results also indicate certain paternal HLA-G alleles are compatible with the maternal immune system while others are less compatible. Combinations of less compatible/incompatible paternal HLA-G alleles with maternal alleles which do not protect against the 15 paternal alleles are likely to cause susceptibility to pre-eclampsia and miscarriage. The fact that second offspring of primagravida normal and pre-eclampsia mothers have the maternal C 93/D-E8 paternal T-93/I-E8 genotype in the absence of pre-eclampsia in the second pregnancy is evidence that maternal education for foetal-paternal antigens occurs during the first pregnancy and that this education is mediated by HLA-G. It is clear from this work that the polymorphisms analysed and/or 20 closely linked polymorphisms in HLA-G or flanking HLA genes contribute directly to enhancing normal pregnancy outcome and to susceptibility to pre-eclampsia and miscarriage. One likely explanation may be that the polymorphisms reported here destabilise HLA-G mRNA and/or alter the splicing pattern and/or glycosylation pattern of HLA-G. The presence or absence of polymorphism is likely to effect the stability and/or alternative splicing of HLA-G mRNA. Thus a protective foetal-maternal HLA-G allele is likely 25 to arise through the transmission of a maternal allele to the foetus which may or may not be expressed in the embryo. A protective foetal-paternal allele is likely to arise through the transmission of a paternal allele to the foetus which may or may not be expressed in the embryo. At least twelve different haplotypes have been described for the HLA-G gene. Considering the link observed between HLA-G and recurrent miscarriage, it is likely that the combination of HLA-G alleles in 30 the early foetus and/or the combination of the HLA-G alleles in the mother has serious effects on the outcome of implantation in general and is likely to account for cases of unexplained or idiopathic infertility as well as miscarriage. The previously reported link between pre-eclampsia and intra-uterine WO 99/43851 PCT/IE99/00012 38 growth retardation indicates that the latter is also likely to be linked to parent of origins effects of foetal HLA-G alleles and indicate that maternal HLA-G alleles also play a role in the foetal growth outcome. HLA-G is capable of protecting otherwise susceptible target cells from natural killer cell mediated lysis through its interaction with inhibitory receptors on natural killer cells. HLA-G is also capable of 5 stimulating an HLA-G restricted lymphocyte response, HLA-G molecules can serve as target molecules in lytic reactions with lymphocytes, and HLA-G is involved in education of the lymphocytic repertoire. Thus, pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation is likely to arise through a mechanism involving blood mononuclear cells such as natural killer cells and cytotoxic T lymphocytes whereby interaction between the female mating partner's T cells and foetal 10 antigens is compromised by comparison with normal pregnancy. Thus, compromised interaction leading to the lack of tolerance leads to cell killing. Compromised interaction also can lead to lack of stimulation of cells expressing HLA-G molecules and/or lack of stimulation of cells interacting with cells expressing HLA-G molecules. The fact that the maternal C-93/D-E8 paternal T-93/I-E8 HLA-G genotype can occur in the second pregnancy of a primagravida pre-eclampsia case without pre-eclampsia indicates that 15 education mediated by foetal HLA-G to foetal antigens occurs in the first pregnancy of such mothers which overcomes compromised interactions in second and subsequent pregnancies. The fact that a deficit of maternal C-93/D-E8 genotypes and an excess of T-93/I-E8 genotypes are transmitted to control offspring but not to pre-eclampsia offspring indicates that selection for foetuses that express antigens for which the mother is educated occurs in normal pregnancy. The fact that pre-eclampsia rarely occurs in a 20 second pregnancy when the first pregnancy has been normal indicates that induction of education to foetal antigens mediated by HLA-G also occurs during and prior to the first pregnancy in normal mothers and that pre-eclampsia, miscarriage, miscarriage-related infertility and intra-uterine growth retardation arises from lack of education and/or inadequate induction of education to the foetal antigens in the female mating partner during and/or prior to pregnancy. Lack of and/or compromised induction of education to 25 paternal antigens such as HLA-G in the foetus and/or a defective HLA-G interaction with natural killer cells could lead to lysis of trophoblasts and/or lack of stimulation of trophoblasts leading to reduced trophoblast function and/or lack of stimulation of cells interacting with trophoblasts. Thus, HLA-G linked conditions such as pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation are likely to arise through blood mononuclear cell mediated killing of accessible foetal tissues 30 such as trophoblasts and/or lack of stimulation of trophoblastic cells because of compromised HLA-G interaction with blood mononuclear cells trophoblasts and/or lack of stimulation of blood mononuclear cells because of compromised HLA-G interaction with trophoblastic cells. Since major histocompatibility (MHC) molecules like HLA-G interact with blood mononuclear cells including cytotoxic T cells and natural killer cells, there is likely to be abnormal interaction between maternal blood mononuclear cells 35 and foetal cells presenting MHC / MHC-antigen complexes and/or MHC / MHC-antigen complexes WO 99/43851 PCT/IE99/00012 39 secreted from foetal cells in pre-eclampsia, miscarriage and intra-uterine growth retardation by comparison with normal pregnancies. Thus, the blood mononuclear cell response and/or the trophoblast response to such an interaction is likely to be abnormal in the HLA-G associated disorders. In particular, the cytokine response produced as a result of such an interaction is likely to be abnormal by comparison 5 with the normal situation. Thus, diagnosis of susceptibility to pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation and prediction of pregnancy outcomes may be achieved by direct and indirect measurement of the education in the female mating partner to foetal antigens and/or direct and indirect measurement of the interaction between blood mononuclear cells and HLA-G and/or HLA-G 10 expressing cells. Furthermore, direct and indirect measurement of the education in the female mating partner to foetal antigens and/or direct and indirect measurement of the natural killer cell activity in the female mating partner to HLA-G expressing cells and/or direct and indirect measurement of the interaction between blood mononuclear cells and HLA-G and/or HLA-G expressing cells offers a means to monitor the course of pregnancy. 15 Induction of education to foetal antigens in the female mating partner by treatment with HLA-G and/or peptides known to bind to HLA-G constitutes a therapeutic means for prevention and/or treatment of pre eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation and any other HLA-G related disorders. The finding that combinations of HLA-G variants in the foetus are closely associated with pre-eclampsia 20 and miscarriage coupled to the fact that HLA-G interacts with blood mononuclear cells offers a further means to prevent and/or treat pre-eclampsia miscarriage, miscarriage-related infertility and intrauterine growth retardation by inhibition and/or alteration of the interaction of HLA-G and/or HLA-G variants with blood mononuclear cells. This may be achieved by any one or combination of approaches including treatment with one or more molecules which recognise HLA-G and/or variants of HLA-G and/or one or 25 more HLA-G receptors on blood mononuclear cells, and/or inactivation of the HLA-G gene and/or HLA G gene variants and/or one or more HLA-G receptors on blood mononuclear cells. For example, this would be achieved by treatment with HLA-G specific and/or HLA-G receptor specific antibodies which interfere with HLA-G - blood mononuclear cell interaction and/or treatment with one or more enzymes which recognise and alter HLA-G and/or HLA-G receptors on blood mononuclear cells and/or treatment 30 with one or more peptides which bind to HLA-G and/or HLA-G receptors on blood mononuclear cells. Alternatively, inhibition of the interaction of one or more HLA-G variants with blood mononuclear cells may be achieved by inactivating the HLA-G gene or HLA-G gene variant and/or one or more HLA-G receptors on blood mononuclear cells. This may be achieved through the use of one or more gene WO 99/43851 PCT/IE99/00012 40 inactivating approaches such as treatment with one or more nucleic acid antisense and/or ribozyme molecules which inhibit expression of the HLA-G gene and/or HLA-G gene variant and/or one or more HLA-G receptors on blood mononuclear cells. This may be also be achieved by inactivating the HLA-G gene in one or both partners of a mating couple somatically or in the germ line through the use of gene 5 therapy approaches whereby inhibitory nucleic acid based molecules such as antisense, and/or ribozyme are introduced into an individual. This may be also be achieved by inactivating the HLA-G gene in one or both partners of a mating couple somatically or in the germ line through the introduction of all or part of a HLA-G gene in such a way that it recombines with the endogenous HLA-G in the cell and inactivates it. Alternatively, the HLA-G gene and/or variants of the HLA-G gene and/or any of it's receptors may be 10 employed in gene therapy methods in order to increase the amount of expression products of such genes in an individual allowing compensation of any deficiency of HLA-G and/or it's receptors in an individual. Thus, alteration of the interaction of HLA-G and/or HLA-G variants with blood mononuclear cells may be achieved by introduction of one or more HLA-G gene variants into somatic cells and/or into the germline of one of both partners of a mating couple or into the fertilised egg or cells arising from the 15 fertilised egg prior to implantation. This is of particular importance for increased fertility for animal breeding purposes. For example, introduction of one or more HLA-G gene variants into the germline of one of both partner of a mating couple or into the fertilised egg or cells arising from the fertilised egg where the HLA-G variant is compatible with the prospective mother offers a means to improve fertility and pregnancy outcome arising from any incompatibility between foetal HLA-G and maternal cells in the 20 mother. HLA-G binds a diverse but limited array of peptides in a manner similar to that found for classical class I molecules and it has been reported that HLA-G is expressed in the human thvmus raising the possibility that maternal unresponsiveness to HLA-G expressing foetal tissues may be shaped in the thymus by central presentation of this MHC molecule on the medullary epithelium (Crisa et al. 1997) HLA-G is 25 known to be capable of stimulating a HLA-G restricted cytotoxic T lymphocyte response and HLA-G molecules can serve as target molecules in lytic reaction with cytotoxic T lymphocytes and HLA-G expressed internally in vivo in transgenic animals is involved in education of the lymphocytic repertoire (Schmidt et al., 1997). The invention shows that the induction of education to foetal antigens occurs during pregnancy and arises from exposure of the mother to foetal antigens during pregnancy. HLA-G 30 allele combinations that were unacceptable in first pregnancy and/or were linked to pre-eclampsia were acceptable in second pregnancy without any associated pregnancy complications. Thus, induction of education to foetal antigens is likely to arise from a process involving HLA-G. Thus the invention offers a means of inducing education including tolerance to HLA-G and/or peptides bound to HLA-G in an individual through mimicking the exposure to foetal antigens that occurs during pregnancy. Thus 35 treatment of an individual with HLA-G and or / peptides known to bind to HLA-G constitutes a means to WO 99/43851 PCT/IE99/00012 41 induce education in an individual to antigens. In particular, this offers a means to induce tolerance to antigens that cause susceptibility to pre-eclampsia, susceptibility to miscarriage, autoimmune disease and transplant rejection. In normal pregnancy, direct and indirect alteration of the level and/or activity of molecules arising from 5 the interaction of HLA-G expressing foetal cells with blood mononuclear cells such as lymphocytes and natural killer cells permit pregnancy to progress properly. In pre-eclampsia, miscarriage, miscarriage related infertility and intrauterine growth retardation and any other HLA-G related disorders, the alteration of the level and/or activity of molecules arising from the interaction of HLA-G and/or HLA-G expressing foetal cells with blood mononuclear cells such as lymphocytes and natural killer cells is likely 10 to be compromised by comparison with that occurring during normal pregnancy. Thus, mimicry of the alteration of the level and/or activity of one or more molecules arising from the interaction of HLA-G and/or HLA-G expressing foetal cells with blood mononuclear cells in an individual constitutes a therapeutic means for prevention and/or treatment of pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation and any other HLA-G related disorders. 15 The deficit of maternal HLA-G C-93/D-E8 genotypes and the excess of T-93/I-E8 genotypes transmitted to control offspring but not to pre-eclampsia offspring implies selection of foetuses in normal pregnancy dependent on HLA-G genotype. For fertility purposes, and especially in vitro fertilisation and embryo transfer in animals, selection of one or both mating partners, sperm and/or egg donors and/or embryo recipients based on male and/or female HLA-G and/or HLA-G homologue genotypes and/or serotypes 20 and/or activity associated with a successful normal first pregnancy with a specific mating partner offers a means to improve fertility and the success rate of in vitro fertilisation and embryo transfer in animals and improve pregnancy outcome. Since HLA-G protects trophoblasts from blood mononuclear cell mediated killing, direct and indirect measurement of measurable substances which originate from trophoblast cell killing should allow 25 diagnosis of susceptibility to pre-eclampsia, miscarriage, intra-uterine growth retardation, and monitoring of pregnancy for normal progress, and progress towards pre-eclampsia, miscarriage and intra-uterine growth retardation in humans and animals. More specifically, the interaction between MHC molecules such as HLA-G and blood mononuclear cells is known to directly and indirectly alter the synthesis and levels of several cytokines. Similarly, trophoblasts are known to synthesise and secrete several cytokines. 30 In particular, the altered regulation of some of these cytokines would be expected to compromise the foetal - maternal immune interaction and could be manifest as pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or miscarriage and/or miscarriage-related infertility. For example, the interaction of HLA-G expressing cells with blood mononuclear cells increases the amount of interleukin- WO 99/43851 PCT/IE99/00012 42 3 (IL-3) and interleukin-1 beta (IL-1 beta) and decreases the amount of tumour necrosis factor-alpha (TNF-alpha) release from the blood mononuclear cells. Trophoblasts are known to produce the immunosuppressive cytokine interleukin 10 - a cytokine that potently inhibits alloresponses in mixed lymphocyte reactions. Trophoblasts are also known to produce interleukin 2, a cytokine that both protects 5 the foetus and in involved in activation of maternal killer cells to protect against invading trophoblasts, interleukin 4 and its receptor, which play a role in regulation of umbilical blood flow mediated through the induction of cyclooxygenase-2, indicating a role for interleukin 4 in vascular tone and blood flow modulation during pregnancy, interleukin 6, which is likely to play a role in tissue remodelling associated with placentation. 10 Since the indications are that pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation arise through a HLA-G mediated mechanism, there are several obvious methods for screening for agents which can potentially be used as diagnostic indicators and therapeutic agents. Screening of gene expression profiles using DNA probe arrays allows identification of genes expressed in HLA-G expressing cells and in blood mononuclear cells and genes whose expression changes as a result 15 of HLA-G interaction with blood mononuclear cells. Comparison of the gene expression profile in HLA G expressing cells and/or blood mononuclear cells and/or HLA-G expressing cells interacting with blood mononuclear cells and/or in blood mononuclear cells interacting with HLA-G allows identification of agents which can potentially be used as diagnostic indicators and therapeutic agents for pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation. 20 HLA-G function and HLA-G expression can be measured. Thus screening for agents which alter the expression and/or function and/or which mimic the function of HLA-G provide a method for screening for potential pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation therapeutic agents. The words "comprises/comprising" and the words "having/including" when used herein with reference to 25 the present invention are used to specify the presence of stated features, integers, steps or components but does not preclude the presence or addition of one or more other features, integers, steps, components or groups thereof WO 99/43851 PCT/IE99/00012 43 References Amrngrimsson. R., Hayward, C., Nadaud, S., Baldursdottir, A., Walker, J.J., Liston, W.A., Bjarnadottir, R.I.. Brock. D.J., Geirsson, R.T., Connor, J.M., Soubrier, F. Evidence for a familial pregnancy-induced hypertension locus in the eNOS-gene region. Am. J. Hum. Genet 61,354-362 (1997) 5 Bellingard, V., Hedon, B.. Eliaou, J., Seignalet, J., Clot, J. and Viala, J. Immunogenetic study of couples with recurrent spontaneous abortions. Eur. J. Obstet. Gynecol. Reprod. Biol., 60, 53-60 (1995). Caudle, M., Rote, N., Scott, J., DeWitt, C. and Barney, M. Histocompatibility in couples with recurrent spontaneous abortion and normal fertility. Fertil. Steril.. 39, 793-198 (1983). Christiansen, O., Mohapeloa, H., Steffensen, R. and Jersild, C. HLA-C and -Bw typing of couples with 10 unexplained recurrent miscarriages. J. Reprod. Immunol., 37, 63-77 (1997). Colbern. G.T.. Chiang, M.H. & Main, E.K. Expression of the nonclassic histocompatibility antigen HLA-G by pre-eclampsia placenta. Am. J. Obstet. Gynecol. 170:1244-1250 (1994). Cooper, D.W.. Hill, J.A., Chesley, L.C. & Iverson Bryans, C. Genetic control of susceptibility to eclampsia and miscarriage. Br. J. Obstet. Gynaecol., 95:644-653 (1988). 15 Cooper, D.W., Brennecke, S.P. & Winton, A.N. Genetics of Pre-eclampsia. Hypertension in Pregnancy, 12(1), 1-23 (1993). Crisa, L., McMaster, M., Ishii, J., Fisher, S. and Salomon, D. Identification of a thymic epithelial cell subset sharing expression of the class Ib HLA-G molecule with foetal trophoblasts. J. Exp. Med. 186, 289-298 (1997) 20 Hara, N., Fuji. T. Yamashita, T., Kozuma, S., Okai, T. & Taketani, Y. Altered expression of human leukocvte antigen (HLA-G) on extravillous trophoblasts in pre-eclampsia: immunohistological demonstration with anti-HLA-G specific antibody "87G" and anti-cvtokeratin antibody "CAM5.2". Am J Reprod Immunol 36:349-358 (1996). Harrison. G.A., Humphrey, K.E., Jakobsen, I.B. & Cooper, D.W. A 14bp deletion polymorphism in the 25 HLA-G gene. Hum. Mol. Genet. 2:2200 (1993). Harrison. G.A., Humphrey, K.E., Jones, N., Badenhop, R., Gua, G., Elakis, G., Kaye, J.A., Turner, R.J., Grehan. M.. Wilton, A.N., Brennecke S.P. & Cooper, D.W. A genomewide linkage study of pre- WO 99/43851 PCT/IE99/00012 44 eclampsia/eclampsia reveals evidence for a candidate region of 4q. Am.J.Hum.Genet., 60, 1158-1167 (1997). Hayward. C., Livingstone, J., Holloway, S., Liston, W.A. & Brock, D.H. An exclusion map for pre eclampsia: assuming autosomal recessive inheritance. Am. J. Hum. Genet. 50, 749-757 (1992). 5 Humphrey. K.E.. Harrison, G.A.. Cooper, D.W., Wilton, A.N., Brennecke S.P. & Trudinger, B.J. HLA G deletion polymorphism and pre-eclampsialeclampsia. Br. J. Obstet. Gynaecol. 102:707-710 (1995). Jin, K., Ho, H., Speed, T. and Gill, T. Reproductive failure and the major histocompatibility complex. Am. J. Hum. Genet., 56, 1456-1467 (1995). Jurisicova, A., Casper, R., MacLusky, N., Mills, G. and Librach, C. HLA-G expression during 10 preimplantation human embryo development. PrC NatlAcadSci U.SA. 93, 161-165 (1996). Karhukorpi, J.. Laitinen. T. & Tiilikainen, A.S. HLA-G polymorphism in Finnish couples with recurrent spontaneous miscarriage. Br. J. Obstet. Gynaecol. 104:1212-1214 (1997). Lee, N., Malacko, A.R., Ishitani, A., Chen, M., Bajorath, J., Marquardt, H. & Geraghty, D.E. The membrane bound and soluble forms of HLA-G bind identical sets of endogenous peptides but differ with 15 respect to TAP association. Immunity, 3, 591-600 (1995). Lie, R.T., Rasmussen, S., Brunborg, H., Gjessing, H.K., Lie-Neilsen, E. & Irgens, L.M. Fetal and maternal contributions to risk of pre-eclampsia: population based study. Br. Med. J. 316, 1343-47 (1998). Loke, Y.W. & King, A. Immunology of human placental implantation: clinical implications of our 20 current understanding. Mol. Med. Today 3(4): 153-159 (1997). Maejima, M., Fuji, T., Kozuma, S., Okai, T., Shibata, Y. & Taketani, Y. Presence of HLA-G expressing cells modulates the ability of peripheral blood mononuclear cells to release cytokines. Am. J. Reprod. Immunol., 38, 79-82 (1997). Ober. C. et al. Human leukocyte antigen matching and fetal loss: results of a 10 year prospective study. 25 Hum. Reprod., 13, 33-38 (1998). Sasaki. T., Yamada, H., Kato, E., Sudo, S., Kishida, T., Sasaki, T.. Nishigaki, F. and Fujimoto, S. Increased frequency of HLA-DR4 allele in women with unexplained recurrent spontaneous abortions, detected by the method of PCR-SSCP. J. Reprod. Immunol., 32, 273-279 (1997).

WO 99/43851 PCT/IE99/00012 45 Sbracia. M., Mastrone, M., Scarpelline, F. and Grasso J. Influence of histocompatibility antigens in recurrent spontaneous abortion couples and on their reproductive performances. Am. J. Reprod. Immunol., 35, 85-92 (1996). Schmidt. C.M., Garrett, E. & Orr, H.T. Cytotoxic T lymphocyte recognition of HLA-G in mice. Hum 5 Immunol., 55, 127-139 (1997). Spinillo. A., Capuzzo, E, Piazzi, G., Nicola, S., Colonna, L., lasci, A. Maternal high-risk factors and severity of growth deficit in small gestational age infants. Early Hum Dev. 38, 35-43 (1994). Timbers. K. and Feinberg, R. Recurrent Pregnancy Loss: A Review. Nurse Practitioner Forum. 8. 77-88 (1997). 10 Van der Ven, K. & Ober, C., HLA-G Polymorphism in African Americans, J. Immunol., 153, 5628 5633 (1995). Vaughan, P. & McCarthy, T.V. A novel process for mutation detection using uracil DNA-glycosylase. Nuc. Acids Res., 26, 810-815 (1998). Weinberg, C.R., Wilcox, A.J. & Lie, R.T. A log-linear approach to case-parent-triad data: assessing 15 effects of disease genes that act either directly or through maternal effects and that may be subject to parentalimprinting. Am J Hum Genet 62, 969-978 (1998). Wilcox ALJ, Weinberg CR, O'Connor JF, Baird DD, Schlatterer JP, Canfield RE, Armstrong EG, Nisula BCIncidence of early loss of pregnancy. N. Eng. J. Med, 319, 189-194 (1988). Roman, E. Fetal loss rates and their relation pregnancy order. J. Epidemiol., 38, 29-43 (1984). 20 Winton, A.N., Cooper, D.W., Brennecke, S.P., Bishop, S.M. & Marshall, P. Absence of close linkage between maternal genes for susceptibility to pre-eclampsia and HLA DR-beta. Lancet, 336, 653-657 (1990). Zuker, M. Prediction of RNA secondary structure by energy minimization, in computer analysis of sequence data A.M. Griffin and H.G. Griffin eds. Methods in Molecular Biology, Humana Press Inc., 25 267-294 (1994).

WO 99/43851 PCT/IE99/00012 46 Table 1: Genotype and allele distribution of the HLA-G C/T-93 (C1488T) and I/D-E8 (exon deletion) polvmorphisms in pre-eclamptic and control offsrping. Polymorphism cenotype control preeclamptic p offspring offspring Aa 22 44 HLA-G C/T-93 aa 8 3 6.11 <0.02 polymorphism AA 18 7 11.01 <0.001 Aa 30 31 HLA-G CIT-93 aa 12 8 0.69 0.3 polymorphism _____ i AA 13 12 0.06 <0.8 Table 2: Genotype and allele distribution of the HLA-G C/T-93 and I/D-E8 polymorphisms in primigravida trios. C/T-93 Frequency n C/C (%) C/T (%) T/T (%) C/T Mothers 90 19(21.1) 50(55.6) 21(23.3) 0.49/0.51 Fathers 90 34(37.8) 39(43.3) 17(18.9) 0.59/0.41 Offspring 90 24(26.7) 41(45.5) 25(27.8) 0.49/0.51 - Male 46 13(28.3) 20(43.4) 13(28.3) 0.5/0.5 -Female 44 11(25.0) 21(47.7) 12(27.3) 0.49/0.51 I/D-Exon 8 Frequency n II (%) I/D(%) D/D(%) I/D Mothers 90 17(18.9) 58(64.4) 15(16.7) 0.51/0.49 Fathers 90 14(15.6) 49(54.4) 27(30.0) 0.43/0.57 Offspring 90 21(23.3) 47(52.2) 22(24.5) 0.49/0.51 - Male 46 10(21.7) 26(56.6) 10(21.7) 0.5/0.5 -Female 44 11(25.0) 21(47.7) 12(27.3) 0.49/0.51 WO 99/43851 PCT/IE99/00012 47 Table 3: Genotype mating outcomes for the HLA-G polymorphisms in control and pre eclampsia trios. Mother Father Offspring mt/mnt/pt/pnt 93*** Exon 8**** AA AA AA AAAA 7 4 aa aa aa aaaa 6 3 AA aa Aa AAaa 6 7 aa AA Aa aaAA 8 2 AA Aa AA AAAa 4 1 AA Aa Aa AAaA 2 5 aa Aa aa aaaA 4 8 aa Aa Aa aaAa 3 2 Aa AA AA AaAA 7 7 Aa AA Aa aAAA 12 1 Aa aa aa aAaa 4 3 Aa aa Aa Aaaa 1 14 Aa Aa AA AaAa 6 9 Aa Aa Aa AaaA* 0 10 aAAa* 3 0 (Aa)** 6 6 Aa Aa aa aAaA 11 8 total 90 90 mt/mnt/pt/pnt = maternally transmitted / maternally non-transmitted / paternally 5 transmitted / paternally non-transmitted. * Allele transmitted assigned from haplotype analysis, ** not possible to determine allele transmitted, ***for C/T-93 matings, A = C 93, a= T-93, ****for I/D-E8 matings, A = I-E8, a = D-E8 Table 4: Relative risk of foetal, maternal and parent of origin effects in a log linear model 4-factor model Stepwise reduced model' Risk Factor Relative risks C-93 I-E8 C-93 I-E8 1 or 2 alleles in offspring 0.38* 0.43 0.29" 0.39' 1 or 2 alleles in mother 0.85 1.17 ---- maternal origin 0.70 4.03" ---- 4.68 paternal origin 1.59 0.88 2.12* --- Chi-square improvement in fit 2 14.6 23.8 12.9 24.5 overall p-value 0.006 0.0001 0.002 0.000005 significance at the 5% level significance at the 1% level *significance at the 0.1% level The model was reduced by stepwise elimination of parameters whose significance was greater than 0.10. 2 Compared to a model where none of the four factors shown are fitted i.e. only a term for mating-type stratification is fitted as it is for all models here.

WO 99/43851 PCT/IE99/00012 48 Table 5: Comparisons within primigravida trios. Maternal transmitted vs. paternal transmitted C/T-93 pl= 0 .009 I/D-E8 pl=0.000001 93-E8 haplotype p3=0.000003 Maternal non-transmitted vs. paternal non-transmitted alleles C/T-93 p9= 0

.

7 5 I/D-E8 pl=0.

0 16 93-E8 haplotype P3=0.0 2 8 Allele transmitted to offspring C-93 vs. T-93 p=0.062 (transmission disequilibrium test) I-E8 vs. D-E8 p=0.37 (transmission disequilibrium test) Maternal transmitted vs. non-transmitted alleles C-93 vs. T-93 p=0.032 (transmission disequilibrium test) I-E8 vs. D-E8 p=0.0005 (transmission disequilibrium test) Paternal transmitted vs. non-transmitted alleles C-93 vs. T-93 p=0.87 (transmission disequilibrium test) I-E8 vs. D-E8 p=0.01 (transmission disequilibrium test) Probability values (p) are presented with the numbers of degrees of freedom as a subscript.

WO 99/43851 PCT/IE99/00012 49 Table 6: Extended haplotype transmission and frequency. Haplotype MT MNT PT PNT Frequency polymorphic sites 31-93-107-110-E8 a-a-a-a-b 17 30 36 25 0.321 a-b-a-a-a 41 18 18 23 0.298 a-a-a-a-a 6 11 10 19 0.137 a-b-a-a-b 6 16 14 9 0.134 a-a-a-b-b 2 5 2 2 0.032 a-a-b-a-a 4 2 2 2 0.030 a-a-a-b-a 3 0 0 1 0.012 a-a-b-a-b 1 1 0 1 0.009 a-b-b-a-b 0 1 1 1 0.009 a-b-b-a-a 2 0 0 0 0.006 a-b-a-b-b 1 0 1 0 0.006 a-b-a-b-a 1 0 0 0 0.003 a-a-b-b-b 0 0 0 1 0.003 37, 93, 107, 110, and E8 refer to the polymorphic sites in codon 31, 93, 107, 110 and E8. "a" and "b" represent the most common and least common allele respectively of each polymorphic site. MTIMNT/PT/PNT= maternally transmitted / maternally non transmitted / paternally transmitted / paternally non-transmitted. Table 7: Transmitted and non-transmitted HLA-G haplotypes to offspring in primigravida trios. HLA-G Maternal Maternal non- Paternal Paternal non Haplotype transmitted transmitted transmitted transmitted haplotype haplotype haplotype haplotype C-93/I-E8 13 (0.16) 11(0.13) 11(0.13) 22(0.26) C-93/D-E8 21(0.25) 38 (0.45) 40 (0.48) 29 (0.34) T-93/D-E8 6 (0.07) 18 (0.22) 19 (0.22) 9 (0.11) T-93/I-E8 44 (0.52) 17 (0.20) 14 (0.17) 24 (0.29) n 84 84 84 84 WO 99/43851 PCT/IE99/00012 50 Table 8: Transmitted and non-transmitted HLA-G haplotypes in trios. Haplotype Offspring Haplotype Mothers Fathers n MT-PT n T - NT n C-I C-I 3 C-I C-I 2 3 C-I C-D 6 C-I C-D 4 3 C-D C-I 2 C-D C-I 2 10 C-I T-D 2 C-I T-D 6 1 T-D C-I 2 T-D C-I 2 7 C-I T-I 2 C-I T-I 1 4 T-I C-I 4 T-I C-I 5 2 C-D C-D 13 C-D C-D 11 19 C-D T-D 5 C-D T-D 3 2 T-D C-D 0 T-D C-D 0 3 C-D T-I 0 C-D T-I 5 9 T-I C-D 21 T-I C-D 23 4 T-D T-D 4 T-D T-D 2 3 T-D T-I 1 T-D T-I 2 6 T-I T-D 8 T-I T-D 7 3 T-I T-I 11 T-I T-I 9 5 n= 84 84 84 T: Haplotype transmitted to offspring, NT: Haplotype non-transmitted to offspring. MT: Haplotype transmitted from mother to offspring, PT: Haplotype transmitted from father to offspring Table 9: Genotype and allele distribution of the HLA-G C/T-93 and I/D-E8 polymorphisms in pre-eclampsia primigravida trios. C/T-93 Frequency I/D-Exon 8 Frequency Controls n C/C (%) C/T (%) T/T (%) C/T I/I (%) I/D(%) D/D(%) I/D Mothers 90 19(21.1) 50(55.6) 21(23.3) 0.49/0.51 17(18.9) 58(64.4) 15(16.7) 0.51/0.49 Fathers 90 34(37.8) 39(43.3) 17(18.9) 0.59/0.41 14(15.6) 49(54.4) 27(30.0) 0.43/0.57 Offspring 90 24(26.7) 41(45.5) 25(27.8) 0.49/0.51 21(23.3) 47(52.2) 22(24.5) 0.49/0.51 PE Mothers 79 30(37.9) 36(45.6) 13(16.5) 0.61/0.39 13(16.5) 47(59.5) 19(24.0) 0.46/0.54 Fathers 76 15(19.7) 48(63.2) 13(17.1) 0.51/0.49 19(25.0) 47(61.8) 10(13.2) 0.56/0.44 Offspring 82 18(22.0) 57(69.5) 7(8.5) 0.57/0.43 14(17.1) 55(67.1) 13(15.8) 0.51/0.49 WO 99/43851 PCT/IE99/00012 51 Table 10:Transmitted and non-transmitted HLA-G haplotypes to offspring in control and pre-eclampsia trios. HLA-G Maternal Maternal non- Paternal Paternal non Haplotype transmitted transmitted transmitted transmitted haplotype haplotype haplotype haplotype Control Trios C-93fI-E8 13(0.16) 11(0.13) 11(0.13) 22 (0.26) C-93/D-E8 21(0.25) 38 (0.45) 40 (0.48) 29 (0.34) T-93/D-E8 6 (0.07) 18 (0.22) 19 (0.22) 9 (0.11) T-93/I-E8 44 (0.52) 17 (0.20) 14 (0.17) 24 (0.29) n 84 84 84 84 PE Trios C-93/I-E8 5(0.07) 13(0.18) 12(0.17) 11(0.16) C-93/D-E8 42 (0.60) 27 (0.39) 22(0.32) 24 (0.35) T-93/D-E8 4 (0.06) 5 (0.07) 3(0.04) 11(0.16) T-93/I-E8 19 (0.27) 25 (0.36) 33(0.47) 22 (0.33) n 70 70 70 68 Table 11: Genotype mating outcomes for the HLA-G polymorphisms in control and pre eclampsia trios. Control Trios PE Trios Mother Father Offspring mt/mnt/pt/pnt 93' Exon 8' 93' Exon 8' AA AA AA AAAA 7 4 4 1 aa aa aa aaaa 6 3 0 0 AA aa Aa AAaa 6 7 8 2 aa AA Aa aaAA 8 2 0 7 AA Aa AA AAAa 4 1 7 5 AA Aa Aa AAaA 2 5 9 4 aa Aa aa aaaA 4 8 4 5 aa Aa Aa aaAa 3 2 9 6 Aa AA AA AaAA 7 7 4 0 Aa AA Aa aAAA 12 1 7 11 Aa aa aa aAaa 4 3 0 4 Aa aa Aa Aaaa 1 14 5 4 Aa Aa AA AaAa 6 9 2 6 Aa An Aa AaaA* 0 10 6 3 aAAa* 3 0 2 5 (Aa) ' 6 6 6 7 An Aa aa aAaA 1.1 8 2 3 mt/mnt/pt/pnt = maternally transmitted / maternally non-transmitted / paternally transmitted / paternally non-transmitted. 'For C/T-93 matings, A = C-93, a = T-93, 'for I/D-E8 matings, A= I-E8, a = D-E8. *Allele transmitted assigned from haplotype, analysis, I not possible to determine allele transmitted.

WO 99/43851 PCT/IE99/00012 52 Table 12: Comparisons between Control and pre-eclampsia trios Mothers Fathers Offspring Allele frequency C/T-93 pl= 0.03 pl= 0.14 pl= 0.18 I/D-E8 pl = 0.37 pl = 0.02 pl= 0.83 93-E8 haplotype frequency p3-0.03 p3= 0

.

0 0 8 p3--0.03 Genotype distribution C/T-93 p2 = 0.05 p2= 0.02 p2 = 0.001 I/D-E8 p2= 0.49 p2 = 0.02 p2 = 0.14 Deviation from Hardy-Weinberg equilibrium Control Control Control PE PE PE Offspring Mothers Fathers Offspring Mother Fathers s C/T-93 pl-0.40 pl=0.29 p 1 =0.34 pl=0.0002 p=0.69 pl=0.021 I/D-E8 pi=0.

6 7 pl=0.006 pl=0.29 pl= 0 .002 pi=0.08 p,=0.027 Parental transmission to offspring MT MNT PT PNT C/T-93 pi=0.000 7 p-0.88 pi=0.13 pi=0.

2 5 I/D-E8 pi=0.00006 p-l=0.009 pl=0.00002 p=0.

4 4 93-E8 haplotype p3-0.0002 p3-0.02 p3-0.00003 p3=0.43 MT = maternally transmitted, MNT = maternally non-transmitted PT = paternally transmitted, PNT = paternally non-transmitted Probability values (p) are presented with the numbers of degrees of freedom as a subscript.

WO 99/43851 PCT/IE99/00012 53 Table 13: Comparisons within control trios and within pre-eclampsia trios. a) Comparison of transmitted and non-transmitted alleles Controls PE Heterozyogte vs. homozygote mating outcome CIT-93 p-0.

2 5 6 pi=0.002 I/D-E8 pi=0.

3 1 7 pi=0.014 Allele transmitted to offspring C-93 v. T-93 (TDT) table 5 p=0.49 I-E8 v. D-E8 (TDT) table 5 p=0.77 Maternal transmitted vs. non-transmitted alleles C-93 v. T-93 (TDT) table 5 p=0.65 I-E8 v. D-E8 table 5 p=0.09 (transmission disequilibrium test) Paternal transmitted vs. non-transmitted alleles C-93 v. T-93 table 5 p=0.60 (transmission disequilibrium test) I-E8 v. D-E8 table 5 p=0.

2 4 (transmission disequilibrium test) b) Test of difference between parent of origin Maternal transmitted vs. paternal transmitted C-93 v. T-93 table 5 pl=0.03 I-E8 v. D-E8 table 5 p=0.0007 93/E8 haplotypes table 5 p3=0.005 Maternal non-transmitted vs. paternal non-transmitted alleles C-93 v. T-93 table 5 p=0.5 I-E8 v. D-E8 table 5 p=0.49 93/E8 haplotypes table 5 p3=0.43 Probability values (p) are presented with the numbers of degrees of freedom as a subscript.

WO 99/43851 PCT/IE99/00012 54 Table 14: Transmitted and non-transmitted HLA-G haplotypes in control and pre eclampsia trios. HLA-G Control PE HLA-G Control Control PE PE Haplotype Offspring Offspring Haplotype Mothers Fathers Mothers Fathers MT - PT n n T- NT n n n n C-I C-I 3 1 C-I C-I 2 3 0 2 C-I C-D 6 3 C-I C-D 4 3 3 3 C-D C-I 2 2 C-D C-I 2 10 11 1 C-I T-D 2 0 C-I T-D 6 1 0 5 T-D C-I 2 3 T-D C-I 2 7 1 1 C-I T-I 2 1 C-I T-I I 4 2 2 T-I C-I 4 4 T-I C-I 5 2 1 7 C-D C-D 13 8 C-D C-D 11 19 16 8 C-D T-D 5 3 C-D T-D 3 2 2 1 T-D C-D 0 1 T-D C-D 0 3 0 0 C-D T-I 0 22 C-D T-I 5 9 13 12 T-I C-D 21 9 T-I C-D 23 4 8 13 T-D T-D 4 0 T-D T-D 2 3 1 0 T-D T-I 1 0 T-D T-I 2 6 2 1 T-I T-D 8 0 T-I T-D 7 3 2 5 T-I T-I 11 6 T-I T-I 9 5 8 7 n= 84 63 84 84 70 68 MT: haplotype transmitted from mother to offspring, PT: haplotype transmitted from father to offspring T: haplotype transmitted to offspring, NT: haplotype non-transmitted to offspring. Table 15: Relative risk of foetal, maternal and parent of origin effects in a log linear model Risk Factor Relative risks PE C-93 I-E8 I or 2 alleles in offspring 3.5 l* 1.7 1 or 2 alleles in mother 0.66 0.98 maternal origin 1.12 0.59 paternal origin 0.74 1.11 * significance at the 5% level WO 99/43851 PCT/IE99/00012 55 Table 16: Genotype and allele distribution of the HLA-G C/T-93 and I/D-E8 polymorphisms in recurrent miscarriage couples. C/T-93 Frequency n C/C C/T T/T C/T Females 22 13 7 2 0.75/0.25 Males 20 6 12 2 0.6/0.4 I/D-Exon 8 Frequency n I/I (%) I/D(%) D/D(%) IID Females 22 3 11 8 0.39/0.61 Males 21 5 12 3 0.55/0.45 C-93/I-E8 C-93/D-E8 T-93/D-E8 T-93/I-E8 Females 7 25 1 9 Males 7 17 1 15 5 Table 17: HLA-G haplotypes in recurrent miscarriage couples. Couple Female partner Male partner no. HLA G haplotype HLA-G haplotype 1 t-i / c-d t-i / c-d 2 c-i / c-d c-i / c-d 3 c-i / c-d t-i / c-d 4 c-i / c-d t-i / c-d 5 c-d / c-d t-i / c-d 6 c-i / c-d c-i / c-i 7 c-d / c-d c-i / t-i 8 c-i / t-i t-d / c-i 9 t-i / c-d t-i / t-i 10 t-i / t-i t-i / c-d 11 c-d / c-d c-i / c-i 12 c-d / c-d t-i / c-d 13 t-i / t-i t-i / c-d 14 t-d / c-i c-d / c-d 15 c-d / c-d t-i / t-i 16 c-i / c-d t-i / c-d 17 c-d / c-d c-d / c-d 18 c-d / c-d c-d / c-d 19 t-i / c-d t-i / c-d 20 t-i / c-d t-i / c-d WO 99/43851 PCT/IE99/00012 56 Table 18: Transmitted and non-transmitted HLA-G haplotypes (extended genotypes) in first and second offspring of normal mothers haplotypes CMT / CFT / first offspring second offspring MNT FNT C-I / C-D 2 2 2 0 C-D/C-I 1 2 5 5 C-I / T-D 1 0 0 0 T-D / C-I 0 1 1 1 C-I / T-I 2 5 3 1 T-I / C-I 4 0 0 0 C-D/T-D 3 1 2 0 T-D / C-D 5 3 3 2 C-D / T-I 6 3 0 5 T-I / C-D 2 2 1 5 T-D / T-I 3 3 4 3 T-I / T-D 2 6 6 6 C-I / C-I 0 0 0 0 C-D / C-D 12 14 14 11 T-D / T-D 1 1 1 0 T-I / T-I 4 5 5 8 T-I / T-I 4 5 5 8 MT: haplotype transmitted from mother to offspring, PT: haplotype transmitted from father to offspring T: haplotype transmitted to offspring, NT: haplotype non-transmitted to offspring. Table 19: Transmitted and non-transmitted ILA-G extended genotypes in first and second offspring of primigravida pre-eclampsia mothers first offspring second mother father pre-eclampsia offspring normal pregnancy MT / PT MTIPT 1 C-D / C-D C-D / C-D C-I I C-D C-D / C-D 2 T-I/ C-D C-D / C-D C-D / T-1 C-1 / C-D 3 T-il C-D T-I/ C-D T-l/ T-1 C-D / T-I 4 C-1 / C-D C-D / C-D C-I / C-D C-D / T-1 5 C-D / T-I* C-D / T-I * C-D / C-D C-D / T-1 6 C-D / T-I* C-D / T-I* C-D / T-D T-1 / T-I 7 C-D / T-l* C-D / T-I* C-D / C-D C-1 / T-1 8 T-D / C-I T-D I C-I T-D / C-1 C-1 / C-I 9 T-I/ C-D T-II C-D T-I / T-I C-D I T-1 MT: haplotype transmitted from mother to offspring, PT: haplotype transmitted from father to offspring

Claims

1. A method for diagnosing susceptibility to normal pregnancy, pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility comprising the steps of: 5 a) obtaining a fluid and/or tissue sample from a female and/or male and/or foetus; and either b) determining the sequence of all or part of the HLA-G nucleic acid, and/or HLA-G linked nucleic acid; or c) detecting variant forms of all or part of the HLA-G protein, and/or proteins encoded by HLA-G linked genes or: 10 d) measuring the functional activity of all or part of the HLA-G encoding protein and/or proteins encoded by HLA-G linked genes or: e) measuring the size and /or level of all or part of HLA-G mRNA or mRNA transcribed from HLA-G linked genes or: f) measuring the size and /or level of all or part of HLA-G protein and/or protein encoded by HLA-G 15 linked genes or: g) quantifying cells or molecules whose concentration changes as a result of HLA-G action; and h) comparing any of the parameters b) to g) with those of a female and/or male and/or foetus of a normal pregnancy and/or a pregnancy with pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related fertility outcome. 20

2. A method as claimed in claim 1 wherein the HLA-G nucleic acid is analysed for the presence of the C and/or T allele of codon 93 in exon 3 and/or the insertion and/or deletion allele of exon 8.

3. A method as claimed in claim I wherein the effect of one or more of the HLA-G sequence variants on the functional activity of HLA-G and / or on the size and /or level of all or part of HLA-G mRNA and/or its encoded polypeptide is measured. 25

4. A method as claimed in claim 1 or 2 wherein all or part of any HLA-G sequence and/or HLA-G linked sequences is amplified, preferably by a method or combination of methods selected from the polymerase chain reaction, nucleic acid sequence based amplification, self sustained sequence replication, transcription-mediated amplification, strand displacement amplification, and the ligase chain reaction.

5. A method as claimed in claim 1-4 wherein comparing of one or more variants identified is performed 30 by association and/or linkage analysis and/or transmission analysis. WO 99/43851 PCT/IE99/00012 58

6. A method as claimed in any preceding claim wherein all or part of the HLA-G sequence is cloned into a vector.

7. A method as claimed in any preceding claim wherein all or part of the nucleic acid sequence is identified by a method or combination of methods selected from DNA sequencing, glycosylase mediated 5 polymorphism detection, restriction fragment length polymorphism analysis, enzymatic or chemical cleavage analysis, hybridisation to DNA and /or RNA probes and /or DNA probe arrays and/or allele specific DNA and /or RNA probes, allele specific amplification analysis, electrophoretic mobility analysis and 5' nuclease assay analysis.

8. A method as claimed in any preceding claim wherein all or part of HLA-G and /or all or part of one 10 or more variants thereof is expressed as a polypeptide in vitro and/or in a prokaryotic and / or eukaryotic cell.

9. A method as claimed in claim 1 wherein the cells of step (g) are blood mononuclear cells and / or T cell and /or natural killer cell subsets thereof and/or HLA-G expressing cells.

10. A method as claimed in any preceding claim wherein the activity of HLA-G and/or any combination 15 of variants thereof and/or blood mononuclear cells and /or a subset of such cells, selected from T cells and/or natural killer cells, is measured by one or more of the following procedures: (a) measuring the interaction of HLA-G and /or variants thereof with blood mononuclear cells and/or subsets thereof by assessing one or more of the following with respect to HLA-G expressing cells and /or blood mononuclear cells; cell proliferation, transformation, cytotoxic response, surface marker 20 expression, cytokine production, conjugate formation and target specificity, (b) measuring the size and / or level of all or part of HLA-G mRNA and/or its encoded polypeptide, (c) measuring the peptide binding capability of all or part of HLA-G and /or variants thereof, (d) measuring the binding capability of all or part of the HLA-G and /or variants thereof to a HLA-G receptor, 25 (e) measuring one or more molecules whose level is altered as a result of the interaction of the HLA-G and /or variants thereof and /or cells expressing HLA-G with blood mononuclear cells, (f) measuring the expression levels of one or more genes and/or proteins in the HLA-G expressing cells.

11. A method as claimed in any preceding claim wherein blood mononuclear cells and/or subsets thereof and/or HLA-G and/or HLA-G linked variants thereof and/or cells expressing all or part of the variants 30 fully and/or partially matching a female and/or male and/or foetus are selected from a test panel.

12. A method as claimed in any preceding claim wherein the HLA-G is partially or fully purified from a cell expressing HLA-G. WO 99/43851 PCT/IE99/00012 59

13. A method as claimed in any preceding claim wherein the HLA-G is detected by immunoassay using one or more antibodies specific for HLA-G and/or variants thereof.

14. A method as claimed in any preceding claim wherein all or part of the HLA-A, HLA-B, HLA-C, HLA-E, HLA-F and HLA-H genes are analysed in the female and/or male and/or foetus. 5

15. A method as claimed in claim 1 wherein the molecules of step (g) are selected from IL-1 beta, IL-2, IL-3, IL-4, IL-6, IL-10 and tumour necrosis factor-alpha, or trophoblast specific markers selected from cytokeratins, pregnancy specific glycoprotein 1, human chorionic gonadotrophin and human placental lactogen.

16. A method for screening for agents which can potentially be used as diagnostic indicators and/or drug 10 targets for pre-eclampsia, miscarriage, miscarriage-related infertility and intrauterine growth retardation by: a) measuring the expression level of one or more genes and/or proteins in HLA-G expressing cells and/or blood mononuclear cells and/or T cell and/or natural killer cells subsets thereof following interaction with HLA-G and/or HLA-G expressing cells; 15 b) comparing the expression level identified in step a) with the expression level in HLA-G expressing cells and/or the blood mononuclear cells and/or T cell and/or natural killer cell subsets thereof following interaction with HLA-G and/or HLA-G expressing cells in normal pregnancy and/or pre-eclampsia pregnancy and/or intrauterine growth retardation pregnancy and/or miscarriage pregnancy and/or miscarriage-related infertility. 20

17. A method as claimed in claim 10 or 16 wherein gene expression and/or protein expression is measured by any one or combination of methods selected from hybridisation between cDNA and/or RNA from the cells and DNA probes and/or RNA probes and/or nucleic acid probe arrays, quantitative amplification methods, reverse transcriptase - polymerase chain reaction (RT-PCR), 5' nuclease assay, ribonuclease protection assay and S I nuclease assay, one dimensional and/or two dimensional gel 25 electrophoresis and staining of proteins, detection of one or more proteins using, enzyme linked immunosorbent assays (ELISA), radioimmunoassays (RIA), protein truncation test (PTT), immunoradiometric assays (IRMA) and immunoenzymatic assays (IEMA), sandwich assays and Western blotting using monoclonal and/or polyclonal antibodies.

18. A pharmaceutical composition comprising a pharmaceutically effective amount of HLA-G and/or 30 cells expressing HLA-G and/or one or more peptides which binds to HLA-G, blood mononuclear cells from a donor and/or test panel known to interact with HLA-G variants, cytokines and any combination thereof including IL-1 beta, IL-2, IL-3, IL-4, IL-6, IL-10 and tumour necrosis factor-alpha and/or WO 99/43851 PCT/IE99/00012 60 inhibitors of cytokines and/or tumour necrosis factor alpha and/or derivatives of cytokines and/or tumour necrosis factor-alpha, optionally with pharmaceutically-acceptable carriers or excipients.

19. A method for screening potential pre-eclampsia and eclampsia and intrauterine growth retardation and miscarriage and miscarriage-related infertility therapeutic agents selected from: 5 a) identifying agents which alter the expression of HLA-G; b) identifying agents which alter the activity of HLA-G; c) identifying agents which mimic the action of HLA-G; d) identifying agents which bind to HLA-G; e) identifying peptides which bind to HLA-G; 10 f) identifying agents which bind to HLA-G receptors; g) identifying expressed genes using DNA probe arrays in a cellular background in HLA-G expressing cells and / or blood mononuclear cells interacting with HLA-G and / or cells expressing HLA-G interacting with blood mononuclear cells; h) identifying expressed proteins using mass spectrometry methods in HLA-G expressing cells and / or 15 blood mononuclear cells interacting with HLA-G and / or cells expressing HLA-G interacting with blood mononuclear cells; i) identifying expressed proteins using mass spectrometry methods in HLA-G expressing cells and / or blood mononuclear cells interacting with HLA-G and / or cells expressing HLA-G interacting with blood mononuclear cells; 20 j) screening sperm and/or semen and/or female reproductive tissue for agents: i) which alter the expression of HLA-G in fertilised eggs and/or embryos; ii) which alter the cell cleavage rate of fertilised eggs and/or embryos; iii) which induce cellular factors in cells in culture and/or cells in vivo that alter the cell cleavage rate of fertilised eggs and/or embryos. 25

20. A method for the prevention of pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility selected from: a) treatment of a female with all or part of a pharmaceutically effective amount of an effective HLA-G and /or peptides which bind to HLA-G and / or cells expressing HLA-G; b) treatment of a female with all or part of a pharmaceutically effective amount of molecules or 30 inhibitors of molecules whose level or activity is directly or indirectly altered by HLA-G action; c) treatment of a female with all or part of a pharmaceutically effective amount of molecules which inhibit the interaction between HLA-G and one or more of its receptors; d) treatment of a female with all or part of a pharmaceutically effective amount of an agent which alters HLA-G expression: WO 99/43851 PCT/IE99/00012 61 e) treatment of a female with all or part of a pharmaceutically effective amount of an agent which alters HLA-G related blood mononuclear cell activity; f) treatment of a female with all or part of a pharmaceutically effective amount of an agent which mimics all or part of HLA-G action: 5 g) treatment of a female with blood mononuclear cells that recognise foetal and / or self HLA-G; h) treatment of a female with HLA-G and / or cells expressing HLA-G or variants thereof; i) treatment of a female with one or more antibodies which bind to HLA-G and I or cells expressing HLA-G and / or any receptor for HLA-G; j) introduction of one or more variants of the HLA-G gene and /or its receptor into a female and / or 10 male; k) introduction of an inhibitor of expression of the HLA-G gene and/or its receptor into a female and/or male; 1) inactivation of one or more variants of the HLA-G gene and/or its receptor in a female and/or male.

21. A method for improving pregnancy success selected from: 15 a) pre-treating the female with sperm and/or attenuated forms thereof, and/or semen and/or fractions thereof from a male with a known HLA-G genotype, prior to mating with a male of a different HLA-G genotype, and/or in vitro fertilisation using sperm from a male of a different HLA-G genotype and/or embryo transfer where the male HLA-G is of a different HLA-G genotype; b) mixing sperm of a known HLA-G genotype with sperm and/or attenuated forms thereof. and/or 20 semen and/or fractions thereof from a male with a different HLA-G genotype prior to in vitro fertilisation.

22. A method as claimed in claim 21 wherein fertility and / or pregnancy outcome are improved by selection of male and / or female partners and / or sperm and / or ova and / or recipients of fertilised eggs and / or zygotes / and / or embryos so that (a) their HLA-G and /or HLA genotypes and /or serotypes or (b) the activity of their HLA-G and / or blood mononuclear cells interacting with HLA-G are indicative of 25 normal pregnancy outcomes and / or not associated with pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility.

23. A test kit for the diagnosis of susceptibility to normal pregnancy, pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility or for monitoring progress of pregnancy comprising: 30 a) oligonucleotide primers for amplification of all or part of the HLA-G gene and /or HLA-G linked DNA; WO 99/43851 PCT/IE99/00012 62 b) amplification reagents for amplification of genomic DNA and / or RNA segments, selected from a DNA / RNA polymerase, a reverse transcriptase, the deoxyribonucleotides dATP, dCTP, dGTP, dTTP and dUTP. and /or ribonucleotides ATP, CTP, GTP, TTP and UTP, and reaction buffer; c) reagents for identifying sequence variants in DNA and / or RNA; 5 d) control DNA and /or RNA.

24. Use of a DNA sequence selected from any one of Sequence I.D.s 1 to 21 for diagnosis of susceptibility to or in a test kit for the diagnosis of susceptibility to normal pregnancy, pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility, for monitoring progress of pregnancy, for use in the manufacture of a 10 medicament, in a method for screening potential therapeutic agents, in a method for screening for potential diagnostic indicators and/or drug targets, in a method for improving pregnancy success or in a method for the prevention of pre-eclampsia and/or eclampsia and/or intrauterine growth retardation and/or susceptibility to miscarriage and/or miscarriage-related infertility, for monitoring progress of pregnancy.

25. A method for induction of tolerance in a host to a non-self tissue which comprises administering 15 HLA-G and /or HLA-G loaded with peptides from the non-self tissue and /or HLA-G expressing cells derived from or related to the non-self tissue, and/or a non-self tissue bearing an introduced HLA-G so that HLA-G is expressed in all or part of the tissue.

26. A method for the treatment of autoimmune disease which comprises administering HLA-G and /or HLA-G loaded with peptides from a self and/or non-self tissue and / or with specific autoimmune antigen 20 and /or HLA-G expressing cells from a self and/or non-self tissue and/or a self and/or self tissue bearing an introduced HLA-G gene so that HLA-G is expressed in all or part of the tissue. WO 99/43851 PCT/IE99/00012 SEQUENCE LISTING <110> National University of Ireland, Cork <120> HLA Linked Pre-Eclampsia and Miscarriage Susceptibility Gene <130> PL977PCT <140> Not Yet Allocated <141> 1999-02-25 <150> IE980134 <151> 1998-02-25 <150> IE980668 <151> 1998-08-12 <160> 23 <170> Patentln Ver. 2.1 <210> 1 <211> 22 <212> DNA <213> Homo sapiens <300> <400> 1 tactcccgag tctccgggtc tg 22 22<210> 2 <211> 2 <211> 23 <212> DNA <213> Homo sapiens <400> 2 aggcgcccca ctgcccctgg tac 23 <210> 3 <211> 25 <212> DNA <213> Homo sapiens <400> 3 gaccaagggg gtggggccag gttct 25 1/9 WO 99/43851 PCT/IE99/00012 <210> 4 <211> 460 <212> DNA <213> Homo sapiens <400> 4 tactcccgag tctccgggtc tgggatccac cccgaggccg cgggacccgc ccagaccctc 60 tacctgggag aaccccaagg cgcctttacc aaaatccccg cgggtgggtc cgggcgaggg 120 cgaggctcgg tgggcggggc tgaccgaggg ggtggggcca ggttctcaca ccctccagtg 180 gatgattggc tgcgacctgg ggtccgacgg acgcctcctc cgcgggtatg aacagtatgc 240 ctacgatggc aaggattacc tcgccctgaa cgaggacctg cgctcctgga ccgcagcgga 300 cactgcggct cagatctcca agcgcaagtg tgaggcggcc aatgtggctg aacaaaggag 360 agcctacctg gagggcacgt gcgtggagtg gctccacaga tacctggaga acgggaagga 420 gatgctgcag cgcgcgggta ccaggggcag tggggcgcct 460 <210> 5 <211> 460 <212> DNA <213> Homo sapiens <400> 5 tactcccgag tctccgggtc tgggatccac cccgaggccg cgggacccgc ccagaccctc 60 tacctgggag aaccccaagg cgcctttacc aaaatccccg cgggtgggtc cgggcgaggg 120 cgaggctcgg tgggcggggc tgaccgaggg ggtggggcca ggttctcata ccctccagtg 180 gatgattggc tgcgacctgg ggtccgacgg acgcctcctc cgcgggtatg aacagtatgc 240 ctacgatggc aaggattacc tcgccctgaa cgaggacctg cgctcctgga ccgcagcgga 300 cactgcggct cagatctcca agcgcaagtg tgaggcggcc aatgtggctg aacaaaggag 360 agcctacctg gagggcacgt gcgtggagtg gctccacaga tacctggaga acgggaagga 420 gatgctgcag cgcgcgggta ccaggggcag tggggcgcct 460 <210> 6 <211> 319 <212> DNA <213> Homo sapiens <400> 6 gaccgagggg gtggggccag gttctcacac cctccagtgg atgattggct gcgacctggg 60 gtccgacgga cgcctcctcc gcgggtatga acagtatgcc tacgatggca aggattacct 120 cgccctgaac gaggacctgc gctcctggac cgcagcggac actgcggctc agatctccaa 180 gcgcaagtgt gaggcggcca atgtggctga acaaaggaga gcctacctgg agggcacgtg 240 cgtggagtgg ctccacagat acctggagaa cgggaaggag atgctgcagc gcgcgggtac 300 caggggcagt ggggcgcct 319 <210> 7 <211> 319 <212> DNA <213> Homo sapiens 2/9 WO 99/43851 PCT/IE99/00012 <400> 7 gaccgagggg gtggggccag gttctcatac cctccagtgg atgattggct gcgacctggg 60 gtccgacgga cgcctcctcc gcgggtatga acagtatgcc tacgatggca aggattacct 120 cgccctgaac gaggacctgc gctcctggac cgcagcggac actgcggctc agatctccaa 180 gcgcaagtgt gaggcggcca atgtggctga acaaaggaga gcctacctgg agggcacgtg 240 cgtggagtgg ctccacagat acctggagaa cgggaaggag atgctgcagc gcgcgggtac 300 caggggcagt ggggcgcct 319 <210> 8 <211> 32 <212> DNA <213> Homo sapiens <400> 8 gaccgagggg gtggggccag gttctcacac cc 32 <210> 9 <211> 27 <212> DNA <213> Homo sapiens <400> 9 gaccgagggg gtggggccag gttctca 27 <210> 10 <211> 21 <212> DNA <213> Homo sapiens <400> 10 tgtgaaacag ctgccctgtg t 21 <210> 11 <211> 21 <212> DNA <213> Homo sapiens <400> 11 aaggaatgca gttcagcatg a 21 <210> 12 <211> 151 <212> DNA <213> Homo sapiens <400> 12 tgtgaaacag ctgccctgtg tgggactgag tggcaagatt tgttcatgcc ttccctttgt 60 gacttcaaga accctgactt ctctttgtgc agagaccagc ccacccctgt gcccaccatg 120 3/9 WO 99/43851 PCT/IE99/00012 accctcttcc tcatgctgaa ctgcattcct t 151 <210> 13 <211> 137 <212> DNA <213> Homo sapiens <400> 13 tgtgaaacag ctgccctgtg tgggactgag tggcaagtcc ctttgtgact tcaagaaccc 60 tgacttctct ttgtgcagag accagcccac ccctgtgccc accatgaccc tcttcctcat 120 gctgaactgc attcctt 137 <210> 14 <211> 26 <212> DNA <213> Homo sapiens <400> 14 caaagggaag gcatgaacaa atcttg 26 <210> 15 <211> 25 <212> DNA <213> Homo sapiens <400> 15 gttcttgaag tcacaaaggg acttg 25 <210> 16 <211> 2442 <212> DNA <213> Homo sapiens <400> 16 tactcccgag tctccgggtc tgggatccac cccgaggccg cgggacccgc ccagaccctc 60 tacctgggag aaccccaagg cgcctttacc aaaatccccg cgggtgggtc cgggcgaggg 120 cgaggctcgg tgggcggggc tgaccgaggg ggtggggcca ggttctcaca ccctccagtg 180 gatgattggc tgcgacctgg ggtccgacgg acgcctcctc cgcgggtatg aacagtatgc 240 ctacgatggc aaggattacc tcgccctgaa cgaggacctg cgctcctgga ccgcagcgga 300 cactgcggct cagatctcca agcgcaagtg tgaggcggcc aatgtggctg aacaaaggag 360 agcctacctg gagggcacgt gcgtggagtg gctccacaga tacctggaga acgggaagga 420 gatgctgcag cgcgcgggta ccaggggcag tggggcgcct ccctgatctc ctgtagacct 480 ctcagcctgg cctagcacaa ggagaggagg aaaatgggac caacactaga atatcgccct 540 ccctctggtc ctgagggaga ggaatcctcc tgggtttcca gatcctgtac cagagagtga 600 ttctgagggc ccgtcctgct ctctgggaca attaagggat gaagtctctg agggagtgga 660 ggggaagaca atccctggaa gactgatcag gggttccctt tgaccccaca gcagccttgg 720 caccaggact tttcccctca ggccttgttc tctgcctcac actcaatgtg tgtgggggtc 780 tgactccagc tcctctgagt cccttggcct ccactcaggt cagaaccgga ggtccctgct 840 4/9 WO 99/43851 PCT/IE99/00012 cccccgctca gagactagaa ctttccaagg aataggagat tatcccaggt gcccgtgtcc 900 aggctggtgt ctgggttctg tgctcccttc cccaccccag gtatctggtt cattcttagg 960 atggtcacat ccaggtgctg ctggagtgtc ccatgagaga tgcaaagtgc ttgaattttc 1020 tgactcttcc tttcagaccc cccaagaca cacgtgaccc accaccctgt ctttgactat 1080 gaggccaccc tgaggtgctg ggccctgggc ttctaccctg cggagatcat actgacctgg 1140 cagcgggatg gggaggacca gacccaggac gtggagctcg tggagaccag gcctgcaggg 1200 gatggaacct tccagaagtg ggcagctgtg gtggtgcctt ctggagagga gcagagatac 1260 acgtgccatg tgcagcatga ggggctgccg gagcccctca tgctgagatg gagtaaggag 1320 ggagatggag gcatcatgtc tgttagggaa agcaggagcc tctctgaaga cctttaacag 1380 ggtcggtggt gagggctggg ggtcagagac cctcaccttc acctcctttc ccagagcagt 1440 cttccctgcc caccatcccc atcatgggta tcgttgctgg cctggttgtc cttgcagctg 1500 tagtcactgg agctgcggtc gctgctgtgc tgtggagaaa gaagagctca ggtaaggaag 1560 gggtgacaag tggggtctga gttttcttgt cccactgggg gtttcaagcc ccaggtagaa 1620 gtgtgccctg cctggttact gggaagcacc atccacactc atgggcctac ccagcctggg 1680 ccctgtgtgc cagcaccttc tcttttgtaa agcacctgtg acaatgaagg acagatttat 1740 taccttgatg attgtagtga tggggacctg atcccagtaa tcacaggtca ggagaaggtc 1800 cctggctaag gacagacctt aggagggcag ttggtcgagg acccacatct gctttccttg 1860 tttttcctga tcgccctggg tctgcagtca cacatttctg gaaacttctc gagggtccaa 1920 gactaggagg ttcctctagg acctcatggc cctgccacct ttctggcctc tcacaggaca 1980 ttttcttccc acagattgaa aaggagggag ctactctcag gctgcaagta agtatgaagg 2040 aggctgatcc ctgagatcct tgggatcttg tgtttgggag ccatggggga gctcacccac 2100 cccacaattc ctcctctggc cacatctcct gtggtctctg accaggtgct gtttttgttc 2160 tactctaggc agtgacagtg cccagggctc taatgtgtct ctcacggctt gtaaatgtga 2220 caccccgggg ggcctgatgt gtgtgggttg ttgaggggaa caggggacat agctgtgcta 2280 tgaggtttct ttgacttcaa tgtattgagc atgtgatggg ctgtttaaag tgtcacccct 2340 cactgtgact gatatgaatt tgttcatgaa tatttttctg tagtgtgaaa cagctgccct 2400 gtgtgggact gagtggcaag atttgttcat gccttccctt tg 2442 <210> 17 <211> 2442 <212> DNA <213> Homo sapiens <400> 17 tactcccgag tctccgggtc tgggatccac cccgaggccg cgggacccgc ccagaccctc 60 tacctgggag aaccccaagg cgcctttacc aaaatccccg cgggtgggtc cgggcgaggg 120 cgaggctcgg tgggcggggc tgaccgaggg ggtggggcca ggttctcata ccctccagtg 180 gatgattggc tgcgacctgg ggtccgacgg acgcctcctc cgcgggtatg aacagtatgc 240 ctacgatggc aaggattacc tcgccctgaa cgaggacctg cgctcctgga ccgcagcgga 300 cactgcggct cagatctcca agcgcaagtg tgaggcggcc aatgtggctg aacaaaggag 360 agcctacctg gagggcacgt gcgtggagtg gctccacaga tacctggaga acgggaagga 420 gatgctgcag cgcgcgggta ccaggggcag tggggcgcct ccctgatctc ctgtagacct 480 ctcagcctgg cctagcacaa ggagaggagg aaaatgggac caacactaga atatcgccct 540 ccctctggtc ctgagggaga ggaatcctcc tgggtttcca gatcctgtac cagagagtga 600 ttctgagggc ccgtcctgct ctctgggaca attaagggat gaagtctctg agggagtgga 660 ggggaagaca atccctggaa gactgatcag gggttccctt tgaccccaca gcagccttgg 720 caccaggact tttcccctca ggccttgttc tctgcctcac actcaatgtg tgtgggggtc 780 tgactccagc tcctctgagt cccttggcct ccactcaggt cagaaccgga ggtccctgct 840 5/9 WO 99/43851 PCT/IE99/00012 cccccgctca gagactagaa ctttccaagg aataggagat tatcccaggt gcccgtgtcc 900 aggctggtgt ctgggttctg tgctcccttc cccaccccag gtatctggtt cattcttagg 960 atggtcacat ccaggtgctg ctggagtgtc ccatgagaga tgcaaagtgc ttgaattttc 1020 tgactcttcc tttcagaccc ccccaagaca cacgtgaccc accaccctgt ctttgactat 1080 gaggccaccc tgaggtgctg ggccctgggc ttctaccctg cggagatcat actgacctgg 1140 cagcgggatg gggaggacca gacccaggac gtggagctcg tggagaccag gcctgcaggg 1200 gatggaacct tccagaagtg ggcagctgtg gtggtgcctt ctggagagga gcagagatac 1260 acgtgccatg tgcagcatga ggggctgccg gagcccctca tgctgagatg gagtaaggag 1320 ggagatggag gcatcatgtc tgttagggaa agcaggagcc tctctgaaga cctttaacag 1380 ggtcggtggt gagggctggg ggtcagagac cctcaccttc acctcctttc ccagagcagt 1440 cttccctgcc caccatcccc atcatgggta tcgttgctgg cctggttgtc cttgcagctg 1500 tagtcactgg agctgcggtc gctgctgtgc tgtggagaaa gaagagctca ggtaaggaag 1560 gggtgacaag tggggtctga gttttcttgt cccactgggg gtttcaagcc ccaggtagaa 1620 gtgtgccctg cctggttact gggaagcacc atccacactc atgggcctac ccagcctggg 1680 ccctgtgtgc cagcaccttc tcttttgtaa agcacctgtg acaatgaagg acagatttat 1740 taccttgatg attgtagtga tggggacctg atcccagtaa tcacaggtca ggagaaggtc 1800 cctggctaag gacagacctt aggagggcag ttggtcgagg acccacatct gctttccttg 1860 tttttcctga tcgccctggg tctgcagtca cacatttctg gaaacttctc gagggtccaa 1920 gactaggagg ttcctctagg acctcatggc cctgccacct ttctggcctc tcacaggaca 1980 ttttcttccc acagattgaa aaggagggag ctactctcag gctgcaagta agtatgaagg 2040 aggctgatcc ctgagatcct tgggatcttg tgtttgggag ccatggggga gctcacccac 2100 cccacaattc ctcctctggc cacatctcct gtggtctctg accaggtgct gtttttgttc 2160 tactctaggc agtgacagtg cccagggctc taatgtgtct ctcacggctt gtaaatgtga 2220 caccccgggg ggcctgatgt gtgtgggttg ttgaggggaa caggggacat agctgtgcta 2280 tgaggtttct ttgacttcaa tgtattgagc atgtgatggg ctgtttaaag tgtcacccct 2340 cactgtgact gatatgaatt tgttcatgaa tatttttctg tagtgtgaaa cagctgccct 2400 gtgtgggact gagtggcaag atttgttcat gccttccctt tg 2442 <210> 18 <211> 2441 <212> DNA <213> Homo sapiens <400> 18 tactcccgag tctccgggtc tgggatccac cccgaggccg cgggacccgc ccagaccctc 60 tacctgggag aaccccaagg cgcctttacc aaaatccccg cgggtgggtc cgggcgaggg 120 cgaggctcgg tgggcggggc tgaccgaggg ggtggggcca ggttctcaca ccctccagtg 180 gatgattggc tgcgacctgg ggtccgacgg acgcctcctc cgcgggtatg aadagtatgc 240 ctacgatggc aaggattacc tcgccctgaa cgaggacctg cgctcctgga ccgcagcgga 300 cactgcggct cagatctcca agcgcaagtg tgaggcggcc aatgtggctg aacaaaggag 360 agcctacctg gagggcacgt gcgtggagtg gctccacaga tacctggaga acgggaagga 420 gatgctgcag cgcgcgggta ccaggggcag tggggcgcct ccctgatctc ctgtagacct 480 ctcagcctgg cctagcacaa ggagaggagg aaaatgggac caacactaga atatcgccct 540 ccctctggtc ctgagggaga ggaatcctcc tgggtttcca gatcctgtac cagagagtga 600 ttctgagggc ccgtcctgct ctctgggaca attaagggat gaagtctctg agggagtgga 660 ggggaagaca atccctggaa gactgatcag gggttccctt tgaccccaca gcagccttgg 720 caccaggact tttcccctca ggccttgttc tctgcctcac actcaatgtg tgtgggggtc 780 tgactccagc tcctctgagt cccttggcct ccactcaggt cagaaccgga ggtccctgct 840 6/9 WO 99/43851 PCT/IE99/00012 cccccgctca gagactagaa ctttccaagg aataggagat tatcccaggt gcccgtgtcc 900 aggctggtgt ctgggttctg tgctcccttc cccaccccag gtatctggtt cattcttagg 960 atggtcacat ccaggtgctg ctggagtgtc ccatgagaga tgcaaagtgc ttgaattttc 1020 tgactcttcc tttcagaccc ccccaagaca cacgtgaccc accaccctgt ctttgactat 1080 gaggccaccc tgaggtgctg ggccctgggc ttctaccctg cggagatcat actgacctgg 1140 cagcgggatg gggaggacca gacccaggac gtggagctcg tggagaccag gcctgcaggg 1200 gatggaacct tccagaagtg ggcagctgtg gtggtgcctt ctggagagga gcagagatac 1260 acgtgccatg tgcagcatga ggggctgccg gagcccctca tgctgagatg gagtaaggag 1320 ggagatggag gcatcatgtc tgttagggaa agcaggagcc tctctgaaga cctttaacag 1380 ggtcggtggt gagggctggg ggtcagagac cctcaccttc acctcctttc ccagagcagt 1440 cttccctgcc caccatcccc atcatgggta tcgttgctgg cctggttgtc cttgcagctg 1500 tagtcactgg agctgcggtc gctgctgtgc tgtggagaaa gaagagctca ggtaaggaag 1560 gggtgacaag tggggtctga gttttcttgt cccactgggg gtttcaagcc ccaggtagaa 1620 gtgtgccctg cctggttact gggaagcacc atccacactc atgggcctac ccagcctggg 1680 ccctgtgtgc cagcaccttc tcttttgtaa agcacctgtg acaatgaagg acagatttat 1740 taccttgatg attgtagtga tggggacctg atcccagtaa tcacaggtca ggagaaggtc 1800 cctggctaag gacagacctt aggagggcag ttggtcgagg acccacatct gctttccttg 1860 tttttcctga tcgccctggg tctgcagtca cacatttctg gaaacttctc gagggtccaa 1920 gactaggagg ttcctctagg acctcatggc cctgccacct ttctggcctc tcacaggaca 1980 ttttcttccc acagattgaa aaggagggag ctactctcag gctgcaagta agtatgaagg 2040 aggctgatcc ctgagatcct tgggatcttg tgtttgggag ccatggggga gctcacccac 2100 cccacaattc ctcctctggc cacatctcct gtggtctctg accaggtgct gtttttgttc 2160 tactctaggc agtgacagtg cccagggctc taatgtgtct ctcacggctt gtaaatgtga 2220 caccccgggg ggcctgatgt gtgtgggttg ttgaggggaa caggggacat agctgtgcta 2280 tgaggtttct ttgacttcaa tgtattgagc atgtgatggg ctgtttaaag tgtcacccct 2340 cactgtgact gatatgaatt tgttcatgaa tatttttctg tagtgtgaaa cagctgccct 2400 gtgtgggact gagtggcaag tccctttgtg acttcaagaa c 2441 <210> 19 <211> 2441 <212> DNA <213> Homo sapiens <400> 19 tactcccgag tctccgggtc tgggatccac cccgaggccg cgggacccgc ccagaccctc 60 tacctgggag aaccccaagg cgcctttacc aaaatccccg cgggtgggtc cgggcgaggg 120 cgaggctcgg tgggcggggc tgaccgaggg ggtggggcca ggttctcata ccctccagtg 180 gatgattggc tgcgacctgg ggtccgacgg acgcctcctc cgcgggtatg aacagtatgc 240 ctacgatggc aaggattacc tcgccctgaa cgaggacctg cgctcctgga ccgcagcgga 300 cactgcggct cagatctcca agcgcaagtg tgaggcggcc aatgtggctg aacaaaggag 360 agcctacctg gagggcacgt gcgtggagtg gctccacaga tacctggaga acgggaagga 420 gatgctgcag cgcgcgggta ccaggggcag tggggcgcct ccctgatctc ctgtagacct 480 ctcagcctgg cctagcacaa ggagaggagg aaaatgggac caacactaga atatcgccct 540 ccctctggtc ctgagggaga ggaatcctcc tgggtttcca gatcctgtac cagagagtga 600 ttctgagggc ccgtcctgct ctctgggaca attaagggat gaagtctctg agggagtgga 660 ggggaagaca atccctggaa gactgatcag gggttccctt tgaccccaca gcagccttgg 720 caccaggact tttcccctca ggccttgttc tctgcctcac actcaatgtg tgtgggggtc 780 tgactccagc tcctctgagt cccttggcct ccactcaggt cagaaccgga ggtccctgct 840 7/9 WO 99/43851 PCT/IE99/00012 cccccgctca gagactagaa ctttccaagg aataggagat tatcccaggt gcccgtgtcc 900 aggctggtgt ctgggttctg tgctcccttc cccaccccag gtatctggtt cattcttagg 960 atggtcacat ccaggtgctg ctggagtgtc ccatgagaga tgcaaagtgc ttgaattttc 1020 tgactcttcc tttcagaccc ccccaagaca cacgtgaccc accaccctgt ctttgactat 1080 gaggccaccc tgaggtgctg ggccctgggc ttctaccctg cggagatcat actgacctgg 1140 cagcgggatg gggaggacca gacccaggac gtggagctcg tggagaccag gcctgcaggg 1200 gatggaacct tccagaagtg ggcagctgtg gtggtgcctt ctggagagga gcagagatac 1260 acgtgccatg tgcagcatga ggggctgccg gagcccctca tgctgagatg gagtaaggag 1320 ggagatggag gcatcatgtc tgttagggaa agcaggagcc tctctgaaga cctttaacag 1380 ggtcggtggt gagggctggg ggtcagagac cctcaccttc acctcctttc ccagagcagt 1440 cttccctgcc caccatcccc atcatgggta tcgttgctgg cctggttgtc cttgcagctg 1500 tagtcactgg agctgcggtc gctgctgtgc tgtggagaaa gaagagctca ggtaaggaag 1560 gggtgacaag tggggtctga gttttcttgt cccactgggg gtttcaagcc ccaggtagaa 1620 gtgtgccctg cctggttact gggaagcacc atccacactc atgggcctac ccagcctggg 1680 ccctgtgtgc cagcaccttc tcttttgtaa agcacctgtg acaatgaagg acagatttat 1740 taccttgatg attgtagtga tggggacctg atcccagtaa tcacaggtca ggagaaggtc 1800 cctggctaag gacagacctt aggagggcag ttggtcgagg acccacatct gctttccttg 1860 tttttcctga tcgccctggg tctgcagtca cacatttctg gaaacttctc gagggtccaa 1920 gactaggagg ttcctctagg acctcatggc cctgccacct ttctggcctc tcacaggaca 1980 ttttcttccc acagattgaa aaggagggag ctactctcag gctgcaagta agtatgaagg 2040 aggctgatcc ctgagatcct tgggatcttg tgtttgggag ccatggggga gctcacccac 2100 cccacaattc ctcctctggc cacatctcct gtggtctctg accaggtgct gtttttgttc 2160 tactctaggc agtgacagtg cccagggctc taatgtgtct ctcacggctt gtaaatgtga 2220 caccccgggg ggcctgatgt gtgtgggttg ttgaggggaa caggggacat agctgtgcta 2280 tgaggtttct ttgacttcaa tgtattgagc atgtgatggg ctgtttaaag tgtcacccct 2340 cactgtgact gatatgaatt tgttcatgaa tatttttctg tagtgtgaaa cagctgccct 2400 gtgtgggact gagtggcaag tccctttgtg acttcaagaa c 2441 <210> 20 <211> 80 <212> DNA <213> Homo sapiens <400> 20 accctccagt ggatgattgg ctgcgacctg gggtccgacg gacgcctcct ccgcgggtat 60 gaacagtatg cctacgatgg 80 <210> 21 <211> 14 <212> DNA <213> Homo sapiens <400> 21 atttgttcat gcct 14 <210> 22 <211> 70 8/9 WO 99/43851 PCT/IE99/00012 <212> DNA <213> Homo sapiens <400> 22 gatatgaatt tgttcatgaa tatttttctg tagtgtgaaa cagctgccct gtgtgggact 60 gagtggcaag 70 <210> 23 <211> 80 <212> DNA <213> Homo sapiens <400> 23 tccctttgtg acttcaagaa ccctgacttc tctttctgca gagaccagcc cacccctgtg 60 cccaccatga ccctcttcct 80 89/9 9/9