HLA types in celiac disease patients not carrying the DQA1*05-DQB1*02 (DQ2) heterodimer: results from the European Genetics Cluster on Celiac Disease

Genetic susceptibility to celiac disease is strongly associated with HLA-DQA1*05-DQB1*02 (DQ2) and HLA-DQA1*03-DQB1*0302 (DQ8). Study of the HLA associations in patients not carrying these heterodimers has been limited by the rarity of such patients. This European collaboration has provided a unique opportunity to study a large series of such patients. From 1008 European coeliacs, 61 were identified who neither carry the DQ2 nor DQ8 heterodimers. Fifty seven of these encoded half of the DQ2 heterodimer. The remaining 4 patients had a variety of clinical presentations. Three of them carried the DQA1*01-DQB*05 haplotype as did 20/61 of those carrying neither DQ2 nor DQ8. This may implicate a role of the DQA1*01-DQB*05 haplotype. None of these four patients carried the DQB1*06 allele that has previously been reported in this sub-group of patients. Of the 16 DQ2 heterodimer negative patients without DRB1*04 or DRB1*07 haplotypes, it was inferred that none encoded the previously implicated DRB4 gene as none had a DRB1*09 haplotype. These results underline the primary importance of HLA-DQ alleles in susceptibility to celiac disease, and the extreme rarity of celiac patients carrying neither the DQ2 or DQ8 heterodimers nor one half of the DQ2 heterodimer alone.     

A high frequency of the A30, B18, DR3, DRw52, DQw2 extended haplotype in Sardinian celiac disease patients: Further evidence that disease susceptibility is conferred by DQ A1*0501, B1*0201

This study characterizes by serological and molecular methods the HLA class I and class II alleles in a group of celiac disease children, their parents and a control group of Sardinian descent. We found the DR3-DQw2 haplotype in all patients which was, in almost all cases (84%), associated with the HLA-A30, B18, DR3, DRw52, DQw2 extended haplotype named "Sardinian haplotype" because of its frequency (12-15%) in this Caucasian population. This is the first time that this DQw2-linked haplotype has been reported with such a high frequency in CD. However, no different distribution of "Sardinian haplotype" was found comparing CD patients with 91 haplotyped DQw2-positive controls. This finding indicates that the DQw2 antigen in Sardinians is almost always associated with the A30, B18, DR3, DRw52, DQw2 extended haplotype. The DQA1 and DQB1 second exon sequence analysis of the B18,DR3 and B8,DR3 haplotypes showed the DQA1*0501 and DQB1*0201 alleles which shared the already published sequences. DPB1 subtyping showed the DPB1*0301 allele more frequently (p less than 0.005) in CD patients but this difference was no longer significant when patients and controls, both heterozygous for the DR3-DQw2 haplotype, were compared. We suggest that the divergent HLA extended haplotypes and DP allele associated with CD, described in different Caucasian populations, can be explained by the particular DQw2 linkage disequilibrium in each population.     

Asp57-negative HLA DQ beta chain and DQA1*0501 allele are essential for the onset of DQw2-positive and DQw2-negative coeliac disease.

The genetic predisposition to coeliac disease is associated with the HLA DQw2 allele. Coeliac patients lacking the DQw2 allele are very rare and always exhibit the DR4-DQw3 haplotype. We performed oligotyping of polymerase chain reaction (PCR)-amplified DQA1 and DQB1 genes in six DQw2-negative and 30 DQw2-positive coeliac patients. The DQB analysis showed that all six DQw2-negative patients possessed the DQB1*0302 allele. The other DQB alleles found in five of these patients were DQB1*0501, DQB1*0604 and DQB1*0302. The DQ beta chains encoded from all these alleles have the replacement of aspartic acid residue at position 57 (Asp57), as well as the DQB1*0201 allele which was found in all 30 DQw2-positive coeliac patients. The DQw2-negative proband who lacked the homozygous Asp57 replacement exhibited the DQA1*0501 allele in the DQA1 gene. The DQA1*0501 allele was also found in 27 of the 30 DQw2-positive coeliac patients. Among this group of coeliacs, the four cases lacking the DQA1*0501 allele exhibited the homozygous Asp57 replacement in the DQ beta chain. Our results indicate that Asp57-negative DQ beta alleles are involved in both DQw2-positive and -negative coeliac patients. Moreover, when the Asp57-negative DQ beta chain is encoded from only one of the two DQB1 genes the DQA1*0501 allele is always present.     

Substitution of aspartic acid at beta57 with alanine alters MHC class II peptide binding activity but not protein stability: HLA-DQ (alpha1*0201, beta1*0302) and (alpha1*0201, beta1*0303)

In class II major histocompatibility complex (MHC) proteins, residue beta57 is usually aspartic acid. Alleles carrying serine, valine, or alanine at this position are strongly correlated with the development of insulin-dependent diabetes mellitus (IDDM). Asp(beta)57 participates in a conserved salt bridge that bridges the alpha and beta subunits in the peptide-binding site. It has been proposed that the correlation between IDDM and MHC alleles lacking Asp(beta)57 may be due to an instability of the protein caused by loss of this salt bridge. Using a pair of HLA-DQ proteins (alpha1*0201, beta1*0302) and (alpha1*0201, beta1*0303) differing only in having aspartic acid or alanine at position beta57, we show that the polymorphism does not have a significant effect on protein stability for either the empty or peptide-loaded forms. However, the circular dichroism spectra indicate that empty and peptide-loaded Alabeta57 proteins display slightly different secondary structures relative to their Aspbeta57 counterparts. A set of three peptides shows different binding affinities for DQ(alpha1*0201, beta1*0302) relative to DQ(alpha1*0201, beta1*0303). We propose that substitution of Asp(beta)57 residue causes a local rearrangement within the DQ peptide-binding site that alters the peptide-binding specificity. This rearrangement may help to explain the previously observed differences in SDS stability between Asp and non-Asp(beta)57 DQ proteins.     

Binding of peptides to HLA-DQ molecules: peptide binding properties of the disease-associated HLA-DQ({alpha}1*0501, {beta}1*0201) molecule

Peptide binding to DQ molecules has not previously been described.Here we report a biochemical peptlde-blndlng assay specificfor the BQ2 [I.e. DQ(1^*0501, ß1^*0201)] molecule. Thismolecule was chosen since It shows a strong association to diseasessuch as celiac disease and insulin-dependent diabetes mellitus.Initially we radlolabelled some selected peptides and testedthem for binding to affinity-purified DQ2 molecules. One ofthe peptides, a Mycobacterium bovis (MB) 65 kDa 243–255Ypeptide, displayed a good slgnal-to-noise ratio and was thuschosen as an indicator peptide in the DQ2 binding assay. TheMB 65 kDa 243–255Y peptide bound to DQ2 In a strictlypH-dependent fashion, with optimal binding around pH 5 and onlyweak binding at pH 7.4. The association of the MB 65 kDa 243–255Ypeptide to DQ2 was slow, but once formed, the peptide-HLA complexeswere very stable. The binding of peptides to DQ2 was specific,as shown in Inhibition experiments with a panel of 47 peptides,differing in length, sequence, and origin. The binding of peptidesto DR3 was tested in a similar assay with a Mycobacterium tuberculosis65 kDa 3–13 peptide as the binding indicator. DQ2 andDR3 molecules bound to different sets of peptides. However,the peptide binding to DQ2 and DR3 showed, In general, similarcharacteristics with respect to pH dependence and kinetic parameters,Indicating that the overall rules for peptide binding to DQmolecules are the same as those previously shown for human DRand murlne I-A and I-E molecules.     

The peptide binding motif of the disease associated HLA-DQ (α 1* 0501, β 1* 0201) molecule

To identify the binding motifs of peptides which bind to the celiac disease and insulin-dependent-diabetes-mellitus (IDDM)-associated DQ2 molecule, peptides were eluted from affinity-purified DQ2 molecules. The eluted peptides were separated by reverse-phase HPLC. Prominent peptide peaks and the remaining pool of peptides were sequenced by Edman degradation. Truncated variants of eight different peptides with a length of 9–19 amino acids were identified; among them class II-associated invariant chain peptides (CLIP) and peptides that stem from HLA class I α, HLA-DQα1*0501, Ig and CD20 molecules. Data from the pool sequencing and the biochemical binding analyses of synthetic variants of an eluted high-affinity ligand (HLA class I α 46–60), indicate that the side chains of amino acid residues at relative position P1 (bulky hydrophobic), P4 (negatively charged or aliphatic), P6 (Pro or negatively charged), P7 (negatively charged) and P9 (bulky hydrophobic) are important for binding of peptides to DQ2. Computer modeling of the DQ2 with variants of the high-affinity ligand in the groove suggests that peptides bind to DQ2 through the primary anchors P1, P7 and P9 and making additional advantageous interactions using the P4 and P6 positions.     

Dermatitis herpetiformis and celiac disease are both primarily associated with the HLA-DQ (α1*0501, (β1*02) or the HLA-DQ (α1*03, (β1*0302) heterodimers

HLA-DRB1, -DQA1, and -DQB1 genomic typing of 50 patients with dermatitis herpetiformis and of 290 healthy blood donors was performed. Genes encoding the DQ (α1*0501, (β1*02) heterodimer were carried by 43 (86%) of the patients and 72 (25%) of the controls. Of die remaining seven patients six (12% of all the patients) carried genes encoding the DQ (α1*03, β1*0302) heterodimer. These HLA associations are very similar to those observed in patients with celiac disease. We thus conclude that dermatitis herpetiformis and celiac disease are associated to the very same HLA-DQαβ heterodimers.     

A study of HLA-DR and -DQ alleles in 588 patients and 562 controls confirms that HLA-DRB1*03 is associated with recurrent miscarriage

BACKGROUND: Previous studies have demonstrated an association between recurrent miscarriage (RM) and the maternal HLA-DRB1*01 and -DRB1*03 alleles. The primary aim of the present study was to confirm or reject the hypothesis about this association in a larger case-control study. METHODS: HLA-DRB1, -DQA1 and -DQB1 genotyping was carried out by the PCR-sequence-specific primer (SSP) method in 354 patients with unexplained RM and 202 fertile controls. These results were combined with the results from a previous study of 234 RM patients and 360 controls. RESULTS: The prevalence of patients with HLA-DRB1*03 was significantly increased compared with controls [odds ratio (OR) = 1.4, 95% confidence interval (CI) = 1.1-1.9, P = 0.01, P corrected for the number of comparisons (Pc) = 0.02]. In patients with at least four previous miscarriages or with secondary RM, the association became even stronger (OR = 1.8, 95% CI = 1.3-2.5, P = 0.0005, Pc = 0.004; and OR = 1.8, 95% CI = 1.3-2.5, P = 0.0007, Pc = 0.006, respectively). There was no significant difference between patients and controls with regard to HLA-DRB1*01. CONCLUSION: The HLA-DRB1*03 allele or genes in linkage disequilibrium with it confer susceptibility to RM.     

A case-control study of tyrosine phosphatase (PTPN22) confirms the lack of association with Crohn's disease

In Crohn's disease (CD), the whole gastrointestinal tract can be affected by discontinuous and transmural inflammation. The terminal ileum and colon are especially prone to inflammation that comprises granulomata and later intestinal and perianal fistulas. Genome-wide linkage and epidemiological studies established genetic predisposition factors to CD. Recently, a variation of the intracellular protein tyrosine phosphatase nonreceptor-type 22 (PTPN22) was associated with several autoimmune diseases. Here, we analysed the functionally relevant polymorphism R620W (rs 2476601) of the PTPN22 gene in 146 patients suffering from CD using restriction fragment length polymorphism (RFLP) analyses. This study revealed evidence that PTPN22 variation may have no influence in the genetic predisposition to CD, at least not in another well-characterized Caucasian cohort.     

Distribution of HLA class II alleles among Scandinavian patients with systemic lupus erythematosus (SLE): An increased risk of SLE among non[DRB1*03,DQA1*0501,DQB1*0201] Class II homozygotes?

HLA-DRB1, -DRB3, -DQA1 and -DQB1 alleles were determined by DNA typing in 51 Scandinavian patients with systemic lupus erythematosus (SLE) and 129 controls. DRB1*03,DRB3*0101,DQA1*0501,DQB1*0201 were significantly increased in the patient group, with relative risks (RR) of 2.80, 3.07, 3.55 and 2.12, respectively. These alleles are in strong linkage disequilibrium, and their possible relative contributions in predisposition to SLE are difficult to distinguish. The strongest association was found for DQA1*0501, which is in linkage disequilibrium with DRB1*03 as well as DRB1*11,12 (DR5). An increased frequency of DRB1*11,12 was observed (RR = 1.89, ns). No association with DRB1*15,16 (DR2) was found. The patients had a higher frequency of HLA class II homozygosity than the controls (RR = 5.05, p = 0.0005). When compared to the low-risk group (nonDRB1*03 class II heterozygotes), the cases homozygous for DRB1*03,DQA1*0501,DQB1*0201, known to be in linkage disequilibrium with the complement allele C4A*Q0, had the highest relative risk of developing SLE (RR = 16.39, p = 0.0002). However non[DRB1*03,DQA1*0501,DQB1*0201] class II homozygotes had a higher relative risk (RR = 4.68, p = 0.0147) than DRB1*03,DQA1*0501,DQB1*0201 heterozygotes, known to carry the C4A*Q0 allele (RR = 2.72, p = 0.0088). This may suggest that HLA class II molecules are directly involved in susceptibility to SLE.     

A family-based study does not confirm the association of MYO9B with celiac disease in the Italian population

Association between Myosin IXB (MYO9B) gene polymorphisms and celiac disease (CD) was recently detected by a case-control association study in the Dutch, but not confirmed in the British and Swedish/Norwegian populations. We tested the association between CD and the three most associated single nucleotide polymorphisms (SNPs) in the Dutch study by the transmission disequilibrium test in the Italian population. A total of 252 pediatric patients and 504 parents were genotyped. No transmission distortion was detected either for the single SNPs or for their haplotypic combinations. Control allele frequencies, calculated from untransmitted alleles, were significantly different from those of the Dutch control population. Conversely, allele frequencies were very similar in Italian, British, Swedish/Norwegian and Dutch patients. In conclusion, MYO9B is not involved in CD susceptibility in the Italian population. The difference with the Dutch result might be explained by an imperfect selection of the Dutch controls.     

Diverse locations of amino acids in HLA-DR beta chains involved in polymorphic antibody binding epitopes on DR(alpha, beta 1*0101), DR(alpha, beta 1*1101), and DR(alpha,beta 3*0202) molecules.

In a previous study, we used transfectants expressing hybrid HLA-DR(beta 1*0403)/DR(beta 1*0701) chains to map sequences involved in polymorphic antibody binding epitopes on DR(alpha, beta 1*0403) or DR(alpha, beta 1*0701) molecules. Amino acids 1-40 of the beta 1 domain were found to make the major contributions to most of the antibody binding epitopes studied. To begin to localize sequences that contribute to polymorphic antibody epitopes on DR(alpha,beta 1*0101), DR(alpha,beta 1*1101) and DR(alpha,beta 3*0202) molecules, we used indirect immunofluorescence and flow cytometry to assess the binding of mAb to transfectants expressing hybrid DR(beta 1*0101)/DR(beta 1*1101) or DR(beta 1*1101)/DR(beta 3*0202) chains that divide the DR beta chain into three segments: amino acids 1-40, 41-97, and the beta 2 domain. The results indicate that amino acids 41-97 of the beta 1 domain on DR(beta 1*0101), DR(beta 1*1101), or DR(beta 3*0202) are critical in most of the epitopes, including those recognized by human antibodies MP4 and MP12, and mouse mAb GS88.2, I-LR1, 21r5, and 7.3.19.1, whereas amino acids 1-40 of DR(beta 1*1101) are critical in the epitope recognized by the MCS-7 mAb, and both segments 1-40 and 41-97 of DR(beta 1*1101) are important in the epitopes recognized by the I-LR2 and UL-52 mAbs. Based on these data and comparison of DR beta allelic protein sequences, the residues that may play critical roles in these antibody binding epitopes are predicted.     

The presence of gliadin specific, HLA DQ2 restricted T cells in the small intestinal mucosa is a common phenomenon of celiac disease.

The gastrointestinal disorder celiac disease is strongly associated with genes encoding the HLA-DQ2 heterodimer, DQ(1^*0501,β1^*02). We have isolated activated lamina propria derived lymphocytes from small intestinal biopsies of 20 DR3-DQ2+ celiac disease patients (17 treated and 3 untreated) by immunomagnetic positive selection of CD25+ cells after in vitro challenge of biopsies with wheat flour gliadin protein. Gliadin-specific, polyclonal T cell lines were established from all of the 20 patients. Inhibition studies of T cell lines with anti-HLA monoclonal antibodies indicated predominant antigen presentation by DQ2 in patients. Furthermore we generated 9 gliadin-specific T cell clones from 5 patients which all were DQ2 restricted. These data confirm our previous reports in a more limited material of preferential DQ2-restriction of gliadin-specific T cells from the small intestine. In conclusion, our findings strongly indicate that gliadin-specific T cells is a common phenomenon in the small intestinal mucosa of celiac disease patients and they support the notion that most of them recognize gliadin-peptide when presented by the disease associated HLA-DQ heterodimer (1^*0501,β1^*02).     

Evolution of T cell receptor (TCR) alpha beta heterodimer assembly with the CD3 complex.

T cell antigen receptors (TCR) are composed of an antigen-recognizing unit, the TCRalpha beta heterodimer, and a signal transduction ensemble, the CD3 complex. Whereas mammals possess three CD3 dimers (delta epsilon, gamma epsilon, and zeta2), birds and amphibians have only two (delta/gamma-epsilon and zeta2). To understand evolutionary changes in TCR/CD3 assembly,a phylogenetic approach was employed to dissect the interaction of TCRalpha beta heterodimers with the CD3 components. While sheep and mouse TCRalpha and TCRbeta chains could replace the corresponding human chains in mutant human T cells to restore surface TCR/CD3 expression and function, chicken TCRalpha, TCRbeta and CD3delta/gamma chains were unable to replace the corresponding human chains in forming a chimeric TCR/CD3 complex. The inability of chicken TCR/CD3 components to replace the human molecules in T cells was found to result from the lack of interaction between chicken TCRalpha beta heterodimers and the human CD3 complex. In contrast, if no CD3 molecules are present (non-T cells), TCRalpha -TCRbeta chain pairing can take place in an apparently non-controlled way. Thus, the TCR-CD3 interactions have changed with the evolutionary divergence of two mammalian CD3gamma and CD3delta genes from a single prototypic chicken delta/gamma gene. Our data suggest that the structures in mammalian TCR.C regions, which distinguish between CD3delta and CD3gamma chains, have evolved with the appearance of two separate CD3delta and CD3gamma functions.     

CD9 is expressed on human endometrial epithelial cells in association with integrins alpha(6), alpha(3) and beta(1)

Recently we reported that CD9 is involved in the invasion of a trophoblast-like choriocarcinoma cell line, BeWo, probably through the regulation of integrin functions. Integrins have also been reported to be expressed in the human endometrium and it has been suggested that they play important roles in blastocyst implantation. This study used immunohistochemistry to investigate the expression of CD9 in the endometrium during the menstrual cycle. CD9 was found to be intensely expressed on the cell surface of the glandular epithelium throughout the menstrual cycle without any apparent differences in staining intensity. In addition, Western blotting analysis of the affinity-purified proteins confirmed that CD9 was associated with integrins beta(1), alpha(3) and alpha(6) in the human endometrium. Therefore it can be concluded that CD9, in association with integrins alpha(6), alpha(3) and beta(1), is a constitutive molecule of the endometrial glandular epithelium. These results also suggest that CD9 may be an important regulator of these integrins in the human endometrium.     

Biochemical analysis of GABA(A) receptor subunits alpha 1, alpha 5, beta 1, beta 2 in the hippocampus of patients with Alzheimer's disease neuropathology

Alzheimer's disease (AD) is characterized by selective vulnerability of specific neuronal populations within particular brain regions. For example, hippocampal glutamatergic cell populations within the CA1/subicular pyramidal cell fields have been found to be particularly vulnerable early in AD progression. In contrast, hippocampal GABA-ergic neurons and receptors appear resistant to neurodegeneration. Despite relative sparing of GABA(A) receptors in AD, it is possible that the specific subunit composition of these receptors may undergo alterations with disease progression. In order to address this issue, we employed quantitative Western blot analysis to examine protein levels of GABA(A) receptor subunits alpha 1, alpha 5, beta 1, beta 2 in the hippocampus of subjects displaying increasing severity of AD neuropathology. Subjects were categorized into three groups based upon Braak staging pathologic criteria: pathologically mild (stages I/II, n=9); moderate (stages III/IV, n=8); and severe (stages V/VI, n=7). Across all subject groups, levels of subunit protein were heterogeneously distributed throughout the five hippocampal subregions analyzed (subiculum, CA1-3, dentate gyrus). Statistical analyses revealed differential preservation of GABA(A) receptor subunits in AD. In particular, alpha 1, beta 1, and beta 2 displayed little difference in protein levels among pathologically mild, moderate, and severe subject groups. In contrast, although relatively modest, protein levels of the alpha 5 subunit were significantly reduced between subjects with severe neuropathology compared with pathologically mild subjects (13.5% reduction). Collectively, our data provide evidence for heterogeneous distribution and relative sparing of GABA(A) receptor subunits in the hippocampus of AD patients.     

Binding of gluten-derived peptides to the HLA-DQ2 (α1*0501, β1*0201) molecule,assessed in a cellular assay

The nature of the immunopathogenic relationship underlying the very strong association of coeliac disease (CD) to the HLA-DQ (A1*0501, B1*0201) genotype is not known, but probably relates to binding of gluten-derived epitopes to the HLA-DQ (α1*0501, β1*0201) heterodimer (DQ2). These epitopes have not yet been defined. In this study we have tested the binding of various gluten-derived peptides to DQ2 in a cellular assay using Epstein–Barr virus (EBV)-transformed B lymphocytes and murine fibroblast transfectants. One of these peptides (peptide A), which has previously been shown to exacerbate the CD lesion in vitro and in vivo, was found to bind to DQ2, albeit only moderately, lending further credence to its possible role in the pathogenesis of CD. The nature of peptide A's binding to DQ2 was explored with truncated and conservative point substituted analogues and compared with the published DQ2 binding motif, the results of which explain the observed level of binding.     

Allele and haplotype frequencies for HLA class II (DQA1 and DQB1) loci in patients with celiac disease from Spain

Celiac disease (CD) is a complex and multifactorial disease, defined as a malabsorptive disorder of the small intestine resulting from ingestion of gluten. Genetic susceptibility to CD has been associated with human leukocyte antigen (HLA)-DQ2 heterodimer, encoded by the DQA1*0501 and DQB1*02 genes. However, HLA risk factors do not explain the whole genetic predisposition: not all DQ2-encoding haplotypes confer equal susceptibility to CD. The aim of the present work was to confirm the aforementioned findings in a southern European population. With this purpose, 136 unrelated children diagnosed with CD were typed at the DNA level for HLA-DQA1 and -DQB1 loci. Patients are currently attended at the Donostia Hospital (province of Guipúzcoa, Spain). HLA class II typing was performed by polymerase chain reaction-sequence specific primer procedures. Conspicuous frequencies of the alleles associated with susceptibility to CD were observed (DQA1*0501: 0.592, DQB1*0201: 0.471). Accordingly, the haplotypes DQA1*0501-DQB1*0201 and DQA1*0201-DQB1*0202 revealed a strong linkage disequilibrium (18.84% and 18.75%, respectively) when compared with the Spanish general population. Of the total sample, 93.4% (127 individuals) were carriers of DQ2 heterodimer, either in homozygosis or in heterozygosis. This percentage coincides with figures reported in previous studies, implying the effect of other genes in the development of CD.     

Alloimmunization to platelet antigen HPA-1a (PIA1) is strongly associated with both HLA-DRB3*0101 and HLA-DQB1*0201.

Antibodies to the platelet HPA-1a antigen can elicit in the newborn a condition known as neonatal alloimmune thrombocytopenic purpura (NAITP). Previous studies based on RFLP analysis showed that 100% of HPA-1a-negative women who produced anti-HPA-1a antibodies (responders) were HLA-DRw52a (DRB3*0101). However, this specificity could also be found in some HPA-1a-negative women not producing anti-HPA-1a antibodies (nonresponders). We have analyzed in detail by PCR-SSOP the HLA-DR, -DQ, and -DP loci of 36 responders and 10 nonresponders. We found that while the allele DRB3*0101 was present in the vast majority of responders (91%), there were exceptions. Furthermore, the DQB1*0201 allele was found to be present in almost all responders (94%), but again was also found in nonresponders. The risk of alloimmunization to HPA-1a in an HPA-1b homozygous mother significantly increases with the presence of either allele, the odds ratio being 39.7 for DQB1*0201 and 24.9 for DRB3*0101. Sequencing of exon 2 of these two alleles from responders indicated no sequence difference when compared with the consensus sequences. This indicates that they do not represent variants when compared with the same alleles found in some nonresponders.