P. Morales

A new HLA-G allele (HLA-G * 0105N) and its distribution in the Spanish population

Six different HLA-G subtypes ( HLA-G*01011, -G*01012, -G*01013, -G*0102, -G*0103, and -G*0104) have been defined by the WHO Nomenclature Committee for Factors of the HLA System.HLA-G is an interesting nonclassical class I gene with the following characteristics: 1) a low degree of polymorphism (Morales et al. 1993; Yamashita et al. 1996); 2)HLA-G expression is restricted to trophoblast cells where it is the only Mhc molecule expressed; low mRNA transcript levels have been detected in a wide variety of cells in fetal and adult tissues (Onno et al. 1994); 3) alternative splicing of HLA-G transcripts gives rise to four membrane-bound and two soluble isoforms (Fujii et al. 1994); 4) polymorphic changes do not follow the three hypervariable regions per domain model and do not affect either the antigen or T-cell receptor binding sites (Arnaiz-Villena et al. 1996); and 5) stop codons have been found in homozygousity at theMhc-G exon 3 in the Cercopithecinaefamily of primates. Altogether this suggests that alternative splicing forms without exon 3 could play a role in cellular physiology other than antigen presenting (Castro et al. 1996). In the present study a new HLA-G subtype is described in the Spanish population; this allele shows linkage disequilibrium with the HLA A30-B13 haplotype and bears a nucleotide deletion either in the third position of codon 129 or the first of codon 130. Genomic DNA from 114 unrelated healthy Spanish individuals was isolated from peripheral blood lymphocytes using standard methods. Amplification of exon 2, exon 3, or exon 2 + intron 2 + exon 3 was performed using HLA-Gspecific primers as previously described (Morales et al. 1993). A total of 20 samples were further studied by singlestrand conformational polymorphism because they showed blanks or typing anomalies. Four showed a rare mobility pattern; polymerase chain reaction (PCR) products from exon 2 + intron 2 + exon 3 were purified, inserted into the pGEM-T vector, and sequenced in an Applied Biosystems (Foster City, CA) DNA automated sequencer. Two new HLA-G sequences were obtained: one ( HLA-G*0104, present in two unrelated samples) has been recently described in a Japanese population study (Yamashita et al. 1996), and the other (present in two other unrelated samples and namedHLA-G*0105N) has not yet been described. HLA-G*0104 has a non-synonymous leucine to isoleucine substitution at the first base of codon 110 (exon 3) compared withG*01013 (Fig. 1); this change is the first non-synonymous variation found at this codon in humans. This position does not seem to be important in the binding site region for processed antigen or in T-cell receptor interactions. A study of this allele in 114 unrelated Spanish individuals using a PCR-restriction fragment length polymorphism (RFLP) method (Morales et al. 1993) indicates that it is present in 100% of HLA-A23(n = 7)and 70% of HLA-A24 (n = 17)-positive Spaniards and has a frequency of 21.05% (n = 114, see Figure 1 legend). It is also found in four HLA-A19and oneHLA-A2/-A11-positive samples; strong linkage disequilibrium has already been described betweenHLA-A and -G alleles (Morales et al. 1993). The newHLA-G*0105Nallele shows a cytosine deletion either at the third base of codon 129 or the first base of codon 130 (exon 3; Fig. 1). This deletion changes the reading frame from this position, yielding a different amino acid sequence of the protein, which most likely would finish with a stop codon at the beginning of the α3 domain (exon 4), ifHLA-G*0105Nexon 4 DNA sequence is identical toHLA-G*01011 exon 4 DNA sequence; the cytosine deletion at codon 129/130 is not a PCR artefact because it has been observed in two different samples and in different PCR-RFLP assays. Stop codons in homozygousity have been found at exon 3 of Mhc-G sequences from The nucleotide sequence data reported in this paper have been subitted to the GenBank the nucleotide sequence database and have been assigned the accession number L78073. The name G*0105N was offficially assigned by the WHO Nomenclature Committee in October 1996

Allelic diversity at the primate Mhc-G locus: Exon 3 bears stop codons in all Cercophitecinae sequences

Twenty-seven major histocompatibility complex (Mhc)-G exon 2, exon 3, and exon 2 and 3 allelic sequences were obtained together with 12 different intron 2 sequences. Homo sapiens, Pan troglodytes, Pan paniscus, Gorilla gorilla, Pongo pygmaeus, Macaca fascicularis, Macaca mulatta, and Cercopithecus aethiops individuals were studied. Polymorphism does not follow the classical pattern of three hypervariable regions per domain and is found in all species studied; exon 3 (equivalent to the alpha 2 protein domain) shows stop codons in the Cercopithecinae group but not in the Pongidae and human groups. Dendrograms show that cotton top tamarin (Saguinus oedipus) Mhc-G sequences are closer to Homo sapiens and Pongidae than to Cercopithecinae, probably due to the stop codons existing at exon 3 of the latter. There is a clear trans-species evolution of allelism in Cercopithecinae and also in exon 2 of all the other apes studied, but a generation of allelism within each species may be present on exon 3 sequences. This discrepancy may be due to the preferential use of exon 2 over exon 3 at the mRNA splicing level within each species in order to obtain the appropriate functional G product. Mhc-G intron 2 shows conserved motifs in all species studied, particularly a 23 base pair deletion between positions 161 and 183 which is locus specific, and some of the invariant residues, important for peptide presentation, conserved in classical class I molecules from fish and reptiles to humans were not found in Mhc-G alleles; the intron 2 dendrogram also shows a particular pattern of allelism within each species. In summary, Mhc-G has substantial differences from other classical class I genes: polymorphism patterns, tissue distribution, gene structure, splicing variability, and probably an allelism variability within each species at exon 3. The G proteins may also be different. This indicates that the Mhc-G function may not be peptide presentation to the clonotypic T-cell receptor.