Identification of the Tapasin gene in the chicken major histocompatibility complex

 The Tapasin molecule plays a role in the assembly of major histocompatibility complex (Mhc) class I molecules in the endoplasmic reticulum, by mediating the interaction of class I-β₂-microglobulin dimers with TAP. We report here the identification of the Tapasin gene in the chicken Mhc (B complex). This gene is located at the centromeric end of the complex, between the class II B-LBI and B-LBII genes. Like its human counterpart it comprises 8 exons, but features a significantly reduced intron size as compared to the human gene. Chicken Tapasin codes for a transmembrane protein with a probable endoplasmic reticulum retention signal. Exons IV and V, and possibly exon III, code for separate domains that are related to the immunoglobulin (Ig) superfamily (this relationship was so far unrecognized for human Tapasin domain IV which has lost its two cysteines). Two different cDNAs corresponding to the Tapasin gene were isolated, possibly related to alternative splicing events; the Ig-like domain encoded by exon IV is missing in one of the cDNAs, suggesting either that this domain is not necessary for the protein to perform its function, or that the two alternatively spliced cDNAs are translated into two functionally different forms of the protein. Received: 8 July 1998 / Revised: 5 October 1998     

Sequence variation at the major histocompatibility complex DRB loci in beluga (Delphinapterus leucas) and narwhal (Monodon monoceros)

 The variation at loci with similarity to DRB class II major histocompatibility complex loci was assessed in 313 beluga collected from 13 sampling locations across North America, and 11 narwhal collected in the Canadian high Arctic. Variation was assessed by amplification of exon 2, which codes for the peptide binding region, via the polymerase chain reaction, followed by either cloning and DNA sequencing or single-stranded conformation polymorphism analysis. Two DRB loci were identified in beluga: DRB1, a polymorphic locus, and, DRB2, a monomorphic locus. Eight alleles representing five distinct lineages (based on sequence similarity) were found at the beluga DRB1 locus. Although the relative number of alleles is low when compared with terrestrial mammals, the amino acid variation found among the lineages is moderate. At the DRB1 locus, the average number of nonsynonymous substitutions per site is greater than the average number of synonymous substitutions per site (0.0806 : 0.0207, respectively;P<0.01). Most of the 31 amino acid substitutions do not conserve the physiochemical properties of the residue, and 21 of these are located at positions implicated as forming pockets responsible for the selective binding of foreign peptide side chains. Only DRB1 variation was examined in 11 narwhal, revealing a low amount of variation. These data are consistent with an important role for the DRB1 locus in the cellular immune response of beluga. In addition, the ratio of nonsynonymous to synonymous substitutions is similar to that among primate alleles, arguing against a reduction in the balancing selection pressure in the marine environment. Two hypotheses may explain the modest amount of Mhc variation when compared with terrestrial mammals: small population sizes at speciation or a reduced neutral substitution rate in cetaceans. Received: 15 July 1997 / Revised: 24 March 1998     

A major histocompatibility complex class I allele shared by two species of chimpanzee

 Little is known regarding the rates at which natural selection can modify or retain antigen presenting alleles at the major histocompatibility complex (MHC). Discovery of identical [1101 base pairs (bp)] coding regions at the MHC class I C locus in Pan troglodytes and Pan paniscus, chimpanzee species that diverged ∼2.3 million years ago, now indicates that a class I allotype can survive for at least this period. Remarkable conservation was also reflected in the (1799 bp) introns where a maximum of only six substitutions distinguished five alleles (three from P. troglodytes and two from P. paniscus) that encoded the identical heavy chain allotype. Analysis of a more distantly related human allele, HLA-Cw ^* 0702, corroborated that intron variation was non-uniform along the gene. Thus we provide a clear reference frame for the lifetime of an MHC class I allotype, a direct estimate of allelic substitution rates, and evidence for an unusual evolution of MHC class I introns. Received: 13 August 1997     

Analysis of polymorphisms in the major expressed class I locus (B-FIV) of the chicken

 We analyzed the polymorphic nature of eleven alleles expressed by the major class I locus (B-FIV) in chickens. Similar to mammalian class I loci, the nucleotide substitutions with high variability occur in exons 2 and 3 encoding the α1 and α2 domains. However, the nonsynonymous to synonymous ratio of nucleotide substitutions in exon 3 encoding the α helix and β sheets is reversed compared with HLA. The region of exon 3 encoding the α2 helix demonstrates a much lower nonsynonymous to synonymous ratio, suggesting evolutionary selection of a more conserved α2 helix in B-FIV compared with HLA. Amino acid residues with high Wu-Kabat variability are typically located in positions predicted to impact antigen presentation. B-FIV amino acid residues predicted to interact with the CDR1α region of the T-cell receptor (Tcr) demonstrate less variability than in mouse and human class I alleles. The combination of a reduced nonsynonymous to synonymous ratio in exon 3 encoding the α2 helix and the limited variability in CDR1α contact residues is discussed with regard to concerted evolution between a minimal major histocompatibility complex and compaction of Tcr variable gene segments in the chicken. Received: 18 Juli 1997 /  Revised: 21 November 1997     

Cloning, localization, and structure of new members of the butyrophilin gene family in the juxta-telomeric region of the major histocompatibility complex

 New members of the butyrophilin (BT) gene family have been identified by cDNA and genomic cloning. Six genes are described: BT2.1, 2.2, 2.3, and BT3.1, 3.2, and 3.3. BT2, BT3, and BT form three distinct subfamilies sharing about 95% amino acid identity at the intra subfamily level and 50% identity at the interfamily level. All the BT2 and BT3 subfamily members map close to BT in the juxta-telomeric region of the major histocompatibility complex. The BT2 members have the canonical structural organization of BT, i.e., two immunoglobulin domains followed by a transmembrane anchor and a B30.2 intracytoplasmic domain. In the BT3 subfamily, only BT3.3 has the structural organization of BT. The two other genes, BT3.1 and BT3.2, code for putative protein without the B30.2 domain. In the case of BT3.2, this is due to an Alu insertion in the B30.2 coding exon, leading to a new polyadenylation site. Received: 21 April 1997 / Revised: 13 June 1997     

Gene conversion of major histocompatibility complex genes is associated with CpG-rich regions

 We examined 32 DNA sequences of mouse and human major histocompatibility complex (MHC) genes believed to have been subjected to gene conversion events. All regions of the mouse H2 genes as well as the human HLA genes which have been implied to be involved in gene conversion events had elevated levels of CpG dinucleotides, whereas the rest of the genes showed extensive CpG suppression. Mouse MHC genes which have been suspected but not directly implied to be involved in gene conversion events also showed elevated levels of CpG dinucleotides. Moreover, both mouse and human MHC genes which have never been suspected of undergoing gene conversion had low levels of CpG throughout the genes. These results indicate that high CpG levels are correlated with gene conversion rather than with polymorphism, as non-polymorphic genes that have been implicated as gene conversion donors also have elevated levels of CpG dimers in the involved regions, whereas polymorphic genes which have never been considered to undergo gene conversion events have a low level of CpG dinucleotides. We also studied the methylation pattern of CpG dimers in the Abk gene by restriction enzyme digestion of mouse testis DNA followed by Southern blot and hybridization to an Abk-specific probe. The examined CpG dimers in prepubescent mice, where the latest germline stages are spermatogonia, leptene, or pachytene, are respectively non-methylated. Accordingly, the CpG dimers appear to be non-methylated in germline DNA from the testis of prepubescent mice, where gene conversions have been reported to occur. Received: 6 May 1998 / Revised: 2 September 1998     

Mhc class I and non-class I gene organization in the proximal H2-M region of the mouse

 A bacterial artificial chromosome (BAC) contig was constructed across the proximal part of the H2-M region from the major histocompatibility complex (Mhc) of mouse strain 129 (H2 ^bc ). The contig is composed of 28 clones that span approximately 1 megabasepair (Mb), from H2-T1 to Mog, and contains three H2-T genes and 18 H2-M genes. We report the fine mapping of the H2-M class I gene cluster, which includes the previously reported M4-M6, the M1 family, the M10 family, and four additional class I genes. All but two of the H2-M class I genes are conserved among haplotypes H2 ^k , H2 ^b , and H2 ^bc , and only two genes are found in polymorphic HindIII fragments. Six evolutionarily conserved non-class I genes were mapped to a 180 kilobase interval in the distal part of the class I region in mouse, and their order Znf173-Rfb30-Tctex5-Tctex6-Tctex4-Mog was found conserved between human and mouse. In this Znf173-Mog interval, three mouse class I genes, M6, M4, and M5, which are conserved among haplotypes, occupy the same map position as the human HLA-A class I cluster, which varies among haplotypes and is diverged in sequence from the mouse genes. These results further support the view that class I gene diverge and evolve independently between species. Received: 27 April 1998 / Revised: 4 June 1998     

The genomic organization and evolution of the natural killer immunoglobulin-like receptor (KIR) gene cluster

Natural killer (NK) immunoglobulin-like receptors (KIRs) are a family of polymorphic receptors which interact with specific motifs on HLA class I molecules and modulate NK cytolytic activity. In this study, we analyzed a recently sequenced subgenomic region on chromosome 19q13.4 containing eight members of the KIR receptor repertoire. Six members are clustered within a 100-kb continuous sequence. These genes include a previously unpublished member of the KIR gene family 2DS6, as well as 2DL1, 2DL4, 3DL1, 2DS4, 3DL2, from centromere to telomere. Two additional KIR genes, KIRCI and 2DL3, which may be located centromeric of this cluster were also analyzed. We show that the KIR genes have undergone repeated gene duplications. Diversification between the genes has occurred postduplication primarily as a result of retroelement indels and gene truncation. Using pre- and postduplication Alu sequences identified within these genes as evolutionary molecular clocks, the evolution and duplication of this gene cluster is estimated to have occurred 30–45 million years ago, during primate evolution. A proposed model of the duplication history of the KIR gene family leading to their present organization is presented. Received: 25 November 1999 / Revised: 10 January 2000     

Evolution of the Mhc class I region: the framework hypothesis

 A comparison of the major histocompatibility complex (Mhc) region between human and mouse highlights both stability and differences. The class II and class III regions are orthologous; they probably existed in the ancestor in a similar organization and were not subjected to major rearrangement. The class I genes, by contrast, are definitely paralogous, having been reorganized several times. As long as only class I genes were identified, the class I regions of human and mouse were difficult to compare directly. The identification of non-class I genes has allowed a comparative map to be drawn, which shows that the class I region is orthologous between human and mouse as well. The lack of orthology specifically applies to the class I sequences. However, the comparative map shows that the non-orthologous class I sequences occupy homologous locations with regard to the conserved genes. I propose a model to explain this paradox. The conserved genes may represent samples of a dense "framework" of genes whose alterations are deleterious. The homologous positions occupied by class I genes would thus represent the few permissive places allowing major perturbations. The evolution of the class I sequences, by duplication and deletion, independently in the two species, has taken place within the scope defined by the framework: insertion at the permissive places, and expansion by creation of class I-related DNA by duplication, thus pushing back the boundaries of the framework. Received: 23 March 1998 / Revised: August 14 1998     

Allelic and haplotype variation of major histocompatibility complex class II DRB1 and DQB loci in the St Lawrence beluga (Delphinapterus leucas)

In order to assess levels of major histocompatibility complex (Mhc) variation within the St Lawrence beluga (Delphinapterus leucas) the variation at the beluga Mhc DRB1 class II locus was assessed by single-strand conformation polymorphism (SSCP) analysis of the peptide-binding region for 313 whales collected from 13 sampling locations across North America. In addition, samples from west Greenland and the St Lawrence were also typed at the DQB locus, allowing comparison to a previous study and assessment of linkage disequilibrium of alleles at the two loci. Comparisons of DRB1 and DQB allele frequencies among all sampling locations indicated genetic structure (α < 0.005). Most of this structure resulted from differences between the different wintering groups. Significant genetic structure (α = 0.05) exists among each pair of the following groups at both the DRB1 and DQB loci; St Lawrence, Hudson Strait, Bering Sea, Cunningham Inlet, and Davis Strait (minus Cunningham Inlet), except the St Lawrence and Hudson Strait for the DQB locus. In the St Lawrence population, six of the eight DRB1 alleles are present representing all five known allelic lineages. Evidence of linkage disequilibrium between the DRB1 and DQB is present in two sampling locations, the St Lawrence and Nuussuaq (α = 0.05). Analysis of probable DRB1–DQB haplotypes among groups of beluga suggests a haplotype reduction in the St Lawrence.     

Natural selection of the major histocompatibility complex (Mhc) in Hawaiian honeycreepers (Drepanidinae)

The native Hawaiian honeycreepers represent a classic example of adaptive radiation and speciation, but currently face one the highest extinction rates in the world. Although multiple factors have likely influenced the fate of Hawaiian birds, the relatively recent introduction of avian malaria is thought to be a major factor limiting honeycreeper distribution and abundance. We have initiated genetic analyses of class II beta chain Mhc genes in four species of honeycreepers using methods that eliminate the possibility of sequencing mosaic variants formed by cloning heteroduplexed polymerase chain reaction products. Phylogenetic analyses group the honeycreeper Mhc sequences into two distinct clusters. Variation within one cluster is high, with dN > dS and levels of diversity similar to other studies of Mhc (B system) genes in birds. The second cluster is nearly invariant and includes sequences from honeycreepers (Fringillidae), a sparrow (Emberizidae) and a blackbird (Emberizidae). This highly conserved cluster appears reminiscent of the independently segregating Rfp-Y system of genes defined in chickens. The notion that balancing selection operates at the Mhc in the honeycreepers is supported by transpecies polymorphism and strikingly high dN/dS ratios at codons putatively involved in peptide interaction. Mitochondrial DNA control region sequences were invariant in the i'iwi, but were highly variable in the 'amakihi. By contrast, levels of variability of class II beta chain Mhc sequence codons that are hypothesized to be directly involved in peptide interactions appear comparable between i'iwi and 'amakihi. In the i'iwi, natural selection may have maintained variation within the Mhc, even in the face of what appears to a genetic bottleneck.