Immunogenetics

The Simian immunodeficiency virus (SIV)-infected Indian rhesus macaque (Macaca mulatta) is the most established model of HIV infection and AIDS-related research, despite the potential that macaques of Chinese origin is a more relevant model. Ongoing efforts to further characterize the Chinese rhesus macaques’ major histocompatibility complex (MHC) for composition and function should facilitate greater utilization of the species. Previous studies have demonstrated that Chinese-origin M. mulatta (Mamu) class I alleles are more polymorphic than their Indian counterparts, perhaps inferring a model more representative of human MHC, human leukocyte antigen (HLA). Furthermore, the Chinese rhesus macaque class I allele Mamu-A1*02201, the most frequent allele thus far identified, has recently been characterized and shown to be an HLA-B7 supertype analog, the most frequent supertype in human populations. In this study, we have characterized two additional alleles expressed with high frequency in Chinese rhesus macaques, Mamu-A1*02601 and Mamu-B*08301. Upon the development of MHC–peptide-binding assays and definition of their associated motifs, we reveal that these Mamu alleles share peptide-binding characteristics with the HLA-A2 and HLA-A3 supertypes, respectively, the next most frequent human supertypes after HLA-B7. These data suggest that Chinese rhesus macaques may indeed be a more representative model of HLA gene diversity and function as compared to the species of Indian origin and therefore a better model for investigating human immune responses.

The genes of the major histocompatibility complex (MHC) are attractive candidates for investigating the link between adaptive variation and individual fitness. High levels of diversity at the MHC are thought to be the result of parasite-mediated selection and there is growing evidence to support this theory. Most studies, however, target just a single gene within the MHC and infer any evidence of selection to be representative of the entire gene region. Here we present data from three MHC class II beta genes (DPB, DQB, and DRB) for brown hares in two geographic regions and compare them against previous results from a class II alpha-chain gene (DQA). We report moderate levels of diversity and high levels of population differentiation in the DQB and DRB genes (Na = 11, D est = 0.071 and Na = 15, D est = 0.409, respectively), but not for the DPB gene (Na = 4, D est = 0.00). We also detected evidence of positive selection within the peptide binding region of the DQB and DRB genes (95% CI, ω > 1.0) but found no signature of selection for DPB. Mutation and recombination were both found to be important processes shaping the evolution of the class II genes. Our findings suggest that while diversifying selection is a significant contributor to the generally high levels of MHC diversity, it does not act in a uniform manner across the entire MHC class II region. The beta-chain genes that we have characterized provide a valuable set of MHC class II markers for future studies of the evolution of adaptive variation in Leporids.

There is presently much interest in utilizing patterns of linkage disequilibrium (LD) to further genetic association studies. This is particularly pertinent in the class III region of the human major histocompatibility complex (MHC), which has been extensively studied as a disease susceptibility locus in a number of ethnic groups. To date, however, few studies of LD in the MHC have considered non-Caucasian populations. With the advent of large-scale haplotyping of the human genome, the question of utilizing LD patterns across populations has come to the fore. We have previously used LD mapping to direct an MHC class III association study in a UK Caucasian population. As an extension of this, we sought to determine to what extent the pattern of LD observed in that study could be used to conduct a similar study in a West African Gambian population. We found that broad patterns of LD were similar in the two populations, resulting in similar candidate region delineations, but at a higher resolution, marker-specific patterns of LD and population-dependent allele frequencies confounded the choice of regional tagging SNPs. Our results have implications for the applicability of large-scale haplotype maps such as the HapMap to complex regions like the MHC.

We investigated the Nco I restriction fragment length polymorphism (RFLP) of the tumor necrosis factor beta (TNFB) gene in 173 patients with systemic lupus erythematosus (SLE), 192 unrelated healthy controls, and eleven panel families, all of German origin. The phenotype frequency of the TNFB*1 allele was significantly increased in patients compared to controls (63.6% vs 47.1%, RR = 1.96, p <0.002). The results of a two-point haplotype statistical analysis between TNFB and HLA alleles show that there is linkage disequilibrium between TNFB*1 and HLA-A1, Cw7, B8, DR3, DQ2, and C4A DE. The frequency of TNFB*1 was compared in SLE patients and controls in the presence or absence of each of these alleles. TNFB*1 is increased in patients over controls only in the presence of the mentioned alleles. Therefore, the whole haplotype A1, Cw7, B8, TNFB*1, C4A DE, DR3, DQ2 is increased in patients and it cannot be determined which of the genes carried by this haplotype is responsible for the susceptibility to SLE. In addition, two-locus associations were analyzed in 192 unrelated healthy controls for TNFB and class I alleles typed by serology, and for TNFB and class II alleles typed by polymerase chain reaction/oligonucleotide probes. We found positive linkage disequilibrium between TNFB*1 and the following alleles: HLA-A24, HLA-B8, DRB1*0301, DRB1*1104, DRB1*1302, DQA1*0501, DQB1*0201, DQB1*0604, and DPB1*0101. TNFB*2 is associated with HLA-B7, DRB1*1501, and DQB1*0602.

An insertion of 6.4 kb is present in intron 9 of 60% of the human complement C4 genes, as well as in the C4 genes of a number of Old World primates. This insertion has the typical genomic organization of endogenous retroviruses, with the three major genes gag, pol and env flanked by long terminal repeats (LTRs). This human endogenous retrovirus K [HERV-K(C4)] insertion is in reverse orientation to the C4 coding sequence. Using RT-PCR as well as RNase protection assays, retroviral transcripts could be detected in different human cell lines which were only present in the antisense orientation of the retrovirus. Furthermore, C4 expression as well as intermediate transcripts comprising both HERV-K(C4) and C4 coding sequences was observed in these cells. These findings were confirmed using real-time PCR to quantitate the number of specific mRNA transcripts. Using reporter gene assays, it could be demonstrated that only the 3′LTR exhibits promoter activity, but in the sense orientation of the retrovirus. It has been suggested earlier that expression of C4 could lead to the transcription of a retroviral antisense RNA, which might protect against exogenous retroviral infections. In a previous study, it was shown that the expression of retroviral-like constructs was significantly downregulated in mouse cells transfected with human C4 genes, and that this downregulation was further modulated after IFN-γ stimulation of C4 expression. In a new series of experiments, we have now confirmed these observations, using human hepatoma cells constitutively expressing C4. A dose-dependent downregulation of up to 45% caused by hybridization of retroviral sense and genomic HERV-K(C4) antisense RNA was observed. The functional 3′LTR promoter, the presence of retroviral antisense RNA transcripts and the functional detection of HERV-K(C4)-specific antisense activity provide strong evidence for a major role of the HERV-K(C4) insertion in the control of gene expression, resulting in a selective advantage favouring the presence of this element in human and primate C4 genes.

Full-length cDNAs encoding the DQB genes expressed by three BoLA class II haplotypes (DH8A, DH15B, and DH24A) were amplified by reverse-transcription polymerase chain reaction, cloned, and sequenced. The sequence data revealed that the DH8A haplotype expressed two DQB genes (DQB*1005 and DQB*1201) while the DH15B and DH24A haplotypes expressed the same single gene (DQB*0101). Comparison of the three alleles showed that the 3′ untranslated (3′UT) sequence of the DQB*1201 allele contained a duplication of about 200 bp. This repeat was also found in other DQB alleles from cattle and sheep, but only in haplotypes with duplicated DQB genes. This 200-bp repeat and other features of the 3′UT may provide useful markers of DQB evolution, allowing us to distinguish and selectively amplify the different DQB loci.

Previous studies have shown that immunization of MAXX rats with spleen and lymph node cells from the MHC-identical BN strain results in the formation of antibodies that react with the renal endothelial alloantigen Eag-1. In the present study, the reactivity of MAXX anti-BN sera was further characterized. No reactivity of the antisera was detected with unseparated spleen, lymph node, thymus and bone marrow cell suspensions, peripheral blood, or cells obtained from lung lavages. The antisera did, however, react with splenic macrophages, as well as with peritoneal granulocytes and macrophages from BN, BMA, BN.1L, and PVG rats. Genetic studies revealed that the antigen, provisionally designated Pag-1, segregates independently of Eag-1, the RT1 complex, sex, and the hooded and albino traits. Pag-1 appears to be absent in the kidney, since absorption of MAXX anti-BN sera with BN kidney homogenates did not remove the reactivity against Pag-1, and antisera raised against BN peritoneal cells did not bind with the renal endothelium. Pag-1 is expressed on bone marrow-derived cells, since peritoneal cells from lethally irradiated MAXX rats that were reconstituted with bone marrow cells from BN donors reacted with MAXX anti-BN sera, whereas peritoneal cells from BN rats reconstituted with MAXX bone marrow did not.