Unprecedented Polymorphism of Mhc-DRB Region Configurations in Rhesus Macaques

Animals

In this study, 210 rhesus macaques belonging to the BPRC self-sustaining colony were analyzed. All animals were serologically typed for their Mamu-A, -B, and -DR Ags, and by pedigree analysis over four generations’ haplotypes could be determined. Selective breeding (consanguineous matings) permitted the establishment of a panel of monkeys fully homozygous for their Mamu-A, -B, and -DR Ags which were used as homozygous typing cells.

DNA isolation and PCR

Genomic DNA was extracted from heparinized blood samples or from immortalized B cells of the rhesus macaques. PCR of exon 2 of Mamu-DRB was conducted with generic primers as previously described (27) (N. Otting, N. G. de Groot, M. C, Noort, G. G. M. Doxiadis, and R. E. Bontrop, manuscript in preparation). For amplification with sequence-specific primers (SSP) of Mamu-DRB6 genes with the 62-bp deletion (25), the following primers were used with restriction sites for SalI and XbaI, respectively: (5′-MDRB6) 5′-TTG GTC GAC GCT AAG TGY GAS TGT CMT A-3′ and (3′-MDRB6) 5′-CTC TCT AGA CCS RYA ATT GTA AYT CTG T-3′.

Denaturing gradient gel electrophoresis

Separation of Mamu-DRB alleles with DGGE was conducted as described, with minor modifications (27). Briefly, GC-clamped PCR products of DRB exon 2 were electrophoresed in a 9% acrylamide (acrylamide-bisacrylamide, 37.5:1) gel with a 40–65% parallel denaturing gradient of urea and formamide at a constant temperature of 57°C for ∼3.5 h. When DNA fragments are electrophoresed through an increasing gradient of denaturants, the alleles dissociate at different rates because of sequence-specific differences in denaturation. As a consequence, each allele migrates at a unique rate through the gel. Alleles could be defined in a parallel run with already described banding patterns and/or in comparison with reamplified ssDNA clones of known content. Bands that could not be identified in the manner described above were cut out of the gel; then the DNA was eluted, reamplified, and subjected to direct sequencing (27).

Cloning and sequencing

M13 cloning and sequencing (31) as well as direct sequencing was achieved as described before on a Perkin-Elmer ABI Prism 310 genetic analyzer with the use of ABI PRISM dRhodamine Terminator Cycle Sequencing Ready Reaction Kit according to the manufacturer’s instructions.

Mamu-DRB6 typing by sequence-specific oligotyping (SSO) and sequencing

SSO was used to descriminate the Mamu-DRB6 alleles containing a 62-bp deletion (25) according to the method described before (32). The biotinylated oligonucleotides used were the following: Mamu-DRB6*0102 (triplet (tr) 47–52), 5′-TT CCA GGA GGT GAG GGA A-3′; Mamu-DRB6*0103 (tr 37–41), 5′-AA CCT GGT CTT CCA C-3′; Mamu-DRB6*0104A (tr 38–42), 5′-CT GCG CTA CAA CAG C-3′; Mamu-DRB6*0104B (insert), 5′-GG GAG GAG AAC CCT G-3′; Mamu-DRB6*0104/0105 (tr 53–57), 5′-CT GGGGTG GCC TGT C-3′; Mamu-DRB6*0105 (tr 39–44), 5′-CG CTT CCA CAG CGA CCT G-3′; Mamu-DRB6* 0106 (tr 49–53), 5′-GC GGT GAT GGA ACT G-3′; Mamu-DRB6*0107 (tr 21–25), 5′-TC TGA GCA GGT GCA G-3′; Mamu-DRB6*0115 (tr 32–36), 5′-CA TAA ACA GGA GGA G-3′. The biotinylated oligonucleotide detecting all Mamu-DRB6 alleles with deletion was 5′-CC TGT CAC AGA RTT ACA ATT-3′.

In the case of ambiguous SSO typing results, direct sequencing or M13 cloning and sequencing of the -DRB6-specific PCR products of exon 2 was performed as described above.

Nomenclature of Mamu-DRB alleles

The allocation of Mhc class II alleles to a certain lineage is based on the similarity of nucleotide sequence motifs within the first part of exon 2 and/or shared clustering in phylogenetic trees. Mamu-DRB alleles that cluster within lineages present in humans and macaques are depicted by the same lineage numbers. The last two or three digits are arbitrary and reflect the order in which the alleles were detected. Alleles with three digits after the lineage number differ from each other only by a synonymous substitution. Loci/lineages for which no apparent human equivalent has been identified are designated by workshop numbers. Alleles that do not cluster within any known equivalent as Mamu-DRB*W2601 and Mamu-DRB*W2701/02 receive a new lineage number.

Mamu-DRB typing and molecular characterization

Serology. Alloantisera allow the serological definition of 13 Mamu-A, 13 Mamu-B, and 9 Mamu-DR specificities (15). The Mamu-DR specificities have been designated Mamu-DR1 to -DR8 and -DR101. Furthermore, another -DR group represents so-called serological “blank” specificities because the corresponding alloantisera are lacking. BPRC houses a self-sustaining colony of rhesus macaques, and four generations of animals have been born in captivity. The serological markers have been shown to segregate according to mendelian rules, and therefore haplotypes could be defined. Based on the combination of different serologically defined Mamu-A, -B, and -DR specificities, at least 264 haplotypes were observed within the breeding colony of ∼800 animals.

DGGE. On the basis of experience in the HLA field, it is known that novel Mhc-DRB haplotypes are frequently detected in individuals that possess rare combinations of class I and class II seromarkers. For that reason, ∼210 pedigreed monkeys with 190 different serologically defined haplotypes were selected to analyze Mamu-DRB region associated polymorphisms by means of DGGE. As in humans, the number of Mamu-DRB genes is not constant (25) and therefore Mamu-DRB DGGE profiles are complex (27). As reference markers, DNA from MHC homozygous typing cell lines (Fig. 1⇓, lanes 2 and 4) was used in combination with DNA from animals with the same serotype (Fig. 1⇓, lanes 1 and 10). This approach demonstrated that each Mamu-DR serotype represents multiple DGGE profiles. In the case of true Mamu-DR homozygous cell lines, the DGGE profile provides information on the number of Mamu-DRB genes present per haplotype (Fig. 1⇓, lanes 2 and 4). If one analyzes Mamu-DR heterozygous animals, the number of Mamu-DRB genes present per haplotype can be deducted from the DGGE profile only if segregation analysis is performed or if DNA obtained from Mamu-DR reference cells is available. For only a few reference cell lines have particular DGGE profiles been matched with alleles (27). If a DGGE profile revealed an unknown pattern, the samples were subjected to sequence analysis to identify the Mamu-DRB alleles in question. In most cases, DGGE analysis and nucleotide sequencing provided unequivocal results. However, some alleles appear to be underrepresented in the PCR amplification and therefore may be missed in DGGE as well as in sequencing analysis. This is the case for some Mamu-DRB6 members (Mamu-DRB6*0102–Mamu-DRB6*0107) that share a deletion of 62 nucleotides spanning the position 181–242 (25). For those Mamu-DRB6 alleles, a specific PCR amplification was performed (SSP), and the PCR products were analyzed by either SSO or sequencing. An overview of the methodologies used to type the 68 Mamu-DRB alleles discussed in this study is given in Table I⇓.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

PCR products of Mamu-DRB exon 2 separated by DGGE. PCR products of -DRB exon 2 of 10 different rhesus macaques separated in DGGE are shown in lanes 1–10. Lane 2, -DRB alleles -DRB6*0101,- DRB1*0309, -DRB*W201 of a Mamu-DR 3,3 homozygous typing cell; lane 4, -DRB alleles -DRB1*0303, -DRB1*1007 of a Mamu-DR1,1 homozygous typing cell.

Analysis of Mamu-DRB haplotypes

From the 264 serologically defined Mamu-A, -B, and -DR haplotypes, 190 were tested for their -DRB content at the molecular level. A Mamu-DRB haplotype was defined when at least one of the following three criteria was met: 1) the rhesus macaque in question was serologically typed as Mamu-A, -B, and -DR homozygous and originates from consanguineous offspring (Fig. 1⇑, lanes 2 and 4); 2) segregation of allelic polymorphism could be followed within a family; 3) the combination of -DRB alleles had been detected in several unrelated rhesus macaques, preferably with a shared serotype. An example is given in Fig. 1⇑, where the -DRB DGGE pattern of three heterozygous rhesus macaques shown in lanes 1, 5, and 10 share the haplotype -DRB1*0303, -DRB1*1007 with the DR 1,1 homozygous monkey (Fig. 1⇑, lane 4).

A total of 33 Mamu-DRB haplotypes can be differentiated which can be divided into 24 -DRB region configurations differing in their number and composition of loci. Table II⇓ shows these Mamu-DR region configurations with their allelic haplotypes listed according to the number of Mamu-DRB genes detected. None of them is identical in its organization to an HLA-DRB equivalent. There are five region configurations comprising only two loci, whereas most of the -DRB region configurations consists of a combination of three or four -DRB genes. Only two -DRB region configurations could be observed with five -DRB loci, one with six and one with seven loci. By convention, in humans every HLA-DRB haplotype encodes one -DRB1 locus (11). This does not hold for the Mamu-DRB haplotypes. On one hand, there are several -DRB region configurations without a conventional -DRB1 locus, namely, groups 2c to 2e, 3f to 3h, 4f, 4g, and 5b. One of them, 2e, appears to have only two macaque-specific loci. On the other hand, there are -DRB groups, 2a, 3a, 3b, 4a, and 4b, encoding two Mamu-DRB1 genes which, however, belong to different lineages (Table II⇓). The Mamu-DRB1 loci/lineages are old entities predating the speciation of the contemporary living primates (17). As such, these loci may have been placed in a cis-configuration due to crossing-over effects (25).

Only one -DRB region configuration, 3a, displays extensive polymorphism, whereas all the others appear to be monomorphic or at best oligomorphic. Within the oligomorphic -DRB region configuration 3e, the haplotype -DRB1*0701, -DRB3*0405, -DRB5*0303 comes from rhesus macaques of Indian origin, whereas the allelic haplotype -DRB1*0702, -DRB3*0406, -DRB5*0305 is observed in a Chinese monkey (Table II⇑). Thus, allelic variation of -DRB haplotypes may reflect to a certain extent different origins of the rhesus macaques. Nevertheless, the -DRB region configuration, 3a, with a high degree of allelic variation and the invariant one, 3c, are both frequent in our colony. Therefore, the frequency of a given -DRB region configuration seems to be independent of its degree of allelic polymorphism.

The majority of -DRB region configurations share the presence of a Mamu-DRB6 gene (Table II⇑). Some of the apparently truncated -DRB regions without a -DRB6 gene (Table II⇑, 2a and 2e;) appear to be present in an extended configuration with an additional Mamu-DRB6 gene (Table II⇑, 3a and 3g;). The same holds for the -DRB region configurations 4a and 7, which are like 3a and 6, respectively, with a duplication of the -DRB6 gene.

The question arises whether all these genes are functional. For some loci, however, it is known that they encode gene products. In 2D gel electrophoresis the Mamu-DRB1*03, -DRB1*10, -DRB1*04, -DRB*W3, -DRB*W4, and -DRB*W5 were shown to encode DR molecules (17, 33). In vitro studies revealed that gene products of -DRB1*03, -DR1*04, and -DRB*W201 are able to present peptides to CD4⁺ cells, the latter being significant as restriction elements in cellular response to conserved regions of the HIV (18, 34). Mamu-DRB3, -DRB4, -DRB5, and -DRB*W101 gene products are also present on the cell surface and may have a role in the alloimmune proinflammatory cytokine response (35). These results indicate that at least 11 Mamu-DRB loci/lineages are important for the immune response.

Conservation of the Mhc-DRB6 pseudogene

The HLA- and Patr-DRB6 genes are considered to be pseudogenes because they lack exon 1 coding for the leader peptide of the corresponding gene product. Moreover, these genes share other features such as premature stop codons that are thought to interfere with proper translation. Some studies have documented, however, the presence of mRNA. This can be explained by the integration of a retroviral insert the long terminal repeat of which provides a promotor and a stretch of hydrophobic amino acids that could function as a leader (36). In addition, transcription and translation of human and chimpanzee -DRB6 genes could be demonstrated (37, 38). The Mamu-DRB6 exon 2 sequences are characterized by various characteristics such as inserts, stop codons, and deletions that would render it as a pseudogene. In this context, it is remarkable that nearly all of the 24 different Mamu-DRB groups harbor at least one -DRB6 gene locus. The Mhc-DRB6 gene has probably been a pseudogene >58 million years (39). This would be consistent with the finding that a high number of Mamu-DRB6 alleles have been identified. An explanation for the maintenance of the -DRB6 gene could be that its gene product has a completely different function from the classical MHC class II gene products. One such possibility would be, for instance, that Mhc-DRB6-encoded peptide segments can be bound by other MHC class I and/or II molecules and as such play a key role in thymic eduction. This would be consistent with the fact that translation has been documented. An alternative explanation is that the Mhc-DRB6 locus is flanked by genes, which are under strong positive selection.

Interspecies comparisons

The primate species whose DR region has been studied most extensively is the human, and five major HLA-DRB region configurations have been defined by molecular mapping studies (11). In humans, the number of Mhc-DRB genes varies from one to four depending on the -DRB region configuration (Fig. 2⇓). In a previous communication, the Mamu-DR region was reported to be subject to contraction and expansion and seven -DRB region configurations were recorded (25). In rhesus macaques the number of -DRB genes appears to vary from two to seven per haplotype (Table II⇑). The detailed study of a much larger group of rhesus macaques allowed the definition of 24 Mamu-DRB region configurations (Fig. 2⇓). No other primate species thoroughly investigated at the population level, such as humans, chimpanzees, and common marmosets, has been documented to possess such a level of variation at the DR region with regard to gene content and make-up. This number of 24 -DRB region configurations seems to be only the tip of the iceberg, because not all -DRB lineages described, for example -DRB*W28 and-DRB*W31, and only 68 -DRB alleles of 116 reported have been allocated to a certain haplotype. Furthermore, five other Mamu-DRB region configurations that apparently are not present in our breeding colony have been defined by another research group (40), bringing the total number of distinct Mamu-DRB region configurations up to 29. For our analyses, monkeys mainly of Indian origin were used. It is anticipated that more Mamu-DRB region configurations will be detected when animals from other geographic areas are studied. The question arises why rhesus macaques may exhibit much more polymorphism concerning their Mhc-DRB region configurations than other primate species. An explanation would be that the older a species is, the more complex the haplotype organization may be. The modern human species is considered to be <250,000 years old, whereas rhesus macaques diverged from other macaque species >700,000 years ago (41, 42). Based on these data, the time period to generate different -DRB region configurations was 3 times as long for the rhesus macaques as for humans. A second explanation is that distinct species utilize different strategies to cope with pathogens. The primate Mhc-DR region is unique with regard to its variation in gene numbers, and allelic variation adds an extra level of complexity. Some primate species such as humans have a low number of -DRB region configurations but display a high level of allelic heterogeneity, e.g., the HLA-DR53 region, comprising 39 alleles (11). Rhesus macaques, on the contrary, have a high number of -DRB region configurations, but allelic variation appears to be relatively low. With regard to allelic variation, the DR region of the chimpanzee seems to mimic the human situation, and until now nine DR region configurations have been detected. Hence, one would predict that primate species possessing limited or no polymorphism of -DRB region configurations would encode highly variable Mhc-DRB genes that are frequently exchanged by recombination. This has been documented for common marmosets, a New World monkey from the neotropical rain forests (32). In addition, the chicken has been reported to possess a “minimal essential Mhc” with two class II genes (43). These examples illustrate that evolutionary strategies for a functional Mhc-DR region are different between distinct species. It is remarkable that even within primate species different approaches are observed.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Organization of the Mhc-DRB region configurations in humans and rhesus macaques. Shared loci/lineages are indicated by identical colors. In the case of HLA-DRB region configurations, the order of genes was determined by genomic mapping (17 ). In the case of the rhesus macaque, neither the exact order of genes nor whether the lineages represent different loci is known.

Furthermore, the present results have importance for the application of rhesus macaques in biomedical research, given that these primates are model species in many types of immunological experiments. This communication clearly illustrates the complexity of the Mamu-DR region, knowledge of which is important to facilitate the proper interpretation of the results of diverse experiments. MhcMamu typing is, for instance, needed to control the degree of MHC matching in transplantation studies and to determine restriction elements in vaccine and disease studies. Therefore, it is essential that concerted efforts will be made to obtain a comprehensive characterization of the entire Mamu region including classical and nonclassical Mhc genes.