|
|
||||||||




* Wisconsin National Primate Research Center, Madison, WI 53706;
University of Wisconsin, Department of Pathology and Laboratory Medicine, Wisconsin National Primate Research Center, Madison, WI 29208; and
Department of Biological Sciences, University of South Carolina, Columbia, SC 53706
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
The supply of Indian Rhesus macaques, however, is dwindling (3). Since 1978, India has banned the export of feral Rhesus macaques (4), making breeding programs the sole source of research animals. The supply of macaques produced by breeding has not kept pace with the demand (5). Additionally, the demand for Indian Rhesus macaques by biodefense and transplant researchers (6) is constraining the availability of animals for HIV research.
HIV vaccine research is disproportionately affected by the Indian Rhesus macaque shortage. Vaccine studies are inherently expansive because populations large enough to detect statistically meaningful differences between vaccinated macaques and vaccine naive controls must be included. For trials studying CD8+ T lymphocyte responses, both groups of animals are often MHC class I (MHC I)2 allele matched to quantitate immune responses against the same epitopes (7, 8). This has exacerbated the shortage and demand for MHC-defined Indian Rhesus macaques (5) because "high frequency" alleles such as Mamu-A*01 are only present in 2025% of macaques (9). Moreover, postacute viral loads are increasingly being used as vaccine trial endpoints (10, 11, 12, 13). These studies are particularly susceptible to differences in MHC I genotypes, as certain alleles, such as Mamu-A*01 (14, 15), Mamu-B*17 (16), and Mamu-A*1303 (17) predispose animals to favorable SIV outcomes.
Fortunately, Indian Rhesus macaques are not unique in their susceptibility to pathogenic SIV infection. Chinese Rhesus macaques can also be infected with SIV (18, 19), though these animals do not possess MHC-I
alleles, such as Mamu-A*01, that are common in Indian Rhesus macaques (20). Cynomolgus macaques (Macaca fascicularis) represent another potential nonhuman primate model for SIV. More than 9,000 Cynomolgus macaques are imported for research each year (T. Demarcus, unpublished observations). Cynomolgus macaques, like Rhesus macaques, are geographically distributed throughout Asia (21). The most widely imported Asian Cynomolgus macaques originate in Vietnam, China, Indonesia, and the Philippines. Though the natural range of Cynomolgus macaques extends throughout Asia, the largest exporter is the Indian Ocean island of Mauritius. Between 5500 and 8000 feral and captive bred Cynomolgus macaques are exported from Mauritius each year (22).
Very little is known about MHC I genes of Cynomolgus macaques. In the 1980s, >30 MHC I serotypes were described (23), but these serotypes were never resolved at the molecular level. One-dimensional isoelectric focusing (1D-IEF) of MHC I proteins immunoprecipitated from Cynomolgus macaques suggests that these animals express between five and seven MHC I proteins (24). Recently, 14 MHC I A locus allele sequences (25) and 26 MHC I B locus alleles (26) were described. Additionally, there is strong evidence that the B locus can be duplicated, as has been previously observed in Rhesus macaques (27).
Others have observed that Cynomolgus macaque origin can influence the success of kidney allografts (28) and susceptibility to Plasmodium coatneyi infections (29). We hypothesized that these outcome disparities are due, at least in part, to differences in the regional distribution of MHC I allele repertoires. If proven correct, the origin of Cynomolgus macaques may be a critical and overlooked variable that prejudices the host response to pathogens including SIV. To test this hypothesis, we identified >60 new MHC I alleles from Cynomolgus macaques of Vietnamese, Chinese, and Mauritian origin. We did not detect the previously described MHC I alleles in any of the animals examined, suggesting that these alleles were identified in animals from a different origin. Mauritian animals exhibited an unexpected degree of allele sharing, prompting the development of genotyping assays for five common alleles. More than 50% of Mauritian Cynomolgus macaques possessed a combination of three MHC I alleles, at least one of which is expressed as protein on the cell surface. We speculate that animals possessing this allele combination may be uniquely valuable for SIV and biodefense projects that require monitoring of CD8+ T lymphocyte responses or MHC I matching.
| Materials and Methods |
|---|
|
|
|---|
Vietnamese origin Cynomolgus macaques (Cy0051-Cy0090) were purchased from Covance Research Products. Chinese origin Cynomolgus macaques (Cy0091-Cy0110) were purchased from Central State Primates. The Vietnamese and Chinese Cynomolgus macaques were housed at the Wisconsin National Primate Research Center. All of these animals were cared for according to protocols approved by the University of Wisconsin Research Animal Resource Committee.
All experiments on Mauritian Cynomolgus macaques were performed on whole blood provided by Charles River Laboratories. Blood from eight monkeys (A1MA8M) was initially obtained for MHC I allele discovery. Subsequently, blood from an additional 48 animals, obtained in three different shipments, was procured and used for MHC I allele genotyping.
MHC I allele cloning and sequencing
MHC I alleles were amplified by RT-PCR, cloned, and sequenced. The criteria that we used to define novel alleles by cloning and sequencing was intentionally stringent; an allele was defined by at least two full-length, nucleotide identical sequences. A sizable fraction of PCR-amplified MHC alleles contain sequence artifacts, including intramolecular recombinants and single nucleotide substitutions (30). To minimize these effects, the PCR conditions were carefully optimized, and high fidelity Phusion DNA polymerase was used in all amplifications. Nonetheless, a number of nonauthentic sequences were obtained from each animal. For example, in animal A1M, 13 clones perfectly matched the Mafa-B*430101 sequence. An additional 16 clones were nonperfect matches to the Mafa-B*430101 sequence that were more closely related to this allele than to any other allele. Authentic MHC I alleles that are not as prominent in our transcript pool (31) may be overlooked with our stringent approach, however, we feel that this is preferable to describing "allele sequences" that may not be biologically relevant.
PBMC were collected from EDTA-treated whole blood by Ficoll-Paque (GE Health Sciences) density gradient purification. Total RNA was harvested from the PBMC with a Qiagen Total RNEasy kit (Qiagen) following the manufacturers instructions. cDNA was synthesized with either SuperScript II or SuperScript III Reverse Transcriptase (Invitrogen Life Technologies). Oligo(dT)-primed reverse transcription was performed according to the manufacturers protocol. The cDNA was amplified using PCR primers designed to maximize amplification of known Indian Rhesus macaque MHC I sequences. Each cDNA was amplified in two PCR, both of which used the same universal forward primer (5'NotI-MHC-full; 5'-GCGGCCGCATGSSSGTCATGGCGCCSSG-3'). The 5'NotI-MHC-full oligonucleotide contains a NotI restriction site in addition to sequence from the first exon of MHC I sequences. Both reverse oligonucleotides, 3'Alocus-KpnI (5'-GGTACCTCACACTTTACAAGCCGTGAGAGACAC-3') and 3'Blocus-KpnI (5'-GGTACCTCAAGCCGTGAGAGACWCATCAGAGCC-3'), contain KpnI restriction sites. These two primers differ in their MHC I exon 7 sequences, as MHC I A locus alleles possess nine additional nucleotides adjacent to the translational stop codon. Because the differences between these two primers are located centrally in the oligonucleotide, both reverse primers amplify MHC I A locus and B locus sequences. The PCR was performed using the DNA polymerase Phusion (Bio-Rad) and the following reaction conditions: 98°C for 30 s, 25 cycles of (98°C for 5 s, 65°C for 1 s, 72°C for 20 s), 72°C for 5 min, 4°C until time-of-use. Amplified cDNA was purified with either a QIAQUICK Gel Extraction kit or a QIAQUICK PCR Purification kit according to the manufacturers protocol. The purified DNA was ligated into pCR-Blunt using the Invitrogen Zero Blunt cloning kit (Invitrogen Life Technologies). Between 48 and 192 clones/animal were miniprepped and sequenced using the primers T7 (5'-TAATACGACTCACTATAGGG-3'), M13 (5'-CAGGAAACAGCTATGAC-3'), 5'RefStrand (5'-GCTACGTGGACGACACGC-3'), and 3'RefStrand (5'-CAGAAGGCACCACGACAGC-3') on an ABI 3730 DNA Analyzer (Applied Biosystems). Sequences were analyzed using software from Genecodes and Accelrys. Novel MHC I sequences were given GenBank accession nos. AY958087AY958152.
Phylogenetic tree construction
Phylogenetic trees of the MHC I alleles were constructed by the neighbor-joining method (32) based on Kimuras two-parameter distance (33).
Generation of custom DNA size standards
A ROX-labeled size standard was prepared using amplicons based on the vector pcDNA3.1+ (Invitrogen Life Technologies). A universal ROX-labeled primer ROX-pcDNA3.1 + 2215-F (5'-ROX-AGACAATCGGCTGCTCTGAT-3') and a series of nested, unlabeled oligonucleotides were used to produce fragments of 110 (5'-ctcgtcctgcagttcattca-3'), 203 (5'-caatagcagccagtcccttc-3'), 310 (5'-gtagccggatcaagcgtatg-3'), 411 (5'-cctgatgctcttcgtccag-3'), 525 (5'-ggccattttccaccatgata-3'), and 621 (5'-cgccaagctcttcagcaata-3') bp. Twenty-six larger size fragments, irrelevant to the MHC I genotyping, and ranging from 661 to 2324 bp, were also included in the size standard.
MHC I genotyping
Genotyping was performed by reference-strand conformational analysis (RSCA). Allele clones containing Mafa-A*200101 (GenBank no. AY958088), Mafa-B*370101 (GenBank no. AY958132), and Mamu-B*07 (GenBank no. U41829) were PCR amplified using the labeled primer 6FAM-5'ShortRSCA (5'-[6FAM]AGGGGCCGGAGTATTGGG-3') and the 5' phosphate-modified primer Short3'RSCA-P (5'-[Phos]TTCAGGGCGATGTAATCC-3'). Clones containing Mafa-B*430101, -B*440101, -B*460101, -B*470101, and -B*510101 were PCR amplified using the 5' phosphate modified primer 5'ShortRSCA-P (5'-[Phos]AGGGGCCGGAGTATTGGG-3') and the unlabeled primer Short3'RSCA (5'-TTCAGGGCGATGTAATCC-3'), generating a 217-bp amplicon. cDNA for genotyping was prepared as described above and was amplified with the 5'ShortRSCA-P and Short3'RSCA primers. All PCR were performed using the DNA polymerase Phusion and the reaction conditions described above. Following PCR, the antisense strands of the reference strand amplicons and the sense strands of the clone and cDNA amplicons were digested with
exonuclease (Epicentre Technologies) according to the manufacturers protocol.
exonuclease selectively digests phosphorylated strands of dsDNA, reducing the formation of homoduplex and nonlabeled heteroduplex products in the heteroduplexing reaction. To form heteroduplexes, 1.5 µl of exonuclease treated, FAM-labeled reference strand was mixed with either 1.5 µl of exonuclease-treated product amplified from a DNA clone or 3 µl of exonuclease-treated product amplified from Cynomolgus macaque cDNA. The heteroduplexing reactions were heated to 95°C for 4 min, cooled to 55°C for 5 min, and chilled to 15°C for 5 min. Eight microliters of the custom ROX size standard were added to each heteroduplex reaction, and these mixtures were purified over Sephadex-G50 columns and concentrated by vacuum. The concentrated products were resuspended in 2.0 µl Dextran Blue/EDTA solution, and 1.5 µl of the resuspension was applied to a 96 lane RapidLoad 2.0 membrane comb (Gel Company). The heteroduplexes were resolved on a 3% nondenaturing acrylamide gel run on an ABI 377 DNA Analyzer (Applied Biosystems). Electrophoresis was performed for 12 h at 1200 V, with the gel temperature set to 30°C.
After extracting data from the ABI 377 raw data files using Genescan 3.1 (Applied Biosystems), the Dax software package (Van Mierlo Software Consultancy) was used for all fragment analysis and sizing. Peak mobilities were converted from seconds to apparent base pairs by fitting the samples to a standard curve generated by the custom ROX size standard, thus correcting for lane-to-lane variation in electrophoretic distance. Depending on the allele: reference strand combination, peaks with an apparent base pair size of between ± 0.25 and ± 1.0% of the clone DNA heteroduplex were scored as positive. The variable threshold for peak qualification is necessary because certain allele: reference-strand heteroduplexes exhibit greater intrinsic variation in their mobilities. After automatic scoring by DAX, a list of peak matches was imported into Microsoft Excel (Microsoft). Animals that scored positive for a given allele in all three reference strands were automatically considered positive for that allele. Fragment profiles from animals that scored positive in two of three reference strands for a particular allele were manually reexamined for peaks that may have been missed by the automatic peak detection (e.g., shoulder peaks adjacent to large, primary peaks).
To estimate the total number of Mauritian Cynomolgus macaque MHC class I alleles, we reexamined the RSCA data from all three reference strands, grouping the heteroduplex peaks from all 56 animals by apparent nucleotide size. Peaks with an apex intensity 20-fold greater than background were used in this analysis. Those peaks whose mobilities differed from one another by <0.5% were categorized as belonging to a single allele.
Determination of MHC I cell surface expression
1D-IEF separates the entire complement of expressed MHC I alleles on the basis of differential endpoint migration of individual
-chains in a discontinuous pH gradient within a polyacrylamide gel. The procedure was performed essentially as described in Watkins et al. (34). A stable transfectant of Mafa-B*440101 was produced in 721.221 cells as described previously (35). B lymphoblastoid cell lines were derived from Cynomolgus macaque PBMCs and from the Mafa-B*440101 transfectant. A total of 2 x 107 cells/sample was metabolically labeled with 35-S. The labeled cells were lysed and immunoprecipitated with w6/32 Ab. The immunoprecipitates were split into two fractions, and half were treated with neuraminidase type VIII to digest sialic and neuraminic acid moieties that decorate MHC molecules on the cell surface. The immunoprecipitates were loaded on a pH discontinuous acrylamide gel for isoelectric focusing. The gels were dried and MHC class I signal was detected by autoradiography following a 3- to 7-day exposure to film.
The alleles Mafa-A*230101, -A*350101, and -B*310101 are identical to Indian rhesus alleles Mamu-A*11, -A*13, and -B*01, respectively. We do not currently know whether these matches are due to subtle experimental contamination with rhesus samples or true trans-species allele sharing.
| Results |
|---|
|
|
|---|
MHC class I alleles identified in Rhesus macaques share significant homology at both the 5' and 3' termini. Therefore, we designed primers against these regions to amplify nearly full-length MHC I alleles from cDNA of Cynomolgus macaques. The PCR primers were broadly cross-reactive and successfully amplified MHC I alleles from Rhesus macaque, Cynomolgus macaque, Pig-tailed macaque (Macaca nemestrina), and human cDNA (data not shown). The PCR amplifications used the high-fidelity DNA polymerase Phusion, few amplification cycles (<25), and brief primer annealing steps (1 s) to minimize the formation of PCR amplification artifacts (30).
Between 48 and 192 individual amplicons/animal were cloned and sequenced. At least eight macaques were examined from each origin. We operationally defined MHC I alleles on the basis of at least two complete, identical sequences, although most alleles were found in at least three clones (36). Sixty-six MHC I alleles were identified. Thirty-five of these alleles were homologues of Rhesus MHC I A locus alleles, and 31 were MHC I B locus homologues. We adopted a modified nomenclature for naming the newly discovered alleles that prioritizes amino acid identity in the
1 and
2 domains that control peptide binding and TCR recognition (37). Although previous nonhuman primate MHC I nomenclatures are based either on ad hoc, arbitrary designations of sequence novelty or concordance with serological data (25, 27, 38), the goal of the new nomenclature is to cluster alleles with identical peptide-binding domains within the same top-level domain. In this nomenclature, any nonsynonymous variation in the sequence encoding the
1 and
2 domains is given a new top-level designation (e.g., Mafa-A*150101 vs Mafa-A*200101). Alleles that have nonsynonymous sequence variation outside of the regions encoding
1 and 2, but sequence identity within
1 and
2 are given a new subgroup designation (e.g., Mafa-A*150101 vs Mafa-A*150201) while any synonymous variants of a given allele are sub-subgrouped (Mafa-A*150101 vs Mafa-A*150102). By naming alleles in this way, related alleles with identical peptide-binding specificities are grouped.
MHC I allele sharing in Cynomolgus macaques from different origins
We identified 35 MHC I alleles in Vietnamese origin macaques, 25 in Chinese origin macaques, and 17 in Mauritian origin macaques. None of the alleles identified by Uda et al. (25, 26) were found in animals from these three origins. Moreover, there were no overlapping alleles present in animals from all three origins that we tested. The alleles Mafa-A*170101, -A*380101, -B*290101, -B*310101, -B*360101, -B*380101, -B*380301, -B*400101, and -B*410101 were found in both Vietnamese and Chinese origin macaques, while only Mafa-A*250201 was found in both Chinese and Mauritian origin macaques (Table I). Vietnamese and Mauritian origin macaques did not share any MHC class I alleles. From this analysis, we conclude that the MHC I allele repertoires of Cynomolgus macaques are largely origin specific.
|
|
|
Common MHC class I alleles, analogous to the Mamu-A*01 allele in Indian Rhesus macaques, will be valuable in the development of Cynomolgus macaques as alternative nonhuman primate models for AIDS. Approximately 20% of captive bred Indian Rhesus macaques express Mamu-A*01. We are particularly interested in common MHC class I alleles that exist in more than one regional population and alleles that are exceptionally common within a single population. Twenty-five alleles were found in two or more of the Cynomolgus macaques studied. Of these 25 alleles, five were found in multiple animals from different origins. An additional 6 alleles were present in at least three animals from a single population. A subset of these alleles, Mafa-A*250201, -B*430101, -B*440101, and -B*460101, were identified in >50% of the animals examined. Remarkably, all of these alleles were found in Mauritian Cynomolgus macaques (Table II). Mafa-B*430101, -B*440101, and -B*460101 are consistently detected in combination; the same five animals possessed all three of these alleles.
|
We then determined the frequency of the common MHC I alleles in a cohort of 48 additional Mauritian origin macaques. It is difficult to develop allele-specific primers for allele-specific PCR in the absence of a reasonably complete allele database. Therefore, we used RSCA for genotyping (39). RSCA assesses the unique heteroduplex conformation that is assumed by a sequence-mismatched MHC I allele and a fluorescently labeled reference strand. A critical advantage of this method is that every heteroduplex will have a characteristic mobility signature. Although multiple alleles can potentially form identically migrating heteroduplexes with a single reference strand, it is unlikely that these migration similarities will be maintained if different reference strands are used. Therefore, increasing the number of reference strands improves the reliability of RSCA interpretation. Each RSCA experiment was internally controlled by including heteroduplexes of individual allele clones along with amplified cDNA from the eight animals used for allele identification. Three RSCA genotyping for the common alleles Mafa-B*430101, -B*440101, and -B*460101, as well as for the less common alleles -B*470101 and -B*510101, matched the typing obtained by sequencing individual MHC clones (Fig. 2). The exceptionally high frequencies of Mafa-B*430101, -B*440101, and -B*460101 were verified in the additional 48 animals. More than 50% of the animals possessed Mafa-B*430101 (59%), -B*440101 (63%), and -B*460101 (53%). Moreover, the combination of Mafa-B*430101, -B*440101, and -B*460101 was found in 52% of the animals. The alleles Mafa-B*470101 and Mafa-B*510101 were found in 20 and 29% of the animals, respectively (Fig. 3). Therefore, we conclude that extraordinary MHC I allele sharing is a characteristic feature of Mauritian Cynomolgus macaques.
|
|
The previous analyses measured the frequency and sequence of transcribed MHC I alleles. There is still considerable uncertainty about the expression of macaque MHC I genes. Sequencing of the complete Rhesus MHC region identified >20 potentially active MHC I genes, but it appears that only a subset of these genes are transcribed (40). Furthermore, it is possible that only a subset of the transcribed genes are translated, or that many of the transcribed genes are expressed at very low levels (23).
Fewer total MHC class I alleles were identified in Mauritian Cynomolgus macaques than in either Chinese or Vietnamese animals, but it is unlikely that our approach uncovered all of the alleles given in each population. We estimated the sensitivity of allele discovery by examining how many heteroduplex peaks are present in each animals RSCA profile. In the eight Mauritian Cynomolgus macaques for which both clone sequence and RSCA are available, an average of 10.1 heteroduplex peaks were identified by RSCA, while an average of only 5.3 alleles were identified by clone sequencing.
How many alleles would be present in a complete Mauritian Cynomolgus macaque MHC I database? We counted the total number of unique RSCA heteroduplex signatures in the 56 genotyped macaques. A total of 37, 36, and 39 heteroduplex peaks formed with the Mafa-A*200101, Mamu-B*07, and Mafa-B*370101 reference strands, respectively. Therefore, we estimate that there are
40 total MHC I alleles in Mauritian Cynomolgus macaques. Our estimate, however, is based on a sample of only 56 animals. Additional alleles may be revealed if a larger population of animals is analyzed.
MHC I expression in Mauritian Cynomolgus Macaques
If the common MHC I genes in Mauritian Cynomolgus macaques are not expressed, the value of these animals may be diminished. To address this issue, we transfected a MHC I-null B cell line (721.221) with a clone containing the sequence of Mafa-B*440101. We detected expression of Mafa-B*440101 by surface staining the transfectant with a pan-MHC class I Ab (W6/32) conjugated to FITC. Expression of Mafa-B*440101 in Mauritian Cynomolgus macaques was verified by immunoprecipitating MHC I proteins from the surface of seven immortalized B cell lines and separating the immunoprecipitated proteins by 1D-IEF. The Mafa-B*440101 signature from the transfectant is present in the MHC molecules precipitated from A1M, A2M, A4M, A5M, and A6M cell lines, in both the presence and absence of neuraminidase (Fig. 4). These same animals were Mafa-B*440101-positive by both RSCA and individual clone sequencing. Though a band with a slightly more acidic isoelectric point is also observed in the neuraminidase-treated sample Mafa-B*440101-negative animal A8M, the banding pattern differs in the neuraminidase-untreated samples. Although we do not currently have MHC I transfectants for the other common alleles, at least two additional bands are shared between these animals, suggesting that at least two more of the common alleles are expressed (Fig. 4). We also calculated the isoelectric point for each of the common MHC class I alleles from its translated nucleotide sequence (protein calculator; C. Putnam, The Scripps Research Institute, La Jolla, CA) and found that Mafa-B*440101 had the most basic isoelectric point, in agreement with the position of this protein in the focusing matrix.
|
| Discussion |
|---|
|
|
|---|
We speculate that MHC I allele repertoire diversity may be a general feature of regional subpopulations of macaques. Anecdotal support for this speculation is found in the observation that the extremely common Indian Rhesus macaque MHC I allele Mamu-A*01(10) is not detected in Chinese Rhesus macaques (20). This result, coupled with our findings, may provide an impetus for the reporting of macaque origins in pathogenesis studies and spur research into the functional consequences of these regional differences. Currently, macaque origins are rarely specified in methods sections of immunology and pathogenesis publications (42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54), though this data may be profoundly influential.
Importantly, we found that Mauritian origin Cynomolgus macaques possess multiple extremely common MHC I alleles, with >50% of animals having the allele combination Mafa-B*430101, -B*440101, and -B*460101. To our knowledge, these are the highest-frequency MHC I alleles ever described in a population of macaques.
Mauritian Cynomolgus macaques may represent an extraordinarily valuable resource for pathogen research. Because the MHC loci are among the most polymorphic in primate genomes (55), it is reasonable to predict that non-MHC genetics in these animals may also be simplified relative to Asian origin macaques. This may be important, as HIV disease progression is likely to be influenced by polymorphisms in other loci, such as Kir receptors (56, 57), cytokine receptors and their ligands (58), in addition to MHC products (59). Vaccine trials that rely on comparisons between groups stand to benefit from the availability of animals with reduced genetic variability. Moreover, mapping cellular immune responses bound by common MHC I alleles will facilitate vaccine evaluation and studies of SIV pathogenesis.
The unusual degree of allele sharing in Mauritian Cynomolgus macaques is predictable given the natural history of this population. Anthropological and historical evidence suggests that the macaques were introduced to the island
400 years ago by European seafarers (60). mtDNA analysis of Mauritian origin macaques supports the existence of a population bottleneck at the time of introduction; in fact, the founding population may have included only a single female (61). Although low mtDNA divergence does not necessarily correlate with low nuclear gene variability, other studies have found the genetic diversity of nuclear genes in this population is lower than in Asian origin Cynomolgus macaques (62). Additionally, a recent study of MHC II sequences detected 20 Mafa-DRB sequences in a survey of 10 Chinese origin Cynomolgus macaques, while only 15 Mafa-DRB sequences were found in a population of 58 Mauritian origin macaques (63).
A significant unanswered question is whether the Mauritian Cynomolgus macaques we studied are representative of the entire population on the island. We do not have any information on the heritage of the animals that we studied because the blood samples were purchased commercially. It may be informative to study MHC haplotypes with microsatellite DNA markers, as recently described in Rhesus macaques (64), to infer the relationships among animals and to define the region configurations of common MHC I haplotypes. It is possible that the Mafa-B*430101, -B*440101, -B*460101 alleles are linked on a frequent MHC haplotype.
We do know, however, that the animals used in this study were wild-caught and, thus, unlikely to be related. Additionally, the blood samples were obtained on two different occasions, further increasing the likelihood that the results are extensible to the general population of Mauritian macaques. Cynomolgus macaques, as a species, are highly adaptable and can thrive in a variety of ecological niches, particularly those disrupted by human activity (60). Mauritius is <2000 km (2), with unequal monkey distribution throughout the small island. Previous analysis of mtDNA using monkeys captured from different sites did not observe genetic stratification by habitat, providing anecdotal evidence that subpopulations of macaques on the island are genetically homogenous (61).
Do Mauritian origin Cynomolgus macaques represent a reasonable alternative animal model for AIDS research? More than 3000 Mauritian Cynomolgus macaques are imported for research yearly (T. Demarcus, unpublished observations), eclipsing the number of Cynomolgus macaques imported from any other country. How does this compare with the demand for Rhesus macaques in AIDS vaccine research? According to the HIV/SIV Vaccine Trials Database, a total of 332 Rhesus macaques was used in 18 vaccine trials published in 2004. Even if only a fraction of all ongoing vaccine trials generate publishable data during a single year, a sizable population of these Mauritian Cynomolgus macaques should be available for vaccine projects that require genetically defined animals.
For the value of this model to be realized, Mafa-B*430101, -B*440101, -B*460101-positive animals must be susceptible to pathogenic SIV infections, without exhibiting the unusual resistance associated with certain MHC class I alleles in Indian Rhesus macaques (15, 16, 17). Previous work has shown that Cynomolgus macaques generally (65, 66), and Mauritian origin Cynomolgus macaques in particular (67, 68, 69), can be infected with pathogenic SIV and develop sAIDS. When compared with Indian Rhesus macaques, Cynomolgus macaques require a higher dose of virus to consistently establish SIV infection (68) and maintain lower viral loads more similar to HIV-infected humans (66). This could complicate the evaluation of vaccines that seek to reduce set point viral load because differences between vaccinees and naive controls may be more difficult to detect. Conversely, the exceptionally high viral loads witnessed in Indian Rhesus macaques may represent an unnaturally high barrier for vaccination, and may be the consequence of deriving viral strains by serial passage of SIV through multiple Rhesus macaques (70). It may be possible to recapitulate high viral loads in Cynomolgus macaques, if desired, by serial passage.
A related concern is whether the common MHC I alleles are likely to bind and present SIV-derived peptides to CD8+ T-lymphocytes. ID-IEF confirms that Mafa-B*440101 is expressed, and strongly suggests the expression of at least two additional, common MHC I alleles. What is the likelihood that expressed MHC I alleles restrict SIV-specific immune responses? In the Indian Rhesus macaque, epitope panning has been performed on four MHC I alleles. In these studies, the peptide binding motif for an allele is determined; the SIV proteome is scanned for peptides that fit the motif; the peptides are experimentally tested for binding affinity; and peptides that exceed the binding threshold are tested for antigenicity in SIV-infected animals. For each allele, at least five SIV-specific CD8+ T lymphocyte responses were identified (41, 71, 72). Therefore, it is reasonable to forecast that at least one (and perhaps all) of the three common MHC I alleles identified in Mauritian Cynomolgus macaques will restrict SIV-specific CD8+ T lymphocyte responses.
In summary, we have established the foundation for performing MHC I-dependent research projects in Cynomolgus macaques. We show that macaques from different origins have largely independent MHC I allele repertoires and should not used interchangeably for studies that may be influenced by MHC I genetics. Mauritian origin animals may be uniquely valuable for this type of research because of their atypical and unprecedented MHC I repertoire: >50% of these animals share a combination of three alleles.
| Acknowledgments |
|---|
| Disclosures |
|---|
|
|
|---|
| Footnotes |
|---|
1 Address correspondence and reprint requests to Dr. David H. OConnor, University of Wisconsin, Department of Pathology and Laboratory Medicine, 1300 University Avenue, Madison, WI 53706. E-mail address: doconnor{at}primate.wisc.edu ![]()
2 Abbreviations used in this paper: MHC I, MHC class I; 1D-IEF, one-dimensional isoelectric focusing; RSCA, reference-strand conformational analysis. ![]()
Received for publication April 13, 2005. Accepted for publication August 2, 2005.
| References |
|---|
|
|
|---|
2/V
2 T cell repertoire in Macaca fascicularis: functional responses to phosphoantigen stimulation by the V
2/J
1.2 subset. Immunology 115:197.-205. [Medline]
This article has been cited by other articles:
![]() |
R. W. Wiseman, J. A. Wojcechowskyj, J. M. Greene, A. J. Blasky, T. Gopon, T. Soma, T. C. Friedrich, S. L. O'Connor, and D. H. O'Connor Simian Immunodeficiency Virus SIVmac239 Infection of Major Histocompatibility Complex-Identical Cynomolgus Macaques from Mauritius J. Virol., January 1, 2007; 81(1): 349 - 361. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||