Characterization of full-length MHC class II sequences in Indonesian and Vietnamese cynomolgus macaques

Abstract

In recent years, the use of cynomolgus macaques in biomedical research has increased greatly. However, with the exception of the Mauritian population, knowledge of the MHC class II genetics of the species remains limited. Here, using cDNA cloning and Sanger sequencing we identified 127 full-length MHC class II alleles in a group of 12 Indonesian and 12 Vietnamese cynomolgus macaques. Forty-two of these were completely novel to cynomolgus macaques while 61 extended the sequence of previously identified alleles from partial to full-length. This more than doubles the number of full-length cynomolgus macaque MHC class II alleles available in GenBank, significantly expanding the allele library for the species and laying the groundwork for future evolutionary and functional studies.

Macaques are useful experimental hosts for a variety of human diseases and are frequently used in pathogenesis and vaccine research. Rhesus macaques (Macaca mulatta; Mamu) from India have traditionally been the preferred species for such studies, but the growing demand for these animals has led to shortages, especially in the field of HIV/AIDS research (Cohen 2000). Researchers are increasingly utilizing the cynomolgus macaque (Macaca fascicularis; Mafa) as an alternative non-human primate species. Cynomolgus macaques are closely related to rhesus macaques, but are smaller and more widely available. They have been used for studies of infectious diseases such as HIV/AIDS, tuberculosis, plague, and dengue as well as in transplantation and drug toxicity research (Capuano et al. 2003; Guirakhoo et al. 2004; Van Andel et al. 2008; Aoyama et al. 2009; Chamanza et al. 2010; Greene et al. 2010; Willer et al. 2010).

Gene products of the major histocompatibility complex (MHC) play a critical role in host immune responses to different pathogens. The most studied gene products encoded by the MHC are the classical class I and class II proteins. MHC class II molecules are expressed as heterodimers consisting of alpha and beta chains that are encoded by A and B genes, respectively. Found on antigen-presenting cells, these molecules display peptides to CD4 T-cells. A number of studies have described associations between specific human DP, DQ, and DRB alleles; HIV susceptibility; and HIV disease progression (Roe et al. 1999; MacDonald et al. 2000; Malhotra et al. 2001; Vyakarnam et al. 2004; Hardie et al. 2008). Some studies also document correlations between MHC class II genetics and simian immunodeficiency virus (SIV) disease progression in both rhesus and cynomolgus macaques (Sauermann et al. 1997; Sauermann et al. 2000; Giraldo-Vela et al. 2008; Mee et al. 2009). MHC class II genetics presumably also affect immune responses to pathogens (eg. Mycobacterium tuberculosis, influenza virus) against which a CD4 T-cell response in known to play an important role (Flynn 2004, Sant et al. 2007). They therefore have the potential to influence the results of vaccine and pathogenesis studies. In addition, characterization of additional MHC class II alleles may improve our understanding of what constitutes an effective CD4 T-cell response against these pathogens.

Researchers distinguish between several genetically distinct geographic populations of cynomolgus macaques: mainland Southeast Asian, Indonesian, Filipino, and Mauritian (Blancher et al. 2008; Stevison and Kohn 2008). Of all of these geographic populations, only the MHC genetics of the Mauritian population have been comprehensively characterized; this group has limited MHC genetic diversity, making it extremely useful for studies requiring cohorts of MHC-identical animals (O’Connor et al. 2007; Wiseman 2007; Greene et al. 2008; Budde et al. 2010). Nonetheless, biomedical researchers continue to use non-Mauritian cynomolgus macaques. During FY09, for example, only 19% of the cynomolgus macaques imported to the U.S. came from Mauritius (R. Mullan, personal communication, 2010. In spite of this widespread use of non-Mauritian cynomolgus macaques, only a handful of studies have examined the MHC class II genetics of these animals, and none of these studies investigated the DPA locus (Leuchte et al. 2004; Blancher et al. 2006; Sano et al. 2006; Aarnink et al. 2010; Ling et al. 2011).

In order to expand the MHC class II allele library for cynomolgus macaques, we used complementary DNA (cDNA) cloning and sequencing to characterize the full-length sequences of alleles at all six MHC class II loci in 12 Indonesian and 12 Vietnamese cynomolgus macaques. These two populations directly account for 13% of cynomolgus macaques imported to the United States during FY2009. Additionally, they represent a likely original source of the non-native cynomolgus macaques imported from Chinese facilities (59% of FY2009 total) (R. Mullan, personal communication, 2010). In order to increase our chances of identifying novel alleles, we aimed to study a diverse set of animals. The wide variety of STR haplotypes present in the Indonesian animals and MHC class I haplotypes in the Vietnamese animals suggests that the individuals were not closely related to one another (data not shown). We can therefore assume that the animals examined are relatively representative of their respective populations, though the cohorts’ small size limits the degree to which data can be generalized to the population level.

Samples from the Indonesian animals were provided by the Washington National Primate Research Center, Cerus Corporation, and the Wake Forest University Primate Center. One animal came from Jakarta while the rest were imported from a natural habitat breeding facility on Tinjil island, Indonesia and from two breeding colonies at which the majority of animals had Sumatran origins. Samples from 12 Vietnamese cynomolgus macaques, imported by Covance, Inc. were provided by Battelle Biomedical Research Center. Methods used were similar to those described previously (O’Connor et al. 2007). Briefly, mRNA was isolated from whole blood or peripheral blood mononuclear cells, reverse transcribed to form cDNA, and then amplified using a subset of the locus-specific primers listed in O’Connor et al. (2007). PCR products were then ligated into a cloning vector, propagated in E. coli, and sequenced using four primers—two internal primers specific to each locus and two primers flanking the cDNA inserts. These sequences were assembled and analyzed using CodonCode Aligner software (CodonCode, Dedham, MA). Novel full-length sequences were submitted to GenBank (accession numbers are listed in Table 1) as well as to the IMGT/MHC Non-human Primate Immuno Polymorphism Database-MHC (IPD-MHC) for official nomenclature (Robinson et al. 2003). The submitted DRB sequences lacked the first base of the open reading frame because it was included as the 3′ nucleotide of the forward primer. In order to avoid characterizing PCR artifacts as novel alleles, putative novel sequences were only considered legitimate when at least three identical full-length clones were observed.

Table 1.

Novel full-length Mafa alleles

Mafa allele name	Accession #	Animals	Animal origin(s)	Accession #(s)–identical Mafa sequence(s)	Origin(s)–identical Mafa
DPA1*02:06	HM580025	DR011	Vietnam	None
DPA1*02:07	HM580026	DR012	Vietnam	none
DPA1*02:08:01	HM580029	DR017	Vietnam	none
DPA1*02:08:02	HM579970	CE06	Indonesia	none
DPA1*02:09	HM580030	DR014, DR019	Vietnam	none
DPA1*02:10	HM580031	DR018	Vietnam	none
DPA1*02:11	HM579965	CE28	Indonesia	none
DPA1*02:12	HM579966	IN10	Indonesia	none
DPA1*02:13	HM579967	IN10	Indonesia	none
DPA1*02:14	HM579968	IN04, CE06	Indonesia	none
DPA1*02:15	HM579971	CE03	Indonesia	none
DPA1*02:16	HM579972	CE16, WF01	Indonesia	none
DPA1*02:17	HM579974	WF05	Indonesia	none
DPA1*04:02	HM579969	CE23, IN01, WF01	Indonesia	none
DPA1*06:01	HM580027	DR012	Vietnam	none
DPA1*07:03	HM580028	DR015	Vietnam	none
DPA1*07:04	HM579964	CE12, CE28	Indonesia	none
DPA1*08:01	HM580032	DR015	Vietnam	none
DPA1*09:01	HM580024	DR009	Vietnam	none
DPB1*01:01:01	HM580034	DR009	Vietnam	AB235856 (1)	In, Vi
DPB1*01:01:02a	HM580042	DR010	Vietnam	AB235856 (1)	In, Vi
DPB1*02:01:01	HM580036	DR012, DR016	Vietnam	AB235857 (1), AM086061 (2)	Vi
DPB1*06	HM579979	CE06	Indonesia	AB235861 (1)	In
DPB1*09	HM580039	DR016	Vietnam	AB235864 (1), D13335 (3)	Vi
DPB1*10	HM580037	DR013, DR014, DR020	Vietnam	AB235865 (1), D13336 (3)	Vi
DPB1*13	HM580041	DR019	Vietnam	AB235868 (1), HM371245 (4)	In, Ph, Vi
DPB1*14	HM579977	CE28, IN10, IN11	Indonesia	AB235869 (1), AM086066 (2)	In
DPB1*17	HM580038	DR017	Vietnam	AB235872 (1), HM371252 (4)	Vi
DPB1*20	HM580035	DR011, CE06	Both	AB235875 (1), AM086062 (2), HM153016 (4)	In, Ph, Vi
DPB1*22	HM579982	CE03	Indonesia	AB235877 (1), AM086165 (2)	In, Ph
DPB1*32	HM579986	IN11	Indonesia	AB235887 (1), HM371251 (4)	In, Vi
DPB1*51	HM579978	IN10	Indonesia	HM153013 (4)	Vi
DPB1*56	HM580040	DR010	Vietnam	none
DPB1*57	HM579981	DR020, CE23, IN01, WF01	Both	none
DPB1*58	HM579983	CE16, WF01	Indonesia	none
DQA1*01:03:02	HM580044	DR009, DR010, DR017, CE26	Both	AM086164 (2)
DQA1*01:09	HM579992	CE03	Indonesia	none
DQA1*01:10	HM579993	CE23	Indonesia	none
DQA1*01:11a	HM579995	IN04, WF01	Indonesia	AM086164 (2)
DQA1*01:12	HM580050	DR016	Vietnam	none
DQA1*05:01	HM580046	DR011, DR018, CE06	Both	AM086053 (2)
DQA1*05:03:02	HM579988	DR012, DR020, IN10, CE23	Both	none
DQA1*05:05	HM580047	DR012	Vietnam	none
DQA1*05:06	HM580051	DR020	Vietnam	none
DQA1*24:04	HM579990	IN01	Indonesia	none
DQA1*24:05	HM579991	DR009, DR018, CE12	Both	none
DQA1*26:01	HM580049	DR013	Vietnam	M76208 (5)
DQA1*26:04	HM580048	DR014, DR019	Vietnam	none
DQA1*26:06a	HM580045	DR011, CE16	Both	M76208 (5)
DQB1*06:07:01	HM579998	IN04, WF01	Indonesia	AJ308055 (6)
DQB1*06:13	HM580005	WF05	Indonesia	AJ308061 (6), HM153002 (4)	Vi
DQB1*06:14	HM579999	DR016, IN04, WF05	Both	AJ308062 (6), HM371235 (4)	Vi
DQB1*06:16	HM580052	DR009	Vietnam	GQ266371 (7), HM371234 (4)	Vi
DQB1*06:18	HM580003	CE16, CE23, WF01	Indonesia	GQ266374 (7)
DQB1*06:19	HM580060	DR010	Vietnam	GQ266375 (7), HM153001 (4)	Vi
DQB1*06:26	HM580004	CE23	Indonesia	GQ266382 (7), HM371232 (4)	Vi
DQB1*06:31	HM580059	CE23	Vietnam	none
DQB1*06:32	HM580062	DR019	Indonesia	none
DQB1*15:03	HM580057	DR010, DR020	Vietnam	AJ308063 (6), HM371225 (4)	Vi
DQB1*15:04	HM580061	DR020	Vietnam	none
DQB1*17:02	HM580055	DR012	Vietnam	X57753 (8), GQ266364 (7), HM153009 (4)	Vi
DQB1*17:03	HM580001	DR011, DR017, CE06	Both	AJ308065 (6), HM371227 (4)	Vi
DQB1*17:07:01	HM580054	DR011, CE16	Both	GU130498 (7), HM153012 (4)	Vi
DQB1*18:05	HM580000	CE12, IN01	Indonesia	AJ308072 (6)
DQB1*18:09	HM580053	DR009, DR015	Vietnam	GU130499 (7), HM371239 (4)	Vi
DQB1*18:13	HM580056	DR013	Vietnam	GU130503 (7)
DQB1*18:20	HM580058	DR019	Vietnam	none
DQB1*18:21	HM580002	CE03	Indonesia	none
DRA*01:01:07	HM580064	DR013	Vietnam	AB306655 (7), AY591919 (9), EU877205 (10)
DRA*01:01:08	HM580067	DR015	Vietnam	EU921294 (10)
DRA*01:03:05	HM580066	DR016	Vietnam	EU877206 (10)
DRA*01:04:01:01	HM580065	DR010, DR012, DR013, DR017, DR019, DR020	Vietnam	AB306656 (7)
DRA*02:01:04	HM580011	CE16	Indonesia	none
DRB*W1:01	HM580068	DR009, DR011, DR012, DR013, DR017, IN11	Both	AF492309 (11), AY340677 (12)	Ch, Ph
DRB*W1:02	FJ919740	DR017, DR018, CE06	Both	AF492306 (11), AY340683 (12)	Ch, Ph
DRB*W2:06	FJ919741	DR011, DR017, DR018, CE06	Both	EF175105 (7), FN433714 (7)	In, Vi
DRB*W3:03:01	HM580016	CE23	Indonesia	AM086041 (2), EF175140 (7)	In
DRB*W3:05	HM580073	DR013, DR014	Vietnam	AM086043 (2), EF175129 (7)	In
DRB*W3:06	FJ919732	IN04	Indonesia	AM086044 (2), EF175139 (7)	In
DRB*W4:03	HM580075	DR019	Vietnam	DQ363298 (7), DQ156968 (13)	Vi
DRB*W4:04	HM580074	DR012	Vietnam	DQ363267 (7), DQ777754 (15), HQ148690 (7)	Ch, Vi
DRB*W7:02	FJ919739	CE12	Indonesia	AM086046 (2), EF175122 (7), HM153072 (7)	Ch, In, Vi
DRB*W7:05	HM580013	CE03	Indonesia	FN433704 (16)
DRB*W7:07a	HM580014	CE16	Indonesia	AM086046 (2), EF175122 (7)	In, Vi
DRB*W25:03	HM580076	DR019	Vietnam	DQ156970 (13), EF175126 (7)	In, Vi
DRB*W26:01:01	HM580078	DR015	Vietnam	DQ363270 (7), FN433706 (7), HM153074 (7)	Ch, Vi
DRB*W26:02:01	FJ919733	IN04, WF01	Indonesia	AM911048 (7), DQ363300 (7), EF175132	In
DRB*W27:01	FJ919726	CE28	Indonesia	AY340671 (12), DQ363277 (7), HM153055 (7)	Ch, Vi
DRB*W33:01:02	FJ919736	IN01	Indonesia	none
DRB*W36:02:02	FJ919730	DR19, IN01, IN10	Both	AM910925 (14)
DRB*W40:01	HM580017	CE23	Indonesia	DQ156971 (13)
DRB*W64:01	HM580072	DR012	Vietnam	DQ363274 (7), FN433702 (7)	Vi
DRB*W66:01	FJ919727	CE28	Indonesia	AM910926 (14)
DRB*W73:01	FJ919731	IN10	Indonesia	none
DRB1*03:03	HM580069	DR009, DR014, DR017	Vietnam	AY340673 (12), DQ363296 (7)	Ch, Vi
DRB1*03:06:02	HM580071	DR011	Vietnam	none
DRB1*03:29	HM580070	DR011	Vietnam	none
DRB1*03:30a	HM580020	IN11	Indonesia	AY340673 (12), DQ363296 (7)	Ch, Vi
DRB1*04:05	HM580077	DR020	Vietnam	DQ363260 (7)	Vi
DRB1*04:11	FJ919729	IN01, IN10	Indonesia	AM911050 (14), EF175117 (7)	In
DRB1*04:14	HM580015	CE16	Indonesia	none
DRB1*07:02:02	FJ919738	CE12	Indonesia	DQ363264 (7), FN433726 (7)	Vi
DRB1*10:12	FJ919737	IN01	Indonesia	HM236222 (7)

^aDenotes alleles which extended previously available sequence but were not given the same name because it had already been assigned to a different full-length sequence with the same exon 2.

Mafa alleles for which full-length sequence was not previously available, including extensions of previously named alleles. “Both” denotes alleles found in both the Indonesian and Vietnamese cohorts. Identical Mafa sequences for which accession numbers are listed are limited to exon 2 with the exception of the DRA alleles. The origin information for identical Mafa sequences was taken from the corresponding publication(s) and/or GenBank record(s). Abbreviations used are as follows: Ch=China, In=Indonesia, Ph=Philippines, Vi=Vietnam. References for accession numbers are (1) Sano et al. 2006, (2) Doxiadis et al. 2006, (3) Hashiba et al. 1993, (4) Ling et al. 2011, (5) Kenter et al. 1992, (6) Otting et al. 2002, (7) unpublished, various sources (8) Gaur et al. 1992, (9) Senju et al. 2007, (10) Aarnink et al. 2010, (11) Blancher et al. 2006, (12) Leuchte et al. 2004, (13) Mee et al. 2008, (14) De Groot et al. 2008, (15) Wei et al. 2007, (16) Doxiadis et al. 2010.

In the 24 animals studied, we identified 127 distinct full-length MHC class II alleles, 43 of which were detected in more than one animal. Since these were cDNA sequences, all are actively transcribed and have the potential to form functional class II molecules. Twenty-four of these 127 transcripts were identical to previously identified full-length Mafa sequences (Supplemental Table 1). The remaining 103 alleles represented novel full-length Mafa sequences. Of these, 61 extended the sequence of known alleles for which the entire open reading frame was not previously available (Table 1). In five instances, two distinct full-length sequences extended a single previously described exon 2 sequence. The other 42 sequences (Table 1) were completely novel to cynomolgus macaques and were named by the NHP Nomenclature Committee based on alignment with known sequences (Robinson et al. 2003). Twelve of these novel transcripts were identical to previously identified Mamu class II nucleotide sequences and three perfectly matched alleles previously identified in stump-tailed macaques (Macaca arctoides; Maar) (Supplemental Table 2). This result is consistent with previous work documenting a high level of MHC class II allele sharing between rhesus and cynomolgus macaques and further supports the hypothesis that the macaque MHC class II has undergone conservative selection (Doxiadis et al. 2006). More of these alleles are potentially shared with other macaque species, but our ability to detect such sharing is hindered by the currently limited allele libraries for these species.

The data presented here triples the number of full-length nucleotide sequences of Mafa alleles in GenBank at all MHC class II loci except DRB. As of November 2010, a total of 36 full-length Mafa coding sequences were available for the DPA, DPB, DQA, DQB, and DRA loci combined. We describe 73 additional full-length sequences at these loci, bringing the total to 109. The DRB locus has been more extensively studied to date, with 37 full-length DRB coding sequences available in GenBank as of November 2010. We report an additional 30 DRB allele sequences, nearly doubling the total in GenBank. These full coding sequences are a prerequisite to the construction of cell lines expressing a single MHC class II molecule and MHC class II-peptide tetramers. Such reagents have informed the study of CD4 T-cell responses to in both humans and rhesus macaques (Kuroda et al. 2000; Dzuris et al. 2001; Giraldo-Vela et al 2008). To our knowledge, studies of SIV-specific CD4 T-cell responses have not yet been conducted in cynomolgus macaques. The expanded number of full-length Mafa MHC class II coding sequences presented here should aid researchers in developing the reagents necessary for investigation of CD4 T-cell responses in cynomolgus macaques.

Previous studies of the DPB, DQA, DQB, and DRB loci have largely been limited to the second exon, which encodes the highly variable peptide-binding region of these MHC molecules (Leuchte et al. 2004; Blancher et al. 2006; Doxiadis et al. 2006; Sano et al. 2006; Ling et al. 2011). However, sequencing of exon 2 alone may not always provide a complete picture of host MHC class II genetics since alleles may differ from one another only outside of this exon. We used the expanded library of full-length Mafa MHC class II alleles resulting from this study to evaluate the degree of resolution of unique alleles which can be achieved by exon 2-based genotyping for the current database of known alleles. For the analysis, we combined our data with all unique sequences available in GenBank as of November 2010 that extended beyond exon 2. A visual representation of the nucleotide variability across each MHC class II locus is shown in Figure 1. These plots demonstrate that all six MHC class II loci contain different degrees of variation outside of exon 2. It was more difficult to distinguish between the A alleles by looking only at exon 2 because there was extensive variability throughout the coding region. In the most extreme example, 26 of 33 DRA sequences (79%) were identical across exon 2 to at least one other allele. This observation was not surprising given the limited polymorphism at this locus (de Groot et al. 2004; Aarnink et al. 2010). Even the DPA and DQA loci, which are more variable, showed significant levels of ambiguity when only exon 2 was examined. Twelve of 27 (44%) DPA and 12 of 24 (50%) DQA alleles could not be distinguished from one another on the basis of exon 2 sequences alone. Although the variability of the class II B loci is more concentrated in exon 2, complete differentiation was not possible for any of the loci. The DQB alleles were easiest to distinguish; only two out of a total of 27 DQB alleles (7%) shared exon 2 sequence. Differentiating between DRB alleles was slightly more difficult as 10 of 78 (13%) sequences were identical to at least one other allele across exon 2. At the DPB locus, over one fifth – 6 of 27 (22%) – of exon 2 sequences could not be assigned to a single specific allele. As the library of full-length MHC class II nucleotide sequences grows, the number of alleles with identical exon 2 sequences will likely increase. Though the biological significance of differences outside of exon 2 remains unclear, our analysis suggests that full-length sequencing can be more precise and informative than that of exon 2 alone, particularly at the DPA, DPB, DQA, and DRA loci.

Fig. 1.

Nucleotide polymorphism across MHC class II coding sequences at each locus. We aligned all available unique sequences covering a minimum of exons 2 and 3 and calculated the Shannon entropy at each nucleotide position. The Shannon entropy (H) was computed for each nucleotide site by the following formula:

$$\text{H}=-\mathrm{\sum }{p}_i{log}_2(p_i)$$

where p_i is the proportion of the ith nucleotide at the site. Given four possible nucleotides, H takes values between 0 and 2; the latter value, equivalent to 2 bits of information, represents the maximum amount of entropy possible given 4 nucleotides. Shaded region denotes exon 2.

We also sought to examine the sharing of MHC class II alleles between geographic populations of cynomolgus macaques. This was most pronounced between the Indonesian and Vietnamese groups. In total, we detected evidence of 22 alleles shared between Indonesian and Vietnamese cynomolgus macaques that had not been previously associated with animals from both regions. Eighteen alleles were detected in both cohorts of animals used in this study, fifteen of which were not previously known to be common to both populations (Table 1, Supplemental Table 1). An additional seven were found in only one of the two populations described here, but had previously been documented in the other. The shared alleles may have originated prior to the initial isolation of the Indonesian population or may have resulted from gene flow across the Sunda shelf during later periods of lowered sea levels (Voris 2000; Sathiamurthy and Voris 2006). Significant allele sharing was also evident between animals from Indonesia and Mauritius; within our small Indonesian cohort, we documented twelve class II cDNA sequences identified previously in cynomolgus macaques from Mauritius. Only two of these (Mafa-DPB1*21 and Mafa-DPB1*29) had been previously associated with the Indonesian population. Such sharing further supports the hypothesis that the Mauritian population was founded by cynomolgus macaques from Indonesia (Tosi and Coke 2007; Bonhomme et al. 2008).

In summary, given the potential of the MHC to serve as a confounding variable in experiments, it is prudent to consider class I and II genotyping, preferably using full-length sequences, of animals used in vaccine or pathogenesis studies. The 103 novel full-length sequences described here greatly expand the Mafa MHC class II allele library and will aid the development of reagents for MHC genotyping. This growing library will also improve future disease-association studies and provides an important foundation for functional studies of CD4 T-cell responses to a diversity of pathogens.