Phylogenetic tree of human, chimpanzee ( Patr ), gorilla ( Gogo ), and cynomolgus macaque ( Mafa ) MIC alleles. Human MICA*001 and MICB*001 alleles were included as references. The human non- 

Contexts in source publication

Context 1

... MHC class I chain-related ( MIC ) genes define a distinct lineage of MHC-encoded class I genes. In humans, these non- conventional MHC-I genes comprise a family of seven members ( MICA-G ) of which only MICA and MICB produce functional transcripts (Bahram et al. 1994). MICA and MICB genes encode highly glycosylated MHC class I-like heavy chains, expressed at the cell surface independently of β 2 microglobulin and TAP-derived peptides. They possess a quasi-ubiquitous transcription pattern, are induced up- on cell stress, but at the protein level, appear to be selectively expressed within mucosae (Schrambach et al. 2007). MICA and MICB engage NKG2D, an activating C-type lectin-like immunoreceptor expressed on natural killer cells, γδ T cells and CD8 + αβ T cells (Raulet 2003). By triggering the cytolysis mediated by NKG2D-bearing cells, they participate in the first line of defense against pathogens and tumors (Lodoen and Lanier 2006). The genomic structure of MICA and MICB is similar to that of conventional MHC class I genes and harbors six exons. Exon 1 encodes the leader sequence, exons 2 – 4 the three extracellular, α 1, α 2, and α 3 domains, whereas exons 5 and 6 correspond to the transmembrane segment and cytoplasmic tail, respectively. The coding sequence of human MICA differs from MICB , most notably by the presence of a short-tandem repeat (STR) polymorphism in exon 5 (GCT n including a frameshift mutation GGCT). Unlike other non-conventional MHC-I (cf. HLA-E/F/G , HFE , MR1 , CD1 , ZAG , EPCR , MILL , FcRn , ULBP/RAET1 ), MICA and MICB have a high degree of allelic diversity. In man, 100 and 40 alleles (corre- sponding to 79 and 26 distinct protein sequences) have been published, thus far, for MICA and MICB , respectively (see MIC polymorphism has been associated with a number of diseases, as is a general characteristic of polymorphic HLA genes and haplotypes. These include auto-immune/auto-inflammatory diseases (e.g., Amroun et al. 2005; Kirsten et al. 2009), cancer (e.g., Douik et al. 2009; Jumnainsong et al. 2007), but also allograft rejection and graft-versus-host disease (Sumitran-Holgersson 2008). The origin, evolution, and the functional relevance of MIC diversity remains an open question (Bahram 2000; Elsner et al. 2001; Fodil et al. 1999; Perez-Rodriguez et al. 2002; Stephens 2001). Unlike MHC class II genes, which show a near colinear orthology between species as evolutionary distant as mouse and man, MHC class I genes, especially non- classical/non-conventional MHC-I genes, exhibit very different chromosomal location, even among the more closely related primate species. Having previously reported on the genomic organization of MHC genes among related non- human primate species (Shiina et al. 2006), in this study, we aimed to extend knowledge of allelic diversity of individual MIC genes within these same species. Members of the MIC gene subfamily have been described in non-human primate species (Anzai et al. 2003; Pellet et al. 1999; Seo et al. 1999; Steinle et al. 1998) and other mammals (Bahram et al. 1994), but are notably absent in rodents (Bahram et al. 1994; Kumanovics et al. 2002). Within mammals, the MIC genes, along with their cognate MHC-I coun- terparts, are quite plastic, with different gene numbers and genomic architectures. For instance, in the common chimpanzee ( Pan troglodytes ), there is only a single MICA/B gene, which resembles a hybrid of human MICA with MICB and has a low degree of polymorphism (Anzai et al. 2003; Kumanovics et al. 2002). By contrast, for rhesus macaque ( Macaca mulatta ), three MIC genes have been described (Averdam et al. 2007; Doxiadis et al. 2007). Whereas rhesus macaque is the most studied animal model in terms of MIC polymorphism, comparatively little is known of MIC variability in the related cynomolgus macaque ( Macaca fascicularis ), the non-human primate most commonly used in biomedical research. The aim of the present work was to study the allelic diversity, the genomic structure, and the phylogeny of MIC genes in three non-human primates: cynomolgus macaque, common chimpanzee, and gorilla. For this purpose, genomic DNA from 72 M. fascicularis (Mafa), 63 P. troglodytes (Patr), and 18 Gorilla gorilla (Gogo) unrelated animals were collect- ed at the authors ’ laboratories and analyzed by DNA sequencing. After PCR amplification of exons 2 to 5 with the Expand Long Template PCR System (Roche, Germany), exons 2, 3, 4, and 5, as well as intron 3, were sequenced using the standard protocol of the BigDye Terminator v3.1 Cycle Sequencing kit (Life Technologies, USA) with 400 ng of genomic DNA as template (primer sequences are provided in Table 1). Ampli- fication and sequencing parameters were those recommended by the respective manufacturers. Where several genes could have been amplified (Mafa) and/or more than one SNP per gene was detected (Patr and Gogo), the amplicons were cloned prior to sequencing. Multiple alignments of the sequences were performed using MUSCLE (Edgar 2004) with a maximum of 20 iterations. Phylogenetic trees were generat- ed using the maximum likelihood algorithm (Sullivan 2005) from PhyML and the LG matrix model (Le and Gascuel 2008). Bootstrap branch support values were calculated from 100 replicates. Calculation of dN/dS was performed using PAML (Yang 1997) which estimates the parameter ω by maximum likelihood. According to genomic structures of the genes and official nomenclature of the HLA system (Marsh et al. 2010), we propose to name MIC genes as follows: Mafa- MICA , Mafa- MICB , Mafa- MICB/A , Patr- MICA/B , and Gogo- MICA (supplementary Table 1). Figure 1 displays the repertoire of allelic sequence for the Mafa, Patr, and Gogo MIC genes. For Mafa, three different MIC genes — Mafa- MICA , Mafa- MICB , and Mafa- MICB/A — were distinguished with 10, 13, and 12 alleles, respectively, being identified (Table 2). Aside from their variation by single amino acid differences, Mafa- MICB and Mafa- MICB/A are distinguished from Mafa- MICA by a three-amino acid deletion in the α 2 domain (residues 139 – 141). Moreover, Mafa- MICB has also a four-amino acid deletion in the α 1 domain (residues 56 – 59). As in human MICA , an STR is present in exon 5 of Mafa- MICA and Mafa- MICB/A that encodes the transmembrane segment. Each reported allele of Mafa- MICA and Mafa- MICB corresponds to a different protein sequence, whereas only 8 unique protein sequences are observed for a total of 12 alleles for Mafa- MICB/A (Table 2). Gorilla and chimpanzee have only one MIC gene with 4 and 18 different alleles being identified, respectively. These genes were named Gogo- MICA and Patr- MICA/B . The sequences of Gogo- MICA*04 ) and Patr- MICA/B*01 , Patr- MICA/B*04 , Patr- MI- CA/B*10 , and Patr- MICA/B*16 alleles are identical to previously published sequences (see supplementary Table 1) (de Groot et al. 2005; Pellet et al. 1999; Shiina et al. 2006; Steinle et al. 1998). From analysis of dN/dS ratio (Table 2), Patr- MICA/B seems to be under positive Darwinian selection, similarly to human MICB (dN/dS=1.44; data not shown) (Messier and Stewart 1997). On the other hand, Mafa- MIC genes appear to have been subject to neutral or purifying selection process like human MICA (dN/dS=0.96; data not shown) (McDonald and Kreitman 1991). Different haplotypic configurations were observed for Mafa- MIC genes. Among the 72 individuals tested, all were positive for Mafa- MICB (as identified by the deletion in exon 2), whereas Mafa- MICA and Mafa- MICB/A were successfully amplified from only 42 (58.3 %) and 68 (94.4 %) individuals, respectively. All three Mafa-MIC genes were present in 38 (52.8 %) cynomolgus macaques. For rhesus macaque, a similar conclusion was drawn in the initial description of three MIC genes: Mamu- MIC1 , Mamu- MIC2 (Seo et al. 1999; Steinle et al. 1998), and Mamu- MIC3 (Seo et al. 1999; Seo et al. 2001). More recently, however, it was shown that no individual animal had more than four different MIC sequences and that Mamu- MIC1 and Mamu- MIC3 were distantly related alleles of the same gene. These data are consistent with rhesus macaque having only two functional MIC genes (Averdam et al. 2007). In our analysis, five Mafa- MIC alleles were observed in three different animals, indicat- ing that Mafa- MICA and Mafa- MICB/A are independent MIC loci , in addition to Mafa- MICB . The genomic structure of Mafa- MICB/A is however similar to that of Mamu- MIC3 , in that it appears to be a recombinant hybrid of Mafa- MICB and Mafa- MICA (Averdam et al. 2007; Doxiadis et al. 2007). A possible recombination breakpoint around nucleotide position 1,110 could indeed be deduced from the alignment of intron 3 sequences from 32 animals (Fig. 2). This presumptive breakpoint is identical to the one reported in the rhesus macaques (Doxiadis et al. 2007). The hybrid nature of Mafa- MICB/A is also apparent from the protein sequence alignment (Fig. 1), in which Mafa- MICB/A shows highest homology with the α 1 and α 2 domains of Mafa- MICB and with the α 3 domain of Mafa- MICA . To investigate further the evolution of MIC genes, we con- ducted phylogenetic analyses of all the non-human MIC allele sequences along with human MICA*001 and MICB*001 as references. Fig. 3 shows that hominid (human, Patr, and Gogo) and old world monkey (Mafa) MIC genes cluster within two separate clades. The analysis also indicates that all the ...

Context 2

... analyzed here are descendants of a single common ancestor. As the human MICA gene is more closely related to the gorilla and chimpanzee MIC genes than to human MICB gene, and as Mafa- MIC genes are closer to human MICB than to human MICA (as confirmed by phylogenetic analyses including all publicly available primate MIC genes; unpublished data), the common ancestor was most probably a human MICB -like gene with a STR in exon 5. This ancestral gene would then have lost its STR before the divergence of old world monkeys and hominids. This hypothesis is supported by the fact that most human MICB and Mafa- MICB alleles retain two GCT repeats (supplementary Fig. 1). It should be noted that the fact that Mafa- MICB is more distantly related to human MICB than Mafa- MICA is related to human MICB (as seen in Fig. 3) does not, in our opinion, invalidate the hypothesis of a common human MICB -like ancestor harboring a STR in exon 5, as this could be in part explained in addition to bootstrap data on all available MIC alleles (Ott et al. unpublished data) — by the fact that Mafa-MICA has not evolved as fast as Mafa-MICB from the ancestral gene. Finally, it is important to notice that the categorization of primate MIC genes in “ A ” or “ B ” is based on the presence or absence of an STR in exon 5. Thus, Mafa-MICA is dubbed “ MICA , ” even if it is closer to human MICB than to human MICA . In conclusion, we report a total of 52 novel MIC alleles in non-human primates. This study further demonstrates the existence of three MIC genes in Mafa and strongly suggests that the ancestral MIC gene was most likely a human MICB like locus. These observations expand knowledge of MIC diversity in non-human primates and contribute to our understanding of the allelic repertoire and diversity of primate MIC ...