Index INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION References

INTRODUCTION

The Asiatic black bear, Ursus thibetanus, is mainly distributed throughout eastern Asia. The Japanese black bear is sometimes identified as the subspecies, Ursus thibetanus japonicus. Mitochondrial DNA analysis and microsatellite markers have been previously used to study the genetic structure of this species (Ishibashi and Saitoh, 2004; Ohnishi et al., 2007, 2009; Yasukochi et al., 2009). The findings of Yasukochi et al. (2009), revealed that bears in western Japan exhibited a lower genetic diversity and a higher level of genetic differentiation than those of eastern Japan. This discovery played a vital contribution to the conservation policy for these isolated populations.

The major histocompatibility complex (MHC) directs the expression of a set of molecules on cell surfaces that are responsible for antigen presentation to lymphocytes. As such the MHC is one of the most important genetic systems for infectious disease resistance in vertebrates (Klein, 1987; Brown et al., 1988; Hedrick, 1999). Therefore a low level of MHC genetic variation is thought to decrease a species adaptive ability towards various pathogens (Klein and Takahata, 1990). In this sense such a genetic diversity factor should be an important index in conservation genetics.

In recent years, an increasing number of studies have focused on the important role of the MHC in wild animals (for example, finless porpoises, New Zealand sea lion, Scandinavian wolf, American bison, sea otter and others) (Lento et al., 2003; Seddon and Ellegren, 2004; Traul et al., 2005; Bowen et al., 2006). However, for the Ursidae, MHC studies have been conducted for only a few species. MHC studies of the giant panda, Ailuropoda melanoleuca, have been relatively extensive and results have played an important role in determining the optimum conditions for captive management (Zeng et al., 2005; Wan et al., 2006; Zhu et al., 2007; Zeng et al., 2007; Pan et al., 2008). Sequences of the DRB locus in the polar bear, Ursus maritimus, have also been submitted to Genbank (Wei and Happ, unpublished). The genetic diversity of the DQA gene has been examined for the brown bear, Ursus arctos (Goda et al., 2009).

Yasukochi et al. (2009) demonstrated genetically endemic that the black bear in Japan was, and isolated from the black bear on the Asian continent. Thus, it is important to examine the genetic diversity of the MHC gene for the conservation and management of the Japanese black bear. Since the existence of unexpressed pseudogenes has been documented in the MHC region (The MHC Sequencing Consortium, 1999), it is essential to focus on an only expressed MHC gene (i. e. except for pseudogenes) for polymorphic analysis of the MHC region. However, to date there is no previously published study that analyzes an entire allelic sequence in a single locus using an expressed sequence from Ursidae. This includes the lack of such published results for the giant panda. We therefore attempted to identify an expressed DQB locus using mRNA from the Asiatic black bear by RACE (rapid amplification of cDNA ends; Frohman et al., 1988) PCR method. Polymorphisms of MHC genes are maintained by balancing selection that enhances the rate of nonsynonymous nucleotide substitutions in codons encoding the peptide binding residues (PBRs) of the molecule (such as in overdominant selection) (Hughes and Nei, 1988, 1989; Hughes, 1995; Hedrick, 1999). We also designed a primer pair for amplifying the DQB locus entire exon 2 region including the PBRs, to allow for further MHC genetic diversity analysis in the future.

MATERIALS AND METHODS

Total RNA was isolated from the fresh blood of a single Japanese black bear using the Catrimox-14^TM RNA Isolation Kit Ver. 2.11 (TaKaRa Bio Inc.). Only full-length mRNA was converted into cDNA using the GeneRacer^TM kit (Invitrogen). Our objective was to amplify the whole sequence of the MHC locus using RACE adapter primers and the designated primers, which were designed to anneal within the exon 2 region of DQB locus, based on a sequence acquired from genomic DNA of the individual. The designated primer sequences were as follows: URRACE-F5 (5’-CCAGTTTAAGGGCGAGTGCTACTTCACCAAC-3’) for 3’RACE-PCR, URRACE-R5 (5’-CTCGCCGCTGCAGGATGAACCTGTCCTCAAT-3’) and URRACE-R8 (5’-GTTGGTGAAGTAGCACTCGCCCTTAAACTGG-3’) for 5’RACE-PCR. Primer ZcDQBUEX1D, previously reported by Bowen et al. (2002), was also used for 5’RACE-PCR. RACE-PCR amplification was carried out using a DNA Thermal Cycler in 25 μl reaction mixture containing 10 × Ex Taq Buffer, 2.5 mM dNTP Mixture, 2 pmol/μl of each primer and 0.625 Units of Ex Taq Hot Start Version (TaKaRa Bio Inc.). The PCR program was performed according to the manufacture’s instructions of GeneRacer^TM kit. To determine whether the amplified MHC DQB locus was homozygous or heterozygous, the PCR product was cloned using the TOPO TA cloning kit (Invitrogen). Approximately 10 white colonies obtained by TA cloning were analyzed by PCR to verify insertion of the target sequence. For positive identification of an allele, at least two identical sequences from each PCR product were required, since it was possible that artifactual point mutations could occur during PCR (Gyllensten et al., 1990; Murray et al., 1995; Hayashi et al., 2003). Purification of PCR products and cycle sequencing was performed following the procedure of Yasukochi et al. (2009).

Seven samples (6 tooth samples, 1 hair sample) from the black bears of the Eastern Chugoku population in western Japan were used to determine the entire exon 2 region (270 bp) at DQB locus. A protein digestion and DNA extraction were performed by following the procedure of Yasukochi et al. (2009). The extracted DNA was amplified by PCR. Primers URDQBex2-F1 (5’-GAGGCCTTCGAGTTCATCAG-3’) and URDQBex2-R1 (5’-GAGACTCCGCAGGCGGAAG-3’) were used to amplify the entire exon 2 region. These primers were designed based on sequences of intron 1 and intron 2 around exon 2. PCR amplification was carried out in same reaction mixture as RACE-PCR. After incubation at 94°C for 1 min, 40 cycles were performed as follows: 30 s at 94°C, 45 s at 58°C, and 45 s at 72°C, with a post-cycling extension at 72°C for 30 s. TA cloning and sequencing was performed by same method as described above.

Nucleotide sequences were aligned and translated into amino acids using MEGA Ver. 4 (Tamura et al., 2007). A neighbor-joining (NJ) tree (Saitou and Nei, 1987) was reconstructed based on the p-distance (Nei and Kumar, 2000). Bootstrap analysis was performed using 1000 replications.

RESULTS AND DISCUSSION

Using cDNAs converted from full-length mRNAs, we sought to initially determine the entire sequence of the DQB allele. Consequently, a near full-length sequence (1150 bp) of a single MHC allele (Urth-DQB) was obtained (Fig. 1), and the sequence was deposited in the DDBJ/EMBJ/GenBank database (accession number: AB473936). The nucleotide sequence of Urth-DQB showed a similarity to that of the California sea lion, Zalophus californianus, DQB allele (90%), dog, Canis lupus familiaris, DQB (85%) and human, Homo sapiens, DQB (81%). The nucleotide sequence of the Urth-DQB coding region (to the stop codon: 798 bp) had a similarity with the California sea lion DQB (94%), dog DQB (90%) and human DQB (87%). Bowen et al. (2002) reported that the California sea lion DQB amino acid sequences contained DQ-specific motifs of 3 amino acids in the exon 3 region. These motifs are especially useful for distinguishing the DQB from DRB sequences in several species of mammal. The expressed amino acid sequence of the black bear also contained this pattern (Fig. 1). This supports that the expressed MHC nucleotide sequence of the bear is a DQB sequence rather than a DRB sequence. The near full-length DQB sequence had no putative frameshifts or deletions. Therefore, our findings suggest that the black bear has at least one functional DQB locus.

View Details

Fig. 1 Nucleotide sequences of the near full-length DQB cDNA (Urth-DQB) and DQB exon 2 region (Urth-DQB*a to Urth-DQB*e) of the Asiatic black bear (this study), California sea lion (Bowen et al., 2002) and the dog DQB sequences (Wagner, unpublished; Debenham et al., 2005). Dots indicate identity with the nucleotides of Urth-DQB. Dashes indicate alignment gaps. The predicted amino acid sequence as translated from the nucleotide sequence of the black bear. An asterisk indicates the stop codon. Boxes indicate codons used to distinguish DQB from DRB. Gray boxes indicate positions of putative peptide binding residues (PBRs) of the human DR molecule (Brown et al., 1993). Arrows designate the borders of the putative β chain domains (TM, transmembrane region; CY, cytoplasmic tail). The borders of exon 1 to exon 6 and the putative β chain domains are according to those of the California sea lion DQB sequences.

We also compared amino acids among several mammalian species, including comparison with human DQB amino acids where the crystal structure is known (Fig. 2). Peptide binding residues (PBRs) in the β1 domain were heterogeneous among all the mammalian species. On the other hand, amino acids known to be coding important functions for human DQB genes, tended to be also conserved among other mammalian species. Amino acids in position 81–88 are thought to be involved in homodimerization of MHC class II αβ heterodimers (Paliakasis et al., 1996). In HLA-DQ amino acids, position 84 is occupied by P or L, while position 87 is occupied by R or L. In the DQB of the Japanese black bear, amino acids at position 84 and 87 were E and R, respectively (Fig. 2). The ability of DQ molecules to form homodimers of MHC class II αβ heterodimers is modulated by the presence of the polymorphic patch at position 81–88. The difference of homodimerization ability may be expressed in the difference of amino acid sequences between human and bear DQB molecules. Amino acids at position 166–180 in the HLA-DQ/DR molecules are involved in the binding of CD4. Their amino acid sequences are conserved in this position because of their importance for CD4 binding (König et al., 1992, 1995; König, 2002; Riberdy et al., 1998; Maroto et al., 1999). In several HLA-DQ/DR amino acids, A→T at position 172 at the mid point of the CD4 binding loop, some unusual functional properties are observed (Bondinas et al., 2007). An amino acid at this position in the black bear DQB had A. Since the residue is identical to that in humans, the MHC sequence of the black bear appears to have normal function in terms of formation of the CD4 binding loop. Amino acids at position 199–201 formed the RGD loop. This conformation is similar to RGD conformation of proteins involved in cell adhesion interactions (Hynes, 1992; Paliakasis et al., 1996; Papadopoulos et al., 1998; Bondinas et al., 2007). This kind of RGD motif was found in the Japanese black bear DQB amino acids. In the intracytoplasmic tails of some HLA-DQB amino acids, there are eight-residue insertions (-PRGPPPAG-). Bondinas et al. (2007) suggested that the three conservative prolines might be part of a special structure, as is the case in the extracellular MYPPPY sequence of T lymphocyte proteins CD28 and CTLA-4. However, these three prolines were found to be not only absent in the DQB amino acids of the black bear but also absent in the other mammalian species. Therefore, the effect of the insertion is not still clear. Amino acids in the leader peptide region were extremely heterogeneous among mammalian species but not within each species. This might reflect a different kind of peptide for each species. Some unusual mutations of the amino acid residue at position 189 (D, S and A) have been reported from human DQB (Bondinas et al., 2007). In the Japanese black bear and the other mammalian species DQB amino acids, such unusual mutations remained unobserved. The intramembranous segment of HLA-DQB molecule is in position 231–251. All residues of the Japanese black bear in the intramembranous segment were hydrophobic, except for S at 233, indicating the segment was inside the membrane. These results from the comparison of amino acids among mammalian species suggested that the predicted amino acid sequence translated from the nucleotide sequence of the black bear possessed most of the features expected to be found in a regular functional MHC class II β chain.

View Details

Fig. 2 DQB amino acid sequences of some mammalian species. The known properties and functions are annotated according to Bondinas et al. (2007). Human DQB amino acid sequence with known crystal structures are in bold. Dots indicate identity with the amino acids of Urth-DQB. Dashes indicate alignment gaps. Light gray boxes indicate positions of putative PBRs of the human DR molecule (Brown et al., 1993). A dark gray box indicates an RGD-like loop. Numbers at the bottom of the aligned sequences indicate relative positions of residues which interact with a bound antigen. The boxes with broken lines indicate homodimerization patches. The residues involved in CD4 binding are indicated by a bold lined box. Residues forming hydrogen bonds with antigenic peptide backbone. #: residues participating in the formation of the homodimer of heterodimers. S-S: disulfide bridges. + and -: charge signs participating in the intra-salt bridge.

An NJ tree, based on 266 amino acids translated from 798 bp nucleotide sequences, showed that the translated Urth-DQB amino acid sequence of the black bear formed a monophyletic group with the DQB sequences of a number of other mammals (Fig. 3). DQB sequences could be clearly separated from DRB sequences, and each DQB and DRB sequences in each order formed a single clade. This strongly supports our conclusion that the sequence obtained from our study is an allele of the DQB locus. Multiple DQB loci exist in humans, pigs and cattle MHC class II regions (Marello et al., 1995; The MHC Sequencing Consortium, 1999; Gelhau et al., 1999; Barbosa et al., 2004). In the NJ tree, five DQB amino acid sequences of the California sea lion may be derived from at least two DQB loci, whereas five DQB sequences of the Indo-Pacific bottlenosed dolphin, Tursiops aduncus, and the bottlenosed dolphin, T. truncates, might exist in a single locus (Bowen et al., 2002; Yang et al., 2007). The DQB sequences from two loci of the California sea lion formed a monophyletic group in the NJ tree although other mammalian DQB sequences were derived from a single locus (i. e. HLA-DQB1, DLA-DQB1 and Tutr-DQB/Tuad-DQB in Fig. 3). Therefore, there exists a discrimination of differing MHC sequences of different DQB loci.

View Details

Fig. 3 Neighbor-joining tree for the mammalian near full-length DQB, DRB and DPB locus based on 266 amino acid sequences. A p-distance was used as distance measure. Only bootstrap values over 70% are shown in this figure. Numbers in parentheses are GenBank Accession Numbers.

Using the primer pair designed in our study, the entire exon 2 region (270 bp) in DQB locus of the black bear was determined. The genomic DNA sequence was consistent with the sequence of the DQB exon 2 converted from full-length mRNA from the same individual. This suggested that we could amplify the exon 2 of the functional DQB locus. Five DQB alleles (Urth-DQB*a to Urth-DQB*e) were detected from the seven bears collected from Eastern Chugoku. Their nucleotide sequences were determined to have 90 to 92% similarity with the California sea lion DQB allele. No more than two different alleles were detected from the clones of the PCR products from each of the seven individuals. Thirteen variable sites among 5 alleles from Urth-DQB*a to Urth-DQB*e were identified without deletion and insertion, and a mean genetic distance among these alleles was 0.027 ± 0.007. The thirty-six DQB exon 2 alleles (267 bp) taken from the dog, from the Immuno Polymorphism Database-MHC (IPD-MHC) at http://www.ebi.ac.uk/ipd/ (Robinson et al., 2005), had a mean genetic distance of 0.070 ± 0.011. The distances for Chinese yellow cattle, Bos taurus, from 19 alleles (270 bp) and the Chinese pig, Sus scrofa, from 68 alleles (260 bp) were 0.083 ± 0.010 and 0.054 ± 0.008, respectively (Wang et al., 2005; Li et al., unpublished). In comparison, the distance values of DQB exon 2 alleles for the Japanese black bears were lower than those of other mammalian species. The result suggests that the DQB genetic diversity of the Japanese black bear might have previously declined due to non-random mating and/or genetic drift. Either of such factors could have possibly been caused by a prior fragmentation and isolation of the bear’s habitat. However, the sample size for the Japanese black bear is so far too small to draw strong conclusions. Future work on the DQB gene of the Japanese black bear should clarify this question more fully. An NJ tree, based on 89 amino acid sequences translated from 267 bp nucleotide sequences of exon 2 alleles, showed that five sequences from the black bears formed a monophyletic group with DQB sequences of Carnivora from other mammalian species (Appendix 1).

View Details

Appendix 1 Neighbor-joining tree for the mammalian DQB, DRB and DPB locus entire exon 2 region based on 89 amino acid sequences. A p-distance was used as distance measure. Only bootstrap values over 70% are shown in this figure. Numbers in parentheses are GenBank Accession Numbers.

As mention above, even if there are no frameshifts or deletions within a MHC sequence, the sequence has the possibility to possess abnormal mutations. Therefore, it is essential to estimate whether the analyzed MHC sequence has normal functions. In our study, no obvious abnormal mutations were found from the DQB amino acid sequence of the Japanese black bear, indicating that we succeeded to isolate an apparently functional black bear DQB locus. In the future, studies designed to investigate the genetic diversity of MHC DQB genes will further contribute to strategies important for the conservation of the isolated populations of this species.

The authors are indebted the Ministry of the Environment of Japan, the Japan Wildlife Research Center, the Aso Cuddly Dominion and Culture, and the International Affairs Division Department of the Environment and Civic Affairs Shimane Prefectural Government for their kind contribution in sample collection and to Chris Wood of the GCOE, Kyushu University, Japan, for kind reading through and editing of English.