HIGH LEVEL OF MHCPOLYMORPHISM IN THE BAIJI AND FINLESS PORPOISE, WITH SPECIAL REFERENCE TO POSSIBLE CONVERGENT ADAPTATION TO THE FRESHWATER YANGTZE RIVER - Biology, Evolution and Conservation of River Dolphins Within South America and Asia - page 411

total of two DRA , eight DQB , six MHC-I alleles were identified in 15 samples. In contrast, a
total of 5 DRA , 14 DQB and 34 MHC-I unique sequences were identified in 195 finless
porpoises. In addition, a high level of sequence variation between sequences also indicates
that the MHC genes have significant genetic diversity.
Similar to other mammalian species, the DRA gene showed very low sequence
divergence in the baiji and the finless porpoise, with 3.17% sequence variation in the baiji and
4.8% in the finless porpoise at the amino acid level. In contrast, a considerable sequence
variation was detected at the DQB locus. This was evidenced by the high level of nucleotide
sequence variation among pairwise comparisons, which ranged from 0.58% to 9.30% in both
the baiji and the finless porpoise. However, of the three genes investigated in this study,
MHC-I showed the most extensive variability. For example, nucleotide sequence variation
among all pairwise comparisons of Live-I sequences, corrected for multiple substitutions,
ranged from 12.24% to 61.22%. The high divergence between alleles of MHC-I was also
supported by the relatively higher ratio of d N / d S than those of DQB and DRA as shown in
Table 2.
Variability was higher in the functionally important antigen recognition and binding sites
of the MHC-I and DQB genes, as supported by more nonsynonymous than synonymous
substitution rates. This is a clear indication of balancing selection (positive selection)
maintaining new variants and increasing allelic polymorphism in the baiji and finless
porpoise. However, for the DRA gene, no such phenomenon was found in the PBR, which
was contrary to the normal pattern of substitution in the PBR for MHC class II genes (Hughes
& Nei, 1989), and may suggest no balancing selection on this gene.
Population Expansion and Intragenic Recombination
Evolution by random bifurcation, without population expansion, recombination and/or
convergence, is expected to yield a multimodal histogram with many peaks and ragged
appearance as a result of differentiation of sequences into allelic lineages and extinctions of
intermediates (Go et al., 2002; Figueroa et al., 2000; Suárez et al. et al., 2006). In this chapter,
the mismatch distribution analyses for the baiji were clearly multimodal, suggesting that this
species did not undergo population expansion or total recombination at the MHC-I, DRA and
DQB loci. In contrast, the mismatch distribution analyses supported that the finless porpoise
underwent a historical population expansion, which was congruent with the analysis by the
mitochondrial control region sequences (Yang et al., 2008).
Additional analysis following the method implemented in GENECONV software
revealed that three Neph-DQB and eight Live-DQB alleles were found to be involved in
intragenic recombination events in some sequence blocks (i.e., DNA block 30-152 and DNA
block 30-172, see Table 5). However, the intragenic recombination was not detected at the
DRA and MHC-I loci. As suggested by Yeager and Hughes (1999), intragenic recombination
was related to the search for MHC genes' diversity as an adaptive response. Thus, the present
high polymorphism at the DQB locus might be an outcome of intragenic recombination.