Large-scale MHC class II genotyping of a wild lemur population by next generation sequencing

Abstract

The critical role of major histocompatibility complex (MHC) genes in disease resistance, along with their putative function in sexual selection, reproduction and chemical ecology, make them an important genetic system in evolutionary ecology. Studying selective pressures acting on MHC genes in the wild nevertheless requires population-wide genotyping, which has long been challenging because of their extensive polymorphism. Here, we report on large-scale genotyping of the MHC class II loci of the grey mouse lemur (Microcebus murinus) from a wild population in western Madagascar. The second exons from MHC-DRB and -DQB of 772 and 672 individuals were sequenced, respectively, using a 454 sequencing platform, generating more than 800,000 reads. Sequence analysis, through a stepwise variant validation procedure, allowed reliable typing of more than 600 individuals. The quality of our genotyping was evaluated through three independent methods, namely genotyping the same individuals by both cloning and 454 sequencing, running duplicates, and comparing parent-offspring dyads; each displaying very high accuracy. A total of 61 (including 20 new) and 60 (including 53 new) alleles were detected at DRB and DQB genes, respectively. Both loci were non-duplicated, in tight linkage disequilibrium and in Hardy-Weinberg equilibrium, despite the fact that sequence analysis revealed clear evidence of historical selection. Our results highlight the potential of 454 sequencing technology in attempts to investigate patterns of selection shaping MHC variation in contemporary populations. The power of this approach will nevertheless be conditional upon strict quality control of the genotyping data.

Fig. 1

Distribution of mean per-amplicon frequency (MPAF) of the first to sixth variants found in the same amplicon for 707 DRB and 644 DQB amplicons. MPAF distributions of the most common allele (MCA) for DQB (a) and DRB (b), of the second (c DQB and d DRB), third (e DQB and f DRB), fourth (g DQB and h DRB), fifth (i DQB and j DRB) and sixth MCA (k DQB and l DRB) are shown. Vertical line stands for MPAF = 0.05

Fig. 2

Plot of mean per-amplicon frequency (MPAF) in relation to allelic frequency (here indexed by the number of individuals carrying this allele) for 827 DQB (a) and 706 DRB (b) unique variant sequences, before the stepwise allelic validation procedure

Fig. 3

Phylogenetic tree of the Mimu-DQB and Mimu-DRB sequences observed and described in this study. Previously published sequences are marked with a black triangle —others are new. Mimu-DQB sequences exhibiting a 6-bp deletion are marked with plain grey circles. The tree configuration was derived from nucleotide sequences using the neighbour-joining method implemented in MEGA. HLA-DQB1*02 (accession number: FR798950.1) and HLA-DRB1*01 (accession number: FN821964.1) sequences are shown for comparative purposes. Accession numbers and nucleotide sequence of the Mimu-DQB and Mimu-DRB alleles are presented in Appendix

Fig. 4

Amino-acid variation plot for a DQB and b DRB alleles. Human antigen-binding sites (ABS) are indicated with the letter ‘h’, whereas positively selected sites (PSS) are indicated with black triangles. In the DRB plot there are no amino acids between the positions 24–25, because these alleles were only 163 bp long, whereas 19 DQB alleles were 169 bp long, with an insertion of two codons at positions 24–25

Fig. 5

Allelic distribution for a DQB alleles and b DRB alleles in the study population

Fig. 6

Estimation of population allelic richness through a resampling procedure counting the number of DQB alleles detected in relation to the number of individuals sampled. The dotted lines indicate the standard deviation of the estimated mean. The corresponding pattern is not shown for DRB, because it looks very similar

Fig. 7

Alignment of the DRB and DQB alleles retrieved from no more than one individual. Antigen-binding sites (ABS) are indicated with *, whereas positively selected sites (PSS) are highlighted in grey. Alignment gaps are indicated by dots