The MHC (Major Histocompatibility Complex) class I and II genes, also known as HLA genes in humans (Human Leukocyte Antigens), are highly polymorphic and there is a substancial body of evidence showing that they have been a target of balancing selection. In addition, these loci have also been associated with more diseases than any other region in the genome. While diversity and differentiation at HLA coding variants have been extensively studied, less is known about their regulatory variation and population-level expression patterns. This is partly because high-throughput technologies of gene expression analysis have serious limitations when applied to HLA genes, e.g. the mapping of transcripts to a reference genome is problematic, since the high degree of polymorphism results in a great divergence between an individual's alleles and the reference sequence, resulting in a high degree of bias and/or inaccuracy when estimating expression patterns for these genes. Here we propose specific approaches to overcome these difficulties, allowing an unbiased quantification of expression levels of HLA genes from generic RNAseq assays. This information on expression levels will allow us to 1) draw a comprehensive multi-population map of HLA gene expression; 2) identify genetic variants which regulate HLA expression levels (i.e., expression QTLs or eQTLs); 3) investigate if HLA loci show patterns of allelic-specific expression (i.e., unequal expression levels between two alleles of a gene). These results will allow us to test the evolutionary hypotheses that HLA balancing selection results from the coexpression of the two alleles of a heterozygote, which maximizes the response to pathogens. Additionally, we will use the eQTLs to test whether the regulatory variants at HLA loci are also targets of strong balancing selection, as are the coding variants. We aim to release our methods as a bioinformatic package to be used by the scientific research community.
News published in Agência FAPESP Newsletter about the scholarship: