The immunogenetic diversity of the HLA system in Mexico correlates with underlying population genetic structure

Abstract

We studied HLA class I (HLA-A, -B) and class II (HLA-DRB1, -DQB1) allele groups and alleles by PCR-SSP based typing in a total of 15,318 mixed ancestry Mexicans from all the states of the country divided into 78 sample sets, providing information regarding allelic and haplotypic frequencies and their linkage disequilibrium, as well as admixture estimates and genetic substructure. We identified the presence of 4268 unique HLA extended haplotypes across Mexico and find that the ten most frequent (HF > 1%) HLA haplotypes with significant linkage disequilibrium (Delta'>= 0.1) in Mexico (accounting for 20% of the haplotypic diversity of the country) are of primarily Native American ancestry (A*02 similar to B*39 similar to DRB1*04 similar to DQB1*03:02, A*02 similar to B*35 similar to DRB1*08 similar to DQB1*04, A*68 similar to B*39 similar to DRB1*04 similar to DQB1*03:02, A*02 similar to B*35 similar to DRB1*04 similar to DQB1*03:02, A*24 similar to B*39 similar to DRB1*14 similar to DQB1*03:01, A*24 similar to B*35 similar to DRB1*04 similar to DQB1*03:02, A*24 similar to B*39 similar to DRB1*04 similar to DQB1*03:02, A*02 similar to B*40:02 similar to DRB1*04 similar to DQB1*03:02, A*68 similar to B*35 similar to DRB1*04 similar to DQB1*03:02, A*02 similar to B*15:01 similar to DRB1*04 similar to DQB1*03:02). Admixture estimates obtained by a maximum likelihood method using HLA-A/-B/-DRB1 as genetic estimators revealed that the main genetic components in Mexico as a whole are Native American (ranging from 37.8% in the northern part of the country to 81.5% in the southeastern region) and European (ranging from 11.5% in the southeast to 62.6% in northern Mexico). African admixture ranged from 0.0 to 12.7% not following any specific pattern. We were able to detect three major immunogenetic clusters correlating with genetic diversity and differential admixture within Mexico: North, Central and Southeast, which is in accordance with previous reports using genome-wide data. Our findings provide insights into the population immunogenetic substructure of the whole country and add to the knowledge of mixed ancestry Latin American population genetics, important for disease association studies, detection of demographic signatures on population variation and improved allocation of public health resources.