Class I Human leukocyte antigen (HLA-I) is a major driver of HIV evolution, both at the individual and population level, promoting HIV adaptation to cellular immune responses. The extent to which HIV adaptation to HLA-I plays a role in transmission is not well understood. Here, we examined associations between HLA-I profiles and HIV transmission in the Mexico City HIV epidemic.
1,049 HIV-1 subtype B pol sequences sampled between 2016 and 2018 from unique, HLA-I-typed individuals in Mexico City were analyzed. Genetic transmission networks were inferred using HIV-TRACE, establishing putative transmission linkage below a genetic distance threshold of 1.5%. High-resolution HLA profiles were determined using next-generation sequencing. Newman’s assortativity coefficients were estimated using igraph. Fisher’s exact tests were used to determine whether there was enrichment of specific HLA alleles in clustering vs. non-clustering individuals. P-groups, known to bind similar peptides, were used for HLA-match analyses.
286/1,049 (27%) individuals were genetically linked with at least one other person, forming 120 clusters (range: 2-8 individuals). All but 2 clustering individuals were male. Clustering and non-clustering individuals did not differ by age, baseline CD4 or HIV RNA level. HLA-C*02:02 was enriched in clustering individuals (p=0.02). Overall 30% (86/286) of clustering individuals shared ≥2 HLA-I P-groups/alleles in any of the three loci (Fig. 1A), and 26 had fully concordant (i.e. two matching alleles) HLA-A (13; 4.5%), HLA-B (2; 0.7%) or HLA-C (11; 3.8%) loci (Fig. 1B). Rates of HLA-I allelic concordance among clustering individuals were significantly higher than among the full cohort at all three loci (p<0.01).
HLA-I haplotypes were significantly more concordant than expected among clustering individuals in Mexico City. These findings suggest that viral adaptations may enhance transmissibility. However, further work is needed to determine if this increased concordance is due to viral factors (i.e. adaptation) or sociodemographic factors (i.e. ancestry, racial assortativity).