The impact of HLA polymorphism on herpesvirus infection and disease

Abstract

Human Leukocyte Antigens (HLA) are cell surface molecules, central in coordinating innate and adaptive immune responses, that are targets of strong diversifying natural selection by pathogens. Of these pathogens, human herpesviruses have a uniquely ancient relationship with our species, where coevolution likely has reciprocating impact on HLA and viral genomic diversity. Consistent with this notion, genetic variation at multiple HLA loci is strongly associated with modulating immunity to herpesvirus infection. Here, we synthesize published genetic associations of HLA with herpesvirus infection and disease, both from case/control and genome-wide association studies. We analyze genetic associations across the eight human herpesviruses and identify HLA alleles that are associated with diverse herpesvirus-related phenotypes. We find that whereas most HLA genetic associations are virus- or disease-specific, HLA-A*01 and HLA-A*02 allotypes may be more generally associated with immune susceptibility and control, respectively, across multiple herpesviruses. Connecting genetic association data with functional corroboration, we discuss mechanisms by which diverse HLA and cognate receptor allotypes direct variable immune responses during herpesvirus infection and pathogenesis. Together, this review examines the complexity of HLA-herpesvirus interactions driven by differential T cell and Natural Killer cell immune responses.

Keywords: EBV; HCMV; Herpesvirus; Human Leukocyte Antigen; Immunogenetics; VZV.

Fig. 1

Shared and distinct HLA alleles are associated with herpetic risk. Odds ratios from 38 case/control and GWAS studies were used for a meta-analysis of HLA associations with herpesvirus infection or disease. A, B Odds ratios (ORs) and 95% confidence intervals are plotted from any HLA class I (A) or class II (B) allele with at least two significant associations. Non-significant associations are faded, while significant associations are bold (as determined from p values reported in source publication). Arrows indicate the confidence intervals extend through the bounds of the plot. Points are sized according to the number of participants in each study. Two analyses were performed on published ORs: using only two-field allele data or using both one and two-field data. Significant HLA:virus phenotype associations from two-field only data are provided in the margin in A, B. C Model output from the meta-analysis of herpesvirus-HLA associations using one and two-field data. For each allele having a significant association with at least one virus, the posterior probability distribution of the log odds ratio is plotted, estimated from association with all measured herpesvirus phenotypes. Colored lines and points show the estimated effect size of the individually significant virus-specific genetic associations. Associations specific to an allele at two-field resolution are indicated. All virus-specific effects are shown for A*01 and A*02, each of which had significant general associations across herpesviruses. D A subset of alleles differentially impacts herpesvirus infection or disease. Shown are specific genotypes (right) that have significant, but opposite, impacts on herpesvirus infection and subsequent disease. Measured infection (i.e., IgA seropositivity, shedding, and herpes zoster (HZ)) and disease (NPC, KS, and PHN) phenotypes are shown below the plot and colored by virus. Genotypes with a semicolon are joint KIR/HLA genotypes. C1 and Bw4 are HLA ligands that interact with KIR