Role of HLA Adaptation in HIV Evolution

Abstract

Killing of HIV-infected cells by CD8(+) T-cells imposes strong selection pressure on the virus toward escape. The HLA class I molecules that are successful in mediating some degree of control over the virus are those that tend to present epitopes in conserved regions of the proteome, such as in p24 Gag, in which escape also comes at a significant cost to viral replicative capacity (VRC). In some instances, compensatory mutations can fully correct for the fitness cost of such an escape variant; in others, correction is only partial. The consequences of these events within the HIV-infected host, and at the population level following transmission of escape variants, are discussed. The accumulation of escape mutants in populations over the course of the epidemic already shows instances of protective HLA molecules losing their impact, and in certain cases, a modest decline in HIV virulence in association with population-level increase in mutants that reduce VRC.

Keywords: CD8+ T cells; HIV-1; HLA class I; viral adaptation; viral fitness; viral replicative capacity.

Figure 1

HLA-B*27 footprints in B clade HIV show a predictable pattern of escape mutation among six different epitopes. The 12 HLA-B*27-associated escape variants within six HLA-B*27-restricted epitopes are shown in red. Shaded are the common escape positions at P2 in the epitope in each case, in all cases from Arg → Lys or Gly. Data from Carlson et al. (44, 75).

Figure 2

Impact of transmission of escape mutant to HLA-matched and to HLA-mismatched recipient. Two examples are shown to illustrate differing impact of transmission of escape variants. Left: transmission of escape mutation (R264K) where the virus is fully compensated by the compensatory mutation S173A. Right: transmission of escape mutation (T242N), where the compensatory mutant H219Q does not fully restore viral fitness (H219Q).

Figure 3

Variable speed of variant accumulation. (A) Frequency of HLA-B*35:01-driven D260E variant in a C clade-infected population in which HLA-B*35:01 is at low prevalence (phenotypic frequency 4%) compared to a B clade-infected population in which HLA-B*35:01 is at high prevalence (15%). (B) Frequency of HLA-B*07:02-driven S357G variant in two comparable Durban cohorts, enrolled in 2002–2005 and 2012–2013. These examples shown in (A,B) illustrate escape mutants selected early in the course of infection by a high proportion of subjects expressing the relevant HLA allele, which confer little fitness cost on the virus. Therefore, the variant accumulates rapidly in the population. (C) Frequency of HLA-B*57/58:01 A83G variant in two cohorts. The third example of rapid accumulation of an escape variant selected early in the course of infection, conferring little fitness cost that accumulates rapidly in the population. (D) Frequency of HLA-B*81:01-driven variant T186X. This example illustrates a variant conferring a high fitness cost on the virus, but for which compensatory mutations have only very modest effects. Therefore, the T186X variant accumulates very slowly in the population. Data from Kawashima et al. (47), Leslie et al. (93), and Payne et al. (96).