Mechanisms of Abrupt Loss of Virus Control in a Cohort of Previous HIV Controllers

Abstract

Elite and viremic HIV controllers are able to control their HIV infection and maintain undetectable or low-level viremia in the absence of antiretroviral treatment. Despite extensive studies, the immune factors responsible for such exclusive control remain poorly defined. We identified a cohort of 14 HIV controllers that suffered an abrupt loss of HIV control (LoC) to investigate possible mechanisms and virological and immunological events related to the sudden loss of control. The in-depth analysis of these subjects involved the study of cell tropism of circulating virus, evidence for HIV superinfection, cellular immune responses to HIV, as well as an examination of viral adaptation to host immunity by Gag sequencing. Our data demonstrate that a poor capacity of T cells to mediate in vitro viral suppression, even in the context of protective HLA alleles, predicts a loss of viral control. In addition, the data suggest that inefficient viral control may be explained by an increase of CD8 T-cell activation and exhaustion before LoC. Furthermore, we detected a switch from C5- to X4-tropic viruses in 4 individuals after loss of control, suggesting that tropism shift might also contribute to disease progression in HIV controllers. The significantly reduced inhibition of in vitro viral replication and increased expression of activation and exhaustion markers preceding the abrupt loss of viral control may help identify untreated HIV controllers that are at risk of losing control and may offer a useful tool for monitoring individuals during treatment interruption phases in therapeutic vaccine trials.IMPORTANCE A few individuals can control HIV infection without the need for antiretroviral treatment and are referred to as HIV controllers. We have studied HIV controllers who suddenly lose this ability and present with high in vivo viral replication and decays in their CD4⁺ T-cell counts to identify potential immune and virological factors that were responsible for initial virus control. We identify in vitro-determined reductions in the ability of CD8 T cells to suppress viral control and the presence of PD-1-expressing CD8⁺ T cells with a naive immune phenotype as potential predictors of in vivo loss of virus control. The findings could be important for the clinical management of HIV controller individuals, and it may offer an important tool to anticipate viral rebound in individuals in clinical studies that include combination antiretroviral therapy (cART) treatment interruptions and which, if not treated quickly, could pose a significant risk to the trial participants.

Keywords: HIV-1 control; HIV-1 progression; cell tropism; host genetics; in vitro virus inhibition; loss of control.

FIG 1

HLA class I allele frequencies. HLA-B allele frequencies are shown for the general Spanish population (64), a cohort of elite controllers (EC, n = 38), viremic controllers (VC, n = 27), and individuals that have abruptly lost their capacity to control the infection (LoC, n = 14).

FIG 2

Evolutionary relationships of taxa. The evolutionary history was inferred using the neighbor-joining method (65). The optimal tree with the sum of branch length of 1.91362276 is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) is shown next to the branches (66). The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Tamura-Nei method (67) and are in the units of the number of base substitutions per site. The rate variation among sites was modeled with a gamma distribution (shape parameter = 1). The analysis was based on a 66-nucleotide amplicon in the V3-loop of HIV gp120. All ambiguous positions were removed for each sequence pair. Evolutionary analyses were conducted in MEGA5 (68).

FIG 3

Number of HLA escape mutations in Gag (23) detected in viral population from samples before and after loss of control.

FIG 4

HIV-specific T-cell responses. (A) Breadth of HIV-specific T-cell responses before and after loss of control (n = 12). (B) Magnitude of responses, expressed as SFC/million PBMC, is shown longitudinally (C) Gag dominance (magnitude of Gag responses/total magnitude) is shown before and after evolution in the 12 LoC subjects and compared to a cohort of elite (EC) and viremic (VC) controllers (n = 65). Individuals that maintained an CCR5 tropism are shown in black dots, individuals that changed to a CXCR4 are shown in gray dots, and the individual from which tropism was not available is shown in white dots. P values are shown for Wilcoxon signed rank test in a comparison of paired data and for Mann-Whitney test to compare between groups.

FIG 5

Viral inhibition capacity. Levels of CD8⁺ viral inhibitory capacity are shown for IIIB E/T 1:1 (A) and E/T 1:10 (B) in individuals who presented a loss of control (n = 8) and long-term controllers (n = 8). Comparison of antiviral capacity is expressed as % inhibition = [(fraction of p24⁺cells in CD4⁺ T cells cultured alone) − (fraction of p24⁺cells in CD4⁺ T cells cultured with CD8⁺ T cells)]/(fraction of p24⁺ in CD4⁺ T cells cultured alone) × 100.

FIG 6

Levels of activation and exhaustion markers. (A) Percentage of expression of CD38, HLA-DR, PD-1, CD25, and CD69 on total CD8⁺ T cells in individuals that experience a loss of control (n = 8) tested before [pre (C)] or after [post (LoC)] loss of viral control. Values are compared to staining in samples from persistent HIV controllers taken at a median of 102 months apart. Markers were assessed following the gating strategy shown in Fig. S1. P values are shown for Wilcoxon signed rank test. (B) Levels of expression of PD-1 in subsets of CD8⁺ T cells populations based on CCR7 and CD45RA expression. Median and interquartile range values are shown for a group of 8 individuals that loss of control compared to 8 individuals maintained control over time. The gating strategy used is shown in Fig. S1. P values are shown for Wilcoxon signed rank test.