Background: Antiretroviral therapy (ART) suppresses HIV-1 replication, but an inducible viral reservoir persists in resting CD4+ T (rCD4) cells. We have found that the amount of inducible HIV-1 RNA following ex vivo activation of purified rCD4 cells varies 1000-fold across donors. Here, we sought to identify the reasons for this wide variation in inducible HIV-1.
Methods: We studied 10 consecutive HIV-1 infected subjects on suppressive ART with HIV-1 RNA in plasma <50 cps/mL for ≥1.4 years (median: 11 years). Plasma and PBMC were collected and rCD4 cells were purified by negative selection. rCD4 cells were treated for 7 days with anti-CD3/CD28 coated beads and the amount of virus produced was quantified in culture supernatants by qPCR for HIV-1 RNA. We evaluated 9 continuous variables for potential relationships with the amount of inducible HIV-1. Correlations between continuous variables and the inducible reservoir were evaluated by Spearman’s correlation. A principle component analysis (PCA) and regression (PCR) were performed using the same 9 variables to identify qualities associated with the amount of inducible HIV-1.
Results: Table 1 shows the 9 variables measured along with the amount of induced HIV-1 per million rCD4 cells. No individual variable correlated significantly with the amount of inducible HIV-1. PCA/PCR was performed to determine the proportion of variation in inducible HIV-1 that could be explained by linear combinations of the 9 variables analyzed. The first 3 dimensions of the PCA explained >80% of the variance in inducible HIV-1. Inducible HIV-1 was positively associated with combinations of pre-ART plasma HIV-1 RNA, residual plasma viremia, and cellular HIV-1 RNA and DNA in PBMC; and inversely associated with CD4 counts at nadir and on suppressive ART.
Conclusions: 80% of the inter-patient variation in the amount of inducible HIV-1 from resting CD4+T-cells was explained by 3 principle components that included pre-ART plasma HIV-1 RNA, residual plasma viremia on ART, cell-associated HIV-1 DNA and RNA levels, and CD4 cell counts pre- and post-ART. This finding supports a model in which no one variable predicts the size of the latent inducible reservoir, but combinations of immunologic and virologic measures explain much of the observed variation in inducible HIV-1.