Abstract
A major barrier to finding a cure for human immunodeficiency virus type-I (HIV-1) infection is the existence and persistence of the HIV-1 latent reservoir. Although the size of the reservoir is shown to be extremely stable under effective antiretroviral therapy, multiple lines of evidence suggest that the reservoir is composed of dynamic and heterogeneous subpopulations. Quantifying the dynamics of these subpopulations and the processes that maintain the latent reservoir is crucial to the development of effective strategies to eliminate this reservoir. Here, we constructed a mathematical model to consider four latently infected subpopulations, according to their ability to proliferate and the type of virus they are infected. Our model explains a wide range of clinical observations, including variable estimates of the reservoir half-life and dynamical turnover of cytotoxic T lymphocyte (CTL) escape viruses in the reservoir. It suggests that very early treatment leads to a reservoir that is small in size and is composed of less stable latently infected cells (compared to the reservoir in chronically infected individuals). The shorter half-lives estimated from individuals treated during acute infection is likely driven by cells that are less prone to proliferate; in contrast, the remarkably consistent estimate of the long half-lives in individuals who are treated during chronic infection are driven by fast proliferating cells that are likely to be infected by CTL escape mutants. Our model shed light on the dynamics of the reservoir in the absence and presence of antiretroviral therapy. More broadly, it can be used to estimate the turnover rates of subpopulations of the reservoir as well as to design and evaluate the impact of various therapeutic interventions to purge the HIV-1 reservoir.
