Project summary
Antiretroviral agents reduce morbidity and mortality in HIV-infected individuals; however, mutations in the viral
genome often result in clinical resistance to their effects. Due to the random nature of the mutations, the
emergence of therapy-resistant mutants is mostly considered unpredictable. Consequently, antiviral treatment
strategies are mainly empiric. The premise of this proposal is that the emergence rate of therapy-resistant
mutants in each patient is largely predictable. Our published and preliminary studies suggest that changes in
virus proteins can be accurately forecasted based their sequence properties before initiation of treatment. In
the proposed study, we will use the example of the HIV-1 envelope glycoproteins (Envs). Several broadly
neutralizing antibody (BNAb) therapeutics that target Env have shown great promise in clinical trials. However,
escape from these agents often occurs after treatment, at different rates for different patients. The goal of this
work is to advance our ability to personalize BNAb therapeutics to people living with HIV, by establishing the
tools to determine the likelihood of each patient to develop resistance to each agent.
The central hypothesis of this proposal is that each swarm of HIV-1 that infects a patient has an inherent and
measurable likelihood to escape from each therapeutic. This likelihood is shared by the viruses that circulate in
the blood and the reservoir of latent viruses, which is often the source of resistant mutants that emerge after
therapy.
To test the above hypothesis, we will first determine if the rate of HIV-1 escape from BNAbs is specific for each
swarm of the virus that infects a patient. To this end, we will test in vitro the escape of strains from different
patients that were isolated from samples collected at different time points. We will then test the ability of our
models to forecast the rate and site of escape for each strain. Based on these studies, our models will be
refined and applied to determine their ability to forecast resistance rates in four clinical trials of BNAbs in
humans. Next, we will determine whether the appearance of resistant mutants is driven by their fitness (i.e.,
higher likelihood to appear before treatment) or by their resistance (i.e., higher replicative capacity after
treatment). Such knowledge is critical for our ability to use patient samples before treatment, which inform of
the viral fitness profiles, to predict escape from the treatment. We will then examine whether the fitness profile
of BNAb escape sites is a persistent property of each virus swarm by measuring the changes that occur over
time in patients. Finally, we will induce outgrowth of latent HIV-1 from peripheral blood cells and compare their
fitness profiles at BNAb escape sites with those of viruses that circulate in the blood.
The models to be developed have the potential to make important contributions to the treatment of patients by
antiviral agents. They will lay the foundations for personalized antiviral medicine that is based on the likelihood
of each virus swarm in a patient to develop resistance to each agent.