Modelling the dynamics of immune escape and latency in HIV infection: implications for vaccination and treatment

Untreated human immunodeficiency virus (HIV) infection causes Acquired Immune Deficiency Syndrome (AIDS) that ultimately leads to death. The main obstacle to HIV eradication in a treated patient is the ability of virus to stay latent for many years in hosts even after the administration of antiretroviral therapy (ART). The cessation of ART leads to rebound of viral replication. The development of an effective vaccine is problematic due to HIV ability to mutate rapidly. This thesis investigates important aspects of latency and evolution of host immune response in HIV and its model infection in macaques. It addresses the problems related to vaccination against and treatment of HIV infection. I used mathematical modelling and statistical analysis to interpret and analyse the complex results of experiments.Virus-specific CD8 T-cells can detect and suppress HIV-infected cells, but they achieve only a temporary success. As soon as the virus mutates, viral variants can escape the established immune response. I demonstrated that the appearance of CD8 T-cell response, rather than viral variability, plays the dominant role in determining the time of immune escape. This discovery highlights a need of vaccines inducing immune response to conserved regions of HIV genome. Then, I investigated the antiviral mechanisms of CD8 T-cells in HIV patients. Previous studies revealed a variety of the possible antiviral mechanisms, whilst the contribution of each has never been estimated. I found that virus-specific CD8 T-cells kill infected cells fused with virus before they become productively infected. An effective CD8 T-cell vaccine should stimulate this “natural” response to conserved epitopes in order to suppress virus-containing cells, before they generate viral progeny. This can reconcile the previous contradictory results that pointed to both cytolytic and non-cytolytic mechanisms as dominant. With regard to HIV latency, I demonstrated a positive correlation between HIV DNA turnover in resting cells and chronic viral load for SIV epitopes with diversified quasispecies. This result suggests that under high viral loads there is no true latency. Using this result and mathematical modelling, I showed that the outcome of treatment with latency-reversing drugs can be improved by their administration at the start of ART