Optimizing gene therapies to suppress human immunodeficiency virus with RNA interference: Integrated molecular-level simulations and experimental implementation predict and elucidate the evolution of viral resistance.

摘要

This dissertation describes the integration of computational and experimental investigations of the processes by which the human immunodeficiency virus (HIV) evolves when challenged with a novel, potent antiviral agent. Current chemotherapeutic agents have proven insufficient for controlling the global HIV-AIDS epidemic, and the efficacy of existing therapies is threatened by the continuous, rapid emergence of drug resistant HIV strains. Genetic therapies present a highly promising alternative with the potential to engineer a patient's immune system to specifically suppress the invading virus. One of the most promising approaches is RNA interference (RNAi)---an innate cellular function that can be harnessed to specifically silence HIV genes. Although RNAi can effectively block viral replication, HIV's ability to rapidly evolve could eventually render treatments based on this technology ineffective.; In this study, we sought to gain a mechanistic understanding of how HIV replicates and evolves when targeted by potent RNAi. We constructed a novel type of agent-based stochastic computer simulation that incorporates the molecular-level mechanisms of these processes, and these simulations made several clinically relevant predictions. In parallel, we developed a novel antiviral RNAi inhibitor in an experimental system that allowed us to test several key hypotheses and thereby also directly validate our model. Together, these systems allowed us to predict and confirm the existence of a critical efficiency threshold for induction of RNAi in a cell population, such that when this threshold is crossed the probability of HIV escape suddenly increases from very low to very high. Moreover, our data suggest that HIV escapes from this challenge by a novel complex and cooperative mechanism that differs substantially from the evolution of drug resistance. We also demonstrated that RNAi can be used in combination with antiviral drugs to more effectively suppress HIV. In related work, we developed a novel process for improving the production of adeno-associated virus-based gene delivery vectors, which should help to increase overall delivery efficacy. These results should prove helpful as RNAi moves from the laboratory to the clinic, and they illustrate that evolutionary considerations can and must be rigorously and quantitatively incorporated into the design of effective antiviral therapies.