Stochastic Models of HIV Escape from CTL Attack

Georgetown University, Washington DC

Investigators

Abstract

During HIV infection, cytotoxic T lymphocytes (CTLs), a type of immune system cell, exert selective pressure on the HIV population through attack at multiple epitopes. Several recent datasets demonstrate that CTL attack can occur simultaneously at multiple epitopes, and that HIV escapes such attack through multiple mutation pathways that lead to rapid selective sweeps of mutant variants. However, HIV escape dynamics from CTL attack and the effect of such attack on HIV genetic diversity are not well understood. Current models of HIV escape from CTL attack typically involve deterministic dynamical systems that model CTL selection but not HIV mutation. In a different context, HIV genetic diversity is often explored through phylogeny construction using evolutionary models that include mutation but not CTL mediated selection. This project bridges these two approaches by developing stochastic birth-death models that include both CTL attack and HIV mutation. Using such models as a basis, this project involves construction of mathematical and computational tools through which the effect of CTL attack on HIV genetic diversity can be explored. Further, novel statistical methods and algorithms will be developed to infer HIV escape dynamics under multi-epitope CTL attack by exploiting both population dynamics and genetic sequence data. Cytotoxic T lymphocytes (CTLs) are immune system cells thought to play a central role in controlling HIV infection. CTLs kill cells infected by HIV, but HIV escapes this attack through mutation. HIV escape from CTL attack is complex as CTLs mount many simultaneous attacks while the HIV population escapes through many different mutations. Currently we do not have a quantitative understanding of HIV escape from CTL attack, especially in the context of multiple, simultaneous CTL attacks. One reason for this is that current mathematical and statistical techniques typically fully model either CTL attack or HIV mutation, not both. This project will develop mathematical, statistical, and computational tools for understanding HIV escape by bridging existing approaches that consider CTL attack with those that consider HIV mutation. Using these newly developed tools, HIV escape will be quantitatively explored through data-based inference and theoretical analysis. A quantitative description of HIV escape from CTL attack is useful in understanding the role of the immune system in controlling HIV infection and, potentially, in designing vaccines. Further, lessons learned from HIV may be transferred to the numerous other viruses exposed to CTL attack.