Spatiotemporal dynamics of HIV propagation.

Although viral propagation is a localized process, mathematical models of viral replication kinetics have been limited to systems of ordinary differential equations describing spatially averaged behavior. In this paper, we introduce a cellular automaton model of viral propagation based on the known biophysical properties of HIV. In particular, we include the competition between viral lability and Brownian motion. The model predicts three testable effects not present in previous descriptions. First, we find a profound dependence of viral infectivity on cell concentration; virion instability decreases infectivity more than 100-fold under typical experimental conditions, resulting in misleading estimates of the number of infectious particles. Second, we find that, in a large parameter regime, infection extinguishes itself due to insufficient target cell replenishment. Finally, we find that propagation is limited by viral stability at low cell density and by geometry at high cell density. The geometry-limited regime can be modulated by downregulation of CD4. These different properties are analysed quantitatively and compared with previous experimental results.