| Abstract | Influenza A is an annual public health threat which has also been responsible for some of the deadliest pandemics in history. Rapid evolution and reassortment of its genetic structure allow the virus to subvert or delay an effective immune response, merge with strains natural to different host species (a characteristic of this years swine-avian-human outbreak), and develop resistance to anti-viral drugs. What had been the most reliable anti-influenza drug only two years ago, oseltamivir (or Tamiflu), faced resistance in 20% of H1N1 strains last year and preliminary data on this year's flu season shows a nearly 100% resistance to the drug. In order to understand the development of this resistance and to potentially avoid or reverse these effects in the future, it is important to understand how the genetic differences between resistant and non-resistant strains translate to differences in their infection kinetics. The virulence of a particular virus strain is typically assessed via in vitro plaque assay experiments. The initial infection of a single cell results in a circular plaque of infected or dead cells, whose rate of growth is an indication of the infection's virulence. In an effort to study how this growth depends on virus-cell interaction parameters including adsorption rate, production rate and time of latent infection, we have developed a reaction-diffusion model and a related spatially-discrete stochastic simulation. We apply the model results to the experimental plaque growth of three circulating influenza A H1N1 strains and their oseltamivir-resistant counterparts (including both natural and synthetic isolates). We map growth differences between the two strains to specific changes in key infection parameters and discuss the implications for the emergence of drug resistance. |
|---|
