A predictive model of immunodominance hierarchies in influenza infection

— We present a model to account for the cellular phenomenon of immunodominance—the differential expansion of antigen-specific populations in response to infection. Specifically, our model addresses CD8+ T cell immunodominance in influenza infection. This model is based on experimental data and is responsive to two main parameters—the epitope density of antigen presented on the cell surface of infected or presenting cells and the precursor frequency of the responding T cell population. Novel methods for accurately quantifying both of these parameters are presented and the ability of the model to account for observed immunodominance phenomeneon is discussed. Immunodominance is the phenomenon of differential expansion of antigen-specific lymphocyte populations. It is observed in B cells as well as CD4+ and CD8+ T cells, often with a single specificity dominating 70-80% of the total response [1]. This phenomenon has direct implications for vaccine design and the control of highly mutable pathogens, which may be able to escape from the dominant response. Several efforts have been made to model immunodominance, though none have been experimentally verified [1]. Our model system, influenza infection in C57BL/6 mice, contains a strong immunodominance structure in the CD8+ T cell response, mirroring those seen in humans. We have attempted to model this response and rigorously quantify the parameters in the model, with the aim of developing a robust method for predicting immunodominance hierarchies and rationally designing vaccines. II. EXPERIMENTAL DESIGN There are equivalent numbers of the immunodominant D b NP 366 , D b PA 224 and K b PB1 703 specific responses in the memory phase following primary influenza infection, but challenge of such memory mice establishes a dramatic immunodominance hierarchy, with the D b NP 366 response