DESCRIPTION (provided by applicant): The overall goal of this proposal is to develop a quantitative understanding of how prior immunity affects the dynamics and evolution of recall immune responses to viruses. This requires developing mathematical models for the dynamics of virus and immunity following immunization and infection and validating the models experimentally. We choose the influenza system because of tractability (the ability to modulate the antigenic properties of the virus and measure the immune responses in model infections of mice), access to samples following immunization or infection of humans, and its direct relevance to human health. We have assembled a multi-disciplinary team with expertise in both theory and experimental studies in mice and humans influenza. The PI's of this multi-PI proposal Rustom Antia and Rafi Ahmed are long-term collaborators with experience in integrating modeling with experimental data in immunology. The team includes: Joshy Jacob (B & T cell responses to influenza), Brian Evavold & Veronika Zarnitsyna (T cell responses /affinity/diversity), Jacob Kohlmeier (resident memory T cells/influenza), Anice Lowen and John Steel (influenza virus engineering) at Emory University joined by Andreas Handel (UGA: modeling/statistics/T cell diversity/influenza); Paul Thomas (St. Judes: influenza in mice/humans), and Trevor Bedford and Erick Matsen (FHCRC: phylodynamics/statistics). In Project 1 Aim 1 we will develop models for the dynamics of humoral responses following immunization. In Aim 2 we model CD8 T cell responses to influenza viruses with focus on resident memory CD8 T cells in the lungs. Aim 3 combines models from Aim 1, 2 to generate a predictive model for how pre-existing immunity to influenza affects the dynamics of virus and immunity following challenge with new strains. We will use phylodynamics approach to infer the dynamics of B and T cell clones competition from sequencing virus- specific B and T cells. Aim 4 is devoted to the development and dissemination of user-friendly modeling tools that can be used by the wider research community for immunological modeling. In Project 2 Aims 1-3 we will validate our models from the corresponding Aims 1-3 of Project 1 in the mouse system by following the immune responses to reverse genetics-derived influenza viruses with modified B and T cell epitopes and recombinant hemagglutinin (HA) protein as viral antigens. This will lead to further model development in Project 1. Finally, in Aim 4 of Project 2 we will test the consistency of our verified and refined n mouse system models using humans samples banked from studies on vaccination with trivalent influenza vaccines and natural influenza infection. The validated models we develop will lay the foundation to explore different strategies for improving vaccination to influenza and other pathogens.

PUBLIC HEALTH RELEVANCE: Understanding the dynamics and evolution of recall immune responses is the foundation of effective vaccine development. We will use a combination of models and experiments to develop a quantitative framework for how the pre-existing humoral and cellular immunity alters the dynamics of virus and immunity following vaccination and infection. Project 1 - Modeling the dynamics and evolution of immune responses to influenza viruses (Description as provided by applicant): The overall goal of this proposal is to develop quantitative models that describe how prior immunity affects the dynamics and evolution of recall immune responses to influenza viruses. Our goal is to understand the fundamental rules underlying the dynamics of virus and immunity in pathogens that exhibit strain variation, and address the following questions: Why do responses to conserved epitopes not get boosted with each new influenza strain and generate strain-transcending immunity? What is the role of CD8 T cells during recall response, what is the nature of the protection from infection and pathology different cell subsets provide, and why do influenza-specific T cell epitopes show little variation We will answer these questions through three Specific Aims. In Aim 1, we will build and test compartmental mechanistic models to understand how prior humoral immunity affects recall response to different hemagglutinin molecules of influenza strains. We will develop multi-epitope models that track the dynamics of clones of B cells, plasma cells, antibodies and CD4 T cell help to different virus epitopes on the HA molecule. Our models describe how epitope masking by secreted antibodies raised against previous strains plays a key role in understanding competition between responses to different epitopes. In Aim 2, we will build and test models to understand how prior CD8 T cell immunity affects recall response to influenza. We will develop multi-epitope models for CD8 T cell responses and use them to identify the key parameters that regulate proliferation, competition and differentiation of CD8 T cells. For Aim 3, we will combine the models from Aims 1 and 2 to predict how the combined effect of preexisting humoral and T cell immunity affects the diversity and evolution of preexisting and stimulation of new clones of immune cells following sequential challenge with different influenza strains and vaccines. We will also develop models based on phylodynamic approaches to analyze next-generation sequences of the antigen receptors on virus-specific T and B cells. All our models will be validated through experiments that will be done concurrently in Project 2 of this application. Finally, Aim 4 of this project is devoted to the development and dissemination of user-friendly and powerful modeling tools that can be used by the wider research community for immunological modeling. We will design and write a new R package that allows graphical model building and analysis. We will also develop several tools for sequence analysis based on the BEAST and Galaxy platforms. By tapping into the infrastructure of existing, widely used modeling tools, we will be able to produce tools that are at the same time very user friendly and highly flexible.