Project Summary Immunological memory is a defining characteristic of vertebrate immune systems, and is encoded in part by populations of T lymphocytes that respond rapidly and potently to repeat exposures to pathogens. These populations are remarkably diverse in their phenotypes and trafficking patterns, and collectively form an ecosystem that reflects a lifetime's exposure to pathogens and to our environment. Yet a fundamental and pressing problem remains - we have a very limited quantitative grasp of how immunological memory is established and maintained, how new memories to infections are integrated with existing memories, and how and why these wane over time. A detailed understanding of these phenomena would guide the design of vaccines that induce durable immunity, and the development of therapies to revive and sustain T cell responses in the settings of chronic infections and cancer. The memory encoded by circulating T cells is highly dynamic. It is sculpted by self-renewal, differentiation, loss, and supplementation or competition with newly-recruited cells. Further, in both mice and humans, memory induced by pathogens co-exists with abundant 'natural' memory T cells, which are specific for commensal and environmental antigens but provide cross-protection to new infections; and with regulatory T cells, that limit inappropriate responses. We have very little understanding of how these three populations relate. In this project we will integrate together a powerful set of modeling and experimental approaches to confront these uncertainties. Specifically, we will develop a structured population modeling (PDE) approach to use with a fate-mapping system that tracks T cell populations throughout their life-histories. We will then combine the fate mapping system with a DNA labeling method and ODE models to define memory T cell dynamics in detail. Together, these approaches will map the development, structure, and rules of replacement of circulating memory T cell subsets in mice. We will also use data from a novel division-reporter mouse strain to model the emergence and long-term dynamics ofT cell memory to influenza infection. Finally, we will combine modeling with fate-mapping and division-reporter mice to understand how regulatory T cells are maintained, and how they impact both natural and influenza-specific memory populations.