A multilevel logic for antigen interpretation governs tissue-specific T cell responses 2259289

Abstract Introduction The immune system must balance between protecting the host and limiting collateral damage. However, this tradeoff is almost certainly not uniform as different tissues operate under distinct physiological constraints, pressures, and regenerative capabilities. How tissue-specific tradeoffs are communicated to the immune system remains unclear. This question is particularly relevant for T cells, which possess unique antigen-binding receptors generated through random gene recombination. Thus, a T cell has no intrinsic knowledge of the context in which it will encounter its cognate antigen. Therefore, its fate–whether to become inflammatory or tolerogenic–is determined by how it interprets antigen within its local microenvironment. Since most T cell responses are initiated in lymph nodes (LNs), which collect and integrate information from upstream peripheral tissues, we hypothesized that LNs encode a logic of antigen interpretation that reflects the constraints and tradeoffs of these draining tissues. Methods I conducted high-resolution multiplexed imaging of homeostatic LNs that drain different tissues, measured their baseline cytokine profiles, and leveraged multiple transgenic and KO mouse models, along with adoptive T cell and dendritic cell (DC) transfer experiments. Results Tissue-specific immune regulatory logic follows a hierarchical structure within lymph nodes: DCs → regulatory T cells → conventional T cells. In particular, the state of migratory DCs–derived from peripheral tissues–and the cytokine environment varied substantially across homeostatic LNs, establishing tissue-specific biases in T cell fate. These biases were revealed upon acute regulatory T cell depletion or by activating monoclonal T cells in different sites. Conclusion Our findings reveal a multilevel logic of immune regulation in which migratory DCs and local cytokine milieus shape regulatory T cell and conventional T cell interactions, enabling LNs to translate peripheral tissue constraints into distinct immune programs. Funding Source Department of Energy Computational Science Graduate Fellowship, Mark and Lisa Schwartz Foundation - Ragon Institute Initiative in Artificial Intelligence Machine Learning, and Immunology Topic Categories Computational and Systems Immunology (COMP)