| Abstract | The important role of stochastic effects in the immune system in general has been investigated for a long time, and the interest in this topic has increased further by the finding that the search of B and T cells search for antigen in lymph nodes is apparently realized by a random walk. Several groups have constructed computational models of this microscopical search process in lymphoid tissue. Our modelling approach links these efforts to stochastic effects observed on the macroscopic scale: Lymphocytes continuously recirculate through lymph nodes, blood and lymph, the lung and other tissues, and thus the microscopic random walk through lymph nodes is coupled with a macroscopic random walk through the organism. We derive a quantitative model that integrates the two levels. Using techniques borrowed from randomized algorithm analysis, we show how both tissue-level and anatomical factors determine the efficiency and robustness of the search process. A good agreement between the optimal T cell residence time predicted by our model and the value found in mice indicated that the kinetics of lymphocyte recirculation have evolved as a solution to a two-level combinatorial optimization problem. Finally, we discuss implications of our work for the currently debated problem of how lymphocytes exit from the lymph node. |
|---|
