Multi-scale models of T cell activation

Abstract

The overarching theme of this thesis is to develop and apply multi-scale computational techniques adopted from physical sciences to study a key phenomenon underlying the adaptive immune response: the activation of T cells. The specific objectives are: 1) develop efficient and versatile computational frameworks to study multi-scale biological systems in silico; 2) obtain mechanistic insights into how T cells are triggered in vivo. The first problem investigated in this thesis addressed a controversy regarding when and how T cells alter migratory patterns in lymphoid tissues, as observed in intravital microscopy experiments. By developing a lattice-based model for T cell migration coupled with a mechanistically motivated simple scheme for T cell activation, I showed that the quantity and quality of cognate antigen (Ag) presented by dendritic cells (DC) dictate such changes. The results from theoretical and computational analyses were not only in agreement with synergistic experiments, but also made predictions that have been tested positively. Furthermore, I identified a consolidated measure of Ag quantity and quality, which provides a unifying conceptual framework for considering diverse future experimental results. The results from this study also suggested that T cells may integrate sub-optimal signals derived from successive encounters with DCs to achieve full activation. However, an underlying molecular mechanism that may confer such "short term memory" of exposure to Ag is not known. I explored the possibility that the hysteresis resulting from positive feedback regulation of the catalytic conversion of a G-protein RasGDP to RasGTP in the T cell receptor (TCR) membrane-proximal signaling network may enable such "short term memory". I developed a multiscale computational model that combines stochastic simulations of the TCR membrane-proximal signaling network with T cell migration. The results showed that this hysteresis can enable T cells to integrate signals derived from weakly stimulatory DCs and may greatly enhance the detection sensitivity during disease onset when Ag presentation is low. The computational framework developed in this study can be readily adapted to examine diverse biological systems where signaling and cell motion need to be studied simultaneously. For example, the model was modified to investigate a DC-mediated mechanism for signal integration, and our results suggest that this mechanism is less likely. Initial steps were also taken to construct a macroscopic model that aims to study how T cell activation impacts observations at the organismic level. Preliminary results for how microscopic receptor-ligand interactions affect the proliferation of different T cell types are presented. Directions for future research are suggested based on these findings.