(393h) Redesigning the Response of T Cell Signaling Networks Using in silico Evolution

Cells have evolved an array of signal transduction pathways that allow them to detect and respond to their environments. T lymphocytes orchestrate adaptive immunity and act as cellular detectors of infection, using the T cell receptor (TCR) to identify molecular signatures of pathogens. The TCR signaling pathway exhibits a sharp activation threshold as a function of the binding strength of the TCR with peptide-MHC (pMHC) ligands presented on other cells. We are interested in the following question: Given a fixed network topology such as the TCR signaling network, can one find sets of parameters (kinetic rates) that give desired network responses to input? Answering such questions can help to shed insight into T cell signaling and may provide a means for designing artificial networks with desired signaling properties. Here we consider a deterministic, well-mixed model of the TCR signaling network and utilize an in silico evolutionary algorithm to search for sets of kinetic parameters that give altered output responses as a function of the off-rate between pMHC and TCR. We begin by shifting the activation threshold by orders of magnitude and find that the algorithm converges to solutions by altering kinetic rates in the network. Interestingly, we find that the TCR signaling network topology can exhibit even more dramatic output profiles, such as an inversion of the activation pattern in which weak TCR-pMHC binding induces activation and strong binding does not. By running multiple independent instances of the evolutionary algorithm for each desired output, we find many different sets of parameters consistent with the desired output, with distinct patterns of solutions become apparent when analyzing the resulting data. Overall, through moderate adjustments of kinetic parameters, the TCR signal transduction network has the potential to produce a wide array of input-output relations.