Variability in cellular signal transduction networks

Cellular variability is fundamental to physiological reality but usually unattended in signaling models. This thesis introduces the new approach of cell ensemble models, which describe biochemical signal transduction networks in heterogeneously behaving cells. Cell ensemble models comprise sets of coupled ordinary differential equations describing protein concentration trajectories in different cells, which are linked by boundary conditions restricting models to physiological limitations. Simultaneous description of single-cell and population data facilitated model discrimination and improved the accuracy of parameter estimations. The approach was applied in two biochemical systems, programmed cell death and the intracellular traffic of erythropoietin receptors. An experimental method was developed to quantify the enzymatic activity of caspase-8, which initializes programmed cell death, in single cells. The analytic solution of a death receptor oligomerization model was combined with cell ensemble models of caspase-8 activation. An activation mechanism, which implies positive feedback, was predicted and experimentally validated. Simulations based on estimated multivariate log-normal distributions of initial cellular protein concentrations clarified the functional roles of involved signaling proteins. In a similar manner, a cell ensemble model was applied to characterize cell-to-cell variability in intracellular erythropoietin receptor transport. The new approach might support optimization of therapeutic applications targeting heterogeneous populations of cancer cells