Heterogeneity and Decision-Making in Cellular Signaling Networks

Abstract

Signaling networks are the means by which cells adapt to their environment. Over the last decade, cellular decision-making has been shown to exist in the midst of substantial heterogeneity, even among isogenic cells. The presence of such variability is generally assumed to be an obstacle that cells overcome in order to precisely resolve information about their environment. In this work, we investigated the effects of two specific types of intracellular heterogeneity on signal transduction in cells. The first is the presence of substantial compositional heterogeneity in the sets of macromolecular complexes used for signal transduction, which we observe in a model of the yeast pheromone signaling system. In spite of this, the model is able to reliably reproduce experimentally observed dynamical and dose-response trends. We then contrasted this model with one that employs a hierarchically assembled, stable signaling complex and found that the two signaling paradigms can exhibit distinctive behaviors. These differences can be attributed in part to the role of the scaffold protein in signal complex assembly, which is required for signal transduction in the pheromone network. We found that features such as signal amplification and crosstalk prevention vary depending on how the assembly of scaffold-based signaling species occurs. Our results clearly show that a dynamical understanding of signal transduction must take place in the context of compositional heterogeneity. The second form of heterogeneity we consider, biochemical noise, occurs at a more fundamental level. We examined how variability in the response to signal impacts the ability of cells to make reliable decisions by quantifying signal transduction using concepts from information theory. Our results revealed the existence of a fundamental trade-off: increased noise in individual cells corresponds to increased information available to control cellular populations. To provide context for the general application of information theory to cell signaling, we characterized the upper limits of information transmission in models of simple signaling motifs. Our results also revealed that certain features of signaling networks, such as enzyme saturation and molecular copy number, are central to regulation of information transmission through networks of arbitrary size. With formal, systematic modeling approaches, we were able to elucidate many non-intuitive behaviors resulting from variability in signal transduction. Thus, we expect that our treatment of heterogeneity in signaling networks will form the basis for the development of a comprehensive theory of cellular decision-making.