O'Shea Lab > Projects

Systems Analysis

Many cellular signaling systems exhibit greater complexity in their input-output relationships than what one would expect based on the properties of their protein components. Although most proteins behave in a Michaelian manner, some signaling pathways act as irreversible switches, others as oscillators, and many exhibit remarkable robustness to variations in parameters. We seek to discover how the architecture of signaling pathways generates this complexity. Much of our current effort is focused on homeostatic systems involved in the regulation of intracellular nutrient levels, including the phosphate-responsive signaling (Pho) pathway involved in phosphate homeostasis in budding yeast. Our approach to this problem involves monitoring signaling in single, living cells using fluorescent reporters, combined with computational modeling to make predictions about systems behavior. Our experimental analysis has uncovered interesting properties in the phosphate homeostasis pathway, including buffering, differences in thresholds of expression of genes, bistability, and hysteresis. We are now focused on investigating the origins of these properties and how they relate to the ability of this system to maintain homeostasis. Through a comparative analysis of nutrient regulatory systems in yeast, we hope to develop a computational model to explain how the wiring of nutrient regulatory systems gives rise to homeostasis.

Genetic, Cell Biological and Biochemical Analysis

The Pho pathway has been an excellent model system with which to investigate basic mechanisms of signal transduction and the regulation of transcription factor activity. Although this work has revealed much about the protein kinase, Pho80-Pho85, and the regulation of its transcription factor substrate Pho4, we know little about how phosphate is sensed by cells and how changes in phosphate levels lead to changes in the kinase activity of Pho80-Pho85. We are taking two approaches to address these questions. First, to understand how cells sense phosphate levels, we have identified and are characterizing genes involved in phosphate sensing that function upstream of Pho80-Pho85. Second, we are investigating connections between metabolism and phosphate sensing. Our recent work indicates that inositol pyrophosphates play a role in the regulation of Pho80-Pho85 by the CDK inhibitor Pho81. We are studying the mechanism of this regulation using in vivo and in vitro approaches and will employ more general methods to globally investigate changes in metabolites in response to phosphate limitation.

Transcriptional Regulatory Network Architecture and Function

Regulatory networks control gene expression in a manner reflecting the level and combination of many signals. Although many components of regulatory networks have been identified, we do not understand how such networks process external signals and achieve fine-tuned changes in the expression of individual genes. In collaboration with Aviv Regev (MIT/Broad Institute) and Nir Friedman (Hebrew University) we are studying in depth a specific regulatory system - the budding yeast response to osmolarity involving the evolutionarily conserved HOG MAPK pathway, which regulates the activity of several transcription factors to control expression of hundreds of genes. We are focused on the following questions: How does a single kinase, Hog1, control a diverse fine-tuned response? What is the specific role of each transcription factor? What is the functional significance of employing multiple transcription factors? How does the effect of the same signal translate to different transcriptional responses for individual genes? To answer these questions we are combining genomics, molecular, and computational approaches. Our goal is to develop, for the first time, a comprehensive model of gene regulation – from the level of gene promoters up to the biological and molecular conditions that initiate the response.

Transcriptional Regulatory Network Evolution

In collaboration with Aviv Regev (MIT/Broad Institute) we are using the HOG system as a model to investigate the evolution of a transcriptional regulatory network. This work is enabled by the genome sequences of a number of yeast species, and also by our construction of a quantitative transcriptional network of the S. cerevisiae HOG-dependent response to salt stress. We will use a combination of experiments and computational approaches to first characterize how this network has evolved from S. pombe to S. cerevisiae and then to explain the mechanisms allowing these regulatory changes.

Organisms exhibit oscillations in metabolism and behavior that are entrained by the environmental light-dark cycle. Circadian rhythms are present even in single-celled cyanobacteria, which use these oscillations to alternate photosynthesis with nitrogen fixation, two mutually exclusive biochemical processes. It is generally thought that circadian rhythms rely on an autoregulatory transcription and translation feedback cycle to generate oscillations. In the cyanobacterial circadian oscillator, genetic and biochemical studies have revealed that three proteins (Kai A, B, and C) play an important role in the oscillator and that KaiC undergoes changes in phosphorylation state with periodicity that corresponds to the circadian cycle. Remarkably, the Kondo group has recently demonstrated that these three proteins (plus ATP) are sufficient to produce sustained oscillations in the phosphorylation state of KaiC in vitro. We are collaborating with Daniel Fisher (Stanford Applied Physics) to combine experiments with modeling to uncover the basis of oscillations. We are also interested in developing a biochemical understanding of how the clock receives environmental input and directs output to regulate processes such as gene expression.