Date Published: May 11, 2004
Publisher: Public Library of Science
Author(s): Ying Wang, Michael Pierce, Lisa Schneper, C. Gökçe Güldal, Xiuying Zhang, Saeed Tavazoie, James R Broach
Abstract: Addition of glucose to starved yeast cells elicits a dramatic restructuring of the transcriptional and metabolic state of the cell. While many components of the signaling network responsible for this response have been identified, a comprehensive view of this network is lacking. We have used global analysis of gene expression to assess the roles of the small GTP-binding proteins, Ras2 and Gpa2, in mediating the transcriptional response to glucose. We find that 90% of the transcriptional changes in the cell attendant on glucose addition are recapitulated by activation of Ras2 or Gpa2. In addition, we find that protein kinase A (PKA) mediates all of the Ras2 and Gpa2 transcriptional effects. However, we also find that most of the transcriptional effects of glucose addition to wild-type cells are retained in strains containing a PKA unresponsive to changes in cAMP levels. Thus, most glucose-responsive genes are regulated redundantly by a Ras/PKA-dependent pathway and by one or more PKA-independent pathways. Computational analysis extracted RRPE/PAC as the major response element for Ras and glucose regulation and revealed additional response elements mediating glucose and Ras regulation. These studies provide a paradigm for extracting the topology of signal transduction pathways from expression data.
Partial Text: Complex intracellular networks inform a cell’s developmental and growth decisions in response to external nutrients or signaling molecules. Defining the topology of such networks has generally relied on combinations of genetic epistasis and biochemical techniques to establish the linear order of components that convey information on the presence of a particular stimulus. Generally, only one or a few endpoints, such as enhanced transcription of a responsive gene, are monitored in gauging the output of a pathway. More recently, global transcriptional analysis has allowed reseachers to capture the entire transcriptional output of a signaling process and assess the consequence of eliminating individual components of the signaling network on the entire response (Fambrough et al. 1999; Roberts et al. 2000). This approach has the potential to extract a complete description of a network from a relatively limited set of experimental perturbations.