UCSF

Living systems operate at many scales, from biochemical reactions of individual atoms and molecules to complex behaviors of cells and organisms, and even evolutionary adaptation of entire ecosystems. Understanding the relationships between these processes, namely how changes at one scale propagate to other scales, is a fundamental pursuit of biology. One such complex propagation is called a genotype-phenotype map, defined here as how a protein mutation impacts its function in the context of its molecular interaction network to ultimately alter cellular fitness. Our generally poor understanding of this propagation limits our prediction of the effects of disease mutations and our ability to rationally engineer mutations for precisely tuning protein function in the dynamic cellular environment. In this dissertation, I present two studies of the small GTPase switch Gsp1, the S. cerevisiae homolog of human Ran, which uncover novel allosteric mechanisms governing how the effects of point mutations propagate from the molecular to the cellular scale.In Chapter 1, I outline the systems biology approach to studying molecular interaction networks, introduce the components of the network of Ran/Gsp1, and motivate the use of mutagenesis in the study of protein structure and cellular function. In Chapter 2, I describe the genetic and physical interaction profiling of point mutations in Gsp1 partner interfaces, which led to the discovery of novel allosteric sites coupled to the GTPase switch, as confirmed by enzyme kinetics and 31P nuclear magnetic resonance. Analysis of the genetic interaction profiles showed that distinct cellular processes were sensitive to changes in either the rates of GTPase hydrolysis or nucleotide exchange, prompting a model for a single GTPase selectively and independently controlling different downstream pathways by regulated tuning of its switching. In Chapter 3, I describe a mutational scanning study which quantitatively measured the fitness effect of all possible point mutations in Gsp1. The scan revealed an unexpected widespread toxic/gain-of- function response, in which mutations were more deleterious than loss of gene function by truncation of Gsp1 via internal STOP codon. Sites enriched for toxic/gain-of-function mutations included a novel allosteric cluster of residues which stabilize the GDP-bound state of Gsp1, confirmed by enzyme kinetics. The study defined a functional map of allosteric regulatory sites in Gsp1 which generalizes to other GTPases and confirmed that perturbation of the switch mechanism is the dominant factor in the effect Gsp1 mutations exert at the cellular level. Finally, in Chapter 4, I discuss the implications of these findings for future studies of molecular switches and their interaction networks, as well as for the use of high-throughput genome-wide measurements to guide the engineering of protein function.