(65b) Deep Mutational Scanning As a New Tool for Protein Structure/Function Relationships

What does the sequence-function fitness landscape of a protein look like?  It has been appreciated for some time that most single point mutations are neutral to deleterious for overall protein fitness, but the fine details at the individual mutation level are scant in the literature.  In this talk I present a new method called deep mutational scanning, which assesses the binding and thermodynamic folding energies of all possible single point mutations in a protein sequence using the enabling technologies of synthetic DNA libraries, yeast surface display, and deep sequencing.  I show the application of the method to two proteins that were designed to bind a neutralizing epitope of the influenza hemagglutinin (HA) from the 1918 H1N1 pandemic virus.  From the scanning results, new variants of these proteins were constructed to bind specific HA subtypes as well as to rapidly tighten affinity by over an order of magnitude to the 1918 H1 HA.  Such construction and subsequent traversing of fitness landscapes should have wide application in the engineering of useful proteins as well as aid in the understanding of the determinants of protein sequence to function.