Print

Roy, Liton

Chemistry

2011

PhD

The dominant force in protein folding is the burial of hydrophobic amino acid side-chains to form a hydrophobic core; however it is not clear how the amino acid composition of the core determines stability and native structure. The only way to rigorously examine how core sequence corresponds to stability is to make all possible core variants and measure their folding stabilities. Analysis of such large numbers of proteins is difficult.
Our approach to this problem is to use dynamic combinatorial libraries. A dynamic combinatorial library (DCL) consists of a covalent library of building blocks that can associate reversibly under thermodynamic control to give a much larger ensemble of multimers. We developed a strategy for the selection of optimally stable proteins from recursively enriched dynamic combinatorial libraries (REDCLs). Small libraries of <100 peptides undergo self-assembly in aqueous solution to form DCLs of upwards of 50,000 trimeric association. The stability of each of these trimers is governed by the formation of a well-packed hydrophobic core. We describe the evolution of a system that selects for optimally stable parallel and antiparallel three-helix bundles. The study leads to the following important conclusions.
A) Between 0.1 and 0.2% of all possible core packing arrangements have high folding stabilities. b) These arrangements are stabilized by intimate 'jigsaw' packing, not by maximizing sequestration of hydrophobic surface area. c) Abundance of highly stable entities do not scale with library size. d) Optimally stable associations are heterotrimeric.