Increasing protein stability by inferring substitution effects from high-throughput experiments

Protein stability is an important parameter in almost all protein-engineering efforts. Evaluating the effects of the many possible amino acid changes to guide such projects is a significant task, even with recent advances in experimental and computational approaches. Here, we apply a computational model, GMMA, to extract substitution effects from a cost-effective genetic screen of a randomly mutated protein library. Using a high mutation frequency, the method can map stability effects of even very stable proteins for which conventional selection systems have reached their limit. Thus, we screened a mutant library of a highly stable and optimised model protein using an in vivo genetic sensor for folding and assigned a stability effect to 374 of 912 possible single amino acid substitutions. Combining the top 9 substitutions increased the thermodynamic stability by almost 50% in a single engineering step. This illustrates the capability of the method, which is applicable to any screen for protein function.