In vivo protein function optimization

Abstract

We developed a synthetic cellular system for in vivo protein function optimization through directed evolution. Unlike conventional protein directed evolution techniques whose mutagenesis and selection are carried out in vitro, in our system a target gene is mutated and selected in vivo. Because the mutagenesis and selection are carried out by cells, the process to optimize a protein is just to culture cells. Such a simple process enables protein evolution to be automated and as a result such automated system could be used for massive protein evolution. For selection we used the fact that cells expressing different level of β-lactamase show different growth rate and susceptibility to antibiotics, such as aztreonam. In the developed system, the β-lactamase is designed to be expressed proportionally to the activity of a target gene. After selection, cells are treated with the chemical mutagen, ethylmethane sulphonate, for mutagenesis. During mutagenesis, not only target gene but also host chromosome would be mutated, which may alter cell susceptibility and growth rate. In order to eliminate false selection, a target gene-containing plasmid is rescued as a phage from the host and then the phage is infected into fresh host cell. These selection and mutagenesis procedure are repeated until the target gene is optimized. To evaluate our system and to find optimal experiment conditions, we developed a mathematical model of protein directed evolution, and showed that the system works efficiently. We constructed the designed synthetic in vivo evolution system and evolved a transcription factor, luxR, as a model system. The system optimized the gene efficiently. The system could be used for other types of protein optimization if a protein’s activity is converted to transcriptional efficiency.