Protein engineering via in vitro coevolution

The creation of novel protein functions represents a significant challenge in protein engineering. To address this challenge, I have developed a simple and efficient method---in vitro coevolution---for the creation of novel protein functions in an existing protein scaffold. The in vitro coevolution method involves the design of a hypothetical pathway toward the target protein function followed by stepwise directed evolution of an appropriate protein scaffold along the pathway. As proof of concept, I have applied this strategy to the engineering of two proteins. In the first example, in vitro coevolution was used to engineer variants of the human estrogen receptor a ligand binding domain (hERalphaLBD) with novel corticosterone activity. A yeast two hybrid system based high throughput selection/screening method was developed to assay the transactivation activity of hERalphaLBD. Two steroidal ligands---testosterone and progesterone---that provide a progressive structural bridge between the natural ligand 17beta-estradiol and the target ligand corticosterone, were chosen to assist the directed evolution of hERalphaLBD. A total of approximately 106 randomly point-mutated variants were phenotypically screened in four rounds of directed evolution, resulting in two hERalphaLBD variants that respond to corticosterone. In the second example, in vitro coevolution was used to the engineer variants of homing endonuclease I-SceI with novel DNA cleavage activity. Homing endonucleases have great potential for applications such as targeted gene correction in gene therapy and gene alteration in systems biology and metabolic engineering. However, the limited repertoire of recognition sequences in naturally available homing endonucleases severely hampers their usefulness. I have developed a highly sensitive selection method for the directed evolution of homing endonucleases that couples enzymatic DNA cleavage with the survival of host cells. Using this selection method and in vitro evolution, I attempted to engineer I-SceI variants that can selectively cleave a target sequence found in a mutant transmembrane conductance regulator (CFTR) gene involved in cystic fibrosis. So far, I have successfully completed two rounds of directed evolution toward the first intermediate target sequence, and obtained a variant that has 5000 fold increased activity toward the new target sequence compared to the wild type enzyme.