Evolving split inteins using bacterial autodisplay

Becher K-S, Nienberg C, Jose J, Mootz HD

Abstract

Site-specific, regio- and chemo-selective modifications of proteins are becoming increasingly important for applications in therapy, diagnostics and cellular imaging. The split intein technology based on protein trans-splicing has several advantages compared to other classical modification strategies. For example, the modification is rapid, specific, orthogonal to cellular chemistry, and can be performed at low protein concentrations.[1] Furthermore, the target protein is not exposed to any harsh reaction conditions. Split inteins have also gained much attention due to their potential for numerous biotechnological applications, e.g. segmental isotopic labeling and production of cyclic polypeptides.[2,3] Nevertheless, the practical advantages are associated with some problems. Many inteins show a certain dependency of their native exteins sequences, do not splice to completion, some require reducing conditions due to catalytic cysteines, and they may have solubility problems. To overcome these limitations, we aim to use directed evolution to develop a new split intein with superior properties. The work presented here explores the autodisplay technology, previously described.[4] The autotransporter can translocate one half of the split intein to the surface of the host cell, here E. coli. Thus, a direct linkage between genotype and phenotype is achieved. After incubating the presenting cell with a fluorescent labeled splicing partner containing the complementary intein half, the splicing reaction takes place and the fluorescent moiety is transferred to the E. coli cell. By usage of fluorescence activated cell sorting (FACS), a large library of randomized intein mutants can be screened fast and easily.