UC Berkeley

Multi-domain proteins are a common and essential component of the cellular proteome and contribute complex functions that single domains cannot achieve. The field of protein engineering uses sophisticated techniques to modify proteins for a given purpose, but generally focuses on single domains without consideration of the protein's overall topology. In this work, we set out to develop tools that enable the high-throughput sampling of different domain connections. A novel method of random transposon-mediated domain insertion was devised, which efficiently created large plasmid libraries of domain fusions. The utility of this method is first demonstrated with insertions of circularly permuted GFP into maltose-binding protein, coupled with fluorescence-activated cell sorting (FACS), to isolate biosensor proteins that are allosterically regulated by maltose. We next demonstrate the generality of this method by constructing a pool of functional dCas9 variants with insertions of a PDZ protein-protein interaction domain. Ultimately, these dCas9 variants may prove useful as a scaffold for the recruitment of engineered proteins to specific sites in the genome. These successful protein-engineering efforts illustrate the advanced functions possible with multi-domain constructs and how this transposon-based insertion method can facilitate the creation and study of this new class of synthetic proteins.