Combinatorial protein engineering identifies potent CRISPR activators with reduced toxicity

Abstract

Abstract Current protein engineering methods are inadequate to explore the combinatorial potential offered by nature’s vast repertoire of protein domains–limiting our ability to create optimal synthetic tools. To overcome this barrier, we develop an approach to create and test thousands of chimeric proteins and employ it to probe an expansive combinatorial landscape of over 15,000 multi-domain CRISPR activators. Our findings indicate that many activators produce substantial cellular toxicity, often unrelated to their capacity to regulate gene expression. We also explore the biochemical features of activation domains and determine how their combinatorial interactions shape activator behavior. Finally, we identify two potent CRISPR activators, MHV and MMH, and demonstrate their enhanced activity across diverse targets and cell types compared to the gold-standard MCP activator, synergistic activation mediator (SAM).