A strategy for genome-wide seamless tagging of human protein-coding genes

Comprehensive and systematic proteome-wide experiments require financial and technical resources unavailable to most researchers. Here we describe a universal and scalable method based on CRISPR/Cas9-non-homologous end joining (NHEJ) for Pooled Recombinant Integration of Seamless Markers (PRISM) into protein coding genes. We created two genome-wide gRNA libraries for 5'- and 3'-tagging of 18,489 human protein-coding genes. Selection for in-frame integration of the donor cassette can be favored by fusing it to an antibiotic resistance enzyme (ARE) and P2A self-cleaving peptide, resulting in tagged proteins and free ARE to select for clones with inserted donor cassettes. We achieved a library integration rate of ~20% and tagging of at least 36% of genes expressed in HEK 293T cells including, at least 28% of essential genes, with donor DNA. Our strategy is universal, scalable, specific, and selective, paving the way for genome-scale construction of human cell lines tagged with different types of reporter genes for protein functional characterization.