Self-cutting and integrating CRISPR plasmids (SCIPs) enable targeted genomic integration of large genetic payloads for rapid cell engineering

Since observations that CRISPR nucleases function in mammalian cells, many strategies have been devised to adapt them for genetic engineering. Here, we investigated self-cutting and integrating Cas9/CRISPR plasmids (SCIPs) as easy-to-use gene editing tools that insert themselves at CRISPR-guided locations. Leaky sgRNA/Cas9 expression in bacteria initially prevented SCIP assembly and production; this was ameliorated by inserting a mammalian intron into the Cas9 gene. SCIPs demonstrated similar expression kinetics and gene disruption efficiency in mouse (EL4) and human (Jurkat) cells, with stable integration in 3-6% of transfected cells. Clonal sequencing analysis indicated that integrants showed bi- or mono-allelic integration of entire CRISPR plasmids in predictable orientations and with limited indel formation. Interestingly, including longer homology arms (HAs) (500 bp) in varying orientations only modestly increased knock-in efficiency (~2-fold), indicating that SCIP integration favours homology-independent mechanisms. Using a SCIP-payload design (SCIPay) which liberates a promoter-less sequence flanked by HAs thereby requiring perfect homology-directed repair (HDR) for expression, longer HAs resulted in higher integration efficiency and precision of the payload but did not affect integration of the remaining plasmid sequence. As proofs-of-concept, we used SCIPay to 1) insert a gene fragment encoding tdTomato into the CD69 locus of Jurkat cells, thereby creating a NOVEL cell line that reports T cell activation, and 2) insert a chimeric antigen receptor (CAR) gene into the T cell receptor alpha constant (TRAC) locus. Here, we demonstrate that SCIPs function as simple, efficient, and programmable tools useful for generating gene knockout/knock-in cell lines and suggest future utility in knock-in site screening/optimization, unbiased off-target site identification, and multiplexed, iterative, and/or library-scale automated genome engineering.