Parallelized engineering of mutational models using piggyBac transposon delivery of CRISPR libraries

Abstract

Novel gene and variant discoveries have reached unprecedented scale with the emergence of exome and genome sequencing studies across a spectrum of human disease initiatives. Highly parallelized functional characterization of these variants is now paramount to deciphering disease mechanisms, and approaches that facilitate editing of induced pluripotent stem cells (iPSCs) to derive otherwise inaccessible tissues of interest (e.g., brain) have become critical in genomics research. Here, we sought to facilitate scalable editing of multiple genes and variants by developing a genome engineering approach that incorporates libraries of CRISPR/Cas9 guide RNAs (gRNAs) into a piggyBac (PB) transposon system. To test the efficiency of inducing small indels, targeted deletions, and large reciprocal copy number variants (CNVs), we simultaneously delivered to human iPSCs both Cas9 and a library including 59 single gRNAs targeting segmental duplications, 70 paired gRNAs flanking particular genic regions, and three single gRNAs targeting the coding sequence of an individual gene, MAGEL2. After editing, we isolated single cells, expanded resultant colonies, and genotyped their gRNA contents and mutational outcomes. We observed that 97.7% of gRNA constructs were integrated into at least one colony, with 85.6% of colonies containing three or fewer PB integrations. This PB editing method generated 354 cell lines with 57.8% of sequenced gRNA cleavage sites modified in at least one line, 14.4% of these lines altered at multiple targets, and single-copy indel mutagenesis predominating. Among the edits generated were eight targeted genomic deletions, including pathogenic microdeletions at chromosome 15q11-q13 ([~]5.3 Mbp), chromosome 16p11.2 ([~]740 kbp), and chromosome 17q11.2 ([~]1.4 Mbp). These data highlight PB editing as a powerful platform for gene inactivation and testify to its strong potential for oligogenic modeling. The ability to rapidly establish high-quality mutational models at scale will facilitate the development of near-isogenic cellular collections and catalyze comparative functional genomic studies, better positioning us to investigate the roles of hundreds of genes and mutations in development and disease.