Joint single-cell profiling of Cas9 edits and transcriptomes reveals on- and off-target effects on gene expression (rhAmpSeq)

A longstanding barrier in genome engineering with CRISPR-Cas9 has been the inability to characterize editing outcomes at single-cell resolution. We present "Superb-seq", a new scRNA-seq method that measures on-target and off-target genome edits and associated transcriptomes. In contrast to Perturb-seq that captures single-cell guide information, Superb-seq directly captures single-cell edit profiles by T7 in situ transcription. We demonstrate and validate this method through two Superb-seq experiments in 30,000 cells of three cell types, amplicon sequencing (rhAmpSeq), and bulk RNA-seq. Superb-seq uses off-the-shelf kits, standard laboratory equipment, and requires no virus, which will improve CRISPR screens in diverse tissues and functional characterization of CRISPR therapeutics. Overall design: RhAmpseq was performed on two sets of eight genomic DNA samples. The first sample of each set was a control sample of unedited (electroporation only) K562 cells. The subsequent seven test samples were K562 cells edited with CRISPR-Cas9 and T7 promoter, using seven guides targeting chromatin remodeler genes ARID1A, SMARCA4, CHD3, and CHD4, one guide per sample. These eight samples were processed into rhAmpSeq libraries using one of two primer sets: either a multiplex panel of 52 primer pairs targeting Superb-seq detected and in silico predicted edit sites, or a singleplex primer pair targeting ADSS1. Sixteen total libraries, one replicate per sample and primer set.