Fig. 1

The editing efficiency of the programmed scaffold. a The relative CRISPR KO efficiency was estimated using spacer sequence targeting human EMX1 gene. The KO efficiency of variant scaffolds was all normalized to the WT scaffold. “Tail”, tail position in the scaffold. “Tetra,” tetraloop position in the scaffold. “L2,” loop2 position in the scaffold. “30A,” capture sequence with 30 consecutive adenosine. “8A8G,” a 30-nt length capture sequence composes eight leading A and mixed with G every seven continuous A (AAAAAAAAGAAAAAAAGAAAAAAAGAAAAA) (*P ≤ 0.05, compared with WT in unpaired T test). b The CRISPR KO efficiencies of three spacers targeting CXCR4, VEGFA, and DMD were estimated. The “8A8G” capture sequence was inserted to the Tail position on each of these scaffolds. c The capture efficiencies of the gRNA transcripts were estimated using RT-qPCR. In all of the Tail, Tetra, and L2 positions, both 30A variant and 8A8G variant scaffolds were examined. In each position, the capture efficiencies of the 30A and 8A8G variant scaffold were normalized to the WT scaffold. d The CRISPR activation assay was conducted on different targeting genes, and expression increase was estimated using RT-qPCR. e The activation efficiencies of the single gRNA and the multiplexed two gRNAs were compared. The brown bar indicated the fold activation when single gRNA was used, and the light green bar indicated the fold activation when a co-expressed (HBG1 gRNA and IL1B gRNA) cassette was used. The scaffold used here was the same as d