Fig. 1: Duplex Sequencing based analysis of SpCas9-induced off-target and translocation effects in the human PCSK9-KI mouse model.

A Experimental setup for SpCas9 off-target analysis in vivo using targeted amplicon sequencing and Duplex Sequencing. The humanized PCSK9-KI mouse model was injected with adenovirus expressing SpCas9 with gMH sgRNA or SpCas9 plus GFP as a control. After 1 week, the liver tissue was sampled to measure DNA editing activity. Initial off-target analysis was performed by multiplexed amplicon sequencing with a detection limit of 0.1%. Re-assessment of off-target editing at higher sensitivity (0.01% detection limit) was done by Duplex-Seq. B Schematic presenting the gMH sgRNA sequence targeting the mouse Pcsk9 locus with perfect base-pair complementarity and the human PCSK9 locus with a single mismatch at position 11 (highlighted in red). C Comparison of amplicon sequencing (Amplicon-Seq) vs. Duplex-Seq-based off-target analysis. Gene editing analysis at off-target sites²⁸ by amplicon sequencing or Duplex-Seq followed by CRISPResso2 analysis. Heatmaps present the editing efficiency for the two target sites (highlighted in green), as well as the five detected off-target sites (highlighted in magenta). Amplicon sequencing and Duplex-Seq were performed with n = 3 biologically independent mice (1, 2, 3) using genomic DNA isolated from the liver of mice treated with adenoviral vector expressing sgRNA gMH together with SpCas9. DNA from the same mice (numbered 1–3) were used for both sequencing approaches. Values show editing efficiency as percent indel-containing reads normalized to mice treated with adenoviral vector encoding SpCas9 without an sgRNA as a control. Extended tables showing non-normalized data of all assayed off-target sites, as well as the two target sites for both amplicon sequencing and Duplex-Seq can be found in Supplementary Figs. S2 and S3, respectively. The detection limits reached were 0.1% and 0.01% for amplicon sequencing and Duplex-Seq, respectively. D Schematic representation of the possible outcomes of the gene editing at the Pcsk9 locus in humanized PCSK9-KI mouse model. E, F SpCas9-induced translocations between mouse and human PCSK9 genes after in vivo editing. Frequency of the specific translocation event where the 3’ end of mouse exon 3 becomes fused to the 5’ end of human exon 3 assayed by ddPCR. In (E), the probe was placed in the human exon, whereas in (F) the probe was placed in the mouse exon. Graph shows individual values, as well as mean ± SD from n = 3 biologically independent mice per group. A two-sided Student’s t test was performed to evaluate statistical significance (**P value = 0.002 (E) and 0.0026 (F)).