Genome-wide target specificities of CRISPR-Cas9 nucleases revealed by multiplex Digenome-seq

Abstract

We present multiplex Digenome-seq to profile genome-wide specificities of up to 11 CRISPR-Cas9 nucleases simultaneously, saving time and reducing cost. Cell-free human genomic DNA was digested using multiple sgRNAs combined with the Cas9 protein and then subjected to whole-genome sequencing. In vitro cleavage patterns, characteristic of on- and off-target sites, were computationally identified across the genome using a new DNA cleavage scoring system. We found that many false-positive, bulge-type off-target sites were cleaved by sgRNAs transcribed from an oligonucleotide duplex but not by those transcribed from a plasmid template. Multiplex Digenome-seq captured many bona fide off-target sites, missed by other genome-wide methods, at which indels were induced at frequencies <0.1%. After analyzing 964 sites cleaved in vitro by these sgRNAs and measuring indel frequencies at hundreds of off-target sites in cells, we propose a guideline for the choice of target sites for minimizing CRISPR-Cas9 off-target effects in the human genome.

Figure 1.

Improving Digenome-seq analysis. (A) Genome-wide Circos plots of in vitro DNA cleavage scores. Human genomic DNA (red) or RGEN-digested genomic DNA (green) was subjected to whole-genome sequencing. The few false-positive sites found in intact genomic DNA are not visible because their cleavage scores are not high. (B) Schematic overview of Digenome-seq using sgRNA transcribed from an oligonucleotide duplex or a plasmid. (C) Sequence logos of cleavage sites obtained using sgRNA transcribed from an oligonucleotide duplex or a plasmid.

Figure 2.

Multiplex Digenome-seq. (A) Schematic overview of multiplex Digenome-seq. (B) A Venn diagram showing the number of in vitro cleavage sites captured by monoplex and multiplex Digenom-seq analyses. (C) In vitro DNA cleavage scores across Chromosome X obtained by monoplex or multiplex Digenome-seq.

Figure 3.

Analysis of multiplex Digenome-captured sites. (A) Venn diagrams showing the number of sites captured by Digenome-seq, GUIDE-seq, and HTGTS. (B) Fractions of sites captured by Digenome-seq according to the total mismatch number (top) and the mismatch number in the seed region (bottom). (C, top) Scatterplot of the number of sites with six or fewer mismatches in the human genome versus the number of Digenome-captured sites. (Bottom) Eleven RGEN target sites were divided into two groups, G1 and G2 (those with fewer than 13,000 and 16,000 sites, respectively, harboring six or fewer mismatches in the human genome). Error bars, SEM. The P-value was calculated by Student's t-test. (D) Scatterplot of the number of GUIDE-seq captured sites versus the number of Digenome-seq captured sites.

Figure 4.

Indel frequencies determined using targeted deep sequencing at off-target sites. (A) Schematic of conventional sgRNAs (GX₁₉ sgRNA) and modified sgRNAs (ggX₂₀ sgRNA). Indel frequencies at NGS-validated on- and off-target sites for the EMX1 (B), HEK293-3 (C), and RNF2 sgRNAs (D). Specificity ratios calculated by dividing the indel frequency at the on-target site with that at an off-target site for the EMX1 (E), HEK293-3 (F), and RNF2 sgRNAs (G).

Figure 5.

Analysis of NGS-validated and NGS-invalidated off-target sites. Plots of relative indel frequencies (log scale) at off-target sites harboring the number of mismatches indicated in the entire 20-nt sequence (A) or the 10-nt seed sequence (B,C). NGS-tested sites (A) were divided into two groups: validated sites (B) and invalidated sites (C). NGS-validated sites and NGS-invalidated sites were those with indel frequencies above and below, respectively, noise indel levels.