Genome-wide specificity of plant genome editing by both CRISPR-Cas9 and TALEN

Abstract

CRISPR and TALENs are efficient systems for gene editing in many organisms including plants. In many cases the CRISPR-Cas or TALEN modules are expressed in the plant cell only transiently. Theoretically, transient expression of the editing modules should limit unexpected effects compared to stable transformation. However, very few studies have measured the off-target and unpredicted effects of editing strategies on the plant genome, and none of them have compared these two major editing systems. We conducted, in Physcomitrium patens, a comprehensive genome-wide investigation of off-target mutations using either a CRISPR-Cas9 or a TALEN strategy. We observed a similar number of differences for the two editing strategies compared to control non-transfected plants, with an average of 8.25 SNVs and 19.5 InDels for the CRISPR-edited plants, and an average of 17.5 SNVs and 32 InDels for the TALEN-edited plants. Interestingly, a comparable number of SNVs and InDels could be detected in the PEG-treated control plants. This shows that except for the on-target modifications, the gene editing tools used in this study did not show a significant off-target activity nor unpredicted effects on the genome, and did not lead to transgene integration. The PEG treatment, a well-established biotechnological method, in itself, was the main source of mutations found in the edited plants.

Figure 1

CRISPR–Cas and TALEN targeted editing on the APT gene. (a) Structure of the APT gene and sgRNAs and TALEN positions. Boxes in white represent the exons and black lines represent the introns. Black arrows represent the primers used for PCR and sequencing. (b) Target sequences for sgRNA#1 and sgRNA#2 on the APT gene and sequence of CRISPR–Cas mutants. PAM (protospacer adjacent motif) sequence is in bold, green letters plus arrow indicate the sequences targeted by sgRNA1 and sgRNA2. (c) Target sequences for TALEN402-TALEN405 on the APT gene. Blue and purple letters plus arrows indicate the sequences targeted by TALEN402 and TALEN405 respectively, in red target sequence for the FOK1 nuclease dimer. DNA insertions are shown in orange and deletions with dashes. In the blue frames, the microhomologies potentially involved in Alt-EJ-mediated repair of the induced DSBs.

Figure 2

Workflow of whole-genome SNV and InDel analysis. SNV and small InDels (< 50 bp) detection was done using HaplotypeCaller algorithm from GATK tools. Large InDels detection was done using Pindel tool.

Figure 3

Distribution of spontaneous and transfection mediated mutations. Circos diagram illustrating the distribution along the genome of the different types of mutations detected in the treated plants (SupData 2). The genes affected by these mutations are also illustrated in this Circos. The size of the labels indicates the ratio of variant alleles (AR) in each sample. A large size corresponds to an AR close to 1, and a small size corresponds to an AR at the subclonal level in case of chimeric spontaneous mutations.

Figure 4

Homology based detection of potential off-targets. BLAT tool was used to predict potential off-target sequences. Results were crossed with detected mutations from WGS variant calling.