Activity and specificity of TRV-mediated gene editing in plants

Abstract

Plant trait engineering requires efficient targeted genome-editing technologies. Clustered regularly interspaced palindromic repeats (CRISPRs)/ CRISPR associated (Cas) type II system is used for targeted genome-editing applications across eukaryotic species including plants. Delivery of genome engineering reagents and recovery of mutants remain challenging tasks for in planta applications. Recently, we reported the development of Tobacco rattle virus (TRV)-mediated genome editing in Nicotiana benthamiana. TRV infects the growing points and possesses small genome size; which facilitate cloning, multiplexing, and agroinfections. Here, we report on the persistent activity and specificity of the TRV-mediated CRISPR/Cas9 system for targeted modification of the Nicotiana benthamiana genome. Our data reveal the persistence of the TRV- mediated Cas9 activity for up to 30 d post-agroinefection. Further, our data indicate that TRV-mediated genome editing exhibited no off-target activities at potential off-targets indicating the precision of the system for plant genome engineering. Taken together, our data establish the feasibility and exciting possibilities of using virus-mediated CRISPR/Cas9 for targeted engineering of plant genomes.

Efficient technologies for targeted engineering of plant genomes are highly needed to discover and develop novel traits in key plant species important for food security. Site-Specific nucleases (SSNs) have been used to generate targeted double strand breaks (DSBs) and harnessing the DNA repair machinery of the imprecise non-homologous end-joining (NHEJ) and precise homology-directed repair (HDR).^1-4 Zinc finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs) have been used for targeted editing of plant genomes.^1,5-7 Customization of ZFNs and TALENs require protein engineering for each user-selected targeted, a resource intensive and time-consuming process.⁸ Different repeat assembly protocols have been developed for TALENs engineering but the requirement of using 2 TALENs monomer to bind to the sense and antisense strands simultaneously complicates its application in plants.³ Recently, bacterial and archaeal natural immunity system that targets and destroys invading nucleic acids has been adapted for genome engineering across eukaryotic species. The clustered regularly interspaced palindromic repeat (CRISPR)/ CRISPR associated (Cas) 9 system has been used in diverse plant species such as rice, Nicotiana benthamiana, and Arabidopsis for targeted genome editing.^9-11 The CRISPR/Cas9 system is comprised of the Cas9 endonuclease of Streptococcus pyogenes and a synthetic guide RNA (gRNA), which combines functions of CRISPR RNA (cRNA) and trans-activating cRNA (tracrRNA) to direct the Cas9 protein to the DNA target sequence preceding the protospacer-associated motif (PAM) (NGG).^12,13 Because the specificity of the system is determined by the 20-nucleotide sequence of the gRNA, it allows for unprecedented and facile genome engineering. Further, the CRISPR/Cas9 system is used to simultaneously edit multiple genomic targets.¹²

Delivery of genome engineering reagents into plant cells is a major barrier for efficient and effective use of these technologies for targeted improvement of crop traits. Tobacco rattle virus (TRV) is used as an efficient vector for virus-induced gene silencing (VIGS) for functional genomics studies in diverse plant species.^14,15 TRV has a bipartite RNA1 and RNA2 genomes, and the RNA2 genome can be modified to carry exonic gene fragments for post-transcriptional VIGS.¹⁶ The cargo capacity of TRV is limited to 2–3 kb, and cannot be used to deliver Cas9 endonuclease into plants but it can be used to deliver one or more gRNAs. Recently, we developed TRV as vector to systemically deliver gRNA and demonstrated high efficiency of targeted modification in Nicotiana benthamiana.¹⁷ Here, we report on the activity and specificity of the TRV-mediated CRISPR/Cas9 system for targeted genome editing. To determine whether the TRV was capable of producing and delivering the gRNA molecules systemically and the persistence of genomic modification in growing tissues, the TRV virus was delivered by agroinfiltration and reconstituted in the leaves of N.benthamiana expressing Cas9. This was performed using mixed Agrobacterium cultures harboring the RNA1 genome (pYL192) in combination with an RNA2 vector, in which a gRNA with binding specificity for the PDS gene was driven by the PEBV promoter (pRNA2.PEBV::PDS.gRNA) (Fig. 1A). A gRNA empty vector clone was used as the negative control. Samples were collected at 7, 15 and 30 d post-infiltration (dpi), and targeted editing of the PDS target sequence was assessed in 3 independent plants by PCR amplifying a 797 bp fragment flanking the target site. Then, PCR products were directly subjected to the T7EI assay. Our results showed high levels of genome editing at 7, 15 and 30 d (Fig. 1B). The modification efficiency was analyzed using ImageJ software as described previously (http://imageJ.nih.gov/ij/).¹⁷ The high efficiency of TRV-mediated CRISPR/Cas9 editing indicated a persistent activity of the system, which is important for the modification of germline cells and recovery of mutant seed progeny.

Figure 1.

Persistence of TRV-mediated CRISPR/Cas9 targeted mutagenesis of the PDS3 gene. (A) Establishment of TRV infection in B14 plants, Nicotiana benthamiana Cas9 overexpression line. Agrobacterium cultures containing RNA1 and engineered RNA2 harboring gRNA for targeting PDS gene were mixed 1: 1 (OD₆₀₀ 0.1 each) and co-infiltrated to 2 fully expanded true leaves. Systemic leaves were collected 7, 15 and 30 d post-infiltration (dpi). (B) T7EI assay for indels detection. Genomic DNA was extracted from the systemic leaves and purified PCR product (200ng) of PDS fragment flanking the targeted locus was subjected to T7EI analysis. All three leaves collected at different time points showed high efficiency of targeted mutagenesis (43– 61 %) compared to vector control systemic leaves collected at 30 dpi. (C) Analysis of progeny plants for the presence of targeted modification using NcoI recognition site loss assay. DNA was extracted from progeny plants (50 seedlings pooled in one tube) and PCR was performed with a primer set to amplify 404 bp fragment encompassing the target site. Purified PCR product (300 ng) was treated with NcoI and separated on 2% agarose gel. Progeny pool 1 and 2 clearly showed a resistant DNA fragment of 404 bp indicating the targeted mutagenesis. (D) Alignment of Sanger sequencing reads showing the presence of indels at the PDS target sequence. Numbers to the right of sequence alignment indicates the number of nucleotides deleted by targeting the PDS genomic target.

Since TRV is capable of infecting whole plant parts including meristematic tissues, we tested whether the seed progeny of plants infected with pRNA2.PEBV::PDS.gRNAs would carry genomic modifications in the PDS gene target.¹⁸ Three early-matured seed capsules next to the infiltrated leaves were collected as one pool. Small leaf discs from 50 progeny plants of each pool were collected in one tube as one pool for a total of 4 pools including a WT control. The corresponding 404 bp DNA PCR fragments were subjected to NcoI restriction digestion to assess for targeted sequence modification in these pools. An NcoI enzyme-resistant band (404 bp) appeared only in pools derived from plants infected with the pRNA2:pEBV::gRNA.PDS construct compared to the WT plants, indicating the presence of genomic modification in seed progeny (Fig. 1C). To confirm these data, the PCR product was cloned into a topo cloning vector and the resultant clones were subjected to Sanger sequencing, which confirmed the presence of modification at the intended target site (Fig. 1D). These data indicate the feasibility of recovering plants carrying the targeted modification, albeit at very low efficiency, and bypassing the need for transformation and tissue culture (Fig. 1D). The detection of germinal transmission only in early flowers indicates that TRV infection and persistence in meristematic cells need to be optimized to improve the recovery of mutated plants from the seed progeny. We did not recover plants carrying the targeted mutagenesis from capsules produced later in development. Further improvements might increase the frequency of germinal transmission and recovery of mutant plants from the seed progeny.^17,18

A primary concern in applications of CRISPR/Cas9 genome editing is off-target activities.^19,20 Although this issue is less important in plant than in human applications, we attempted to determine whether our system exhibited off-target activities. To identify candidate unintended targets of genome editing, the draft genome of N. benthamiana was screened for imperfect matches (i.e., allowing several mismatches) to the 20-nucleotide gRNA sequence (Fig. 2A), and 13 candidates were then subjected to T7EI and restriction-protection assays. Consistent with previous reports, no genomic modifications were detected at any of the predicted unintended targets (Fig. 2B).^4,10,21 Therefore, we concluded that either our system exhibits no off-target activities, or that any such activities occurred at levels too low to be detected by the modification-detection assays used.

Figure 2.

TRV-based CRISPR/Cas9 system exhibited no apparent off-target effects in N. benthamiana genome. (A) Table showing combinatorics approach to identify potential off-target binding of PDS3 gRNA in N. benthamiana genome. The putative off-target binding sites were subjected to further annotation, where sequences were split into 2 groups of conserved and not conserved NcoI restriction site directly preceding the PAM sequence and sites containing mutation in seed and non-seed sequence. Based on dissimilarities of 1 to 7 nucleotides with PDS3 gRNA, a total of 4265 potential off targets were detected, out of which 375 have NcoI site proximal to PAM sequence. (B) T7EI assay for the presence of indels at potential off target sites. DNA fragment flanking potential 13 targets were amplified by PCR with their respective primers. TRV-mediated CRISPR/Cas9 system exhibited no detectable activities at all tested targets using the T7EI assays. The respective contig number of each of 13 potential off target site are represented on the top of gel

In conclusion, our work demonstrates the persistence of TRV-mediated CRISPR/Cas9 editing and the possibility of optimizing this method to recover progeny plants carrying the targeted modifications thereby bypassing the need for tissue culture or repeated transformation. This method will expand the utility of the CRISPR/Cas9 system for plant functional genomics and targeted improvements of crop traits. Further, the use of heterozygous Cas9 overexpressing plants with this facile and versatile genome-editing platform allows the engineering and production of plants free of foreign DNA. This might overcome the regulatory hurdles that impede the commercialization of engineered plants.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We would like to thank members of the laboratory for genome engineering at KAUST for helpful discussions and comments.

Funding

This study is supported by King Abdullah University of Science and Technology (KAUST).