Knockout of VvCCD8 gene in grapevine affects shoot branching

Background Shoot branching is an important trait of plants that allows them to adapt to environment changes. Strigolactones (SLs) are newly identified plant hormones that inhibit shoot branching in plants. The SL biosynthesis genes CCD7 (carotenoid cleavage dioxygenase 7) and CCD8 have been found to regulate branching in several herbaceous plants by taking advantage of their loss-of-function mutants. However, the role for CCD7 and CCD8 in shoot branching control in grapevine is still unknown due to the lack of corresponding mutants. Results Here we employed the CRISPR/Cas9 system to edit the VvCCD7 and VvCCD8 genes in the grape hybrid 41B. The 41B embryogenic cells can easily be transformed and used for regeneration of the corresponding transformed plants. Sequencing analysis revealed that gene editing has been used successfully to target both VvCCD genes in 41B embryogenic cells. After regeneration, six 41B plantlets were identified as transgenic plants carrying the CCD8-sgRNA expression cassette. Among these, four plants showed mutation in the target region and were selected as ccd8 mutants. These ccd8 mutants showed increased shoot branching compared to the corresponding wild-type plants. In addition, no off-target mutation was detected in the tested mutants at predicted off-target sites. Conclusions Our results underline the key role of VvCCD8 in the control of grapevine shoot branching.


Background
The control of shoot branching is an adaptive strategy that allows plants to optimize their growth to adapt to environmental changes. Shoot branching is determined by the number and outgrowth of axillary buds, and bud outgrowth contributes to the flexibility in branching [1]. Auxin and cytokinin are master regulators that control shoot branching in plants. Auxin was considered as inhibitor in bud outgrowth [2,3], whereas cytokinin was found to promote this process [2]. However, the established hormone signaling pathways cannot fully explain the control of bud outgrowth [4], suggesting the existence of other regulators.
Recently, SLs were proposed to control scion development in response to nitrogen availability in grafted grapevine plants [25]. Additionally, overexpression of grape CCD7 or CCD8 gene in Arabidopsis max3 or max4 mutants background partly reverted their phenotypes [25], suggesting a potential role for CCD7 and CCD8 in grapevine shoot branching. However, up to date, in grapevine no experimental evidence supporting the role of these two genes in the control of shoot branching exists. This role has therefore still to be demonstrated in grapevine. CRISPR/Cas9 (clustered regulatory interspaced short palindromic repeats/CRISPR-associated protein 9) system is a powerful tool for targeted mutagenesis that has been successfully applied in many plant species to achieve genome editing. In grape, this system was efficiently used to edit IdnDH (L-idonate dehydrogenase), PDS (phytoene desaturase), and VvWRKY52 genes [26][27][28]. This indicates that the CRISPR/Cas9 system can be used for precise genome editing in grapevine.
In this study, we used the CRISPR/Cas9 technology to edit the VvCCD7 and VvCCD8 genes in 41B grapevine rootstock, respectively. As 41B embryogenic cell transformation, selection and regeneration are easy to perform, these cells were chosen to perform gene editing experiments. After regeneration, four VvCCD8 knockout lines were obtained. The recovered ccd8 mutants exhibited increased shoot branching when compared to wildtype plants. Sanger sequencing results showed that VvCCD8 mutant plants carried the targeted mutations, and that no mutation occurred at the putative off-target sites. Altogether, these results underline the efficiency of grape genome editing and provide evidence that VvCCD8 plays a key role in the control of shoot branching in grapevine.

Target design and CRISPR/Cas9 vector construction
The VvCCD7 (VIT_15s0021g02190) and VvCCD8 (VIT_ 04s0008g03380) genes contain 6 and 5 exons, respectively. Considering that targeted mutagenesis caused by CRISPR/ Cas9 generally resulted in frameshifts or generation of stop codons [26,27], the upstream exons would be better targets for gene editing to produce non-functional proteins. Thus, the first exon (Exon1) of VvCCD7 and the second exon (Exon2) of VvCCD8 were selected as the targets for CRISPR-Cas9 gene editing, respectively (Fig. 1a). The target regions of these two genes were cloned and verified by Sanger sequencing prior to sgRNA design. The results showed that the amplified sequences of VvCCD7 and VvCCD8 are almost identical to their reference sequences (Additional file 1: Figure S1). The sgRNAs used for targeting VvCCD7 (CCD7-sgRNA) and VvCCD8 (CCD8-sgRNA) were designed accordingly (Fig. 1a). Both sgRNAs were driven by the Arabidopsis U6 promoter (AtU6), while the expression of Streptococcus pyogenes Cas9 was under the control of CaMV35S promoter (35S). The EGFP (enhanced green fluorescent protein) gene was used as a reporter gene to rapidly select efficiently transformed cells (Fig. 1b).

Targeted mutagenesis of VvCCD7 and VvCCD8 genes in 41B cells
The constructed CRISPR/Cas9 expression vectors were introduced into 41B grape cells by Agrobacterium-mediated transformation. Successfully transformed cells were selected by EGFP fluorescence, whereas no fluorescence signal could be detected in untransformed cells (Fig. 2a). The 41B cells exhibiting EGFP signal were sampled and subjected to Sanger sequencing in order to reveal the presence of mutations at the target sites. The sequencing chromatograms were manually analyzed for the presence of double tracing peaks at the target regions, considering that the presence of overlapping peaks was a typical indicator of targeted mutations [29]. Our sequencing results ( Fig. 2b) revealed the presence of overlapping peaks in the positively transformed 41B cells by contrast to the single peaks of wild-type (WT) cells chromatograms. These results clearly indicate the existence of targeted mutagenesis in VvCCD7 and VvCCD8 genes in transformed 41B cells.
In these mutant plants, VvCCD8 target sequences were analyzed by Sanger sequencing. Twenty clones of PCR amplicons were sequenced for each mutant plant. The results showed that Plant #1 and Plant #3 contained two types of mutations at the target site. The first one corresponds to an insertion of one nucleotide and the second one to a deletion of several nucleotides (20 bp for Plant #1 and 11 bp for Plant #3) (Fig. 3e). These results suggest that these two mutant plants might be biallelic. According to the sequencing results, Plant #5 and Plant #6 only contained one type of mutation (Fig. 3e). Plant #6 mutant might be homozygous as almost all its sequenced clones (19/20) contained the same mutation (1-bp insertion). By contrast, Plant #5 mutant might be heterozygote or chimeric as both wild-type and mutated (1-bp deletion) sequences were identified within the sequenced clones (Fig. 3e). These different mutations led to frameshifts changes, resulting in new mutated amino acid sequences (Fig. 3f) or in the production of stop codons (Fig. 3f) that would result in premature termination of translation.
We also investigated the expression profiles of VvCCD8 in these mutant plants. The results showed that the transcript abundance of VvCCD8 in the four  mutants was significantly decreased in comparison with wild-type plants (Fig. 4), suggesting that the targeted mutagenesis observed in VvCCD8 resulted in transcript decay in these ccd8 mutants.

Off-target effect was not detected in ccd8 mutants
To make sure that the enhanced shoot branching phenotype observed in ccd8 mutants was not due to offtarget effects, we investigated the putative genomic offtarget loci of CCD8-sgRNA in Plant #1 and Plant #6. Putative off-target sites were predicted according to their sequence homology with CCD8-sgRNA. Among the 5 highest ranked potential off-target sites, 2 sites were predicted to localize in the exons of VIT_03s0091g00830 and VIT_13s0019g01150 sequences (Additional file 5: Table S2). These two putative off-target sites were therefore selected for further analysis. DNA fragments containing the 2 putative off-target sites were amplified from Plant #1 and Plant #6 by PCR using specific primers (Additional file 4: Table S1). PCR products were cloned into pLB vector and verified by Sanger sequencing. No mutation was detected at the potential off-target sites (Additional file 2: Figure S2), supporting the fact that VvCCD8 editing was efficient in grapevine and that the increased branching phenotype observed in ccd8 mutants was specifically due to the mutation in VvCCD8 and not to off-target effects.

Discussion
The CRISPR/Cas9 system has emerged as a powerful tool for genome editing, and it shows great potential in generating mutants in plants. In Arabidopsis, CRISPR/ Cas9 system was successfully used to produce cbfs mutants. The characterization of these mutants revealed the important role played by CBF2 in cold acclimationdependent freezing [30]. Intriguingly, targeted mutagenesis of SBP-CNR and NAC-NOR transcription factors, which are thought to be master regulators of tomato ripening, resulted in tomato delayed ripening or partial non-ripening. This phenotype was surprisingly distinct from the previously used original tomato mutants [31], suggesting a great potential of CRISPR/Cas9 system in gene functional research. SLs were found to retard bud outgrowth [5,6], and SL biosynthetic genes CCD7 (MAX3) and CCD8 (MAX4) had been demonstrated to be involved in branching control in multiple herbaceous plants, including Arabidopsis, rice and tomato [8,[32][33][34]. In woody plant, poplar MAX4 (CCD8) knockdown lines exhibited altered branching patterns [24]. In grape, it has been suggested that SL could be involved in the control of scion architecture in grafted grapevine plants, based on the observation that exudate from grape CCDs-overexpressing transgenic cells could stimulate the germination of Phelipanche ramosa seeds, and that overexpression of grape CCD7 or CCD8 gene in corresponding Arabidopsis mutant can partly revert the mutant phenotype [25]. However, no direct and clear evidence supporting this role exists in grapevine.
In the present study, we employed the CRISPR/Cas9 system to edit VvCCD7 and VvCCD8 genes in grapevine (41B rootstock). After transformation of 41B embryogenic cells, Sanger sequencing assay was performed to detect the targeted mutations. The results showed that the designed sgRNAs could effectively direct the targeted (See figure on previous page.) Fig. 3 Identification of VvCCD8 knockout mutants. a Overview of identification of regenerated plants. b Identification of exogenous T-DNA insertions in regenerated plants by PCR. The specific primers designed for Cas9 gene were used for PCR identification. Only CCD8-sgRNA plants were identified with exogenous T-DNA insertions. Lanes 1-6 represent different individual CCD8-sgRNA plants. The plasmid was used as the positive control (P), while the wildtype genomic DNA was used as the negative control (N). M, DNA marker. The cropped gel image is shown here, and the original, uncropped image is available in Additional file 3: Figure S3. c Phenotypes of VvCCD8 knockout mutants. The shoot branches of VvCCD8 knockout mutants were indicated in black arrows. Scale bars: 0.5 cm. d The branch number of the four VvCCD8 knockout mutants. e Sequencing results of the target sites in the four VvCCD8 knockout mutants. The gene fragments were amplified from each mutant plant and were cloned into pLB vector for Sanger sequencing assay. A number of 20 clonal amplicons for each plant were analyzed. The mutated sequences identified from the mutants were shown. The plant IDs are shown on the left. Mutation types (colored in red) and the corresponding number (indicated in black) of clones were shown on the right. Those undesired sequences were omitted from the analysis. f Mutations of amino acids in mutated sequences shown in e. The altered amino acids are colored in red and the premature stop codons are indicated in red asterisks (*). The number of amino acids (aa) that are not shown in the figure is indicated in parentheses Fig. 4 Expression profiles of VvCCD8 in ccd8 mutant plants. The expression of VvCCD8 was determined by quantitative real-time PCR, and the Actin 1 (accession no. AY680701) was used as internal control. The experiment was repeated three times and the data are shown as mean ± SD. The significance of differential expression level was examined by Student's t-test with P < 0.01 indicating highly significant (**) editing in both genes (Fig. 2b). The whole plants were obtained through regeneration, and transgenic plants were identified and selected by PCR. Only 6 CCD8-sgRNA plants were found to contain exogenous Cas9 gene (Fig. 3b). The lack of antibiotics-dependent selection and low regeneration rate of 41B cells in this experiment are probably responsible for low rate of transgenic plants obtained. Among the 6 transgenic plants, 4 were identified as ccd8 mutants (Fig. 3a). As expected, the ccd8 mutants exhibited increased shoot branching, which is in agreement with previous reports [13,15,16,24]. Except for Plant #5, all mutants had at least 3 shoots, whereas WT plants generally had only one shoot (Fig. 3d). According to the sequencing results, Plant #5 might be heterozygous or chimeric (Fig. 3e), suggesting a possible relationship between the number of shoots and SLs concentration in grapevine. Off-target effect is a major concern when applying CRISPR/Cas9 technology. We therefore investigated the putative genomic off-target loci of CCD8-sgRNA, and no off-target mutation was observed (Additional file 2: Figure S2). These results ruled out the possibility that the altered shoot branching observed in ccd8 mutants was caused by the presence of off-target mutations.
Interestingly, the ccd8 mutants obtain with 41B rootstock could serve for grafting experiments in order to further study the role of SLs in the control of grapevine shoot branching. Finally, whether VvCCD7 plays a same role in shoot branching is remains unclear and has yet to be investigated. To go further in this direction, new attempt to regenerate ccd7 mutant plants could be achieved in the future.

Conclusions
Collectively, our results showed that CRISPR/Cas9 system can be successfully used to knock out VvCCD7 and VvCCD8 genes in grape. Additionally, the study of VvCCD8 knockout grapevine plants revealed the key role of this gene in the control of shoot branching, therefore providing a first clue to investigate the mechanisms involved in the regulation of shoot architecture in grapevine.

Design of sgRNA and construction of genome editing vectors
The target regions of VvCCD7 and VvCCD8 genes were amplified from 41B embryogenic cells by PCR with primers CCD7-F/R and CCD8-F/R, respectively. The amplified fragments were verified by Sanger sequencing. The verified sequences were used as an input for sgRNA design with the online tool CRISPR-P v2.0 [35]. The potential off-target sites were predicted simultaneously with this tool. The designed sgRNAs were then ligated into the pCACRISPR/Cas9 vector via homologous recombination (HR). PCR cloning, sgRNA design and plasmid construction were conducted as previously described [26]. The pCACRISPR/Cas9 vector was digested with SmaI and XhoI to remove the hptII (hygromycin phosphotransferase II) gene, and the EGFP gene (NCBI accession: NC_025025) amplified from pCAMBIA2300-EGFP vector was inserted into the linearized pCACRISPR/Cas9 vector via HR using the Clo-nExpress II One Step Cloning Kit (Vazyme, China). The primers used in the experiment are available in Additional file 4: Table S1.

Plant material, transformation and regeneration
The embryogenic grape cells derived from 41B rootstock (Vitis vinifera cv. Chasselas × Vitis berlandieri) were graciously provided by Dr. F. Lecourieux (EGFV, Université de Bordeaux), and the cells were cultured as previously described [36]. In brief, the suspension cells were subcultured weekly in 25 mL of liquid glycerol-maltose (GM) medium containing 1 mg L − 1 naphthoxy acetic acid (NOA) in the dark.
The constructed binary vectors were introduced into Agrobacterium tumefaciens strain EHA105 by the freezethaw method, and the 41B embryogenic cells were transformed using the A. tumefaciens co-cultivation method [37]. After co-cultivation, the grape cells were first washed twice with liquid GM medium and then subcultured every other day in GM medium supplemented with 200 mg/L timentin for 1 week. Then the cells were collected and divided into small groups (~0.5 cm 2 ) for EGFP detection.
For induction of embryogenesis, 41B cells were transferred onto solid hormone-free regeneration medium (GM medium without NOA) under a 16-h photoperiod with white fluorescent lights. Plants regenerated on McCown woody plant medium (Duchefa) supplemented with 3% sucrose, 0.2 mg/L naphthalene acetic acid (NAA), 0.5 mg/L activated charcoal, 7.5 g/L agar under long-day (16 h light/8 h dark) conditions.

Extraction of genomic DNA and PCR identification of exogenous T-DNA insertion
Genomic DNA was prepared using the CTAB plant genomic DNA extraction kit (Aidlab, China) according to the manufacturer's instructions. The isolated DNA was used as the template for PCR. The PCR reaction was performed with Cas9-specific primers (Additional file 4: Table S1) using Es Taq DNA polymerase (CWBIO, China) according to the manufacturer's protocol. The PCR products were detected by 1% agarose gel electrophoresis and were further confirmed by Sanger sequencing.

Sanger sequencing assay
The DNA fragments containing the target sites were amplified from 41B cells or regenerated plants by PCR with the primers CCD7-F/R and CCD8-F/R, respectively. The PCR products amplified from grape cells were purified and directly used for Sanger sequencing analysis (Tsingke, Beijing). The amplified fragments from 41B plants were cloned into pLB-Simple vector (TIANGEN, China), and a total of 20 clones for each sample were sequenced.

Quantitative real-time PCR assay
The expression profiles of VvCCD8 gene were investigated using quantitative real-time PCR (qRT-PCR) with VvCCD8 specific primers (Additional file 4: Table S1). The grape Actin 1 (accession no. AY680701) was used as internal control, and the relative expression level was determined using the 2 −ΔΔCT method [38]. The qRT-PCR assay was performed as previously reported [26].

Off-target analysis
Off-target analysis was performed in VvCCD8 knockout lines. Two top-ranking putative off-target sites that localize in gene exons were chosen for off-target analysis. The potential off-target regions were amplified using their specific primers (Additional file 4: Table S1), and the fragments were cloned into pLB vector and at least 6 clones were analyzed by Sanger sequencing.

EGFP detection
The EGFP signal was detected using the Eclipse Ni-U fluorescence microscope (Nikon, Japan) with excitation at 487 nm, emission at 505 nm. The wild-type cells were used as the negative control.