Rapid breeding of parthenocarpic tomato plants using CRISPR/Cas9

Abstract

Parthenocarpy in horticultural crop plants is an important trait with agricultural value for various industrial purposes as well as direct eating quality. Here, we demonstrate a breeding strategy to generate parthenocarpic tomato plants using the CRISPR/Cas9 system. We optimized the CRISPR/Cas9 system to introduce somatic mutations effectively into SlIAA9-a key gene controlling parthenocarpy-with mutation rates of up to 100% in the T0 generation. Furthermore, analysis of off-target mutations using deep sequencing indicated that our customized gRNAs induced no additional mutations in the host genome. Regenerated mutants exhibited morphological changes in leaf shape and seedless fruit-a characteristic of parthenocarpic tomato. And the segregated next generation (T1) also showed a severe phenotype associated with the homozygous mutated genome. The system developed here could be applied to produce parthenocarpic tomato in a wide variety of cultivars, as well as other major horticultural crops, using this precise and rapid breeding technique.

Figure 1

Site-directed mutagenesis in SlIAA9 using CRISPR/Cas9. (a) Schematic representation of T-DNA regions of the CRISPR/Cas9 binary vectors used in this study. U6-26 gRNA: Arabidopsis U6 snRNA-26 promoter and the gRNA sequence, 2 × 35SΩ: 2 × CaMV35S promoter with the omega enhancer sequence, Pubi4: parsley ubiquitin 4-2 promoter, AtCas9: Arabidopsis-codon optimized SpCas9, 2 A: 2 A self-cleavage peptide, Km: the kanamycin resistant marker expression cassette, RB: right border of T-DNA, LB: left border of T-DNA. (b) Target sites for SlIAA9. 17 b/18 b-length target sequences are in red. The 2–3 base extensions of the 5′ end for the 20 b-length target sequences are in green. The PAM is in blue. A bent arrow indicates the translational start site. (c) Visual GFP-selection of transgenic calli with introduced gRNA and Cas9-2A-GFP. Tissues indicated by white circles (white dotted-circles in lower panels “GFP”) were used for mutation analysis or plantlet-regeneration. (d) Heteroduplex mobility assay with the MultiNA electrophoresis system. Multiple heteroduplex peaks (red arrows) were detected in PCR amplicons from the CRISPR/Cas9 transgenic tomato calli, whereas a single peak was detected from the wild-type control (blue arrow). VC; vector control. M; internal marker. (e) PCR-RFLP analysis of the genome editing in Micro-Tom calli (RNA2-20b). +; Acc I digested PCR products, −; non-digested PCR products. Numbers show the regenerated T0 plant lines. (f) Determination of mutation ratios in transgenic calli and shoots. The mutation ratios were calculated by dividing number of mutant calli or shoots by the total number of transformed calli or shoots.

Figure 2

CRISPR/Cas9-induced SlIAA9 mutations in transgenic tomato calli and shoots. (a) Comparison of the rates of high-efficiency mutations (100% mutation at somatic levels detected by PCR-RFLP) using different promoters for Cas9 expression, or different lengths of gRNAs. The mutation rates were calculated by dividing number of 100% mutation shoots by the total number of all-types of mutated shoots. (b) Mutation sequences in transgenic calli transformed with pEgPubi-gRNA2-20b (line #8 in Fig. 2) or pEgP237-gRNA2-20b (line #9 in Fig. 2). The WT sequences are shown on top. gRNA target sequences are indicated in blue boxes. Red; mutations generated by CRISPR/Cas9, Magenta; stop codons generated by the CRISPR/Cas9-induced mutations. (c) Summary of mutation rates analyzed by NGS in SlIAA9-crispr plants. The mutation rates and patterns around the PAM sequence were shown in circle and bar graphs, respectively. Mutation rates were calculated using total read numbers at sequence position. NHEJ; non-homologous end joining. PAM; green nucleotides. (d) Mutation rates of off-target sites of gRNA2. Off-target candidates were analyzed by the “focas” website.

Figure 3

CRISPR/Cas9-induced SlIAA9 mutations in transgenic tomato plants and their parthenocarpic phenotypes. (a) Putative mutant form of the SlIAA9 protein produced by the CRISPR/Cas9 gRNA2. The conserved domains I, II, III and IV of AUX/IAA proteins are indicated. (b) Simple leaf morphology in the SlIAA9-crispr Micro-Tom plants (#10 and #11 were 100% in PCR-RFLP, #12 showed mild mutation; data not shown). bar = 2 cm. (c) Regenerated transgenic T0 plants with fruit formation. bar = 3 cm. (d) PCR-RFLP analysis of leaves from pEgPubi4_237-gRNA2-20b T0 #8 (see Supplementary Fig. 5a) (e) PCR-RFLP analysis of Ailsa Craig mutants (pEgPubi4_237-gRNA2-20b T0) shown in panel (f). (f) Abnormal leaf morphology in the SlIAA9-crispr Ailsa Craig plants. bar = 2 cm. (g) The SlIAA9-crispr Ailsa Craig plants. bar = 3 cm. (h) Seedless fruit formation in the SlIAA9-crispr T0 Micro-Tom. Seeds in the WT fruit are indicated by white arrows. bar = 5 mm. (i) Seed formation rates in SlIAA9-crispr Micro-Tom plants. The average seed numbers were calculated in fruits (N = 5–18) in the individual plant lines. (j) Fruit formation showing parthenocarpy in SlIAA9-crispr Micro-Tom plants. Three plants of each construct and WT were used for the measurements. + ; Acc I digested PCR products, −; non-digested PCR products. Error bars indicate SE. ND; not detected.