Identification of Transgene-Free CRISPR-Edited Plants of Rice, Tomato, and Arabidopsis by Monitoring DsRED Fluorescence in Dry Seeds

Abstract

Efficient elimination of the editing machinery remains a challenge in plant biotechnology after genome editing to minimize the probability of off-target mutations, but it is also important to deliver end users with edited plants free of foreign DNA. Using the modular cloning system Golden Braid, we have included a fluorescence-dependent transgene monitoring module to the genome-editing tool box. We have tested this approach in Solanum lycopersicum, Oryza sativa, and Arabidopsis thaliana. We demonstrate that DsRED fluorescence visualization works efficiently in dry seeds as marker for the detection of the transgene in the three species allowing an efficient method for selecting transgene-free dry seeds. In the first generation of DsRED-free CRISPR/Cas9 null segregants, we detected gene editing of selected targets including homozygous mutants for the plant species tested. We demonstrate that this strategy allows rapid selection of transgene-free homozygous edited crop plants in a single generation after in vitro transformation.

Keywords: Arabidopsis thaliana; CRISPR/Cas9; DsRED; Oryza sativa; Solanum lycopersicum; genome editing.

Figure 1

(A) Position of each CRISPR target in selected genes At5g55250 (Arabidopsis thaliana), Solyc07g64990 and Solyc12g14500 (Solanum lycopersicon), and Os04g56950 (Oryza sativa). (B) Description of transcriptional units (arrows) assembled in the vectors generated for plant transformation using the modular system GoldenBraid. Specific promoters were selected for expression of Cas9 protein in each species, pAtUBQ10 in Arabidopsis thaliana, p35S for Solanum Lycopersicum, and pZmUBQ for Oryza sativa. Adequate promoter for expression of sgRNA was selected: pAtU6-26 for dicotyledonous species (Solanum and Arabidopsis) and pOsU3 for monocotyledonous species (rice).

Figure 2

Diagram describing the steps for plant transformation and selection. Transformation of Arabidopsis thaliana was done by floral dip while in vitro transformation was used for Solanum lycopersicum and Oryza sativa. Selection of DsRED T1 seeds of Arabidopsis thaliana and DsRED-negative selection of segregant seeds from the three species was done by direct visualization under a stereoscope equipped with DsRED filter. Detection of fluorescence in seed is very clear, easy, and fast in all three species, and it allowed perfect separation of positive and negative fluorescent seeds.

Figure 3

(A) Absence of transgene in plants from DsRED-negative selected seeds was confirmed by PCR using Cas9-specific primers. Examples are shown for tomato, Arabidopsis and rice, selecting different number of seeds and number of independent lines. In all cases, fluorescence and Cas9 PCR band did correlate perfectly (an image indicates the presence/absence of DsRED fluorescence in the original selected seed). (B) Sequence alignment of mutations detected in homozygosis in transgene-free T2 individual plants. CRISPR target sequence in red and PAM sequence in green. Nucleotide insertions are indicated in blue; deletions are also indicated with blue dashes. Deletion “del193” starts in target 2 and ends around 150 nts downstream target 3. It also includes an insertion of six nts that interestingly match with six nucleotides upstream target 2 (Supplementary Figure 5B).