No abstract

Summary The CRISPR/Cas9 system has been demonstrated to efficiently induce targeted gene editing in a variety of organisms including plants. Recent work showed that CRISPR/Cas9‐induced gene mutations in Arabidopsis were mostly somatic mutations in the early generation, although some mutations could be stably inherited in later generations. However, it remains unclear whether this system will work similarly in crops such as rice. In this study, we tested in two rice subspecies 11 target genes for their amenability to CRISPR/Cas9‐induced editing and determined the patterns, specificity and heritability of the gene modifications. Analysis of the genotypes and frequency of edited genes in the first generation of transformed plants (T0) showed that the CRISPR/Cas9 system was highly efficient in rice, with target genes edited in nearly half of the transformed embryogenic cells before their first cell division. Homozygotes of edited target genes were readily found in T0 plants. The gene mutations were passed to the next generation (T1) following classic Mendelian law, without any detectable new mutation or reversion. Even with extensive searches including whole genome resequencing, we could not find any evidence of large‐scale off‐targeting in rice for any of the many targets tested in this study. By specifically sequencing the putative off‐target sites of a large number of T0 plants, low‐frequency mutations were found in only one off‐target site where the sequence had 1‐bp difference from the intended target. Overall, the data in this study point to the CRISPR/Cas9 system being a powerful tool in crop genome engineering.

The CRISPR (clustered regularly interspaced short palindromic repeat)/Cas (CRISPR-associated) system has emerged as a powerful tool for targeted gene editing in many organisms, including plants. However, all of the reported studies in plants focused on either transient systems or the first generation after the CRISPR/ Cas system was stably transformed into plants. In this study we examined several plant generations with seven genes at 12 different target sites to determine the patterns, efficiency, specificity, and heritability of CRISPR/Cas-induced gene mutations or corrections in Arabidopsis. The proportion of plants bearing any mutations (chimeric, heterozygous, biallelic, or homozygous) was 71.2% at T1, 58.3% at T2, and 79.4% at T3 generations. CRISPR/Cas-induced mutations were predominantly 1 bp insertion and short deletions. Gene modifications detected in T1 plants occurred mostly in somatic cells, and consequently there were no T1 plants that were homozygous for a gene modification event. In contrast, ∼22% of T2 plants were found to be homozygous for a modified gene. All homozygotes were stable to the next generation, without any new modifications at the target sites. There was no indication of any off-target mutations by examining the target sites and sequences highly homologous to the target sites and by in-depth whole-genome sequencing. Together our results show that the CRISPR/Cas system is a useful tool for generating versatile and heritable modifications specifically at target genes in plants.

Summary The Streptococcus-derived CRISPR/Cas9 system is being widely used to perform targeted gene modifications in plants. This customized endonuclease system has two components, the single-guide RNA (sgRNA) for target DNA recognition and the CRISPR-associated protein 9 (Cas9) for DNA cleavage. Ubiquitously expressed CRISPR/Cas9 systems (UC) generate targeted gene modifications with high efficiency but only those produced in reproductive cells are transmitted to the next generation. We report the design and characterization of a germ-line-specific Cas9 system (GSC) for Arabidopsis gene modification in male gametocytes, constructed using a SPOROCYTELESS (SPL) genomic expression cassette. Four loci in two endogenous genes were targeted by both systems for comparative analysis. Mutations generated by the GSC system were rare in T1 plants but were abundant (30%) in the T2 generation. The vast majority (70%) of the T2 mutant population generated using the UC system were chimeras while the newly developed GSC system produced only 29% chimeras, with 70% of the T2 mutants being heterozygous. Analysis of two loci in the T2 population showed that the abundance of heritable gene mutations was 37% higher in the GSC system compared to the UC system and the level of polymorphism of the mutations was also dramatically increased with the GSC system. Two additional systems based on germ-line-specific promoters (pDD45-GT and pLAT52-GT) were also tested, and one of them was capable of generating heritable homozygous T1 mutant plants. Our results suggest that future application of the described GSC system will facilitate the screening for targeted gene modifications, especially lethal mutations in the T2 population.