Engineering Herbicide-Resistant Rice Plants through CRISPR/Cas9-Mediated Homologous Recombination of Acetolactate Synthase

Genome editing technologies enable precise modifications of DNA sequences in vivo and offer great promise for crop improvement. CRISPR/Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats/CRISPR-associated Cas9) has revolutionized genome editing because of its simplicity and versatility (Cong et al., 2013Cong L. Ran F.A. Cox D. Lin S. Barretto R. Habib N. Hsu P.D. Wu X. Jiang W. Marraffini L.A. et al.Multiplex genome engineering using CRISPR/Cas systems.Science. 2013; 339: 819-823Crossref PubMed Scopus (9981) Google Scholar). However, the majority of the editing events reported in plants were through error-prone non-homologous end joining (NHEJ) to generate mutations and gene knockouts (Miao et al., 2013Miao J. Guo D. Zhang J. Huang Q. Qin G. Zhang X. Wan J. Gu H. Qu L.J. Targeted mutagenesis in rice using CRISPR-Cas system.Cell Res. 2013; 23: 1233-1236Crossref PubMed Scopus (585) Google Scholar, Voytas and Gao, 2014Voytas D.F. Gao C. Precision genome engineering and agriculture: opportunities and regulatory challenges.PLoS Biol. 2014; 12: e1001877Crossref PubMed Scopus (279) Google Scholar, Xie et al., 2014Xie K. Zhang J. Yang Y.N. Genome-wide prediction of highly specific guide RNA spacers for CRISPR-Cas9-mediated genome editing in model plants and major crops.Mol. Plant. 2014; 7: 923-926Abstract Full Text Full Text PDF PubMed Scopus (215) Google Scholar, Ma et al., 2015Ma X. Zhang Q. Zhu Q. Liu W. Chen Y. Qiu R. Wang B. Yang Z. Li H. Lin Y. et al.A robust CRISPR/Cas9 system for convenient, high-efficiency multiplex genome editing in monocot and dicot plants.Mol. Plant. 2015; 8: 1274-1284Abstract Full Text Full Text PDF PubMed Scopus (1160) Google Scholar). While complete knockouts and loss-of-function mutations are very valuable in defining gene functions, their applications in crop improvement are somewhat limited because many agriculturally important traits are conferred by point mutations or change of gene expression levels. Development of a technique that enables gene replacement rather than gene inactivation will greatly facilitate plant breeding (Weeks et al., 2016Weeks D.P. Spalding M.H. Yang B. Use of designer nucleases for targeted gene and genome editing in plants.Plant Biotechnol. J. 2016; 14: 483-495Crossref PubMed Scopus (157) Google Scholar). Homology-directed repair (HDR) after CRSIPR/Cas9 that generates double-stranded breaks (DSB) at specific locations can potentially provide a feasible approach to achieve gene replacement. However, it has been very challenging to use HDR in plants because the frequency of targeted integration by HDR remains much lower than that of NHEJ (Puchta, 2005Puchta H. The repair of double-strand breaks in plants: mechanisms and consequences for genome evolution.J. Exp. Bot. 2005; 56: 1-14Crossref PubMed Google Scholar). Another major obstacle in HDR in plants is the delivery of repair templates. Recently, particle bombardment was employed to deliver Cas9/gRNA and repair templates simultaneously into soybean and maize in attempts to achieve HDR-mediated targeted gene replacement. It was reported that a P178S mutation in the Acetolactate Synthase 1 (ALS1), a key enzyme for the biosynthesis of branched chain amino acids and a major target for important herbicides including chlorsulfuron and bispyribac sodium (BS) (Mazur et al., 1987Mazur B.J. Chui C.F. Smith J.K. Isolation and characterization of plant genes coding for acetolactate synthase, the target enzyme for two classes of herbicides.Plant Physiol. 1987; 85: 1110-1117Crossref PubMed Google Scholar), was successfully introduced in soybean through CRISPR/Cas9-mediated gene editing. But such strategy only resulted in one chlorsulfuron-resistant callus event (Li et al., 2015Li Z. Liu Z.B. Xing A. Moon B.P. Koellhoffer J.P. Huang L. Ward R.T. Clifton E. Falco S.C. Cigan A.M. Cas9-guide RNA directed genome editing in soybean.Plant Physiol. 2015; 169: 960-970Crossref PubMed Scopus (295) Google Scholar). Substitution of P165 with Ser in the ALS2 in maize using either single-stranded oligonucleotides or double-stranded DNA vectors as repair templates yielded chlorsulfuron-resistant maize plants (Svitashev et al., 2015Svitashev S. Young J.K. Schwartz C. Gao H. Falco S.C. Cigan A.M. Targeted mutagenesis, precise gene editing, and site-specific gene insertion in maize using Cas9 and guide RNA.Plant Physiol. 2015; 169: 931-945Crossref PubMed Scopus (449) Google Scholar). It is not clear whether similar strategies will work in other plant species including Arabidopsis and rice. Furthermore, it is important to investigate whether multiple discrete point mutations can be simultaneously introduced into plant genes through CRISPR/Cas9-mediated gene editing. In this study, we report an efficient method to introduce multiple discrete point mutations in the rice ALS gene using CRISPR/Cas9-mediated homologous recombination. We first used single-stranded oligonucleotides as repair templates, which were successful in maize. We co-transformed the rice calli with either pCXUN-Cas9-gRNA1 and the W548L oligonucleotide donor, or pCXUN-Cas9-gRNA2 and the S627I oligonucleotide donor, through particle bombardment (Supplemental Figure 1). Among the 480 calli bombarded, only one had the W548L HDR event and none with the edited ALS was recovered, demonstrating that this strategy works much less efficiently in rice than that reported in maize (Svitashev et al., 2015Svitashev S. Young J.K. Schwartz C. Gao H. Falco S.C. Cigan A.M. Targeted mutagenesis, precise gene editing, and site-specific gene insertion in maize using Cas9 and guide RNA.Plant Physiol. 2015; 169: 931-945Crossref PubMed Scopus (449) Google Scholar). We then decided to use two gRNAs instead of one gRNA and a repair template in an attempt to simultaneously substitute two amino acid residues (W548 to L and S627 to I) in rice ALS (Figure 1A ). We designed a plasmid that harbors cassettes expressing Cas9 and two gRNAs (Figure 1A and 1B). In addition, the plasmid contained a donor fragment was designed as a template for HDR (Figure 1B). Cas9 was under the control of the maize ubiquitin promoter and both gRNAs were driven by the rice U3 promoter (Figure 1B). The two gRNAs were located at opposite strands of the two targeting sites (target 1 and target 2) (Figure 1B). The 476-bp donor fragment contained several features (Figure 1B and Supplemental Figure 2). Firstly, the fragment contained the desired mutations (W548L and S627I substitutions) in the middle of the 264-bp core sequence of the donor fragment, which corresponded to the region (1625–1888 bp) of the wild-type ALS gene. Secondly, the donor fragment had several synonymous substitutions that would not change amino acid residues, but could prevent Cas9 from cutting the donor fragment again once HDR was successfully achieved. Thirdly, this 264-bp core sequence in the donor fragment was flanked with the 100-bp left arm and the 46-bp right arm, which were identical to the stretches of ALS sequences. The two arms were then flanked with two gRNA target sequences including the PAM motifs, respectively, so that the donor DNA can be released by Cas9/gRNAs. Moreover, an EcoRV restriction site between the two gRNA targeting sites in the donor fragment was mutated as a diagnostic marker for HDR events (Figure 1B and Supplemental Figure 2). In order to provide enough donor fragments for HDR, the free 476-bp double-stranded DNA donor fragment and the plasmid were co-introduced into rice (Japonica cv. Nipponbare) calli through particle bombardment. Among 320 calli bombarded, 116 independent lines survived the selection pressure of both hygromycin and BS. Among these lines, 52 were randomly selected for further analysis. PCR primer set 753F/R was designed to amplify an ALS fragment from both the wild-type ALS locus and the edited ALS, but not from the plasmid (Figure 1C). As shown in Figure 1C and Supplemental Table 1, EcoRV failed to cut the PCR products from all of the lines except for a few lines, which actually produced four bands after the EcoRV digestion, while wild-type DNA would only produce two bands (Figure 1C), indicating that most of the generated plants underwent HDR. Sequencing results showed that CRISPR/Cas9-mediated HDR occurred in all the plants analyzed. However, the lines with four EcoRV-digested bands had different HDR events (Figure 1C–1F and Supplemental Table 1). For all the lines analyzed, W548 was successfully changed into L in ALS as designed, and only one type of recombination event was observed at this locus and was designated as HDR548 (Figure 1D and 1F). But for the S627 locus, three different haplotypes of HDR (HDR627-1, HDR627-2, and HDR627-3) were observed (Figure 1D and 1F). Whereas HDR627-1 was exactly the same as the designed sequence, HDR627-2 and HDR627-3 were slightly different from the designed donor core sequence with one or two SNPs identical to the wild-type ALS gene (Figure 1F). For batch B97, all lines had the HDR627-1 genotype (Supplemental Table 1). In contrast, all lines produced in batch B98 had the HDR627-3 genotype (Supplemental Table 1). Interestingly, the 24 lines produced in batch B99 contained all three HDR haplotypes at the S627 locus. The other four- lines derived from batch B99 showed a different HDR genotype (Supplemental Table 1). They were bi-allelic variations: one allele had HDR548, while no HDR occurred at the S627 locus and EcoRV site in one DNA strand (B99-12–39). If this allele was used as a template, the EcoRV-digested PCR fragment amplified by using 753F/R primers would generate 488-bp and 265-bp fragments. Another allele resulted from NHEJ at both target 1 and target 2, with the sequence between the two target sites replaced by its reverse-complemented wild-type sequences (Figure 1E and Supplemental Table 1). EcoRV digestion of the PCR fragment using this allele as a template generated fragments of 424-bp and 331-bp because the distance from the EcoRV restriction site to the location of the primer had changed after integration of the reverse-complemented wild-type sequences at this locus (Figure 1C and 1E). Overall, the bi-allelic plants would generate four EcoRV bands. We also delivered the pCXUN-Cas9-gRNA1-gRNA2-armed donor vector (Figure 1B) into rice calli through Agrobacterium infection. Among 240 calli used for infection, 80 independent lines were obtained and 40 lines were randomly selected for further analysis. PCR/RE assays indicated that no homozygous HDR line was obtained, whereas 30 lines were heterozygous and the other 10 lines were wild-types. Sequencing results indicated that HDR event occurred only in one strand with HDR types of HDR548 and HDR627-1 (Figure 1F); whereas NHEJ occurred in the other strand between the two gRNA targeting sites in all 30 heterozygous lines investigated (Supplemental Table 1). Combining the results of co-bombardment of both the vector donor and free donor fragments that generated 48 homozygous lines in the T0 population among the 52 lines investigated, we believe that the strategy of using two sources of DNA repair templates simultaneously increased HDR efficiency in rice. To determine whether the plants with edited ALS had gained herbicide resistance, we sprayed both the non-transformed wild-type plants derived from tissue culture without selection and plants containing the edited ALS gene with 100 μM BS at the five-leaf stage. Ten days after the spray, new leaves of the wild-type rice plants became withered, whereas the edited plants behaved normally. After 36 days, the wild-type rice plants died, whereas the plants with the modified ALS locus exhibited tolerance to BS and grew normally (Figure 1G). In conclusion, the previously reported strategy using single-stranded oligos as repair templates did not yield edited rice plants as designed in this study, suggesting that a successful strategy in one plant species needs to be modified for other plant species. Our strategy of using two gRNAs and providing repair templates from both the plasmid and free double-stranded DNAs worked efficiently in precisely substituting two discrete amino acid residues in the ALS through particle bombardment.Thus, an agriculturally important trait in rice was successfully introduced. We not only generated homozygous herbicide-resistant rice plants in one generation but also demonstrated that the strategy using the CRISPR/Cas9 system presented here is feasible and effective in precise gene replacement and thus might facilitate crop genetic improvement. This work is supported by grants from the Chinese Ministry of Agriculture (grant no. 2016ZX08010003, 2016ZX010-2, 2014ZX08010003) and a startup from the Huazhong Agricultural University.