Expanding the Scope of CRISPR/Cpf1-Mediated Genome Editing in Rice

The CRISPR/Cas gene editing system offers great potential for functional genomics in plants and crop improvement. The specificity of Cas-directed DNA cleavage is strictly determined by a chimeric single guide RNA (sgRNA) and a short protospacer adjacent motif (PAM) in the genome (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, Zetsche et al., 2015Zetsche B. Gootenberg J.S. Abudayyeh O.O. Slaymaker I.M. Makarova K.S. Essletzbichler P. Volz S.E. Joung J. van der Oost J. Regev A. et al.Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system.Cell. 2015; 163: 759-771Abstract Full Text Full Text PDF PubMed Scopus (2493) Google Scholar). The widely used SpCas9 and its variants have been shown to recognize PAM sequences in the canonical form NGG and non-canonical NGA, NAG, or NGCG in plants (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, 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, Hu et al., 2016Hu X.X. Wang C. Fu Y.P. Liu Q. Jiao X.Z. Wang K.J. Expanding the range of CRISPR/Cas9 genome editing in rice.Mol. Plant. 2016; 9: 943-945Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar). CRISPR/Cpf1, a new class 2 CRISPR/Cas system, was recently exploited as an alternative tool for genome editing in various organisms, including plants (Zetsche et al., 2015Zetsche B. Gootenberg J.S. Abudayyeh O.O. Slaymaker I.M. Makarova K.S. Essletzbichler P. Volz S.E. Joung J. van der Oost J. Regev A. et al.Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system.Cell. 2015; 163: 759-771Abstract Full Text Full Text PDF PubMed Scopus (2493) Google Scholar, Kim et al., 2017Kim H. Kim S.T. Ryu J. Kang B.C. Kim J.S. Kim S.G. CRISPR/Cpf1-mediated DNA-free plant genome editing.Nat. Commun. 2017; 8: 14406Crossref PubMed Scopus (306) Google Scholar, Tang et al., 2017Tang X. Lowder L.G. Zhang T. Malzahn A.A. Zheng X. Voytas D.F. Zhong Z. Chen Y. Ren Q. Li Q. et al.A CRISPR-Cpf1 system for efficient genome editing and transcriptional repression in plants.Nat. Plants. 2017; https://doi.org/10.1038/nplants.2017.18Crossref Scopus (332) Google Scholar, Wang et al., 2017Wang M. Mao Y. Lu Y. Tao X. Zhu J.K. Multiplex gene editing in rice using the CRISPR-Cpf1 system.Mol. Plant. 2017; 10: 1011-1013Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar, Xu et al., 2017Xu R.F. Qin R.Y. Li H. Li D.D. Li L. Wei P.C. Yang J.B. Generation of targeted mutant rice using a CRISPR-Cpf1 system.Plant Biotechnol. J. 2017; 15: 713-717Crossref PubMed Scopus (181) Google Scholar). Cpf1 utilizes a thymidine-rich PAM site, TTTN, and is guided by a single CRISPR RNA (crRNA) (Zetsche et al., 2015Zetsche B. Gootenberg J.S. Abudayyeh O.O. Slaymaker I.M. Makarova K.S. Essletzbichler P. Volz S.E. Joung J. van der Oost J. Regev A. et al.Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system.Cell. 2015; 163: 759-771Abstract Full Text Full Text PDF PubMed Scopus (2493) Google Scholar). The Cpf1 crRNA (∼43 nt) is shorter than that of SpCas9 sgRNA by 60 nucleotides and no trans-acting crRNA (tracrRNA) is needed (Fonfara et al., 2016Fonfara I. Richter H. Bratovič M. Le Rhun A. Charpentier E. The CRISPR-associated DNA-cleaving enzyme Cpf1 also processes precursor CRISPR RNA.Nature. 2016; 532: 517-521Crossref PubMed Scopus (536) Google Scholar). These advantages of Cpf1 make it easy to generate constructs and to multiplex gene editing (Wang et al., 2017Wang M. Mao Y. Lu Y. Tao X. Zhu J.K. Multiplex gene editing in rice using the CRISPR-Cpf1 system.Mol. Plant. 2017; 10: 1011-1013Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar). Thus, the CRISPR/Cpf1 system is an attractive tool, enabling editing AT-rich regions such as 5′ and 3′ UTR and promoter regions (Zetsche et al., 2015Zetsche B. Gootenberg J.S. Abudayyeh O.O. Slaymaker I.M. Makarova K.S. Essletzbichler P. Volz S.E. Joung J. van der Oost J. Regev A. et al.Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system.Cell. 2015; 163: 759-771Abstract Full Text Full Text PDF PubMed Scopus (2493) Google Scholar). However, the target sequences for genome editing by CRISPR/Cpf1 are restricted to sites containing a TTTN motif, thus limiting the availability of suitable target sites and reducing the practical utility of Cpf1 in plants. To overcome this constraint, recent work in human cells has revealed that Cpf1 can be modified to recognize alternative PAM sequences (Gao et al., 2017Gao L.Y. Cox D.B.T. Yan W.X. Manteiga J.C. Schneider M.W. Yamano T. Nishimasu H. Nureki O. Crosetto N. Zhang F. Engineered Cpf1 variants with altered PAM specificities.Nat. Biotechnol. 2017; 35: 789-792Crossref PubMed Scopus (257) Google Scholar). Yet, the feasibility of expanding the range of genome editing via the CRISPR/Cpf1 system in plants remains to be investigated. In the Cpf1 family, LbCpf1 from Lachnospiraceae bacterium ND 2006 and AsCpf1 from Acidaminococcus sp. BV3L6 act more effectively in human cells compared with several other orthologs investigated (Zetsche et al., 2015Zetsche B. Gootenberg J.S. Abudayyeh O.O. Slaymaker I.M. Makarova K.S. Essletzbichler P. Volz S.E. Joung J. van der Oost J. Regev A. et al.Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system.Cell. 2015; 163: 759-771Abstract Full Text Full Text PDF PubMed Scopus (2493) Google Scholar). Both LbCpf1 and AsCpf1 utilize a PAM site, TTTV. Comparison of the editing efficiency of AsCpf1 and LbCpf1 in rice indicated that LbCpf1 showed higher activity (Tang et al., 2017Tang X. Lowder L.G. Zhang T. Malzahn A.A. Zheng X. Voytas D.F. Zhong Z. Chen Y. Ren Q. Li Q. et al.A CRISPR-Cpf1 system for efficient genome editing and transcriptional repression in plants.Nat. Plants. 2017; https://doi.org/10.1038/nplants.2017.18Crossref Scopus (332) Google Scholar). To expand the range of CRISPR/Cpf1-mediated genome editing in rice, we generated two LbCpf1 variants at the corresponding positions as reported in AsCpf1 (Gao et al., 2017Gao L.Y. Cox D.B.T. Yan W.X. Manteiga J.C. Schneider M.W. Yamano T. Nishimasu H. Nureki O. Crosetto N. Zhang F. Engineered Cpf1 variants with altered PAM specificities.Nat. Biotechnol. 2017; 35: 789-792Crossref PubMed Scopus (257) Google Scholar) and investigated their capacities of genome editing in rice. We demonstrate that LbCpf1 can be engineered to target and edit the genome sequences containing alternative non-canonical Cpf1 PAMs in rice, significantly broadening the range of genome editing. We first introduced two point mutations in plant codon-optimized LbCpf1 to generate the G532R/K595R variant (hereafter referred to as the RR variant) by overlapping PCR using appropriate primer sets (Figure 1A; Supplemental Table 1). To test the endogenous genome editing capability and multiplexing of the RR variant in rice, we selected OsPDS, which encodes a phytoene desaturase, and OsSBEIIb, which encodes a starch branching enzyme IIb, as the target genes and designed two crRNAs for each target gene, or two crRNAs with one crRNA targeting OsPDS and the other targeting OsSBEIIb (Figure 1A; Supplemental Figure 1 and Supplemental Table 2). Because the first nucleotide of Lb-crRNA is a U, it cannot be accommodated by RNA polymerase III (Pol III) promoters such as U6 (transcription starts with G) and U3 (transcription starts at A). Pol III-transcribed crRNAs usually contain 3’ terminal poly U sequences, which may affect their recognition by the CRISPR complex in plants (Tang et al., 2017Tang X. Lowder L.G. Zhang T. Malzahn A.A. Zheng X. Voytas D.F. Zhong Z. Chen Y. Ren Q. Li Q. et al.A CRISPR-Cpf1 system for efficient genome editing and transcriptional repression in plants.Nat. Plants. 2017; https://doi.org/10.1038/nplants.2017.18Crossref Scopus (332) Google Scholar, Xu et al., 2017Xu R.F. Qin R.Y. Li H. Li D.D. Li L. Wei P.C. Yang J.B. Generation of targeted mutant rice using a CRISPR-Cpf1 system.Plant Biotechnol. J. 2017; 15: 713-717Crossref PubMed Scopus (181) Google Scholar). To overcome this sequence specificity, the crRNAs ribozyme cassettes were driven by different promoters such as Ubi promoter (Tang et al., 2017Tang X. Lowder L.G. Zhang T. Malzahn A.A. Zheng X. Voytas D.F. Zhong Z. Chen Y. Ren Q. Li Q. et al.A CRISPR-Cpf1 system for efficient genome editing and transcriptional repression in plants.Nat. Plants. 2017; https://doi.org/10.1038/nplants.2017.18Crossref Scopus (332) Google Scholar). Here, we took advantage of the nuclease activity of ribozymes (Gao and Zhao, 2014Gao Y. Zhao Y. Self-processing of ribozyme-flanked RNAs into guide RNAs in vitro and in vivo for CRISPR-mediated genome editing.J. Integr. Plant Biol. 2014; 56: 343-349Crossref PubMed Scopus (357) Google Scholar) and put the two crRNA ribozyme cassettes in a single array, named as RCRs (ribozyme-crRNA1-ribozyme-ribozyme-crRNA2-ribozyme), under the drive of a single OsU3 promoter to achieve coordinated expression and precise release of the designed crRNAs without additional poly U sequences at the 3′ terminus through ribozyme self-catalyzed cleavage. We assembled the LbCpf1 variant and two RCRs to generated three pCXUN-LbCpf1(RR)-OsU3-RCRs vectors: pCXUN-LbCpf1(RR)-OsU3-RCR1-RCR2 (PDS), pCXUN-LbCpf1(RR)-OsU3-RCR1-RCR2(SBEIIb), and pCXUN-LbCpf1(RR)-OsU3-RCR1(PDS)-RCR2(SBEIIb) (Supplemental Figure 1A–1C). We then transformed rice calli by Agrobacterium-mediated transformation to generate stable lines and tested the LbCpf1 RR variant vectors for their feasibility, efficacy, and multiplexing activity in genome editing. For OsPDS (Figure 1A-1), among 99 independent transgenic lines recovered, we detected 51 plants with targeted mutagenesis (Figure 1B and 1C). Mutations around target one were detected in 20 plants, among which 19 were heterozygous lines and one was a bi-allelic line (Figure 1C). Mutations around target two were detected in 22 plants, and 19 were heterozygous lines and three were chimeric lines (Figure 1C). Interestingly, nine plants were identified to have two target sites mutated simultaneously, and one plant had a deletion between the two targets, indicating that it is feasible to multiplex editing using the RR variant (Figure 1C). Different from the common 1–2 bp short indels generated by Cas9 in rice, most of the mutations generated by the LbCpf1 RR variant in this study were larger indels >2 bp at 3′ of the target sequence (Figure 1B). The mutation spectrum was similar to those recently reported for single gene editing using LbCpf1 in rice (Xu et al., 2017Xu R.F. Qin R.Y. Li H. Li D.D. Li L. Wei P.C. Yang J.B. Generation of targeted mutant rice using a CRISPR-Cpf1 system.Plant Biotechnol. J. 2017; 15: 713-717Crossref PubMed Scopus (181) Google Scholar). The editing efficiencies at target one, target two, and both targets were 20.2% (20/99), 22.2% (22/99), and 9.1% (9/99), respectively, suggesting reduced activity of RR variant. Loss of function mutations of OsPDS in the bi-allelic lines led to albino phenotype (Supplemental Figure 2). For OsSBEIIb (Figure 1A-2), among 90 independent transgenic lines recovered, we detected 29 plants with targeted mutagenesis (Figure 1C and 1D). Mutations only around target one were detected in 28 plants, whereas one heterozygous plant was identified to have deletion of fragment between two targets (Figure 1C). No other mutations in target two were detected. Of the plants only with mutations at target one, 25 were heterozygous, and three were chimeric. The editing efficiencies at target one and both targets were 31.1% (28/90) and 1.1% (1/90), respectively (Figure 1C). Combined with the results of OsPDS editing, the variations in editing efficiencies were probably due to differences in target sequences. For crRNAs targeting both OsPDS and OsSBEIIb (Figure 1A-3, Supplemental Figure 1C), among 97 independent transgenic lines recovered, 42 plants were detected with targeted mutations (Figure 1C and 1E). Mutations only at OsPDS locus were detected in 17 plants, mutations only at OsSBEIIb locus were detected in 16 plants, whereas nine plants had two loci mutated simultaneously (Figure 1C). The editing efficiencies at OsPDS, OsSBEIIb, or both OsPDS and OsSBEIIb loci were 17.5% (17/97), 16.5% (16/97), and 9.3% (9/97), respectively (Figure 1C). No off-target effects were found at potential off-target sites in these tested lines (Supplemental Table 3). We then genotyped four T0 heterozygous lines, two for OsPDS and OsSBEIIb, respectively, by PCR using the DNA extracted from three independent leaves from each line. All sampled leaves from individual plants carried the same mutations, suggesting these mutations in the lines tested will probably transmit to the next generation. Taken together, these results indicate that the RR variant may be used to edit genome sequences bearing TYCV (Y = C/T) PAMs with comparable efficacy. To further broaden the genome editing scope of CRISPR/Cpf1 in rice, we then modified the 532G, 538K, and 542Y to generate G532R/K538V/Y542R variant (RVR variant) (Supplemental Figure 1D–1F; Supplemental Table 1), which was shown to be able to recognize TATV (V = A/C/G) PAMs in human cells (Gao et al., 2017Gao L.Y. Cox D.B.T. Yan W.X. Manteiga J.C. Schneider M.W. Yamano T. Nishimasu H. Nureki O. Crosetto N. Zhang F. Engineered Cpf1 variants with altered PAM specificities.Nat. Biotechnol. 2017; 35: 789-792Crossref PubMed Scopus (257) Google Scholar). Similarly, we designed two crRNAs for OsPDS and OsSBEIIb (Supplemental Table 2), respectively, and generated three pCXUN-LbCpf1(RVR)-OsU3-RCRs vectors, including pCXUN-LbCpf1(RVR)-OsU3-RCR1-RCR2(PDS), pCXUN-LbCpf1(RVR)-OsU3-RCR1-RCR2(SBEIIb), and pCXUN-LbCpf1(RVR)-OsU3-RCR1(PDS)-RCR2 (SBEIIb) (Supplemental Figure 1D–1F). Among 98, 95, and 108 independent transgenic lines recovered for OsPDS and OsSBEIIb targets, respectively, we failed to detect any mutant lines, indicating that although the RVR variant worked well in human cells (Gao et al., 2017Gao L.Y. Cox D.B.T. Yan W.X. Manteiga J.C. Schneider M.W. Yamano T. Nishimasu H. Nureki O. Crosetto N. Zhang F. Engineered Cpf1 variants with altered PAM specificities.Nat. Biotechnol. 2017; 35: 789-792Crossref PubMed Scopus (257) Google Scholar), it was not capable of genome editing in rice in our study. These results demonstrate that gene editing efficiency/activities of nuclease variants need to be tested in a specific organism before large-scale experiments are conducted. The LbCpf1(RR) variant described in this study can in principle target any sequences containing TYCV PAMs. Computational analysis of the rice genome sequence (Os-Nipponbare-Reference-IRGSP-1.0) revealed that among approximately 55 986 annotated genes, 96% (53 764) of genes have the TTTV PAMs, whereas 99.6% (55 762) harbor the TYCV PAMs. Of 2222 genes that do not have the TTTV PAMs, 93.7% (2080) have the TYCV PAMs. Thus, with the development of the RR variant, the number of endogenous sites in rice accessible to CRISPR/Cpf1 reagent for genome editing increased to approximately 99.7% (55 844) (Figure 1F). Given that the GC content and crRNA secondary structure may affect the efficacy of genome editing as demonstrated with the SpCas9 nuclease (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), the development of the RR variant will provide more choices in selecting suitable targets and greatly facilitate the practical application of the CRISPR/Cpf1 system in genome editing of crop plants. In conclusion, we generated two modified LbCpf1 variants and demonstrated that only LbCpf1(RR) variant enables efficient editing or multiplex editing of target genes containing non-canonical TYCV PAMs in rice. This expands the scope of LbCpf1-mediated genome editing in rice. The LbCpf1(RR) variant would be useful for both basic research and crop breeding in other plant species. This work is partly funded by the Ministry of Science and Technology of China (grant no. 2016YFD0102003) and the Chinese Ministry of Agriculture (grant no. 2018ZX0801016B and 2016ZX08010003). Y.D.Z. is supported by a startup project from Hua-Zhong Agricultural University, China.