Highly Efficient A·T to G·C Base Editing by Cas9n-Guided tRNA Adenosine Deaminase in Rice

The newly developed CRISPR/Cas9-mediated base editing technology with cytosine deaminase is capable of precisely and efficiently introducing point mutations at the target genomic locus, which does not require double-stranded DNA breaks or any donor templates and thus exhibit a great potential for gene correction and genetic diversification in yeasts, plants, and mammalian and human cells (Komor et al., 2016Komor A.C. Kim Y.B. Packer M.S. Zuris J.A. Liu D.R. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage.Nature. 2016; 533: 420-424Crossref PubMed Scopus (2428) Google Scholar, Nishida et al., 2016Nishida K. Arazoe T. Yachie N. Banno S. Kakimoto M. Tabata M. Mochizuki M. Miyabe A. Araki M. Hara K.Y. et al.Targeted nucleotide editing using hybrid prokaryotic and vertebrate adaptive immune systems.Science. 2016; 353: aaf8729Crossref PubMed Scopus (747) Google Scholar, Lu and Zhu, 2017Lu Y. Zhu J.K. Precise editing of a target base in the rice genome using a modified CRISPR/Cas9 system.Mol. Plant. 2017; 10: 523-525Abstract Full Text Full Text PDF PubMed Scopus (274) Google Scholar, Ren et al., 2017Ren B. Yan F. Kuang Y. Li N. Zhang D. Lin H. Zhou H. A CRISPR/Cas9 toolkit for efficient targeted base editing to induce genetic variations in rice.Sci. China. Life Sci. 2017; 60: 516-519Crossref PubMed Scopus (68) Google Scholar). However, compared with AID/APOBEC1 members of the cytosine deaminase family that are widely utilized in base editing to induce cytidine (C) to thymine (T) conversion, adenosine deaminase is far from being applicable since TadA/ADAR members act strictly on duplex RNA, or DNA/RNA hybrids with mismatches, instead of single-stranded DNA (Zheng et al., 2017Zheng Y. Lorenzo C. Beal P.A. DNA editing in DNA/RNA hybrids by adenosine deaminases that act on RNA.Nucleic Acids Res. 2017; 45: 3369-3377PubMed Google Scholar). To address this problem, great efforts have been invested recently in identifying Escherichia coli TadA variants that accept DNA as a substrate through rounds of protein evolution and engineering, ultimately leading to a number of adenine base editors (ABEs) with great efficiencies and broadened sequence compatibility in inducing nucleotide changes at a wide range of target genomic loci in human cells (Gaudelli et al., 2017Gaudelli N.M. Komor A.C. Rees H.A. Packer M.S. Badran A.H. Bryson D.I. Liu D.R. Programmable base editing of A·T to G·C in genomic DNA without DNA cleavage.Nature. 2017; 551: 464-471Crossref PubMed Scopus (1806) Google Scholar). In these ABE systems, the TadA:TadA* heterodimer is guided by the Cas9n/single guide RNA (sgRNA) complex to the target site, and the engineered TadA*, but not the wild-type TadA, functions as the active tRNA adenosine deaminase that turns adenine (A) to inosine (I) in single-stranded genomic DNA, subsequently resulting in A to guanine (G) mutation in genome during DNA repair or DNA replication (Gaudelli et al., 2017Gaudelli N.M. Komor A.C. Rees H.A. Packer M.S. Badran A.H. Bryson D.I. Liu D.R. Programmable base editing of A·T to G·C in genomic DNA without DNA cleavage.Nature. 2017; 551: 464-471Crossref PubMed Scopus (1806) Google Scholar). These tools, together with previous base editors, enable programmable introduction of all four transitions (C to T, G to A, A to G, and T to C) at the target loci in the genome, greatly expanding the capabilities of base editing. Here, we report the development of fluorescence-tracking base editing systems with E. coli TadA variants and Cas9 variants in rice. We first codon-optimized the wild-type E. coli TadA gene, which encodes the non-catalytic monomer acting as a docking station for the substrate, and TadA*7.10 version, which encodes a catalytic monomer with the most robust deaminase activity being reported at protospacer position −17 to −14 (counting the first nucleotide upstream of the protospacer adjacent motif PAM as position −1) for expression in rice (Supplemental Figures 1 and 2). Both TadA and TadA*7.10 were fused together to the N-terminus of Cas9(D10A) nickase (Cas9n) and catalytically dead Cas9 protein (dCas9) with two 32-amino-acid XTEN2 linkers, resulting in rBE14 and rBE15, respectively (Figure 1A and Supplemental Table 1). Furthermore, A142N and P152R mutations were imported into TadA*7.10 to generate TadA*7.8, and used to create rBE17 and rBE18 in the same manner as described for rBE14 and rBE15 (Figure 1A), presumably broadening the editing window, which is from position −17 to −11 within the protospacer. Meanwhile, the mGFP5-ER expression cassette was incorporated into the entry vector pENTR4:sgRNA5 (Zhou et al., 2014Zhou H.B. Liu B. Weeks D.P. Spalding M.H. Yang B. Large chromosomal deletions and heritable small genetic changes induced by CRISPR/Cas9 in rice.Nucleic Acids Res. 2014; 42: 10903-10914Crossref PubMed Scopus (386) Google Scholar), resulting in pENTR4:sgRNA8 (Figure 1A). The mGFP5-ER cassette and the sgRNA cassette can be gatewayed together to pUbi:rBE vectors for Agrobacterium-mediated rice transformation. Therefore, transgenes can be readily tracked in rice plants under a handheld UV lamp. We next introduced rBE14, rBE15, rBE17, and rBE18, together with an sgRNA targeting the pathogen-responsive phosphorylation site in the endogenous OsMPK6 gene (Hou et al., 2016Hou Y. Tong X. Wang Y. Qiu J. Li Z. Zhang W. Huang S. Zhang J. Data set from a comprehensive phosphoproteomic analysis of rice variety IRBB5 in response to bacterial blight.Data Brief. 2016; 6: 282-285Crossref PubMed Scopus (4) Google Scholar), into rice cells to investigate the feasibility and efficacy of our ABEs, respectively (Figure 1B). A total of 68, 61, 60, and 69 independent transgenic callus lines from 12 batches of rice transformation were generated for rBE14, rBE15, rBE17, and rBE18, respectively. The target region of OsMPK6 was individually PCR amplified with gene-specific primers and subjected to Sanger sequencing. Interestingly, 12 lines in total were detected with base editing events and all were related to rBE14 (16.67% efficiency), with no mutants being identified for rBE15, rBE17, or rBE18 (Figure 1C and 1D). Sequencing results also indicated that all mutant lines were heterozygous or monoallelic with one OsMPK6 allele carrying the desired Y227P substitution, which results from a pure A to G conversion at protospacer position −15 (Supplemental Figure 3A). Of note, neither any other nucleotide changes in the putative editing window nor any indel mutations were identified (Supplemental Figure 3B). Therefore, adenosine deaminase-engineered rBE14 is fully functional in rice, and induces point mutations much more cleanly than cytosine deaminase-based rBE systems we reported previously (Ren et al., 2017Ren B. Yan F. Kuang Y. Li N. Zhang D. Lin H. Zhou H. A CRISPR/Cas9 toolkit for efficient targeted base editing to induce genetic variations in rice.Sci. China. Life Sci. 2017; 60: 516-519Crossref PubMed Scopus (68) Google Scholar). In general, dCas9 and Cas9 undergo progressive structure transitions in a slightly different way as they bind to DNA, and both TadA*7.10 and TadA*7.8 have been evolved from TadA, nonfunctional wild-type unable to deaminate adenine in DNA, through protein engineering on the base of Cas9n (Gaudelli et al., 2017Gaudelli N.M. Komor A.C. Rees H.A. Packer M.S. Badran A.H. Bryson D.I. Liu D.R. Programmable base editing of A·T to G·C in genomic DNA without DNA cleavage.Nature. 2017; 551: 464-471Crossref PubMed Scopus (1806) Google Scholar). Together with our data, this suggests that the H840A substitution in dCas9 might trigger improper conformation change, not allowing the 39.9 kDa TadA:TadA* heterodimer to interact with the target A within the editing window of the protospacer. Similarly, the conformation of Cas9n/sgRNA complex might affect the functionality of TadA* in base editing. TadA*7.8 has been reported with slightly decreased deaminase activity and a broadened editing window in human cells compared with TadA*7.10 (Gaudelli et al., 2017Gaudelli N.M. Komor A.C. Rees H.A. Packer M.S. Badran A.H. Bryson D.I. Liu D.R. Programmable base editing of A·T to G·C in genomic DNA without DNA cleavage.Nature. 2017; 551: 464-471Crossref PubMed Scopus (1806) Google Scholar). Therefore, we further examined the efficacy of rBE14 and rBE17 by targeting the translation initiation site in the endogenous OsMPK13 gene (Supplemental Figure 4A). After screening a total of 93 and 76 independent transgenic callus lines obtained for rBE14 and rBE17, respectively, we identified 10 mutant lines only for rBE14 (Supplemental Figure 4B). Among them, a pure A to G conversion was detected at protospacer position −15 in six lines as well as at position −13 in four lines. All mutant lines were heterozygous or monoallelic and had no indels detected (Supplemental Figure 4C–4E). Taken together, our data suggest that TadA*7.8 has poor performance at two loci tested here, while TadA*7.10 is more suitable for base editing of A to G in the rice genome. We then only included rBE14 in the downstream experiments. Next, the pathogen-responsive phosphorylation sites in the endogenous OsSERK2 and OsWRKY45 genes (Ueno et al., 2015Ueno Y. Yoshida R. Kishi-Kaboshi M. Matsushita A. Jiang C.J. Goto S. Takahashi A. Hirochika H. Takatsuji H. Abiotic stresses antagonize the rice defence pathway through the tyrosine-dephosphorylation of OsMPK6.PLoS Pathog. 2015; 11: e1005231Crossref PubMed Scopus (61) Google Scholar, Hou et al., 2016Hou Y. Tong X. Wang Y. Qiu J. Li Z. Zhang W. Huang S. Zhang J. Data set from a comprehensive phosphoproteomic analysis of rice variety IRBB5 in response to bacterial blight.Data Brief. 2016; 6: 282-285Crossref PubMed Scopus (4) Google Scholar) were targeted to further explore the action of rBE14 (Figure 1E and 1G). For OsSERK2, out of 78 independent transgenic callus lines obtained, 25 lines (32.05% efficiency) were identified with nucleotide changes. We found only one type of mutant allele in which an A to G conversion occurred at position −15, replacing the phosphorylatable serine residue with proline residue as desired (Figure 1F; Supplemental Figure 5A and 5B). As for OsWRKY45, 33 of 53 callus lines (62.26% efficiency) were identified with point mutations at position −14 or/and −17 (Figure 1H and 1I; Supplemental Figure 6A). Among them, 20 lines carried a single A to G conversion at the desired site, and the additional lines had two point mutations. To examine whether the two point mutations were located together or separately on chromosomes, we cloned the PCR products and sequenced the individual clones. The results showed that four lines contained diallelic mutations and nine lines contained monoallelic mutations of the target site (Figure 1I and Supplemental Figure 6B). Similarly, no indel mutations were found for both OsSERK2 and OsWRKY45. TadA*7.10 was derived from native E. coli TadA, which prefers deaminating the A in the UAC anticodon of tRNAArg, and the engineered TadA:TadA7.10 heterodimer overcomes the sequence preference and functions in a context-independent manner in human cells (Gaudelli et al., 2017Gaudelli N.M. Komor A.C. Rees H.A. Packer M.S. Badran A.H. Bryson D.I. Liu D.R. Programmable base editing of A·T to G·C in genomic DNA without DNA cleavage.Nature. 2017; 551: 464-471Crossref PubMed Scopus (1806) Google Scholar). To gain more insights into the functionality of TadA*7.10 in targeted base editing in rice, we summarized the efficiencies of mutations in different DNA contexts at all target sites tested in this study, including Tms9-1 locus, which was resistant to rBE14 (Figure 1J). The variation in base editing efficiency on the same sequence composition of different genomic loci implies that the performance of TadA*7.10-based rBE14 is likely affected by chromatin states of the target region, besides the conformation of Cas9n/sgRNA complex we proposed above. Meanwhile, these data, together with the results of ABEs in human cells, suggest that the three nucleotides at position −16 to −14 in the protospacer would be strongly preferred by rBE14. In addition, different from the complex output of cytidine base editors in rice (Li et al., 2017Li J. Sun Y. Du J. Zhao Y. Xia L. Generation of targeted point mutations in rice by a modified CRISPR/Cas9 system.Mol. Plant. 2017; 10: 526-529Abstract Full Text Full Text PDF PubMed Scopus (200) Google Scholar, Ren et al., 2018Ren B. Yan F. Kuang Y. Li N. Zhang D. Zhou X. Lin H. Zhou H. Improved base editor for efficiently inducing genetic variations in rice with CRISPR/Cas9-guided hyperactive hAID mutant.Mol. Plant. 2018; https://doi.org/10.1016/j.molp.2018.01.005Abstract Full Text Full Text PDF Scopus (124) Google Scholar), the exclusive A to G mutation and the lack of indels in the editing window for rBE14 suggest that inosine, yielded from adenosine deamination in genomic DNA, would not trigger double-strand breaks and somehow be strictly processed during DNA repair. More endogenous loci in the rice genome will be further targeted to exploit these features. We also examined the potential off-target effect of rBE14 for OsWRKY45- and OsSERK2-targeting sgRNA due to its great on-target efficiencies. Bioinformatics analysis identified three and one potential off-target sites in the rice genome for these (Supplemental Figure 7). Sequencing data analysis indicated no trace of any nucleotide changes at these sites (data not shown). Finally, transgenic calli were visualized under a handheld UV lamp and a confocal microscope. Bright green fluorescence of mGFP5-ER in rice cells (Figure 1K) suggests that it will be very convenient to isolate the transgene-free, base-edited rice plants in the next generation through detection of fluorescence. In summary, we have successfully developed a fluorescence-tracking adenine base editor using the Cas9n-guided TadA:TadA7.10 heterodimer, which can efficiently and cleanly introduce A to G conversion in rice. rBE14, together with the other rBE vectors in the same toolkit (Ren et al., 2017Ren B. Yan F. Kuang Y. Li N. Zhang D. Lin H. Zhou H. A CRISPR/Cas9 toolkit for efficient targeted base editing to induce genetic variations in rice.Sci. China. Life Sci. 2017; 60: 516-519Crossref PubMed Scopus (68) Google Scholar, Ren et al., 2018Ren B. Yan F. Kuang Y. Li N. Zhang D. Zhou X. Lin H. Zhou H. Improved base editor for efficiently inducing genetic variations in rice with CRISPR/Cas9-guided hyperactive hAID mutant.Mol. Plant. 2018; https://doi.org/10.1016/j.molp.2018.01.005Abstract Full Text Full Text PDF Scopus (124) Google Scholar), would be very useful in the generation of DNA variations in rice to a large extent for both gene function study and rice improvement in the future. This study was supported by grants from the National Key Research and Development Program of China (2017YFD0200900) and the Agricultural Science and Technology Innovation Program of The Chinese Academy of Agricultural Sciences to H.Z., and a grant from the National Natural Science Foundation of China (31701780) to F.Y.