Targeted G‐to‐T base editing for generation of novel herbicide‐resistance gene alleles in rice

Deaminase-based cytosine base editors (CBEs) and adenine base editors (ABEs) induce C-to-T and A-to-G transitions, respectively, enabling single-nucleotide variants (SNVs) in plants for research and crop enhancement (Li et al., 2023a). The C-to-G base editors (CGBEs) and A-to-Y base editors (AYBEs), developed by incorporating glycosylases with CBEs and ABEs, expand the repertoire of base editing products, allowing C-to-T/G and A-to-T/G transversions in plants (Li et al., 2023b, 2023c; Sretenovic et a., 2021; Wu et al., 2023; Zeng et al., 2022). Yet, no current base editors could induce efficient C-to-A or A-to-C conversions in plants. Very recently, a novel base editor constructed by directly fusing an engineered N-methylpurine DNA glycosylase (gMPGv6.3) with Cas9 nickase (nCas9), namely glycosylase-based guanine base editor (gGBE), allows for the direct excision of guanine to generate abasic sites (Tong et al., 2023). The abasic sites are repaired by endogenous translesion DNA synthesis (TLS) and/or DNA replication, producing efficient G-to-T and G-to-C (C-to-A and C-to-G in the opposite strand) base conversion in mammalian cells. Development of gGBE to broaden the editing scope is highly desirable for plant basic research and breeding applications (Figure S1). However, its implementation in plants remains unexplored. In this study, we developed a rice guanine base editor (OsgGBE) by fusing a rice codon-optimized gMPGv6.3 (RgMPG) with the nCas9 (D10A). This new base editor generates G-to-T base conversion in transgenic rice plants with up to 39.1% efficiency. Similar to the gGBE in mammalian cells, G-to-C (up to 8.7%) and G-to-A (up to 2.4%) editing are also detected in rice. Interestingly, G-to-T is the predominant editing type generated by OsgGBE in rice, with the purity (rate of G-to-T edits over total base edits) exceeding 80%, showing a quite different editing pattern from that in mammalian cells (Tong et al., 2023). Hence, we termed the OsgGBE as OsGTBE (Figure 1A) to better reflect its editing characteristics. Oryza sativa G-to-T base editor (OsGTBE) enables targeted G-to-T editing in rice (A) Plasmid constructs used in this study. (B) Gene structure of OsALS1. The single guide RNA (sgRNA) targets of OsCGBE03 and OsGTBE were underlined, with the protospacer adjacent motifs (PAMs) marked in bold. (C) Editing outcomes of different base editors in rice callus. Frequencies were calculated as desired reads/total reads. (D) Distributions of base editing products at specific positions. The first nucleotide of the protospacer distal to the PAM was recognized as position 1. Proportions base conversion types were calculated as desired base-edited reads/total base-edited reads. Base-edited reads counted only edits without InDel combinations. (E) Amino acid changes induced by OsGTBE in rice callus. (F, G) Genotypes of representative OsGTBE-edited plants at OsALS1 (F) and OsACC1 (G). Base conversions are indicated with red arrows. Alleles determined with HiTOM were listed under corresponding chromatograms. (H) Summary of editing outcomes induced by OsGTBE in rice T0 plantlets. (I) Summary of base-edited rice T0 plantlets across protospacers from five endogenous targets. (J) Proportion of allele patterns of base-edited T0 plants. (K) Diagram illustrating desired base conversion with available base editors in rice. (L) Herbicide-tolerance assay on OsGTBE-edited plants with amino acid substitution at OsALS1 or OsACC1. Amino acid substitution types are annotated below corresponding plants. Phenotypes were recorded 20 d post-foliar spraying with herbicides. Scale bar, 3 cm. In mammalian cells, gGBE is capable of efficient G-to-T/C (C-to-A/G in the opposite strand) editing. Combined with CGBE (induce C-to-G/T editing), the G:C base-pair could be converted to any desired base-pair (Tong et al., 2023). To assess the complementary nature of CGBE and OsgGBE (OsGTBE), we investigate and compare the editing profiles of OsGTBE and OsCGBE03 at endogenous loci in rice. Two pairs of single guide RNAs (sgRNAs) were designed to target the P171 and G628 residues of OsALS1 (Os02g0510200). As shown in Figure 1B, sgRNA#1 and sgRNA#2 direct OsCGBE03 to target the P171 and G628 residues, respectively, while sgRNA#3 and sgRNA#4 were utilized to guide OsGTBE to target the P171 and G628 residues, respectively. For the comparison of editing activity between OsGTBE and OsCGBE03, approximately 50 hygromycin-resistant calli from each transformation were pooled separately for DNA extraction and subjected to genotyping through next-generation sequencing (NGS). Analysis of a total of 310,349 NGS reads revealed that OsCGBE03 could induce both C-to-T and C-to-G base editing within C4-C10, with up to 13.0% base editing efficiencies at sgRNA#1 site and 25.2% at sgRNA#2 site (Figure 1C, D). The C-to-G editing purities reached 29.7% and 24.0% at sgRNA#1 and sgRNA#2 (Figure S2), respectively. Differing from OsCGBE03, the OsGTBE targeting the opposite strands of OsCGBE03 targets predominantly generated G-to-T base conversion within G4-G11, with up to 10.3% and 21.9% base editing efficiencies at sgRNA#3 and sgRNA#4 sites, respectively (Figure 1C, D). Notably, OsGTBE exhibited the highest editing efficiencies at the G6 site of sgRNA#3 (7.02%) and the G8 site of sgRNA#4 (13.41%). The G-to-T editing purities reached 98.4% and 86.4% at sgRNA#3 and sgRNA#4 (Figure S2), respectively. As previously reported, mutations at the P171 and G628 residues of rice OsALS1 confer tolerance to sulfonylurea (SU) and imidazolinone (IMI) herbicides, respectively (Zhang et al., 2021; Wang et al., 2023). Editing these two positions with OsCGBE03 generated several known herbicide-resistance alleles (Table S1), that is, P171S, P171F, P171A, G628D, and G628E (Zhang et al., 2021). In contrast to conventional CBE and CGBE, the OsGTBE exhibited a distinct editing profile, generating over a dozen novel missense mutations at these two sites (Figure 1E), including P171T, P171H, P171N, P171R, G628W, G628V, and G628L. These results demonstrated the potential of OsGTBE in creating genetic diversity. Complementing conventional CBE and CGBE, it greatly expands the application prospects of base editing tools in saturation mutagenesis of amino acids. To evaluate the performance of OsGTBE in stable-transgenic rice, 86 and 23 T0 rice plants were regenerated from the positive calli transformed with OsGTBE-sgRNA#3 and OsGTBE-sgRNA#4 (Table S2), respectively. Sanger sequencing results for these T0 plants revealed that most T0 plantlets exhibited superimposed sequencing chromatograms (Figure 1F). To better illustrate the editing outcomes of OsGTBE in stable-transgenic rice, we re-analyzed these T0 plantlets using HiTOM. The valid edited plants were characterized by HiTOM assays with a 10% threshold to ensure their inheritability. These results indicated that up to 33.7% (29/86) and 39.1% (9/23) of T0 plants contained G-to-T base editing at sgRNA#3 and sgRNA#4 sites, respectively. Notably, G-to-C base conversions were also detected at these two sites, with editing efficiencies of 1.2% (1/86) at sgRNA#3 and 8.7% (2/23) at sgRNA#4 (Table S2). Furthermore, consistent with Sanger sequencing results, a total of 32 (29.4%) T0 plants harbored indels (Table S2), primarily characterized by the deletion and/or duplication of sequences between the targeted Gs and the Cas9 cleavage site (Figure 1F). To further validate the utility of OsGTBE in rice, we designed three additional sgRNAs, that is, sgRNA#5, sgRNA#6, and sgRNA#7, to assess the activity of OsGTBE at the OsALS1-W548, OsACC1-W2125/R2126 and OsACC1-G2194 sites (Figures 1G, S3), respectively. A total of 42, 81, and 31 T0 plants were obtained, respectively (Table S2). Genotyping results revealed that OsGTBE successfully induced G-to-T conversion (chimerism rate >10%) at all three targets, with frequencies of 16.7%, 21%, and 3.2% (Figure 1H), respectively. Different from the editing outcomes at the sgRNA#3 and sgRNA#4 sites, there was no detectable G-to-C conversion at these three sites, but one plant (sgRNA#5-No.29) contained G-to-A conversion at the sgRNA#5 site (Figure S3A). In total, genotyping of 263 T0 rice plants at these five targets revealed that OsGTBE induced an average of 25.1% (66/263) guanine base editing with a wide editing window of G2–G11 (Figure 1I). Even though 51.5% (34/66) of the base-edited plants were chimeric, homozygous/biallelic base conversions were also found at sgRNA#3 (6.9%, 2/29), sgRNA#5 (12.5%, 1/8), and sgRNA#6 (17.6%, 3/17) sites (Figure 1J; Table S3). Compared to conventional deaminase-based CBEs (Li et al., 2023a), OsGTBE exhibited a slightly broader editing window and markedly different editing characteristics. In rice, OsGTBE induced targeted G-to-T editing, addressing the limitations of traditional CBE or CGBE in effectively inducing G:C-to-T:A mutations. Combining GTBE with CBE and CGBE, the G:C base-pair could be converted to any desired base-pair (Figure 1K). Considering the missense mutations induced by OsGTBE at conserved residues (OsALS1-P171, OsALS1-G628/G629, and OsACC1-W2125/R2126) responsible for herbicide resistance (Zhang et al., 2021; Wang et al., 2022), T0 plants with desired base edits were selected for herbicide tolerance assays (Figure S4). As shown in Figure 1L, the plants containing P171T, P171N, P171H, or P171R survived 0.2 g(ai)/L flucarbazone-sodium treatment, suggesting that these amino acid substitutions confer tolerance to SU herbicides. Consistent with previous findings (Wang et al., 2022), G628W conferred high tolerance to imazethapyr. However, plants harboring novel alleles G628V and G629C generated in this study did not survive 0.3 g(ai)/L imazethapyr treatment. At the sgRNA#6 target of OsACC1 (Os05g0295300), plants with W2125L, W2125C, and W2125L/R2126I mutations survived exposure to 0.2 g(ai)/L haloxyfop-P-methyl, indicating that these novel alleles could confer tolerance to aryloxy-phenoxy-propionate (APP) herbicides (Xu et al., 2021). Overall, these results fully demonstrate the feasibility of our OsGTBE system in creating useful genetic diversity. Introducing these novel herbicide-resistant alleles in elite rice cultivars through OsGTBE would be a simple and rapid strategy, paving the way for potential agricultural applications in the future. In general, base editors remain the most efficient and user-friendly genome editing tools to convert specific nucleotides without donor templates, allowing efficient saturation mutagenesis. However, existing base editors are still incapable of achieving all possible base conversions in plants (Li et al., 2023a). In this study, we constructed a novel deaminase-free base editor (OsGTBE) that directly excises the guanine with the engineered glycosylase (RgMPG). The rice intrinsic TLS system appears to preferentially pair As opposite abasic sites, thereby inducing targeted G-to-T transversion (Figure 1I). This unique mechanism achieves desired G:C-to-T:A base substitutions by converting the Gs opposite the desired Cs (Figure 1K). By employing OsGTBE to target two herbicide-resistance genes OsALS1 and OsACC1, we successfully generated several novel herbicide-resistance alleles (Figure 1L). These findings fully illustrate the promising application value of OsGTBE in creating genetic diversity and generating germplasm for breeding. This study was supported by the National Key R&D Program of China (2021YFA1300404 to J.-K.Z.) and the National Natural Science Foundation of China (32188102 to J.-K.Z.), the China Postdoctoral Science Foundation (BX20220098 and 2022M720973 to Y.T.), the Hainan Seed Industry Laboratory (B22C1000P to Y.T.), and Nanfan special project, CAAS (ZDXM2314 to M.W.). The authors declare no competing interests. Y.T. and J.-K.Z. designed the research; C.D., X.L., J.X., Z.Z., and R.S. performed the rice transformation; Y.T., C.D., and M.W. genotyped the plants and conducted the herbicide tolerance assay; Y.T., J.X., and X.C analyzed the data; Y.T. and J.-K.Z. wrote the manuscript. Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.13657/suppinfo Figure S1. The experimental procedures for evaluating G:C-to-Y:R base conversion. Figure S2. Editing profiles of OsCGBE03 and OsGTBE in rice calli. Figure S3. Representative Sanger sequencing chromatograms of selected T0 plants. Figure S4. Genotyping results of the representative T0 plants that survived herbicide treatment. Sequence S1. Amino acid sequence of the OsGTBE. Table S1. Amino acid substitutions at OsALS1-P171/G628/G629. Table S2. Summary of base editing outcomes in rice T0 plantlets. Table S3. Summary of the allele patterns of base-edited plants. Table S4. Primers and oligos used in this study Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.