Fangquan Wang, Yang Xu, Wenqi Li, Zhihui Chen, Jun Wang, Fangjun Fan, Yajun Tao, Yanjie Jiang, Qian-Hao Zhu, Jie Yang,e,

aInstitute of Food Crops,Jiangsu Academy of Agricultural Sciences/Nanjing Branch of Chinese National Center for Rice Improvement,Nanjing 210014,Jiangsu, China

bJiangsu Co-Innovation Center for Modern Production Technology of Grain Crops,Yangzhou University,Yangzhou 225009,Jiangsu,China

cProvincial Key Laboratory of Agrobiology, Jiangsu Academy of Agricultural Sciences,Nanjing 210014,Jiangsu,China

dCSIRO Agriculture and Food,GPO Box 1700,Canberra,ACT 2601,Australia

eInstitute of Life Science,Jiangsu University,Zhenjiang 212013,Jiangsu,China

Keywords:

ABSTRACT Weeds and weedy rice plague commercial rice fields in many countries.Developing herbicide-tolerance rice is the most efficient strategy to control weed proliferation.CRISPR/Cas9-mediated gene editing, which generates small InDels and nucleotide substitutions at and around target sites using error-prone non-homologous end joining DNA repairing, has been widely adopted for generation of novel crop germplasm with a wide range of genetic variation in important agronomic traits.We created a novel herbicide-tolerance allele in rice by targeting the acetolactate synthase (OsALS) gene using CRISPR/Cas9-mediated gene editing.The novel allele(G628W)arose from a G-to-T transversion at position 1882 of OsALS and conferred a high level of herbicide tolerance.Transgene-free progeny carrying homozygous G628W allele were identified and showed agronomic performance similar to that of wild-type plants, suggesting that the G628W allele is a valuable resource for developing elite rice varieties with strong herbicide tolerance.To promote use of the G628W allele and to accelerate introgression and/or pyramiding of the G628W allele with other elite alleles, we developed a DNA marker for the G628W allele that accurately and robustly distinguished homozygous from heterozygous segregants.Our result further demonstrates the feasibility of CRISPR/Cas9-mediated gene editing in creating novel genetic variation for crop breeding.

1.Introduction

Direct seeding of rice is favored by farmers in many countries for its saving in labor and money over transplanting.However,weedy rice, which is taxonomically closely related to and physiologically similar to cultivated rice, spreads in directseeded paddy fields, resulting in delays in growth and loss of grain yield and quality [1-4].Development of herbicidetolerance rice and corresponding breeding methods to keep abreast of newly released herbicides is one of the promising strategies for facing the challenge of weedy rice[5].

Herbicides targeting acetolactate synthase (ALS, also known as acetohydroxyacid synthase,AHAS),which catalyzes the initial step of the biosynthesis of branched-chain amino acids such as valine, leucine, and isoleucine, are widely applied for crop protection [6,7].Imidazolinone (IMI) herbicides are a class of high-efficiency, broad-spectrum and lowtoxicity herbicides targeting ALS.IMI herbicides effectively control annual grass weeds and broadleaf weeds in soybean fields [8,9], but are phytotoxic to many crops, including rice,maize and wheat, greatly limiting their use in protection of these important staple crops.Breeding crops tolerant to IMI herbicides can reduce their damage to crops and broaden their use.

Mutations in the ALS gene confer plant tolerance to IMI herbicides by disabling the interaction between the ALS protein and IMI herbicides [10,11].Several natural rice ALS mutants tolerant to IMI have been found and herbicidetolerance rice mutants have also been created by artificial mutation and selection of the ALS gene [12-14].Rice plants containing mutation at amino acids S627 or G628 of OsALS have been shown to be tolerant to IMI [12,13].The level of herbicide tolerance corresponds to the number and location of mutations in the ALS gene [2].Creation of new mutations in the ALS gene will enrich the genetic diversity of herbicidetolerance ALS alleles and provide genetic materials for breeding novel rice germplasm strongly tolerant to herbicides.

Genome-editing technologies provide an efficient method for creating nucleotide mutations in target genes.Homologydirected repair (HDR) and targeted base editing with cytidine base editors(CBEs)and adenine base editors(ABEs)have been widely used in precision plant breeding, including for herbicide-resistant crops [15-18].For example, herbicidetolerance rice mutants have been generated via CRISPR/Cas9-mediated homologous recombination to simultaneously replace two amino acids (W548L and S627I) in ALS [19].Targeted base editing (C to T) in the rice ALS gene mutated its 627th amino acid from serine(S)to asparagine(N)to create a herbicide-tolerance rice mutant [20].These strategies enabled precise gene replacement or target site editing to obtain a desired allele.In contrast, most editing events created in plants by CRISPR/Cas9 are generated by an errorprone non-homologous end-joining (NHEJ) DNA repair process.It can produce small InDels and/or nucleotide substitution(s), such as G to T, that may not be easily generated by HDR and base-editing.

The objective of this study was to generate an herbicidetolerance rice mutant by mutating the rice ALS gene (OsALS)using the NHEJ-dominant CRISPR/Cas9 approach.

2.Materials and methods

2.1.Selection of a sequence targeting the OsALS gene and offtarget analysis

Nangeng 9108 (Oryza sativa L.japonica), the target variety of this study,is a rice cultivar with good eating quality and high yield that is widely planted in Jiangsu province,China.A guide RNA (5′-TCCTTGAATGCGCCCCCACT-3′) tiled on the reverse DNA strand with a protospacer-adjacent motif (PAM) of TGG(Fig.1A) was designed using CRISPR-GE (http://skl.scau.edu.cn/targetdesign/) [21].The expected Cas9 cleavage site was located between positions 1881(part of the 627th amino acid)and 1882 (part of the 628th amino acid) of OsALS.The predicted off-target loci based on CRISPR-GE (OT1, OT2, OT3,OT4, and OT5) with an off-score greater than 0.04 were amplified with the locus-specific-primers ALS-L-OT1-F/ALSL-OT1-R, ALS-L-OT2-F/ALS-L-OT2-R, ALS-L-OT3-F/ALS-L-OT3-R, ALS-L-OT4-F/ALS-L-OT4-R, and ALS-L-OT5-F/ALS-L-OT5-R,and examined by sequencing.Genomic DNA was extracted from Nangeng 9108 leaves.The OsALS gene was amplified using the primers ALS5-F: 5′-TCGCCCAAACCCAGAAACCC-3′and ALS5-R:5′-CTCTTTATGGGTCATTCAGGTC-3′,and verified by sequencing.PCR products were sequenced by TsingKe Biological Technology (Nanjing, China).

2.2.Vector construction

The CRISRP/Cas9-ALS transformation vector was generated following an established method[22].The target adaptor was prepared by natural annealing of primer pair ALS-U3-F/ALSU3-R (Table S1).A restriction-ligation reaction (10 μL) was prepared with 1×DNA ligase buffer, 15 U of T4 DNA ligase(Takara, Dalian, China), 20 ng of the pYLsgRNA-OsU3 vector,2.5 U of Bsa I, and 0.05 μmol L?1of the target adaptor.The reactions were incubated for 5 cycles (37 °C, 5 min; 20 °C 5 min).The sgRNA-ALS expression cassette driven by the OsU3 promoter was synthesized using overlapping PCR with the restriction-ligation reaction product as template.The first PCR was performed using primer set U-F/gRNA-R.Products of the first PCR were then used as templates for the second PCR reaction with the primer pair Uctcg-B1′/gRcggt-BL to generate the sgRNA-ALS expression cassette.The sgRNA-ALS cassette was then cloned into the CRISPR/Cas9-encoding scaffold with restriction and ligation.The restriction-ligation reaction(15 μL) was performed with 1× DNA ligase buffer, 15 U of T4 DNA ligase (Takara), 10 U of Bsa I, 60 ng of the intact binary vector pYLCRISPR/Cas9-MH, 2.5 U of Bsa I, and 20 ng of the purified product of the second PCR reaction.The ligated products were transformed into E.coli competent cells(DH 5α)using the heat shock method.Colonies grown on an LB-agar medium plate containing 50 mg L?1kanamycin were tested by PCR and those containing CRISRP/Cas9-ALS were selected based on sequencing verification(Fig.1B).

2.3.Rice transformation

The CRISRP/Cas9-ALS construct was introduced into Agrobacterium tumefaciens strain EHA105.Bacteria containing the CRISRP/Cas9-ALS vector were transformed into Nangeng 9108 to generate T0plants for identification of imazethapyr(IMT)-tolerance mutants.Agrobacterium-mediated transformation of rice calli was performed as previously described[23].

2.4.Mutation detection in edited plants

T0plants were obtained from medium plates containing 50 mg L?1hygromycin.The T0plants were transferred to and grown in a growth chamber and T1seeds were harvested for further analysis.Rice genomic DNA from leaf tissue was extracted using a DNA Quick Plant System (Tiangen, Beijing,China).PCRs were performed using primer pairs ALS-1F/ALS-1R matching the flanking regions of the designed target site to amply sequence fragments.The PCR products were sequenced for detection of gene editing.Sequence chromatograms were decoded using the DSDecode Web-based Tool(http://skl.scau.edu.cn/dsdecode/)[21].

2.5.Transgene-free detection

The presence or absence of the transfer DNA (T-DNA)containing CRISPR-associated 9 (Cas9), sequence-specific single guide RNA (sgRNA), and hygromycin phosphotransferase (HPT) in T1progeny was determined based on gene-specific PCRs.The primer pairs Cas9T-F/Cas9T-R, pYLU3-F/pYLsgRNA-R and hyg283-F/hyg283-R were used to amplify the Cas9, sgRNA,and HPT genes,respectively(Table S1).Plants were considered to be free of transgenes when amplicons were concurrently lacking for all three genes.

2.6.Herbicide spray assays

IMT at the recommended concentration of 70 (1×) g active ingredient per hectare (a.i.ha?1) and 210 (3×) g a.i.ha?1, and imazapic(IMP)at the recommended concentration of 240(1×)g a.i.ha?1were used in an herbicide spray assay to test T1and T3progeny.For T1progeny,two-week-old rice seedlings were sprayed with IMT suspension with a concentration at 210(3×)g a.i.ha?1using a nebulizer at 0.3 MPa pressure.For T3progeny, two-week-old rice seedlings were sprayed with IMT suspension at a concentration of 70 (1×), 700 (10×), or 1400(20×) g a.i.ha?1or with imazapic suspension at a concentration of 240(1×),2400(10×),or 4800(20×)g a.i.ha?1.Phenotypes were recorded 14 days after herbicide application.Wild-type(WT) plants were included as controls.These experiments were repeated three times.

2.7.Design of allele-specific primers for marker-assisted selection of herbicide-tolerance progeny with the G628W allele

The nonsynonymous nucleotide substitution (G to T) that changes the 628th amino acid of ALS from glycine (G) to tryptophan (W) in the herbicide-tolerance rice mutant is at position 1882 of OsALS.To distinguish mutant (herbicidetolerance) from WT (herbicide-susceptible) plants and to facilitate introgression of the herbicide-tolerance OsALS allele(G628W), we designed two pairs of primers that shared a common forward primer (ALS-1F) and had a different allelespecific reverse primer (ALS-4R or ALS-6R) with the newly created single-nucleotide polymorphism at position 1882 of OsALS as their last base.To increase the specificity of the reverse primers, we introduced artificial mismatches in ALS-4R(C to A)and ALS-6R(C to T)at the third position from the 3′end.The artificial nucleotide substitution eliminated background fragments that led to ambiguous genotyping results.The size of the PCR product amplified by ALS-1F/ALS-4R and ALS-1F/ALS-6R is 329 bp.

The PCR amplification was performed in a reaction volume of 15 μL, consisting of 1.5 μL PCR buffer (10×),1.5 μL MgCl2, 0.15 μL Taq DNA polymerase, 1.5 μL forward primer, 1.5 μL reverse primer, 1.5 μL genomic DNA, and 7.35 μL ddH2O.The PCR conditions were: 94 °C, 5 min;32 cycles of 94 °C, 30 s; 60 °C 30 s; 72 °C, 30 s; and a final extension of 2 min at 72 °C.The PCR products were separated in 2% agarose gel and photographed using a gel imager system.

3.Results and discussion

3.1.CRISPR/Cas9-mediated editing of OsALS

In total,58 independent T0lines were obtained.To identify gene editing events,a sequence fragment containing the target site was amplified from 25 T0lines using the primer set ALS-1F/ALS-1R and sequenced.Sequencing revealed eight edited plants,for an editing efficiency of 32%(8/25),including five(20%)heterozygous plants,one(4%)biallelic plant,and two(8%)mosaic plants(Table S2).Of the five edited heterozygous plants,A3,A15 and A18(designated WT/-2)had 2-bp deletions at positions 1882 and 1883,A9(WT/-1)had a 1-bp deletion at position 1882,and A24(WT/-2)had a 2-bp deletion at positions 1881 and 1882.The biallelic plant (A51)contained a G-to-T transversion and a 1-bp deletion(G ＞T/-1)at position 1882(Fig.2A).Interestingly,none of the 25 T0lines was a homozygous frameshift mutant,suggesting that no homozygous frameshift editing event had occurred or that such an event is lethal,implying an essential role of ALS in plant survival.

3.2.A novel rice OsALS allele conferring herbicide tolerance

T1seeds were harvested from 58 independent selfed T0lines.Two-week-old T1seedlings from each T0line and WT seedlings were sprayed with IMT suspension with a concentration of 210(3×)g a.i.ha?1to screen mutants for resistance to herbicide.After 14 days, all 18 T1seedlings from T0-A51 survived, whereas all progeny from other T0lines and WT died(Fig.2B).T0-A51(G ＞T/-1)was a biallelically edited plant.Sequencing the edited site in the 18 surviving T1progeny showed that 11 were homozygous G ＞T and 7 were biallelic G ＞T transversion and 1-bp deletion.No progeny showed a homozygous 1-bp deletion, suggesting that the 1-bp deletion allele found in the T0-A51 could not be transmitted to the next generation in homozygous form (Fig.2D).In fact, to date, no functional ALS knockout has been reported [16,17].We accordingly speculate that knockout of ALS is lethal to plants.The nucleotide change of G ＞T at position 1882 converted the 628th amino acid of OsALS from glycine(G)to tryptophan(W)to create an allele hereafter designated G628W,while the 1-bp deletion at position 1882 caused a frameshift and resulted in premature translation termination(Fig.2C).The G628W allele has not been reported previously, indicating that we have generated a novel herbicide-tolerance ALS allele in plants.

Fig.2– Identification of an herbicide-tolerance mutant carrying a novel OsALS allele.A.The T0 plants with mutated(A3,A9,A24,and A51)or unmutated(A5)OsALS.The PAM and target sequences are highlighted in green and bold,respectively.Changed nucleotide(s)in the putative editing window are shown in blue with deletion(s)indicated by“-”.WT,wild-type.B.Herbicide selection of T1 plants using an imazethapyr(IMT)concentration of 210 g(a.i.)ha?1.C.Chromatograms showing the mutated OsALS alleles in two representative IMT-tolerance T1 plants.The amino acid changes caused by nonsynonymous nucleotide substitution and deletion are highlighted in blue and yellow,respectively.FM,frameshift mutation.D.The genotype of gene editing and T-DNA insertion of the IMT-tolerance T1 progeny derived from the T0 plant A51.R, resistance.

In transgene-based crop breeding, it is vital for the final products to carry the desired trait while being transgene-free.Transgene-free products would be free of the risk management mandatory for genetically modified products and would be more acceptable by the general public.We accordingly investigated the presence and absence of T-DNA in the surviving T1progeny of T0-A51.Three sets of primers specific to HPT,Cas9 or sgRNA were used to check for their presence in the surviving plants.Of the 18 T1progeny,three were found to be transgene-free, including one (T1-A51-1) homozygous for the G628W allele and two (T1-A51-12 and T1-A51-18)heterozygous (G ＞T/-1) (Fig.2D, Fig.S1).In T1-A51-1 we found no editing or nucleotide change in any of the five predicted off-target loci (OT1, OT2, OT3, OT4, and OT5)(Fig.S2).

Fig.3– Herbicide-tolerance assay of T3 progeny carrying the G628W allele.Two-week-old rice seedlings were sprayed with imazethapyr(IMT)suspension at concentrations of 70,700,or 1400 g(a.i.)ha?1(A)or imazapic(IMP)suspension at concentrations of 240,2400,or 4800 g(a.i.)ha?1(B).Phenotypes were recorded 21 days after herbicide application.WT,wild-type.

Fig.4–Agronomic traits of T4 progeny carrying the G628W allele.Plant height(A),tiller number per plant(B),fertility(C),1000-kernel weight(D),and morphology(E)of T4 progeny carrying G628W and wild-type(WT)are displayed.The plants were grown under natural light conditions in a greenhouse in Nanjing.Values are mean± SD of ten biological replicates.Means were compared by Student’s t-test.***,P＜0.001;non-significant(NS), P ＞0.05.

Use of the G628W allele in agricultural production depends largely on the strength of its herbicide tolerance.We treated transgene-free T3progeny homozygous for the G628W allele with varying doses of IMT and IMP.The IMT concentrations used were 70, 700, and 1400 g a.i.ha?1, and the IMP concentrations used were 240, 2400, and 4800 g a.i.ha?1.After 14 days, all WT plants died at all concentrations,whereas G628W plants survived at 1400 g a.i.ha?1IMT and 2400 g a.i.ha?1IMP.Thus, the G628W plants showed strong tolerance to both IMT and IMP(Fig.3).

3.3.Agronomic trait analysis

To evaluate the effect of the G628W allele on agronomic traits,we grew T4progeny carrying the homozygous G628W allele under natural growth conditions in the greenhouse in Nanjing,China.No significant differences in agronomic traits,including morphology, tiller number per plant, fertility, and 1000-kernel weight were observed between the G628W plants and WT, although the height of the G628W plants was reduced(by about 7 cm)compared with that of WT(Fig.4).

3.4.Development and application of a DNA marker for introgression of the G628W allele

To facilitate introgression of the G628W allele into elite rice cultivars, particularly those tolerant to certain herbicides,marker-assisted selection would be essential.We accordingly developed an allele-specific PCR (AS-PCR) marker for the G628W allele.The marker consists of two allele-specific reverse primers (ALS-4R and ALS-6R for the WT and the G628W allele, respectively) and a single common forward primer (ALS-1F) (Fig.5A).We demonstrated the usefulness of the marker in two populations, one with 26 rice cultivars or lines that are herbicide-sensitive and another is an F2population segregating for herbicide tolerance.As shown in Fig.5B, in all 26 herbicide-sensitive rice cultivars or lines, a single PCR product was amplified only by the primer set ALS-1F/ALS-4R and not by the primer set ALS-1F/ALS-6R.In contrast, the G628W allele was amplified only by ALS-1F/ALS-6R and not by ALS-1F/ALS-4R (Fig.5B).

Fig.5– Development of allele-specific PCR markers for detection of the G628W allele.

(A) Allele-specific primers for detection of wild-type and G628W alleles.The last nucleotide of ALS-4R and ALS-6R matches the nucleotide at the position where the nonsynonymous nucleotide substitution occurred in plants with the G628W allele.The underlined nucleotides in ALS-4R and ALS-6R are artificially added mismatches to ensure the specificity of the primers.WT, wild-type.(B) PCR amplification results of selected rice cultivars or lines using the AS-PCR marker.(C)Genotyping and phenotyping results of 19 randomly selected F2progeny derived from Xudao 9×G628W.R, IMT tolerance,S,IMT sensitivity.(D)Segregation of the IMT-tolerance phenotype in the Xudao 9×G628W F2population.

The F2population segregating for herbicide tolerance was derived from crossing Xudao 9 (herbicide sensitive) with plants homozygous for the G628W allele.Xudao 9 is a rice cultivar with good eating quality that is widely cultivated in the Huang-Huai area of China.Of 132 F2progeny that were genotyped and phenotyped, respectively, 102 and 30 were tolerant and sensitive to IMT (Fig.5D).The 102 IMT-tolerance F2progeny were homozygous (32 plants) or heterozygous (70 plants) for the G628W allele, whereas all 30 IMT-sensitive plants were found to be homozygous for the WT allele(Fig.5C and D).The AS-PCR genotyping results were fully consistent with the phenotyping outcome, indicating the specificity and robustness of the AS-PCR marker for breeding applications.In addition, the AS-PCR marker accurately distinguished homozygous from heterozygous G628W alleles.

4.Conclusions

We generated a novel OsALS allele (G628W) conferring high herbicide tolerance using CRISPR/Cas9-mediated gene editing,enriching the genetic diversity of herbicide-tolerance ALS alleles.Transgene-free progeny with homologous G628W alleles showed agronomic characteristics similar to those of WT plants except for the reduced plant height.Given that shorter plants are more resistant to lodging, so we expect that the G628W allele would be valuable for breeding elite rice varieties with a high level of herbicide tolerance.We also developed a robust DNA marker for G628W allele pyramiding in rice breeding.

Author contributions

Jie Yang, Fangquan Wang designed the research; Fangquan Wang, Yang Xu, Wenqi Li, Zhihui Chen, Jun Wang, Fangjun Fan, Yajun Tao, and Yanjie Jiang performed the research;Fangquan Wang analyzed the data; and Fangquan Wang,Qianhao Zhu,and Jie Yang wrote the paper.

Declaration of competing interest

The authors have submitted a patent application based on the results reported in this paper.

Acknowledgments

This study was supported by the National Transgenic Science and Technology Program (2018ZX08001-02B), the Jiangsu Agricultural Science and Technology Innovation Fund (CX(19) 3059), and the Jiangsu Province Key Research and Development Program(Modern Agriculture, BE2017345-2).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.06.001.