Lei Chen ,Gng Gu ,Chengxu Wng ,Zhufeng Chen ,Wei Yn ,Mn Jin ,Gng Xie ,Junli Zhou ,Xing Wng Deng ,c,*,Xioyn Tng ,,*

a Guangdong Provincial Key Laboratory of Biotechnology for Plant Development,School of Life Sciences,South China Normal University,Guangzhou 510631,Guangdong,China

b Shenzhen Institute of Molecular Crop Design,Shenzhen 518107,Guangdong,China

c Institute of Plant and Food Sciences,Department of Biology,Southern University of Science and Technology,Shenzhen 518055,Guangdong,China

ABSTRACT Herbicide resistance in crop plants is valuable for integrated weed management in agriculture.Herbicide resistant rice,in particular,is important to management of weedy rice,a close relative of cultivated rice and a noxious weed prevalent in rice fields that remains challenging to farmers worldwide.Herbicide resistant plants can be obtained through transgenic approach or by mutagenesis of regular plant and screening of mutants with elevated resistance to herbicide.In this study,we conducted ethyl methyl sulfonate mutagenesis (EMS) to elite indica cultivar Huanghuazhan (HHZ) and screened for mutants resistant to imazapic,a herbicide that can inhibit the acetolactate synthase (ALS) in plants.We obtained three mutants of OsALS gene that have not been reported previously in rice.One of the mutants,with Trp548 changed to Met (W548M),was analyzed in more details in this study.This mutation had no negative effect on the plant physiology and morphology as well as rice yield.Compared with the imidazolinone-resistant mutant S627N (Ser627 changed to Asn) that has been deployed for Clearfield rice development,W548M mutant showed high levels of resistance to a broad spectrum of five families of ALSinhibiting herbicides,in addition to a higher level of resistance to herbicides of the imidazolinone family.The herbicide-resistance was stably inherited by crossing into other rice lines.Thus,the W548M mutation provides a valuable resource for breeding of herbicide resistant rice and weed management.

Keywords:ALS-inhibiting herbicide Herbicide tolerance Acetolactate synthase Rice Mutant

1.Introduction

Weedy rice is one of the most destructive weeds that is also most difficult to control in the rice growing fields worldwide,particular the direct-sowing fields [1].According to the definition of the United Nations Food and Agriculture Organization (FAO),weedy rice refers to all self-reproducing plants of theOryzagenus that invade farmland ecosystems and compete with cultivated rice for resources and water,including wild rice,the hybrid offspring of cultivated rice and wild rice,and the shattered cultivated rice(Oryza sativa) [2].Almost all rice planting areas are affected by weedy rice,with varying degrees of impact.The total annual loss of rice production caused by weedy rice in the United States amounts to more than 50 million US dollars [3].In Europe,～65%of rice fields are infested by weedy rice,while in Malaysia,weedy rice has been reported to reduce rice production by 75% [3,4].Weedy rice not only reduces the yield but also the quality of cultivated rice.Because of the long dormancy of weedy rice seeds under natural conditions,rice field initially infested by a few weedy rice seeds often incurs serious infestations after a few growing seasons and eventually becomes yield-less [1].

As labor costs in China increase and labor flows from rural areas to cities,rice farming in China has gradually changed from intense and meticulous transplanting to direct-seeding cultivation.Consequently,the problem of weedy rice has become more and more serious.According to incomplete statistics,more than 3 Mha of rice fields across the country have been infested by weedy rice [5].Jiangsu and Guangdong provinces in China are areas suffering from serious weedy rice damages [5].

Herbicides need to selectively kill weeds without harming the crop before they can be used in rice farming.However,weedy rice is a close relative of cultivated rice,and their toxicological and physiological characteristics are very similar.Therefore,there is no herbicide that can effectively control weedy rice in paddy fields[6].As of January 2020,there were 2293 herbicide formulation products registered for rice field in China,according to the database online of China Pesticide Information Network (http://www.chinapesticide.org.cn/),but none of them can effectively suppress the spread of weedy rice in China.

Genetic alteration of cultivated rice can enhance its tolerance to certain herbicides,so that the corresponding herbicides can selectively kill weedy rice without harming the cultivated rice.At present,researchers have obtained various herbicide-resistant rice lines through transgenics,gene editing,and mutant screening.For example,Liberty Link rice (glufosinate-resistant) was generated by transforming aStreptomyces hygroscopicusgene encoding phospinothricin acetyl transferase (PAT) into rice [7].Glyphosateresistant rice was obtained by gene editing of the rice endogenous gene 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) [8].Acetyl coenzyme-a carboxylase (ACCase) inhibitor-resistant rice was obtained by both mutagenesis/mutant screening and gene editing technology [9].Rice resistant to the protoporphyrinogen IX oxidase (PPX) inhibitor herbicide was obtained by mutagenesis and mutant screening [10].Hydroxyphenylpyruvate dioxygenase(HPPD)herbicide-resistant rice was obtained from naturally occurring resources [11].Commercialized herbicide-resistant rices include Clearfield (imidazolinone-resistant) [12] and Provisia(quizalofop-p-ethyl-resistant) [13].By far,the most widely used herbicide-resistant rice is the imidazolinone-resistant rice Clearfield.

Acetolactate synthase(ALS) is a key enzyme in the synthesis of branched chain amino acids such as valine,leucine,and isoleucine[14].The active enzyme complex consists of two identical catalytic subunits and one small regulatory subunit [15].Inhibition of the ALS enzymatic activity hinders the synthesis of branched chain amino acids,in turn affecting protein synthesis and inhibiting plant growth[16].Therefore,ALS is an important herbicide target.ALS-inhibiting herbicides are divided into five families according to their molecular structure,including sulfonylureas (SU),imidazolinones(IMI),pyrimidinylthiobenzoates(PTB),sulfonylaminocarbonyltriazolinones (SCT),and triazolopyrimidines (TP) [17].These herbicides bind to distinct yet overlapping amino acids in the herbicide docking pocket formed by the interface between the two catalytic subunits [15].Mutations at different sites on the ALS catalytic subunit can cause the plants to develop different levels of resistance to different types of ALS-inhibiting herbicides[18].Researchers in the United States have discovered two mutations,Gly628Glu (G628E) and Ser627Asn(S627N),in rice ALS that can make rice acquire IMI herbicide resistance [19,20].Then,BASF took the lead in applying IMI herbicide-resistant rice complemented with IMI herbicide to control weedy rice,and they named it Clearfield rice [1,6,12].Because S627N exhibited～5-fold higher resistance compared with G628E,S627N is more widely used in Clearfield rice at present [1].

Although the application of S627N mutant significantly improved weedy rice management,there were still problems such as field safety and narrow herbicide spectrum.We therefore conducted EMS mutagenesis to an eliteindicacultivar in China and screened for new mutations resistant to IMI herbicides.We obtained a total of 25 mutant plants.Sequencing analysis revealed three new mutations that have not been reported previously.One of the new mutations,W548M,was analyzed in more details in this study.Compared with S627N mutation in rice,the W548M mutant has broad-spectrum resistance to other ALS inhibitors in addition to the IMI herbicides,and the level of resistance has been greatly improved without affecting the yield.The W548M mutant provides thus a valuable resource for herbicide-resistant rice breeding and weedy rice management.

2.Materials and methods

2.1.Mutagenesis and mutant screening

Theindicarice variety Huanghuazhan(HHZ),a core germplasm for rice breeding in Guangdong province,China,was used as material in this study.This variety has the advantages of high and stable yield,excellent grain quality,good stress tolerance,and wide adaptability[21].For mutagenesis,50 kg of HHZ seeds were firstly soaked in water for 22 h at 28°C and then drained to clean the surface water.The seeds were then soaked in 0.7% EMS solution at 28°C for 12 h to form the mutant library of M1generation.The germination rate of M1seeds after EMS-treatment was 63%,and the albino rate at seedling stage was 3.5%.M1seedlings were thrown-planted at planting density of～36 plants m-2with regular field management.When the seeds were yellow ripe,plants in every 30 m2were mix-harvested and stored separately as a pool of the M2mutant library.For mutant screening,～1 kg of seeds from each pool of the M2mutant library were directly sowed in separate plot,and the plants at 3–4-leaf stage were uniformly sprayed with imazapic solution at a dose of 108 g a.i.ha-1.The plant phenotype was evaluated 15 days after spray,and the plant with a green healthy heart leaf was regarded as the resistant plant.Another mutant screening method was to soak the M2seeds in 0.02% (w/v) imazapic solution for 24 h at 25 °C.The seeds were drained,raised with water,and then germinated on wet paper at 25 °C.Sprouts turning green during elongation were regarded as resistant candidate.

2.2.Genetic analysis and cleaning of other mutations irrelevant to herbicide resistance in the herbicide-resistant mutants

Backcross of mutant with wild type HHZ was conducted to clean other EMS-induced mutations in order to eliminate the effect of these mutations on growth and development of the herbicide resistant mutant.We used the herbicide resistant mutants KH-9 and KS-2 as the male parents,wild-type HHZ as the recurrent female parent,and backcrossed to BC3F1generation.Each generation was sprayed with 108 g a.i.ha-1imazapic to remove the non-resistant plants.The BC3F1plants were self-pollinated to obtain the BC3F2seeds.The BC3F2plants in the field were sprayed with 108 g a.i.ha-1imazapic 30 days after sowing,and the number of surviving and dead plants was counted 20 days after spraying.Chi-square test was performed to calculate the segregation ratio of surviving and dead plants to determine the genetic nature of herbicide resistance.

2.3.Identification of the causal mutation in KH-9 mutant

To identify the causal mutation,30 M3plants derived from the homozygous KH-9 mutant were selected for genomic DNA extraction.Equal amount of DNA were mixed together for bulked DNA re-sequencing on the Illumina HiSeq platform.The clean data were aligned to the Nipponbare reference genome (MSU v7.0) with SOAP2 [22],allowing 5 mismatches at most.The candidate single nucleotide polymorphisms (SNPs) for KH-9 were extracted and compared with the SNPs in other HHZ mutants using SIMM [23].Mutations presented in the Rice3K project [24] were also discarded.Only SNPs presented in clusters with SNP index ≥0.8 and depth ≥10 in KH-9,and SNP index ≤0.2 and depth ≥10 in other HHZ mutants were retained.SNPs causing amino acid variations or splicing variations on non-transposon genes were considered as candidates.To the candidate site,linkage between the genotype and phenotype was verified with high resolution melting(HRM) experiment [25].

2.4.Linkage verification of W548M site with herbicide-resistance phenotype

The linkage between W548M mutation and imazapic-resistance was tested using HRM method with PCR primers (5′-GTGTTGAA CAACCAACA-3′and 5′-AAGTATGTATGCGCCCT-3′) flanking the mutation site and genomic DNA isolated from the F2segregating population generated by crossing the W548M mutant with wild type HHZ.Briefly,PCR reactions were set up by mixing 10× PCR buffer 1.0 μL,2.5 mmol L-1dNTP Mix 0.1 μL,10 μmol L-1primers 0.1 μL for each,rTaq 0.1 μL,template DNA 1 μL,ddH2O 7.5 μL,and 0.1 μL EvaGreen.To avoid evaporation,20 μL of mineral oil was added to each reaction.The PCR program was 95 °C for 3 min;40 cycles of 95 °C for 30 s,54 °C for 30 s,and 72 °C for 10 s;then 72 °C for 1 min and 95 °C for 1 min.When the PCR reaction was over,add 1 μL of low-temperature internal standard to each sample for calibration.The low-temperature internal standard sequence is:forward primer 5′-TTAAATTATAAAATATTTATAATAT TAATTATATATATATAAATATAATA-3′and reverse primer 5′-TATTATATTTATATATATATAATTAATATTAATATTTTATAATTTAA-3′.The PCR product was placed i n Li ghtscanner HR 96 (Idaho Technology Inc.) for genotyping analysis.

2.5.Identification of mutation sites in ALS gene by sequencing

Genomic DNA was extracted from herbicide resistant mutants using the CTAB method [26].PCR primers (5′-CCATCCGAGCCACA CATCGCCTC-3′and 5′-ACAAACATCATAGGCATACCACTC-3′) were designed according to theALScDNA sequence (accession number AB049822).TheALSgene in each mutant was amplified by PCR.The PCR products were separated by agarose gel electrophoresis,and the target band was extracted from the gel for DNA sequencing.DNASTAR software was used for DNA sequence alignment to identify the mutation site.

2.6.Transfer the W548M site into other rice varieties by crossing

JTD-001 derived from KH-9 was crossed with differentindicaorjaponicavarieties to obtain BC1F1,and these varieties were used as recurrent parents for further backcrossing.The seeds obtained after each backcrossing were sowed,and 20 days after sowing,the seedlings were sprayed with imazapic (108 g a.i.ha-1).Resistant plants with vegetative growth and appearance similar to the recurrent parent were selected for further backcrossing in each generation until BC4F1.The resistant BC4F1individuals most similar in appearance and production traits to the recurrent parent were selfed to obtain the BC4F2seeds,and 20 days after sowing,the BC4F2plants were further sprayed with imazapic (108 g a.i.ha-1).The resistant BC4F2individuals most similar in appearance to the recurrent parent were selected for HRM analysis of the W548M mutation,and plants of homozygous mutation were further selfed to propagate the BC4F3seeds.BC4F3plants were evaluated for herbicide resistance and stability of other important agronomic traits under field conditions by growing >500 plants in each plot.

2.7.Comparative analysis of W548M and S627N resistance to different herbicides

Rice seedlings were grown in trays,each tray was equally divided into 3 plots,which were respectively planted with JTD-001 (W548M),HHZ(WT),and KS-2 (S627N,with background mutations cleaned).At three-leaf stage,the IMI herbicides imazapic,imazamox,and imazathapyr;the SU herbicides mesosulfuronmethyl,tribenuron-methyl,and nicosulfuron;the SCT herbicide flucarbazone;the TP herbicide pyroxsulam;and the PTB herbicide bispyribac-sodium at 0.5×,1×,2×,4×,and 8× of the recommended field dosage were sprayed on the plants at 450 L ha-1.Three trays of seedlings were used for each treatment,and the experiment was repeated 3 times.The fresh weight of the plants was measured 20 days after treatment for calculation of the fresh weight inhibition rate.The inhibition rate was calculated as(Wf-Wt)/WfwhereWfindicated the fresh weight of control plants without herbicide treatment,andWtindicated the fresh weight of treated plants.According to the fresh weight inhibition rate,the resistance level of each plant to different herbicides was determined.GR50indicated the herbicide rate causing 50%plant growth reduction.The resistance index (RI) was calculated by the GR50of the resistant population divided by GR50of the sensitive population to estimate the resistance levels.

3.Result

3.1.The isolation of imazapic-resistant mutants

To isolate the herbicide resistant mutants,we initially sowed the M2seeds in the field and then sprayed the seedlings with 108 g a.i.ha-1imazapic solution(Fig.1A).From a total of 200 pools of M2seeds,two resistant plants were obtained.Because field screening was time-consuming and low efficient,we later deployed another method by soaking the M2seeds in imazapic solution and then germinating the seeds on wet towel paper.Most of the M2seeds germinated but the sprouts stopped elongation after a couple of days (Fig.1B).The sprouts continuing elongation and turning green were transplanted into soil for further growth,and the survived plants were regarded as candidate resistant mutants.From a total of 100 M2pools,we obtained 67 candidate resistant individuals that were able to set seeds.

To confirm the resistance in these mutants,M3seeds harvested from each candidate mutant were sowed in field,and the seedlings were sprayed with imazapic(108 g a.i.ha-1).Some of the mutants showed segregation of herbicide resistant and sensitive phenotypes,while some did not show segregation.Totally,we obtained 25 confirmed mutants with imazapic resistance.The M3plants derived from mutant KH-9 all showed normal growth after imazapic-treatment,and this plant was chosen for further study.

3.2.Genetic study and identification of the mutation site for imazapicresistance in KH-9

The uniform resistance of the M3plants derived from KH-9 suggested that KH-9 was likely a homozygous mutant.To determine the genetic nature of the resistance trait,the M3plants were crossed with WT HHZ,and the F1plants were further selfpollinated to produce the F2populations.The F2plants were sprayed with 108 g a.i.ha-1imazapic at three-leaf stage,and the number of herbicide-resistant and sensitive plants was counted 20 days after spraying.The total number of plants in three F2populations was 160,118,and 103,respectively,of which 123,90,and 79 were resistant,while 37,28,and 24 were sensitive (Table S1).The resistant and sensitive plants were separated at a ratio of 3:1,suggesting that the resistance trait in KH-9 was controlled by a single dominant gene.

Fig.1.Mutant screening.(A) Mutant screening by spraying seedlings in field.The candidate mutant is shown in insert.(B) Mutant screening by soaking seeds in imazapic solution and germinating on wet towel paper.The candidate mutant plant is indicated by red arrow.

The whole genome re-sequencing method was deployed to identify the mutant gene responsible for herbicide-resistance in KH-9.Genomic DNA was isolated from 30 M3individuals derived from KH-9,and equal amount of DNA was mixed together and sequenced using the Illumina Hiseq platform.The re-sequencing data were then processed with SIMM algorithm,a pipeline for simultaneous identification of causal mutations in multiple mutants [23].A total of 118,486,076 clean reads were obtained from sequencing,and 74.51% of the reads were properly mapped to the Nipponbare reference genome.The resulting 2,539,731 SNPs between KH-9 and Nipponbare were then filtered with the SNPs presented in other HHZ mutants and SNPs presented in the Rice3K project [24],resulting in 7421 SNPs.Because the resistance locus was homozygous,the SNPs were further filtered with SNP index(SI),an indicator of the ratio of mutant reads to the total reads of a SNP in bulk-sequenced data.A homozygous mutant was expected to have a SI=1.However,considering that the Illumina Hiseq platform often generates random sequencing errors,and SNPs supported by a small number of reads are less reliable,we set SI ≥0.8 and reads ≥10 as cutoff criteria to filter the SNPs,leading to 98 SNPs in total.Of these,four SNPs located on chromosome 2 caused amino acid changes or RNA splicing variations(Table S2).

Two adjacent SNPs were found on geneLOC_Os02g30630,inducing substitution of amino acid Trp548(TGG) by Met (ATG) in the protein (Fig.2).LOC_Os02g30630encodes acetolactate synthase OsALS.A number of ALS mutants have been reported in various plant species that confer IMI-resistance [6].The other two SNPs were localized inLOC_Os02g32030andLOC_Os02g33180that were not known with a function in herbicide resistance.

We then determined the linkages between the mutation inLOC_Os02g30630with the resistance phenotype in F2populations.As expected,all resistant individuals harbored either homozygous or heterozygous mutation,while the sensitive individuals did not carry mutation onOsALS(Fig.S1).

3.3.Identification of mutation sites in other imazapic-resistant mutants

Besides KH-9,we also obtained 24 other rice mutants resistant to imazapic.Because previous studies showed that resistance to IMI herbicides was mostly caused by mutations inALSgenes,we decided to test if theOsALSgene in these mutants had a mutation.OsALSin these mutants was PCR-amplified and subjected to sequencing analysis.Compared with the wild typeOsALSsequence,mutation was identified inOsALSgene in each of these mutants(Table 1).Among them,one mutant line had a G to A mutation at position 286,causing substitution of amino acid Ala96(GCG)by Thr (ACG);10 mutant lines had a C to T mutation at position 287,resulting in the conversion of Ala96(GCG) to Val (GTG);7 mutants,including KH-9,had TG to AT mutations at positions 1642 and 1643,resulting in the mutation of Trp548(TGG) to Met(ATG);5 mutant lines had a G to T mutation at position 1644,resulting in the substitution of Trp548(TGG) by Cys (TGT);and 2 mutant strains had a G to A mutation at position 1880,resulting the change of Ser627(AGT) to Asn (AAT) (Table 1).Alignment of the amino acid sequence of these mutants with ALS proteins from various other plant species indicated that the mutations all occurred to the amino acids that were known to be important in herbicide interaction (Fig.S2).

Fig.2.Identification of the causal mutation in KH-9.SNPs are indicated by the dots in the diagram.Candidate SNPs have higher SI and ED6 values.The causal mutation site marked with red arrow is located in gene LOC_Os02g30630 with TGG (Trp) changed to ATG (Met).

Table 1 Mutation information of the 25 herbicide resistant mutants.

3.4.Background mutation cleaning and breeding of JTD-001

HHZ is an excellentindicacultivar that has been widely cultivated in China for years [21].A herbicide-resistance trait would significantly enhance its value.However,EMS treatment can induce hundreds of mutations in the plant,and some of the mutations may have adverse effects on the variety.The whole genome re-sequencing analysis of KH-9 identified 7421 SNPs that might be induced by EMS.Thus,it is necessary to clean the background mutations irrelevant to herbicide resistance.

We used the M3progeny of KH-9 and wild type HHZ as the recurrent parent in backcross.After three generations of backcross,the BC3F1plant was selfed to produce BC3F2,and 100 BC3F2seeds were planted to propagate the BC3F3generation.Each generation was sprayed with 108 g a.i.ha-1imazapic,and the surviving plants were selected for crossing to produce next generation.Finally,the BC3F3plants were compared with HHZ in plots,and a strain with stable inheritance of herbicide-resistance and uniform physiological traits was obtained,named JTD-001 (Fig.3A).

To test if the W548M mutation has an effect on rice growth and development,we conducted two years of field test in 2015–2016,and compared the morphological and physiological characteristics of JTD-001 and HHZ under the same planting conditions.There was no significant difference between JTD-001 and HHZ in terms of plant architecture,leaf morphology,panicle shape,and grain morphology (Fig.3B–D).

We also compared a number of yield traits between JTD-001 and HHZ,including growth period,plant height,panicle length,grain number per panicle,seed-setting rate,thousand-grain weight,effective panicles per hectare,and yield per hectare,and did not detect any significant difference as well (Table 2).These results indicated that the W548M mutation did not have a detectable impact on rice growth and production.

3.5.Comparison of W548M and S627N mutations on the level and spectrum of herbicide resistance

S627N has been used in Clearfield rice varieties across different regions in the world for years [1].To determine if there is any difference between W548M and S627N in term of the level and spectrum of herbicide resistance,we applied different concentrations of various types of ALS inhibiting herbicides to the plants.As shown in Fig.4,HHZ,JTD-001,and KS-2 (S627N,with background mutations cleaned) plants were grown in the same tray.The herbicides included the IMI herbicides imazapic,imazamox,and imazathapyr;the SU herbicides mesosulfuron-methyl,tribenuron-methyl,and nicosulfuron;the SCT herbicide flucarbazone-sodium,the TP herbicide pyroxsulam,and the PTB herbicide bispyribac-sodium.Each of the herbicides were sprayed at 0.5×,1×,2×,4×,and 8× of the commercially recommended dose for field application.GR50(g a.i.ha-1) and RI were calculated for each of the herbicides.

It was clear that HHZ was sensitive to most of the chemicals except bispyribac,because the GR50values for HHZ were much lower than the recommended doses for field application(Table 3).JTD-001 showed a high level of resistance to all the herbicides tested.The S627N mutant KS-2 was resistant to imidazolinones,nicosulfuron,flucarbazone,and pyroxsulam,but the level of resistance level was lower than that of JTD-001 (Fig.4).

Compared with HHZ,the RIs of JTD-001 were 152.41,451.97,and 118.35,respectively,for the three IMI agents,which was extremely resistant.However,the RIs of KS-2 for the three imidazolinone agents were 11.70,28.84,and 11.52,respectively,much lower than that of JTD-001 (Table 3).As for the SU herbicides mesosulfuron-methyl,tribenuron,and nicosulfuron,the RIs of JTD-001 were 18.67,84.70,and 113.23,respectively,while the RIs of KS-2 were 1.04,1.02,and 4.56,respectively.Apparently,KS-2 did not have a significant resistance to mesosulfuronmethyl and tribenuron.While KS-2 showed some resistance to nicosulfuron,but the level of resistance was much lower than that of JTD-001(Table 3).The RI of JTD-001 for the SCT herbicide flucarbazone was 106.21,which was also very strong.However,the RI of KS-2 to this herbicide was only 2.14,much weaker than JTD-001(Table 3).For the TP herbicide pyroxsulam,the RI of JTD-001 was 88.70,while RI of KS-2 was 8.81,also much lower than that of JTD-001(Table 3).HHZ itself had a good resistance to the PTB herbicide bispyribac.The RI of JTD-001 to bispyribac was 6.81,indicating the W548M mutation elevated the resistance to this chemical as well.However,the RI of KS-2 was 0.98,equivalent to that of HHZ.

Table 2 Comparison of yield traits between HHZ and JTD-001.

Table 3 Comaprison of herbicide resistance levels between JTD-001 and KS-2.

These results indicated that W548M has a much broader spectrum of resistance to ALS inhibiting herbicides than that of S627N.Besides,to the chemicals that both W548M and S627N exhibited resistance,W548M showed a much higher level of resistance than did S627N.

3.6.Breeding of new rice varieties using the W548M mutation

Fig.3.Herbicide response and morphological comparison between JTD-001 and HHZ.(A) JTD-001 and HHZ seedlings after sprayed with imazapic.(B–D) Comparisons of plant architecture (B),panicles (C),and seeds (D) between HHZ and JTD-001 at yellow-ripening stage.

The above results indicated that W548M had a high level of resistance to a broad spectrum of ALS inhibiting herbicides without negative impact on plant growth and production.Thus it is likely a valuable resource for breeding of herbicide resistant rice varieties.To test if the performance of W548M mutation is stable in different rice background as well as to breed more rice varieties with the herbicide-resistance trait,we transferred the W548M locus from JTD-001 to 72 rice varieties (45indicaand 27japonica) through backcrossing.Takingindicavarieties Fengruanzhan,Xinganzao 10,Huangxiuzhan,andjaponicavarieties Xuhan 1,Huaidao 5,and Songjing 12 as examples,we crossed JTD-001 with these varieties as recurrent parents,and backcrossed them to BC4F1generation.Homozygous lines with stable physiological traits were selected by herbicide-spraying and molecular verification of the W548M locus using HRM analysis.The stable lines were then sprayed with a high dose of imazapic (576 g a.i.ha-1) at the three-leaf stage,and the phenotype was observed 20 days after spraying.All the W548M varieties grew normally without significant herbicide damage,while the original parent lines were all killed (Fig.5).The results showed that the herbicide resistance of W548M was stable,regardless of the rice background.

4.Discussion

Herbicide-resistant mutant crop plants are valuable to agriculture.Previous to this study,three rice mutants conferring resistance to ALS inhibiting herbicides had been isolated,including Gly628Glu,Ala96Thr,and S627N mutants of ALS,and S627N had been deployed for commercial weed management [1,27].In this study,we isolated five ALS mutants in theindicarice HHZ background,including Ala96Thr,Ala96Val,W548M,Trp548Cys,and S627N.Among them,Ala96Val,W548M,and Trp548Cys are new mutations that have not been reported before in rice.Further studies showed that W548M mutation had no visible impact on rice growth and production but with higher resistance to a broader spectrum of ALSinhibiting herbicides than S627N.These results indicated that the W548M mutant provides another valuable resource for rice breeding program for weed management.

Fig.4.Comparison between JTD-001 and KS-2 in resistance spectrum and resistance levels to various ALS-inhibiting herbicides.Plants from left to right in each tray were JTD-001,HHZ,and KS-2.1× means that the applied dose is equivalent to the recommended dose of field application.The recommended field dose is 108 g a.i.ha-1 for imazapic,45 g a.i.ha-1 for imazamox,75 g a.i.ha-1 for imazethapyr,135 g a.i.ha-1 for mesosulfuron-methyl,22.5 g a.i.ha-1 for tribenuron -methyl,54 g a.i.ha-1 for nicosulfuron,31.5 g a.i.ha-1 for flucarbazone,9 g a.i.ha-1 for pyroxsulam,and 33.75 g a.i.ha-1 for bispyrifen.

Fig.5.W548M locus confers herbicide resistance after transferred to other rice varieties.(A) Plants sprayed with water.(B) Plants sprayed with 576 g a.i.ha-1 imazapic.All the plants were photographed 20 days after spraying.JTD-001,HHZ/KH-9 BC3F4;HHZ,Huanghuazhan;JT-FR,FRZ/JTD-001 BC4F3;FRZ,Fengruanzhan;JT-XA,XAZ-10/JTD-001 BC4F3;XAZ-10,Xinganzao-10;JT-HX,HXZ/JTD-001 BC4F3;HXZ,Huangxiuzhan;JT-XH,XH-1/JTD-001 BC4F3;XH-1,Xuhan-1;JT-HD,HD-5/JTD-001 BC4F3;HD-5,Huaidao-5;JT-SJ,SJ-12/JTD-001 BC4F3;SJ-12,Songjing-12.

Since the application of ALS inhibiting herbicides,there are 30 amino acid substitutions in eight positions of plant ALS proteins(equivalent to Gly95,Ala96,Ala129,Pro171,Ala179,Asp350,Arg315,Trp548,Ser627and Gly628of the rice ALS)that have been discovered to cause resistance to ALS-inhibiting herbicides in crop plants and resistant weeds [6,28,29].The most commonly occurring mutations are at positions of Ala96,Pro171,Ala179,Trp548,and Ser627[6].A number of ALS mutants have been used in cultivation of crop varieties resistant to ALS inhibitor herbicides,such as S627N in rice and wheat,S627N and Trp548Leu in rapeseed,S627N,Trp548Leu,Ala129Thr,and Ala96Thr in corn,and Ala179Val in sunflower [6].S627N is currently the most widely used ALS mutation in commercial rice farming[6].This mutant confers resistance to the IMI family herbicides.IMI herbicides have a broad herbicidal spectrum with high efficiency.Most of the Clearfield rice varieties use S627N because it confers higher resistance to IMI herbicides than the other two available mutants[1].However,IMI herbicides such as imazethapyr,imazapyr,imazapic,and imazaquin have a long turnover time,and the residues are harmful to the next crop.This problem is particularly serious in Asian countries where the rice field is usually crop-planted 2–3 times a year [1].Thus,a mutant that can replace S627N in rice is deemed to be useful for weed management.Mutants on Trp548have been isolated in a number of plants including corn and rapeseed [30].Different from S627N,the mutants of Trp548site are generally cross-tolerant to the different families of ALS-inhibiting herbicides including the IMI family [6].Trp548Leu mutants have been used in rapeseed and corn [6],but mutant at Trp548was not available to rice before this study.

Upon isolation of the W548M mutant in HHZ,we compared W548M with S627N to all five families of ALS-inhibiting herbicides,and found that W548M conferred resistance to all families of ALSinhibiting herbicides in addition to the IMI family.Besides,the resistance index of W548M rice far exceeded the S627N rice.This characteristic gives W548M the following advantages in commercial applications in comparison with S627N.

First,W548M can improve the safety in field application.During field spray of herbicides,it is common that local application exceeds the recommended dose due to repeated spray or mechanical problems,which may cause crop injuries.However,the RI of W548M locus is mostly several to dozens folds of the recommended dosage.The high level of resistance can protect the plants from high-dose herbicide toxicity,and thus it is safer to use compared with plants of low resistance.

Second,W548M can be used to improve the safety of subsequent cropping.The S627N site is only resistant to IMI herbicides.Although IMI herbicides are a class of excellent herbicides with a broad herbicidal spectrum and high herbicidal efficiency,most of the IM herbicides,such as imazethapyr,imazapyr,imazapic,and imazaquin,have a long turnover time.However,in Asian countries,rice field is often crop-planted 2–3 times a year.For example,after rice planting in southern China,the fields are often planted with corn,rapeseed,vegetables,fruit melons,and other crops.Many crops are very sensitive to IMI herbicides.The W548M rice is resistant to many other types of ALS-inhibiting herbicides besides the IMI herbicides,including many types with short turnover time and high herbicidal efficiency.These types of herbicides can be used in conjunction with the W548M rice to solve the residue problem for rotation crops.

Third,W548M can be used more safely to improve the hybrid rice purity.Hybrid seeds are produced by cross-pollination of male sterile lines by male parents.Some male sterile lines,particularly the male sterile lines controlled by photoperiod/thermalsensitive genes,are usually unstable in male sterility,and they can set seeds by self-pollination under environmental conditions restoring the male fertility.If the paternal line carries W548M mutation,herbicide resistance can be deployed to remove the impurities produced by self-pollination of the maternal line or by cross-pollination with pollens from herbicide-sensitive nonpaternal sources.The herbicide resistance conferred by ALS mutations is a dominant trait,but the level of resistance is accumulative,i.e.,the level of herbicide resistance of heterozygous plant is much lower than that of the homozygous mutant.However,due to the high resistance of W548M,even heterozygotic state can survive the dosage of field application.In field production,impurities in hybrids can be effectively removed from the hybrids by herbicide-soaking of the hybrid seeds or herbicide-spray of seedlings in field.

Finally,W548M can be applied to hybrid seed production by mixed-seeding of the parental lines.At present,hybrid seed production is carried out solely by planting the parental lines in separate rows,then using manual or wind pollination,and finally removing the male plants manually after pollination.This method is very laborious and costly and is not suitable for mechanized production of hybrid seeds.Researchers have been seeking for ways for mix-planting both parents and mix-harvesting the seeds,and then using a simple method to get rid of the non-hybrid seeds.This can be realized by introducing the W548M locus into the male sterile line and keeping the paternal line sensitive to the ALS-inhibiting herbicides,and treating the mix-harvested seeds with herbicide to kill the non-hybrids produced by male parent.

Clearfield rice based on S627N has been widely planted in the United States,Brazil,Argentina,Colombia,Uruguay,Italy,Malaysia and Vietnam,with planting acreage exceeding 1,500,000 ha in total by the year of 2012 [1].The deployment of Clearfield rice has achieved good results in weed control including weedy rice management [1].JTD-001 has the W548M site,and at present,it has been grown at more than 500 test sites in 16 provinces in China,with more than 2000 rice growers participated in the application trials.In areas with serious weedy rice infestation,such as Leizhou in Guangdong province,plot tests showed that the use of JTD-001 in combination with supporting herbicides could effectively control the incidence of weedy rice to 0.07% in direct-seeded rice fields.Compared with the incidence of weedy rice of 13.31% in the control field,the weedy rice eradication effect was remarkable [5].The effective control of weedy rice in the fields significantly reduced the cost in weed management and also increased the rice yield.As a result,the application of JTD-001 increased the farmers’income by$200 ha-1[5].There are more than 3,000,000 ha of rice fields infested by weedy rice in China.Therefore,W548M rice has an enormous application prospect.

Herbicide-resistant crops are a frontier in agricultural biotechnology.At present,the most widely grown herbicide-resistant crops are genetically modified corn,rapeseed,and soybean.However,whether in China or any other countries in the world,genetically modified crops are still facing strict regulation and long approval process,as well as negative public opinions.In this scenario,herbicide-resistant crops obtained by traditional mutation breeding methods are a better choice.The W548M rice mutant is expected to provide a powerful tool for weed control.

CRediT authorship contribution statement

Xiaoyan Tang and Xing Wang Deng:conceived the experiments.Lei Chen,Gang Gu,and Chengxu Wang:designed and conducted the herbicide analyses.Chengxu Wang:screened and obtained the mutants.Gang Xie:did rice breeding.Zhufeng Chen:constructed the mutant library and sequenced the mutant genes.Wei Yan:did SIMM assay of KH-9 mutant.Man Jin:analyzed the data,prepared the figures,and revised the paper.Junli Zhou:studied the intellectual properties.Lei Chen and Xiaoyan Tang:wrote and revised the paper.Xing Wang Deng:reviewed and edited the manuscript.All authors read and approved the manuscript.

Declaration of competing interest

Authors declare that there are no conflicts of interest.

Acknowledgments

HHZ seeds used in this study were propagated and provided by Shenzhen Xing Wang Biological Seed Industry Co.,Ltd.The other parental lines were provided by Sichuan Zhongyan Seed Industry Co.,Ltd.This work was supported by Major Program of Guangdong Basic and Applied Research (2019B030302006),National Natural Science Foundation of China (U1901203 and 31901532),Natural Science Foundation of Guangdong Province (2018B030308008 and 2018A0303130270),Shenzhen Commission on Innovation and Technology Programs (JCYJ20180507181837997),and China Postdoctoral Science Foundation (2018M633069 and 2019M652920).

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.11.003.