CRISPR/Cas systems:The link between functional genes and genetic improvement

Yong Huang ,Huirong Dong ,Meiqi Shang,Kejian Wang*

State Key Laboratory of Rice Biology,China National Rice Research Institute,Chinese Academy of Agricultural Sciences,Hangzhou 310006,Zhejiang,China

Keywords:CRISPR/Cas Genome editing Rice Application Heterosis

ABSTRACT With the ever-increasing human population and deteriorating environmental conditions,there is an urgent need to breed environmentally friendly and resource-conserving rice cultivars to achieve sustainable agricultural development and food security.However,conventional rice improvement strategies,such as hybrid breeding,are time-consuming and laborious processes and may not be able to keep pace with increasing food demand in the future.Targeted genome-editing technologies,especially clustered regularly interspaced short palindromic repeat(CRISPR)/CRISPR-associated protein(CRISPR/Cas),permit efficient targeted genome modification and offer great promise for the creation of desired plants with higher yield,improved grain quality,and resistance to herbicides,diseases,and insect pests.There is also great potential for tapping heterosis using the CRISPR/Cas technology.In this review,we focus on the most essential applications of CRISPR/Cas genome editing to rice genetic improvement,considering traits such as yield,quality,and herbicide,disease and insect-pest resistance.We discuss applications of CRISPR/Cas to the exploitation of heterosis.Finally,we outline perspectives for future rice breeding using genome-editing technologies.

1.Introduction

Global food security is suffering because of severe challenges posed by population growth,reduced arable land,water scarcity,and global climate change[1].Rice is a staple food for much of the world’s population[2].During the last half-century,world rice production has increased substantially because of an increase in harvest index and the exploitation of heterosis[3].However,since the mid-1990s,the average yield of rice has reached a plateau[4].Although conventional cross-breeding still plays a pivotal role in maintaining food security,it has several limitations,such as the requirements for multiple generations of selfing and artificial selection to introduce desirable alleles and the frequent introduction of deleterious along with desirable alleles because of inevitable linkage drag[5].Such laborious,time-consuming,untargeted,and complicated breeding programs cannot meet the increasing demand for crops in the future[6].At the current rate of growth,the global population is expected to reach～10 billion by 2050[7].To feed this population,it is estimated[8]that agricultural production levels need to increase by up to 60%by 2050.Novel approaches to rice breeding are needed to break the yield ceiling.

Since Charpentier and Doudna discovered the clustered regularly interspaced short palindromic repeats(CRISPR)/CRISPR-associated(CRISPR/Cas)genetic scissors in 2012[9,10],CRISPR/Cas-based techniques have been successfully used as efficient tools for genome editing in a variety of species because of their simplicity,efficiency,and versatility[11].The production of DNA doublestrand breaks(DSBs)at target loci is the key characteristic of CRISPR/Cas gene editing,with two main repair pathways:nonhomologous end joining(NHEJ)and homology-directed repair(HDR)[12].DSBs are usually repaired by the NHEJ machinery,which often creates insertions/deletions of one or several bases at target sites with extremely high efficiency.However,in terms of generating large deletions of genomic fragment,microhomology-mediated end joining(MMEJ)may be more effective than the NHEJ[13,14].Since no homologous repair template is required,NHEJ is a preferred and popular repair pathway[5].Although the error-prone end-joining processes,that is NHEJ,are highly efficient for knockout studies,they do not support precise modification,and some randomly generated unexpected mutations may occur[15].HDR-mediated genome editing can be used to introduce desired mutations precisely into a target site.However,HDR is generally thought to occur at a very low frequency in plants because of limitations in donor-template delivery to plant cells[16].Recently,several strategies,such as transcripttemplated HDR(TT-HDR)and tandem repeat HDR(TR-HDR)[17,18],have been devised to perform HDR in rice.In these approaches,a precisely customized gene can be generated with the help of an appropriate repair sequence.Beyond DSB-mediated genome editing,base editors,which provide a highly efficient and simple strategy for engineering nucleotide substitutions at target sites,directly induce specific base changes without requiring DSBs or donor DNA and without depending on HDR.Although this highly efficient base-editing technology is currently limited to C·G to T·A and A·T to G·C substitutions,it has been rapidly expanding and is applied in rice for its precise nucleotide substitutions[19-26].Prime editing is a technology that can introduce all possible types of point mutations,insertions,or deletions in a precise and targeted manner without reliance on DSBs or donor DNA templates,but it currently exhibits rather low editing efficiencies,limiting its application in rice[27,28].Using these new genetic scissors,researchers can realize the ultimate aspiration of being able to make any desired sequence alteration at any position in the genome[12].Such systems have driven advances in a variety of plant species,including model plant species and agriculturally important crops such as rice,in turn providing a huge stage for CRISPR/Cas research.This virtuous circle leads us to a new path toward increasing rice yield potential and global food security.To date,although many genes for agronomic traits have been identified in rice,the achievements of functional genomics research have not been applied well to rice improvement.As the bridge between functional genes and genetic improvement,CRISPR/Cas-based genome editing technologies have attracted increasing attention from rice scientists and breeders.These technologies provide rice researchers with the opportunity to introduce specific and explicit changes at target loci and can be used for knocking out multiple genes simultaneously in the rice genome.

In this review,we describe the most recent application of CRISPR/Cas technologies to rice improvement,considering yield increase,quality improvement,improvement of herbicide,disease,and insect-pest resistance,and exploitation of heterosis.Our goal is to promote the use of this genome-editing tool for breeding highyield,high-quality,and multiply resistant rice cultivars.

2.Genome editing to increase rice yield

Grain yield is determined mainly by three components:panicle number,grain number per panicle,and grain weight[29].In recent decades,great progress has been made in cloning yield-associated genes with the help of functional genomic research.Multiple functional genes could be stacked in elite rice cultivars by gene-editing technologies,to improve rice yield.As current genome editing is used mainly to introduce gene-targeted knockouts,yieldincreasing functional genes have not been well applied in rice breeding.We accordingly focus on the application of gene editing to functional genes or regulators that affect yield.Several strategies that employ gene-editing technology have been reported to improve rice yield(Fig.1).

2.1.Knockout of yield-reducing genes to increase rice yield

Knocking out functional genes that reduce yield in elite rice cultivars via gene-editing technology is a promising approach to the creation of ideal rice cultivars.For example,four yield-reducing functional genes:GRAIN NUMBER 1a(Gn1a),DENSE AND ERECT PANICLE1(DEP1),GRAIN SIZE 3(GS3)and IDEAL PLANT ARCHITECTURE1(IPA1),were edited using the CRISPR/Cas9 system to improve yield traits[30].Wang et al.[31]deleted large fragments of DEP1 in indica rice and obtained yield-increasing dep1 mutants similar to those reported previously[32].Yield increase has been achieved by the simultaneous genome engineering of three deleterious functional genes for grain size and grain number:GS3,GRAIN WIDTH 2(GW2),and Gn1a[33,34].Shen et al.[35]used a multiplex CRISPR/Cas9 system to target eight rice agronomic genes and quickly obtained abundant and varied breeding materials.The eight genes included three adverse functional genes for panicle architecture(DEP1,EP3,and Gn1a),two functional genes for grain size(GS3 and GW2)and three major genes affecting plant architecture(LPA1),rice aroma(BADH2),and heading date(Hd1).Knockout mutants of GW5 and GW5L both resulted in significantly increased grain width and weight[36,37].Mutagenesis of RGG2,a deleterious functional rice gene for plant growth and organ size,using the CRISPR/Cas9 system yielded elongated internodes,increased 1000-grain weight and plant biomass,and increased grain yield per plant[38].Targeted mutagenesis of OsFWL4,an adverse functional rice gene for tiller number and plant yield,using the CRISPR/Cas9 system led to increased tiller number and plant yield[39].Recently,Lyu et al.[40]used CRISPR/Cas9 to perform targeted mutagenesis of POLYAMINE OXIDASE 5(OsPAO5),which negatively regulates mesocotyl elongation,and obtained direct-seeding rice with markedly increased grain weight,grain number,and yield potential.These findings suggest that deletion of negative regulators of yield is a rapid way to generate ideal germplasm.

2.2.Manipulation of regulators of cytokinin homeostasis and signal response to increase rice yield

The manipulation of regulators of cytokinin homeostasis and signal response is a practical approach to increasing rice yield.New rice germplasm with increased salinity tolerance without loss of grain yield was obtained by CRISPR/Cas9-targeted mutagenesis of the OsRR22 gene,which is involved in both cytokinin signal transduction and metabolism[41].CRISPR-edited variants at the 3′-end of OsLOGL5,which encodes a cytokinin-activation enzyme,increased grain yield under various field environments[42].Cytokinin oxidase/dehydrogenase(CKX)is the main enzyme that inactivates cytokinin.Disruption of OsCKX11 using the CRISPR/Cas9 system increased cytokinin content and grain yield compared with wild-type rice[43].Similarly,knockout of OsCKX2 in a similar manner to that described above for the Gn1a gene generated higheryield phenotypes in rice[44].In these contexts,the role of OsCKX is promising for increasing rice yield because it may regulate the level of cytokinins in vivo.CRISPR/Cas9-mediated genome editing of the drought and salt tolerance gene,which directly regulates the expression of OsCKX2/Gn1a,yielded rice plants with high yield[45].Plants derived by mutagenesis of ERECTA1,which acts upstream of the OsMKKK10-OsMKK4-OsMPK6 cascade to regulate cytokinin metabolism in rice,led to increased spikelet number per panicle[46].The number of tillers increased significantly in knockout lines of amino acid permease 3(OsAAP3),which affects tiller bud outgrowth via the cytokinin pathway,thereby increasing rice yield[47].Similar phenotypes characterized by increased tiller number and grain yield,were also observed in OsAAP5-knockout plants,lacking a gene that regulates tiller bud outgrowth by affecting cytokinin levels[48].OsNAC2 knockout plants derived via the CRISPR/Cas9 system exhibited a stronger root system and higher rice yield via the auxin-OsNAC2-cytokinin pathway[49].Knockout of a basic helix-loop-helix transcription factor,OsPIL15,using the CRISPR/Cas9 technology increased grain size and yield by increasing cytokinin content in OsPIL15-knockout spikelets[50].These results suggest that cytokinin homeostasis and signal response genes have great potential for conferring superior rice yield.

Fig.1.Several strategies for improving rice yield byusing gene-editing technology.The five ellipses represent five strategies for improving rice yield by using gene-editing technology.OsCKX2/Gnla is the common target gene of two strategies(functional genes for decreasing yield and cytokinin-homeostasis and signal-response regulators).

2.3.Editing of plant-growth and environmental-response regulators to increase rice yield

The use of CRISPR/Cas9 to edit regulators that control plant growth and environmental response is an effective approach to promoting rice growth and productivity.In natural paddy field conditions,rice plants need to adapt to dynamic environmental factors over very long periods.Accordingly,researchers have attempted to create novel high-yield rice germplasm with resilience against various stress factors.Simultaneous mutation of rice genes encoding the abscisic acid(ABA)receptors PYRABACTIN RESISTANCE 1-LIKE 1(PYL1),PYL4,and PYL6 yielded plants with robust growth and increased grain yield [51].CRISPR/Cas9-induced modification of PYL9,which encodes one of the rice ABA receptors,conferred increased drought tolerance and grain yield[52].New rice cultivars with both high yield and high cold tolerance were obtained by simultaneously editing two yield-related negative genes(OsPIN5b and GS3)and one cold-tolerance gene(OsMYB30)using the CRISPR/Cas9 system[53].Rice PARAQUAT TOLERANCE 3 (OsPQT3)knockout mutants generated using CRISPR/Cas9 technology conferred higher yield and increased resistance to oxidative and salt stress[54].These examples represent a strategy for improving rice yield in specific environments.

2.4.Disruption of microRNA regulators to increase rice yield

Disruption of microRNA regulators may be a promising approach to improving rice yield.Several miRNAs have been reported to control rice yields by targeting their direct transcription factors,growth-regulating factors(GRFs).For example,knockout of MIR396e and MIR396f via CRISPR/Cas9 increased both grain size and panicle branching,resulting in increased grain yield[55,56].Mutation in OsGRF4 introduced by CRISPR/Cas9 perturbed the miR396-directed regulation of OsGRF4 and generated plants with larger grain size and increased grain yield[57]similar to those reported previously[58,59].Knockout of UCLACYANIN 8(OsUCL8),the downstream target of miR408,led to increased grain yield[60].Mutagenesis of genes in the MIR156 subfamily using CRISPR/Cas9 in rice led to increased seed dormancy without compromising grain yield[61].MIR396d[62],MIR529a[63],and MIR530[64]were elite targets for the breeding of high-yielding rice via genome-editing technologies.

3.Genome editing to improve rice grain quality

Rice grain quality is evaluated mainly based on four quality standards:processing quality,appearance quality,eating and cooking quality(ECQ),and nutritional quality[65].Because the last three have attracted the most attention in rice-consuming countries,we focus on gene-editing research aimed at these traits(Fig.2).

3.1.Improvement of rice grain appearance quality

Rice grain appearance,which affects the market acceptability of rice,is determined mainly by grain size,chalkiness,and transparency[66].Grain size is also an important yield component[29].In recent decades,numerous major QTL with large effects on grain size have been cloned and characterized in rice[67-76].Several authors have described gene editing for orientation improvement in these major grain-size QTL.For example,the two major QTL GS3 and GL3.1,which are negative regulators of grain size,were simultaneously edited via CRISPR/Cas9-mediated multiplex genome editing in the typical japonica rice cultivar Nipponbare.In the T1generation,gs3 formed slender grains with a lower chalkiness percentage,while gs3gl3.1 produced larger grains with a higher chalkiness percentage[77],indicating that knockout of GS3 and GL3.1 rapidly improved grain size.GS3 and Gn1a,which controls grain number,were successfully edited in four rice cultivars.Both gs3 and gs3gn1a mutants exhibited greater grain length and 1000-grain weight than their wild-type counterparts,and the gs3gn1a double mutants had more grains per panicle than the gs3 mutants[78,79].Three major genes,GW2,GW5,and TGW6,that negatively regulate grain weight were selected as targets for multiplex editing using the CRISPR/Cas9 system.The grain size of the CRISPR/Cas9-generated T-DNA-free homozygous gw2gw5tgw6 T1mutants was markedly greater than that of wild-type plants[80].These results indicate that the genome editing system is well suited for the rapid generation and pyramiding of beneficial alleles.

Fig.2.Genome editing to improve rice grain quality.For genes in blue font following each quality,the corresponding quality traits can be regulated using gene-editing technology.The scissors are Cas9,the orange line is single guide RNA(sgRNA),the green line is target sequence,and yellow parts are protospacer adjacent motifs(PAMs).The purple solid arrow indicates that rice quality can be improved.

Among the determinants of appearance quality,grain chalkiness is a highly undesirable quality trait that impairs eating and cooking and grain appearance,resulting in low market acceptability[81].Chalk5,the first cloned positive regulator of grain chalkiness[82],is a potential candidate gene and might be used for gene editing to obtain new germplasms with reduced grain chalkiness.Recently[83],the homozygous null gs9 mutant generated using the CRISPR/Cas9 editing system in the high-yield Yandao 8 cultivar showed a decrease in grain chalkiness without a negative effect on grain weight or final yield,suggesting that the null gs9 allele has high potential to improve grain appearance quality in rice cultivars with high yield but poor appearance.

3.2.Improvement of rice grain eating and cooking quality

Rice ECQ is determined mainly by three physiochemical properties:amylose content(AC),gel consistency,and gelatinization temperature,among which AC is the most important[84,85].Rice AC is governed by the Waxy(Wx)gene,which encodes granule-bound starch synthase I,an enzyme that controls amylose synthesis in the endosperm[86,87].In recent decades,multiple alleles(e.g.Wxlv,Wxa,Wxb,Wxin,Wx°p/hp,Wxmp,Wxmq,Wxmw,and wx)identified at the Wx locus accounted for high diversity in rice AC and affected consumer preference[88,89].Loss-of-function mutants generated using the CRISPR/Cas9 system to target the Wx coding region in multiple rice cultivars all showed decreased AC and produced glutinous rice[90,91],which is far from meeting the diverse demands of ECQ.New approaches to creating favorable allelic variation at the Wx locus are sought.Recently,several research teams in China used the CRISPR/Cas9 editing system to generate novel Wx alleles that yield fine-tuned grain AC by different strategies.Xu et al.[92]used a base-editing system to create a series of mutants with fine-tuned rice AC(0-12%),which have broadened the range of breeding materials available to breeders.Zeng et al.[93]developed high-efficiency CRISPR/Cas9-mediated Wx promoter/5′UTR intronic splicing site-engineered strategies for generating new quantitative trait alleles with fine-tuned transcriptional and posttranscriptional regulation of Wx expression activity.These new Wx alleles produced grain-improved lines with desirable AC levels,which might meet consumer preferences.Huang et al.created novel Wx alleles that yielded fine-tuned AC and improved grain quality in rice via promoter editing using the CRISPR/Cas9 system[94].These results indicate that these novel transgene-free Wx alleles with fine-tuned AC will help fill the AC gaps in the natural Wx alleles.

The strategies outlined above all aimed to produce new rice germplasm with fine-tuned AC by editing the Wx gene.But desired AC levels can also be obtained by editing genes upstream of Wx.Knockout plants for the heterotrimer complex NF-YB1-YC12-bHLH144,which directly activates Wx transcription,showed lower AC levels than the wild-type counterpart[95].Several genes that encode transcription factors:FLO2,OsBP-5,OsEBP-89,OsbZIP58,and OsMADS7,and are involved in transcriptional regulation of rice Wx[96-99]may also be potential targets of modification by the CRISPR/Cas system,to achieve a desired rice grain AC.Targeted mutagenesis of amino acid transporter genes(OsAAP6 and OsAAP10)using the CRISPR/Cas9 system can rapidly reduce grain protein content and improve rice ECQ[65],thus providing a new strategy for the breeding of cultivars with desired ECQ.

Fragrant rice cultivars are becoming popular because of their unique aroma and taste.There are many flavor compounds in rice,the most important being 2-acetyl-1-pyrroline(2AP),which is regulated mainly by the OsBADH2 gene.Function-knockout mutants of OsBADH2 obtained in the background of the non-aromatic rice Zhonghua 11 using CRISPR/Cas9 showed increased accumulation of 2AP and increased rice fragrance[100].Similar results were obtained for novel alleles of OsBADH2 via CRISPR/Cas9 mutagenesis[101,102].

3.3.Improvement of rice grain nutritional quality

The nutritional quality of rice,a staple food,strongly influences consumers’health.With an increase in worldwide attention to ricenutritional quality,target-induced mutagenesis of negative genes involved in the synthesis or accumulation of target nutrients in grain via the CRISPR/Cas gene-editing technology may be a fast and efficient approach for rice grain nutritional improvement.Several studies have been conducted to improve rice grain nutritional quality using the CRISPR/Cas9 system.Zhu et al.[103]used CRISPR/Cas9-mediated functional recovery of the recessive rc allele to develop red rice,which provides nutrients for improving human health.Marker-free rice plants with high carotenoid content in the seeds were obtained with an optimized CRISPR-Cas9-based method[104].Knockout of the phospholipase D gene(OsPLDα1)using the CRISPR/Cas9 system produced rice grains with reduced phytic acid content compared with their wild-type counterparts[105].Development by gene editing of elite rice cultivars with reduced heavy metal accumulation is a practical strategy for protecting human health.Knockout of the OsNramp5 gene using the CRISPR/Cas9 system produced promising rice cultivars with extremely low cadmium accumulation in grains,without compromising yield[106].Similarly,rice plants with a low cesium content were obtained by CRISPR/Cas9 knockout of OsHAK1[107].These approaches could lead to safe food production in regions with heavy metal contamination.

4.Genome editing to improve rice resistance

4.1.Herbicide resistance

Breeding herbicide-resistant rice can reduce the damage produced in rice by herbicides,increase the efficiency of chemical weeding,and reduce the cost of weeding(Fig.3A,B).The conventional method of breeding herbicide-resistant rice consists of introducing exogenous herbicide-resistance genes,such as Bar,into elite rice cultivars using transgenic technology.However,because of safety concerns,these herbicide-resistant rice cultivars cannot be used in production.CRISPR/Cas technologies enable precise modifications of herbicide-targeted genes in vivo and confer endogenous herbicide resistance on rice,thus offering great promise for rice improvement.Acetolactate synthase(ALS)catalyzes the first step of the biosynthetic pathway of branched-chain amino acids(valine,leucine,and isoleucine)[108,109]and is a target enzyme of ALS-inhibiting herbicides such as sulfonylureas and imidazolinones)Point mutations in the ALS gene sequence lead to specific amino acid substitutions that confer resistance to ALS-inhibiting herbicides[110].Accordingly,in recent years,the use of CRISPR/Cas to introduce specific base substitutions into the rice ALS gene has become a research target.Rice mutants with a change in the amino acid at position 96 from alanine to valine showed sulfonylurea resistance after being subjected to a fusion of CRISPR/Cas9 with activation-induced cytidine deaminase[111].The P197F amino acid substitution identified in a strong OsALS1 allele endowed rice plants with resistance to bispyribac-sodium by base-editing-mediated artificial evolution,and created a new herbicide-resistant rice cultivar[112].Introduction of point mutations into the rice ALS gene using CRISPR/Cas9-mediated homologous recombination also conferred herbicide tolerance on rice plants.The 548th and 627th amino acids of the rice ALS gene were edited to yield novel rice genotypes with bispyribac-sodium resistance[113,114].Acetyl coenzyme A carboxylase(ACCase)is the rate-limiting enzyme in the fatty acid biosynthetic pathway and is the target of ACCase-inhibiting herbicides,such as aryloxyphenoxypropionic acid(APP)and cyclohexenedione(CHD)[115].The C2186R amino acid substitution in the ACCase gene endows rice plants with resistance to haloxyfop-R-methyl[116].Similar haloxyfop-R-methyl-resistant rice genotypes were produced by induction of I1879V and W2125S amino acid substitutions in ACCase[117].Other amino acid substitutions in ACCase,such as W2125C and P1927F,were also obtained by CRISPR-based saturated targeted endogenous mutagenesis editors,which conferred haloxyfop resistance in rice[118].Tthe T102I and P106S aminoacid mutations in EPSPS and the M268T mutation in TubA2 have been reported[119,120]to confer rice resistance to glyphosate and trifluralin,respectively.Several novel amino-acid substitutions in the SF3B1 gene were discovered through CRISPR/Cas-based directed evolution and confer herboxidiene resistance in rice[121].These herbicide-resistance alleles generated using the CRISPR/Cas9 system hold great potential to accelerate the development of novel rice germplasm.

4.2.Disease and insect-pest resistance

Plant pathogens and pests and cause estimated losses of 20%-40%in global food production[122].Knockout of susceptibility genes using CRISPR/Cas technology is a promising strategy to increase rice resistance to disease and insects,reducing the impact of diseases and pests on rice development and yield(Fig.3A,C,D).In rice,several catastrophic diseases affect grain yield.Bacterial blight,caused by the main pathogen Xanthomonas oryzae pv.Oryzae(Xoo),is a prevalent and destructive rice disease.SWEET genes were reported[123]to be induced by Xoo and serve as susceptibility genes for bacterial blight.These genes are thus attractive targets for genome editing for disease resistance.Rice genotypes with broad-spectrum resistance to bacterial blight were generated by systematic interference with SWEET gene induction in the promoter region[124,125].Similar results were obtained with Xa13.Li et al.[126]used the CRISPR/Cas9 system to edit the Xa13 promoter and obtained transgene-free bacterial-blight-resistant rice.In addition to editing in the promoter region,which can increase resistance to bacterial blight in rice, CRISPR/Cas9-directed mutagenesis in the Xa13/Os8N3/OsSWEET11 coding region can also confer resistance to Xoo[127].CRISPR/Cas9 has been employed to produce plants resistant to rice blast,another disease causing large-scale reduction in rice production.Edited rice plants with knockout mutations created by CRISPR/Cas9 in an ethyleneresponsive factor,OsERF922,which is a negative regulator of blast resistance,showed increased resistance to Magnaporthe oryzae without alteration to agronomic traits[128,129].Similarly,rice plants resistant to blast disease were generated by disruption of the OsSEC3A gene using CRISPR/Cas9[130].Homozygous mutant plants for Pi21 obtained by CRISPR/Cas9-based targeted mutagenesis showed increased resistance to rice blast without changes in major agronomic traits[131].Macovei et al.[132]developed new sources of resistance to rice tungro spherical virus by mutagenesis of the endogenous eukaryotic translation initiation factor 4G(eIF4G)in rice plants.Genome editing strategies for multiplexed recessive susceptibility genes have great potential to yield superior cultivars with broad-spectrum resistance.

Compared with the widespread application of gene-editing technologies in disease resistance,progress in insect-pest resistance has been slower,perhaps owing the more complicated resistance mechanism of the latter.Only a few brown planthopper(BPH)resistance genes have been cloned.These BPH genes conferred BPH resistance,resulting in difficulty in their use in rice cultivars via CRISPR/Cas9.Recently,Lu et al.[133]used CRISPR/Cas9 technology to knock out the cytochrome P450 gene CYP71A1,encoding tryptamine 5-hydroxylase,which catalyzes the conversion of tryptamine to serotonin,thereby leading to rice resistance to insect pests via suppression of serotonin biosynthesis.This study gives promise for insect-resistant rice breeding and suggests a direction for the use of gene-editing technologies to develop insect-resistant cultivars in other crops.

Fig.3.Genome editing to improve rice resistance.(A)represents the process of obtaining resistant plants using the CRISPR/Cas system.The blue line is a target sequence,red line is sgRNA,yellow spots in the Petri dish are callus,seedlings surviving in dishes under selection pressure are resistant plants.(B-D)Phenotype comparison of control(wild-type plants)with resistant plants.(B)Herbicide resistance;(C)Disease resistance;(D)insect pest resistance.

5.Exploitation and utilization of heterosis

Hybrid rice breeding for exploiting hybrid vigor,or heterosis,has greatly boosted rice yield.In the”three-line”and”two-line”hybrid systems,the male-sterile line is the foundation used to exploit the heterosis of hybrid rice.However,development of male-sterile lines by conventional breeding is a time-consuming and laborious process.CRISPR-based genome editing technologies not only accelerate the breeding of male-sterile lines but facilitate the exploitation of heterosis.Shen et al.[134]rapidly created marker-free photoperiod-/thermosensitive genic male sterile(P/TGMS)rice materials by editing the male fertility gene PTGMS2-1 in two widely compatible rice cultivars.Japonica photosensitive genic male-sterile rice lines were developed by targeted editing of the Carbon Starved Anther gene in japonica rice using CRISPR/Cas9[135,136].Thermosensitive male sterile lines were also created by mutagenesis of the TMS5 gene by the CRISPR/Cas9 system[137,138].Li et al.[139]developed disease-resistant thermosensitive male-sterile rice by simultaneous genome engineering of the TMS5,Pi21,and Xa13 genes.Knockout of the SaF and SaM alleles by CRISPR/Cas9 produced hybrid-compatible lines that overcame Sa-mediated hybrid male sterility in rice[140].Recently[141-143],a new system of third-generation hybrid rice technology was created by clever use of a cytoplasmic sterility gene(ORFH79)and a genic male sterile gene(CYP703A3)via CRISPR/Cas9.Using this third-generation hybrid rice system,stable sterility and free combination genic male sterility(GMS)rice materials were obtained and have been used to breed superior combinations of hybrid rice.GMS rice can fill the gaps in low resource utilization for the cytoplasmic male-sterility system(first generation of hybrid rice)and susceptibility to environmental effects of the P/TGMS system(second generation of hybrid rice).

Hybrid ric seeds cannot be replanted.Breeders must produce hybrid seeds on a yearly basis,a costly and laborious task.Finding a way to bypass genetic segregation and allow hybrids to propagate clonally through their own seeds has been pursued by breeders worldwide.This is also the highest strategic goal of hybrid rice breeding.The emergence of modern tools,such as CRISPR/Cas9,has brought hope for the prospect of achieving asexual reproduction and fixing hybrid vigor.The combination of genome editing to substitute mitosis for meiosis(MiMe)with the expression of BABY BOOM1(BBM1)in the egg cell can yield clonal progeny that retain genome-wide parental heterozygosity[144].Wang et al.[145]produced clonal seeds from hybrid rice by simultaneous genome engineering of meiosis genes(REC8,PAIR1,and OSD1)and a fertilization gene(MTL)[146],in which knockout of MTL induces haploid seed formation.Although these synthetic-apomictic strategies cannot yet be commercialized for rice because of low fertility,they open a promising path for the application of‘‘one-line”hybrid rice.

6.Conclusions and prospects

During the last decade,CRISPR/Cas-based genome editing techniques have revolutionized biotechnology.Genome-editing technology coupled with functional genomics has greatly advanced the genetic improvement of rice.Progress has been made in creating rice genotypes with higher yield,improved grain quality,and resistance to herbicides,diseases,and insect pests.Such improved rice cultivars might play an important role in addressing theincreasing global food demand in the face of population growth and climate change.However,several challenges remain.

First,nonhomologous end joining(NHEJ),which causes mainly random insertions or deletions(indels)near the target sequence,is the predominant repair pathway.HDR-mediated site-specific transgene integration or point mutations remains technically challenging because of its low efficiency.In addition,to date,the protospacer-adjacent motif sequences recognized by Cas variants still cannot cover the whole-genome sequence,somewhat limiting the application of CRISPR in rice.These are challenges that will need to be addressed if the application of CRISPR/Cas technology to rice breeding is to be further expanded.

Second,the application of the CRISPR/Cas technology to rice breeding has been achieved mainly via knockout of unfavorable genes to produce desired traits,an approach that cannot meet the needs of breeders.Many cloned yield-increasing functional genes or regulators in rice have not been effectively used in breeding via CRISPR/Cas technology.There is an urgent need to develop efficient CRISPR/Cas-mediated targeted-insertion and chromosomerearrangement technologies to combine or‘‘stack”superior alleles.

Given the increasing attention being paid to transgenic food,caution must be exercised in the commercial deployment of genome-edited rice cultivars.Although genome-integrated exogenous DNA fragments can be eliminated by selfing or hybridization,undetected foreign elements may remain in the genome and might lead to safety problems.Recently,Liu et al.[147]developed a userfriendly,highly sensitive,open-source tool able to detect tens of thousands of exogenous DNA fragments using whole-genome sequencing data,and such a tool could ensure the safety of genome-edited products in agriculture.With the use of foreignelement detection tools,genome editing technology will have broader application prospects in rice breeding.

CRediT authorship contribution statement

Kejian Wangconceived and designed the manuscript,Yong HuangandHuirong Dongwrote the initial manuscript,Meiqi Shangassisted with drawing the figures,Kejian Wangrevised the manuscript,and all authors read and approve the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China(U20A2030),Central Public-interest Scientific Institution Basal Research Fund(Y2020YJ12 and Y2020XK17),Key Research and Development Program of China National Rice Research Institute(CNRRI-2020-01),and Foreign Cooperation Project of Ningxia Academy of Agricultural and Forestry Institute(DW-X-2018004).