Jun Li, Jing Yng, Yn Li, Ligeng M,*

aBeijing Key Laboratory of Plant Gene Resources and Biotechnology for Carbon Reduction and Environmental Improvement, College of Life Sciences, Capital Normal University, Beijing 100048, China

bState Key Laboratory of North China Crop Improvement and Regulation, College of Life Sciences, Hebei Agricultural University, Baoding 071001, Hebei, China

Keywords:Wheat Functional genomics Gene cloning Agronomical trait Rational design-based molecular breeding

A B S T R A C T The characterization of agronomically important genes has great potential for the improvement of wheat. However, progress in wheat genetics and functional genomics has been impeded by the high complexity and enormous size of the wheat genome.Recent advances in genome sequencing and sequence assembly have produced a high-quality genome sequence for wheat. Here, we suggest that the strategies used to characterize biological mechanisms in model species, including mutant preparation and characterization, gene cloning methods, and improved transgenic technology, can be applied to wheat biology. These strategies will accelerate progress in wheat biology and promote wheat breeding program development. We also outline recent advances in wheat functional genomics. Finally, we discuss the future of wheat functional genomics and the rational design-based molecular breeding of new wheat varieties to contribute to world food security.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880

2. Strategies used in wheat biology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880

2.1. Mutagenesis,mutant preparation,and characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880

2.2. Gene cloning methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881

2.2.1.Progress in wheat genome sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881

2.2.2.Traditional gene cloning methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882

2.2.3.Genetic mapping based on sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882

2.2.4.MutMap-based cloning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884

2.3. Genetic transformation of wheat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885

3. Advances in wheat functional genomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885

3.1. Vernalization and flowering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885

3.2. Disease resistance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886

3.3. Wheat yield-and quality-related genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886

3.4. Male sterility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886

3.5. Nutrient uptake in wheat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886

4. Conclusions and future perspectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887

Declaration of competing interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887

Author contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887

1. Introduction

Wheat (Triticum aestivum L.) is an important staple food crop worldwide; it is cultivated in very different environments and serves as a food source for 30%of the human population[1].It is estimated that the world's population will reach 9 billion by 2050;thus, a substantial increase in wheat yield is needed to ensure future global food security [2]. The grain quality of wheat also requires improvement; however, conventional breeding approaches have reached their limit in terms of increasing grain quality and yield through the enhancement of current wheat varieties.Therefore,a new breeding strategy for wheat is needed.

Rational design-based molecular breeding takes advantage of the genetic basis of agronomically important traits;thus,it is predicted to increase selection accuracy and shorten breeding times [3]. Indeed, scientists have successfully used this strategy to develop super-rice lines with high yields and superior-quality grains in less than five years, which is significantly shorter than the time that would be required using traditional breeding methods[4].Thus,rational designbased molecular breeding strategies hold great promise for crop breeding.

The success of rice breeding using molecular design is mainly due to advances in our understanding of the genetic basis of important agronomic traits; this has enabled scientists to create different trait combinations and produce numerous elite lines [5,6]. To utilize this strategy to create desirable new wheat cultivars, the most important step is to identify genes that control important agronomic traits. Much effort has been made to uncover the genetic basis for important traits in allohexaploid wheat; however, due to its complex and monstrously large genome, our ability to conduct functional genomic analyses in wheat is limited compared to that in rice or maize. Nevertheless, given recent progress in genome sequencing and assembly technologies,draft and near-complete genome assemblies of wheat have been generated [7] that will be useful for the functional annotation of the wheat genome.

In this review, we propose that the strategies used for functional genomic analysis in other model organisms should be adopted by wheat researchers. These strategies include revealing the biological function of genes by characterizing their corresponding mutants,identifying the gene responsible for an interesting phenotype by gene cloning, and verifying the function of a target gene by transforming the candidate wild-type gene into the corresponding wheat mutant.Additionally, gene cloning will enable the characterization of the molecular mechanisms in wheat, and the cloned genes will be used for rational design-based molecular breeding of wheat varieties. We also summarize recent progress in determining the sequence of the wheat genome and in functional genetic studies of wheat.

2. Strategies used in wheat biology

Modern biology has developed rapidly by following a strategy of mutant preparation and characterization, cloning of the gene responsible for a mutant phenotype,and transformation to confirm the identified candidate gene.However,due to the enormous size and complexity of the wheat genome and difficulties with the genetic transformation of wheat, certain strategies used in other model organisms have not been widely applied to wheat research. Consequently, our understanding of wheat biology is limited. Still, given recent advances in genome sequencing and assembly, as well as improved methods for gene cloning and transformation in wheat, it is time for wheat biologists to utilize the same research tools as their colleagues in other fields.

2.1. Mutagenesis, mutant preparation, and characterization

Mutants are the major driving force in genetics; they are critical for the analysis of gene function and the genetic improvement of crops.As species evolve,natural mutants are generated; such mutants have played an important role in functional genomics and molecular breeding [8,9]. For example, the global wheat yield has been greatly improved by the adoption of a natural mutant of the ‘green revolution’ gene[10].However,the probability of natural variation is extremely low at 10?5to 10?8events per base pair per generation in higher plants [11]. Therefore, various strategies for mutagenesis, including chemical mutagenesis (e.g., using ethyl methanesulfonate [EMS] or sodium azide) and physical mutagenesis (e.g., using fast neutron or gamma radiation),have been utilized to generate random mutations on a genome-wide scale in a variety of plant species.The resulting populations are then screened for mutants with an interesting phenotype,and the gene thought to be responsible for this phenotype is revealed by mutant characterization. Notably,these mutants are also valuable resources for crop improvement.

EMS is a commonly used chemical agent for seed mutagenesis[12]. The mutations it creates are high in density and randomly distributed in the genome, usually as single nucleotide changes (G to A or C to T) [13]. Several groups have generated EMS-mutagenized populations of different wheat varieties and screened and characterized the resulting mutants of interest [14,15]. EMS-based mutagenesis offers several advantages to researchers, including high-efficiency,large-scale, irreversible, and non-transgenic mutagenesis.

When seeds are bombarded by ionizing particles such as fast neutrons or gamma rays,their DNA is damaged by strand breaks. Seeds sent to space can also be mutated by cosmic radiation. The patterns observed in such mutants include large DNA sequence deletions and chromosomal rearrangements of different sizes, ranging from one base pair to millions of base pairs [16]. This approach has produced new plant varieties with improved properties, but it is difficult to locate the mutations in the genome.

These methods of mutagenesis have proven to be effective at generating mutants that can be used to characterize the relationship between genotype and phenotype. However, the mutagenesis is random; therefore, large-scale screens for the intended mutants are necessary.In addition,common wheat,as an allohexaploid plant,has three copies of most genes in its genome. This results in functional redundancy among the homologous genes from the three subgenomes such that no phenotype may be observed in response to the mutation of a single homolog.However,the recent development of genome editing technologies, especially clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated (Cas) systems, has enabled the production of targeted mutations in wheat [17–20]. Such approaches will play an important role in uncovering the functions of wheat genes.However, the application of CRISPR/Cas systems relies on sequence information about the targeted gene. Therefore,random mutagenesis and CRISPR/Cas-induced targeted mutations should be considered complementary approaches for wheat mutant production.

2.2. Gene cloning methods

After characterizing the phenotype of an interesting mutant,the next step is to isolate the affected gene responsible for the observed phenotype.As genomic sequence information is the basis for gene cloning, we will first provide a summary of recent advances in wheat genome sequencing(Table 1)before introducing the methods used for gene cloning.

2.2.1.Progress in wheat genome sequencing

High-quality genomic sequences and assemblies are beneficial to basic research, including gene cloning, and genetic improvement in wheat. Therefore, enormous effort has been made to study the wheat genome, and great progress has been achieved in the last few years. Due to the large genome size of bread wheat, sequencing and mapping the expressed portion of the genome is a priority for gene discovery.

In 2005, the International Wheat Genome Sequencing Consortium (IWGSC), with 2400 members from 68 countries,was established to obtain a high-quality genome sequence of hexaploid bread wheat. Because of the complexity and high similarity among the three subgenomes(A,B,and D;2n=6x=42), the 21 chromosomes of wheat (Chinese Spring) were separated by flow cytometry; bacterial artificial chromosome(BAC)libraries were then constructed for each chromosome or chromosome arm,and the construction of physical maps and sequencing of the BACs were undertaken by IWGSC members[21].A physical map of 3B, the largest chromosome in wheat,was generated using a chromosome-based BAC-by-BAC sequencing strategy[22].Later,the sequences of the chromosomes(or parts of the chromosomes)were released.

Table 1–Recent advances in wheat genome sequencing.

Thanks to these technological advances in sequencing and sequence assembly, complete genomic sequences and assemblies of hexaploid bread wheat (Chinese Spring) and its diploid ancestors (Triticum urartu and Aegilops tauschii) have been reported [23–25]. They have provided a snapshot of the wheat genome and its gene loci; however, they are highly fragmented and contain a large number of unordered scaffolds[27].Using chromosome-based sequencing and shotgun assembly, an ordered draft sequence of bread wheat was prepared; most genes were evenly distributed along the homeologous chromosomes, and highly similar and conserved gene copies were retained in each subgenome [1].Meanwhile, a high-quality reference sequence of chromosome 3B [28] was published that features much more contiguity than the sequence reported by the IWGSC[1].

Several technological and conceptual achievements in genome sequencing and assembly have vastly improved our understanding of the wheat genome. For example, a highquality genome sequence of wild emmer (Triticum turgidum)was obtained [29] and validated using genetic and Hi-C data.Also, an ordered and annotated genome sequence of bread wheat (IWGSC RefSeq v1.0) was released. The quality and contiguity of IWGSC RefSeq v1.0, assessed through alignments with radiation hybrid maps of the subgenomes and independent data from coding and noncoding sequences,exceed those of previous sequences[32].Very recently,IWGSC RefSeq v2.0(i.e.,version 2 of the reference sequence of wheat)has become available (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies). Apart from these draft and near-complete genome sequences of wheat, major advances in the isolation and characterization of genes endowing important agronomic traits for the molecular breeding of wheat are in the pipeline.

2.2.2. Traditional gene cloning methods

In the pre-reference genome era, only genetic information was available for a great number of genes responsible for agronomically important traits in crops. Map-based cloning,or positional cloning, is a useful method to clone genes of interest. The creation of high-resolution mapping populations, genotyped with molecular markers and phenotyped by the trait of interest, is a prerequisite. Using this strategy, one can gradually narrow down the location of a mutated gene by utilizing mapping populations consisting of a few hundred or thousand plants generated through genetic recombination(Fig. 1) [27,33]. By combining the phenotypic and molecular marker data, a genetic map covering the target locus is produced. Then, the corresponding physical map is pieced together by chromosome walking, in which newly developed closest flanking markers are used to screen a BAC library and subclones of the identified BACs [16]. A contig spanning the target locus is established, and the candidate gene is identified and verified by sequencing [34].

Map-based cloning even works when no information about the gene product is available.It has been successfully used to clone a great number of genes,including Xa21,HM1,and Lr10 from rice, maize, and wheat, respectively [8,35,36]. Notably,nearly half of these genes are disease resistance genes(Table 2). However, map-based cloning is laborious and timeconsuming. Moreover, when the genes are located in centromeric regions, it is impossible to clone them using a map-based strategy[27].

2.2.3.Genetic mapping based on sequencing

With the rapid advances in sequencing and bioinformatic technologies, innovative and accelerated gene cloning strategies, including homology-based cloning, mutational mapping (MutMap), mutant chromosome sequencing(MutChromSeq), targeted chromosome-based cloning via long-range assembly (TACCA), and mutagenesis resistance gene enrichment and sequencing (MutRenSeq), have been created to overcome the limitations of traditional gene cloning methods.

Given that genes are evolutionarily conserved,orthologous genes in different species may have similar sequences and conserved functions. Homology-based cloning, which relies on sequence similarities between known and target genes,has been used successfully to isolate wheat genes. For example, the wheat rht-1 mutant exhibits a short stature,impaired responsiveness to gibberellin,and increased gibberellin levels in planta, similar to maize d8 and Arabidopsis gai;based on these similarities, the DNA sequence of Rht-D1 in wheat was obtained [10]. However, when no relevant sequence information is available in other species, it is impossible to clone a target gene using this method.

MutMap is a rapid gene isolation method based on mutagenesis, gene mapping, and whole-genome resequencing. In MutMap, a plant with an interesting phenotype is selected, crossed to its wild-type parent, and then selfed. The F2progeny will exhibit mutant and wild-type phenotypes. Bulked DNAs from F2progeny displaying the mutant phenotype are subjected to whole-genome sequencing, and the sequences are aligned to the reference genome.Single nucleotide polymorphisms (SNPs) showing an SNP index of 1 are thought to harbor the gene responsible for the mutant phenotype [85]. This approach has been used to identify the genomic positions of genes controlling agronomically important traits in rice. For species with a relatively small genome and a high-quality reference genome, MutMap is a good choice for isolating mutated genes.

In comparison, MutChromSeq is a fast method for gene cloning from large genomes. It combines mutagenesis, the reduction of genome complexity by chromosome sorting,and high-throughput sequencing. A candidate gene with a clear phenotype is identified through sequencing and a comparison of wild-type and mutant chromosomes. This method, which does not require positional fine mapping of genes,was used to clone the barley Eceriferum-q gene and wheat Pm2 gene [64].When the requirements of the method(i.e.,mutagenesis,the identification of target genes conferring strong phenotypes,and the isolation of individual chromosomes) are met, it is possible to clone previously intractable genes.

TACCA enables rapid gene cloning from complex polyploid genomes through the use of flow-sorted chromosomes to reduce genome complexity,next-generation DNA sequencing,and cultivar-specific de novo assembly. Using this method,the wheat leaf rust resistance gene Lr22a was isolated from bread wheat[67].When chromosome isolation is feasible and a preliminary gene map is available, TACCA is a good option for gene isolation.

Fig.1–Overview of the isolation of mutated genes from a complex genome by MutMap-based cloning.Seeds of a cultivar are mutagenized by EMS,resulting in M1 plants,which are selfed to generate a large number of M2 seeds.An M2 plant with an interesting phenotype is selected and crossed to its wild-type parent.In the F2 generation,plants exhibiting mutant and wildtype phenotypes are segregated. Bulked DNA from F2 plants displaying interesting phenotypes and wild-type plants is then subjected to whole-genome sequencing(WGS).SNPs between the wild-type and mutant plants are identified through MutMap analysis;these SNPs are used as markers for traditional map-based cloning to narrow down the region of interest.After one to several rounds of mapping,the target gene is pinned down to a small interval.

MutRenSeq is a rapid gene cloning method for the isolation of nucleotide binding and leucine-rich repeats (NLR) genes. It involves chemical mutagenesis,exome capture,and sequencing. Since most resistance genes encode proteins with NLRs,exome capture is used to enrich the NLR-specific bait library and sequencing of the resistant wild-type parent and susceptible loss-of-function mutants is performed. The mutant reads are aligned with those from the parent, allowing the identification of genes responsible for the resistance. This method, which was used to clone two fungal stem rust resistance genes (Sr22 and Sr45) from bread wheat [61], does not require positional fine mapping and can be applied to isolate NLR-type resistance genes from most crops and their wild relatives.

Table 2–Some of the genes cloned and their functions in wheat.

2.2.4. MutMap-based cloning

Map-based cloning can be used to clone any gene; it works even when no information about the gene product is available. However, it is laborious, time-consuming, and requires a large number of markers to narrow down the chromosomal location of the target gene.Thus,occasionally it is impossible to isolate a gene using map-based cloning due to a lack of molecular markers. Considering that map-based cloning is sometimes unusable and MutMap is only suitable for organisms with small genomes, we developed MutMapbased cloning, a rapid gene cloning method for use with complex genomes (Fig. 1). It combines MutMap with traditional map-based cloning[15].

MutMap-based cloning consists of EMS mutagenesis,genome sequencing, and mapping. Initially, SNPs are identified between wild-type plants and homozygous mutants(segregated from the progeny of heterozygous mutants) and the chromosomal region in which the target gene is located is determined by MutMap. Second, the SNPs identified by MutMap are used as markers to narrow down the region of interest.Eventually,the target gene is pinned down to a small interval by traditional map-based cloning using these SNP markers. Using MutMap-based cloning, we successfully mapped Ms1 to a 198-kb interval and cloned Ms1 from allohexaploid bread wheat using 676 F2plants derived from the progeny of heterozygous ms1e mutants [15].

It is estimated that the SNPs generated by EMS mutagenesis are evenly distributed throughout the genome and that the markers generated by MutMap analysis are sufficient for mapping (usually 3–5 SNPs per 1 Mb). Therefore, the required mapping population is smaller than that in traditional mapbased cloning.Furthermore,the mapping population is built from the progeny of heterozygous mutants.This saves one to several generations of wheat breeding time,and it simplifies the genetic background (compared to crossing with different varieties);therefore, it shortens the experimental time and overcomes potential problems due to genetic complexity arising from crosses between varieties. MutMap-based cloning can be used to clone previously intractable genes from any organism(especially organisms with complex genomes),including wheat,which lacks a sufficient marker density to allow the region containing a gene of interest to be narrowed down.

The above-mentioned gene cloning methods exhibit distinct advantages and disadvantages, and each method works in at least some cases (Table 2). However, in general, those methods that take greater advantage of recent advances in genome sequencing have greater potential for high-efficiency gene cloning.

2.3. Genetic transformation of wheat

Following the isolation of a candidate gene from a mutant,the gene should be confirmed to be responsible for the observed phenotype by transforming the corresponding wild-type gene into mutant plants. Biolistic transformation and Agrobacterium-mediated genetic transformation are the two most commonly used methods to stably introduce foreign genes of interest into plant cells[86].The ability to transform plants successfully is a powerful advantage in studies of gene function.

Wheat, however, is difficult to transform and regenerate,so progress in genetic engineering,basic research,and applied genetics has lagged behind that in other plants. After significant effort, however, the first fertile transgenic wheat line was obtained by the microprojectile bombardment of an embryogenic callus [87]. Later, the first Agrobacteriummediated transformation of wheat was reported [88]. However, the efficiency of the process was very low. Since then,significant progress has been made to improve the efficiency of wheat transformation(e.g.,better DNA delivery methods,a broader choice of target tissues, improved culture media components,and new selectable marker genes)[89,90].

Transformation by biolistic bombardment is preferred over Agrobacterium-based methods,mainly because success is less genotype-dependent and the efficiency is higher (approximately 10%). However, the procedure is complex, with multiple copies of the transgene in the host genome and occasional integration of the vector backbone into the genome. In contrast, Agrobacterium-mediated transformation is simple to perform and is cost-effective; moreover, it produces a low copy number, intact integration of foreign genes into the host genome,and stable transgene expression.Recently, the Japan Tobacco Company developed an efficient protocol in which transgenic wheat plants could be obtained from immature embryos of the cultivar Fielder by Agrobacterium-mediated transformation with high efficiency[91]. This protocol was later successfully applied with modifications to additional cultivars[20,92].

The strategies mentioned above have been used to characterize several model species,and they are being applied to wheat functional genomics and wheat biology. Given the many experimental methodologies available,it is time for the scientists who study wheat to use these strategies in their research. This will accelerate progress in wheat functional genomics and wheat biology and promote the development of new alleles for wheat breeding.

3. Advances in wheat functional genomics

Elucidating the functions of genes responsible for important agronomic traits is the most pressing issue in the postgenomic era.However,due to the complex,large genome of bread wheat, progress in wheat biology has fallen behind that in other crops. Still, significant progress in wheat functional genomics has been made recently. Indeed, a growing number of genes controlling, but not limited to,vernalization and flowering, disease resistance, yield and quality improvements, male sterility, and nutrient uptake efficiency have been cloned and elucidated, making rational design-based molecular breeding in wheat possible. Here, we provide a brief introduction to recent achievements in wheat functional genomics(Table 2).

3.1. Vernalization and flowering

Winter wheat requires extended exposure to cold temperatures to induce flowering. This biological process, called vernalization,ensures that wheat plants will not flower before winter. The winter growth habit of winter wheat is determined by both VRN2, which represses flowering, and VRN1,which promotes flowering after vernalization[39,41].

VRN2 is a dominant repressor of flowering. After vernalization, the expression of VRN2 is down-regulated, releasing the expression of VRN3,an ortholog of Arabidopsis FLOWERING LOCUS T(FT),which encodes a protein that is transferred from leaves to the shoot apical meristem[42].FT participates in the floral activation complex, which induces the expression of VRN1 and promotes the transition of the apical meristem from the vegetative stage to the reproductive stage [42].Recently, VRN-D4 was identified in the ancient wheat from South Asia, and mutations in VRN-D4 were shown to cause delayed flowering[55].

For proper flowering induction, VRN1 mRNA must accumulate in response to prolonged cold exposure. A novel mechanism for VRN1 induction during vernalization was recently revealed. Under non-vernalization conditions, GRP2 binds to VRN1 pre-mRNA to repress its accumulation.Vernalization induces VER2 expression and elevates the level of O-GlcNAcylated GRP2(GRP2-G).Next,phosphorylated VER2 interacts with GRP2-G, releasing the repressive effect of GRP2 on VRN1 and resulting in VRN1 mRNA accumulation [93].Moreover,tae-miR408-mediated expression of TIMING OF CAB EXPRESSION is needed for the regulation of heading time[94].

3.2. Disease resistance

Fungus-induced diseases, including rusts, powdery mildew,head blight/scab, and septoria tritici blotch (STB), reduce wheat yields around 25% per annum [80] and could threaten world food security. Enormous effort has been made to identify genes responsible for disease resistance in order to breed durable and broadly disease-resistant wheat varieties.

Rusts, including leaf rust, stripe rust, and stem rust, are the most widespread and devastating diseases in wheat.Lr34 confers resistance to multiple rusts and a mildew.Although it has been used for wheat improvement since the 20th century, it wasn't until 2009 that the Lr34 gene,encoding an ATP-binding cassette transporter,was cloned.The LR34 transporter functions in adult plants;similar to the pleiotropic drug resistance protein PEN3 in Arabidopsis, it confers resistance to pathogens by exporting metabolites to affect fungal growth [47]. Yr36, which provides broad resistance to stripe rusts, was also cloned. It contains a functional enzymatic (kinase) domain and a putative START domain,both of which are necessary for disease resistance[48].Very recently, Yr28, which is an ancestral NB-LRR and confers stripe rust resistance in wheat, was cloned [83]. Further, two genes (Sr33 and Sr35) conferring resistance to wheat stem rust race Ug99 have been cloned from A. tauschii and Triticum monococcum,respectively[52,53].

Powdery mildew, a destructive disease caused by Blumeria graminis, can reduce wheat yields dramatically. To date, 78 powdery mildew resistance alleles and 50 loci have been identified and assigned designations, respectively. However,only a few genes have been molecularly and functionally characterized. For instance, Pm21 encodes a CC-NBS-LRR protein that confers broad-spectrum resistance to powdery mildew [77,78], while Pm60, cloned from T. urartu, was recently found to encode an NB-LRR protein that plays a major role in Bgt E09 resistance[76].

STB caused by Zymoseptoria tritici could result in economic losses worldwide. There are 21 Stb resistance genes in wheat.However,only one gene,Stb6,has been cloned and characterized.It encodes a conserved wall-associated receptor-like kinase that confers resistance to the pathogen without a hypersensitive response[75].

3.3. Wheat yield- and quality-related genes

As the global population continues to increase and living standards improve, it will be essential to identify and characterize genes regulating grain yield and quality to promote the rational design-based molecular breeding of new wheat varieties.

The stems of tall wheat are not strong enough to support heavy grains (they fall over). The introduction of dwarfing or semi-dwarfing genes into wheat is thus crucial to increase world wheat yields. The Rht gene of wheat has been isolated and characterized.It encodes a transcription factor containing an SH2-like domain that participates in gibberellin signaling[10]. New varieties carrying mutated versions of Rht (and which respond abnormally to gibberellin) are short and exhibit increased grain yields; such plants have enabled the Green Revolution. Recently, the mutant allele gw2-A1 was characterized;it increases the thousand grain weight of wheat by producing wider and longer grains[95].

Wheat grain quality is affected by many traits, including starch composition,nutritional value,and end-use,which are determined by numerous genes.However,only a few relevant genes have been cloned. For example, the NAC gene Gpc-B1 was mapped to chromosome arm 6BS and cloned.It encodes a transcription factor that regulates senescence and is responsible for increasing nutrient remobilization, which in turn improves the protein,zinc,and iron contents of wheat grains[43].Wheat lines carrying mutations in starch branching enzyme II genes have elevated levels of amylose and resistant starch,which are beneficial for human health[96].

3.4. Male sterility

Male sterility is a valuable agricultural trait for hybrid seed production. Although five genic male sterility (Ms) loci have been identified in wheat, only Ms1, Ms2, and Ms5 have been cloned and characterized.

Ms1 is a recessive gene cloned by two independent groups.Only Ms1 from the B subgenome is expressed; its homologs from the A and D subgenomes are epigenetically silenced.Ms1, which is specifically expressed in microsporocytes,encodes a phospholipid-binding protein that is necessary for pollen exine development and male fertility[15,74].

The ms2 mutant is male sterile but lacks other unfavorable agronomic phenotypes; thus, it has been extensively used to breed numerous wheat cultivars in China. In 2017, the molecular characterization of Ms2 was reported. A TRIM insertion gives rise to the activation of Ms2, which encodes an orphan protein.It is specifically expressed in the anthers of ms2 plants, where it confers dominant male sterility [72,73].Ms5 has also been cloned. It is a recessive gene expressed from subgenomes A and D that encodes a glycosylphosphatidylinositol-anchored lipid transfer protein necessary for pollen exine development and male fertility[82].

3.5. Nutrient uptake in wheat

Fertilizers containing nitrogen (N) and phosphorus (P) contribute to improved crop yields, including wheat yields.However, the widespread overuse of fertilizers has reduced the nutrient use efficiency of wheat plants and created adverse environmental effects. Therefore, it is important to identify genes regulating N and P use in wheat in order to understand nutrient uptake and further improve use efficiency.

N and P use efficiency is complex and is influenced by diverse physiological traits. Only a few genes, including NAC and NFYB, regulating N and P use in wheat have been identified.For example,in a genome-wide analysis performed to identify TAA1/TAR genes in wheat, 15 TaTAR genes were identified, among which TaTAR2.1 is mainly expressed in roots [70]. TaTAR2.1 expression is up-regulated under low-N conditions, promoting root growth. Therefore, TaTAR2.1 may play an important role in improving wheat yields and N use efficiency.P is also a macronutrient essential for plant growth and crop yields. The knockout of TaPHO2-A1 in wheat could improve P uptake and wheat yields under low-P conditions[97]. As the genes regulating N and P use in wheat are identified, our understanding of nutrient uptake and the N and P use efficiencies of wheat will increase.This in turn may lead to increased wheat yields with less fertilizer input.

4.Conclusions and future perspectives

Rational design-based molecular breeding, which involves the characterization and utilization of agronomically important genes,has great potential for wheat improvement.Considerable effort has been made to elucidate the functions of genes conferring important agronomic traits to wheat. However,progress in functional wheat genomics and wheat biology has been hindered by the complex and monstrously large genome,which is 135 times larger than that of Arabidopsis and 37 times larger than that of rice. As the technologies used for genome sequencing and sequence assembly have advanced and the cost of sequencing has fallen,a near-complete assembly of the bread wheat genome has been generated [32]. In addition, significant progress in genetic mapping techniques and wheat transformation methods has been made over the last two decades. As a result,a growing number of wheat genes have been cloned and dissected(Table 2).

Wheat is an allohexaploid species containing three distinct subgenomes (AABBDD), and most genes show functional redundancy. Recent developments in genome editing have been especially useful in studies of wheat because they have enabled targeted mutations to be produced in all copies of a gene (i.e., not only in a single gene). For instance, the simultaneous knockout of all three copies of Mlo confers strong resistance in wheat to powdery mildew[17].Therefore,this approach offers a shortcut to relate phenotypes to sequence variants or to inactivate genes conferring undesirable traits and thus enable the genetic improvement of wheat.

Several powerful tools are now available to analyze the functions of wheat genes, including mutagenesis, highquality genome sequences,extended genetic maps,improved transgenic technology, and advanced genome editing techniques. The combined use of random mutagenesis and genome editing should enable the mutation of any gene in the wheat genome; other approaches can then be applied to reveal the biological function of the mutated gene.

Strategies and techniques are now in place that can greatly advance wheat biology. The entry of scientists who study other model plants or bioinformatics into the wheat biology field will greatly promote progress in wheat functional genomics. We predict that many more agronomically important genes, which are valuable resources for molecular breeding, will be isolated and characterized by applying the strategies used for functional genomic analysis in model organisms.We also believe that rational design-based molecular breeding is a promising avenue for creating elite wheat varieties that will promote future global food security.

Declaration of competing interest

The authors declare no conflicts of interest.

Acknowledgments

We thank Dr. Jessica Habashi for critical reading of the manuscript. This work was financially supported by the National Key Research and Development Program of China(2017YFD0101001),the Beijing Municipal Government Science Foundation, China (IDHT20170513), and the Starting Grant from Hebei Agricultural University, China(YJ201958).

Author contributions

Ligeng Ma conceived the idea;Jun Li and Ligeng Ma wrote the paper;Jing Yang gathered data;Yan Li prepared the figure.All authors contributed to the revisions.