Abstract

Wheat (Triticum aestivum L.), with a large genome (16000 Mb) and high proportion ( 80%) of repetitive sequences, has been a difficult crop for genomics research. However, the availability of extensive cytogenetics stocks has been an asset, which facilitated significant progress in wheat genomic research in recent years. For instance, fairly dense molecular maps (both genetic and physical maps) and a large set of ESTs allowed genome-wide identification of gene-rich and gene-poor regions as well as QTL including eQTL. The availability of markers associated with major economic traits also allowed development of major programs on marker-assisted selection (MAS) in some countries, and facilitated map-based cloning of a number of genes/QTL. Resources for functional genomics including TILLING and RNA interference (RNAi) along with some new approaches like epigenetics and association mapping are also being successfully used for wheat genomics research. BAC/BIBAC libraries for the subgenome D and some individual chromosomes have also been prepared to facilitate sequencing of gene space. In this brief review, we discuss all these advances in some detail, and also describe briefly the available resources, which can be used for future genomics research in this important crop.

1. Introduction

Wheat is one of the most important staple food crops of the world, occupying 17% (one sixth) of crop acreage worldwide, feeding about 40% (nearly half) of the world population and providing 20% (one fifth) of total food calories and protein in human nutrition. Although wheat production during the last four decades had a steady significant increase, a fatigue has been witnessed during the last few years, leading to the lowest current global wheat stocks ever since 1948/49. Consequently, wheat prices have also been soaring, reaching the highest level of US $ 10 a bushel as against US $ 4.50 a year ago (http://www.planetark.com/dailynewsstory.cfm/newsid/44968/story.htm). As against this, it is projected that, in order to meet growing human needs, wheat grain production must increase at an annual rate of 2%, without any additional land to become available for this crop [1]. In order to meet this challenge, new level of understanding of the structure and function of the wheat genome is required.

Wheat is adapted to temperate regions of the world and was one of the first crops to be domesticated some 10000 years ago. At the cytogenetics level, common wheat is known to have three subgenomes (each subgenome has 7 chromosomes, making ) that are organized in seven homoeologous groups, each homoeologous group has three closely related chromosomes, one from each of the three related subgenomes. The diploid progenitors of the A, B, and D subgenomes have been identified, although there has always been a debate regarding the progenitor of B genome (reviewed in [1]). It has also been found that common wheat behaves much like a diploid organism during meiosis, but its genome can tolerate aneuploidy because of the presence of triplicate genes. These features along with the availability of a large number of aneuploids [particularly including a complete set of monosomics, a set of 42 compensating nullisomic-tetrasomics and a complete set of 42 ditelocentrics developed by Sears [2]] and more than 400 segmental deletion lines [developed later by Endo and Gill [3]] facilitated greatly the wheat genomics research.

Molecular tools have recently been used in a big way for cytogenetic studies in wheat, so that all recent cytogenetic studies in wheat now have a molecular component, thus paving the path for wheat genomics research. However, these studies in the area of molecular cytogenetics have been relatively difficult in bread wheat due to its three closely related subgenomes and a large genome (1C = >16 billion base pairs) with high proportion (>80%) of repetitive DNA. Despite this, significant progress in the area of molecular cytogenetics and cytogenomics of wheat has been made during the last two decades, thus making it amenable to genomics research. For instance, molecular maps in bread wheat, emmer wheat, and einkorn wheat utilizing a variety of molecular markers are now available, where gene rich regions (GRRs) and recombination hotspots have also been identified (for a review, see [4, 5]).

In recent years, a number of initiatives have been taken to develop new tools for wheat genomics research. These include construction of large insert libraries and development of massive EST collections, genetic and physical molecular maps, and gene targeting systems. For instance, the number of wheat ESTs has increased from a mere 5 in 1999 [6] to a massive >1 240 000 in January 2008 (http://www.ncbi.nlm.nih.gov/), thus forming the largest EST collection in any crop as a resource for genome analysis. These ESTs are being used for a variety of activities including development of functional molecular markers, preparation of transcript maps, and construction of cDNA arrays. A variety of molecular markers that were developed either from ESTs or from genomic DNA also helped to discover relationships between genomes [7] and to compare marker-trait associations in different crops. Comparative genomics, involving major crop grasses including wheat, has also been used not only to study evolutionary relationships, but also to design crop improvement programs [8]. Functional genomics research in wheat, which though lagged far behind relative to that in other major food crops like maize and rice, has also recently witnessed significant progress. For instance, RNA interference, TILLING, and “expression genetics” leading to mapping of eQTLs have been used to identify functions of individual genes [9]. This allowed development of sets of candidate genes for individual traits, which can be used for understanding the biology of these traits and for development of perfect diagnostic marker(s) to be used not only for map-based cloning of genes, but also for MAS [9, 10]. In order to sequence the GRRs of wheat genome, a multinational collaborative program named International Genome Research on Wheat (IGROW) was earlier launched, which later took the shape of International Wheat Genome Sequencing Consortium (IWGSC) [11]. This will accelerate the progress on genome sequencing and will allow analysis of structure and function of the wheat genome. Keeping the above background in mind, Somers [12] identified the following five thrust areas of research for wheat improvement: (i) genetic mapping, (ii) QTL analysis, (iii) molecular breeding, (iv) association mapping, and (v) software development. In this communication, we briefly review the recent advances in all these areas of wheat genomics and discuss their impact on wheat improvement programs.

2. Molecular Maps of Wheat Genome

2.1. Molecular Genetic Maps

Although some efforts toward mapping of molecular markers on wheat genome were initially made during late 1980s [13], a systematic construction of molecular maps in wheat started only in 1990, with the organization of International Triticeae Mapping Initiative (ITMI), which coordinated the construction of molecular maps of wheat genome. Individual groups (headed by R Appels, PJ Sharp, ME Sorrells, J Dvorak, BS Gill, GE Hart, and MD Gale) prepared the maps for chromosomes belonging to each of the seven different homoeologous groups. A detailed account on mapping of chromosomes of individual homoeologous groups and that of the whole wheat genome is available elsewhere [14]; an updated version is available at GrainGenes (http://wheat.pw.usda.gov/), and summarized in Table 1. Integrated or composite maps involving more than one type of molecular markers have also been prepared in wheat (particularly the SSR, AFLP, SNP, and DArT markers (see Table 1)). Consensus maps, where map information from multiple genomes or multiple maps was merged into a single comprehensive map, were also prepared in wheat [15, 16]. On these maps, classical and newly identified genes of economic importance are being placed to facilitate marker-assisted selection (MAS). Many genes controlling a variety of traits (both qualitative and quantitative) have already been tagged/mapped using a variety of molecular markers (for references, see [14, 17]). The density of wheat genetic maps was improved with the development of microsatellite (SSR) markers leading to construction of SSR maps of wheat [1820]. Later, Somers et al. [16] added more SSR markers to these earlier maps and prepared a high-density SSR consensus map. At present, >2500 mapped genomic SSR (gSSR) markers are available in wheat, which will greatly facilitate the preparation of high-density genetic maps, so that we will be able to identify key recombination events in breeding populations and fine-map genes. In addition to gSSRs, more then 300 EST-SSR could also be placed on the genetic map of wheat genome [2123]. However, more markers are still needed, particularly for preparation of high-density physical maps for gene cloning [24]. Availability of a number of molecular markers associated each with individual traits will also facilitate marker-assisted selection (MAS) during plant breeding.

In addition to random DNA markers (RDM), gene targeted markers (GTMs) and functional markers (FMs) are also being used in wheat to facilitate identification of genes responsible for individual traits and to improve possibilities of using MAS in wheat breeding. As a corollary, functional markers (FMs) are also being developed from the available gene sequences [10]. These markers were also used to construct transcript and molecular functional maps. Recently, microarray-based high-throughput diversity array technology (DArT) markers were also developed and used for preparing genetic maps in wheat [53, 54]. Large-scale genotyping for dozens to thousands of SNPs is also being undertaken using several high-density platforms including Illumina’s GoldenGate and ABI’s SNaPshot platforms (http://wheat.pw.usda.gov/SNP/new/index.shtml). The genotyping activity may be extended further through the use of Solexa’s high throughput and low-cost resequencing technology.

2.2. Molecular Marker-Based Physical Maps

Molecular markers in bread wheat have also been used for the preparation of physical maps, which were then compared with the available genetic maps involving same markers. These maps allowed comparisons between genetic and physical distances to give information about variations in recombination frequencies and cryptic structural changes (if any) in different regions of individual chromosomes. Several methods have been employed for the construction of physical maps.

2.2.1. Deletion Mapping

In wheat, physical mapping of genes to individual chromosomes began with the development of aneuploids [55], which led to mapping of genes to individual chromosomes. Later, deletion lines of wheat chromosomes developed by Endo and Gill [3] were extensively used as a tool for physical mapping of molecular markers. Using these deletion stocks, genes for morphological characters were also mapped to physical segments of wheat chromosomes directly in case of unique and genome specific markers or indirectly in case of duplicate or triplicate loci through the use of intergenomic polymorphism between the A, B, and D subgenomes (see Table 2 for details of available physical maps). In addition to physical mapping of genomic SSRs, ESTs and EST-SSRs were also subjected to physical mapping (see Table 2). As a part of this effort, a major project (funded by National Science Foundation, USA) on mapping of ESTs in wheat was successfully completed by a consortium of 13 laboratories in USA leading to physical mapping of 16000 EST loci (http://wheat.pw.usda.gov/NSF/progress_mapping.html; [56] (see Table 2)).

2.2.2. In Silico Physical Mapping

As many as 16000 wheat EST loci assigned to deletion bins, as mentioned above, constitute a useful source for in silico mapping, so that markers with known sequences can be mapped to wheat chromosomes through sequence similarity with mapped EST loci available at GrainGene database (http://wheat.pw.usda.gov/GG2/blast.shtml). Using the above approach, Parida et al. [80] were able to map 157 SSR containing wheat unique sequences (out of 429 class I unigene-derived microsatellites (UGMS) markers developed in wheat) to chromosome bins. These bin-mapped UGMS markers provide valuable information for a targeted mapping of genes for useful traits, for comparative genomics, and for sequencing of gene-rich regions of the wheat genome. Another set of 672 loci belonging to 275 EST-SSRs of wheat and rye was assigned to individual bins through in silico and wet-lab approaches by Mohan et al. [79]. A few cDNA clones associated with QTL for FHB resistance in wheat were also successfully mapped using in silico approach [81].

2.2.3. Radiation-Hybrid Mapping

Radiation hybrid (RH) mapping was first described by Goss and Harris [82] and was initially used by Cox et al. [83] for physical mapping in animals/humans. In wheat, the approach has been used at North Dakota State University (NDSU) utilizaing addition and substituition of individual D-genome chromosomes into tetraploid durum wheat. For RH mapping of 1D, durum wheat alien substitution line for chromosome 1D (DWRH-1D), harboring nuclear-cytoplasmic compatibility gene scsae was used. These RH lines initially allowed detection of 88 radiation-induced breaks involving 39 1D specific markers. Later, this 1D RH map was further expanded to a resolution of one break every 199 kb of DNA, utilizing 378 markers [84]. Using the same approach, construction of radiation hybrid map for chromosome 3B is currently in progress (S. Kianian personal communication).

2.3. BAC-Based Physical Maps

BAC-based physical map of wheat D genome is being constructed using the diploid species, Aegilops tauschii, with the aim to identify and map genes and later sequence the gene-rich regions (GRRs). For this purpose, a large number of BACs were first fingerprinted and assembled into contigs. Fingerprint contigs (FPCs) and the data related to physical mapping of the D genome are available in the database (http://wheat.pw.usda.gov/PhysicalMapping/index.html). BACs belonging to chromosome 3B are also being fingerprinted (with few BACs already anchored to wheat bins), and a whole genome BAC-based physical map of hexaploid wheat is proposed to be constructed under the aegis of IWGSC in its pilot studies (see later).

3. In Situ Hybridization Studies in Wheat

In bread wheat, in situ hybridization (ISH) involving radioactively labeled probes was initially used to localize repetitive DNA sequences, rRNA and alien DNA segments [104106]. Later, fluorescence in situ hybridization (FISH), multicolor FISH (McFISH, simultaneous detection of more than one probe), and genome in situ hybridization (GISH, total genomic DNA as probe) were used in several studies. FISH with some repeated sequences as probes was used for identification of individual chromosomes [107110]. FISH was also utilized to physically map rRNA multigene family [111, 112], RFLP markers [110, 113], and unique sequences [114116] and also for detecting and locating alien chromatin introgressed into wheat [117119].

A novel high-resolution FISH strategy using super-stretched flow-sorted chromosomes was also used (extended DNA fibre-FISH; [120122]) to fine map DNA sequences [123, 124] and to confirm integration of transgenes into the wheat genome [125].

Recently, BACs were also utilized as probes for the so called BAC-FISH which helped not only in the discrimination between the three subgenomes, but also in the identification of intergenomic translocations, molecular cytogenetic markers, and individual chromosomes [126]. BAC-FISH also helped in localization of genes (BACs carrying genes) and in studying genome evolution and organization among wheat and its relatives [110, 127, 128].

4. Map-Based Cloning in Wheat

In wheat, a number of genes for some important traits including disease resistance, vernalization response, grain protein content, free threshing habit, and tolerance to abiotic stresses have been recently cloned/likely to be cloned via map-based cloning (see Table 3). The first genes to be isolated from wheat by map-based cloning included three resistance genes, against fungal diseases, including leaf rust (Lr21; [88, 129, 130] and Lr10; [87]) and powdery mildew (Pm3b ; [94]). A candidate gene for the Q locus conferring free threshing character to domesticated wheat was also cloned [92]. This gene influences many other domestication-related traits like glume shape and tenacity, rachis fragility, plant height, spike length, and ear-emergence time. Another important QTL, Gpc-B1, associated with increased grain protein, zinc, and iron content has been cloned, which will contribute in breeding enhanced nutritional value wheat in future [96]. Cloning of three genes for vernalization response (VRN1, VRN2, VRN3) helped in postulating a hypothetical model summarizing interactions among these three genes [8991, 131].

5. EST Databases and their Uses

During the last 8–10 years, more than 1240455 wheat ESTs have become available in the public domain as in January 2008 (http://www.ncbi.nlm.nih.gov/). A number of cDNA libraries have been used for this purpose. These ESTs proved to be an enormous resource for a variety of studies including development of functional molecular markers (particularly SSRs and SNPs), construction of a DNA chip, gene expression, genome organization, and comparative genomics research.

5.1. EST-Derived SSRs

Wheat ESTs have been extensively used for SSR mining (1SSR/10.6 kb; [80]), so that in our own laboratory and elsewhere detected by author, a large number of SSRs have already been developed from EST sequences [22, 78, 80, 132134]. These EST-SSRs served as a valuable source for a variety of studies including gene mapping, marker-aided selection (MAS), and eventually positional cloning of genes. The ESTs and EST-derived SSRs were also subjected to genetic and physical mapping (see above).

Since EST-SSRs are derived from the expressed portion of the genome, which is relatively more conserved, these markers show high level of transferability among species and genera [133, 135]. However, the transferability of wheat EST-SSRs to closely related triticeae species (Triticum and Aegilops species) is higher as compared to more distant relatives such as barley, maize, rice, sorghum, oats, and rye. The EST-SSRs thus also prove useful in comparative mapping, transfer of markers to orphanage wild species, and for genetic diversity estimates [79, 132, 134, 136139].

5.2. EST-Derived SNPs and the International SNP Consortium

In recent years, single nucleotide polymorphisms (SNPs) have become the markers of choice. Therefore, with the aim to discover and map SNPs in tetraploid and hexaploid wheats, an International Wheat SNP Consortium was constituted, and comprehensive wheat SNP database was developed (http://wheat.pw.usda.gov/SNP/new/index.shtml). Approximately 6000 EST unigenes from the database of mapped ESTs and other EST databases were distributed to consortium members for locating SNPs, for designing conserved primers for these SNPs and for validation of these SNP. Considerable progress has been made in this direction in different laboratories; the project data are accessible through http://wheat.pw.usda.gov/SNP/snpdb.html. In May 2006, the database contained 17174 primers (forward and reverse), 1102 wheat polymorphic loci, and 2224 polymorphic sequence tagged sites in diploid ancestors of polyploid wheat. Zhang et al. [140] also reported 246 gene loci with SNPs and/or small insertions/deletions from wheat homoeologous group 5. Another set of 101 SNPs (1SNP/212 bp) was discovered from genomic sequence analysis in 26-bread wheat lines and one synthetic line (http://urgi.versailles.inra.fr/GnpSNP/, [141]).

6. BAC/BIBAC Resources

BAC/BIBAC libraries have been produced in diploid, tetraploid, and hexaploid wheats (see Table 4). Chromosome-specific BAC libraries were also prepared in hexaploid wheat [142144]. These BAC resources proved useful for a variety of studies including map-based cloning (see Table 3), organization of wheat genome into gene-rich and gene-poor regions that are loaded with retroelements [8, 145147], and for physical mapping and sequencing of wheat genome (http://wheatdb.ucdavis.edu:8080/wheatdb/, [11]).

7. Gene Distribution in Wheat: Gene-Rich and Gene-Poor Regions

Genetic and physical maps of the wheat genome, discussed above, have been utilized for a study of gene distribution within the genome [58, 63, 148]. In order to identify and demarcate the gene-containing regions, 3025 loci including 252 phenotypically characterized genes and 17 quantitative trait loci (QTL) were physically mapped with the help of deletion stocks [149, 150]. It was shown that within the genome, genes are not distributed randomly and that there are gene-rich regions (GRRs) and gene-poor regions (GPRs), not only within the wheat genome, but perhaps in all eukaryotes (for reviews, see [4, 151]).

In wheat genome, 48 GRRs containing 94% of gene markers were identified with an average of 7 such GRRs (range 5–8) per homoeologous group. It was also shown that different wheat chromosomes differed for number and location of GRRs, with 21 GRRs on the short arms containing 35% of the wheat genes, and the remaining 27 GRRs on the long arms containing about 59% of the genes. The GRRs also vary in their size and in gene-density with a general trend of increased gene-density toward the distal parts of individual chromosome arms. This is evident from the fact that more than 80% of the total marker loci were mapped in the distal half of the chromosomes and 58% mapped in the distal 20%.

Among 48 GRRs, there were 18 GRRs (major GRRs), which contained nearly 60% of the wheat genes, covering only 11% of the genome, suggesting a very high density of genes in these GRRs, although the number and density of genes in these 18 GRRs was also variable [149, 150]. It has also been shown that the size of GRRs decreases and the number of GRRs increases, as the genome size increases from rice to wheat [4]. For instance, the average size of gene clusters in rice is 300 kb as compared to less than 50 kb in wheat and barley. However, no correlation was observed between the chromosome size and the proportion of genes or the size of the GRRs. For instance, group 3 has the longest chromosomes among the wheat homoeologous groups but contained only 13% of the genes compared to group 5 chromosomes that contained 20% of genes [150].

For the chromosomes of homoeologous group 1, the distribution of genes and recombination rates have been studied in a relatively greater detail. Each chromosome of this group (1A, 1B, 1D) has eight GRRs (ranging in size from 3 Mb to 35 Mb), occupying 119 Mb of the 800-Mb-long chromosome. Using this homoeologous group, it was confirmed that the GRRs differ in the number of genes and gene-density even within a chromosome or its arms. For instance, the “1S0.8 region” is the smallest of all GRRs, but has the highest gene-density, which is 12 times that in the “1L1.0 region.”

The distribution of GRRs has also been compared with the distribution of chromosome breaks involved in the generation of deletion stocks that are currently available and have been used for physical mapping of wheat genome. It was found that the breakpoints are nonrandom, and occur more frequently around the GRRs (one break every 7 Mb; [58, 67]); they seem to occur around GRRs twice as frequently as one would expect on random basis (one break every 16 Mb; [149]). Consequently, GRRs interspersed by <7-Mb-long GPRs will not be resolved and better resolution would be needed to partition the currently known GRRs into mini-GRRs and GPRs.

It has also been inferred that perhaps in eukaryotic genomes, the “gene-poor” regions preferentially enlarged during evolution, as is obvious in wheat, where large, essentially, “gene-empty” blocks of up to 192 Mb are common. Taking polyploidy into account, 30% gene-rich part of the genome is still 4 times larger than the entire rice genome [149]. Therefore, gene distribution within the currently defined GRRs of wheat would probably be similar to that in the rice genome, except that the gene-clusters would be smaller and the interspersing “gene-empty” regions would be larger, similar to barley as described above. It has also been shown that the “gene-empty” regions of the higher eukaryotic genomes are mainly comprised of retrotransposons and pseudogenes [152, 153]. The proportion of retrotransposons is significantly higher than pseudogenes, especially in the larger genomes, like those of maize and bread wheat.

8. Variable Recombination Rates

The recombination rate has also been recently shown to vary in different regions of the wheat genome. This was demonstrated through a comparison of consensus physical and genetic maps involving 428 common markers [149, 150]. Recombination in the distal regions was generally found to be much higher than that in the proximal half of individual chromosomes, and a strong suppression of recombination was observed in the centromeric regions. Recombination rate among GRRs present in the distal half of the chromosome was highly variable with higher recombination in some proximal GRRs than in the distal GRRs [149, 150]. The gene poor-regions accounted for only 5% of recombination.

It has also been reported that the distribution of recombination rates along individual chromosomes is uneven in all eukaryotes studied so far (for more references, see [154, 155]). Among cereals, the average frequency of recombination in rice (with the smallest genome) is translated into a genetic distance of about 0.003 cM per kb with a range of 0 to 0.06 cM per kb (http://rgp.dna.affrc.go.jp/Publicdata.html) and that of wheat (the largest genome) is 0.0003 cM per kb with a range from 0 to 0.007 cM per kb. Non-recombinogenic regions were observed in yeast as well as in rice, but the highest recombination rate for a region appears to be 35-fold lower in rice and 140-fold lower in bread wheat (relative to yeast). It may be due to differences in the resolution of recombination rates, which is 400 kb in rice (in wheat the resolution is much lower than in rice), whereas the resolution in recombination hotspots in yeast may be as high as only <1 kb in length. Due to averaging over larger regions, recombination in hotspots in rice and wheat may appear to be low relative to that in yeast [4, 150, 151, 156].

9. Flow Cytogenetics and Microdissection of Chromosomes in Wheat

Flow cytogenetics and microdissection facilitated physical dissection of the large wheat genome into smaller and defined segments for the purpose of gene discovery and genome sequencing. Flow karyotypes of wheat chromosomes were also prepared [157159]. DNA obtained from the flow-sorted chromosomes has been used for the construction of chromosome-specific large-insert DNA libraries, as has been done for chromosome 4A [157, 159]. Later, all individual 42 chromosome arms involving 21 wheat chromosomes were also sorted out using flow cytometry [160]. In another study, it was also possible to microdissect 5BL isochromosomes from meiotic cells and to use their DNA with degenerate oligonucleotide primer PCR (DOP-PCR) to amplify chromosome arm-specific DNA sequences. These amplified PCR sequences were then used as probes for exclusive painting of 5BL [161].

Flow sorting in wheat has also been used for efficient construction of bacterial artificial chromosome (BAC) libraries for individual chromosomes [143, 162]. The use of these chromosome- and chromosome arm-specific BAC libraries is expected to have major impact on wheat genomics research [1]. For instance, the availability of 3B-specific BAC library facilitated map-based cloning of agronomically important genes such as major QTL for Fusarium head blight resistance [98]. Flow cytometry can also be used to detect numerical and structural changes in chromosomes and for the detection of alien chromosomes or segments thereof (reviewed in detail by [163]). For instance, a 1BL.1RS translocation could be detected by a characteristic change in the flow karyotype [164]. In addition, DNA from flow-sorted chromosomes can be used for hybridization on DNA arrays and chips, with the aim of assigning DNA sequences to specific chromosome arms. This technique will be extensively used now with the availability of Affymetrix wheat GeneChip [165].

10. Wheat Gene Space Sequencing

International Triticeae Mapping Initiative (ITMI), at its meeting held at Winnipeg, Canada during June 1–4, 2003, took the first initiative toward whole genome sequencing (WGS) in wheat and decided to launch a project that was described as International Genome Research of Wheat (IGROW) by B. S. Gill. A workshop on wheat genome sequencing was later organized in Washington, DC during November 11–13, 2003, which was followed by another meeting of IGROW during the National Wheat Workers Workshop organized at Kansas, USA, during Feb 22–25, 2004 [166]. Consequently, IGROW developed into an International Wheat Genome Sequencing Consortium (IWGSC). Chinese Spring (common wheat) was selected for WGS, since it already had ample genetic and molecular resources [1].

Three phases were proposed for sequencing the wheat genome: pilot, assessment, and scale up. The first phase was recommended for 5 years and is mainly focused on the short-term goal of IWGSC, involving physical and genetic mapping along with sample sequencing of the wheat genome aimed at better understanding of the wheat genome structure. The assessment phase will involve determining which method(s) can be used in a cost-effective manner to generate the sequence of the wheat genome. After a full assessment, the scale-up phase will involve the deployment of optimal methods on the whole genome, obtaining the genome sequence and annotation, which is the long-term goal of IWGSC. With the availability of new sequencing technologies provided by 454/Roche and those provided by Illumina/Solexa and ABI SOLiD [167]; sequencing of gene space of the wheat genome, which was once thought to be almost impossible, should become possible within the foreseeable future.

First pilot project for sequencing of gene space of wheat genome, led by INRA in France, was initiated in 2004 using the largest wheat chromosome, 3B (1GB = 2x the rice genome) of hexaploid wheat as a model. As many as 68000 BAC clones from a 3B chromosome specific BAC library [143] were fingerprinted and assembled into contigs, which were then anchored to wheat bins, covering 80% of chromosome 3B. Currently, one or more of these contigs are being sequenced [11], which will demonstrate the feasibility of large-scale sequencing of complete gene space of wheat genome.

11. Functional Genomics

The determination of the functions of all the genes in a plant genome is the most challenging task in the postgenomic era of plant biology. However, several techniques or platforms, like serial analysis of gene expression (SAGE), massively parallel signature sequencing (MPSS), and micro- and macroarrays, are now available in several crops for the estimation of mRNA abundance for large number of genes simultaneously. The microarrays have also been successfully used in wheat for understanding alterations in the transcriptome of hexaploid wheat during grain development, germination and plant development under abiotic stresses [168, 169]. Recently, a comparison was made between Affymetrix GeneChip Wheat Genome Array (an in-house custom-spotted complementary DNA array) and quantitative reverse transcription-polymerase chain reaction (RT-PCR) for the study of gene expression in hexaploid wheat [170]. Also, functional genomics approach in combination with “expression genetics” or “genetical genomics” provides a set of candidate genes that can be used for understanding the biology of a trait and for the development of perfect or diagnostic marker(s) to be used in map-based cloning of genes and MAS [9]. A similar example was provided by Jordan et al. [9], when they identified regions of wheat genome controlling seed development by mapping 542 eQTLs, using a DH mapping popultion that was earlier used for mapping of SSRs and QTL analysis of agronomic and seed quality traits [171]. Expression analysis using mRNA from developing seeds from the same mapping population was also conducted using Affymetrix GeneChip Wheat Genome Array [172].

11.1. RNA Interference for Wheat Functional Genomics

RNA interference (RNAi), which was the subject of the 2006 Nobel Prize in Physiology or Medicine, is also being extensively utilized for improvement of crop plants [173]. This technique does not involve introduction of foreign genes and thus provides an alternative to the most controversial elements of genetic modification. Plans in Australia are underway, where the knowledge gained from RNAi approach will be used for developing similar wheats by conventional method of plant breeding, as suggested by CSIRO scientists for developing high-fibre wheat [174]. In bread wheat, in particular, the technology provides an additional advantage of silencing all genes of a multigene family including homoeoloci for individual genes, which are often simultaneously expressed, leading to a high degree of functional gene redundancy [175]. It has been shown that delivery of specific dsRNA into single epidermal cells in wheat transiently interfered with gene function [176, 177]. Yan et al. [90] and Loukoianov et al. [178] used RNAi for stable transformation and to demonstrate that RNAi-mediated reduction of VRN2 and VRN1 transcript levels, respectively, accelerated and delayed flowering initiation in winter wheat. Similarly, Regina et al. [179] used RNAi to generate high-amylose wheat. However, none of the above studies reported long-term phenotypic stability of RNAi-mediated gene silencing over several generations, neither did they report any molecular details on silencing of homoeologous genes. However, Travella et al. [180] showed RNAi results in stably inherited phenotypes suggesting that RNAi can be used as an efficient tool for functional genomic studies in polyploid wheat. They introduced dsRNA-expressing constructs containing fragments of genes encoding Phytoene Desaturase (PDS) or the signal transducer of ethylene, Ethylene Insensitive 2 (EIN2) and showed stably inherited phenotypes of transformed wheat plants that were similar to mutant phenotypes of the two genes in diploid model plants. Synthetic microRNA constructs can also be used as an alternative to large RNA fragments for gene silencing, as has been demonstrated for the first time in wheat by Yao et al. [181] by discovering and predicting targets for 58 miRNAs, belonging to 43 miRNA families (20 of these are conserved and 23 are novel to wheat); more importantly four of these miRNAs are monocot specific. This study will serve as a foundation for the future functional genomic studies. The subject of the use of RNAi for functional genomics in wheat has recently been reviewed [173].

11.2. TILLING in Wheat

Recently, Targeting Induced Local Lesions IN Genomes (TILLING) was developed as a reverse genetic approach to take advantage of DNA sequence information and to investigate functions of specific genes [182]. TILLING was initially developed for model plant Arabidopsis thaliana [183] having fully sequenced diploid genome and now has also been successfully used in complex allohexaploid genome of wheat, which was once considered most challenging candidate for reverse genetics [184].

To demonstrate the utility of TILLING for complex genome of bread wheat, Slade et al. [185] created TILLING library in both bread and durum wheat and targeted waxy locus, a well characterized gene in wheat encoding granule bound starch synthase I (GBSSI). Loss of all copies of this gene results in the production of waxy starch (lacking amylose). Production of waxy wheat by traditional breeding was difficult due to lack of genetic variation at one of the waxy loci. However, targeting waxy loci by TILLING [185], using locus specific PCR primers led to identification of 246 alleles (196 alleles in hexaploid and 50 alleles in tetraploid) using 1920 cultivars of wheat (1152 hexaploid and 768 tetraploid). This made available novel genetic diversity at waxy loci and provided a way for allele mining in important germplasm of wheat. The approach also allowed evaluation of a triple homozygous mutant line containing mutations in two waxy loci (in addition to a naturally occurring deletion of the third locus) and exhibiting a near waxy phenotype.

Another example of on-going research using TILLING in wheat is the development of EMS mutagenised populations of T. aestivum (cv. Cadenza, 4200 lines, cv. Paragon, 6000 lines), T. durum (cv. Cham1, 4,200 lines), and T. monococcum (Accession DV92, 3000 lines) under the Wheat Genetic Improvement Network (WGIN; funded by Defra and BBSRC in the UK and by the EU Optiwheat programme). The aim of this program is to search noval variant alleles for Rht-b1c,RAR-1, SGT-1, and NPR-1 genes (personal communication: [email protected] and [email protected]).

The above examples provide proof-of-concept for TILLING other genes, whose mutations may be desired in wheat or other crops. However, homoeolog-specific primers are required in order to identify new alleles via TILLING in wheat. In case of waxy, the sequences of the three homoeologous sequences were already known, which facilitated primer designing, but TILLING of other genes may require cloning and sequencing of these specific genes in order to develop homoeolog-specific target primers.

12. Comparative Genomics

In cereals, a consensus map of 12 grass genomes including wheat is now available, representing chromosome segments of each genome relative to those in rice on the basis of mapping of anchor DNA markers [186]. Some of the immediate applications of comparative genomics in wheat include a study of evolution [187] and isolation/characterization of genes using the model genome of rice. The genes, which have been examined using comperative genomics approach include the pairing gene, Ph1 [102, 188], gene(s) controlling preharvest sprouting (PHS; [189]), receptor-like kinase loci [190], gene for grain hardness [191], genes for glume coloration and pubescence (Bg, Rg; [192]), and the Pm3 gene, responsible for resistance against powdery mildew [187].

Conservation of Colinearity and Synteny
Among cereals, using molecular markers, colinearity was first reported among A, B, and D subgenomes of wheat [13, 193], and later in the high-gene density regions of wheat and barley. At the Lrk10 locus in wheat and its orthologous region in barley, a gene density of one gene per 4-5 kb was observed, which was similar to that found in A. thaliana [6]. Conservation of colinearity between homoeologous A genomes of diploid einkorn wheat and the hexaploid was also exploited for chromosome walking leading to cloning of candidate gene for the leaf rust resistance locus Lr10 in bread wheat [194]. Lr10 locus along with LMW/HMW loci of diploid wheat, when compared with their orthologs from tetraploid and hexaploid wheats, was found to be largely conserved except some changes that took place in intergenic regions [195197]. On the basis of divergence of intergenic DNA (mostly transposable elements), tetraploid and hexaploid wheats were shown to have diverged about 800000 years ago [197]. Similarly, the divergence of diploid from the tetraploid/hexaploid lineage was estimated to have occurred about 2.6–3 million years ago [195, 196].
Notwithstanding the above initial demonstration of colinearity using molecular markers, later studies based on genome sequences suggested disruption of microcolinearity in many regions thus complicating the use of rice as a model for cross-species transfer of information in these genomic regions. For instance, Guyot et al. [198] conducted an in silico study and reported a mosaic conservation of genes within a novel colinear region in wheat chromosome 1AS and rice chromosome 5S. Similarly, Sorrells et al. [199] while comparing 4485 physically mapped wheat ESTs to rice genome sequence data belonging to 2251 BAC/PAC clones, resolved numerous chromosomal rearrangements. The above findings also received support from sequence analysis of the long arm of rice chromosome 11 for rice-wheat synteny [200].

More recently, the grass genus Brachypodium is emerging as a better model system for wheat belonging to the genus Triticum, because of a more recent divergence of these two genera (35–40 million years) relative to wheat-rice divergence [201203]. Also, sequence of Brachypodium, which is likely to become available in the near future, may help further detailed analyses of colinearity and synteny among grass genomes. This has already been demonstrated through a comparison of 371 kb sequence of B. sylvaticum with orthologous regions from rice and wheat [204]. In this region, Brachypodium and wheat showed perfect macrocolinearity, but rice was shown to contain 220 kb inversion relative to Brachypodium sequence. Also, in Ph1 region, more orthologous genes were identified between the related species B. sylvaticum and wheat than between wheat and rice, thus once again demonstrating relative utility of Brachypodium genome as a better model than rice genome for wheat comparative genomics [102, 188].

13. Epigenetics in Wheat

Epigenetics refers to a heritable change that is not a result of a change in DNA sequence, but, instead, results due to a chemical modification of nucleotides in the DNA or its associated histone proteins in the chromatin. Several studies have recently been intiated to study the epigenetic modifications in the wheat genome. For instance, methylation-sensitive amplified polymorphism (MSAP) has been used to analyze the levels of DNA methylation at four different stages (2d, 4d, 8d, and 30d after pollination) of seed development in bread wheat [205]. It was found that 36–38% of CCGG sites were either fully methylated at the internal C’s and/or hemimethylated at the external C’s at the four corresponding stages. Similarly, Shitsukawa et al. [206] also studied genetic and epigenetic alterations among three homoeologs in the two class E-type wheat genes for flower development, namely, wheat SEPALLATA (WSEP) and wheat LEAFY HULL STERILE1 (WLHS1). Analyses of gene structure, expression patterns, and protein functions showed that no alterations were present in the WSEP homoeologs. By contrast, the three WLHS1 homoeologs showed genetic and epigenetic alterations. It was shown that WLHS1-B was predominantly silenced by cytosine methylation, suggesting that the expression of three homoeologous genes is differentially regulated by genetic or epigenetic mechanisms. Similar results were reported for several other genes like TaHd1 involved in photoperiodic flowering pathway, Ha for grain hardness, and TaBx for benzoxazinone biosynthesis [207209].

A prebreeding program in wheat (along with barley and canola) based on epigenetically modified genes has also been initiated in Australia at CSIRO, under the leadership of Dr. Liz Dennis and Dr. Jim Peacock, with the support from Dr. Ben Trevaskis (http://www.grdc.com.au/director/events/groundcover?item_id=A5B55D1DED8B9C20860C0CDE8C6EE077&article_id=A97C28B1F1614E34835D6BDB8CBDC75C). This pioneering work will involve vernalization, the mechanism that allows winter crops to avoid flowering until spring, when long days and mild conditions favor seed setting and grain filling. They plan to breed varieties with a wider range of heading dates and improved frost tolerance during flowering. In wheat (as also in other cereals), the epigenetic component is also built around VRN1 gene, which plays a role analogous to that of Flowering Locus C (FLC) in Arabidopsis and canola. VRN1 is one of the most important determinants of heading dates in winter cereals including wheat and also accounts for difference between winter and spring wheat varieties. It has been shown that during vegetative growth, VRN1 is repressed epigenetically; this repression is lifted in spring, allowing the protein encoded by VRN1 to activate other genes involved in reproduction. As many as 3000 wheat varieties are being looked at for variation in their VRN1 gene so as to breed better combinations of heading date and frost tolerance (http://www.grdc.com.au/director/events/groundcover?item_id=A5B55D1DED8B9C20860C0CDE8C6EE077&article_id=A97C28B1F1614E34835D6BDB8CBDC75C).

Wheat Allopolyploidy and Epigenetics
Polyploidization induces genetic and epigenetic modifications in the genomes of higher plants including wheat (reviewed in [210, 211]). Elimination of noncoding and low-copy DNA sequences has been reported in synthetic allopolyploids of Triticum and Aegilops species [212214]. In two other studies, patterns of cytosine methylation were also examined throughout the genome in two synthetic allotetraploids, using methylation-sensitive amplification polymorphism (MSAP; [215, 216]). This analysis indicated that the parental patterns of methylation were altered in the allotetraploid in 13% of the genomic DNA analyzed. Gene silencing and activation were also observed when 3072 transcribed loci were analyzed, using cDNA-AFLP [217, 218]. This study demonstrated new, nonadditive patterns of gene expression in allotetraploid, as indicated by the fact that 48 transcripts disappeared and 12 transcripts that were absent in the diploid parents, appeared in the allotetraploid. These results were found reproducible in two independent synthetic allotetraploids. The disappearance of transcripts could be related to gene silencing rather than gene loss and was partly associated with cytosine methylation. In another similar study involving artificially synthesized hexaploid wheats and their parents, down-regulation of some genes and activation of some other genes, selected in a nonrandom manner, was observed [219]. The genome-wide genetic and epigenetic alterations triggered by allopolyploidy thus suggested plasticity of wheat genome. The reproducibility of genetic and epigenetic events indicated a programmed rather than a chaotic response and suggests that allopolyploidy is sensed in a specific way that triggers specific response rather than a random mutator response [218].

14. Quantitative Trait Loci (QTL) and Protein Quantitative Loci (PQLs) in Wheat

A large number of QTL studies for various traits have been conducted in bread wheat, leading to mapping of QTL for these traits on different chromosomes. In most of these studies, either single marker regression approach or QTL interval mapping has been utilized. Although most of these studies involved mapping of QTL with main effects only, there are also reports of QTL, which have no main effects but have significant digenic epistatic interactions and QTL environment interactions [220222]. A detailed account of studies involving gene tagging and QTL analyses for various traits conducted in wheat is available elsewhere [14, 223]. More up-to-date accounts on QTL studies (summarized in Table 5) are also available for disease resistance [224], for resistance against abiotic stresses [225], grain size, and grain number [226], and for several other traits including yield and yield contributing characters, plant type, and flowering time [222, 227]. Advanced backcross QTL (AB-QTL) analysis, proposed by Tanksley and Nelson [228], has also been utilized in wheat to identify QTL for a number of traits including yield and yield components, plant height, and ear emergence [129, 229]. More recently AB-QTL analysis was practiced for the identification of QTL for baking quality traits in two B populations of winter wheat [230].

Quantitative variation in protein spots was also used for detection of protein quantitative loci (PQL) in wheat. For instance, in a study, 170-amphiphilic protein spots that were specific to either of the two parents of ITMIpop were used for genotyping 101 inbred lines; 72 out of these 170 proteins spots were assigned to 15 different chromosomes, with highest number of spots mapped to Group-1 chromosomes. QTL mapping approaches were also used to map PQL; 96 spots out of the 170 specific ones showed at least one PQL. These PQL were distributed throughout the genome. With the help of MALDI-TOF spectrometry and database search, functions were also assigned to 93 specific and 41 common protein spots. It was shown in the above study that majority of these proteins are associated with membranes and/or play a role in plant defense against external invasions [231].

15. Recent Insights into the Origin/Evolution of Wheat Genomes

In the genomics era, the subject of origin and evolution of bread wheat has also been revisited. This gave new insights into the identity of progenitors of the three subgenomes (A, B, D) of bread wheat, and into the genome alterations, which presumably accompanied the course of its evolution and domestication (see Figure 1). These aspects of evolution of bread wheat will be discussed briefly in this section.

15.1. Origin of A, B, and D Subgenomes

As mentioned earlier, bread wheat is a segmental allohexaploid having three closely related subgenomes A, B, and D. Initial analysis of the three subgenomes of bread wheat was mainly based on studies involving chromosome pairing in interspecific hybrids, and karyotype analysis in bread wheat as well as in the probable donors of the subgenomes (for reviews, see [232236]). However, more recently, molecular markers and DNA sequence data have been used for the analysis of these subgenomes (see [237239]). As a result, we have known with some degree of certainty that T. urartu (2 n = 14) is the donor of subgenome A and Ae. tauschii (synonyms, T. tauschii, Ae. squarrosa) is the donor of subgenome D; this has recently been confirmed through analysis of DNA sequences of two genes, namely, Acc-1 (plastid acetyl-CoA carboxylase) and Pgk-1 (plastid 3-phosphoglycerate kinase) [240]. In contrast to this, although Ae. speltoides was once considered as the probable donor of the B subgenome ([241], for a review, see [237]), studies carried out later showed that Ae. speltoides more closely resembles the subgenome G of T. timopheevii rather than to the subgenome B of bread wheat. DNA sequences of the above genes, Acc-1 and Pgk-1 also proved to be of no help in identification of the progenitor of the subgenome B. There is, thus still no unanimity on the progenitor of the subgenome B of bread wheat (for more details, see [242]), and there are speculations that the donor of the subgenome B might have lost its identity during evolution and may never be discovered.

DNA sequences of genes other than the above two genes have also been used for the study of origin and evolution of the component subgenomes of bread wheat. For instance, in one such study, sequences from 14 loci (2 sequences from each of the 7 chromosomes) belonging to the subgenome B of bread wheat, when compared with those from five diploid species (from section Sitopsis) closely related to the B subgenome of bread wheat, indicated that the B subgenome of bread wheat and the genomes of the above five diploid species diverged greatly after the origin of tetraploid wheat [243]. The above study also received support from the recent evidence of independent origins of wheat B and G subgenomes [244]. In this study, 70 AFLP loci were used to sample diversity among 480 wheat lines collected from their natural habitats, which encompassed the entire range of habitats for all S genome Aegilops species. Also, a comparison of 59 Aegilops representatives of S genome diversity with 2x, 4x chromosome number, and 11 nulli-tetrasomic wheat lines at 375 AFLP loci suggested that B genome chromosomes of 6x wheat were derived from chromosomes of Ae. speltoides, and no other species. Further, an analysis of the haplotypes at nuclear and chloroplast loci ACC1, G6PDH, GPT, PGK1, Q, VRN1, and ndhF for 70 Aegilops and Triticum lines (0.73 Mb sequenced) revealed that both B and G genomes of polyploid wheats are unique samples of A. speltoides haplotype diversity. However, it is likely that due to the outbreeding nature of A. speltoides, no modern A. speltoides lines have preserved the B donor genotype in its ancestral state. The above findings can be incorporated into a broader scheme of wheat genome evolution (see Figure 1) with resolved positions of the B genome relative to S progenitors and G sisters. Similar analysis of the D subgenome and its progenitor showed that the D subgenome had more than one allele for a single locus derived from a progenitor, suggesting that hexaploid wheat perhaps originated from tetraploid wheat more than once utilizing different sources of Ae. tauschii [245]. Also, it was realized that major part of the large genome (16000 Mb) of bread wheat is composed of transposable elements (TEs). Therefore, the role of TEs in the evolution of bread-wheat and allied genomes has also been examined [246, 247]. In these studies, some specific sequences from A and B genomes of diploid species were located, respectively, in B- and A-subgenomes of bread wheat, suggesting the role of TEs in transfer of sequences between A and B subgenomes. A bioinformatics approach was also used on a large genomic region (microgenomic approach) sequenced from T. monococcum (AA) and Ae. tauschii (DD). This approach allowed a comparison of variation within coding regions with that in the noncoding regions of the subgenomes.

15.2. Alterations that Accompanied Domestication

Domestication of most crop plants including wheat involved transition from short day, small-seeded plants with natural seed dispersal to photoperiod insensitive, large-seeded nonshattering plants. A study of genetic loci underlying domestication-related traits in T. dicoccoides was also conduced [430], where seven domestication syndrome factors (DSFs) were proposed, each affecting 5–11 traits. Following conclusions were made with respect to the domestication-related QTL. (i) Some of these QTL had strong effect and were clustered. (ii) Strong QTL were mainly associated with GRRs, where recombination rates are high. (iii) These QTL predominantly occurred in the A genome, suggesting that A genome has played a more important role than the B genome in evolution during domestication; this is understandable, because einkorn diploid wheat (T. monococcum) carrying the A genome was the first wheat to be domesticated, so that most of the domestication related traits in different wheats must have been selected within the A genome. Similar studies involving study of evolution during domestication were also conducted in hexaploid wheats for seed size, free threshing habit, rachis stiffness, photoperiod insensitivity, and so forth (for a review, see [431]). In wheat, a primary component of domestication syndrome was the loss of spike shattering, controlled by Br (brittle rachis) loci on chromosome 3A and 3B [414]. Other traits of wheat domestication syndrome shared by all domesticated wheats are the soft glumes, increased seed size, reduced number of tillers, more erect growth, and reduced dormancy [432]. A gene GPC-B1, which is an early regulator of senescence with pleiotropic effects on grain nutrient content, has also been found to affect seed size [96]. However, in some genotypes and environments, the accelerated grain maturity conferred by functional GPC-B1 allele has been found associated with smaller seeds [433], suggesting that indirect selection for large seeds may explain the fixation of the nonfunctional GPC-B1 allele in both durum and bread wheats [96]. Among many genes relevant to wheat domestication syndrome, only Q and GPC-B1 have been successfully isolated so far, suggesting a need for systematic effort to clone other genes, since it is possible that genetic variation at these loci might have played an important role in the success of wheat as a modern crop.

16. Application of Genomics to Molecular Breeding of Wheat

16.1. Association Mapping in Wheat

Association mapping is a high-resolution method for mapping QTL based on linkage disequilibrium (LD) and holds great promise for genetic dissection of complex traits. It offers several advantages, which have been widely discussed [434, 435]. In wheat, some parts of the genome relative to other parts are more amenable to LD/association mapping for QTL detection and fine mapping, since the level of LD is variable across the length of a chromosome. As we know, LD decay over longer distances will facilitate initial association of trait data with the haplotypes in a chromosome region and LD decay over short distances will facilitate fine mapping of QTL [12].

Several studies involving association mapping in wheat have been conducted in the recent past. For instance, association mapping has been conducted for kernel morphology and milling quality [436] and for the quantity of a high-molecular-weight glutenin [141, 437]. In another study, 242 diversity array technology (DArT) markers were utilized for association mapping of genes/QTL controlling resistance against stem rust (SR), leaf rust (LR), yellow rust (YR), powdery mildew (PM), and those controlling grain yield (GY). Phenotypic data from five historical CIMMYT elite spring wheat yield trials (ESWYT) conducted in a large number of international environments were utilized for this purpose and two linear mixed models were applied to assess marker-trait associations after a study of population structure and additive genetic covariance between relatives [438]. A total of 122, 213, 87, 63, and 61 DArT markers were found to be significantly associated with YR, GY, LR, SR, and PM, respectively. Association analysis was also conducted between markers in the region of a major QTL responsible for resistance to Stagonospora nodorum (causing glume blotch); it was concluded that association mapping had a marker resolution, which was 390-fold more powerful than QTL analysis conducted using an RIL mapping population [439]. Such high-resolution mapping of traits and/or QTL to the level of individual genes, using improved statistical methods, will provide new possibilities for studying molecular and biochemical basis of quantitative trait variation and will help to identify specific targets for crop improvement.

16.2. Marker-Assisted Selection in Wheat

A large number of marker-trait associations determined during the last decades facilitated the use of molecular markers for marker-assisted selection (MAS) in bread wheat, which is gaining momentum in several countries. In particular, major programs involving MAS in wheat are currently underway in USA, Australia, and at CIMMYT in Mexico. In USA, a wheat MAS consortium comprosing more than 20 wheat-breeding programs was constituted at the end of 2001. The objective of this consortium was to apply and to integrate MAS in public wheat breeding programs [440]. Under these programs, MAS has been utilized for transfer of as many as 27 different insect and pest resistance genes and 20 alleles with beneficial effects on bread making and pasta quality into 180 lines adapted to the primary US production regions. These programs led to release of germplasm consisting of 45 MAS-derived lines [441]. Similarly, the program in Australia involved improvement of 20 different traits (including resistance to some abiotic stresses) and has already led to release of some improved cultivars ([442], Peter Langridge personal communication). Among these traits, MAS has become a method of choice for those agronomically important traits, where conventional bioassays were expensive and unconvincing, as was the case in selection for cereal cyst nematodes resistance carried out by Agriculture Victoria [443]. In addition to this, MAS has been incorporated in backcross breeding in order to introgress QTL for improvement of transpiration efficiency and for negative selection for undesirable traits such as yellow flour color [444]. Australian scientists also conducted a computer simulation in order to design a genetically effective and economically efficient marker-assisted wheat-breeding strategy for a specific outcome. This investigation involved an integration of both restricted backcrossing and doubled haploid (DH) technology. Use of MAS at the B followed by MAS in haploids derived from pollen of B (prior to chromosome doubling) led to reduction of cost of marker-assisted breeding up to 40% [445]. Later, this MAS strategy was validated practically in a marker-assisted wheat-breeding program in order to improve quality and resistance against rust disease (for review, see [446]). At CIMMYT, markers associated with 25 different genes governing insect pest resistance, protein quality, homoeologous pairing, and other agronomic characters are currently being utilized in wheat breeding programs in order to develop improved wheat cultivars [447]. Some of the markers used in these programs are perfect markers that have been developed from available nucleotide sequences of these genes. In future, large-scale sequencing of GRRs (gene-rich regions), to be undertaken by IWGSC, will also facilitate isolation of important genes for production of improved transgenic crops, and for development of “perfect markers” for agronomically important traits to be used in MAS [448, 449].

17. Organellar Genomes and their Organization

The genomes of wheat chloroplast and mitochondrion have also been subjected to a detailed study during the last decade. The results of these studies will be briefly discussed in this section.

17.1. Chloroplast Genome

In bread wheat, 130–155 chloroplasts, each containing 125–170 circular DNA molecules (135 kb), are present in each mesophyll cell, thus making 16000–26000 copies of cpDNA within a cell. This makes 5–7% of the cellular DNA in the leaf and 10–14% of the DNA in a mesophyll cell. In the related diploid species, there are 4900–6600 copies and in tetraploid species, there are 9600–12400 copies of cpDNA per mesophyll cell.

The wheat chloroplast genome, like all other plant chloroplast genomes, has two inverted repeat regions, each copy (21-kb-long) separated from the other by two single copy regions (12.8 kb, 80.2 kb). The gene content of wheat chloroplast is the same as those of rice and maize plastomes, however some structural divergence was reported in the gene coding regions, due to illegitimate recombination between two short direct repeats and/or replication slippage; this included the presence of some hotspot regions for length mutations. The study of deletion patterns of open reading frames (ORFs) in the inverted-repeat regions and in the borders between the inverted repeats and the small single-copy regions supports the view that wheat and rice are related more closely to each other than to maize (see [450, 451]). Deletions, insertions, and inversions have also been detected during RFLP analysis of cpDNA, which gave eleven different cpDNA types, in the genus Triticum, the bread wheat sharing entirely the cpDNA type with durum wheats, but not with that of any of the diploid species. The cpDNA of Ae. speltoides showed maximum similarity to those of T. aestivum, T. timopheevii, and T. zhukovskyi, suggesting that Ae. speltoides should be the donor of the B subgenome of common wheat [452].

17.2. Mitochondrial Genome

Wheat mtDNA is larger (430 kb) than cpDNA (135 kb) with a minimum of 10 repeats but encodes only 30–50% polypeptides relative to cpDNA. Thus, large amount of mtDNA is noncoding, there being about 50 genes involved in RNA synthesis [453]. Mitochondrial genome of Chinese Spring has been sequenced using 25 cosmid clones of mitochondrial DNA, selected on the basis of their gene content. This led to the identification of 55 (71) genes including the following: 18 genes (20) for electron transport system, 4 genes for mitochondrial biogenesis, 11 genes for ribosomal proteins, 2 genes for splicing and other function, 3 genes (10) for rRNAs, and 17 genes (24) for tRNAs (the numerals in parentheses represent number of genes, taking multiple copies of a gene as separate genes). When mitochondrial gene maps were compared among wheat, rice, and maize, no major synteny was found between them other than a block of two to five genes. Therefore, mitochondrial genes seem to have thoroughly reshuffled during speciation of cereals. In contrast, chloroplast genes show perfect synteny among wheat, rice, and maize [451].

18. Conclusions

Significant progress during the last two decades has been made in different areas of wheat genomics research. These include development of thousands of molecular markers (including RFLPs, SSRs, AFLPs, SNPs, and DArT markers), construction of molecular genetic and physical maps (including radiation hybrid maps for some chromosomes) with reasonably high density of markers, development of more than 1 million ESTs and their use for developing functional markers, and the development of BAC/BIBAC resources for individual chromosomes and entire subgenomes to facilitate genome sequencing. Functional genomics approaches like TILLING, RNAi, and epigenetics have also been utilized successfully, and a number of genes/QTL have been cloned to be used in future wheat improvement programs. Organellar genomes including chloroplast and mitochondrial genomes have been fully sequenced, and we are at the threshold of initiating a major program of sequencing the gene space of the whole nuclear genome in this major cereal. The available molecular tools also facilitated a revisit of the wheat community to the problem of origin and evolution of the wheat genome and helped QTL analysis (including studies involving LD and association mapping) for identification of markers associated with all major economic traits leading to the development of major marker-aided selection (MAS) programs for wheat improvement in several countries.

Acknowledgments

The authors would like to thank the Indian National Science Academy (INSA) for award of an INSA Honorary Scientist Position to PKG, Department of Biotechnology (DBT) for providing financial support to R. R. Mir, A. Mohan, J. Kumar, and also to the Head of Department of Genetics and Plant Breeding, Chaudhary Charan Singh University, Meerut, for providing the facilities.