Abstract
Wheat is a typical allopolyploid with three homoeologous subgenomes (A, B, and D). The ancestors of the subgenomes A and D had been identified, but not for the subgenome B. The goatgrass Aegilops speltoides (genome SS) has been controversially considered a candidate for the ancestor of the wheat B genome. However, the relationship of the Ae. speltoides S genome with the wheat B genome remains largely obscure, which has puzzled the wheat research community for nearly a century. In the present study, the genome-wide homology analysis identified perceptible homology between wheat chromosome 1B and Ae. speltoides chromosome 1S, but not between other chromosomes in the B and S genomes. An Ae. speltoides-originated segment spanning a genomic region of approximately 10.46 Mb was identified on the long arm of wheat chromosome 1B (1BL). The Ae. speltoides-originated segment on 1BL was found to co-evolve with the rest of the B genome in wheat species. Thereby, we conclude that Ae. speltoides had been involved in the origin of the wheat B genome, but should not be considered an exclusive ancestor of this genome. The wheat B genome might have a polyphyletic origin with multiple ancestors involved, including Ae. speltoides. These novel findings provide significant insights into the origin and evolution of the wheat B genome, and will facilitate polyploid genome studies in wheat and other plants as well.
Introduction
Wheat (Triticum aestivum, 2n=6x=42, genome AABBDD), a major food grain source for humans, has been considered a typical allohexaploid originated from the interspecific hybridization involving three diploid ancestors (Sakamura 1918; Kihara 1919, 1954; Kihara et al. 1959; Sax 1922). This evolutionary theory of allopolyploid has led to successful identification of the ancestors for the wheat subgenomes A and D, but not yet for the subgenome B. The wild grasses T. urartu (2n=2x=14, genome AA) and Aegilops tauschii (2n=2x=14, genome DD) contributed A and D subgenomes to wheat, respectively (Kihara 1944; McFadden and Sears 1946; Dvorak et al. 1993). The ancestor of the subgenome B, however, remains controversial even though tremendous research efforts have been made to tackle this evolutionary puzzle of wheat for nearly a century.
Early studies on meiotic pairing, karyotyping, plant morphology, and geographic distribution of wheat-related wild species and their hybrids with wheat species (Triticum L.) identified the goatgrass Ae. speltoides (2n=2x=14, genome SS) as the closest ancestor of the wheat B genome (Jenkins 1929; Pathak 1940; Sarkar and Stebbins 1956; Riley et al. 1958). Meanwhile, questions had been raised about the meiotic pairing-based assessment of genome homology in these early studies because Ae. speltoides was suspected to contain genetic factors with epistatic effect on the wheat diploidization system, now designated Ph (pairing homoeologous) gene (Jenkins 1929; Sarkar and Stebbins 1956; Riley et al. 1958). The Ph gene limits meiotic pairing to homologous chromosomes in wheat and wheat hybrids with its relatives. Also, it had been assumed that the ancestral form of the B genome might have undergone a series of changes since its incorporation into wheat (Jenkins 1929; Sarkar and Stebbins 1956).
In a later study, Riley et al. (1961) confirmed the presence of the gene(s) in Ae. speltoides that suppresses the effect of the Ph1 gene located on wheat chromosome 5B. Three genotypes with high, intermediate, and low ability to suppress Ph1 activity were identified in Ae. speltoides (Dvorak 1972). According to these findings, Kimber and Athwal (1972) reassessed meiotic pairing in the hybrids and amphiploids involving wheat and Ae. speltoides accessions with different levels of suppression for the Ph1 activity. They determined that the variation of meiotic pairing in the hybrids resulted from the presence of different Ph1 suppressors in the Ae. speltoides accessions. Also, they found that chromosomes predominantly paired as bivalents in the amphiploid involving polyploid wheat and a low pairing Ae. speltoides accession, which was very similar to a normal diploidized allopolyploid. As a result, they concluded that Ae. speltoides could not be considered as the ancestor of the wheat B genome. This was also supported by the evidence of chromosome banding patterns (Gill and Kimber 1974) and protein electrophoretic profiles (Johnson 1972). More recently, three Ph1 suppressor gene loci were identified and mapped to chromosome 3S, 5S, and 7S of Ae. speltoides, respectively (Dvorak et al. 2006).
In contrast, molecular analyses of both nuclear and extranuclear genomic DNAs suggested that Ae. speltoides or a species in the evolutionary lineage of Ae. speltoides could be the most likely ancestor of the wheat B genome as well as wheat plasmon (Ogihar and Tsunewaki 1988; Dvorak and Zhang 1990; Sasanuma et al. 1996; Wang et al. 1997; Kilian et al. 2007). However, comparative analysis of several gene loci and nearby genomic regions across the Triticum and Aegilops species did not reveal clear evidence supporting that conclusion (Hang et al. 2002; Salse et al. 2008). The Ae. speltoides S genome was considered to be evolutionarily closer to the wheat B genome than to the A and D genomes, but its candidacy as the ancestor of B genome remained undetermined in both studies.
The wheat B genome has significantly higher genetic variability than A and D genomes (Chao et al. 1989; Felsenburg et al. 1991; Siedler et al. 1994; Petersen et al. 2006). These findings support the hypothesis that the wheat B genome had diverged from its ancestor through various genomic modifications (Jenkins 1929; Sarkar and Stebbins 1956; Blake et al. 1999). In addition, Ae. speltoides has significantly higher intraspecific genetic variability than any of the other four Aegilops species in the Sitopsis section, which is even comparable to the interspecific variability among the other four Aegilops species in the section (Sasanuma et al. 1996). This appears to support the hypothesis that Ae. speltoides might contribute to the origin of the wheat B genome, but the current version of Ae. speltoides had diverged from the original ancestor of the B genome (Salse 2008). Another hypothesis, proposed by Zohary and Feldman (Zohary and Feldman 1962), states that the wheat B genome is a reconstructed genome resulted from meiotic homoeologous recombination between multiple ancestral genomes of the Aegilops species. This evolutionary recombination process was assumed to occur in the hybrids of the tetraploid amphiploids that combined the different ancestral Aegilops genomes and a common ancestral A genome of T. urartu. In other words, the wheat B genome might have a polyphyletic origin.
A species with a genome more closely related to the wheat B genome than the S genome of Ae. speltoides has not been discovered even though intensive search for the ancestor of B genome has been performed over nearly a century. It seems inevitable to reason that Ae. speltoides might contribute to the origin and evolution of the wheat B genome to some extent according to the previous studies. The present study aimed to assess the homology of individual wheat B-genome chromosomes with their homoeologous counterparts in the S genome of Ae. speltoides and to detect the Ae. speltoides genomic components in the wheat B genome if there are any. A novel integrative cytogenetic and genomic approach was taken to accomplish this research, which could not be done in the previous studies due to the lack of the genomics/cytogenetics tools and resources. This work shed new light on the origin and evolution of the wheat B genome, and will facilitate further studies of the complex polyploid genomes in wheat and other plants.
Materials and Methods
Plant Materials
Six “Chinese Spring” (CS) wheat B genome-Ae. speltoides disomic substitution lines [DS 1S(1B), DS 2S(2B), DS 4S(4B), DS 5S(5B), DS 6S(6B), and DS 7S(7B)] (Friebe et al. 2011), one substitution line involving chromosome 3S and 3A [DS 3S(3A)], and the CS ph1b mutant were the initial genetic stocks used in this research. They were provided by the Wheat Genetics Resource Center at Kansas State University, USA. DS 3S(3A) was included in this study because DS 3S(3B) is not available (please see results in this study). In addition, six CS wheat B genome-Thinopyrum elongatum (2n=2x=14, genome EE) disomic substitution lines [DS 1E(1B), DS 2E(2B), DS 3E(3B), DS 5E(5B), DS 6E(6B), and DS 7E(7B)] and one substitution line involving chromosome 4E and 4D [DS 4E(4D)] were used as controls to assess B-S genome homology in this study. DS 4E(4B) was not available. They were kindly supplied by J. Dvorak at UC Davis. Each of these disomic substitution lines has a pair of wheat chromosomes replaced by their homoeologous counterparts of Ae. speltoides or Th. elongatum. The common and durum wheat (T. turgidum ssp. durum, 2n=4x=28, genome AABB) accessions used in this study were selected from the worldwide diversity panel of the Triticeae Coordinated Agricultural Project (T-CAP). The other wheat species/accessions that contain the B genome were obtained from the U.S. National Plant Germplasm System. A total of 179 accessions under 13 wheat species (Triticum L.) were chosen based on their geographic origin and distribution, representing a diverse worldwide collection of the tetraploid and hexaploid wheat species (File S1). A subset of representative wheat species/accessions (n=88) were selected for single nucleotide polymorphism (SNP) genotyping from the 179 accessions (File S2).
Construction of the special genotypes for meiotic pairing analysis
The CS-Ae. speltoides and CS-Th. elongatum disomic substitution lines were crossed and backcrossed with the CS ph1b mutant to construct the special genotypes monosomic for the individual B/A-S or B/D-E homoeologous pairs in the presence and absence of Ph1, respectively (Fig. S1). The chromosome-specific DNA markers (File S3) were employed to assist selection of the double monosomics for each of the homoeologous pairs. The selected individuals were verified for the monosomic condition by genomic in situ hybridization (GISH). The Ph1-specific DNA markers (Roberts et al., 1999) were used to select the double monosomics with Ph1 as well as those without Ph1 (i.e. homozygous for ph1b deletion mutant) (Fig. S1).
Anthers with meiocytes [pollen mother cells (PMCs)] at metaphase I (MI) were collected for meiotic pairing analysis from the heterozygotes with and without Ph1 following the procedure of Cai and Jones (1997). A total of over 100 meiocytes at MI from 1-6 plants were observed and analyzed for each of the special genotypes. GISH was used to differentially paint chromosomes of Ae. speltoides, Th. elongatum, and wheat for meiotic pairing analysis.
DNA Marker Analysis
DNA samples were prepared as described by Niu et al. (2011). Chromosome-specific DNA markers, including SSRs (simple sequence repeats) and STSs (sequence-tagged sites), were developed and used for the identification of individual B-, S-, and E-genome chromosomes as described by Chen et al. (2007). Two STS markers (PSR128 and PSR574) that tag the Ph1 allele were used to identify individuals homozygous for the ph1b deletion mutant (Roberts et al., 1999). The wheat 90K iSelect SNP arrays were used to perform SNP genotyping assay for CS wheat, the disomic substitution lines, and the 88 representative wheat species/accessions (45 hexaploids and 43 tetraploids) (File S2) using the Illumina iScan instrument. SNP allele clustering and genotype calling were conducted using the GenomeStudio v2011.1 software (Illumina, Inc.) (Wang et al. 2014). The polymorphisms for each of the homoeologous pairs at the SNP loci were calculated as the percentages of the polymorphic loci out of the total loci genotyped. The graphical view of the genotypes for the 88 wheat species/accessions at the 68 SNP loci within the distal end of 1BL/1SL was constructed using the Flapjack software (Milne et al. 2010). Genetic diversity was calculated based on the SNP genotyping data of the 88 representative wheat species/accessions (Nei 1973) and plotted against the SNP consensus linkage map of wheat chromosome 1B (Wang et al., 2014). The SNP genotype-based cluster dendrogram was developed using R package “ape” (https://cran.r-project.org/).
Fluorescent in situ hybridization (FISH)
Fluorescent genomic in situ hybridization (FGISH) was performed to differentiate wheat B-genome and S/E-genome chromatin from each other as described by Cai et al. (1998). Total genomic DNAs of Ae. speltoides and Th. elongatum were labeled with biotin-16-dUTP by nick translation as probe DNA for detecting Ae. speltoides and Th. elongatum chromatin, respectively. Total genomic DNA of CS wheat was used as blocking DNA. Ae. speltoides/Th. elongatum chromatin was painted with fluorescein isothiocyanate-conjugated avidin (FITC-avidin) as yellow-green and wheat chromatin was counter-stained with propidium iodide (PI) as red. Multicolor FISH was conducted following the procedure of Liu et al. (2006). The clone pTa71, a wheat 9 kb rDNA repeating unit that contains the 18S, 5.8S, and 26S rRNA genes and intergenic spacer (Gerlach and Bedbrook, 1979), was supplied by Peng Zhang at The University of Sydney, Australia. It was labeled with dig-11-dUTP and detected by anti-dig-rhodamine as red. This rDNA probe was used to tag the nucleolar organizer region on wheat chromosomes 1B and 6B. Total genomic DNA of Ae. speltoides was labeled with biotin-16-dUTP and detected by FITC-avidin as yellow-green. This genomic probe was used to identify Ae. speltoides chromatin in the wheat genome. Wheat chromatin was counter-stained with 4’,6-diamidino-2-phenylindole (DAPI) as blue. The fluorescence microscopy system BX51 (Olympus, Japan) was used to visualize GISH/FISH-painted chromosomes.
DNA Sequence Analysis
The DNA sequences of wheat chromosome 1B was extracted from the IWGSC RefSeq v1.0 (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies). The contextual sequences of the SNP loci at the distal ends of both CS wheat 1BL and Ae. speltoides 1SL were aligned to the DNA sequences of chromosome 1B using the Splign software (Kapustin et al. 2008). The physical order of the SNP loci was determined based on the DNA sequence alignment.
Results
Homology analysis of the individual B/A-S homoeologous chromosome pairs
Meiotic pairing has been considered direct cytological evidence for genome homology. It can, however, be influenced by the genetic factors in addition to homology, such as Ph1 gene in wheat and Ph1 suppressors in Ae. speltoides. To take account of the non-homology factors in the B-S genome homology analysis, we investigated meiotic pairing of individual B-S homoeologous pairs under the same genetic background of CS wheat in the presence and absence of Ph1. The CS wheat B genome-Ae. speltoides S genome disomic substitution lines dissect the S genome of Ae. speltoides into individual chromosomes in the CS wheat background. They were used to construct double monosomics for the individual B- and S- genome chromosomes. Meanwhile, the ph1b deletion mutant of Ph1 was introduced into the double monosoimcs with assistance of the Ph1-specific DNA markers. Thus, we were able to investigate meiotic pairing of individual B-S homoeologous chromosome pairs in the presence as well as absence of Ph1. In addition, we investigated meiotic pairing of the individual B-E homoeologous chromosome pairs as controls for B-S genome homology analysis.
Chinese Spring wheat chromosome 1B was found to pair with Ae. speltoides chromosome 1S in 67 of the 134 PMCs analyzed (50.00%), while other B/A-S homoeologous pairs had a relatively low meiotic pairing frequency ranging from 0.00 to 8.63% in the presence of Ph1 (Fig. 1). In addition, we noticed that 1B-1S meiotic pairing predominantly involved the long arms of chromosome 1B (1BL) and 1S (1SL). Surprisingly, wheat 1BL was found to contain a small Ae. speltoides S genome-derived chromosomal segment at its distal end, where meiotic pairing initiated (Fig. 2a). Also, the same Ae. speltoides-derived segment was observed on the unpaired 1BL (univalent) (Fig. 2b). The Ph1 suppressor genes mapped to Ae. speltoides chromosomes 3S, 5S, and 7S (Dvorak et al. 2006). We found that meiotic pairing involving chromosome 5S was noticeably higher than that involving 2S, 3S, 4S, 6S, and 7S in the presence of Ph1 (Fig. 1). Thus, there might be a Ph1 suppressor on this particular Ae. speltoides chromosome 5S, but not on chromosomes 3S and 7S involved in this study. In the absence of Ph1 (i.e. ph1bph1b), chromosomes 1B and 1S paired at a frequency of 60%, which was higher than the 1B-1S pairing frequency (50%) in the presence of Ph1. Meiotic pairing of other homoeologous pairs (2B-2S, 3A-3S, 4B-4S, 5B-5S, 6B-6S, and 7B-7S) was dramatically enhanced by ph1b mutant (Fig. 1). 5Bph1b-5S also exhibited a high pairing frequency (42.16%), suggesting absence of Ph1 on Ae. speltoides chromosome 5S (Griffiths et al. 2006).
Meiotic pairing was not observed between CS wheat chromosome 1B and Th. elongatum chromosome 1E in the 105 PMCs analyzed under the presence of Ph1. The other B/D-E homoeologous chromosome pairs also showed a low meiotic pairing frequency when Ph1 was present (Fig. 1). Meiotic pairing of all B/D-E homoeologous chromosome pairs was enhanced by the ph1b mutant, but not as extensively as that with the B/A-S homoeologous chromosome pairs except 4D-4E and 5B-5E. Apparently, CS wheat chromosome 5Bph1b had a high meiotic pairing affinity with Ae. speltoides chromosome 5S as well as Th. elongatum chromosome 5E (Fig. 1).
GISH/FISH analysis of wheat B-genome chromosomes
Meiotic pairing analysis demonstrated a notable homology between CS wheat chromosome 1B and Ae. speltoides chromosome 1S (Fig. 1). Apparently, the Ae. speltoides-derived chromosomal segment at the distal end of 1BL contributed to the high 1B-1S pairing. To confirm the Ae. speltoides segment on 1BL and determine whether any additional Ae. speltoides segments are present on the CS B-genome chromosomes, we performed multicolor FISH/GISH to the mitotic chromosomes of CS wheat. The CS wheat chromosomes 1B and 6B were tagged by FISH using the rDNA probe pTa71; and Ae. speltoides chromatin was simultaneously painted by Ae. speltoides genomic DNA-probed GISH. An Ae. speltoides chromosomal segment was clearly detected at the distal end of CS 1BL, but not in other regions of chromosome 1B and other B- genome chromosomes (Fig. 2c). To further verify the origin of the distal segment on 1BL, we performed Th. elongatum genomic DNA-probed GISH to all three chromosome sets of the CS wheat genome. No Th. elongatum-derived GISH signals were observed on any of the CS wheat chromosomes, including 1BL (Fig. 3A). Thus, the distal segment on CS chromosomal arm 1BL is Ae. speltoides S genome-specific, not a common chromosomal region shared by CS wheat and its relatives. It was derived from the S genome of Ae. speltoides.
We surveyed the B genome of 179 representative accessions under 13 wheat species (Triticum L.) for the presence of Ae. speltoides chromatin by GISH. They were collected from the different geographic regions around the world and represented a diverse collection of the hexaploid and tetraploid wheat species/accessions that contain the B genome. All of these wheat species/accessions were found to contain an Ae. speltoides chromosomal segment at the distal end of 1BL as what we observed on 1BL of CS wheat, but not in the other regions of chromosome 1B and other B-genome chromosomes (Table 1; Fig. 2c). Therefore, the Ae. speltoides-derived chromosomal segment is universally present at the distal end of 1BL in both tetraploid and hexaploid wheat. It has been part of chromosome 1B probably since the incorporation of the B genome into tetraploid and hexaploid wheat.
Comparative analysis of the individual B-S homoeologous pairs
Both CS wheat and CS-Ae. speltoides disomic substitution lines were genotyped using wheat 90K iSelect SNP arrays. The SNP genotyping results indicated that the substitution line originally designated DS 3S(3B) (Friebe et al. 2011) should be DS 3S(3A), which was further confirmed by SSR markers and chromosome C-banding. Thus, DS 3S(3A), instead of DS 3S(3B), was included in the SNP assay in addition to the substitution lines involving other six B- genome chromosomes (1B, 2B, 4B, 5B, 6B, and 7B).
High-throughput genotyping of CS wheat and the CS B genome-Ae. speltoides disomic substitution lines at 17,379 SNP loci identified a total of 6,722 SNPs polymorphic in the seven B/A-S homoeologous pairs. The homoeologous pair 2B-2S showed the lowest polymorphism (33.34%) and 7B-7S the highest (43.37%) at the SNP loci surveyed. The polymorphisms of other five B/A-S homoeologous pairs ranged from 33.89% (3A-3S) to 41.86% (4B-4S) (Fig. 4). Plotting of the SNP polymorphisms between CS wheat chromosome 1B and Ae. speltoides chromosome 1S against the SNP consensus linkage map of chromosome 1B (Wang et al. 2014) identified a genomic region that shared the same alleles at 65 of the 68 SNP loci within the distal ends of CS wheat 1BL and Ae. speltoides 1SL (Fig. 3C). Such a monomorphic linkage block was not detected in other chromosomal regions of the 1B-1S homoeologous pair and on other B/A-S homoeologous pairs (Fig. S2).
Genetic and physical characterization of the 1BL distal end
The chromosomal regions at the distal ends of CS wheat 1BL and Ae. speltoides 1SL were found to be highly monomorphic at the 68 SNP loci (65 monomorphic/68 SNPs). The three polymorphic SNP loci within the region on 1BL and 1SL were positioned toward the proximal end of the region on the consensus linkage map (Wang et al. 2014). The entire region defined by the 68 SNPs spans a genetic distance of 12.7 cM (Fig. 3B and File S4). The contextual sequences of these 68 SNPs were aligned to the distal region of 1BL according to the IWGSC RefSeq v1.0 (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies). Two of the three SNP loci polymorphic between 1BL and 1SL, Tdurum_contig41999_2908 and Ex_c1058_1537, were physically assigned to the proximal end of the region. The other polymorphic SNP within the region, RFL_Contig785_1156, was distal to the first two polymorphic SNPs. One monomorphic SNP (RFL_Contig785_1700) was physically positioned within the interval between Tdurum_contig41999_2908 and RFL_Contig785_1156 according to the sequence alignment (Fig. 3B and File S4).
The genomic region that spans the 65 SNP loci monomorphic between CS 1BL and Ae. speltoides 1SL and one polymorphic SNP (RFL_Contig785_1156) was estimated to be 9.61 Mb in length according to the DNA sequence assemblies of 1BL (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies). The two polymorphic SNP loci (Tdurum_contig41999_2908 and Ex_c1058_1537) at the proximal end of that region were not included in the estimate (Fig. 3D). In addition, we identified an extended terminal segment of 0.85 Mb distal to the 68 SNP-defined region on 1BL according to the DNA sequence alignment. As a result, the total physical length of the SNP-defined and extended distal genomic region on 1BL was estimated to be 10.46 Mb (Fig. 3D). The actual physical size of this distal segment on 1BL might be greater than this estimate (10.46 Mb) because the current DNA sequence assemblies (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies) we used in this study cover 689.9 Mb out of the total length 849 Mb of chromosome 1B (Šafář et al. 2010).
Four hexaploid wheat accessions (Cltr8347, PI429624, PI481728, and CI13113), similar to CS wheat, were found to be highly monomorphic with CS DS1S(1B) at the 68 SNP loci within the distal regions of 1BL and 1SL (Fig. S3). They were clustered together with CS wheat and CS DS1S(1B) in the dendrogram (Fig. 5). Cltr8347, PI429624, and PI481728 are the landraces from China, Nepal, and Bhutan, respectively. Both CS and Cltr8347 are the landraces collected probably in southwest China (P.D. Chen, personal communication), which is geographically close to the Himalayan region where Nepal and Bhutan are located. CI13113 is a winter wheat germplasm line with CS involved in the pedigree (http://www.arsgrin.gov/npgs/). Thus, these five hexaploid wheat accessions seem to share a similar origin of this particular Ae. speltoides-derived genomic region on 1BL. In addition, we found that 11 tetraploid wheat accessions share the same genotypes at the 68 SNP loci on 1BL (Fig. S3), making them clustered together in the dendrogram (Fig. 5). These tetraploid wheat accessions originated from two primary geographical regions, the Mediterranean Basin and South America (File S2). Overall, the tetraploid wheat accessions showed higher genetic variability than hexaploids in the 68 SNP-defined genomic region on 1BL according to the cluster dendrogram and genetic diversity analysis (Figs. 5 and 6).
Four of the 68 SNP loci (46805, 31066, 50867, and 76928) showed no allelic variation at all in the 88 wheat accessions. The other four SNPs (71971, 65270, 71898, and 78965) had very minimal variation in the wheat accessions (Fig. S3; File S4). Apparently, these SNP loci have been very conservative over the evolutionary process of this chromosomal region.
Discussion
Ae. speltoides has been considered the diploid species with a genome most closely related to the wheat B genome according to the previous studies (Jenkins 1929; Pathak 1940; Sarkar and Stebbins 1956; Riley et al. 1958; Dvorak and Zhang 1990; Sasanuma et al. 1996; Wang et al. 1997; Kilian et al. 2007). However, the relationship of the Ae. speltoides S genome with the wheat B genome remains largely obscure. Previous meiotic pairing-based homology analyses had led to inconsistent conclusions about the evolutionary relationship of the B and S genomes due primarily to the influence of the Ae. speltoides-derived Ph1 suppressors on meiotic pairing (Jenkins 1929; Riley et al. 1958; Kimber and Athwal 1972). In the present study, we partitioned the B and S genomes and investigated meiotic pairing of the individual B/A-S homoeologous pairs in the presence as well as absence of Ph1. This allowed us to monitor the effect of Ph1 and Ph1 suppressors in meiotic homoeologous pairing and to precisely assess homology for the individual B/A-S homoeologous pairs.
Substantial meiotic pairing was observed between CS wheat chromosome 1B and Ae. speltoides chromosome 1S in the presence of Ph1 and absence of other S-genome chromosomes. The 1B-1S pairing frequency (50.00%) was significantly higher than any other B/A-S homoeologous pairs (0.00-8.63%). Chromosome 1S does not contain a Ph1 suppressor gene (Dvorak et al. 2006). Therefore, no Ph1 suppressor was involved in the 1B-1S meiotic pairing analysis. It was homology that made CS wheat chromosome 1B pair with Ae. speltoides chromosome 1S in a relatively high frequency. A small Ae. speltoides-derived chromosomal segment was detected at the distal end of CS wheat 1BL by GISH. Also, we found that 1B-1S meiotic pairing occurred mostly on the long arms (i.e. 1BL-1SL) that share the Ae. speltoides segment at the distal ends. Therefore, this Ae. speltoides-derived segment was the homologous counterpart on chromosomes 1B and 1S that initiated high meiotic pairing between them. Extremely low meiotic pairing (0 out of 105 PMCs) was observed between CS wheat chromosome 1B and Th. elongatum chromosome 1E in the presence of Ph1. Also, a Th. elongatum-specific segment was not detected on 1B and other B-genome chromosomes of CS wheat by GISH. As a control, these findings further confirmed the Ae. speltoides origin of the distal 1BL region in CS wheat.
We detected the Ae. speltoides-originated chromosomal segment at the distal end of 1BL in a representative worldwide collection of tetraploid (including wild and cultivated emmer wheat) and hexaploid wheat (n=179) in addition to CS wheat. Apparently, this Ae. speltoides-originated segment on 1BL has been part of the B genome in both tetraploid and hexaploid wheat species probably since the initial incorporation of the B genome into tetraploid wheat. Also, we found that tetraploid wheat had higher genetic diversity than hexaploid wheat within the Ae. speltoides-originated genomic region on 1BL, suggesting an evolutionary pattern similar to other genomic regions of polyploid wheat (Doebley et al. 2006). All these new findings consistently support the conclusion that Ae. speltoides had been involved in the origin of the wheat B genome. The Ae. speltoides-originated chromosomal segment on wheat chromosome 1BL has been retained in the B genome probably throughout the entire evolutionary and domestication process of polyploid wheat.
High-throughput SNP genotyping of individual B/A-S homoeologous pairs and the 88 representative wheat species/accessions identified a sizable highly monomorphic linkage block (12.7 cM) at the distal ends of wheat 1BL and Ae. speltoides 1SL. This was surprisingly consistent with the meiotic pairing and GISH results, supporting the Ae. speltoides origin of the distal segment on 1BL. The SNP-defined monomorphic genomic region and extended distal end on 1BL was estimated to be 10.46 Mb in physical size based on the genomic DNA sequence assemblies of chromosome 1B currently available from IWGSC RefSeq v1.0 (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies). This estimate is about 1.2% of the total length of chromosome 1B (849 Mb) (Šafář et al. 2010). The GISH-detected Ae. speltoides segment on 1BL spans approximately 2.0% of the cytogenetic length of chromosome 1B. In addition, the DNA sequence assemblies of the IWGSC RefSeq v1.0 cover approximately 81.3% of the entire chromosome 1B (689.9 Mb out of 849 Mb) (Šafář et al. 2010). Thus, the monomorphic linkage block is probably a major portion of the Ae. speltoides-originated chromosomal segment on 1BL. The actual physical size of the Ae. speltoides-originated distal segment on 1BL might be greater than the estimate of 10.46 Mb.
Relatively low meiotic pairing was observed with each of other B/A-S homoeologous pairs in the presence of Ph1. In addition, we did not detect any Ae. speltoides-originated chromosomal segments in the other regions of chromosome 1B and on other B-genome chromosomes by GISH. Are there additional Ae. speltoides-originated chromosomal segments that are too small to leverage meiotic pairing and undetectable by GISH in the wheat B genome? We observed small monomorphic linkage blocks at multiple locations in other regions of chromosome 1B and on other B-genome chromosomes, but none was comparable in size to the one at the distal end of 1BL. Also, a clear association of those monomorphic linkage blocks with meiotic pairing could not be established in this study. Thus, we were unable to determine whether those B-genome chromosomal regions originated from Ae. speltoides. Further studies, such as genome-wide sequence comparative analysis, are needed to uncover the evolutionary relationship of those genomic regions with Ae. speltoides.
In summary, we conclude that Ae. speltoides had been involved in the origin and evolution of the wheat B genome. The current form of Ae. speltoides should not be considered an exclusive donor of the B genome. Ae. speltoides is probably one of the diploid ancestors involved in the evolutionary lineage of the B genome as stated in the theory of polyphyletic origin (Zohary and Feldman 1962). The wheat B genome might be a genome reconstructed from the homoeologous meiotic recombination between multiple ancestral genomes of Aegilops species, including Ae. speltoides. Further studies of Aegilops species, especially those in the Sitopsis section, may reveal additional insights into the origin and evolution of the wheat B genome.
Author contributions
W.Z. and M.Z. equally contributed to this work on crossing, meiotic pairing analysis, SNP assays, GISH/FISH, comparative analysis of the homoeologous genomic regions, and manuscript preparation. X.Z. participated in crossing, chromosome-specific marker analysis, GISH, and manuscript preparation. Y.C. participated in GISH analysis. Q.S. performed computation in DNA sequence and comparative analysis. G.M. was involved in material maintenance and crossing. S.H. ran SNP assays and participated in manuscript preparation. C.Y. participated in DNA sequence and comparative analysis. S.S.X. was involved in data analysis and manuscript preparation. X.C. designed and coordinated this work, made crosses, analyzed and interpreted all experimental data, and prepared the manuscript.
Conflict of interest statement
The authors declare that they have no conflict of interest.
Acknowledgements
We thank members of the labs involved for their help to this research and Drs. Lili Qi and Rebekah Oliver for their critical review of the manuscript. This project is supported by Agriculture and Food Research Initiative Competitive Grant no. 2013-67013-21121 from the USDA National Institute of Food and Agriculture.
Footnotes
↵1 Authors with equal contribution to this work.