|Home | About | Journals | Submit | Contact Us | Français|
High molecular weight glutenin subunits (HMW-GSs), encoded by the genes at Glu-1 loci in wheat and its related species, are significant in the determination of grain processing quality. However, the diversity and variations of HMW-GSs are relatively low in bread wheat. More interests are now focused on wheat wild relatives in Triticeae. The genus Aegilops represents an important germplasm for novel HWM-GSs and other useful genes for wheat genetic improvement.
Six novel Glu-1 alleles and HMW-GSs were identified and characterized from three species of Aegilops section Sitopsis (S genome). Both open reading frames (ORFs) and promoter regions of these Glu-1 alleles were sequenced and characterized. The ORFs of Sitopsis Glu-1 genes are approximately 2.9kb and 2.3kb for x-type and y-type subunits, respectively. Although the primary structures of Sitopsis HMW-GSs are similar to those of previously reported ones, all six x-type or y-type subunits have the large fragment insertions. Our comparative analyses of the deduced amino acid sequences verified that Aegilops section Sitopsis species encode novel HMW-GSs with their molecular weights larger than almost all other known HMW-GSs. The Glu-1 promoter sequences share the high homology among S genome. Our phylogenetic analyses by both network and NJ tree indicated that there is a close phylogenetic evolutionary relationship of x-type and y-type subunit between S and D genome.
The large molecular weight of HMW-GSs from S genome is a unique feature identified in this study. Such large subunits are resulted from the duplications of repetitive domains in Sitopsis HMW-GSs. The unequal crossover events are the most likely mechanism of variations in glutenin subunits. The S genome-encoded subunits, 1Dx2.2 and 1Dx2.2* have independent origins, although they share similar evolutionary mechanism. As HMW-GSs play a key role in wheat baking quality, these large Sitopsis glutenin subunits can be used as special genetic resources for wheat quality improvement.
High molecular weight glutenin subunits (HMW-GSs) are important storage proteins in seed endosperms of wheat and its related species [1,2]. Due to their composition effects on the elasticity of wheat dough, HMW-GSs play a significant role in the determination of grain processing quality . HMW-GSs are encoded by the genes at Glu-1 loci on the long arms of the Group 1 chromosomes (1A, 1B and 1D) in bread wheat. HMW-GSs can be further classified into two subfamilies (x-type and y-type), which are thought to have arisen from gene duplication events. Single copy of x-type and y-type gene occurs at two tightly linked loci, Glu-1x and Glu-1y. The HMW-GSs encoded by Glu-1x or Glu-1y can be distinguished from each other by the difference in their peptide lengths [1,4]. Previous studies indicated that allelic polymorphism in wheat HMW-GSs is associated with variations in the gluten viscoelasticity and bread making quality . Up to now, a number of Glu-1 alleles and HMW-GSs have been identified and characterized from wheat and its related species [5-18]. Sequence analyses of HMW-GS coding regions revealed that the primary structure of mature HMW-GSs consists of a central repetitive domain flanked by the conserved N-terminal and C-terminal regions . The repetitive domain is mainly composed of repeat motifs including tripeptide, hexapeptide and nonapeptide. The difference among various HMW-GSs is mainly resulted from variable number of repeat motifs in the repetitive domains [2,19].
The section Sitopsis of genus Aegilops contains five species: Aegilops bicornis, (Forsskal) Jaub. & Spach. (SbSb, 2n = 2x = 14), Ae. longissima (Schweinf. & Muschl.) Á. Löve. (SlSl, 2n = 2x = 14), Ae. sharonensis (Eig) Á. Löve. (SshSsh, 2n = 2x = 14), Ae. searsii, (Feldman & Kislev ex Hammer) Á. Löve. (SsSs, 2n = 2x = 14) and Ae. speltoides, (Tausch) Á.Löve, (SS, 2n = 2x = 14) . Previous reports on cytogenetic and genetic investigations indicated that Aegilops genomes from this section of five species are closely related [21-24]. Although the characterization of HMW-GSs in two accessions of Ae. searsii have been reported , the Glu-1 alleles and HMW-GSs in other four Sitopsis species have not been investigated. Our preliminary study surveyed the expression of HMW-GSs in Ae. bicornisAe. longissima and Ae. sharonensis and realized that the Sitopsis Glu-1 alleles encodes the glutenin subunits with molecular weights much larger than other known HMW-GSs available in public databases. Here, we report the isolation and characterization of novel Glu-1 alleles and HMW-GSs from Ae. bicornisAe. longissima and Ae. sharonensis. The objective of this study is to investigate the structural features of Sitopsis HMW-GSs, understand the evolutionary relationship of HWM-GS gene family within Triticeae, and further explore the potentials of S genome-encoded HMW-GSs in wheat quality breeding.
The SDS-PAGE profiles on three accessions of Ae. bicornisAe. longissima and Ae. sharonensis indicated that Sitopsis HMW-GSs consist of large x-type and y-type subunits which migrate significantly slower than the same type of subunits in Chinese Spring (Figure (Figure1a).1a). Subsequent cloning of the Glu-1 ORFs further verified that the molecular weights of these Sitopsis x-type subunits are close to or larger than that of 1Dx2.2, one of largest HMW-GSs previously reported . The results of N-terminal sequencing suggested that the protein bands with slower and faster mobility are x- and y-type subunits, respectively. The obtained sequences of seed protein are perfectly matched to those deduced from the cloned genes (Table (Table11).
The PCR amplicons of Sitopsis Glu-1 alleles are composed of two DNA fragments (approximately 2.9kb and 2.3kb) for each of three accessions (Figure (Figure1b).1b). All amplified PCR products were cloned and the Glu-1 ORFs at different alleles were determined by the sequence analysis and enzyme digestions. The full length of Glu-1 ORFs was obtained by using the method of primer walking and nested deletion. Six sequences for x-type and y-type HMW-GSs from the S genome of three Aegilops species were designated as 1Sbx2.9 and 1Sby2.3 (Ae. bicornis), 1Slx2.9 and 1Sly2.3 (Ae. longissima), 1Sshx2.9 and 1Sshy2.3 (Ae. sharonensis), respectively. All these DNA sequences have been deposited into the NCBI database with the Genbank accession numbers from JN001481 to JN001486.
After removing the coding sequence for the signal peptide from the ORFs of 1Sshx2.9 and 1Sshy2.3, the modified ORFs were cloned into pET-30. Two bacterial expression constructs (pET-1Sshx2.9 and pET-1Sshy2.3) were chosen to express mature protein in bacterial cells. In the cells harboring pET-1Sshx2.9 and pET-1Sshy2.3, IPTG induction led to the expression of the protein bands with electrophoretic mobility similar to those of the native x and y-type subunits from the seed extract of Ae. sharonensis (Figure (Figure22).
We predicted the amino acid sequences of six Sitopsis HMW-GSs and found that their primary structures are composed of four regions, i.e. a signal peptide, a central repetitive domain, the conserved N-terminal and C-terminal. The distribution and number of cysteine residues are identical to those in typical x-type and y-type subunits (Figure (Figure3,3, ,4;4; Table Table2).2). The deduced protein sequences were firstly aligned with other known HMW-GSs from A, B and D genomes. Such comparison demonstrated that there is a higher similarity between S and D genome. Therefore, we realigned the HMW-GSs identified from three Sitopsis accessions in this study with those encoded by D genome available in public databases to determine their evolutionary relationship (Figure (Figure3,3, ,4).4). Our results indicated that the S genome-encoded glutenin subunits considerably differ from other known HMW-GSs. Compared to 1Dx2, Sitopsis x-type subunits (1Sbx2.9, 1Slx2.9 and 1Sshx2.9) share the insertion of 141 residue with five tripeptides, 15 hexapeptides and four nonapeptides (Figure (Figure5a-c).5a-c). For Sitopsis y-type subunits, both 1Sly2.3 and 1Sshy2.3 have an insertion of 105 residues with seven hexapeptides and seven nonapeptides (Figure (Figure5d),5d), but this duplicated block in 1Sby2.3 only contains five hexapeptides and five nonapeptides (a total of 75 residues). We found that the peptide insertions in both x-type and y-type subunits from three Sitopsis species are copied from the adjacent regions, with some variations in single or more amino acid residues.
The 5’ flanking promoter regions of HMW-GS genes in three Sitopsis species were amplified by using two pairs of PCR primers specific to x-type and y-type subunits, respectively. All amplified PCR products were cloned and sequenced. Based on previous studies, the promoter regulatory elements of HMW-GSs are composed of TATA box, complete and partial HMW enhancers, E and N motifs . Our results indicated that the amplified promoter regions of Sitopsis HMW-GSs cover all recognized promoter regulatory elements. The DNA lengths of 1Sbx2.9, 1Slx2.9 and 1Sshx2.9 promoter are identical (904bp); whereas those of 1Sby2.31Sly2.3 and 1Sshy2.3 varied from 845bp to 919bp. The characterized promoter sequences of Sitopsis x-type and y-type HMW-GSs were aligned to homologous regions of 1Ax2*, 1Bx7, 1Dx2Triticum urartu 1Ay1By9 and 1Dy10, respectively. Multiple sequence alignments showed that both types of glutenin subunits encode the conserved domains and variable parts in their promoter regions. We found that the HMW-GS promoters mainly differentiate with base substitutions, insertions, or deletions (data not shown). All the regulatory elements in the characterized Sitopsis HMW-GS promoters share the high conservation with few substitutions (Table (Table3).3). An 85bp fragment, in which the partial HMW enhancer was included, was deleted in the 5’ flanking promoter regions of 1Sshy2.3 (Table (Table3).3). This deletion has not interrupted the expression of 1Sshy2.3.
To investigate the evolutionary relationship between S genome-encoded HMW-GS alleles and those of A, B, D genomes, we constructed the phylogeny of a network and a neighbor-joining tree (Figure (Figure6a,6a, b). The 5’ flanking promoter sequences, plus the sequences encoding signal peptides and the N-terminal, were selected for phylogenetic analysis because they have demonstrated to be phylogenetically informative. Firstly, our previous study on HMW-GS promoter indicated that the regulatory elements which control the tissue specificity and expression level of different HMW-GS genes are well conserved in diploid species of Triticeae . Secondly, the sequences encoding signal peptides and N-terminal domain are also relative conserved. Therefore, these HMW-GS sequences were suitable for phylogenetic analysis [17,28].
Our network analysis demonstrated that all HMW-GS alleles are clustered into two groups (x-type and y-type subunits) (Figure (Figure6a).6a). Both x-type and y-type subunits showed a star-like phylogeny from principle nodes. In the x-type group, 1Slx2.9 and 1Sshx2.9 are linked to 1Sbx2.9, and then form a close link to the principle node which is composed of 1Dx2, 1Dx2.2 and 1Dx2.2*. For the y-type group, 1Sby2.3, 1Sly2.3 and 1Sshy2.3 are formed a parallel link to the y-type genes encoded by D genome. The resulted NJ tree also indicated that x-type and y-type subunits are divided into two clades which support a close phylogenetic relationship on HMW-GSs between S and D genome (Figure (Figure6b).6b). Therefore, our results from both network and neighbor-joining tree demonstrated that S genome-encoded HMW-GS alleles are evolutionally related to those of D genome.
As HMW-GSs play the key role in determining wheat gluten and dough elasticity, the characterization on novel HMW-GSs from Triticeae wild germplasm will be beneficial not only for improving wheat end-use quality but also for further understanding the structure variation and evolution of this important protein family. Compared to bread wheat, wild Triticeae grasses have more HMW-GS variants. For example, a number of HMW-GS variants with novel structural characteristics have been identified from Aegilops genus [16,18,29]. However, the progress on characterizing the HMW-GS expression from Aegilops section Sitopsis species remains slow and our knowledge on their structure, function, and evolution is still limited. In this study, we report the isolation and characterization of six x-type and y-type HMW-GSs variants from the S genome of three Aegilops species (Ae. bicornisAe. longissima and Ae. sharonensis). These novel variants will be useful to widen or enrich Glu-1 genes and HMW-GSs for wheat quality breeding.
To avoid the potential error in PCR or sequencing, each nucleotide sequence was determined by multiple independent clones. The results of N-terminal sequencing indicated that the first 10 residues of N-terminal protein sequence of all 6 x and y-types subunits deduced from DNA sequences match perfectly to those directly determined by protein sequencing. And further bacterial expression proved that the cloned sequences are indeed accurate representations of the coding genes of HMW-GS in three Aegilops species. Therefore, the molecular information for GluS1 alleles obtained in this study is reliable and suitable for exploring structural differentiation and evolution of Glu-S1 alleles. Our results demonstrated that each of three Aegilops section Sitopsis species has two expressed subunits and the possession of large molecular weights is unique in both x-type and y-type subunit of S genomes (Sb, Sl and Ssh). Previous study reported that 1Dx2.2* and 1Dx2.2 are the largest HMW-GSs and their mature subunits contain 1003 and 950 amino acid residues, respectively [25,30]. In this study, we identified that the length of Ae. sharonensis subunit 1Sshx2.9 is 953 residues, shorter than 1Dx2.2* but longer than 1Dx2.2, which means 1Sshx2.9 is the second largest HMW-GSs characterized so far (Table (Table2).2). In addition, 1Sbx2.9 of Ae. bicornis and 1Slx2.9 of Ae. longissima also have large molecular weights close to that of 1Dx2.2. For y-type HMW-GS genes, the lengths of their complete ORFs are usually less than 2kb. Our previous study reported that a HMW-GS gene variant 1Ay (Ta-e3), isolated from einkorn wheat, encodes its ORF with the length of 2202bp, larger than all other known y-type genes . In this study, however, we identified two novel y-type Glu-1 alleles, 1Sly2.3 from Ae. longissima and 1Sshy2.3 from Ae. sharonensis, and determined that their ORF lengths are 2256 and 2242bp, respectively, much larger than that of 1Ay (Ta-e3) (Table (Table2).2). As the y-type HMW-GS genes with such large molecular weights have not been reported in wheat and its related species, both 1Sly2.3 and 1Sshy2.3 will be special and useful to extend our knowledge on structure, function, evolution of the y-type HMW-GSs.
Four modes have been proposed for the sequence alteration and evolution of HMW-GSs: (1) single residue changes, (2) deletion or insertion in a repeat unit, (3) single repeat changes, and (4) deletions or duplications of repeat blocks . It has been reported that the unequal crossover events and slip-mismatching are the most likely mechanism of the size variations in HMW-GSs [25,31]. In this study, we found that the large molecular weights of S genome-encoded subunits are almost entirely due to the insertion and duplication of these repeat motifs (Figure (Figure3334).4). Previous study on comparative analysis of peptide sequences indicated that 1Dx2.2 and 1Dx2.2* are evolved from the two separate duplications of 132 and 186 residues, respectively . Although three S genome-encoded x-type subunits, 1Dx2.2 and 1Dx2.2* have been resulted from duplication events, they are different in three ways. Firstly, each duplication of 1Sx subunits, 1Dx2.2 and 1Dx2.2* occur at different positions of repetitive domains (Figure (Figure5a-d;5a-d; Table Table4).4). Secondly, the duplicated regions contain the varied numbers of repeat motifs which result in distinct size of inserted fragments among 1Sx subunits, 1Dx2.2 and 1Dx2.2*. Thirdly, the inserted fragments from duplications in 1Sx subunits are not completely identical to that from which it was duplicated (Figure (Figure7).7). On the contrary, the new inserted regions in 1Dx2.2 and 1Dx2.2* are perfect copy of adjacent region without any changes. Based on above discussions, we can conclude that Sitopsis x-type subunits, 1Dx2.2 and 1Dx2.2* may have independent origins, although they share the similar evolutionary mode. We realized that such similar pattern also exists in the S genome-encoded y-type subunits.
Two structural features of HMW-GSs may be relevant to their participation as gluten polymers in the baking quality of dough. Firstly, the number and distribution of cysteine residues determine the forming inter- and intra-molecular disulphide bonds. It is well known that disulphide bonds play a key role in determining the structure and properties of wheat glutenin polymers. The presence of an additional cysteine residue in the repetitive region of subunit 1Dx5 was reported to be responsible for the correlation of this particular HMW-GS with good bread-making quality [32,33]. We found that the S genome-encoded subunits have the conserved cysteine composition, which may be important to keep the normal gluten polymer. Secondly, the properties and interactions of repetitive domains are important in determining the dough viscoelastic properties . The positive relationship between the HMW-GS sizes and their effects on dough strength has been revealed by previous studies. Belton  and Feeney et al.  proposed a model in which the gluten polymers interact via inter-chain hydrogen bonds between the subunit repetitive domains and more stable interactions can be formed with longer subunits. The experiments of incorporating the 1Dx2.2 and 1Dx2.2*subunits into dough indicated that both subunits can lead to yield the dough strength greater than 1Dx2. As both x-type and y-type subunits encoded by the S genome are larger than almost all other known HMW-GSs, we predict that the S genome-encoded HMW-GSs may have an outstanding ability to strengthen the gluten interactions. Based on our results, it will be valuable to further explore the potential values of these novel Sitopsis HMW-GS variants in modifying the structure, composition and function of wheat storage proteins. Furthermore, these special S genome-encoded genes and glutenin subunits will be helpful to overcome the bottleneck of poor genetic diversity of Glu-1 alleles and HMW-GSs in hexaploid wheat. Two approaches are under the way to verify the function of 1Sx subunits. One is to develop wheat transgenic plants which allow the endosperm specific expression of 1Sx alleles; the other is to transfer the 1Sx subunits to tetraploid or hexaploid wheats by the interspecies cross.
We have identified and characterized six novel HMW-GS variants from three Aegilops section Sitopsis species. The possession of large molecular weights is unique feature of S genome-encoded HMW-GSs. These Sitopsis glutenin subunits with large molecular weights have been resulted from the similar duplication of repetitive domains as those in the subunits 1Dx2.2 and 1Dx2.2*. The S genome-encoded subunits, 1Dx2.2 and 1Dx2.2* have independent origins, although they share similar evolutionary mechanism. Because of their molecular weights much larger than all other known HMW-GSs, these novel Sitopsis glutenin subunits can be used as special genetic resources to improve wheat quality breeding.
Sixty-five accessions of Ae. bicornis, Ae. longissima and Ae. sharonensis, kindly provided by USDA-ARS (http://www.ars-grin.gov), were investigated on their HMW-GS profiles by using the SDS-PAGE. Three accessions (CIae 70 of Ae. bicornis, PI 604122 of Ae. longissima and PI 584388 of Ae. sharonensis) with larger HMW-GS combinations were chosen for further cloning and characterization.
HMW-GSs of Ae. bicornisAe. longissima and Ae. sharonensis were extracted from the half of single seed according to Mackie et al. . HMW-GSs from hexaploid wheat cv. Chinese Spring (null, 1Bx7+1By8, 1Dx2+1Dy12) were used as a standard reference for comparison of HMW-GS electrophoretic mobility. Total seed proteins were extracted in the sample buffer containing 0.0625M Tris-HCl (pH 6.8), 2% (w/v) SDS, 1.5% (w/v) DTT, 10% (v/v) glycerol and 0.1%w/v Coomassie Brilliant Blue R250. The extracts were heated at 95°C for 5min and centrifuged for 10min. The supernatant was loaded onto a 10% (w/v) SDS-PAGE gel as described by Shewry et al. . To ensure the experimental accuracy, at least three seeds were analyzed for each accession of these three Aegilops section Sitopsis species.
After electrophoresis, the proteins were transferred from the gel onto a PVDF (Poly vinylidene fluoride) membrane by western blotting. The membrane was saturated with methanol and stained with 0.1% (w/v) Coomassie BBR250. The HMW subunit bands were then excised for protein sequencing. The N-terminal amino acid sequences of the HMW subunits were determined by GeneCore Bio-Technology company (Shanghai, China) using the PROCISETM494CLC amino acid sequencer of Applied Biosystems.
Genomic DNAs were extracted from the leaves of two-week-old seedlings by using the CTAB method . To amplify the complete coding regions of HMW-GSs, a pair of primers, P1 (5’-ATGGCTAAGCGGC/TTA/GGTCCTCTTTG-3’) and P2 (5’-CTATCACTGGCTG/AGCCGACAATGCG-3’), were designed according to nucleotide sequences in the conserved 5' or 3' ends of available HMW-GS ORF sequences. The high fidelity LA Taq polymerase (TaKaRa) with GC buffer for GC-rich template was used in the PCR amplification to minimize the errors which were introduced into the sequences. The PCR cycling parameters was 94°C for 5min, followed by 30 cycles of 94°C for 40sec, 68°C for 5min and a final extension step at 72°C for 12min . PCR products were separated in 1% agarose gels and all DNA fragments were recovered, purified and further ligated into the pMD19-T vector (TaKaRa). The ligated mixtures were transformed into Escherichia coli DH5α competent cells. The strategy of primer walking and the nest deletion method  were used to obtain the full-length of Sitopsis Glu-1 ORFs. The DNA sequencing was performed by the Invitrogen Company (Shanghai, China). Each clone was sequenced in two directions, the final nucleotide sequences for each Glu-1 ORF was determined from the sequencing results of 3 independent clones.
In order to confirm that the novel Glu-S1 genes expressed proteins that corresponded to those in the grain, we choose 1Sshx2.9 and 1Sshy2.3 as the representation of Glu-S1x and Glu-S1y for expressional experiments, as three pairs of x and y-type genes possess highly similar DNA sequences and molecular mass in Ae. bicornisAe. longissima and Ae. sharonensis. To express of the mature proteins of HMW-GS from Ae. sharonensis, we designed two pairs of primers for amplifying the mutant ORF from which the sequence coding for signal peptide was removed and introducing appropriate restriction enzyme sites of NdeI and EcoRI for the mutant ORF to facilitate following cloning and expression. The primers pairs of PET-F1 (CTCACCCATATG GAAGGTGAGGCCTCTGGGCA) and PET-R1 (GGCAATGAATTC CTATCACTGGCTAGCCGACA) were used to amply 1Sshx2.9 while the combination PET-F2 (CTCATCCATATGGAAGGTGAGGCCTCTAGGCA) and PET-R2 (GGCAAT GAATTCCTATCACTGGCTGGCCGACA) were specific for y-type genes of 1Sshy2.3. PCR conditions for amplifying mutant ORF were identical to those described above except that the template was plasmid DNA purified from the determined clones. After the mutant ORF was cloned into the expression vector pET-30a (Novagen), the recombinant construct was selected to express mature protein in the E. coli strain BL21 (DE3). Induction of bacterial expression was performed with 1mM IPTG for 3 to 5 hours. The expressed proteins were purified by extraction with 50% (v/v) propanol containing 2% (w/v) DTT, and then separated by SDS-PAGE .
Two pairs of primers were designed for amplifying promoter regions for both x-type and y-type glutenin subunits in these three Aegilops section Sitopsis species. The P3 primer (5’-AGGGAAAGACAATGGACATG -3’) was designed from the sequence which was strictly conserved in the 5’ flanking regions of all Glu-1 loci, whereas the primer P4 (5’-GTCTCGGAGC/T TGC/TTGGTC-3’) and primer P5 (5'-CATCTGGAGCCCCGTGCTC-3’) was derived from the sequence coding for 6 residues (DQQLRD) and (STGLQM), respectively. Each of sequence residues exists only in x-type and y-type HMW-GSs, respectively. The primer combinations P3 + P4 and P3 + P5 are specific for x-type and y-type promoters. The amplification profile was 94°C for 5min, followed by 35 cycles of 94°C for 40sec, 60°C for 1min, and 72°C for 1min 30sec, and a final extension step at 72°C for 7min. PCR products were purified, cloned into pMD19-T, and then sequenced. The final nucleotide sequences for Glu-1 promoters were also constructed from sequencing at least 3 independent clones.
The prediction of nucleotide sequences was performed by the DNAman software package (V5. 2. 10; Lynnon Biosoft). Multiple alignments were carried out by using Clustal W (V1.83) for comparisons of either DNA or protein sequences . Alignments were further improved by visual examination and manual adjustment. To characterize the phylogenetic relationship of HMW-GS genes, we compared the S genome-encoded Glu-1 alleles from these three Aegliops section Sitopsis species with previously characterized x-type HMW-GSs alleles represented by 1Ax2* (M22208), 1Bx7 (X13927), 1Dx2 (X03346), 1Dx5 (X12928), 1Dx2.1 (AY517724), 1Dx2.2 (AY159367), 1Dx2.2* (AY893508), and also with previously characterized y-type HMW-GSs represented by 1Ay (EU984508), 1By9 (X61026), 1Dy10 (X12929), 1Dy12 (X03041) and 1Dy10.1 (AY695379). The 5’ flanking promoter sequences, plus the sequences encoding signal peptides and the N-terminal, which is considered as phylogenetically informative , were selected to create a multiple alignment by the Clustal W program. All the nucleotide sequences and their alignments have also been listed as supplementary materials (see Additional files 12 and 3).The neighbour-joining (NJ) tree was constructed by using the software MEGA 4.02 with the substitute model of Maximum Composite Likelihood . In the NJ analysis, gaps were treated as missing data. The bootstrap values were calculated based on 1000 replications to estimate the topological robustness. For the network analysis, the sites with base substitution or mutation were used to constructed media-joining network in program Network 184.108.40.206 (http://www.fluxus-engineering.com/). The media-joining network was calculated under the default parameters of weights = 10 and epsilon = 0 .
JQT contributed to design and carry out the experiments and wrote the draft; MJ did the cloning of HWM glutenin ORFs; WYM revised the manuscript; LYX and LXJ made contribution to SDS-PAGE analysis and promoter cloning; LZX conducted the analysis of the data and review the manuscript; ZS and ZQZ finished the cloning of promoter, phylogenetic analysis and bacterial expression; ZYL contributed to improve research program. All authors have read and approved the final manuscript.
Full alignment of y-type HMW-GS ORF sequences ofAegilops specieswith those encoded by D genome.Description: The similarity was showed by different color. The deletions were showed by gaps.
Full alignment of x-type HMW-GS ORF sequences ofAegilops specieswith those encoded by D genome.Description: The similarity was showed by different color. The deletions were showed by gaps.
Full alignment of promoter sequences of Aegilops species with those ofGlu-A1, Glu-B1andGlu-D1.Description: The similarity was showed by different color. The deletions were showed by gaps.
This work was supported by the National Natural Science Foundation of China (31000167 and 31171556), the National High Technology Research and Development Program of China (863 program 2011AA100103), the National Basic Research Program of China (973 Program 2010 CB134400) and China Transgenic Research Program (2011ZX08002).