Introduction

Bread wheat (Triticum aestivum L.) is the most important staple food worldwide contributing nearly 30% to global cereal production, with average productivity of 744.5 million tonnes (FAO 2020). Despite the global importance, wheat production is significantly lower than rice and maize (FAO 2020), and an increase by 70% from the present level is required to feed an additional 3 billion mouths by 2050 (Kaur et al. 2021). Maintaining crop productivity levels under changing climate, increasing vulnerability, and decreasing land is a major challenge in modern agriculture, and there is an immediate need of the hour to streamline suitable strategies to boost the yield under limited resources.

Grain weight is a major determinant of final wheat yield and starch accounts for 65–75% grain dry weight and 80% of the endosperm dry weight (Rahman et al. 2000). Photo-assimilates produced in the vegetative tissue of wheat plants were transported as sucrose (Tuncel and Okita 2013) and sink convert these photo-assimilates into starch, an important limiting factor to wheat grain yield (Zhenlin et al. 1999; Miralles and Slafer 2007). The starch biosynthetic pathway is regulated by a coordinated series of enzyme-catalyzed reactions, including granule bound starch synthase (GBSS), soluble starch synthase (SS), starch branching enzyme (SBE), starch debranching enzyme (SDE), isoamylases, pullulanase, starch phosphorylase (SP) and disproportionating enzyme (DPE) (Keeling and Myers 2010). Of these GBSS and SS catalyze the first unique step in starch synthesis (Jeon et al. 2010). SS is involved in the elongation of amylopectin polymers and is one of the most important enzymes with four isoforms (SSI, SSII, SSIII, and SSIV), (James et al. 2003). SSI isoform contributing 60–70% of the total SS activity is the major isoform responsible for synthesizing its shorter chains (dp 8–12) (Delvallé et al. 2005; Cao et al. 1999; Kumar et al. 2021). The enzyme is constitutively expressed at all stages of seed development (Park and Nishikawa 2012). It also participates in the regulation of other starch biosynthetic enzymes and (Crofts et al. 2017a, b) different SS isoforms play a synergistic role in the synthesis of amylopectin clusters, and a deficiency of an individual isoform have an impact on the grain development and starch accumulation (Fujita et al. 2007).

Reduction in starch quantity is one of the major factors for the reduction in yield and yield components of wheat in response to different biotic and abiotic stresses. Downregulation of starch deposition has been reported in the kernel as high temperature impairs the conversion of sucrose to starch (Keeling et al. 1994; MacLeod and Duffus 1988; Yamakawa and Hakata 2010; Phan et al. 2013; Sehgal et al. 2018; Yang et al. 2019). Among the enzymes involved in the wheat starch biosynthetic pathway, soluble starch synthase (SS) is the most labile to high temperature (Keeling et al. 1993; Hurkman et al. 2003; Kumar et al. 2021) and a decline in SS activity leads to reduced starch accumulation in wheat (Prakash et al. 2004). Terminal heat stress can reduce the yield of wheat crop by 11.1% by 2050 worldwide and negatively influence the heading and grain filling stages (Dubey et al. 2020). Heat stress pertaining to a 5 °C increase in temperature above 20 °C accelerates the rate of grain filling, whereas grain filling duration is reduced by 12 days in wheat (Yin et al. 2009).

Utilization of genetic variation available in the gene pool and targeting the important traits for attaining stress tolerance is one of the key principles for crop improvement. The diversity of cultivated wheat has been eroded due to continuous pressure of domestication and selection, leaving bleak chances for any improvement (Donini et al. 2005; Peleg et al. 2011; Longin et al. 2014). On the other hand, wild germplasm of wheat including Triticum and Aegilops species, old wheat varieties, landraces, and exotic wheats have numerous unknown and useful alleles for plants to resist, tolerate, or avoid extreme temperatures, drought, or flooding, as well as resist against pests and diseases. Identification and transfer of valuable alleles from variable germplasm are like hitting goldmine to introgress novel alleles for wheat improvement (Dwivedi et al. 2008; Pajkovic et al. 2014; Ceoloni et al. 2017; Olivera et al. 2018; Dhillon et al. 2020). The present study was undertaken to look for diversity of SSI-7B isoform of SS gene wild progenitor and cultivated tetraploid and hexaploid wheats and its comparison with their counterpart from A genome (SSI-7A) and D genome (SSI-7D) wild species.

Materials and methods

Plant material

A set of 13 different variable wheat genotypes, including wild and cultivated wheat germplasm, were selected for this study. Wild genotypes included two accessions of diploid Aegilops speltoides (SS) and two accessions of tetraploid Triticum dicoccoides (AABB). Besides, one accession each of T. monococcum (AmAm) and T. boeoticum (AbAb) and two accessions of Ae. tauschii (DD) were selected to amplify and compare the SSI-7A and SSI-7D homeologue of the gene. Cultivated wheat genotypes include one tetraploid (T. durum, AABB) MACS9 and ten hexaploid wheat (T. aestivum AABBDD) cultivars, consisting of two winter wheat and eight spring wheat genotypes. Among the spring wheat, two were old tall Indian wheats and six dwarf wheats. The list of these selected genotypes is given in Table S1.

Primer designing for homoeologous SSI gene and in vitro DNA amplification

Partial nucleotide sequence (5156 bp) of SSI-7B (Traes_7BS_6135B1D85.1) gene was used to design gene-specific primers using the PerlPrimer tool (Marshall 2004). Two sets of different overlapping primers were designed (Table S2). One primer pair SSL1 amplifies 2289 bp sequence from nucleotide positions 130–2419, while the second primer pair SSL2 amplifies 2677 bp sequence between nucleotide positions 2229 and 4906. These primers could not amplify the expected fragments in wild wheat. Thus, the second set of seven overlapping primer pairs was designed for the amplification of gene in seven small fragments as SSOL1 (130–879), SSOL2 (700–1555), SSOL3 (1320–2164), SSOL4 (1970–2864), SSOL5 (2726–3598), SSOL6 (3410–4290) and SSOL7 (4152–4906) with the amplicon length ranging from 830 to 900 bp. The same set of seven overlapping primers was also used for sequencing PCR amplicons. Two additional sequencing primers, SSOL4A, and SSOL4B were also designed for sequencing two fragments of size 2289 bp and 2419 bp.

The genomic DNA was isolated from young leaves of field-grown plants sown in the month of November with standard practices for wheat cultivation and quantified using 0.8% agarose gel electrophoresis. A PCR reaction of 25 μL volume containing 100 ng template DNA (4 μL), 5X PCR buffer plus Mg2+ (2.5 μL), 10 mM dNTPs (2.0 μL), 5 pM forward and reverse primers (1.25 μL), 5 units/μL Takara Ex Taq DNA Polymerase (0.25 μL), and sterile water (13.75 μL) was used for amplification. Amplification for 2289 bp and 2677 bp amplicons was done as initial denaturation at 94 °C for 5 min, followed by denaturation and primer annealing. The first step included 14 cycles of 94 °C for 1 min, 60 °C for 1 min and 68 °C for 3 min followed by the second step of 16 cycles of 94 °C for 1 min, 60 °C for 1 min, and 68 °C for 3 min with 20 s increment per cycle. The last step included a final extension of 10 min at 68 °C. The PCR reaction was loaded and separated on a 0.8 percent (w/v) agarose gel.

Molecular cloning of SSI gene

The PCR products were purified using QIAquick Gel Extraction Kit (Qiagen®) and cloned into a pGEM®-Teasy vector (Promega) and transformed into E. coli, DH-5α host strain. The white recombinant clones were selected on LB-Amp-Xgal agar plates using blue/white screening. Positive clones were identified with colony PCR using the gene-specific primers, and the product was resolved on 1.2 percent agarose gel. Plasmid DNA was isolated from positive clones using R.E.A.L® prep 96 (QIAGEN). Cloning of candidate gene fragments was done in tetraploid and hexaploid genotypes only, while in diploids, namely T. monococcum, T. boeoticum, Ae. speltoides and Ae. tauschii, the PCR product was sequenced directly after purification.

Sequencing of SSI gene

The sequencing of PCR products was done for both the complementary strands using overlapping primers. The sequences were extracted from chromatogram files using CHROMAS Lite 2.1.1 (http://technelysium.com.au/), and a full-length sequence of SSI gene was completed by manual aligning of the sequence of one strand with that of the reverse complementary sequence and high-quality contigs were generated using DNA Baser v4.23.0 (http://www.dnabaser.com/). Multiple sequence alignment of cloned sequences and reference sequence was done using ClustalX 2.1 (Larkin et al. 2007) and based on the alignment, candidate SNPs and InDels were predicted. Exonic and intronic boundaries of the SSI gene were predicted using the Artemis tool (Carver et al. 2008). The sequence of SSI genes obtained from T. monococcum and T. boeoticum were aligned with SSI-7A as reference (Traes_7AS_70ED86B15.1), sequences of two accessions Ae. tauschii aligned with SSI-7D as reference (Traes_7DS_5159E3934.1), while sequences from all other genotypes of Ae. speltoides, T. dicoccoides, T. durum, and T. aestivum, were aligned using SSI-7B as reference (Traes_7BS_6135B1D85.1).

Phylogenetic analysis

Primary structure and amino acid sequences were predicted using the Artemis tool (Carver et al. 2008). The amino acid substitutions in SSI between different wheat genotypes were identified visually with respect to the reference, using the multiple alignment viewer JalView 2.0 (www.jalview.org). Phylogenetic and molecular evolutionary analysis of the multiple sequence alignment of protein sequences from selected genotypes and the monocot and dicot SSI sequences were carried out using the MAFFT tool (Katoh and Standley 2013). The phylogenetic tree of these sequences was constructed using Phylip software using the neighbor-joining method (Felsenstein 2005).

Results

In-silico candidate gene identification

Nucleotide sequences of SSI genes selected from NCBI (https://www.ncbi.nlm.nih.gov/) and EMBL (https://www.ebi.ac.uk/ena/browser/home) databases were clustered to remove redundant sequences using the CD-HIT Suite tool (Fu et al. 2012) at 95% identity, shortlisting ten representative sequences. Three full-length coding sequences (CDS) of SSI gene from the TriFLDB database (Mochida et al 2009) were clustered into these ten representative sequences at 99% identity narrowing down to seven sequences which then used as a query for standalone (offline) blast (Tao 2010) against cDNA and gDNA IWGSC (Ensembl Plants; https://plants.ensembl.org/Triticum_aestivum/) T. aestivum databases. A total of 40 and 56 hits were obtained in cDNA and gDNA databases, respectively. Based on the common hits, maximum query coverage, and bit score, three sequences on homoeologous chromosomes 7A, 7B, and 7D, were selected. The expression of these three homoeologous sequences was more in grain and spike compared to leaf, root, and stem (http://wheat.pw.usda.gov/WheatExp/) (Pearce et al. 2015). Expression of homoeologue 7B (gene ID Traes_7BS_6135B1D85.1) specific transcript was much higher in grains along with a higher expression of this transcript under heat and drought stress. (Fig. S1, S2).

Characteristic of SSI gene

A partial sequence of SSI-7B gene was amplified from all the 13 genotypes, including two di1ploid (SS), three tetraploid (AABB), and eight hexaploid (AABBDD) wheat genotypes. Similarly, partial sequence of SSI-7A and SSI-7D was also amplified from two AA genome and two DD genome-specific diploid accessions, respectively. None of the sequenced loci contained more than one SSI gene (Fig. S3, S4, S5). Thus, the SSI sequence amplified from all the 17 different genotypes has a similar structure with nine exons and eight introns as reference gene. The sequences were deposited in the GenBank with accession numbers MN75891 to MN75907.

Exonic/intronic variation

Length of partial SSI-7B gene in 13 diploid, tetraploid and hexaploid genotypes varied from 3626 to 3709 bp. Total exon length varied from 1059 to 1146 bp, with the first exon being the largest (281-360 bp), followed by the fifth (173 bp) and third (124 bp) exon. Exon-2, 4, 6, 7, and 8 were smaller with 62–91 bp length (Table S3). The total length of eight introns varied from 2563 to 2575 bp, with the third intron being the largest (1079–1083 bp) and the eighth being smallest (70–82 bp). Intron-1, 3, and 8 showed some length variation, while intron-2, 5, 6, and 7 were conserved across diploid, tetraploid, and hexaploid. Within hexaploid genotypes, SSI-7B had similar sizes (3708–3709 bp) with length variations in intron-1 (C306), intron-8 (Arbon and WH542) and exon-9 (WH542). Among three tetraploid (AABB) wheat genotypes, T. durum and T. dicoccoides, length variations were found in intron-1, exon-2, and intron-3. In the case of diploids (SS), intron-1, 3, and 8 showed length variations which were also prominent across the three species used in this study. The differences in the first and last exons were due to differences in the position of the start and stop codons in the amplified sequence as these codons were identified manually.

The partially amplified sequence of SSI-7A gene from T. monococcum and T. boeoticum was longer (3854 bp) than SSI-7B with 1122 bp exonic length and 2732 bp intronic length. Nine exons and eight introns were of similar length in both A-genome species. The exonic length was also identical to SSI-7B amplified sequences, although the seventh intron (293 bp) showed significant length variation. Exonic and intronic lengths in SSI-7D sequence of two Ae. tauschii accessions had a total gene length of 3704–3705 bp with an exon length of 1118 bp and intron length of 2586–2587 bp, respectively. These two accessions were similar except for one base pair difference in the third intron. Exonic regions of the SSI gene amplified from three homoeologues of chromosome 7 showed more proximity than intronic regions as exon 2–7 were conserved across three homoeologues. At the same time, intron-6 was the only intron that had conserved length among three homeologues.

SNPs/indel in genomic sequences

In total, 44 exonic SNPs and 282 intronic SNPs were identified in the partly amplified SSI-7B gene in 13 wheat genotypes (Fig. 1, Table S4). Maximum exonic SNPs were detected from two diploid S-genome accessions of Ae. speltoides, while there was no exonic SNP in T. dicoccoides acc. pau14801 and four hexaploid wheats (Giza, Arbon, C306, PBW 343). Exon-1, being the longest, had a maximum of 18 SNPs, with 2–5 SNPs in the rest of the exons (Fig. 1). C ↔ T (transition) was the most commonly occurring exonic SNP detected across all the genotypes. Several intronic SNPs outnumbered exonic SNPs irrespective of genotype. The first and third introns were highly variable with 99 and 156 SNPs, respectively, as expected from their comparatively longer size. Intron 2, 6, and 8 were conserved without any SNP. No insertions were found in the sequenced SSI-7B, although deletions were present in the intronic regions only.

Fig. 1
figure 1

Exonic SNP positions represented individually on SSI gene sequence as amplified in 17 different wild and cultivated wheat genotypes

Full size image

There were nine exonic and 24 intronic SNPs in the eight hexaploid wheats with an average of 1SNP/824 bp (Fig. 2; Table 1; Table S4). No indel was detected in any of the hexaploid lines. Of the 21 SNPs representing 20 haplotypes in three tetraploid wheats, there were five exonic and 16 intronic SNPs with the frequency of 1SNP/523 bp in 10974 bp total gene length (Table 1). T. dicoccoides acc pau14801 also had one single base pair deletion in intron-3.

Fig. 2
figure 2

Exonic SNP positions in SSI gene represented as a combined figure for each homeologue (from chromosome 7A, 7B and 7D) in amplified sequences from 17 different diploid, tetraploid and hexaploid wheats

Full size image
Table 1 Frequency of exonic and intronic SNPs of the SSI sequences identified in different diploid, tetraploid and hexaploid wheats
Full size table

In two Ae. speltoides accessions, maximum exonic (21), and intronic (126) SNPs were identified with the frequency of 1SNP/50 bp, representing 124 haplotypes (Table 1). The first exon had the maximum number of SNPs (7), while other exons had 2–4 SNPs except for exon-2, which did not have any SNP. Intron-1 (54 SNPs) and intron-3 (62 SNPs) have a large number of SNPs as compared to introns-4, 5, and 7 with 3 SNPs only (Table S4).

There were six exonic and 89 intronic SNPs in two A-genome species with the frequency of one exonic SNP per 374 bp (Fig. 2; Table 1) and one intronic SNP/61 bp, respectively, representing 58 haplotypes. Exonic SNPs were concentrated only in exon-1 and 4, while intronic SNPs were more concentrated in intron-1 and 3, with no SNPs in intron-2,6,7 and 8. Four deletions were also detected in intron-1, 3, and 8. A total of 22 haplotypes representing 25 SNPs, including three exonic and 22 intronic SNPs, were found in D-genome species. These SNPs were found only in exon-1 and intron-1, and 3. The frequency of exonic SNPs was lower with 1SNP/745 bp than intronic 1SNP/235 bp. It represented the second-lowest frequency of exonic SNPs after the 1/1009 SNP of SSI-7B in hexaploid (AABBDD) genotypes followed by 1/651 SNP of tetraploid (AABB), 1/374 SNP of AA, and 1/104 SNP of SS genome species (Fig. 2; Table 1). No Indels were detected in D-genome species.

Thus, the frequency of SNPs was lower in hexaploids than in tetraploids and diploids. Overall, introns were more diverse than exons. Among diploids, the highest diversity was observed in Ae. speltoides (SS) compared to A-genome or D-genome species in both the intronic and exonic region.

Transitions and transversions

SNPs involve both transition (Tr), change between the purine-purine/pyrimidine-pyrimidine, or transversions (Tv), changes between the purines and pyrimidines. In SSI-7B amplified sequences, more transitions than transversions were detected with Tr/Tv ratio of 0.6:1, 0.3:1, and 0.42:1 in diploid, tetraploid, and hexaploid (Table 2). Maximum Tr (80) and Tv (44) were found in Ae. speltoides followed by A-genome species with 37 Tr and 21 Tv and hexaploid with 31 Tr and 4 Tv. G ↔ T Tv was completely absent in tetraploids, while G ↔ C Tv was missing in hexaploid. Exonic regions also lack C ↔ A type Tv in tetraploid and T ↔ A type Tv in tetraploid and hexaploid. Hexaploid genotypes (MACS 9, Giza, Arbon, PBW343, and Halna) had Tr only, C306 has Tv only, while C591, Impala, and WH542 had both Tr and Tv.

Table 2 Transition and transversions in SSI gene corresponding to SNPs identified in different diploid, tetraploid and hexaploid wheats
Full size table

In SSI-7A and SSI-7D homeologue, specific amplified sequences, 37 and 13 Tr and 21 and 9 Tv were found, respectively, with a Tv/Tr ratio of 0.57:1 and 0.69:1 in A- and D-genome diploid species (Table 2).

Phylogenetic and molecular evolutionary analysis

The sequence of full-length SSI gene obtained from 17 genotypes was clustered into three groups based on phylogenetic analysis conducted using Phylip software (Fig. 3). T. monococcum and T. boeoticum accessions were clustered with A-genome reference (Traes_7AS_70ED86B15.1). While, Ae. tauschii acc. pau3747 and Ae. tauschii acc. pau14102 were clustered with D-genome reference (Traes_7DS_5159E3934.1). Rest of the genotypes, Ae. speltoides, T. dicoccoides, T. durum, and T. aestivum were grouped with B-genome reference (Traes_7BS_6135B1D85.1).

Fig. 3
figure 3

Neighbor-joining phylogenetic tree of SSI gene sequence amplified from 17 different diploid, tetraploid and hexaploid wheats along with SSI-7A, SSI-7B and SSI-7D reference sequences. Red: B-genome sequences, Green: A-genome species, Blue: D-genome species (color figure online)

Full size image

To understand the relationship between SSI proteins examined in the current study, a brief phylogenetic analysis of SSI protein was done in the major monocot crops (wheat, rice, maize, barley) along with Glycine max and Arachis hypogaea. The analysis clustered dicots and monocots into two different clusters (Fig. 4, Table S5). As expected, closely related species showed fewer sequence variations than the distantly related species and were clustered together. Among the monocots, the SSI gene also diverged into three major clusters first one with maize (Zm) and sorghum (Sb), the second with rice (Os), and the third with wheat (Ta), barley (Hv), and Brachypodium (Bd) along with few genotypes from T. urartu (Mishra et al. 2017). All the genotypes of the current study were grouped into a third cluster (Figs. 4, 5).

Fig. 4
figure 4

Neighbor-joining phylogenetic tree of SSI proteins corresponding to amplified gene sequences from 17 different diploid, tetraploid and hexaploid wheats in the present study along with representative SSI protein sequences from monocots and dicots. Predicted amino acid sequences of the SSI genes were obtained from the NCBI (List of Accession Numbers is given in Supplementary Table 4) Bootstrap values are shown as percentages for 100 replicates

Full size image
Fig. 5
figure 5

Non-synonymous substitutions (with position) lying in the GT-5 domain region (1-6 exons); of SSI amplified sequences from six different wheat genotypes. Colored box represents exon; bar line (-) represents intron

Full size image

Detection of variations in protein sequences

Out of 34 exonic SNPs in SSI-7B amplified sequence, 24 showed synonymous substitutions (SSt) while 10 were non-synonymous substitutions (NSSt). SSt were absent in T. dicoccoides acc. pau7107, MACS9, and C591, as all the exonic SNPs contribute to non-synonymous substitution. On the other hand, NSSt was absent in hexaploid wheats WH542 and Halna. T. dicoccoides acc. pau7107 genotype contributed maximum NSSt (3) followed by Ae. speltoides acc. pau15081 (2), Impala (2) C591 (2), and MACS9 (1) (Table

Table 3 Details of non-synonymous substitution as identified in seven of the wheat genotypes in exonic regions of SSI gene
Full size table

3).

In SS1-7D, one NSSt was detected in each of two D genome accessions of Ae. tauschii while no NSSt was detected in SS1-7A of A-genome species. The rate of SSt and intronic substitutions were more in the partially amplified SSI gene. The maximum number of NSSt (3) was identified for amino acid D-Aspartic acid, followed by two NSSt for amino acid A-Alanine and E-Glutamic acid. While one for L-Leucine, Q-Glutamine, T-Threonine, N-Asparagine, and P-Proline (Table 3).

Structure prediction of SSI protein

Starch synthase catalytic domain belonging to the glycosyltransferase 5 (GT-5) family was also found in the SSI gene sequences from cultivated and wild genotypes under study. Analysis results using the Pfam program suggested that in the selected genotypes, the alignment of the GT-5 domain starts from amino acid on 35th position and ends at amino acid 293. The predicted GT-5 domain was found to be of similar length in all the selected wheat genotypes. This domain lies from exons 1- 6, and six SNPs corresponding to non-synonymous substitutions fall in this region (Fig. 5; Table S6).

Discussion

Four different isoforms of the SS gene and its chromosomal locations have been reported as SSI and SSII (homeologous group 7) (Li et al. 1999a, b; Gao and Chibbar 2000; Peng et al. 2001; Shimbata et al. 2005; Huang and Brlé-Babel 2010; Li et al. 2013; McMaugh et al. 2014), SSIII and SSIV (homeologous group 1) (Li et al. 2000; Pan et al. 2011). SSI is a major gene in wheat starch synthesis and contributes 60–70% of the SS activity (Cao et al. 1999; Fujita et al. 2006; Zeeman et al. 2010). It is constitutively expressed at all stages of seed development with a higher transcript level than SSII and SSIII during the early seed development stage (6 DAP) and steady expression thereafter throughout endosperm development (15 DAP) (Park and Nishikawa 2012; Kumar et al. 2021).

Polyploid wheat underwent a strong differentiation compared to wild ancestral species, significantly decreasing the genetic diversity on major loci. The progenitors of common wheat are a rich reservoir of various economically important traits, and diverse alleles of these species can be utilized in wheat improvement (Lopes et al. 2015). We compared the diversity and genetic differentiation of the most important isoform SSI-7B among hexaploid (spring wheat, winter wheat, tall wheat), tetraploid (T. dicoccoides and T. durum), and diploid wheats and their comparison with SSI-7A and SSI-7D isoforms from A- and D- genome species. There is an apparent reduction in the diversity of SSI-7B in hexaploid wheat, which is visible in the number of SNPs with the progression from diploids (Ae. speltoides) to tetraploids and hexaploids. The frequency of SNPs (exon + intron) varied from 1SNP/50 bp in Ae. speltoides to 1SNP/523 bp in tetraploid and 1SNP/824 bp in hexaploids. These frequency estimations are clear signs of loss of diversity in both intronic and exonic regions at different ploidy levels.

Similarly, the number of haplotypes was reduced from 124 in Ae. speltoides to 20 in tetraploids and 35 in hexaploids. SNP frequency in A-genome was 1SNP/81 bp, which is more than that of the D-genome species (1SNP/296 bp). Introns were found to be more diversified in terms of size and nucleotide sequences (Breathnach and Chambon 1981; Haga et al. 2002; Ramakrishna et al. 2002) than exons as the exons are under intense selection pressure and eliminate the deleterious changes, whereas introns have a high buffering rate for any change. In SSI high frequency of variations were found in intronic regions than in exonic regions. These variations were more pronounced between the different species.

Similarly, all the deletions observed were in diploid and tetraploid and in the intronic region, while no Indel was found in hexaploid wheat. In another isoform of starch synthase genes, SSIIa represented more allelic variation in the intronic regions (Huang and Brûlé-Babel 2012). Similarly, the sequenced AGP-L-B allele of AGP-L genes had 13 indels and 58 SNPs in the introns compared to only nine SNPs in the exon (Rose et al. 2016).

The rate of transition was more than transversion in all three homologous sequences. Transitions are usually favored over transversions (Sankoff et al. 1976) as transition mutations do not alter the 3D structure of the protein. These mutations result in synonymous substitution because selection pressure acts to conserve the chemical properties of amino acids (Vogel and Kopun 1977). Transitions are about twice as frequent as transversions in rice (Hayashi et al. 2004) and maize (Batley et al. 2003). A higher number of transversions in diploid and tetraploid species than hexaploid indicated clear preservation of diversity, absent in cultivated hexaploids making wild non-cultivated germplasm an important source of new alleles. Luo et al. (2016) provided evidence that universal bias occurs in favor of transitions over transversions. This bias results as the process of transversion mutations, including size conformation, is more complicated than transition. Leterrier et al. (2008) studied homology modeling for protein structure prediction of SS gene using Tool for Incremental Threading Optimization (TITO) and have suggested that valine residue in the K-X-G-G-L conserved motif in SSIII and SSIV isoforms could be an important factor for protein specificity as compared to SSI and SSII isoforms.

SS gene has a conserved function

SS gene is known to encode a protein with a glucosyltransferase domain. In most known SS isoforms, the GT-1 domain occupies the C-terminal half, and the GT-5 domain occupies the N-terminal half of catalytic active motifs. The GT-5 domain is responsible for binding with glucosyl donor, i.e., ADPglucose (Keeling and Myers 2010). Leterrier et al. (2008) studied starch synthase glycosyl transferase (GT-5) domain homology, which was compared to prokaryotic SS and revealed that GT-5 domain (Pfam PF08323) was conserved among all SS (SSI-SSIV) isoforms.

Exon 1–6 corresponds to the GT-5 domain of the N-terminal half, and six crucial non-synonymous SNPs have been identified in this region. Of these six SNP-based alleles, four were from wild progenitor species of Ae. tauschii acc. pau14102, Ae. tauschii acc. pau3747, Ae. speltoides acc. pau15081 and T. dicoccoides acc. pau7107, each present in exon-1 and two from cultivated wheats Impala and C591 present in exon 2 and 4, respectively. SSI involved in starch synthesis losses half of its activity when a temperature ~ 35 °C is experienced by the crop (Keeling et al. 1994) but a higher average starch deposition at high temperature is required to combat terminal heat stress. Waines (1994), Pradhan et al. (2012), and Awlachew et al. (2016) reported that Ae. speltoides could serve as a source of genetic variability for improved thermotolerance in wheat.

PAU, Ludhiana has a collection of > 1500 wild progenitor and non-progenitor species and different germplasm of hexaploid wheat that have been evaluated for the past many years for different biotic and abiotic stress-related parameters and agronomic traits. Wild and cultivated wheats of different ploidy were selected for the current study, as these species carry tremendous variation for biotic and abiotic stresses (Kaur et al. 2018). Ae. tauschii accessions were selected for having stay green and bold grain traits while T. dicoccoides accessions were selected for their bold grain, expecting a better process of starch deposition in these accessions. On the other hand, Ae. speltoides accessions have been identified to be better performing under high temperatures (Awlachew et al. 2016). Impala is a winter wheat collection, and our studies showed it has a stay-green trait under North Indian environmental conditions. C591, on the other hand, derived from local landrace material, belongs to a group of tall traditional cultivars grown under rainfed conditions in Punjab before the advent of semi-dwarf wheat in the 1960s. These tall wheats were found to be resistant to abiotic stress tolerance (Almeselmani et al. 2012). The six haplotypes corresponding to synonymous substitutions, identified in the present study could serve as new haplotypes contributing to better grain filling and starch deposition though only phenotypic confirmation has yet been done.

Conclusion

Starch is the major component of mature wheat grain and significantly determines the ultimate yield of the wheat crop. However, the increments in the global temperature especially during the seed development have raised some serious concerns across the world. The elevated temperatures during these stages affect the major biosynthetic pathways/enzyme (s) involved in starch deposition and grain development. One of these enzymes, Starch Synthase I, was targeted in the present study with an aim to understand the allelic diversity available in the cultivated and wild germplasm of wheat. In total, six SNPs (four from wild species; two from cultivated species) present in the Exon-1, -2, and -4 were identified in the SSI gene and were associated with the GT domain of the starch synthase protein. Exploitation and transfer of allelic variation have proven to be a useful resource for wheat improvement. A number of allelic variants have been identified for various candidate genes in wheat such as Pm3 for powdery mildew resistance (Kaur et al. 2008), Wx-A1 for waxy gene and amylose biosynthesis (Saito and Nakamura 2005), Ap1 and PhyC for vernalization response (Beales et al. 2005), SSII for endosperm starch biosynthesis (Shimbata et al., 2005), Wx-B1 for waxy protein (Monari et al. 2005), PSY-1 and PSY-2 for grain yellow pigment content (Zhang and Dubcovsky 2008), Viviparous-1 for pre-harvest sprouting tolerance (Xia et al. 2008) and TaALMT1 for aluminum resistant (Raman et al. 2008). Identification of diversity and novel variants in the present study indicated that wild and cultivated germplasm is a good resource for useful alleles and the positive effects of these haplotypes could be further validated through molecular and phenotypic evaluation.