Mapping Some Seed Quality Traits in Bread Wheat ( Triticum aestivum L . ) by Association Mapping Using SSR Markers

Improvement of the end-use quality is one of the primary objectives of breeding for researchers working on raising the nutritional and functional quality of wheat. The processing and end-use characteristics of the grain, collectively known as quality traits, are under genetic and environmental control. Molecular genetic studies explaining this control may increase the efficiency of wheat breeding to improve the grain quality. Association mapping which is also referred to as linkage disequilibrium (LD) mapping has gained considerable popularity as an efficient genetic mapping methodology because of improved statistical approaches that increase its proficiency and reduce false positive associations.1 Furthermore, association mapping has been used to identify trait-marker relationships in plants.2,3 Initially, this method was extensively used to dissect human diseases.4 Unlike linkage analysis where mapping populations are used to determine correlations between phenotype and genotype, association mapping relies on unrelated individuals to create population-wide marker-phenotype associations.5 Linkage mapping based on the biparental populations is frequently used to dissect the genetic architecture of wheat quality traits. Several main effect QTLs and major genes have been identified for glutenin and grain hardness locus Ha,6 test weight,7-11 a-amylase activity,10,12 grain protein content,11,13-19 sedimentation volume,10,20-22 and grain weight.23 The main objective of the association mapping studies is detecting the correlations between genotypes and phenotypes in a sample of individuals based on the LD,24 and it is suggested as a promising alternative strategy to linkage mapping.25 There are some examples of applying association mapping via wide genome or candidate genes approaches in wheat. Breseghello and Sorrells26 used association mapping and detected the main effect QTLs for the kernel weight, length, and width. Mapping Some Seed Quality Traits in Bread Wheat (Triticum aestivum L.) by Association Mapping Using SSR Markers


Introduction
Improvement of the end-use quality is one of the primary objectives of breeding for researchers working on raising the nutritional and functional quality of wheat.The processing and end-use characteristics of the grain, collectively known as quality traits, are under genetic and environmental control.Molecular genetic studies explaining this control may increase the efficiency of wheat breeding to improve the grain quality.Association mapping which is also referred to as linkage disequilibrium (LD) mapping has gained considerable popularity as an efficient genetic mapping methodology because of improved statistical approaches that increase its proficiency and reduce false positive associations. 1urthermore, association mapping has been used to identify trait-marker relationships in plants. 2,3Initially, this method was extensively used to dissect human diseases. 4Unlike linkage analysis where mapping populations are used to determine correlations between phenotype and genotype, association mapping relies on unrelated individuals to create population-wide marker-phenotype associations. 5Linkage mapping based on the biparental populations is frequently used to dissect the genetic architecture of wheat quality traits.Several main effect QTLs and major genes have been identified for glutenin and grain hardness locus Ha, 6 test weight, [7][8][9][10][11] a-amylase activity, 10,12 grain protein content, 11,[13][14][15][16][17][18][19] sedimentation volume, 10,[20][21][22] and grain weight. 23The main objective of the association mapping studies is detecting the correlations between genotypes and phenotypes in a sample of individuals based on the LD, 24 and it is suggested as a promising alternative strategy to linkage mapping. 25There are some examples of applying association mapping via wide genome or candidate genes approaches in wheat.Breseghello and Sorrells 26 used association mapping and detected the main effect QTLs for the kernel weight, length, and width.

Mapping Some Seed Quality Traits in Bread Wheat (Triticum aestivum L.) by Association Mapping Using SSR Markers
Neumann et al used 96 bread wheat accessions and detected the main effect QTLs for several agronomic and quality traits such as protein content, sedimentation volume, and 1000-kernel weight. 27Jochen et al investigated the genetic basis of protein content, sedimentation volume, 1000-kernel weight, test weight, and starch content using an association mapping approach and detected the main effect QTLs for these traits. 28n contrast, it was found that the test weight and sedimentation volume were only controlled by epistatic QTLs.Emebiri et al 12 used a whole genome scan with DArT markers to identify chromosomal regions influencing late maturity a-amylase (LMA) in synthetic hexaploid wheat.They found significant markers at the chromosome 7B, a region previously linked to LMA in bread wheat, but not at the chromosome 3B region, they concluded that a region on chromosome 6B has potentially great interest for this trait and have a significant association with LMA phenotypes in the wheat accessions.Abdollahi Mandoulakani et al 29 investigated associations between the ISSR, IRAP, REMAP markers and the agronomic traits of wheat.They found that 94 loci were significantly associated with agronomic traits.Shahzad et al experimented the grain quality traits, genetic diversity, and marker-trait association in a range of wheat species. 30In their study, 8 QTLs were found for 5 traits including protein, gluten contents, a test weight of bread, and chapati making quality.Protein content, test weight, bread quality, and Glu-B1 were found significantly associated respectively with primers WMC419 (32 cM), WMC128 (30 cM), WMC419 (32 cM), WMC818 (17 cM), and WMC416 (44 cM).Kumari et al also identified quantitative trait loci (QTL) regulating grain traits in wheat. 31They found 18 QTLs distributed on 8 chromosomes for 7 grain traits of bread wheat.Karolina et al confirmed the predominant effect of the Glu-D1d allele on the technological properties of wheat grains. 32In a study conducted by Kaur et al it was indicated that micro-sedimentation test values are not much affected by the absence of Glu-B3/Gli-B1, and hence, the lines having better root traits with no Glu-B3/ Gli-B1 and secalin could be used for improving the bread quality and yield in wheat. 33Our study aimed to evaluate the population structure of 92 bread wheat accessions using the association mapping method to detect SSR markers linked to the loci involved in quality characteristics of the bread wheat.

Materials and Methods
Plant Material and Field Experiment A collection of 92 genotypes of bread wheat (Triticum aestivum) and 8 durum wheat (T.durum) (only used in field experiment) were used in this study (Table 1).These genotypes were cultivated in different regions of Iran and widely used in wheat production as well.This population consisted of local and modern cultivars as well as promising lines.The experiment was alpha lattice designed with 2 replications for 100 entries.The 92 bread wheat genotypes used for association mapping were evaluated according to AACC approval methods 39-25, 38-12, 56-81, and 56-60 34 for grain protein, gluten, falling number, and SDS sedimentation

DNA Extraction and Marker Assay
Genomic DNA was extracted from 100 mg fresh frozen leaves of individual plants for all genotypes grown in the plastic pot in greenhouse taking a modified CTAB method. 35DNA quality was checked by electrophoresis on 0.8% agarose gel and DNA concentration was determined by a Pico Drop (Pico200).Sixty-six unlinked SSR markers were selected and synthesized according to the information available in the Grain Genes database (http://wheat.pw.usda.gov/GG2).These markers were randomly distributed across the wheat genome.Furthermore, 36 mapped QTLs linked markers from previous studies were selected.Map positions of some of these markers were based on the linkage map published by Somers et al. 36 Polymerase chain reactions were performed in a Thermal Cycler (Bio-Rad Model thermal cycler) in a volume of 15 µL containing: 3 µL of DNA template (50 ng/mL) and 12 µL of the master mix containing 7.8 µL of ddH 2 O, 1.5 µL of 10x PCR buffer, 0.3 µL of 100 mM MgCl 2 , 0.3 µL of 10 mM dNTPs, 0.5 µL of each forward and reverse primers (1 pmol/mL), and 0.1 µL of Taq polymerase (500 U/mL).The amplification steps were as follows: 1 cycle at 94°C for 4 minutes, then 35 cycles comprising 94°C for 1 minute, annealing of primer at 50-60°C (depending on the primer) for 1 minute and then extension at 72°C for 1 minute.The final extension was carried out at 72°C for 10 minutes.The amplification products were electrophoresed on 3.5% agarose gels (50% Metaphor and 50% LE Agarose), and for staining, 3 µL Gel Red and dye (1:1 ratio) was added to each sample.Gel scanning was performed using Bio-Rad Gel Doc.
Data Analysis and Association Mapping Analysis of variance (ANOVA) for quality traits was carried out using ALPHAGEN software version 1.1. 37Two sets of SSR markers were used for association analysis.First, 66 SSR primers were used for structure analysis in the 92 bread wheat genotypes.These data were also used for a genomewide approach to identify markers linked to seed quality traits.A second set of these 36 SSR QTLs-linked markers on wheat were tested for targeted association mapping of seed quality traits.For determining the population structure and K values, genotypic data was processed by the software program STRUCTURE 2.3. 38Applying a burn-in of 100 000 iterations, followed by 100 000 iterations, 35 K=1-15 and 5 runs per K was tested for targeted association mapping of seed quality traits.The fundamental basis of such clustering methods is allocating every individual genotype to K clusters in such a way that both Hardy-Weinberg equilibrium and linkage equilibrium are valid within clusters, whereas these kinds of equilibrium are missing between clusters.To obtain an association between markers and traits, Q-matrix, quality traits matrix, and scored SSR markers matrix were tested using the mixed linear model (MLM) method of Pritchard et al, 38 where this method is accomplished in the software package TASSEL 2.1 (http://www.maizegenetics.net/).To obtain the permutationbased test of marker significance and the experiment-wise P values for marker significance, the number of permutations was set at 1000 in this software.Only markers with an allele frequency of 5% or higher were included in the association mapping analysis.

Analysis of Variance and Structure Analysis
Analysis of variance showed a significant difference in all quality traits (Table 2).These results indicates that genetic variation exists among genotypes.The Population structure analysis was conducted using genotypic data of 102 SSR markers.To determine the number of subpopulations based on the suggestion of Pritchard and Wen, 39 we set K from 1 to 15.The population structure matrix (Q) was defined by the running structure of K = 6 where the highest likelihood has been obtained.The standard deviation of this group was lower than other groups (Table 3).K is the number of subpopulations consisted of loci in Hardy-Weinberg and linkage equilibrium.The accessions were subdivided into 6 subpopulations (Figure 1).In Figure 1, the majority of spring growth type genotypes were categorized into 3 subgroups (red, yellow, and aqua).Some of the winter growing genotypes were allocated to 2 subgroups (green and fuchsia) and facultative growing genotypes were classified in blue subgroups.Q-matrix outputs of 6 subpopulations were run (K = 6) for the structure based association analysis.

Marker-Trait Associations
Association analysis was conducted based on an MLM method. 38Association analysis was used for determining the SSR markers associated with the quality-related traits in the structured bread wheat population based on the population structure (Q-matrix).The association with SSR markers for the studied traits is described in table 4. The results of this study showed that among 36 QTLs derived primers, 29 SSR primers were polymorphic.These primers amplified 58 polymorphic allele markers, ranging from 1 (Xgwm639) to 4 (Xcfd13) with a mean of 2 alleles per locus.Of the total 34 allele markers linked to 4 quality wheat traits, 8, 9, 6, and 11 of them were related to grain protein content, gluten, falling number, and SDS sedimentation volume, respectively.Twenty-two of the 258 allele markers from the genome-wide SSR markers were found to exhibit a significant (P < 0.01) association with the above-mentioned 4 quality traits along with twelve of the 72 allele markers amplified by QTL-derived SSR primers.Allelic data onto all SSR markers with the significant association is presented in Table 4.

Discussion
Association mapping can identify QTLs by examining the marker-trait associations which can be attributed to the strength of LD between markers and functional polymorphisms across a set of diverse germplasm. 40Seed quality traits of bread wheat, specifically protein content, gluten, falling number, and SDS sedimentation volume are among the main objectives of a bread wheat breeding program and are effective in grain quality of bread wheat.These traits strongly influence the end use wheat and it's nutritional and market value.Marker-assisted selection (MAS) will  enhance the efficiency of the breeding process.Moreover, the accomplishment of MAS allows the selection of individuals carrying the suitable alleles at the target loci.The markertrait associations revealed for all 4 traits QTL distributed throughout the genome.
In previous studies, 18 markers that linked with protein content were used.In this study, we identified only 5 markers including Xcfd13 marker on chromosomes 6B and 6D, Xbarc54 on chromosomes 3A and 6D, Xbarc86 on chromosome 3A, Xbarc320 on chromosome 5D, and finally Xgwm577 on chromosome 7B which were shown to be associated (Table 4).This result confirmed the QTL locations that were identified in some previous studies. 9,18,41adesse et al reported that 2 DArT markers on 5B were highly associated with protein content and alveograph strength. 19n microsatellite consensus map, some of these SSR markers have been mapped as Xgwm577 in the distance 137 cM on the short arm of 7B, Xcfd13 in the distance 17 cM on the short arm 6B and 21 cM on the long arm of 6D, and Xbarc54 on the distance 47 cM on the short arm 6D. 36dditionally, Xcfa2141 on chromosome 5A, 5D, Xgwm121 and Xgwm515 on chromosome 5D, 7B, 2A, and 2D, were linked to grain protein content.These markers were used for population structure.It is possible which indicate a systematic type I error and false positive. 43Jochen et al, using the association mapping in winter wheat cultivars, studied QTLs linked to the protein quality-related traits and detected 4 QTLs associated with this trait located on the chromosomes 3A, 1B, 5D, and 2D. 28Furthermore, an experiment was carried out in different environmental conditions and identified QTLs placed on the chromosome 7A. 44Using recombinant inbred lines, Blanco et al identified 3 QTLs on chromosomes 6AS, 2AS, and 7BL that controlled this trait. 45Joppa et al reported that 66% of the variation of QTLs that control protein content was related to chromosome 6B. 46These loci located on the short arm of chromosome 6B near the centromere.As well as Joppa et al, Chee et al reported that QTLs for the protein quality were on the short arm of chromosome 6B. 46,47It can be concluded that chromosome 6B has an important role in the genetic control of this trait.Glutenin and gliadins are important quality traits in wheat.Nine QTLs were identified for gluten on chromosomes 5D, 2B, 5A, 7B, 2D, 5B, 1A, and 6B.In a previous study, Xcfd18, Xbarc200, and Xcfa2153 markers were known as associated QTLs with this trait located on chromosomes 5D, 2B and 1A, respectively. 6The results of the present study accorded to the results of Zhang et al study. 6Somers et al mapped Xbarc200 marker at a distance of 37 cM on the short arm of chromosome 2B. 36In another study, it was found that the amounts of gluten are controlled by 2 QTLs that located on chromosomes 5B and 7A. 48The genetic control of glutenin and gliadin is relatively well known in wheat. 28,49It can be concluded that association mapping can complement previous QTL information and provide opportunities for further wheat improvement programs.
Late maturity a-amylase is a genetic defect in wheat which results in the production of a-amylase, shown as substandard falling numbers, in the absence of pre-harvest rain and under cool temperatures during ripening. 12Wheat seeds with high a-amylase activity have little economic value because their handling and storage are difficult. 50,51Previous investigations had identified 5 SSR markers with a significant association with falling number.In the present study, these markers were tested for targeted association mapping.The result of the present study indicated that Xgwm80 and Xbarc113 markers as QTLs on chromosomes 1B and 4D were significantly associated with this trait (Table 4).This result was also confirmed by another researcher finding. 10In the Somers wheat consensus map, Xbarc80 marker was located on long arm of chromosome 1B at a distance of 106 cM. 36The present study confirmed a significant correlation between falling number and the markers located on chromosome 6B.Through using association mapping to identify the exact location of a-amylase genes in hexaploid wheat, a significant correlation between these traits and markers located on chromosome 7B was observed. 12urthermore, 4 SSR markers used for population structure assay displayed to be associated with falling number and located on chromosomes 5A, 5B, 5D, 7D, and 2D.These results showed that possibly this association may be false positives.Although, false positives can also arise from situations where the statistical test is valid and the association exists, but there is an association with population structure instead of the trait of interest. 52This matter is recommended to be examined in future studies.
Five SSR markers that in the previous studies were shown to be linked to SDS sedimentation were selected for association analysis.Among them, Xwmc453 9 and Xgwm371 42 markers were found to exhibit a significant association with SDS sedimentation.These markers were located on the chromosomes 2A, 2B, 2D, 5B, and 5D.Xwmc453 marker was mapped on the short arm of chromosome 2D at a distance of 43 cM. 36Furthermore, Xbarc86 marker was associated with SDS sedimentation and protein content.This marker has a pleiotropic effect.Nine SSR markers used for other quality traits or population structure analysis were associated significantly with SDS sedimentation.Using QTL mapping, Blanco et al also found QTLs that located on chromosome 3AS, 3BL, 5AL, 6AL, and 7BS linked with this trait. 40Huang et al reported 3 QTLs on 1B, 2D, and 5D chromosomes. 9In another study using association mapping analysis in bread wheat, QTLs on chromosome 2D, 3A, 5D, and 1A were identified for SDS sedimentation. 28These results are in agreement with the results of the present study.Results of this study and previous investigation confirmed that 2D and 5D chromosomes had a more significant role in controlling this trait.

Conclusions
In the present study, we conclude that association mapping in bread wheat is not only suitable but can also reveal additional QTLs not found in bi-parental populations, because the genetic variation within an association mapping panel is usually much greater than that in a conventional linkage mapping populations.

Figure 1 .
Figure 1.Diagram Derived From the Program Structure 2.3 Showing the Distribution of Bread Wheat Genotypes Into 6 Subpopulations (K=6).

Table 1 .
Information on Wheat Genotypes Used in This Study

Table 2 .
Analysis of Variance for Seed Quality Traits

Table 3 .
Average Logarithm of the Probability of Data Likelihoods (Ln P(D)) of 92 Bread Wheat Genotypes