ALLELE FREQUENCY OF GLUTENIN SUBUNITS AND GLU-1 QUALITY SCORES IN SOME TURKISH BREAD WHEAT LANDRACES

Kultur bitkileri icin sekiz orijin merkezi vardir ve Turkiye, bu merkezlerden ikisinin, Yakin Dogu ve Akdeniz'in, kesistigi yerdedir. Turkiye bu nedenle bugday, arpa, cavdar ve yulaf gibi ana tahil bitkilerinin cesitlendigi ve dagildigi gen merkezi olarak bilinir. Bu calismanin amaci, Turk ekmeklik bugdaylarindan 116 yerel bugday irkinin gluten allel sikligi ve Glu-1 kalite skorunu belirlemektir. SDS-PAGE ve PCR, gluten allellerini tanimlamak icin kullanilmistir. Istatistiksel analizler icin POPGENE 1.31 yazilimi kullanilmistir. Incelenen Turk yerel bugday irklari toplam 19 farkli alt birim ( Glu-A1 'de 3 alt birim, Glu-B1 'de 11 ve Glu-D1 'de 5 alt birim) ve 50 farkli kombinasyon icermektedir. En yuksek ve en dusuk allel kombinasyonlari sirasiyla Dogu Anadolu ve Ege bolgelerinde belirlenmistir. Glu-A1c (% 65,11), Glu-B1b (% 53,60) ve Glu-D1a (% 58,30) en sik gorulen alleller olarak tespit edilmistir. Glu-1 kalite skoru, Turkiye genelinde incelenen genotipler icin 6,07 olarak bulunmustur. Bolgeler arasinda ortalama en yuksek (7,18) ve en dusuk (4,80) Glu-1 skorlari sirasiyla Marmara ve Guneydogu Anadolu bolgelerinde tespit edilmistir. Calisilan 116 aksesyondan 4 tanesi (TR32846-6, TR36948-1, TR45105 ve TR63536) kalite skoru 10 olup en yuksek Glu -1 kalite skoruna sahipken 6 tanesinin ise (TR45398-4, TR48025-3, TR33264-6, TR393-5, TR52021-3 ve TR45094) kalite skoru 9 olarak hesaplanmistir. Daha fazla yerel cesitlerin calismalara eklenmesi yeni Glu-1 allelerinin kesfedilmesine katkida bulunabilir.


Introduction
Vavilov (1951) identified eight centers with gene center status that have long been used for agriculture and Turkey is located in the interception of two of these centers (Near East and the Mediterranean), making the country one of the richest in terms of plant biodiversity. Turkey is also known to be one of the gene centers for diversification and dispersion of main cereal crops such as wheat, barley, rye, and oat. This special future, in addition to being the first place where wheat was cultivated, makes the country for several wild species and landraces of wheat to be available. For instance, the first cultivated forms of diploid and tetraploid wheat originated from Karacadağ (Diyarbakır) in the Southeastern region of Turkey.
These first diploid (AA) and tetraploid genomes (AABB) and their phylogenetic analysis indicate that they had originated in Southeastern Turkey (Dubcovsky & Dvorak 2007). Cultivation shifted from here to near east about 9.000 years ago and the hexaploid wheat appeared for the first time (Feldman 2001). Cereals occupy 55% of total agricultural and 57% of cultivated areas (21.4 million hectares) in Turkey (Grain sector report 2013). Among the cereals in Turkey, wheat is the pioneering crop with 67% share and the country supplies 3% of world wheat production every year. However, because of quality-related issues, it is still a wheat importing country (TMO 2017). Among the regions, the leading wheat producing region is Central Anatolia (32.4%), followed by Marmara (17.1%), Southeastern Anatolia (13.3%), the Mediterranean (12.8%), Black Sea (10.1%), Aegean (7.4%) and Eastern Anatolia (6.9%) regions.
The quantity and composition of high molecular weight gluten subunits (HMW-GS) are important factors in determining wheat baking properties. Localization of HMW subunit genes on long arms of homologous group 1 was reported by Orth & Bushuk (1974) and Bietzh et al. (1975). Each locus contains two linked genes called x and y type which are distinguished by their characteristics and molecular weights (Payne et al. 1981). However, as some of those genes are silent, the common wheat possesses 3 to 5 HMW subunits encoded at the Glu-1 loci on the long arms of group 1 chromosomes (1A, 1B, and 1D). The contribution of D genome, followed by B genome considered to have a significant influence on good baking quality (Uthayakumaran et al. 2002). Especially, two subunits are expressed always by Glu-D1, one or two by Glu-B1, one or none (null allele) is expressed by Glu-A1 loci. If one subunit is expressed by Glu-A1 or Glu-B1, this is always considered as x-type. Rheological properties of the gluten complex are related to the presence or absence of specific subunits of these proteins. The presence of certain HMW subunits is positively correlated with good bread-making quality to determine gluten elasticity (Nakamura 2000, Shewry et al. 2003. The relationships between HMW subunits and dough elasticity were determined 40 years ago (Payne et al. 1979). Significant differences were found among protein components of wheat grain depending on cultivars, environments and their interactions (Horvat et al. 2015, Tok et al. 2011. Allelic variations in each Glu-1 loci were reported in bread wheat genotypes (Lawrence & Shepherd 1980) and an enumeration system was developed to define different allelic subunits. The definition of HMW subunits coded by Glu-D1 and Glu-A1 was described by  and Lafiandra et al. (1997).
Subunits of gluten proteins can be identified by Sodium Dodecyl Sulphate Polyacrylamide Gel Electrophoresis (SDS-PAGE) analysis. Considering the fact that chains of gluten polymers are stabilized by disulfide bonds, using reducing agents such as DTT and mercaptoethanol in SDS-PAGE makes the identification of gluten subunits easier.
Although Turkey comes third in the production of wheat in the world, quality problems still persist. This study was performed to determine the composition and occurrence frequency of HMW glutenin subunits and the potential of the end-use quality in Turkish wheat landraces to be used in future breeding programs. Analysis of broad collections of landraces enabled the definition of rare alleles on Glu-1 loci. Table 1. The wheat specimens studied with their respective distribution regions. All specimens were given using their gen bank accession numbers.

Plant materials
116 bread wheat landraces of Turkey obtained from Ankara Field Crops Central Research Institute and Izmir Aegean Agricultural Research Institute were included in the study ( Table 1). The specimens were so selected to provide representation of all geographical regions (Aegean, Central Anatolia, Marmara, the Mediterranean, Black Sea, Eastern Anatolia and Southeastern Anatolia) in the country. These were supplied from the "Field Crops Central Research Institute, Ankara" and "Aegean Agricultural Research Institute, İzmir. 15 standard genotypes (Chinese Spring, Drago, Lobeiro, Svevo, Lira, Durambo, Ak 702, Bezostaya 1, KateA1, Bayraktar, Mizrak, Yakar, Atay 85, Gerek 79 and Tosunbey) were used as references to define gluten alleles The wheat specimens studied with their respective distribution regions is provided in Table 1.  (Table 2). PCR reaction was realized in 15 µl reaction mixture containing 50 nmol of each primer, 0.3 nmol dNTP, 1-2 unit Taq polymerase, 1.5-2.5 mM MgCl2, 10 µg/µl BSA and 30 ng template DNA. Following the initial denaturation at 94°C for 5 minutes, PCR reaction was carried out in 35 cycles of 94°C for 1 minute, 54-65°C for 1-2 minutes and 72°C for 2 minutes. The last extension step was realized at 72°C for 10 minutes. PCR products were run in 1-1.5% agarose-containing ethidium bromide and imaged with a gel documentation system (BioRad MP5 gel documentation system).

Gluten extraction and SDS-PAGE analyses
Gluten proteins were extracted in accordance with the methods described by Gao et al. (2012) and Temizgul et al. (2018). The methods of Li et al. (2012) with slight modification were used for the electrophoresis of gluten proteins. 30-60 mA current was applied per gel during electrophoresis. Gels were stained (dissolved in 187.5 ml of methanol, 225 mg of CBB-R 250; 750 ml 10% TCA and 62.5 ml glacial acetic acid) for 12-24 hours and kept in washing solution (333 ml methanol, 100 ml 10% TCA and 567 ml distilled water) for 5 hours and imaged under the gel documentation system.

Definition of gluten alleles
Gluten alleles of the wheat landraces were defined based on individual HMW subunit distributions in SDS-PAGE in accordance with the methods specified by  and McIntosh et al. (1994). The verification of alleles defined by allele-specific primers was also performed.

Calculation of Glu-1 quality scores
Based on the SDS-PAGE profile, relationships between individual HMW subunits and quality was determined by using the method described by Payne (1987a) and scored by using subunit scores of Payne (1987a) and Lukow et al. (1989). POPGENE version 1.31 software was used to draw dendrogram based on Nei's original measurement showing the relationships among the genotypes based on individual subunits (Nei 1972).

Statistical analysis
Data analysis was carried out based on the frequencies of HMW gluten gene alleles. Statistical analyses were performed separately in individual variety and populations for individual alleles, individual sub-units, loci (Glu-A1, Glu-B1, and Glu-D1) and geographical origin (7 regions). POPGENE version 1.31 software was used for statistical analysis. Polymorphism percentages were calculated using Equation 1; Eq. 1: Polymorphism% = (number of polymorphic allele/number of total alleles) × 100.
Allelic combination frequencies were calculated using Equation 3; Eq. 3: Allelic combination frequency = (observed total number of each allelic combinations/number of total wheat samples) × 100.

Results
In A genome of studied genotypes, the frequencies of Glu-A1c, Glu-A1b, and Glu-A1a were calculated to be 65.1%, 26.4%, and 8.5%, respectively (Table 3). In B genome, the frequencies of Glu-B1b, Glu-B1e, Glu-B1d, and Glu-B1c were determined to be 44.6%, 16.6%, 10.0%, and 10.7%, respectively. In D genome, the frequencies of Glu-D1a, Glu-D1d, and Glu-D1b were determined to be 59.7%, 27.9%, and 7.7%, respectively. Since quality scores were not determined for these subunits, the contributions of the alleles Glu-B1h, Glu-B1z, and Glu-B1aj to quality score could not be calculated for 7 of the 116 genotypes studied. After determination of rheological characteristics of individual subunits, the contribution of these subunits to the quality can also be determined. The average quality score in the genotypes was calculated to be 6.07. The Marmara and the Southeastern Anatolia regions had 7.18 and 4.8 quality scores as the highest and the lowest, respectively.

Allelic combinations and frequencies
Allelic combinations and their frequencies are given in Table 4. The results indicated the presence of 39 different allelic combinations. Among these combinations, Glu-B1b/Glu-D1a was the most frequent combination with a 22.48% frequency value, followed by Glu-A1c/Glu-B1b/Glu-D1a, Glu-A1b/Glu-B1c/Glu-D1d, and Glu-B1d/Glu-D1a allelic combinations with 11.7%, 9.35%, and 5.45% frequencies, respectively. The rest of the alleles were observed with a frequency less than 5%.

Region-based allelic combinations and Glu-1 scores
The frequencies of HMW gluten subunits and their Glu-1 scores are given in Table 5. In some of the genotypes, the effect of some subunits on quality score could not be calculated since the effect of these subunits on quality score was not determined. The quality score was 10 in four accessions (TR32846-6, TR36948-1, TR45105 and TR63536), and 9 in 6 accessions (TR45398-4, TR48025-3, TR33264-6, 393-5, TR52021-3, TR45094). The highest (17) and the lowest (5) allelic combinations were observed in Eastern Anatolia and Aegean regions, respectively ( Table 5).

Development of dendrogram based on subunits of Glu-1 loci
When the dendrograms (Nei 1972) drawn based on individual subunits were investigated, the Mediterranean region clustered with Black sea (genetic distance 0.32) and Aegean region clustered with Central Anatolia (genetic distance 0.52) (Fig. 1). Marmara (genetic distance 1.35), Eastern Anatolia (genetic distance 0.37), and Southeastern Anatolia (genetic distance 0.52) regions were separately clustered.

Discussion
The determination of HMW-GS composition in wheat cultivar collections from different countries has been studied (Nucia et al. 2019). The contribution of individual subunits on dough quality is quite different (Payne et al. 1987, Lukow et al. 1989). While quality score of Glu-D1d is the highest (quality score 4), the contribution of Glu-A1c, Glu-B1a, Glu-B1d, Glu-B1e, and Glu-D1c is considered to be the lowest (quality score 1). In the present study, the highest Glu-D1d frequency (50%) was observed in Marmara and the lowest was observed in Aegean regions (6.67%). The frequency of this subunit was 27.6% throughout Turkey ( Table 2). The highest Glu-1 score with 7.18 was observed in Marmara followed by Black Sea with 6.65, Eastern Anatolia with 6.57, Mediterranean with 5.70, Central Anatolia with 5.63, Aegean with 5.10, and Southeastern Anatolia regions with 4.8 (Table 5).
Observing the highest quality score in Marmara and the lowest in the Southeastern region is an expected outcome. Southeastern Turkey is on the interception of gene centers for wheat and its wheat is intensively studied and high-quality score wheat genotypes are already selected for breeding. Marmara region is the most industrialized region and agricultural activity is quite poor in this region, therefore, the landraces might not be used to select for high-quality score cultivar breeding. Throughout Turkey, 19 different subunits (3 Glu-A1, 11 Glu-B1 and 5 Glu-D1) were observed in 39 different combinations (Tables 3 and 4). Among the allelic combinations Glu-B1b/Glu-D1a was the highest with 22.48% frequency, followed by Glu-A1c/Glu-B1b/Glu-D1a with 11.70%, Glu-A1b/Glu-B1c/Glu-D1d with 9.35% and Glu-B1d/Glu-D1a with 5.45%. Other alleles were observed with less than 5% frequency. The highest allelic variation with 16 subunits and 17 different combinations was observed in Southeastern Anatolia. Glu-B1b/Glu-D1d allelic combinations with 16.67% frequency were the most frequent allelic combinations in this region (Table 5). Based on Nei's original measurement (Nei 1972), while the highest similarity among regions for individual subunits of Glu-1 loci was observed between Black Sea and the Mediterranean regions (0.9935), the lowest similarity was observed in Marmara and the Aegean regions (0.953). When the dendrogram was drawn based on Nei's (1972) original measurements, the Mediterranean clustered with the Black Sea and the Aegean with the Central Anatolia accessions. Other regions were separately clustered (Fig. 1). Nakamura et al. (1999) studied variations in the HMW subunits of Glu-1 loci of Kapon wheats and identified 14 different alleles, 3 of which were on Glu-A1, 6 on Glu-B1 and 5 on Glu-D1 loci. In the present study, 19 different subunits were identified and 3 of them were on Glu-A1, 11 of them on Glu-B1 and 5 of them on Glu-D1 loci. The frequency of null alleles located on chromosome 1A was reported to be high (74%) on Japanese wheat especially on Norin variety. The frequency of this allele was also found to be high in landraces of Turkish wheat (65.11 %). While Glu-B1a, Glu-B1f, Glu-B1h, Glu-B1j, and Glu-B1k subunits were not observed in Japanese bread wheat (Nakamura et al. 1999), only Glu-B1j and Glu-B1k subunits were not observed in Turkish wheat genotypes used in this study. While the frequency of 2+12 subunit coded by Glu-D1a allele was 55% in Japanese wheat, it was only 1.5% in 5+10 subunit coded by Glu-D1d. The average frequencies of these alleles were found to be 58.3% and 27.6% in Turkish genotypes, respectively. The frequencies of Glu-D1a and Glu-D1d subunits were 86.7% and 50% in Aegean and Marmara regions, respectively. The frequency of Glu-D1d subunit was found to be high in European wheat varieties (Payne et al. 1984). The frequency of this subunit was also found to be high in Turkish wheat landraces, especially in Marmara region (average 27.6%, Marmara region 50.0%). Of the 11 alleles, 3 were coded by Glu-A1, 6 were coded by Glu-B1 and 2 were coded by Glu-D1 loci. Null allele frequency was found to be quite low (1.1%) and Glu-D1d was the highest in frequency. However, Glu-A1c subunit was the highest with the frequency of 65.1% among the alleles in Turkish landraces. Gianibelli et al. (2002) also calculated the quality score taking the Glu-A1 into account and separated the Argentinian wheat into 18 groups. Glu-A1a/Glu-B1bq/Glu-D1d was observed in highest frequency (22%). In Turkish landraces, Glu-B1b/Glu-D1a was observed in highest frequency. Payne (1987b) found the quality score of world wheat collection as 9.5. The quality score of Turkish wheat landraces was found to be significantly lower (6.07). Notwithstanding, quality score is not determined by only HMW-GS, the contribution of LMW-GS and Gliadins should also be taken into account (MacRitchie et al. 1990).  published the catalogue of Glu-1 alleles. They determined 3, 11, and 7 alleles in Glu-A1, Glu-B1, and Glu-D1, respectively. Additional alleles were also determined but most of those alleles were found to be in Glu-B1 loci (Pogna et al. 1990). In the present study, a possible new allele (Dy12*) was determined in Glu-D1 locus. This new subunit was differentiated considering its faster movement in SDS-PAGE. Branlard et al. (1989)
Primitive cultivars and locally grown landraces were considered to be the sources of variation for grain protein quality, disease resistance, and resistance to abiotic stress conditions (Porceddu et al. 1988, Kaplan et al. 2014). HMW subunit variations in Turkish bread wheat landraces were found to be higher compared to Australian, Italian, American, Canadian, French and Spanish wheat samples (Autran & Feillet 1985, Margiotta et al. 1987, Carrillo et al. 1990). This outcome is somewhat expected since intensive breeding works have decreased the variation in western wheat genotypes (Porceddu et al. 1988). Morgunov et al. (1993) and Sultana et al. (2007) stated that the increase in HMW score was related to a decrease in diversity in gluten alleles. A similar phenomenon was also observed in wheats grown in Dobrudzha Agricultural Institute ). That is why it is crucial to use Glu-B1f, Glu-B1h and Glu-B1i subunits to increase quality score. This situation may decrease genetic diversity and increase end-product quality (Liu et al. 2007). In the present study, the observation of these alleles was also quite low in 116 Turkish wheat landraces (approximately 7%). The frequency of Glu-B1h (14+15) was observed to be 55% in the Mediterranean region. Maintenance of Glu-A1b (2*) and Glu-D1d (5+10) alleles is important. These alleles contribute to quality supported with Glu-B1i (17+18), Glu-B1f (13+16) and Glu-B1h (14+15) ). Glu-D1a (81.5%) were the most frequently observed alleles. Although Glu-D1a (46.9%) was the most frequently observed allele in Central Asia, it was lower in all the other regions except Caucasian region. A total of 83 allelic combinations were determined on Glu-1 loci in their studies. Among the allelic combinations, Glu-A1c/Glu-B1b/Glu-D1a was the most frequently observed genotype. The frequency of this allelic combination was found to be 11.70% for Turkish cultivars. The most frequently observed allelic combination was Glu-B1b/Glu-D1a (22.48%). Although western Asian, Afghanistan, and Eastern Asian wheats were exhibiting similar characteristics, Caucasian and Central Asian wheats differed from these three regions. As it is reported by Terasawa et al. (2009) and Lagudah et al. (1987), the most common genotypes were determined to be Glu-A1c (null), Glu-B1b (7+8), and Glu-D1a (2+12) among Western and Eastern Asian genotypes. These alleles were also found to be the highest in frequency in the present study. Those results, the results of Terasawa et al. (2009) and Lagudah et al. (1987) indicated that Glu-A1c, Glu-B1b, and Glu-D1a genotypes were dominant in regions extending from Mesopotamia, Afghanistan, and Far East to Central Asia. In Southern Asia, Glu-A1c, Glu-B1i, and Glu-D1a were the most frequently observed genotypes in a study reported by Terasawa et al. (2010). This genotype was considered to be a modified version of the typical Asian genotype in the sense that only Glu-B1i allele was replacing Glu-B1b allele. Glu-B1i allele was observed rarely in other regions of Southern Asia. This allele was also found to be very rare in the present study (0.78%). Similarly, this allele was also rare among European endemic wheat genotypes (Gregova et al. 1999, Juha'sz et al. 2003. This is why researchers considered that Glu-B1i had appeared in Southern Asia (Terasawa et al. 2010). Glu-B1i allele was providing more firmness to dough compared to Glu-B1b allele , Mondal et al. 2008. Glu-D1d allele is common in Caucasian and Central Asian accessions. The high frequency of this allele in Caucasian and Central Asia is remarkable. In the same region Glu-A1b and Glu-B1a alleles have also been observed in high frequency (Terasawa et al. 2010). Glu-D1d allele is known to contribute to bread-making quality. This allele is introduced to modern wheat genotypes to increase breadmaking quality (Wrigley et al. 2015, online). The frequency of this allele was calculated to be 27.91% in the present study and the allele was observed in 50% of the wheat genotypes. This allele was also in high frequency in European wheat (Gregova et al. 1999, Juha'sz et al. 2003. However, researchers suggested the Caucasian region as the center of origin for this allele and its dispersion to other regions (Dvorak et al. 1998).
In conclusion, the quality score was found to be low in the studied genotypes (6.07). High quality score genotypes might have already been selected for breeding purposes by agricultural institutes. Relatively high quality scores in Black Sea region, where wheat breeding studies is relatively low, and in highly industrialized Marmara region support that claims. Considering the individual alleles in Glu-1 loci, the highest similarity was observed between Black Sea and the Mediterranean regions (0.9953) and the lowest similarity was between Marmara and Aegean regions (0.9472). When individual alleles and subunits are used for cluster analysis in Glu-1 loci, the Mediterranean region clustered with Black Sea and Aegean region clustered with Central Anatolia. Other regions are individually clustered and separated from these regions.