lnfluence of protein characteristics and the proportion of gluten on end-use quality in Korean wheat cultivars

Seong-Woo Cho, Chon-Sik Kang, Hyeon Seok Ko, Byung-Kee Baik, Kwang-Min Cho, Chul Soo Park

1 Department of Crop Science & Biotechnology, Chonbuk National University, Jeonju 54896, Republic of Korea

2 National Institute of Crop Science, Rural Development Administration, Wanju 55365, Republic of Korea

3 Farm & Agribusiness Management Division, Rural Development Administration, Jeonju 54875, Republic of Korea

4 USDA-ARS CSWQRU, Soft Wheat Quality Laboratory, Wooster, OH 44691, USA

5 Research Center of Bioactive Materials, Chonbuk National University, Jeonju 54896, Republic of Korea

Abstract The effects of protein characteristics and the proportion of gluten on end-use quality in 13 Korean wheat cultivars for three years were verified in this study. Year, cultivar, and the interaction between the year and the cultivar influenced protein characteristics, the proportion of gluten except for γ- and ω-gliadin using RP-HPLC (reversed-phase high-performance liquid chromatography), and end-use quality. Protein characteristics and the proportion of gluten in Korean wheat cultivars were between those of Australian standard white (ASW) and hard wheat (AH). Korean wheat cultivars exhibited a higher average α+β gliadin proportion than imported wheat, a γ-gliadin proportion similar to that of dark northern spring wheat,and the same ω-gliadin proportion as AH. They showed a bread-loaf volume intermediate between those of ASW and AH and a texture of cooked noodles similar to that of soft white wheat, but less springiness than imported wheat. The cookie diameter of Korean wheat cultivars was similar to that of hard red winter wheat. There was a correlation between bread-loaf volume and protein characteristics, except for the protein content in Korean wheat cultivars. Springiness and cohesiveness of cooked noodles were not correlated with protein characteristics, while hardness was correlated with the protein content and water absorption of a mixograph. Cookie diameter was negatively correlated with the sodium dodecyl sulfate (SDS)sedimentation volume and water absorption of a mixograph. The end-use quality was not correlated with any proportion of gluten composition. Principal component analysis (PCA) showed that the proportion of gluten was not related to the quality of the bread (both PCs, 81.3%), noodle (77.7%), and cookie (82.4%). PCA explained that Keumkang is suitable for superior bread, while Uri is good for cooked noodles and cookies.

Keywords: wheat quality, flour quality, storage proteins

1. lntroduction

Wheat gluten is the primary factor that determines the enduse quality because of protein content and composition influence dough rheology and baking properties (Carson and Edwards 2009). Glutenins, mainly related to gluten quality in wheat, are divided into two groups, high-molecularweight glutenin subunits (HMW-GSs) and low-molecularweight glutenin subunits (LMW-GSs) (Gianibelliet al.2001). HMW-GSs are the key factors in bread baking, as the major determinants of dough elasticity, and LMW-GSs play significant roles in determining dough resistance and extensibility (D’Ovidio and Masci 2004). LMW-GSs and HMW-GSs can be linkedviaboth inter- and intra-molecular disulfide bonds, forming very large polymeric proteins(Gianibelliet al. 2001; D’Ovidio and Masci 2004). The cohesiveness and extensibility of the gluten are attributed to the monomeric gliadins (Carson and Edwards 2009).These gluten properties, such as the amount of gluten,ratio of glutenin to gliadin, gluten composition, and ratio of HMW-GSs to LMW-GSs, have been closely related to enduse quality (Song and Zheng 2007).

Bread-loaf volume generally increases with increasing protein content and is significantly influenced by the content and composition of protein in wheat grains (Branlardet al. 2001). The bread-loaf volume, as well as gluten properties, depends principally on both the genotype and the environment (Grayboschet al.1996; Panozzo and Eagles 2000; Zhu and Khan 2001). Wheat flour of about 10% protein content is acceptable for the preparation of soft-bite white salted noodles (Huo 2001). The protein content and its related parameters, such as sedimentation volume and the mixing time of mixograph, are known to affect the texture of cooked noodles (Baiket al. 1994;Yunet al. 1996; Parket al. 2003). Cookie diameter is an excellent indicator of general soft-wheat baking quality(Finneyet al. 1987; Hoseneyet al. 1988). Cookie diameter was more affected by protein content than by composition,which fundamentally is genetically controlled, but protein content is largely influenced by environmental conditions(Hoseneyet al. 1988; Slade and Levine 1994; Souzaet al. 1994). The principal environmental factors, such as fertilization and weather, affect the quantitative variations,like the amount and fraction of gluten. Such quantitative characteristics are equally important for the end-use quality.Understanding these effects on gluten properties and enduse quality is important for wheat breeders, growers, and end-use processors.

Korean wheat cultivars have a short maturation period because of temperature and rice transplantation. The temperature rapidly rises after mid-May, so the wheat harvest has to be completed in mid-June, because of overlapping rainy seasons. The Korean wheat-breeding program has long focused on improving grain yield and early maturation. Korean wheat cultivars also had narrow genetic diversity and the bread produced has been inferior to that from foreign wheat cultivars, despite the similar protein content (Parket al. 2012). Thus, quality improvement is receiving more attention from Korean wheat breeders. The objectives of this study are to find out how gluten properties affect end-use quality, including that of pan bread, white slated noodles, and sugar-snap cookies, of 13 Korean wheat cultivars during three growing seasons, and to provide useful information for improving end-use quality in wheat-breeding programs.

2. Materials and methods

2.1. Plant materials

Thirteen Korean representative wheat cultivars, which originated in the 1990s, were sown in randomized complete blocks with three replications in the Upland Crop Experimental Farm of the National Institute of Crop Science,Rural Development Administration (Korea) in 2013/2014,2014/2015, and 2015/2016 on 50% clay/loam soil. The seeds were sown in late October, and each plot consisted of three 4-m rows spaced 25 cm apart. These plots were combine-harvested in mid-June for three years. Prior to sowing, fertilizer in the ratio of 5:7:5 kg 10 yr–1(N:P:K) was applied, while weeds, insects, and disease were stringently controlled. No supplemental irrigation was applied. The means of temperature and precipitation for the three seasons were 10.4°C and 586.0 mm. During the three seasons, the temperature and precipitation of 2015/2016 season were higher (10.8°C, 702.1 mm) than those of 2013/2014 (10.4°C, 481.8 mm) and 2014/2015 (10.0°C,574.1 mm). Grains from each plot were dried using forcedair dryers and bulked from replications to provide grains for milling. The five imported wheat flours used in this study,including Australian hard (AH), Australian standard white(ASW), soft white (SW), dark northern spring (DNS), and hard red winter (HRW), were obtained from Dong-A-One(Dangjin, Korea).

2.2. Analytical methods

Wheat was milled using a Bühler Experimental Mill, based on the American Association of Cereal Chemists International Approved Method 26-31.01 (AACCI 2010). Two kilograms of wheat under 15% moisture content were milled with a feed rate of 100 g min–1and with roll settings of 8 and 5 in break rolls and 4 and 2 in reduction rolls. Protein content(%) of wheat flour was measured according to AACCI Approved Methods 46-30.01 (AACCI 2010). An sodium dodecyl sulfate (SDS) sedimentation test was performed according to AACCI Approved Method 56-70.01 (AACCI 2010). Optimum water absorption and mixing time of wheat flour were measured using a 10-g mixograph (National Mfg.Co., USA) according to AACCI Approved Method 54-40.02(AACCI 2010). Dry gluten content was measured using a Glutomatic 2200 (Perten Instruments AB, Sweden) with a constant volume of Glutomatic wash solution (4.8 mL) based on AACCI Approved Method 38-12.02 (AACCI 2010).

2.3. RP-HPLC

Reversed-phase high-performance liquid chromatography(RP-HPLC) was performed to measure the proportion of glutenin components, including HMW-GSs and LMW-GSs,x- and y-type glutenin subunits, and gliadin components,such as ω, α+β, and γ-gliadin.

Gliadin was extracted from 100 mg of wheat flour, which was incubated in 400 μL of 70% ethanol at root temperature(RT) for 1 h under vortexing and then centrifuged at 13 000 r min–1for 15 min (Zhouet al. 2013). The supernatant was used to analyze gliadins. After extraction of gliadin, the pellet was used to extract glutenin according to the protocol of Yanet al. (2014). It was washed three times with 500 μL of 55%isopropanol (v/v) for 30 min at 65°C and then centrifuged at 1 300 r min–1for 10 min. After centrifuging, the pellet was incubated with 100 μL extraction buffer containing 50% isopropanol (v/v), 80 mmol L–1Tris-HCl (pH 8.0), and 1% dithiothreitol for 30 min at 65°C, and then centrifuged at 13 000 r min–1for 10 min. For alkylation, the pellet was incubated with 100 μL extraction buffer for 30 min at 65°C, wherein 1% dithiothreitol was replaced with 1.4%4-vinylpyridine (v/v). After centrifuging, the supernatant was collected and added to cold acetone at –20°C to a final concentration of 80%, and then stored overnight at –20°C.Precipitated glutenins were obtained by centrifuging at 13 000 r min–1for 10 min, followed by removal of acetone in a fume hood. The pellet was eluted in 200 μL of 20%acetonitrile with 0.1% trifluoroacetic acid. Extracted gliadins and glutenins were cleaned with a 0.2-μm polyvinylidene difluoride membrane filter (Pall Life Sciences, USA). The injection volume for RP-HPLC analysis was 20 μL. RPHPLC analysis was carried out on an Agilent 1100 instrument with a ZORBAX 300StableBond C18 as a reverse-phase column (300? 5 μm, 4.6 mm×250 mm, Agilent Technologies,USA) at the column temperature of 65°C, and the flow rates of 0.8 mL min–1for glutenin or 1.0 mL min–1for gliadin.For separation of glutenin, the initial solvent contained 25% acetonitrile and 75% water, each comprising 0.1%trifluoroacetic acid. The acetonitrile ratio gradually increased to 33% within 13 min, 43% within 54 min, and 50% within 62 min, and then returned to the initial condition within 64 min.For separation of gliadin, the initial solvent contained 25%acetonitrile and 75% water, each with 0.1% trifluoroacetic acid. The acetonitrile ratio gradually increased to 50% within 60 min, 25% within 62 min, and then returned to the initial condition within 70 min. The proteins were detected using UV analysis at a wavelength of 210 nm. Glutenin properties were measured according to the percentage of the area of total gluten or gliadin content.

2.4. End-use quality

Bread was baked according to the optimized straightdough bread-making method according to AACCI Approved Method 10-10.03 (AACCI 2010). The ingredients of the baking formula consisted of 100 g (14% moisture basis)flour, 6 g sugar, 3 g shortening, 1.5 g salt, 5.0 g fresh yeast,50 mg ascorbic acid, and 0.25 g barley malt (about 50 DU g–1,20°C). The optimum water absorption and mixing time were judged by the feel and appearance of the dough during the mixing. The dough was fermented in a cabinet at 30°C and 85% relative humidity for 70 min with two punches and a proof period of 60 min, and then baked at 210°C for 18 min. After cooling for 2 h at room temperature, a slice 2.0 cm thick was cut from the center portion of the bread. Bread-loaf volume was measured immediately by rapeseed displacement and weighed after taking the bread out of the oven.

White salted noodles were prepared as in our previous study (Park and Baik 2002). Flour (100 g, 14% moisture basis) was mixed with a predetermined amount of sodium chloride (NaCl) solution in a pin mixer (National Mfg. Co.,USA) for 4 min, with a head speed of 86 r min–1. The concentration of NaCl solution for making noodles with 34% absorption was adjusted to have 2.0% NaCl in the noodle dough. The crumbly dough was passed through the rollers of a noodle machine (Ohtake Noodle Machine Manufacturing Co., Tokyo, Japan) running at 65 r min–1with a 3-mm gap to form a dough sheet, which was further folded and put through the sheeting rollers. The folding and sheeting were repeated twice. The dough sheet was rested for 1 h and then put through the sheeting rollers thrice at progressively decreasing gaps of 2.40, 1.85, and 1.30 mm.The dough sheet was cut through number 12 cutting rolls to produce noodle strands with a cross section of 3 mm×2 mm, each about 30 cm in length. Raw noodles (20 g)were cooked for 18 min in boiling distilled water (500 mL)and then rinsed with cold water. Two replicates of cooked noodles were evaluated by texture profile analysis (TPA)with a TA.XT2 texture analyzer (Stable Micro Systems,Godalming, UK) within 5 min after cooking. A set of five strands of cooked noodles was placed parallel on a flat metal plate and compressed crosswise twice to 70% of their original height with a 3.175-mm metal blade at a crosshead speed of 1.0 mm s–1. From TPA force-time curves, hardness,springiness, and cohesiveness were measured as described by Parket al. (2003).

Sugar-snap cookies were baked to measure their diameter following the AACCI Approved Method 10-52.01 (AACCI 2010). The ingredients for cookie baking were 40 g flour with 14.0% moisture, 24.0 g sugar, 12.0 g shortening, 1.2 g nonfat dry milk solids, 0.4 g sodium bicarbonate, 4 mL solution with 0.32 g sodium bicarbonate, 2 mL solution with 0.2 g ammonium chloride, 0.18 g sodium chloride, and 2.7 mL of deionized water. The sifted sugar, nonfat dry milk solids,and sodium bicarbonate were combined with the shortening and then creamed by Kitchen Aid Professional KPM5 Mixer(Kitchen Aid, MI, USA) equipped with a flat beater mixing arm (type K45AB) for 4 min. Of the creamed mass, 37.6 g were blended with a water solution of sodium bicarbonate and ammonium chloride. The mixture was remixed again with flour by a National cookie dough micromixer (National Mfg. Co., Lincoln, NE) at 172 r min–1for 3 min. The cookie dough, 7 mm thick, was divided into a scrap using a cookie dough cutter (60 mm inside diameter). The scrap was baked at (205±2)°C for 10 min. Cookie diameter was measured after the cookies were cooled at room temperature for 20 min.Four cookies were baked for each flour.

2.5. Statistical analysis

Statistical analysis of the data was performed by SAS software (SAS Institute, NC, USA) using ANOVA (analysis of variance) with LSD (Fisher’s least significant difference test)and a pairwiset-test. Analysis of variance was conducted using the general linear model, and the cultivar×year component was used as the error term. Sources of variation in the model were considered to be fixed effects.Correlation of end-use quality with the measured traits was analyzed by Pearson’s correlation coefficients, with aP-value of less than 0.05 being considered significant.PCA (principal component analysis) was carried out using the Proc Princomp procedure in SAS to load the measured traits under visualization. All measurements were performed at least in triplicate and all were averaged.

3. Results

3.1. Analysis of variance

Year, cultivar, and their interaction were significant for the protein characteristics, the proportion of gluten, and the end-use quality of the 13 Korean wheat cultivars (Table 1).Cultivar accounted for the largest proportion of the variation in protein characteristics, including protein content (57.4%),SDS-sedimentation volume (75.5%), dry gluten (50.0%), and water absorption and mixing time of the mixograph (82.0 and 97.3%, respectively). The variation in the proportions of HMW-GSs and LMW-GSs, the ratio of HMW-GSs to LMW-GSs, and the proportion of gliadin were significantly influenced by the cultivar rather than the year. The variation of the proportion of x- and y-type and the ratio of x-type to y-type in HMW-GSs were mainly affected by the year.The proportion of γ-gliadin was largely influenced by the cultivars, while the proportion of ω-gliadin was influenced by the cultivar and the interaction between the year and the cultivar. The variation of the bread-loaf volume was mainly affected by the cultivar, while the hardness, springiness, and cohesiveness of cooked noodles as well as cookie diameter were influenced by the year.

3.2. Protein characteristics

The protein characteristics of the 13 Korean wheat cultivars and the five imported wheat flours are summarized in Table 2. The protein content of Korean wheat cultivars was 9.6–14.5%, which was the range of protein content in the imported wheat flours (9.6–14.4%). SDS sedimentation volume was 19.0–73.8 mL in the Korean wheat cultivars,but the imported wheat flour exhibited SDS sedimentation volumes of 23.0–69.0 mL. Dry gluten was 5.7–14.0% in the Korean wheat cultivars and 7.7–13.9% in the imported wheat flours. Among the Korean wheat, Joongmo 2008 showed superior protein content, SDS sedimentation volume, and dry gluten content, while these characteristics were inferior in Goso, Joa, and Uri wheat. These values of Joongmo 2008 were similar to those of DNS, while Goso, Joa, and Uri showed properties similar to those of soft white flours.Water absorption and mixing time of the mixograph were 55.5–66.1% and 1.4–4.1 min, respectively, for the Korean wheat cultivars, and were 58.0–66.0% and 3.9–5.6 min,respectively, for the imported wheat flours.

Table 1 Analysis of variance for protein properties, proportion of gluten composition, and end-use quality of 13 Korean wheat cultivars

Table 2 Protein characteristics of 13 Korean wheat cultivars and five imported wheat flours1)

3.3. Proportion of gluten

The proportions of glutenin composition in the 13 Korean wheat cultivars and five imported wheat flours are summarized in Table 3. The proportions of gluten composition were measured by reversed-phase highperformance liquid chromatography (RP-HPLC); Fig. 1 shows the chromatogram of the Korean wheat cultivars.The proportions of HMW-GSs and LMW-GSs in the 13 Korean wheat cultivars were 42.1–47.3% and 52.7–57.9%,respectively. Younbaek showed the highest proportion of HMW-GSs and the lowest proportion of LMW-GSs, whereas a precisely contrasting tendency was seen in Joongmo 2008(i.e., the lowest proportion of HMW-GSs and the highest proportion of LMW-GSs). Younbaek and DNS showed similar proportions, while those proportions were similar in Joongmo 2008 and HRW wheat. The proportions of x-type, y-type, and the ratio of x- to y-type in HMW-GSs in the Korean wheat cultivars were 60.8–76.6%, 23.4–39.2%,and 1.77–3.31, respectively. The x- and y-type in the HMWGSs of the imported wheat flours could not be measured because different x- and y-types coexisted in each flour,which blended flour from several cultivars. The ratio of HMW-GSs to LMW-GSs in the Korean wheat cultivars was 0.73–0.90. Younbaek showed the highest ratio of HMWGSs to LMW-GSs, while Joongmo 2008 showed the lowest value. Imported wheat flours showed 0.72–1.09 in the ratio of HMW-GSs to LMW-GSs. Younbaek exhibited a ratio similar to that of DNS. Joongmo 2008 was similar to SW or HRW in the ratio of HMW-GSs to LMW-GSs.

The proportions of gliadin compositions in the 13 Korean wheat cultivars and five imported wheat flours are summarized in Table 4. The proportion of α+β gliadin,γ-gliadin, and ω-gliadin in the Korean wheat cultivars was 48.2–60.7%, 27.7–35.1%, and 11.6–21.0%, respectively,while those of the imported wheat flours was 33.3–53.6%,29.9–36.1%, and 14.1–33.3%, respectively. Jojoong showed the highest proportion of α+β gliadin, and the lowest proportions of γ-gliadin and ω-gliadin among all the wheat,including the imported wheat flours. Younbaek showed the lowest proportion of α+β gliadin and the highest proportion of ω-gliadin, while Suan showed a higher proportion of γ-gliadin than other Korean wheat cultivars. Younbaek was similar to AH in the proportion of α+β gliadin and had a higher proportion of ω-gliadin than all the imported wheat flours except for SW. Suan was intermediate between ASW and AH. The ratios of α+β to γ, α+β to ω, and γ to ω in the Korean wheat cultivars were 1.50–2.19, 2.85–5.28,and 1.68–2.73, respectively, which are higher values than those of the imported wheat flours, 1.00–1.79, 1.00–3.60,and 1.00–2.57, respectively. Jojoong showed higher α+β to γ and α+β to ω ratios, while Younbaek showed lower α+β to ω and γ to ω ratios than the other Korean wheat cultivars. These ratios in Jojoong were also higher than in the imported wheat flours, and lower in Younbaek, except for SW. Suan showed a lower α+β to γ ratio than the other Korean wheat cultivars; it was similar to that of Australian standard white wheat. Keumkang exhibited the highest γ to ω ratio among all the Korean wheat cultivars and imported wheat flours.

Table 3 Proportion and ratio of glutenin in 13 Korean wheat cultivars and five imported wheat flours

Fig. 1 The proportion of glutenin composition of Jojoong (A) and Goso (B), and gliadin composition of Younbaek (C) and Goso(D) using RP-HPLC (reversed-phase high performance liquid chromatography). HMW, amount of high molecular weight glutenin subunits; x-type, amount of x-type in HMW; y-type, amount of y-type in HMW; LMW, low molecular weight glutenin subunits; α+β,proportion of α+β gliadin; γ, proportion of γ-gliadin; ω, proportion of ω-gliadin.

Table 4 Proportion and ratio of gliadin composition of 13 Korean wheat cultivars and imported five wheat flours1)

3.4. End-use quality

Bread-loaf volume, texture of cooked noodles, and cookie diameter of the 13 Korean wheat cultivars and five imported wheat flours are summarized in Table 5. In the Korean wheat cultivars, the bread-loaf volume ranged from 609.7–913.9 mL. Keumkang (913.9 mL), Jokyung (870.8 mL), and Joongmo 2008 (884.7 mL) had a bread-loaf volume similar to that of DNS and HRW (916.7 and 833.3 mL, respectively).Bread-loaf volume in the Korean wheat cultivars was significantly correlated with SDS sedimentation volume(r=0.87,P＜0.001) and dry gluten content (r=0.68,P＜0.05),as well as with the water absorption (r=0.57,P＜0.05) and mixing time of the mixograph (r=0.58,P＜0.05) (Table 6).

For the texture of cooked noodles, the hardness for Korean wheat cultivars was 3.21–4.22 N compared to 3.43–5.36 N in the imported wheat flours. Further, the springiness was 0.90–0.93 in Korean wheat cultivars, 0.93–0.95 in the imported wheat flours. The cohesiveness for Korean wheat cultivars was 0.64–0.67, which was similar to the range of cohesiveness in imported wheat flours, 0.63–0.67. Jojoong and Joa were harder, springier and more cohesive than the other cultivars. Jojoong and HRW were about as hard, and Joa was about as springy as AH and about as cohesive as ASW. Korean wheat cultivars exhibited significant correlations between the hardness of cooked noodles and protein content (r=0.67,P＜0.05) and with mixograph water absorption (r=0.75,P＜0.01), but there were no significant correlations between springiness, cohesiveness, and protein properties. The texture properties of cooked noodles for Korean wheat cultivars were not significantly correlated with any proportion of gluten composition.

The cookie diameters for Korean wheat cultivars were 78.40–93.11 mm, which is a narrower range than for the imported wheat flours (70.78 to 96.49 mm). Uri exhibited the largest cookie diameter among Korean wheat and imported wheat flours and was similar in size to SW. The cookie diameter from Korean wheat cultivars was negatively correlated with SDS sedimentation volume (r=–0.70,P＜0.01) and mixograph water absorption (r=–0.71,P＜0.01),but was not significantly correlated with protein content despite a negative relationship (r=–0.53) (Table 6).

3.5. Classification of Korean wheat cultivars

The PCA was loaded to verify multivariate correlations between protein properties, gluten composition, and enduse quality (Fig. 2). For bread (Fig. 2-A), both principal components (PCs) explained 81.4% (PC1, 43.5% and PC2,37.8%). The PC1 was positively associated with protein content, SDS sedimentation, dry gluten, mixograph water absorption, and mixograph mixing time to a lesser extent.Bread-loaf volume was a critical component for good bread quality. PC1 explained the close relationship of breadloaf volume with protein content, SDS sedimentation, dry gluten, and mixograph water absorption and mixing time.However, PC2 explained the vector of bread-loaf volume in positive direction, but the vectors of protein content, SDS sedimentation, dry gluten, and mixograph water absorption and mixing time in negative direction. Both PCs showed no relationship between bread-loaf volume and gluten composition. Keumkang and Joongmo 2008 were suitable for good bread-loaf volume, like DNS or HRW.

Table 5 Bread-loaf volume, texture of cooked noodles, and cookie diameter from 13 Korean wheat cultivars and five imported wheat flours

Table 6 Correlation coefficients between protein and gluten and end-use quality parameters in 13 Korean wheat cultivars

For the texture of cooked noodles (Fig. 2-B), the PCs together explained 77.7% (PC1, 41.9% and PC2,35.8%). PC1 was strongly and negatively correlated with cohesiveness and positively correlated with protein properties, while PC2 was strongly and positively correlated with the proportion of α+β, the ratios of α+β to ω and α+β to γ, and negatively correlated with the proportion of γ- and ω-gliadin. Epsteinet al. (2002) reported that the hardness of cooked noodles was negatively correlated with their springiness, and cohesiveness was correlated with springiness in recombinant inbred lines with various waxy genotypes. Ross (2006) said that hardness was not significantly correlated with springiness, which was not significantly correlated with cohesiveness. Also, hardness was negatively or positively correlated with cohesiveness,depending on different cooking times. Cohesiveness was negatively related to springiness and hardness, and hardness was closely related to springiness for the Korean wheat cultivars. Joa, Suan, and Uri were clustered on the direction of the cohesiveness vector, and most Korean wheat cultivars were very close to them. HRW and AH were on the direction of the springiness vector, and DNS was strongly related to springiness and hardness, which can explain the difference between Korean wheat cultivars and HRW, AH,and DNS for hardness and springiness. PCA explained that Joa, Suan, and Uri were similar to SW in cohesiveness and were better than ASW. Also, both PCs showed that gluten composition was not related to the texture of cooked noodles for Korean wheat cultivars.

For cookies (Fig. 2-C), the PCs together explained 82.4% (PC1, 44.6% and PC2, 37.8%). PC1 was positively correlated with protein properties and negatively correlated with cookie diameter, so cookie diameter was negatively related to protein properties; i.e., cookie diameter decreases as protein content increase. Both PCs explained that cookie diameter was not close with gluten composition. PCA explained that Uri was suitable for large cookie diameter as a cookie quality among Korean wheat cultivars. Uri was clustered with SW; that is, Uri showed cookie quality similar to that of SW, whereas, Keumkang was not suitable for superior cookie quality like HRW and DNS.

Fig. 2 Principal component analysis (PCA) loading of the measured traits and values of components 1 and 2 with clustering corresponding to bread-loaf volume (A), texture of cooked noodles (B), and cookie diameter (C). SDSS, sodium dodecyl sulfate sedimentation volume; Abs, mixograph water absorption; time, mixograph mixing time; HMW, amount of high-molecular-weight glutenin subunits; x-type, amount of x-type in HMW; y-type, amount of y-type in HMW; LMW, low-molecular-weight glutenin subunits;HMW/LMW, ratio of HMW to LMW; α+β, proportion of (α+β)-gliadin; γ, proportion of γ-gliadin; ω, proportion of ω-gliadin; (α+β)/γ,ratio of (α+β)- to γ-gliadin; (α+β)/ω, ratio of (α+β)- to ω-gliadin; γ/ω, ratio of γ- to ω-gliadin; BLV, bread-loaf volume; HD, hardness of cooked noodles; SP, springiness of cooked noodles; CO, cohesiveness of cooked noodles; DIA, cookie diameter; AH, Australian hard; ASW, Australian standard white; SW, soft white; DNS, dark northern spring; HRW, hard red winter. Each number indicates a Korean wheat cultivar, in the following order: 1, Baekjoong; 2, Goso; 3, Hojoong; 4, Joa; 5, Jojoong; 6, Jokyung; 7, Joongmo 2008;8, Joongmo 2012; 9, Jopoom; 10, Keumkang; 11, Suan; 12, Uri; 13, Younbaek.

4. Discussion

4.1. Major effect on variation of protein characteristic and end-use qualities

The protein characteristics and end-use qualities were influenced by the year as well as the cultivar, as reported previously (Rozbickiet al. 2015). The proportion of gluten was also influenced by the year and cultivar (Triboiet al.2000). In this study, the cultivar influenced the protein characteristics, proportion of gluten composition, and end-use quality in Korean wheat cultivars while the year influenced them except for the γ- and ω-gliadin. The interaction between year and cultivar also influenced them except for the γ-gliadin.

4.2. Profiling of protein characteristics and proportion of gluten

Generally, wheat flours with high protein content exhibited high SDS sedimentation volumes and gluten content,because SDS sedimentation volume and dry gluten are influenced by protein content and quality. Protein content was correlated with SDS sedimentation volume(r=0.79,P＜0.01) and dry gluten content (r=0.93,P＜0.001).Mixograph parameters were used to compare wheat flours for their differences in protein quality, because dough-mixing properties of flours are mainly controlled by the quantity and quality of protein (Finney and Shogren 1972). The Korean wheat cultivars had a shorter mixing time in the mixograph than the imported wheat flours with similar water absorption by the mixograph. Protein content was correlated with the water absorption of the mixograph (r=0.95,P＜0.001),but was not significantly correlated with the mixing time of the mixograph, because of the narrow variation in Korean wheat cultivars.

For the correlation between the proportions of glutenin composition and protein characteristics, the proportions of HMW-GSs (r=–0.61,P＜0.05) and LMW-GSs (r=0.61,P＜0.05), and the ratio of HMW-GSs to LMW-GSs (r=–0.59,P＜0.05) were correlated with only SDS sedimentation volume. In contrast, the proportions of glutenin composition were not correlated with protein content, because the variation of protein content was not a linear regression with the variation of the proportions of HMW and LMW, and of xand y-type between the Korean wheat cultivars. Al-Khawani(1989) reported the variation of glutenin content according to class of wheat cultivars. However, Korean wheat cultivars showed similar proportions of glutenin composition, because of the narrow allelic background, almost all of which have a similar allelic composition, such as theGlu-D1fallele,which is composed of Dx2.2 and Dy12 glutenin subunits like Japanese wheat (Parket al. 2006; Leeet al. 2013).Because of the narrow genetic stock of glutenin, there was no correlation between protein content and the proportion of glutenin. Hence, the Korean wheat-breeding program should focus on the diversity of the genetic stock of glutenin.

The gliadin content differed considerably, depending on the class of wheat (Bietzet al. 1984; Al-Khawani 1989).However, Korean wheat cultivars showed similarity of the proportion of gliadin composition like the case of glutenin.It can be explained as a narrow variation of the proportion of gliadin composition similar to glutenin composition, e.g.,five cultivars with the same level as the proportion of α+β gliadin. In Korean wheat cultivars, the allelic composition of gliadin is limited; that is, there is no variation of ω5-gliadin between Korean wheat cultivars (Kimet al. 2016). Endoet al. (1993) reported different fragment areas of gliadin between Australian standard wheat and Japanese wheat cultivars during the same elution time. Most Korean wheat cultivars showed differences of the proportion of gliadin composition compared to that of the imported wheat flours,e.g., a higher proportion of α+β gliadin than in AH and ASW,but a lower proportion of γ-gliadin than in them. For the correlations between the proportions of gliadin composition and protein characteristics, there was a correlation only between the proportion of α+β gliadin and the dry gluten content (r=0.56,P＜0.05). Huebneret al. (1997) discussed correlations of γ-gliadin peak areas, such as the increase in the proportion of bread-loaf volume in HRW, and of γ-gliadin with the increase in total protein, whereas the other gliadins do not increase directly. The proportion of gliadin in the Korean wheat cultivars was not correlated with protein content. Furthermore, some Korean cultivars that had a higher protein content then the other cultivars did not show a higher proportion of γ-gliadin.

4.3. Evaluation of end-use quality

Bread-loaf volume is generally known to have a positive correlation with protein content and quality (Hussain 2009).In HMW and LMW, glutenin was positively correlated with bread-loaf volume (He and Hoseney 1992; Zhanget al.2007). The role of gliadin on bread-making quality is still debated, but many researchers found that gliadins are correlated directly to mixing time and bread-loaf volume(Parket al. 2006; Wanget al. 2007; Ohmet al. 2010).However, the proportion of gluten compositions as well as protein content was not significantly correlated with breadloaf volume, whereas the imported wheat flours with higher protein content showed a larger bread-loaf volume than all the other flours (Tables 2 and 5). For example, DNS showed the largest bread-loaf volume (916.70 mL), corresponding to the highest protein content (14.4%), while SW showed the smallest bread-loaf volume (662.50 mL), corresponding to the lowest protein content (9.6%). That is, the variation of bread-loaf volume is larger than the variation of protein content in Korean wheat cultivars. Glutenin and gliadin composition using RP-HPLC was used to identify suitable wheat cultivars for baking quality (Al-Khawani 1988). The composition and levels of protein expression are also related to dough properties and bread-making quality (Payne 1987;Shewry and Halford 2002). Keumkang, Jokyung, and Joongmo 2008 have theGlu-D1dallele, which is associated with better quality in bread baking than other alleles atGlu-1loci (Shewryet al. 1992), whereas most of the Korean wheat cultivars containing theGlu-D1fallele exhibited poor rheological properties and low bread-loaf volume (Parket al.2006; Leeet al. 2013). Wanget al. (2016) reported that the wheat line with theGlu-B3hallele shows a larger breadloaf volume than the other wheat lines without theGlu-B3hallele given the same level of protein content. Both Jokyung and Keumkang possess theGlu-B3hallele, but most other Korean wheat cultivars for noodles possess theGlu-B3dallele, which is rarely inherited in Argentinean, Australian,French, Japanese, and USA wheat cultivars (Eagleset al.2002; Branlardet al. 2003; Tanakaet al. 2005; Shanet al.2007; Lerneret al. 2009; Parket al. 2011).

The protein content of wheat is positively correlated with the texture of cooked noodles, especially hardness (Yunet al. 1996; Rosset al. 1997). Houet al. (2013) found that the proportions of protein composition as polymeric glutenin,monomeric gliadin, and other proteins using SE-HPLC (size exclusion high-performance liquid chromatography) are not significantly related to hardness, cohesiveness, and springiness of cooked Chinese white salted noodles made from flours from hard and soft wheat of the USA and hard white wheat of Australia. Epsteinet al. (2002) concluded that the texture of cooked noodles mainly depends on starch properties rather than protein contents and qualities.

Ohmet al. (2009) concluded that the correlation between SDS sedimentation volume and cookie characteristics means that the qualitative variation in protein significantly influences cookie characteristics more than quantitative variation under a narrow range of protein content. Huebneret al. (1999) reported that cookie diameter is negatively or not significantly correlated with each protein fraction using SE-HPLC in soft red winter wheat and soft white winter wheat. The cookie diameter for Korean wheat cultivars was also not significantly correlated with any proportion of gluten composition. Cookie diameter is negatively correlated with the particle size of flour and damaged starch content(Kanget al. 2014). Monsalve-Gonzalez and Pomeranz(1993) demonstrated that cookie diameter is negatively correlated with grain hardness because of the increase of water absorption and viscosity by damaged starch. Also,soft wheat show a higher ratio of spreading and longer expansion time than, and are thus superior to, hard wheat(Abboudet al. 1985). Hence, in the Korean wheat-breeding program, grain hardness might be important for the quality of a cookie.

5. Conclusion

The protein characteristics were somewhat correlated with bread-loaf volume, cookie diameter, and the texture of cooked noodles for the Korean wheat cultivars. Furthermore,the proportion of gluten using RP-HPLC was not correlated with them, because Korean wheat had a narrow variation of the protein content and proportion of gluten composition,given the narrow allelic diversity of Korean wheat. Hence,the Korean wheat-breeding program should focus more on enlarging the genetic stock of gluten composition to improve the end-use quality of Korean wheat cultivars.

Acknowledgements

This work was carried out with the support of Cooperative Research Program for Agriculture Science & Technology Development (Project title: Establishment of quality criteria for high uniformity in end-use of Korean wheat cultivars,PJ011009), Rural Development Administration, Republic of Korea.