Wheat storage proteins: changes on the glutenins after wheat infection with different isolates of Fusarium graminearum

Wheat gluten proteins are decisive for the industrial properties of flour, so alterations resulting from grain infection with Fusarium graminearum produce changes in the glutenin content that affect the baking properties. This work analyzes the high-molecular-weight glutenin changes from wheat flour with different degrees of F. graminearum infection at field, since these proteins are determinant for the quality properties of flour. Wheat cultivars — on field trials — infected with F. graminearum isolates of diverse aggressiveness showed severity values between 9.1 and 42.58% and thousand kernel weight values between 28.12 and 32.33 g. Negative correlations between severity and protein content and positive correlations between yield and protein content were observed, employing reversed-phase high-performance liquid chromatography and polyacrylamide gel electrophoresis. Furthermore, the protein signal changes were in agreement for both methodological approaches. Also, the degree of disease observed and the protein changes on infected wheat cultivars varied in relation with the aggressiveness of the isolate responsible for the infection. The principal component analysis showed a close arrangement among protein values obtained by HPLC. For each cultivar, two principal components were obtained, which explained 80.85%, 88.48%, and 93.33% of the total variance (cultivars Sy200, AGP Fast, and Klein Tigre respectively). To our knowledge, the approaches employed for the analysis of protein changes according to the degree of disease, as well as the thorough statistical analysis, are novel for the study of Fusarium Head Blight.


Introduction
Wheat (Triticum aestivum) is one of the most extended crops around the world, along with corn and rice.Nowadays, the principal exporters are the USA, Canada, the European Union, Argentina, and Australia (FAO 2015).Wheat flour-main derivate of graincomes from the endosperm, which is the most abundant structural component of the grain.The endosperm is mainly composed of starch and storage proteins, which function as an energy reservoir during germination (DuPont and Altenbach 2003;Gianibelli et al. 2001;Gras et al. 2001;Shewry et al. 2013).The protein content determines the market value, industrial quality, and end use of the grains and flours, since it forms the viscoelastic net necessary to retain the gas during the fermentation process (Wieser 2007).Gluten is composed of gliadins and glutenins (DuPont and Altenbach 2003;Gianibelli et al. 2001;Gras et al. 2001).Both types of proteins have different effects on dough; while monomeric gliadins determine viscosity and extensibility, glutenins' capacity to form disulfure bonds is responsible for the elasticity and cohesiveness of the dough (Belton 2005;Kamal et al. 2009;Payne et al. 1987).The glutenins are classified into low-and high-molecular-weight glutenin subunits (LMW-GS and HMW-GS respectively) (Belton 2005;D'Ovidio and Masci 2004).HMW-GS proteins are encoded by three genes (GluA1, GluB1, and GluD1) that constitute the Glu-1 gene group.Each gene presents different alleles that code for different proteins, which have a differential influence on the quality of wheat, so alterations in the HMW-GS protein fraction can reduce the quality of the flour (Payne et al. 1987).
Wheat is susceptible to various diseases, Fusarium Head Blight (FHB) being one of the most devastating diseases worldwide, since it produces high economic losses due to yield and quality reduction (Kikot et al. 2011;McMullen et al. 1997;Monds et al. 2005), and the associated presence of mycotoxins in the infected grains (Palacios et al. 2011).The FHB disease is mainly caused by members of the Fusarium genus, Fusarium graminearum sensu stricto (s.s.) being the principal etiological agent in wet and temperate geographic regions (Alvarez et al. 2011;Castañares et al. 2014;Ortega et al. 2016).When the wheat plant is on the phenological state of anthesis, the pathogen can penetrate and infect the exposed tissues (Pritsch et al. 2000).Once the colonization has initialized, the fungal hypha produces enzymes that degrade the structural components of the cell walls, allowing the spread of the disease through the spikes (Kang et al. 2005;Kang and Buchenauer 2000;Kikot et al. 2009;Mesterházy et al. 1999;Wanjiru et al. 2002).On the late infection stage, the protease enzymes released can degrade the gluten, causing a reduction in the quality and industrial properties of the derived flours (Barneix 2007;Brzozowski et al. 2008).The pathogen capacity to release these enzymes can determine the aggressiveness of each isolate.The high intraspecific diversity (Akinsanmi et al. 2008;Tóth et al. 2005) generates variability in the aggressiveness of the isolates, and as a consequence it induces different degrees of infection (Malbrán et al. 2012).
In wheat breeding programs, where the size of the sample is limiting, the electrophoresis and chromatography are among the most suitable techniques of analysis (Eggert et al. 2010 ).Due to the high resolution of reversed-phase and molecular exclusion high-performance liquid chromatography, both techniques are widely employed in the study of the composition of infected and uninfected wheat flours (Larroque et al. 2000;Naeem and Sapirstein 2007;Ueno et al. 2002) To our knowledge, the majority of the researches related to Fusarium infection and wheat protein changes are only focused on the major components of gluten, the studies of single protein type being scarce (Horvat et al. 2014).The present research provides a novel approach to the relation between field infection and protein changes measured by analytical methods.Therefore, the aim of the present work was to analyze the changes on HMW-GS proteins of whole wheat flour from grains with different degrees of Fusarium graminearum s.s.infection, since these proteins are determinant for the quality properties of flours.

Biological Material
Three wheat cultivars that belong to the national wheat breeding program of Marcos Juárez-INTA (Córdoba, Argentina) were employed in this study: Sy 200, AGP Fast, and Klein Tigre, selected according to their growth cycle and resistance to infection.The international highresistant cultivar Sumai 3was employed as a control (Table 1).The wheat plants were sowed during winter 2013 in the experimental field of Marcos Juárez-INTA (32°42′ S; 62°06′ W) under conditions similar to those in the field, but with partial control of humidity using a field microsprinkler, as in previous works (Alberione et al. 2016).
The infection was carried out with three Fusarium graminearum sensu stricto isolates with different degrees  HMW-GS, high-molecular-weight glutenin subunit; LMW-GS, lowmolecular-weight glutenin subunit; P1-P7, proteic peaks; AU, absorbance units at 206 nm of aggressiveness according to a previous characterization (Ortega 2017;Ortega et al. 2018) (Table 2).At anthesis, wheat spikes were independently inoculated with a suspension of 3 × 10 5 conidia/mL of each isolate by means of a hand sprayer (Alberione et al. 2016).Nine treatments were obtained from the combination of each cultivar with an isolate.The control plants were inoculated with distilled water.The experimental design was randomized complete blocks and the experimental units were hill plots.The assay was done by quadruplicate.

Infection parameters: severity and yield
After harvest, the degree of infection was estimated calculating the percentage of infected spikelets by spike, which is related to infection severity, and the thousand kernel weight (TKW) related to yield (Alberione et al. 2016).

Protein extraction
After grain severity and yield measurements, the whole wheat flours were obtained by milling the grains with an electric grinder FW 100 (Arcano, Argentina).The protein extraction was carried out according to Naeem and Sapirstein (2007).Fifty milligrams of flour from each sample was used for the extractions.The gliadin fraction was extracted by three consecutive washes of the flours with 500 μL of 1-propanol (0.5 mL/mL) solution, for 15 min at room temperature, and it was centrifuged for 3 min at 15000×g.The supernatant was discarded.The glutenins contained in the pellet were extracted with 150 μL of buffer (0.08 M Tris-HCl, 0.5 mL/mL1-propanol, pH 7.5) with 10 mg/mL dithiothreitol (DTT).The mix was incubated for 30 min at 60 °C and then 150 μL of the extraction buffer with 40 μL/mL 4-vinylpyridine was added and incubated for 30 min at 60 °C.The mix was centrifuged and fractioned.Lastly, the samples were analyzed by highperformance liquid chromatography by reversed phase (RP-HPLC).For the electrophoretic analysis, the samples were lyophilized and stored at − 4 °C until posterior use.

Chromatographic analysis
The analysis of the samples by RP-HPLC was carried out according to Naeem and Sapirstein (2007) with modifications.The employed equipment was a Waters 717 with reversed-phase column Aquapore RP-300 7m (300-μm pore size, 7-μm particle size).The employed solvents were the following: solvent A-80 μL of trifluoroacetic acid (TFA) in 1 L of bidistilled water; solvent B-80 μL of TFA, 440 mL acetonitrile, 560 mL bidistilled water.The gradient was the following (% of solvent B): 0-7 min 52% B; 7-54 min 100% B; 54-55 min 52% B; 55-65 min 52% B. The injection volume was 20 μL and the flow was 1 mL/min.The detection was carried out with diode array equipment (Waters, MA, USA) at 206 nm.In the chromatograms, the area under peaks was integrated as an estimate of the protein content employing Empower software (Waters, MA, USA).

Electrophoretic analysis
Electrophoresis was carried out on mini-PROTEAN Tetra cell (Bio-Rad, CA, USA) equipment, on 8% polyacrylamide gel employing 10-180-kDa Page Ruler molecular marker (Thermo Fisher, MA, USA).The dried samples previously obtained were resuspended in 300 μL of sample buffer (40 mg/mL SDS, 120 μL/mL glycerol, 6.1 mg/ mL Tris-HCl, 50 μL/mL β-mercaptoethanol, 10 mg/mL DTT and 0.1 mg/mL Coomassie Brilliant Blue R 250, pH 6.8), sonicated for 30 min and then incubated for 5 min at 90 °C according to Eggert et al. (2011).The pictures were taken with Quantity One software (Bio-Rad, CA, USA) and analyzed with the software Gel Analyzer 2010 (Freeware).The density values obtained from the densitograms were employed as an estimate of the protein content of the band.The identification of the different electrophoretic profiles obtained from HMW-GS glutenins was performed according to Payne et al. (1987).

Statistical analysis
The data was analyzed employing STATISTICA 7 (StatSoft, OK, USA).The relations between variables were analyzed by Pearson's correlations, the isolate effect on the severity degree was analyzed by variance analysis (ANOVA), and how proteins clustered was observed by principal component analysis (PCA).

Results and discussion
The artificial infection of wheat plants employing isolates with different aggressiveness allowed us to observe a wide range of disease severity in relation to the changes detected in the HMW-GS protein fraction by analytical techniques.These proteins are responsible for important baking characteristics in the industry (Belton 2005;D'Ovidio and Masci 2004;Kamal et al. 2009).
The effect of the disease on the 9 treatments of field-grown wheat under controlled conditions was measured as the percentage of severity and TKW.As shown in Fig. 1, a diverse degree of infection was obtained from the wheat plants infected with F. graminearum isolates of different aggressiveness.Furthermore, it was observed that the isolate 1 produced the lowest disease effects, which was in agreement with its characterization of aggressiveness.
The RP-HPLC studies were carried out from the protein extracts obtained of whole wheat flour for each treatment.Representative protein elution profiles of wheat cultivars Sy 200, AGP Fast, and Klein Tigre and the control Sumai 3 infected with isolate 2 are shown in Fig. 2.
The regions corresponding to HMW-GS and LMW-GS fractions were identified by the retention time according to previous researches (Eggert et al. 2010;Naeem and Sapirstein 2007).As an estimate of protein content, the areas under the peaks were employed.Table 3 shows the correlation between values of protein peak integrations for each cultivar.It can be observed, for the cultivars Sy 200 and AGP Fast, that most of the peaks correlated positively with each other, whereas for cultivar Klein Tigre, peaks showed some negative  correlations.The positive correlations suggest that protein changes resulting from infection were similar within analyzed protein group.
In order to analyze the effect of the disease, Pearson's correlation between the values of proteic profile obtained by HPLC and disease severity and TKW-representing the degree of infection-was calculated (Table 4).
Almost all correlations between protein content and severity were negative, indicating that the higher the severity, the lower the protein signals.For the yield-measured as TKW-the correlations were positive, but less significant.These results were in concordance with the observations of Wang et al. (2005), who reported that higher levels of infection with Fusarium culmorum reduced the glutenin percentage-especially HMW-GS; meanwhile, Nightingale et al. 1999 found less amount of proteins in Fusarium damaged kernels compared to clean grains.Eggert et al. (2010) compared the effect of natural and artificial infection and found that the latter induced lower protein values.
The isolate effect on the protein values was also analyzed by ANOVA, finding significant differences among them.Isolate 1 was found to have the lowest degree of infection as it produced the lowest protein changes (Table 4).
From the protein data analyzed by PCA for each cultivar, two principal components were obtained, which explained 80.85%, 88.48%, and 93.33% of the total variance (cultivars Sy200, AGP Fast, and Klein Tigre respectively).The representation of the principal components is shown in Fig. 3, in which the peaks were grouped for all cultivars, except for those that present lower Pearson's correlations values.
Lastly, the protein extracts obtained from the flours were analyzed by SDS-PAGE.A representative gel of each analyzed wheat cultivar is shown in Fig. 4.
The density values obtained from the densitograms were employed as an estimate of the band protein content, which lead to the positive correlation with the infection degree.The analysis of Pearson's correlations of the protein bands showed that almost all band values were positively correlated, showing similitude with the results observed by RP-HPLC (Table 5).
Pearson's correlations between the values obtained by SDS-PAGE and severity and TKW were calculated in order to analyze the effect of the disease.These results are shown in Table 6.Negative correlations between protein band values and severity were found for all cultivars, indicating that at a higher severity, a lower signal of proteins was found.In relation to yield, almost all correlations were positive.Similar results were previously observed, where negative effects on HMW-GS by Fusarium infection were reported (Eggert et al. 2010;Horvat et al. 2014;Kreuzberger et al. 2015;Wang et al. 2005).
In conclusion, negative correlations between the protein profile and the disease severity could be observed employing both techniques.This would indicate that the protein fraction is sensitive to alterations caused by the disease.Although the SDS-PAGE methodology is a reference technique for the identification of the HMW-GS profile, the HPLC proved to be a complementary, powerful, and rapid method for the analysis of the severity degree effect in the protein fraction given its resolution and sensibility.To our knowledge, the approaches employed for the analysis of protein changes according to the degree of disease, as well as the thorough statistical analysis, are novel for the analysis of FHB and its implications in the quality of flours.

Fig. 1
Fig. 1 Severity and thousand kernel weight of wheat field treatments.TKW, thousand kernel weight.The same letter means no significant differences, obtained by ANOVA

Table 1
Cultivars employed in field assay PG, polygalacturonase; +, relative enzymatic and toxicologic productions

Table 3
Pearson's correlation matrix from RP-HPLC areas under the peaks for wheat cultivars

Table 4
Infection effect on chromatogram peak areas by RP-HPLC according to Pearson's correlation values between disease degree parameters and ANOVA for the isolate effect

Table 5
Pearson's correlation matrix of bands densitograms from SDS-PAGE for wheat cultivars

Table 6
Infection effect on band densitograms from SDS-PAGE according to Pearson's correlation values between disease degree parameters and ANOVA for the isolates effect