Plant Material

Four transgenic reduced-gliadin lines of Triticum aestivum cv. Bobwhite 208 (‘BW208’) and three transgenic reduced-gliadin lines of T. aestivum cv. Bobwhite 2003 (‘BW2003’), and their respective wild-type lines were assayed using a randomized complete block design with two replicates. Both wild types are spring wheat cultivars obtained by the CIMMYT from the cross CM 33203 with the pedigree Aurora//Kalyan/Bluebird/3/Woodpecker. Cultivar BW2003 has been selected for its high transformation efficiency, and carries the T1BL.1RS translocation from rye. On the other hand, BW208 derives from the SH 98 26 ‘Bobwhite’ line described as highly transformable by Pellegrineschi et al. [39], and does not contain the rye translocation. In the present study BW208 and BW2003 showed a total protein content of ∼11.6% and ∼10.5% of dry weight, respectively. The total gluten protein content (gliadins plus glutenins) was about 8.7% and 7.9% of dry weight, respectively for BW208 and BW2003.

All the transgenic reduced-gliadin lines were reported previously in [25], and contained the inverted repeat (IR) fragment ω/α (vectors pGhp-ω/α and/or pDhp-ω/α) designed to down regulate all the groups of gliadins by RNAi. Transgenic lines were self-pollinated for 4–5 generations, and they showed normal phenotypes in comparison to their corresponding wild types.

Grain Milling

White flour was obtained from each of the two independent repetitions of the reduced-gliadin and wild-type lines. Grains were hydrated to 16.5% humidity by addition of distilled water in two steps (24 h and 20 h before milling) with continuous shaking. Hydrated seeds (1 kg) of each line were milled separately in two steps in a CD1 Chopin (Chopin Technologies, Villeneuve-la-Garenne Cedex, France) standardized test mill. In the first step white flour and wholemeal flour were obtained. The wholemeal flour was reloaded in a second step of milling, and the resultant white flour was blended with that obtained previously, resulting in a total yield of about 60%. Flour was stored at room temperature (RT) for a week. Commercial rice flour supplied by Harinera Derivats del Blat de Moro, S.L. (Parets del Vallés, Spain) was used to make the gluten-free control bread.

Bread Baking

Dough was prepared on a flour weight basis: for 300 g flour, 180 ml water (225 ml water for the rice flour), 3.6 g baker’s yeast (Saf-Instant, Lesaffre, France) and 4.8 g table salt were added. Ingredients were mixed in a Farinograph (Brabender GmbH & Co. KG, Germany) for 4 min, and rested for 10 min with a plastic film cover to avoid drying. Dough was divided manually (50 g) and dough pieces were rolled mechanically in a ball homogenizer (Brabender GmbH & Co. KG, Germany). Dough pieces were placed on aluminum trays and fermented for 45 min at 30°C. Dough pieces were baked in an electrical convection oven (Eurofours, Gommegnies, France). The baking process was performed at fixed oven temperature of 180°C for 16 min with 2 initial steam injections of 10 sec each. After baking, bread loaves were rested for 30 min at RT to cool down.

Bread Characterization

Bread weight and volume were determined in three loaves from each sample. Bread volume was determined by the rapeseed displacement method [40]. Moisture content of the loaves was determined following the ICC Method No. 110/1 [41], with a pre-conditioning step of the bread samples. Three loaves from each sample and their medial slices were scanned (HP Scanjet 4400C, Hewlett–Packard, USA), and height and width were determined to subsequently calculate the width/height ratio. Crust and crumb color was determined by a Chroma Meter CR-400 colorimeter (Konica Minolta Sensing Inc., Japan), and expressed in a CIE-L* a* b* color scale (CIE-Lab). The CIE-Lab color space is composed by three perpendicular axis: L*, a* and b*. These three coordinates indicate the lightness of the color (L*; where L = 100 indicates white color and L = 0 black color), and its position between green and red (a*; where negative values indicate green and positive values indicate red), and between blue and yellow (b*; where negative values indicate blue and positive values indicate yellow). Two independent measurements were made to each of the three loaves to determine crust and crumb color.

Descriptive Sensory Analysis

A panel of 11 trained assessors was selected to evaluate the bread samples (n = 20; 10 samples with two repetitions) corresponding to the reduced-gliadin and wild-type wheat flours, and the rice flour. The range of experience of the test panelists of participating in descriptive analysis and scale rating of a wide range of bread products varied from 3 to 20 years. All the individuals composing the panel gave their informed consent. The panel evaluated appearance, aroma, flavor, and overall acceptance of each sample in a blind tasting. For evaluation, a set of six samples was presented in slices (1 cm thick) on plastic dishes coded and served in a randomized order. In addition, assessors were provided with mineral water in order to cleanse their palate between tastings. Each assessor received a list of sensorial attributes and their definitions to guide them during the sample evaluation.

Amino Acids Profile Characterization

Wholemeal flour was used for the characterization of amino acids profile of reduced-gliadin and wild-type lines. Flour samples were hydrolyzed using 6 N chloric acid and phenol, and then derivatized and analyzed. For the derivatization we used the AccQ Fluor reagent Kit (Waters). First, 20 µl of the hydrolyzed sample were mixed with 60 µl of buffer solution (0.2 M borate buffer), and afterwards 20 µl of derivatization reagent (2 mg/ml 6-aminoquinolyl-N-hydrosysuccinimidyl carbamate, AQC) was added according to the manufacturer’s instructions. After 10 min at 50°C, the solution was directly injected into the high-performance liquid chromatography tandem mass spectrometry (HPLC-MS/MS) system (Varian 320-MS). The amino acid separation was carried out using 2.5 mM ammonium acetate (pH = 5.75) as solvent A, and a solution of 2.5 mM ammonium acetate (pH = 6) and acetonitrile (30∶70, ammonium acetate:acetronitrile) as solvent B. The Pursuit XRs Ultra 2.8 C18 100×2.0 mm column (Agilent) was used as stationary phase, and the flow was 200 µl/min. The detection was performed by mass spectrometry (MS) with the electrospray ionization mode (ESI) (positive and negative). The amount of amino acid is expressed as percentage of the total sample weight.

Scanning Electron Microscopy (SEM)

Samples (dough and bread) were frozen and dried, and then fractured manually by using the tip of a razor blade and coated with gold. A JEOL JSM6300 scanning electron microscope (JEOL, Tokyo, Japan) was used to observe the samples at 15 kV at RT. SEM pictures at 1000x and 3000x magnifications were taken to the newly exposed surface of each sample. Samples from lines D894 (reduced gliadin content), E82 (reduced gliadin and low molecular weight (LMW) glutenin content), and the wild type BW208 were analyzed.

G12 Competitive ELISA

Gluten proteins were extracted according to the manufacturer’s instructions using Universal Gluten Extraction Solution UGES (Biomedal SL, Seville, Spain). Maxisorp microtiter plates (Nunc, Roskilde, Denmark) were coated with Prolamin Working Group (PWG) gliadin solution and incubated overnight at 4°C. The plates were washed with PBS-Tween 20 buffer and blocked with blocking solution (phosphate-buffered saline (PBS)-5% non-fat dry milk) for 1 h at RT. Different dilutions of each sample as well as standard solution of PWG gliadin were made in PBS-bovine serum albumin 3%, to each of which was added horseradish peroxidase–conjugated G12 mAb solution. The samples were pre-incubated at RT and added to the wells. After 30 min of incubation at RT, the plates were washed, and 3,3′,5,5′-tetramethylbenzidine (TMB) substrate solution (Sigma, St. Louis, Missouri, USA) was added. After 30 min of incubation at RT in the dark, the reaction was stopped with 1 M sulfuric acid, and the absorbance at 450 nm was measured (microplate reader UVM340, Asys Hitech GmbH, Eugendorf, Austria). Results were expressed in parts per million (ppm) in dry matter for the flour, and in a 35% humidity basis for the bread loaves. In order to estimate the total gluten content the results obtained were multiplied by two. Data were obtained from three independent experiments with samples run in triplicate.

Statistical Analysis

Data were analyzed with the statistical software R version 2.12.1 [42] using the Graphical User Interface (GUI) R Commander. Major assumptions of analysis of variance (ANOVA) were confirmed by the Shapiro-Wilk’s test for normal distribution (function ‘shapiro.test’, package stats), by the Levene’s test for homocedasticity (function ‘leveneTest’, package car) and the Ramsey’s regression equation specification error test (RESET) for linearity (‘resettest’; package lmtest), and variables were transformed if required. The statistical analysis between the different lines was carried out by using the analysis of variance (ANOVA) model ‘Variable ∼ Line+Block’ (function ‘aov’, package agricolae), followed by Tukey’s Honestly Significant Difference (HSD) post hoc all-pairwise comparison test (function ‘HSD.test’, package agricolae). In all the statistical analyses P values lower than 0.05 were considered significant.

The differences of the amino acids contents between the control and the low-gluten lines were assessed using the ANOVA model ‘Variable ∼ Block+Line’ with the function ‘lm’ (package stats). Mean comparisons were carried out by Dunnett’s post hoc multiple-comparison test (function ‘glht’, package multcomp). The box-and-whisker plot was plotted with the function ‘boxplot’ (package graphics).

Physical Characterization of reduced-gliadin Bread

The reduced-gliadin breads have been characterized and compared with those from wild-type wheat lines and from rice, as a control of market-typical gluten-free bread. All bread loaves were made by using the same formulation, and following the same mixing and baking methods. The overall visual appearance was very similar in all the wheat-bread loaves of both wild types and reduced-gliadin lines (Figure 1A and Figure S1A). The bread slices were cut mechanically, with both the reduced-gliadin lines and the wild types presenting similar visual appearance and porous structure, whereas the rice-bread showed a lighter and less porous surface than the wheat counterparts (Figure 1A and Figure S2A). Crust color was very similar in all the samples, whereas crumb showed greater differences between reduced-gliadin and wild type lines, and the rice control. The parameter a* (red-green range) was the most variable color parameter (Figures S1B and S2B). Although the shape (ratio width/height) and the weight were comparable in all the wheat loaves independent of the flour source, the volume (ml) and bread specific volume (ml/g) were reduced between 20–30% in the reduced-gliadin lines in comparison with the wild types (Figure 1B). This reduction could be explained by a reduced capacity of expansion of the dough during fermentation. Gliadins are known to contribute to the extensibility of the dough [43], and consequently, in the reduced-gliadin lines with all the gliadins down-regulated, the three-dimensional network formed by gluten proteins appears to have a reduced extension capacity. Nonetheless, it is sufficient to allow a significant expansion of the dough during fermentation and baking process, producing bread loaves of similar crust and crumb appearance to those made from wild types (Figures S1A and S2A). As reported by Piston et al. [36] lines D793, E82, D874, and E93 also presented down-regulation of the LMW fraction; however, none of the physical parameters studied in the present paper were significantly affected by this reduction. Current gluten-free breads available in the market are typically made with rice or maize flour, and although they represent economical ingredients, in general the loaf quality, shelf-life and organoleptic properties are substantially reduced compared to the wheat-flour counterparts [44]. Furthermore, the substitution of food with gluten-free alternatives may result in inadequate intakes of important nutrients since the nutritional properties (protein, fiber, and essential micronutrients) of gluten-free breads are frequently reduced [45], producing nutritional deficiencies in patients suffering gluten-related pathologies [46], [47].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Reduced-gliadin bread: physical properties.

(A) Loaves and slices of wild-type BW208, reduced-gliadin line D793, and rice. (B) Physical properties of bread loaves obtained from wild-type lines, reduced-gliadin lines, and rice. Lines with the same letter indicate that no significant differences exist among them as determined by the Tukey HSD post hoc all-pairwise comparison test (P<0.05).

https://doi.org/10.1371/journal.pone.0090898.g001

Organoleptic Properties of the reduced-gliadin Wheat Bread are Comparable to that of Normal Bread

A descriptive sensory analysis was carried out by a panel of 11 trained assessors, and the scores were expressed in a 1 to 9 hedonic scale (Figure 2A). The rice control presented significantly lower scores than the reduced-gliadin and wild-type wheat lines for all the parameters, indicating a higher quality of wheat breads. In addition, most of the reduced-gliadin lines showed statistically comparable levels of quality with their wild-type counterparts. Although the overall acceptance was reduced, with an average score of 7.4 for the wild types and 6.6 for the reduced-gliadin lines, no significant differences were found between most reduced-gliadin lines and wild types. By contrast, the overall acceptance of the rice was significantly lower, presenting a score of 2.4. This indicates, i) there are no differences in terms of quality between these reduced-gliadin breads and normal wheat flour breads, and ii) potential consumers would prefer the reduced-gliadin bread rather than the rice bread. In addition, a further advantage of the reduced-gliadin bread is that it can be made using standard, simple recipes, in contrast to the complex recipes currently used for the manufacture of the rice or maize gluten-free products, which are necessary to give these products an acceptable baking and organoleptic quality [48]–[50].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Reduced-gliadin bread: organoleptic and nutritional properties.

(A) Descriptive sensory analysis. Lines with the same letter indicate that no significant differences exist among them as determined by the Tukey HSD post hoc all-pairwise comparison test (P<0.05). Sensory scale ranges from 1–9. Wheat wild types are represented by black bars, reduced-gliadin lines are represented by white bars, and rice is represented by gray bars. (B) Box-and-whisker plot of lysine content (%) in wild types and reduced-gliadin lines. Dark grey boxes represent the wild-type lines. Significant differences between reduced-gliadin lines and their respective controls are indicated in pale grey (P<0.1), and white (P<0.05).

https://doi.org/10.1371/journal.pone.0090898.g002

No differences were found for the organoleptic parameters between the reduced-gliadin lines with and without reduction of the LMW fraction. This indicates that, although LMW glutenins play an important role in the bread-making quality, the down-regulation of the LMW glutenins in the lines described here does not affect the overall acceptance of the bread.

The reduced-gliadin Wheat Bread has Increased Lysine Content

Flour samples from reduced-gliadin wheat and wild types were compared in terms of amino acid composition (Table S1). Interestingly, the content of lysine was significantly increased in all the reduced-gliadin lines (Figure 2B), with increments that ranged between 24–67% respect to the wild types. In wheat, gliadins are known to contain lower amounts of lysine (around 50% less) than glutenins [51]. Consequently, the lysine increase observed in the reduced-gliadin lines may be an indirect consequence of the down-regulation of the lysine-poor gliadins and the compensatory increase in other more lysine-rich grain proteins, like the high molecular weight (HMW) glutenins or albumins and globulins, that we have observed in most of these lines [52]. Cereal proteins generally exhibit poor nutritional quality because of a lack of balance in amino acid composition and a low content of lysine. Lysine is considered the most important essential amino acid, and due to it is not synthesized in animals it must be acquired through diet. There is great interest in increasing the content of lysine in cereal crops since it has both an economical and humanitarian importance, especially in developing countries where the diet is mainly composed by a single cereal [53], [54]. Mutant high-Lys lines have been obtained in maize opaque-2 mutants [55] and opaque-2-derived quality protein maize (QPM) lines [56], [57]. Genetic engineering approaches have been also used to increase the lysine content in maize [58], [59] and rice [60]. However, so far there has not been any study describing high-Lys wheat lines. Previous studies with opaque-2 mutants and QPM maize lines have demonstrated the enhancement of the nutritional properties in animal (rats, pigs, and chickens) [61]–[64] and human nutrition [65], [66], with higher utilizable protein values as a consequence of the lysine increase and a more balanced amino acid composition. Consequently, the increased content of lysine in the wheat lines described in this paper potentially confers an increased nutritive value to the reduced-gliadin breads.

Dough and Bread Crumb Microstructure

Scanning electron microscopy (SEM) was used to study the microstructure of the dough in reduced-gliadin and wild-type wheat lines (Figure 3). The micrographs from the wild-type dough (Figure 3A) showed a distinct gluten film surrounding the small and large starch granules. Similar microstructures have been previously observed in wheat doughs [67]. However, in the reduced-gliadin lines (Figures 3B and 3C) most of the starch granules appeared naked, especially in the line E82 (Figure 3B). It seems that in the reduced-gliadin lines the gluten film is unable to surround all of the starch granules. A possible explanation for this is that the lack of gliadins can be affecting the formation of the gluten structure. It has been previously reported that gliadins may be involved in the development of the structure of the gluten film networks through covalent and non-covalent bonding with other gluten proteins [68]. In addition, some gliadins may be also form inter-molecular disulfide bonds in the gluten [69], stabilizing the gluten film during dough formation. Consequently, the inability of the gluten network to surround the starch granules in the reduced-gliadin lines can be explained by either the reduction of gliadins, and/or the increase in the ratio glutenins:gliadins. The microstructure of the bread crumb was also analyzed in the reduced-gliadin and the wild-type wheat lines by SEM (Figure S3). In all cases the micrographs showed complex network structures with numerous cavities independently of the line. This network structure corresponds to gluten threads intimately linked to smooth jelly areas resulting from starch gelling, as described previously in [67]. In general the wild types displayed bigger pores than the reduced-gliadin lines, as expected given the higher volume of the wild-type breads. In Figure S3 it can be observed that the reduced-gliadin line E82 had a similar microstructure to the wild type, with bigger pores than the line D894. In addition, in the line D894 elongated protein strands were more frequently observed.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Dough microstructure in reduced-gliadin and wild-type wheat lines.

SEM was carried out at 1000x and 3000x magnifications in wild type BW208 (A), and reduced-gliadin lines E82 (B), and D894 (C). Arrows indicate starch granules covered by gluten in panel (A), and naked starch granules in panels (B) and (C).

https://doi.org/10.1371/journal.pone.0090898.g003

Predicted Tolerable Daily Intake of reduced-gliadin Bread by Celiac Patients

Finally, the gluten content expressed in parts per million (ppm) of flour and bread samples was determined by a competitive ELISA based on the G12 monoclonal antibody (mAb). The G12 mAb specifically recognizes the hexapeptide QPQLPY present in the 33-mer peptide of α2-gliadin, binding also to related peptide variants found in other immunotoxic gluten proteins [70]. The 33-mer peptide has been identified as one of the digestion-resistant gluten peptides and the main immunodominant toxic peptide in celiac patients [71]. The sensitivity and epitope preferences of G12 antibody was found to be useful for detecting gluten-relevant peptides to infer the potential toxicity of food for patients with CD [72]. The reactivity of G12 mAb with cereal storage proteins of different varieties of cereals was correlated with the immunotoxicity of the dietary grains [73]. We observed that the wild type BW2003 had a higher content of gluten than its counterpart BW208 in both flour and bread samples. In addition, all the reduced-gliadin lines showed a large reduction of the immunotoxicity in comparison with their respective wild types, with the BW208 reduced-gliadin lines displaying a lower content of gluten than their BW2003 counterparts (Table 1). The amount of gluten was strongly decreased in the bread loaves compared to the flour samples, especially in the reduced-gliadin lines. The lines with the strongest depletion of gluten in loaves were line D793 (∼ 97% of reduction) and line E82 (∼ 96% of reduction) (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Gluten content in flour and bread loaves, and estimated maximum tolerable daily intake of bread from wild types and reduced-gliadin lines.

https://doi.org/10.1371/journal.pone.0090898.t001

These results were compared with the predicted tolerable daily intake of gluten described by Catassi et al. [74]. In that study, a daily intake of 50 mg gluten was established as the maximum dose in order to avoid damage in the small intestine of celiac patients during a prolonged exposure of 90 days. This means that for the wild types the maximum tolerated amount of bread per day would be, in the best of the scenarios, less than 1.9 g. However, for the line E82 a celiac individual could ingest up to 43.6 g per day of bread, and for the line D793 the tolerated amount would be of up to 66.9 g, as determined by the G12 test. In addition, we have previously reported [25] that the immunotoxicity of the remaining gluten in these reduced-gliadin wheat lines was up to 100 times lower than the gluten of the wild type wheat, as determined by T-cell assays. Consequently, the above estimations of the tolerated daily intake might be even higher, and still be safe for most of the CD patients. However, in order to determine whether or not the product can be consumed by the general celiac and gluten-intolerant population, as well as the actual tolerated amount that can be safely ingested, additional studies including feeding trials with gluten-sensitive/intolerant patients carrying various combinations of the susceptibility genes are still needed.