Towards coeliac-safe bread

Abstract

Gluten-free foods cannot substitute for products made from wheat flour. When wheat products are digested, the remaining peptides can trigger an autoimmune disease in 1% of the North American and European population, called coeliac disease. Because wheat proteins are encoded by a large gene family, it has been impossible to use conventional breeding to select wheat varieties that are coeliac-safe. However, one can test the properties of protein variants by expressing single genes in coeliac-safe cereals like maize. One source of protein that can be considered as coeliac-safe and has bread-making properties is teff (Eragrostis tef), a grain consumed in Ethiopia. Here, we show that teff α-globulin3 (Etglo3) forms storage vacuoles in maize that are morphologically similar to those of wheat. Using transmission electron microscopy, immunogold labelling shows that Etglo3 is almost exclusively deposited in the storage vacuole as electron-dense aggregates. Of maize seed storage proteins, 27-kDa γ-zein is co-deposited with Etglo3. Etglo3 polymerizes via intermolecular disulphide bonds in maize, similar to wheat HMW glutenins under non-reducing conditions. Crossing maize Etglo3 transgenic lines with α-, β- and γ-zein RNA interference (RNAi) lines reveals that Etglo3 accumulation is only dramatically reduced in γ-zein RNAi background. This suggests that Etglo3 and 27-kDa γ-zein together cause storage vacuole formation and behave similar to the interactions of glutenins and gliadins in wheat. Therefore, expression of teff α-globulins in maize presents a major step in the development of a coeliac-safe grain with bread-making properties.

Keywords: gluten; maize; storage vacuole; teff; wheat; α-globulin.

Figure 1

Generation of Etglo3 transgenic maize plants by Agrobacterium‐mediated transformation. (a) Schematic diagram of the Etglo3 transgene construct. The left and right borders (LB and RB) of the T‐DNA binary vector flank the region that was transformed into maize; p10 and p27 represent 10‐kDa δ‐zein and 27‐kDa γ‐zein promoters, respectively; GFP, green fluorescent protein; T10, 10‐kDa δ‐zein terminator; T35S, 35S terminator; p35S, 35S promoter; Bar, bialaphos‐resistance gene; Tvsp, soybean vegetative storage protein terminator. (b) Cobs and transverse kernels of primary transformants (T1) of three independent transgenic lines (#1, #2 and #4) under fluorescent/white light. (c, d) SDS‐PAGE and immunoblotting of Etglo3 protein in total proteins from mature seeds of Etglo3 transgenic plants. Twenty micrograms of total protein per line was loaded onto two SDS‐PAGE gels. The left gel was stained with Coomassie Brilliant Blue, and the right gel was immunoblotted with the anti‐FLAG antibody. The red arrow in panel (c) indicates the visible accumulation of Etglo3 protein in Etglo3#2.

Figure 2

Quantification of Etglo3 and zein accumulation. (a) SDS‐PAGE of total proteins from mature seeds of Etglo3#2 and wild‐type Hi‐A×B. The SDS‐PAGE gel was 15% (w/v). The eight numbers on the right indicate the following bands on the gel: 1, Etglo3; 2, GFP; 3, 27‐kDa γ‐zein (27γ); 4, 22‐kDa α‐zein (22α); 5, 19‐kDa α‐zein (19α); 6, 16‐kDa γ‐zein (16γ); 7, 15‐kDa γ‐zein (15β); and 8, 10‐kDa δ‐zein (10δ). (b) Quantification of the eight bands in panel (a) by AlphaView software. The percentage per protein refers to protein accumulation compared to the total protein. The amounts of proteins ± SD from three replicates of Etglo3 transgenic plants and Hi‐A×B in the above gel are shown.

Figure 3

Transmission electron micrographs of developing endosperm cells of teff, wheat, maize and Etglo3 transgenic maize. SG, starch granule; PB, protein body; V, vacuole; CW, cell wall. In the panel of Etglo3 transgene maize, red arrowheads indicate electron‐dense aggregates.

Figure 4

Subcellular localization of Etglo3, 27‐kDa γ‐zein, 22‐kDa α‐zein and GFP by immunogold labelling TEM in the 20‐DAP starchy endosperm cells of Etglo3 transgenic plants. (a) and (b) show subcellular localization of Etglo3 by labeling the FLAG tag. (c) and (d) show subcellular localization of 27‐kDa γ‐zein. (e) and (f) show subcellular localization of 27‐kDa α‐. (g) and (h) show subcellular localization of GFP as the control. White arrowheads indicate immunogold particles. PB, protein body; ED, electron‐dense aggregates. In (g), the cytoplasmic (C) area marked with a white frame is enlarged and immunogold signals are highlighted with arrowheads. Bars = 200 nm in all images.

Figure 5

Immunoblotting detection of Etglo3 and gluten polymerization. (a) immunoblotting of total proteins of mature seeds of Etglo3 transgenic plants (Etglo3) and Hi‐A×B (Hi‐II) with an anti‐FLAG antibody under non‐reducing conditions. Total seed proteins were extracted and separated on the same two SDS‐PAGE gels (8% w/v). The left gel was stained with Coomassie Brilliant Blue (CBB), and the right gel was immunoblotted. Non‐reducing conditions are those in which the extraction and loading buffer did not contain β‐mercaptoethanol (reducing reagent). A piece of gel marked by the red rectangle on the top of CBB‐stained SDS‐PAGE was excised to perform LC‐MS. The top 30 identified proteins are listed in Table S2. (b) Immunoblotting of total wheat seed proteins with an anti‐gluten antibody. This antibody recognizes all wheat glutens. On the left (L), 5% (w/v) SDS‐PAGE was used to separate the total proteins of wheat mature seeds under non‐reducing conditions (without β‐mercaptoethanol in the extraction and loading buffer). On the right (R), 15% (w/v) SDS‐PAGE was used to separate the total proteins from wheat mature seeds under reducing conditions (containing 5% β‐mercaptoethanol in the loading buffer). The amount of protein loaded in each lane was 20 µg.

Figure 6

Bimolecular fluorescence complementation (BiFC) (a) and yeast two‐hybrid (Y2H) assay (b) For Review Only detection of the interaction between Etglo3 and different zeins. The fluorescence signal intensities represent the interaction in (a). The detailed Y2H results are depicted in Figure S6. ‘+’ represents interaction, and ‘−’ represents no interaction.

Figure 7

The genetic crosses between Etglo3 transgene and α‐/β‐/γ‐zein RNAi and combined RNAi lines. (a) Schematic diagram of the transformation constructs used in this study. (b,c) Immunoblotting of Etglo3 in mature seeds of progeny of the cross between Etglo3 transgene and the α‐zein RNAi line (b), and between β‐/γ‐zein and their combined RNAi line (c). In (b) and (c), Coomassie Brilliant Blue‐stained gels of zein proteins were used for the genotypes indicated, and immunoblotting with the anti‐FLAG antibody was performed using the corresponding non‐zein proteins.