RESUMO
Celiac disease (CD) is a permanent gastrointestinal disorder characterized by the intolerance to a group of proteins called gluten present in wheat, rye, barley, and possibly oats. The only therapy is a strict lifelong gluten-free diet. The standard method for gluten determination in foods produced for CD patients is the R5-enzyme-linked immunosorbent assay (ELISA) as proposed by the recent Codex Alimentarius Draft Revised Standard. This test is based on the determination of prolamins, the alcohol-soluble proteins of gluten, and is available as a sandwich ELISA for intact proteins and as a competitive ELISA for gluten-derived peptides. While the suitability of the sandwich ELISA including a wheat prolamin (gliadin) reference for calibration has been shown by various studies and a ring test, the competitive ELISA still lacks a convenient reference for the quantitation of gluten peptides in fermented cereal foods (e.g., sourdough products, starch syrup, malt extracts, beer). Therefore, the aim of the present study was to prepare a suitable reference for the quantitation of partially hydrolyzed gluten in fermented wheat, rye, and barley products. The prolamin fractions from barley (hordein) and rye (secalin) were isolated from corresponding flours by means of a modified preparative Osborne fractionation. The prolamin fraction from wheat was obtained as reference gliadin from the Prolamin Working Group. The prolamin fractions were successively digested by pepsin and trypsin or pepsin and chymotrypsin procedures, which have been used for CD-specific toxicity tests on cereal storage proteins for many years. The protein/peptide content (N x 5.7) of the prolamin fractions and digests, which was the basis for the calculation of the gluten content by means of ELISA, varied between 67.1% and 96.0%. The prolamin fractions and enzymatic digests were then tested for their response in both sandwich and competitive assays. Intact prolamins responded similarly in both ELISA showing no important differences between the cereals. In the case of digested proteins, however, the sandwich ELISA was considerably less sensitive than the competitive ELISA. The former provided approximately 40% and the latter 70% of the signal intensity obtained with the intact prolamins. Thus, the combination of the competitive ELISA and the enzymatic digests of prolamin fractions as reference was considered to be an adequate system for the analysis of partially hydrolyzed gluten. The limit of detection using a peptic-tryptic hordein digest as reference was 2.3 microg prolamin equivalent per milliliter, and the limit of quantitation was 6.7 microg prolamin equivalent per milliliter. This system was applied for the determination of gluten equivalents in five commercial beverages based on fermented cereals.
