The safety of gluten-free foods is essential for celiac disease (CD) patients to prevent serious complications. Enzyme-linked immunosorbent assays (ELISAs) are recommended for gluten analysis to monitor the compliance of gluten-free products to the Codex threshold of 20 mg gluten/kg. However, due to the specific features of each gluten ELISA test kit, the results often deviate systematically and largely depend on the characteristics of the antibody. This comprehensive study assessed the specificities and sensitivities of three monoclonal (R5, G12, and Skerritt) and two polyclonal antibodies to the alcohol-soluble prolamin and alcohol-insoluble glutelin fractions of gluten from wheat, rye, and barley, all of which harbor CD-active epitopes. Reversed-phase high-performance liquid chromatography served as independent reference method to quantify gluten protein concentrations and allow comparisons of different gluten fractions within one kit and between kits. Wheat prolamins were detected quite accurately by all antibodies, but high variability between antibody specificities and sensitivities was observed for rye and barley prolamins and rye glutelins, and the largest discrepancies were found for wheat and barley glutelins. The gluten content (sum of prolamins and glutelins) was either overestimated up to six times (rye) or underestimated up to seven times (barley). Overestimation of gluten contents may unnecessarily limit the availability of gluten-free products, but underestimation represents a serious health risk for CD patients. It is important to consider these differences between antibodies used in kits and consider what each kit is capable of measuring, especially with samples where the source of gluten is unknown.

No abstract