When analyzing the proteomics data for the presence of celiac epitopes, the celiac epitopes previously described by Sollid et al. [17
] were used. Table 1
summarizes the celiac epitopes detected in the different samples that were used in this study. Celiac epitopes were detected in all three wheat-related ancient grains (einkorn, emmer, and Kamut) analyzed in addition to rye, which is known to elicit the immune response associated with celiac disease. As expected, celiac epitopes were not detected in teff and sorghum as well as the pseudocereals, which are considered gluten-free and safe for celiac subjects. A third of the epitopes evaluated were detected in einkorn and four epitopes were found in Kamut, which are diploid and tetraploid wheat species, respectively, consisting of the AA genome and the AABB genome. In emmer, which is a tetraploid species (AABB), only one epitope was detected. However, this could be due to limitations in the method or the heterogeneity of the sample and does not imply that only this epitope is found in emmer. It is interesting to note that the FRPQQPYPQ (glia-α20) epitope was detected in all three ancient wheat species while the glia-α9 epitope (PFPQPQLPY) and epitope QGSFQPSQQ were detected in two of the three species. In our previous work, the glia-α-20 epitope was the most prevalent epitope analyzed, whereas the glia-α9 epitope was the second most prevalent in historical and modern wheat cultivars released in North Dakota between 1910 and 2013 [9
]. Three of the nine γ-gliadin epitopes were also found in the wheat species, although they were detected in a more random manner. The epitopes in ω-gliadin proteins were not detected in any of the ancient wheat species, which could be explained by the distribution of ω-gliadin T-cell epitopes encoded by the A/D genomes in comparison to the B genome [18
]. In hexaploid bread wheat, celiac epitopes are distributed throughout the chromosomes encoding ω-gliadin in the A/D genomes, whereas the B-encoded ω-gliadins do not show T-cell epitopes. Thus, the lack of the D genome in diploid and tetraploid ancient wheat species could be associated with the non-detection of ω-gliadin celiac epitopes. Additionally, non-detection could be due to genetic variation between the different ancient grain cultivars. According to these results, celiac epitopes can be found in ancient wheat species, which make them unsuitable for consumption by celiac subjects. According to Sollid et al. [17
], the proteins rich in glutamine and proline, collectively termed prolamins of rye are referred to as secalin proteins, and T-cell epitopes in secalin proteins are homologous to that of wheat [19
]. In this analysis, the proteomics data were used to determine the presence of two T-cell epitopes in rye secalin proteins, where one of the two peptides was detected. All three wheat-related ancient grain species and rye fall under the Poaceae
family which may explain why they contain celiac epitopes. In a few of the ancient grains, namely teff and sorghum, and in the pseudocereals that were analyzed, as expected, none of the celiac epitopes were detected, further confirming that they are celiac safe.
All ancient grains and pseudocereals used in the study belong to the division of Magnoliopsida. The ancient wheat species and rye are then categorized as monocots, whereas the pseudocereals are categorized as dicots and this is the point of taxonomic separation for these species.
Gliadin proteins can be separated into α/β, γ, and ω according to the electrophoretic mobility of the different protein fractions. As mentioned above, the α-gliadin fraction is considered to be highly antigenic compared to the other gliadin protein fractions. The different gliadin protein fractions have various roles in functionality as determined by previous studies. The separation of gliadin proteins into its different constituents can be achieved through reverse phase-HPLC (RP-HPLC). The gliadin/ prolamin protein profile of ancient wheat species and rye is shown in Figure 1
, where the different gliadin types are separated according to previous work [20
Using the data from the HPLC analysis, the percent area corresponding to individual gliadin proteins was determined for each sample, and significant differences in the percent area were analyzed. For all three gliadin protein types, the percent area values were significantly (p ≤ 0.001) different between the different ancient grains. For ω-gliadin proteins, rye showed the highest area percent (21%), whereas the Kamut showed the lowest (4%). On average, the percent area for the wheat-related ancient grains was less than half of that of rye. For α/β-gliadins, the highest area percent was found in einkorn (68%), whereas the lowest was found in rye (24%). In general, the area percent of the wheat-related ancient grains was more than double compared to that of rye. Therefore, when considering the abundance of α-gliadin proteins, rye may have lower celiac antigenicity. However, to produce celiac-safe rye, the antigenic epitopes need to be removed. The γ-gliadin protein percentage of the different species showed a trend similar to that of ω-gliadin proteins, with the highest percentage found in rye (55%), and the lowest in einkorn (23%).