Multivariate RP-HPLC analysis as a tool in quality studies in durum wheat

Abstract

Multivariate analysis of RP-HPLC data has been used in this study as a means of relating information derived from gliadin chromatograms to quality characteristics in durum wheat. In a study involving seven cultivars grown under distinct nitrogen and sulfur regimes, it was shown that for (i) grain sulfur content, (ii) grain protein content and (iii) Farinograph tolerance, strong correlations were detected between the observed values and those predicted from the most influential principal components (PC), whereas for a range of other quality characters no such correlations were seen. For (i) grain sulfur content, r predicted values was 0.739 before normalizing the data for total gliadin content and 0.710 after normalization, meaning that total gliadin content was not significantly involved in the observed relationship for this trait; the highest loadings for the first PC were located at 23.6, 32.5 and 39.5 min in the positive sense and at 36.1 min in the negative, indicating that these were the gliadins that most varied with the variation observed in sulfur content. For (ii) grain protein content, there was an expected effect of total gliadin content, reflected in the difference in r 0.411, respectively. Nonetheless, a significant correlation, though of intermediate magnitude, remained for the normalized data, where the principal gliadins involved eluted at 23.6, 26.9 and 39.5 min. For (iii) tolerance, r respectively, indicating that, as for sulfur content, total gliadin content did not strongly influence the observed relationships. In this case, the gliadins eluting at 23.3, 36.1 min (positively) and 32.5 min (negatively) were those with the highest loadings for the first PC. The three characters appear to be related in that they showed some common gliadin peaks with major influences. In contrast, the gliadin differences observed in this study did not appear to affect characters involved in evaluating gluten strength (gluten index and Farinograph energy level), and neither appeared to affect gluten content. These latter observations are consistent with previous evidence that it is the glutenins that exert a greater influence upon this aspect in durum wheat than the gliadins. It is concluded that multivariate PC analysis of gliadin chromatograms provides the means to identify which gliadin components are important in the determination of particular quality characteristics in durum wheat.