Aplicação da metabolômica por RMN na diferenciação de cultivares de sorgo buscando um aumento na eficiência da produção de etanol

Abstract

Sorghum bicolor L. Moench is an excellent alternative to sugar cane and maize in the production of bioethanol since it is a cheaper crop, has a high starch content, high yield potential, low input demand, and relative tolerance to drought. Four sorghum genotypes, two lines (SC084, CMSXS180), and two hybrids (BRS305, BR501) were evaluated for their metabolic profile and efficiency in ethanol production. Metabolomic analysis has grown as a tool to understand how genetic and environmental factors affect phenotypic variations of plants. In this study, the use of NMR-based metabolomics was evaluated in differentiating four sorghum genotypes and its application in correlating these metabolic profiles with better efficiency in ethanol production. The content of total phenolic compounds, tannins, anthocyanins, total starch, ether extract, mineral matter, crude protein, and antioxidant activity were determined for the cultivars. Cultivar SC084 was the one with the highest levels of total phenolic compounds, tannins, anthocyanins, and crude protein, and the lowest total starch content. Non-targeted NMR-based metabolomic analysis showed that the different cultivars have different metabolic profiles, mainly in the aromatic region, and that the SC084 cultivar has a different profile from the others, standing out in the region at δ7.06-5.98. BR501 and CMSXS180 cultivars showed higher yields in bioethanol production through the simultaneous saccharification and fermentation process. Correlations between metabolites, tannin content, and degreasing treatment were possible through univariate and multivariate analyses. Tannin content is highly negatively correlated with ethanol content, while treatment of the grains does not influence this factor. This analysis identified the chemical shifts consistent with the presence of flavan-3-ol units from proanthocyanidins and the ones observed in the metabolic profile of the SC084 cultivar, which presented a lower yield in ethanol production. It was found that the presence of durrin and its catabolism products are present in the metabolome of the line CMSXS180, differentiating it from the others. The supervised random forest classification method showed that characteristic signals of flavan-3-ol units are important in classifying samples with high tannin content, suggesting the possibility of a predictive model for ethanol yield based on the intensity of 1H NMR signals. NMR-based metabolomics presented itself as a quick method for the metabolic differentiation of sorghum cultivars and it lends itself to studying the correlation of metabolic profiles with the process of obtaining ethanol. This study represents a starting point for the application of this tool in the sorghum production chain aiming for its use in the energy matrix.