Steroidal alkaloids (SAs) and steroidal glycoalkaloids (SGAs) are common constituents of plant species belonging to the Solanaceae family. However, the molecular mechanism regulating the formation of SAs and SGAs remains unknown. Here, genome-wide association mapping was used to elucidate SA and SGA regulation in tomatoes: a SlGAME5-like glycosyltransferase (Solyc10g085240) and the transcription factor SlDOG1 (Solyc10g085210) were significantly associated with steroidal alkaloid composition. In this study, it was found that rSlGAME5-like can catalyze a variety of substrates for glycosidation and can catalyze SA and flavonol pathways to form O-glucoside and O-galactoside in vitro. The overexpression of SlGAME5-like promoted α-tomatine, hydroxytomatine, and flavonol glycoside accumulation in tomatoes. Furthermore, assessments of natural variation combined with functional analyses identified SlDOG1 as a major determinant of tomato SGA content, which also promoted SA and SGA accumulation via the regulation of GAME gene expression. This study provides new insights into the regulatory mechanisms underlying SGA production in tomatoes.

Rice (Oryza sativa L.) is one of the most globally important crops, nutritionally and economically. Therefore, analyzing the genetic basis of its nutritional quality is a paramount prerequisite for cultivating new varieties with increased nutritional health. To systematically compare the nutritional quality differences between landraces and cultivated rice, and to mine key genes that determine the specific nutritional traits of landraces, a seed metabolome database of 985 nutritional metabolites covering amino acids, flavonoids, anthocyanins, and vitamins by a widely targeted metabolomic approach with 114 rice varieties (35 landraces and 79 cultivars) was established. To further reveal the molecular mechanism of the metabolic differences in landrace and cultivated rice seeds, four cultivars and six landrace seeds were selected for transcriptome and metabolome analysis during germination, respectively. The integrated analysis compared the metabolic profiles and transcriptomes of different types of rice, identifying 358 differentially accumulated metabolites (DAMs) and 1982 differentially expressed genes (DEGs), establishing a metabolite–gene correlation network. A PCA revealed anthocyanins, flavonoids, and lipids as the central differential nutritional metabolites between landraces and cultivated rice. The metabolite–gene correlation network was used to screen out 20 candidate genes postulated to be involved in the structural modification of anthocyanins. Five glycosyltransferases were verified to catalyze the glycosylation of anthocyanins by in vitro enzyme activity experiments. At the same time, the different mechanisms of the anthocyanin synthesis pathway and structural diversity in landrace and cultivated rice were systematically analyzed, providing new insights for the improvement and utilization of the nutritional quality of rice landrace varieties.