Construction of a non-coding RNA-mediated metabolic regulatory network to explore metabolic variation in tomato breeding

Abstract Tomato (Solanum lycopersicum), an economically important crop rich in secondary metabolites, serves as a model organism for studying secondary metabolism. However, research into how non-coding RNAs (ncRNAs) regulate metabolite accumulation in tomato is limited. Here, we identified and characterized 418 miRNAs in three tomato subgroups: Solanum pimpinellifolium (SP), Solanum lycopersicum var. cerasiforme (SLC), and Solanum lycopersicum var. lycopersicum (SLL). Notably, the Solanum pimpinellifolium genome exhibited the highest number of missing miRNAs, with targets linked to sugar metabolism. Next, we identified 133 miRNAs that were up-regulated during domestication; their target genes were associated with plant development and morphogenesis. Additionally, 2,938 long non-coding RNAs (lncRNAs) were identified and characterized, with intergenic lncRNAs being the most abundant type (70.2% of all lncRNAs). A total of 280 metabolites were identified from multiple tissues (flavonoids, steroidal glycoalkaloids, etc), and the steroidal glycoalkaloid content of all tissues was associated with domestication status. Finally, we constructed a ncRNA-mediated metabolic regulatory network for tomato, which included 369 lncRNA-miRNA pairs, 5,659 miRNA-mRNA pairs, and 136,599 mRNA-metabolite pairs. We found that miR172 isoforms target a UDP-glycosyltransferase (SlUGT71T212), which is involved in the glycosylation of flavonoids, and we validated the role of the miR172a-SlUGT71T212 module in the formation of naringenin-O-glycoside. Our study provides a comprehensive and valuable dataset of ncRNAs and metabolites in tomato. The ncRNA-mediated metabolic regulatory network is a useful theoretical resource for elucidating metabolic variation and its regulation during tomato breeding.