Integrative transcriptomic and metabolomic analysis reveals variety-specific metabolic divergence and regulatory networks in fresh leaves of flue-cured tobacco

The metabolite composition of fresh tobacco leaves determines cured tobacco quality, yet the metabolic diversity among flue-cured tobacco varieties and its underlying regulatory mechanisms remain poorly understood. Here, we performed integrated widely targeted metabolomics (UPLC-MS/MS) and transcriptomics (RNA-seq) on fresh leaves of five major flue-cured tobacco cultivars (YY87, ZC208, KRK26, NC55, and YY301) grown in Shandong, China, to decipher the molecular basis of inter-varietal metabolic variation. We identified 2,522 metabolites, revealing significant metabolic divergence among varieties, with the most pronounced differences between NC55 and YY87 (748 differentially accumulated metabolites). Multi-omics integration highlighted phenylpropanoid and flavonoid biosynthesis as the core pathways driving this differentiation. Weighted gene co-expression network analysis (WGCNA) and K-means clustering identified gene modules strongly associated with specific metabolic phenotypes. Gene–metabolite association analysis identified key associations: the peroxidase 43-like gene LOC107786050 showed a strong positive correlation with the aroma precursor sinapaldehyde, whereas the cytochrome P450 98A2-like gene LOC107797681 was primarily associated with the accumulation of the antioxidant taxifolin. This study reveals distinct metabolic and transcriptional patterns among representative flue-cured tobacco cultivars and identifies candidate gene–metabolite associations related to aroma-precursor and flavonoid antioxidant metabolism. These findings provide potential targets for future validation and may support quality-oriented improvement of flue-cured tobacco cultivars.