Floral scent is a crucial quality trait in ornamental plants, yet research has been hampered by the lack of standardized sensory evaluation and the disconnect between genes, volatile compounds, and human perception. Hedychium is an excellent model for fragrance research due to its diverse fragrance types and rich volatile organic compound (VOC) profiles. This study establishes a sensory-omics framework to connect genetic pathways, VOC chemistry, and fragrance perception in Hedychium flowers. A multidisciplinary approach combined sensory panel analysis (developing a fragrance wheel), VOC profiling (HS–SPME–GC–MS and PTR–ToF–MS), transcriptomics, and functional characterization of key biosynthetic genes in 30 Hedychium accessions representing six fragrance types. Six distinct fragrance types were classified (e.g. strong floral, fruity), linked to specific VOC profiles (e.g. monoterpenoids, esters). PTR–ToF–MS validated rapid detection of key fragrance markers. Supervised partial least squares-discriminant analysis (PLS-DA) modeling of VOC signatures enabled fragrance-type classification and key variable selection. Transcriptomic analysis coupled with weighted gene co-expression network analysis (WGCNA) revealed two key gene modules—MEbrown (terpenoid-associated) and MEyellow (phenylpropanoid-associated)—that underlie fragrance variation. Functional validation through in vitro enzymatic assays and transient overexpression in tobacco leaves confirmed HcTPS1 as a eucalyptol synthase and HmBEAT1 as a benzyl acetate synthase. Collectively, these findings provide a comprehensive framework for Hedychium flowers, thereby elucidating the molecular and chemical basis of their sensory fragrance variation. The study delivers valuable genetic resources and a predictive model that establishes a foundation for the targeted breeding of floral fragrance in ornamental horticulture.

Full text PDF download