Genomic, metabonomic and transcriptomic analyses of sweet osmanthus varieties provide insights into floral aroma formation

Sweet osmanthus is an important ornamental evergreen tree or shrub, which has been widely used in the world's subtropical landscapes and the fragrance industry due to its deliciously fragrant flowers. In this study, whole-genome resequencing, RNA sequencing, and GC-MS analysis were used to compare 30 accessions of Osmanthus fragrans to gain insights into taxonomic classification and the diversity of volatile floral scents. A total of 13,466,389 high confidence SNPs were identified from the accessions with an average sequencing depth of 16.96 × . Phylogenetic analysis of the sequences of the 30 ancient or improved O. fragrans genotypes with diverse origins could be roughly divided into two differentiated sweet osmanthus populations, corresponding to the autumn-flowering and the four-seasons flowering characteristics. The odor activity values and orthogonal projections to latent structures-discriminant analysis of GC-MS metabolomics data indicated that linalool was one of the most important volatile organic compounds (VOCs) which can determine the scent diversity among different sweet osmanthus cultivars. Weighted gene co-expression network analysis (WGCNA) identified 23 distinct modules, four of which were highly associated with important floral VOCs. Based on WGCNA analysis, a critical enzyme gene OfHDR2.1 in the MEP pathway and its 12 co-expression transcription factors from the turquoise module were considered to play important roles in linalool synthesis among differently scented O. fragrans populations. Thus, our study provided new clues for the classification of O. fragrans cultivars and valuable genetic resources for future molecular breeding of O. fragrans accessions improved with respect to floral aroma.