Search

Additional file 1: Table S1. Structures of compounds related to this work. Table S2. List of compounds used as potential C-8 hydroxylase inducers. Table S3. Accession number and origin of sequences used in the phylogenetic analysis. Table S4. Sequences of flavonoid C-8 hydroxylase gene candidates, primer sequences and transcription units. Figure S1. Nucleotide sequences encoding the amino acids fdeE—line 1, and Rg3610long (RgF8H)—line 2. Rg6421, Rg2726. Highly conserved FAD and GD motifs were highlighted in boxes using red and blue underscores, respectively. Figure S2. LC–MS analysis of naringenin hydroxylation by RgF8H. (A) Mass spectrum of naringenin, (B) 8-hydroxynaringenin. Figure S3. LC–MS analysis of eriodictyol hydroxylation by RgF8H. (A) Mass spectrum of eriodictyol, (B) 8-hydroxyeriodictyol. Figure S4. LC–MS analysis of pinocembrin hydroxylation by RgF8H. (A) Mass spectrum of pinocembrin, (B) 8-hydroxypinocembrin. Figure S5. LC–MS analysis of hesperetin hydroxylation by RgF8H. (A) Mass spectrum of hesperetin, (B) 8-hydroxyhesperetin. Figure S6. LC–MS analysis of luteolin hydroxylation by RgF8H. (A) Mass spectrum of luteolin (B) 8-hydroxyluteolin. Figure S7. LC–MS analysis of apigenin hydroxylation by RgF8H. (A) Mass spectrum of apigenin, (B) 8-hydroxyapigenin. Figure S8. LC–MS analysis of chrysin hydroxylation by RgF8H. (A) Mass spectrum of chrysin, (B) 8-hydroxychrysin. Figure S9. LC–MS analysis of diosmetin hydroxylation by RgF8H. (A) Mass spectrum of diosmetin, (B) 8-hydroxydiosmetin. Figure S10. LC–MS analysis of 7,4ʹ-dihydroxyflavone hydroxylation by RgF8H. (A) Mass spectrum of 7,4ʹ-dihydroxyflavone, (B) 7,8,4ʹ-trihydroxyflavone.

The presence of naturally-occurring color mutants in plants has permitted the identification of many regulatory genes implicated in the synthesis of discrete metabolic compounds, mostly anthocyanins and carotenoids. Conversely, transcription factors that coordinate more than one specialized metabolic pathway seem challenging to screen from a forward genetics’ perspective. We explored the relationship between different branches of the phenylpropanoid and isoprenoid pathways while examining an infrequent berry skin color variegation in grapevine. Red and white berry skin sections were compared at the genetic, transcriptomic and metabolomic levels showing that, as in most cultivated white grape varieties, the uncolored skin section convened the non-functional alleles of the anthocyanin regulators MYBA1 and MYBA2, explaining the lack of pigments. In contrast, light-responsive flavonols and monoterpenes increased in anthocyanin-depleted areas. We disclosed an enrichment of the flavonol, terpene and carotenoid pathways among up-regulated genes from white-skin sections, accompanied by increased expressions of flavonol regulators and the still uncharacterized MYB24 gene. We used DAP-seq to examine the in vitro binding of affinity-purified MYB24 protein to genomic DNA and demonstrated its binding in the promoter regions of terpene (22) and carotenoid (6) genes, in addition to more than 30 photosynthesis and light-response genes, including the flavonol-regulator HY5 homologue (HYH). We confirmed the activation of TPS35 and HYH promoter:luciferase reporters in the presence of MYB24 and the grape bHLH MYC2, all of which correlate in their higher expression in white skin variegated sections. The integration of several datasets allowed to define a list of high confidence targets, suggesting MYB24 as a modulator of light responses including the synthesis of flavonoids (flavonols) and isoprenoids (terpenes, and putatively carotenoids). The correspondence between MYB24 and monoterpenes in all conditions surveyed implies that this regulatory network is broadly triggered towards berry ripening, and that the absence of anthocyanin sunscreens accelerates its activation most likely in a dose-dependent manner due to increased radiation exposure.