ODORANT1 regulates fragrance biosynthesis in petunia flowers

Abstract

Floral scent is important to plant reproduction because it attracts pollinators to the sexual organs. Therefore, volatile emission is usually tuned to the foraging activity of the pollinators. In Petunia hybrida, volatile benzenoids determine the floral aroma. Although the pathways for benzenoid biosynthesis have been characterized, the enzymes involved are less well understood. How production and emission are regulated is unknown. By targeted transcriptome analyses, we identified ODORANT1 (ODO1), a member of the R2R3-type MYB family, as a candidate for the regulation of volatile benzenoids in Petunia hybrida cv W115 (Mitchell) flowers. These flowers are only fragrant in the evening and at night. Transcript levels of ODO1 increased before the onset of volatile emission and decreased when volatile emission declined. Downregulation of ODO1 in transgenic P. hybrida Mitchell plants strongly reduced volatile benzenoid levels through decreased synthesis of precursors from the shikimate pathway. The transcript levels of several genes in this pathway were reduced by suppression of ODO1 expression. Moreover, ODO1 could activate the promoter of the 5-enol-pyruvylshikimate-3-phosphate synthase gene. Flower pigmentation, which is furnished from the same shikimate precursors, was not influenced because color and scent biosynthesis occur at different developmental stages. Our studies identify ODO1 as a key regulator of floral scent biosynthesis.

Figure 1.

Petal-Specific Expression of ODO1 Correlates with Scent Emission. Light (L) and dark (D) periods are indicated at the top of the figure. Hybridization with 18S rRNA is shown to illustrate the loading of the gels. These experiments ([A] to [D]) were performed with plants from growth chambers. (A) RNA gel blot analyses of ODO1 in Mitchell petals harvested at 3-h intervals. The asterisks indicate the two transcripts that were detected. (B) Bar graph depicting the emission of three selected volatile benzenoids measured for 1 h around the same time points (mean and se, n = 3). (C) Organ- and tissue-specific expression. R, roots; S, stems; L, leaves; Sp, sepals; PT, petal tube; PL, petal limb; A, anthers; St, stigma. (D) RNA gel blot analysis of ODO1 in Mitchell (M) and W138 at 1800 h.

Figure 2.

ODO1 Belongs to a Distinct Subgroup of MYB Proteins. Phylogenetic tree of ODO1 and MYB proteins from various plant species. The tree was rooted to human (Hs) C-Myb. Bootstrap values are indicated at branch nodes. PhODO1 is boxed. Known functions of several MYBs are indicated.

Figure 3.

Alignment of Amino Acid Sequences of the R2R3 Domains of P. hybrida ODO1, PbMYB from P. brachicarpa, AtMYB42 and AtMYB85 from Arabidopsis, and Four MYBs from P. hybrida. Conserved amino acids are highlighted in dark gray; conserved residues in the hypervariable regions in ODO1, PbMYB, AtMYB42, and AtMYB85 are highlighted in light gray and boxed.

Figure 4.

Suppression of ODO1 Expression Reduces Volatile Benzenoid Emission. (A) Gas chromatography time-of-flight mass spectrometry chromatograms of volatiles emitted by Mitchell and RNAi line 3. (B) Quantified emission of volatile benzenoids by Mitchell (M), four RNAi lines that show reduced emission of volatiles (1, 3, 12, and 35), and one control RNAi line (40). Bars annotated with different letters indicate significant differences among lines (analysis of variance, P < 0.05 according to least significant difference post-hoc analysis; n ≥ 4). All experiments were performed with greenhouse-grown plants.

Figure 5.

Modulation of ODO1 Expression Alters Expression of Volatile Benzenoid Genes. (A) RNA gel blot analysis of Mitchell (M) and RNAi lines 1, 3, 12, 35, and 40 (control) at 1700 h for ODO1, DAHPS, EPSP synthase (EPSPS), CM, two PAL genes (PAL1 and 2), BSMT, BPBT, and SAM-synthase (SAMS). Transcript levels of FLORAL BINDING PROTEIN 1 (FBP1), a MADS box protein involved in specification of floral organ identity (Angenent et al., 1992), are shown to indicate the loading of the gels. (B) Schematic representation of the shikimate pathway that leads to the biosynthesis of Phe and its derivatives trans-cinnamic acid, coumaric acid, caffeic acid, and ferrulic acid. Solid arrows indicate well-described enzymatic reactions, whereas broken arrows indicate routes that are still hypothetical. E-4P, erythrose 4-phosphate; PEP, phosphoenolpyruvate. Numbers indicate the following enzymes: 1, DAHP synthase; 2, EPSP synthase; 3, CM; 4, PAL; 5, chalcone synthase. (C) Activation assays. The columns and error bars denote the mean and standard error of the activity of the EPSPS-GUS and DFR-GUS reporter constructs after bombardment with 35S-ODO1 or a mix of 35S-AN1 and 35S-AN2. Reporter gene activity, measured as glucuronidase enzyme activity, is expressed in relative arbitrary units and was normalized to luciferase enzyme activity (LUC) expressed from a cobombarded reference construct, which contained the luciferase gene driven by the 35S promoter.

Figure 6.

Flavonol and Anthocyanin Biosynthesis Is Not Influenced by Suppression of ODO1 Expression. (A) Section through flowers from Mitchell and RNAi line 3, showing the characteristic purple, anthocyanin-containing stripes in the tube. (B) Thin layer chromatography of flavonols in petal limbs from Mitchell (M) and RNAi line 3, showing that the expected flavonols (dihydrokaempferol, dihydroquercetin, and dihydromyrcetin) are present in both. The silica gel-TLC F254 plate was photographed under UV light (254 nm). Two marker flavonols, kaempferol (K) and quercetin (Q), are indicated. (C) RNA gel blot analysis of ODO1 in developing Mitchell flowers. Fl, flower; CHS, chalcone synthase. Hybridization with 18S rRNA is shown to illustrate the loading of the gels.

Figure 7.

ODO1 Regulates Volatile Benzenoids in the Violet V26. (A) Quantified emission of phenylacetaldehyde and isoeugenol by V26, Mitchell (M), the F1 progeny of a cross between Mitchell and V26 and a cross between RNAi line 3 and V26. An asterisk denotes a significant difference between Mitchell × V26 and RNAi line 3 × V26 at P < 0.05 (Student's t test; n = 4). (B) RNA gel blot analysis for ODO1 of the lines and crosses described in (A). Transcript levels of FLORAL BINDING PROTEIN 1 (FBP1) are shown to illustrate the loading of the gel.