The tomato DWARF enzyme catalyses C-6 oxidation in brassinosteroid biosynthesis

Abstract

Brassinosteroids (BRs) are steroidal plant hormones essential for normal plant growth and development. Mutants in the biosynthesis or perception of BRs are usually dwarf. The tomato Dwarf gene (D), which was predicted to encode a cytochrome P450 enzyme (P450) with homology to other P450s involved in BR biosynthesis, was cloned previously. Here, we show that DWARF catalyses the C-6 oxidation of 6-deoxocastasterone (6-deoxoCS) to castasterone (CS), the immediate precursor of brassinolide. To do this, we first confirmed that the D cDNA complemented the mutant light- and dark-grown phenotypes of the extreme dwarf (dx) allele of tomato. To identify a substrate for the DWARF enzyme, exogenous application of BR intermediates to dx plants was carried out. C-6 oxoBR intermediates enhanced hypocotyl elongation whereas the C-6 deoxoBR, 6-deoxoCS, had little effect. Quantitative analysis of endogenous BR levels in tomato showed mainly the presence of 6-deoxoBRs. Furthermore, dx plants were found to lack CS and had a high level of 6-deoxoCS in comparison to D plants that had CS and a lower level of 6-deoxoCS. Confirmation that DWARF catalyzed the C-6 oxidation of 6-deoxoCS to CS was obtained by functional expression of DWARF in yeast. In these experiments, the intermediate 6alpha-hydroxycastasterone was identified, indicating that DWARF catalyzes two steps in BR biosynthesis. These data show that DWARF is involved in the C-6 oxidation in BR biosynthesis.

Figure 1

Growth responses of the wild type (D), the d^x mutant, and the 35S∷D line (a d^x-complementing line) under light/dark conditions in the presence or absence of BL. (A) Group of four light-grown seedlings (16 h/day) of D, d^x, and 35S∷D. In each group, the two on the left were grown on media only and the two on the right were grown on media containing 10⁻⁶ M brassinolide. (B) Same as A, but dark-grown. (C) Representative phenotypes of 2-month-old light-grown (16 h/day) plants of each line before BR analysis.

Figure 2

(A) Average height of plants (in millimeters) before harvesting for chemical analysis. (B) RNA gel blot analysis. Shown are D transcript levels in D, d^x, and 35S∷D plant material. No hybridization of D sequences to a D transcript was detected in d^x mRNA, suggesting this as a null allele. Low and high transcript accumulation was seen in the D and 35S∷D lines, respectively. 18S RNA hybridization is shown as RNA loading control.

Figure 3

(A) Dose–response curves showing percentage increase in hypocotyl length of d^x (solid line) and D plants (dashed line) for various BRs. Each point is mean ± SEM; n = 6. CT, cathasterone; TE, teasterone; 3DT, 3-dehydroteasterone; TY, typhasterol. (B) Dose–response curves showing increase of hypocotyl length of d^x and D plants for 6-deoxoCS (solid line), CS (small dashed line) and BL (long dashed line). Each point is mean ± SEM; n = 4–7.

Figure 4

Brassinolide biosynthetic pathway and brassinosteroid content of tomato lines. Brassinosteroid amounts, in nanograms per kilogram fresh weight, of D, d^x, and 35S∷D plants are shown in white, gray, and black boxes, respectively. nd, not detected, i.e., the BR concentration was below the detection limit.

Figure 5

Two-step oxidation of 6-deoxocastasterone catalyzed by DWARF (D).