Catabolism of brassinosteroids regulates the endogenous level of bioactive brassinosteroids. In Arabidopsis thaliana, bioactive brassinosteroids such as castasterone (CS) and brassinolide (BL) are inactivated mainly by two cytochrome P450 monooxygenases, CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1; CYP734A1/BAS1 inactivates CS and BL by means of C-26 hydroxylation. Here, we characterized CYP734A orthologs from Oryza sativa (rice). Overexpression of rice CYP734As in transgenic rice gave typical brassinosteroid-deficient phenotypes. These transformants were deficient in both the bioactive CS and its precursors downstream of the C-22 hydroxylation step. Consistent with this result, recombinant rice CYP734As utilized a range of C-22 hydroxylated brassinosteroid intermediates as substrates. In addition, rice CYP734As can catalyze hydroxylation and the second and third oxidations to produce aldehyde and carboxylate groups at C-26 in vitro. These results indicate that rice CYP734As are multifunctional, multisubstrate enzymes that control the endogenous bioactive brassinosteroid content both by direct inactivation of CS and by the suppression of CS biosynthesis by decreasing the levels of brassinosteroid precursors.
Brassinosteroids are endogenous phytohormones that are involved in the regulation of various growth and developmental processes in higher plants, such as cell and stem elongation, dark-adapted morphogenesis (skotomorphogenesis), responses to environmental stress, and differentiation of tracheary elements (reviewed in Clouse and Sasse, 1998; Krishna, 2003; Sasse, 2003). Brassinosteroids are believed to enhance crop production. In Oryza sativa (rice), for example, brassinosteroids induce disease resistance (Nakashita et al., 2003) and abiotic stress tolerance (Koh et al., 2007), and, interestingly, overproduction and deficiency of brassinosteroids both increase grain yield, although by different mechanisms (Morinaka et al., 2006; Sakamoto et al., 2006; Wu et al., 2008). These findings suggest the feasibility of genetic improvement of crop production by the modulation of brassinosteroid metabolism and the contents of bioactive brassinosteroids.
The major pathway for brassinosteroid biosynthesis has been established in Arabidopsis thaliana, and a number of dwarf mutants have been identified (reviewed in Bishop, 2007). Among the six major enzymes involved in brassinosteroid biosynthesis, four are cytochrome P450 monooxygenases (P450s): in rice, C-22 hydroxylase is encoded by CYP90B2/OsDWARF4 and CYP724B1/D11 (Sakamoto et al., 2006); C-23 hydroxylase is encoded by CYP90D2/D2 and CYP90D3 (TS and MM, unpublished results); and C-6 oxidase is encoded by CYP85A1/OsDWARF (Hong et al., 2002; Mori et al., 2002). CYP90As, encoded by CYP90A3/OsCPD1 and CYP90A4/OsCPD2, are also believed to be involved in brassinosteroid biosynthesis in rice, although their catalytic function has not been clarified (Sakamoto and Matsuoka, 2006).
Catabolism of brassinosteroids is another important factor that regulates the endogenous levels of bioactive brassinosteroids. In A. thaliana, inactivation of brassinosteroids is also catalyzed by P450s, and two genes, CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1, have been identified (Neff et al., 1999; Turk et al., 2003, 2005; Nakamura et al., 2005; Takahashi et al., 2005). The biochemical function of CYP72C1/SOB7/CHI2/SHK1 has not yet been clarified; however, CYP734A1/BAS1 inactivates castasterone (CS) and brassinolide (BL) by means of C-26 hydroxylation (Turk et al., 2003). CYP734A1/BAS1 orthologs have been identified in Solanum lycopersicum (tomato), in which at least one of two genes, CYP734A7, encodes a C-26 hydroxylase of brassinosteroids and is probably involved in brassinosteroid catabolism (Ohnishi et al., 2006a).
By using the genes for these brassinosteroid biosynthetic or catabolic enzymes in breeding programs, it should become possible to control the level of bioactive brassinosteroids and thereby increase crop productivity. We attempted to identify rice orthologs of CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1 to enhance our understanding of brassinosteroid metabolism in rice. On the basis of our results, we discuss the evolution of the genes for brassinosteroid catabolic enzymes in rice.
Isolation of CYP734A1/BAS1-like genes from rice
The predicted amino acid sequences of CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1 were used as in silico probes to screen all available rice DNA databases. Candidate sequences detected during this process were also used iteratively as search probes. We found four candidate genes that had relatively high identity with the amino acid sequences of CYP734A1/BAS: CYP734A2 (locus ID Os02g0204700), CYP734A4 (Os06g0600400), CYP734A5 (Os07g0647200), and CYP734A6 (Os01g0388000). We found no orthologs of CYP72C1/SOB7/CHI2/SHK1 in the rice genome. The deduced amino acid sequences of CYP734A2, CYP734A4, and CYP734A6 were very similar (ranging between 70.2 and 82.1% identity), whereas they showed between 58.0 and 58.9% identity with CYP734A5 (Table S1). CYP734A1/BAS1 showed the highest identity with CYP734A6 (62.0%), and lower identity with CYP734A5 (50.5%; Table S1). All four rice CYP734As and CYP734A1/BAS1 showed lower identities with CYP72C1/SOB7/CHI2/SHK1 (ranging between 34.9 and 37.6%; Table S1).
Overexpression of CYP734A genes in transgenic rice
To assess the activity of the CYP734A gene products in rice, the entire coding regions for the rice CYP734As were amplified by means of RT-PCR using total RNA extracted from whole seedlings. Because we could not obtain the full-length cDNA for CYP734A5 by RT-PCR, we ectopically expressed the other three cDNAs in transgenic rice under the control of the rice actin1 promoter (McElroy et al., 1990).
All transformants overexpressing the CYP734A cDNA exhibited abnormal phenotypes, and we categorized them into two groups on the basis of leaf morphology and gross morphology. The control transformants that contained the empty vector were indistinguishable from wild-type rice plants. They flowered about 90 days after the regeneration and the plant height reached about 90 cm (Figure 1a). Plants exhibiting a group-1 phenotype had severe dwarfing (about 10 cm in height) and malformed leaves with twisted leaf blades (Figure 1b). The leaves of these plants were erect, and the ratio of leaf sheath length to leaf blade length was reduced (0.209 ± 0.002, versus 0.601 ± 0.011 in the wild type). The floral organs did not form and internodes did not elongate in the group-1 phenotype plants, and they did not bear any seeds. These phenotypes were indistinguishable from those of the brassinosteroid-deficient mutant brd1-2 (Figure 1c; Hong et al., 2002). Transformants categorized as group-2 phenotype formed only abnormal leaves with stiff, tortuous blades, and their leaf sheaths were scarcely developed (Figure 1d, right plant). These plants did not flower, exhibit internode elongation, or bear seeds. These phenotypes were indistinguishable from those of the brassinosteroid-insensitive mutant d61-3 (Figure 1d, left plant; Nakamura et al., 2006). About 90% (40 plants) of the transformants overexpressing CYP734A2 exhibited the group-2 phenotype and the remaining 10% (five plants) had the group-1 phenotype. About 55% (23 plants) of the CYP734A6 transformants exhibited the group-2 phenotype and the remaining 45% (19 plants) had the group-1 phenotype. In contrast to these two genes, all 38 transformants that overexpressed CYP734A4 exhibited the group-2 phenotype. These results strongly suggest that overexpression of CYP734A2, CYP734A4, and CYP734A6 cDNA decreased the level of bioactive brassinosteroids in the transgenic rice.
We compared the endogenous levels of sterols and brassinosteroids in the control transformants and CYP734A transformants having the group-2 phenotype by using gas chromatography–mass spectrometry (GC-MS) analyses (Table 1). Cathasterone (CT) and BL were not detected in either the control or the CYP734A transformants, suggesting that these compounds are minor components of the total brassinosteroid pool in rice. Although BL was not detected, another bioactive brassinosteroid (i.e. CS) was detected in the control, but not in the CYP734A4 transformants, confirming that the CYP734A4 transformants are brassinosteroid-deficient. Levels of the other eight intermediates downstream of C-22 hydroxylation – 6-deoxocathasterone (6-deoxoCT), 6-deoxoteasterone (6-deoxoTE), teasterone (TE), 3-dehydro-6-deoxoteasterone (6-deoxo3DT), 3-dehydroteasterone (3DT), 6-deoxotyphasterol (6-deoxoTY), typhasterol (TY), and 6-deoxocastasterone (6-deoxoCS) – were also greatly reduced in the CYP734A4 transformants. Similar results were observed in the CYP734A2 and CYP734A6 transformants. These results suggest that overexpression of CYP734A2, CYP734A4, and CYP734A6 increases the inactivation of C-22 hydroxylated brassinosteroids in transgenic rice.
Table 1. Endogenous content of sterols and brassinosteroids in transgenic rice
Sterol or brassinosteroid intermediate
Content (ng g−1 fresh weight)
n.d., not detected.
Catalytic activities of rice CYP734As
To determine the catalytic function of the rice CYP734A gene products, we performed in vitro conversion assays using the recombinant CYP734A proteins produced in insect cells by a baculovirus expression system. Brassinosteroid intermediates with a diol at the C-22 and C-23 positions of the side chain were used for the assays, in which the reaction products were converted into 9-phenanthreneboronate derivatives and analyzed by means of HPLC. In the presence of NADPH, CYP734A2 metabolized 6-deoxo3DT into three new products with retention times (rt) of 6.5, 8.3, and 9.0 min (Figure 2a); these products were not identical to the compounds previously found in the brassinosteroid biosynthetic pathway. Analysis of the products by means of fast-atom-bombardment mass spectrometry (FAB-MS) revealed that the product at rt 6.5 min was 29.97 mass units larger than that of the substrate 6-deoxo3DT; this is consistent with the theoretical mass of the carboxylate form of 6-deoxo3DT (Table 2). On the other hand, the observed mass of the product at rt 8.3 min was 15.99 mass units larger, which was consistent with the theoretical mass of the hydroxylate form of 6-deoxo3DT. These results suggested that CYP734A2 is a multifunctional P450 that catalyzes a three-step successive oxidation of 6-deoxo3DT to produce an alcohol, an aldehyde, and a carboxylate group. Similarly, three products were detected in the assays for CYP734A2 with all the 22,23-hydroxylated compounds in the brassinosteroid biosynthetic pathway, including CS and BL (Figure S1).
Table 2. Fast-atom-bombardment mass spectrometry (FAB-MS) data for the reaction products with 6-deoxo3DT as the substrate and CYP734A2 as the enzyme
Error (p.p.m. per mmu)
mmu, millimass units.
C42 H55 O3 B
C42 H55 O4 B
C28 H46 O4 Na
C42 H53 O5 B
In the case of the brassinosteroid compounds without the diol side chain, the reaction products were converted into trimethylsilyl derivatives and analyzed by means of GC-MS. In the assay with (22S,24R)-22-hydroxy-5α-ergostan-3-one (22-OH-3-one), three products (with retention times of 12.06, 13.38, and 14.48 min) were detected (Figure 3). The C20–C22 fission fragment ion of the derivative of the substrate 22-OH-3-one had a mass-to-charge ratio (m/z) of 187, whereas those of the product derivatives were predicted to be m/z 201, 185, and 289, respectively, but there was no predicted fragment peak with an m/z of 187 (Table 3). These results suggested that one of the methyl groups at C-26, C-27, and C-28 in the side chain is consecutively oxidized into a carboxylated form. The other 22-hydroxylated brassinosteroid compounds were similarly metabolized into three products by CYP734A2 (Figure S2). Campesterol (CR, Figure S2), campest-4-en-3-one, campest-3-one, and campestanol (CN) were not metabolized at all, suggesting that the presence of the hydroxyl group at C-22 is crucial for recognition of the substrate by CYP734A2. These results indicated that CYP734A2 is a multifunctional and multisubstrate enzyme capable of metabolizing C-22 hydroxylated compounds in the brassinosteroid biosynthetic pathway.
Table 3. Gas chromatography-MS data for the products obtained in the CYP734A2 assay
We also performed conversion assays with CYP734A4 and CYP734A6 under the same conditions. We found that the CYP734A4 assays gave results similar to those of the CYP734A2 assays for all the substrates we tested. Interestingly, CYP734A6 acted on a similar array of substrates, but only two products were detected in each of the conversion assays. For instance, CYP734A6 metabolized 6-deoxo3DT into two products, which corresponded to the oxidized products in alcohol and aldehyde forms found in the CYP734A2 assay (Figure 2b). No carboxylated product was detected even after extended incubation, suggesting that CYP734A6 can catalyze only a two-step oxidation to produce an aldehyde form.
Determination of the position oxidized by CYP734A2
To determine the specific position of the hydroxyl group introduced by CYP734A2, we performed a conversion assay with CS as a substrate. Castasterone was metabolized into three products, and one of the products, which corresponded to hydroxylated CS, was identical to the authentic 26-hydroxyCS (26-OHCS) in its GC retention time and mass spectrum (Figure S3).
Next, d7-6-deoxoCS labeled with deuterium atoms at the C-25, C-26, and C-27 positions was applied as a substrate. CYP734A2 metabolized d7-6-deoxoCS into three products, and these metabolites were separated and collected by means of HPLC. The metabolites corresponding to the aldehyde and carboxylate forms were found to have m/z values of 469.38 and 507.36, respectively, in the high-resolution mass spectra (HRMS) analysis (Table 4). The observed values were consistent with the theoretical m/z values of d5-6-deoxoCS-CHO and d4-6-deoxoCS-COOH, respectively. These results suggest that two or three deuterium atoms in either the C-26 or the C-27 methyl group were displaced by successive oxidation. Taken together, these results suggest that CYP734A2 catalyzes the oxidation of brassinosteroids to produce C-26 oxidized metabolites.
Table 4. High-resolution mass spectra data for the reaction products with d7-6-deoxoCS as the substrate and CYP734A2 as the enzyme
Error (p.p.m. per mmu)
mmu, millimass units; n.d., not detected.
C42 1H50 D7 O4 B
C28 1H43 D5 O5
C28 1H44 D4 O6 Na
Substrate specificity of CYP734A2
The present results clearly demonstrated that, in contrast to CYP734A1/BAS1 and tomato CYP734A7, which selectively metabolize CS and BL, rice CYP734A2 could metabolize various brassinosteroid intermediates. Thus, we determined the substrate specificity of CYP734A2 by determining the relative activities for the brassinosteroid intermediates with a diol at the C-22 and C-23 positions (Table 5). Among the compounds that we tested, 6-deoxo3DT provided the best match for CYP734A2, with another early intermediate, 22,23-dihydroxycampesterol, also preferred. On the other hand, CS and BL were not highly suitable substrates for CYP734A2 (with relative activities of 39 and 16%, respectively). These results suggested that rice CYP734As metabolize endogenous brassinosteroid intermediates during the early steps of the brassinosteroid biosynthetic pathway. This hypothesis is consistent with the content of the endogenous brassinosteroid intermediates that we observed in the transgenic rice that overproduced CYP734As.
Table 5. Substrate specificity of CYP734A2
Relative activity (%)a
aThe value obtained using 6-deoxo-3-dehydroteasterone as the substrate was arbitrarily set at 100.
Expression of CYP734A genes in wild-type and mutant rice
Quantitative reverse-transcriptase PCR (qRT-PCR) analysis revealed that all four rice CYP734A genes were expressed at different levels in different organs (Figure 4). CYP734A2, CYP734A4, and CYP734A6 were expressed in all the organs of wild-type rice that we tested, including the vegetative shoot apices, leaf sheaths, leaf blades, elongating internodes, roots, and panicles at flowering time, whereas CYP734A5 expression was not observed in the leaves or internodes, was present at nearly undetectable levels in the shoot apices, leaf sheaths, and panicles, and was expressed at the highest level in the roots.
Previous observations in A. thaliana indicate that the expression of CYP734A1/BAS1 was up-regulated by the application of brassinosteroid (Goda et al., 2002; Tanaka et al., 2005). Thus, we examined whether such feedback regulation also occurs in rice. We also determined the expression of genes for the brassinosteroid biosynthetic enzyme (OsDWARF; Hong et al., 2002; Mori et al., 2002) and for the brassinosteroid receptor (OsBRI1; Yamamuro et al., 2000) as controls. Quantitative RT-PCR analysis revealed that the expression of all four rice CYP734As was rapidly increased by BL treatment (Figure 5a), whereas it was down-regulated in brassinosteroid-deficient brd1-2 (a loss-of-function mutant of OsDWARF; Figure 5b). The level of CYP734A2 transcripts decreased in brassinosteroid-insensitive d61-3 (a loss-of-function mutant of OsBRI1), but the expression of the other three CYP734A genes increased in the d61-3 mutant.
Genetic evidence has uncovered a role for brassinosteroids in the control of skotomorphogenesis, the process by which etiolated seedlings rapidly grow an exaggerated hypocotyl in the dark so they can reach the soil surface. In previous research, brassinosteroid-deficient or brassinosteroid-insensitive mutants did not show an elongated phenotype and developed similarly to plants grown in the light (Clouse et al., 1996; Szekeres et al., 1996). Expression of CYP734A2, CYP734A4, and CYP734A5 was up-regulated in dark-grown seedlings, whereas the level of CYP734A6 transcripts decreased (Figure 5b).
The concentration of bioactive phytohormones is tightly controlled by the regulation of both their biosynthesis and their catabolism. Therefore, characterization of catabolic enzymes is an important part of understanding phytohormone metabolism. Inactivation of bioactive brassinosteroids such as CS and BL, as well as their biosynthetic precursors, occurs through various reaction processes, including epimerization of 2- and 3-hydroxy groups followed by glucosylation or acylation, hydroxylation of C-20 and successive side-chain cleavages, glucosylation of the C-23 hydroxy group, and hydroxylation of C-25 or C-26 followed by glucosylation (Bajguz, 2007). In A. thaliana, a UDP-glucosyltransferase (UGT73C5) catalyzes 23-O-glucosylation of CS and BL (Poppenberger et al., 2005). Two P450s, CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1, also act in the inactivation of brassinosteroids by means of different enzymatic activities (Turk et al., 2005). Biochemical analyses revealed that CYP734A1/BAS1 is a C-26 hydroxylase that utilizes CS and BL as substrates.
Our results showed that there are four rice CYP734A genes, but there appears to be no CYP72C1/SOB7/CHI2/SHK1 ortholog in the rice genome. It is noteworthy that at least three rice CYP734As (CYP734A2, CYP734A4, and CYP734A6) utilized a broad range of C-22 hydroxylated brassinosteroid intermediates as substrates (summarized in Figure 6). In addition, CYP734A2 catalyzed the oxidation of 22-hydroxy-cholesterol, suggesting that rice CYP734As can metabolize both C27-brassinosteroids and C28-brassinosteroids. Consistent with this result, the CYP734A-overproducing rice transformants showed the greatest decrease in levels of the endogenous brassinosteroid intermediates downstream of the C-22 hydroxylation step. These results suggest that rice CYP734As control the endogenous bioactive brassinosteroid content not only by direct inactivation of CS by C-26 oxidation (as in A. thaliana), but also by the suppression of CS biosynthesis through decreases in CS precursors.
Interestingly, some rice CYP734As (CYP734A2 and CYP734A4) can catalyze not only the hydroxylation of C-22 hydroxylated brassinosteroids but also the second and third oxidations to produce an aldehyde and a carboxylate group at C-26 in vitro (Figure 6). Although hydroxylation at C-25 and C-26 and subsequent glucosylation were observed in 24-epi-brassinolide exogenously applied to cultured tomato cells (Hai et al., 1995; Winter et al., 1997), the metabolic processes that follow the successive oxidations are unknown. Previously, demethylation of BL at C-26 was observed when BL was applied to Marchantia polymorpha cells (Kim et al., 2000), and a cell-free tomato solution also converted CS into 26-norCS by means of C-26 demethylation (Kim et al., 2004). It seems that oxidation and subsequent decarboxylation at C-26 should occur before demethylation. In our conversion assays, we did not detect any demethylated product. This indicates that rice CYP734As do not catalyze C-26 demethylation; however, our results do not rule out the possibility that other enzymes are involved in decarboxylation in rice. Because of technical difficulties, we could not confirm the unique catalytic activities of rice CYP734As in vivo. However, we believe that further analyses will support the conclusion that we reached from our in vitro analyses.
Our results indicate that rice CYP734As are multifunctional and multisubstrate P450s, which catalyze the consecutive oxidation of various C-22 hydroxylated brassinosteroids. There are some examples of multifunctional and/or multisubstrate P450s in the primary and secondary metabolism, such as sterol C-14 demethylase (CYP51G) in sterol biosynthesis (Kahn et al., 1996), ent-kaurene oxidase (CYP701A) and ent-kaurenoic acid 7α-hydroxylase (CYP88A) in gibberellin biosynthesis (Helliwell et al., 1998, 2001), CYP79s catalyzing successive N-oxidations of amino acids to form aldoximes (Bak et al., 2006), and abietadienol/abietadienal oxidase (CYP720B1) in diterpene resin acid biosynthesis (Ro et al., 2005). The consensus amino acid sequence motif (A/G)GX(D/E)TT in the I-helix of most P450s are known to be involved in the substrate interaction and the oxygen activation, and in particular, the conserved Asp/Glu and Thr residues in the motif play a crucial role in the hydrogen-bond network that enables the protonation of the distal O2 to promote the heterolytic scission of the O–O bond. Some P450s catalyzing unusual reactions have a unique substitution in the conserved motif (Mizutani and Sato, 2011). In rice CYP734A2, CYP734A4, and CYP734A6, the conserved Asp/Glu residue is replaced by glutamine, suggesting that this amino acid substitution may be crucial for CYP734A activity. In the brassinosteroid biosynthetic pathway, hydroxylation at C-22 and C-23 positions occurs at several intermediate steps, and CYP90B, CYP90C, CYP90D, and CYP724B are multisubstrate enzymes towards brassinosteroid intermediates (Fujita et al., 2006; Ohnishi et al., 2006b,c). Furthermore, Arabidopsis CYP85A2 catalyzes the C-6 oxidation and lactonization reactions of teasterone, typhasterol, and castasterone (Katsumata et al., 2008). Thus the brassinosteroid biosynthetic pathway constitutes a complicated metabolic grid, which provides a broad range of potential bioactive brassinosteroids. The multifunctional and multisubstrate properties of rice CYP734As will enable the regulation of the endogenous content of bioactive brassinosteroids.
The CYP734A subfamily is found in various plant species, including gymnosperms and angiosperms, but not mosses and ferns. This indicates that the CYP734A subfamily is widely distributed among vascular plants. Phylogenetic analysis revealed that CYP734As are divided into different clades between monocot and dicot species (Figure 7). Duplication of CYP734A has also occurred in tomato, but tomato CYP734As function as a C-26 hydroxylase that utilizes CS and BL as substrates, as in the case of CYP734A1/BASI (Ohnishi et al., 2006a). These results suggest that the rice CYP734As acquired their novel function in brassinosteroid inactivation after the differentiation of monocots and dicots, and then increased their number in the rice genome. Three CYP734A genes in Zea mays (CYP734A-Zm1, CYP734A-Zm2, and CYP734A-Zm3) are clustered with rice CYP734A6, CYP734A4, and CYP734A2, respectively (Figure 7). This suggests that CYP734As were duplicated before speciation in the monocots. In contrast to this finding, the CYP85A gene has been duplicated in A. thaliana and tomato but not in rice. The rice CYP85A enzyme catalyzes the C-6 oxidation of 6-deoxoCS to produce CS, a bioactive brassinosteroid (Hong et al., 2002; Kim et al., 2008), whereas one of the two enzymes in A. thaliana and tomato, AtCYP85A2 and LeCYP85A3, has evolved to catalyze a Baeyer–Villiger oxidation at C-6 to form BL, the most biologically active brassinosteroid, from 6-deoxoCS through CS (Kim et al., 2005; Nomura et al., 2005). These findings suggest that divergent evolutionary processes have occurred in brassinosteroid metabolism.
Expression of all four CYP734A genes was up-regulated by the BL treatment and down-regulated in the brassinosteroid-deficient mutant brd1-2, suggesting that the levels of endogenous bioactive brassinosteroids control the expression of these genes, as was the case for CYP734A1/BAS1 in A. thaliana. However, the levels of CYP734A4, CYP734A5, and CYP734A6 transcripts increased in d61-3 (a loss-of-function mutant of the rice brassinosteroid receptor gene, OsBRI1). In d61-3, lack of the brassinosteroid signal stimulates feedback up-regulation of the brassinosteroid biosynthetic enzyme gene, OsDWARF (Figure 5b), resulting in accumulation of bioactive CS in the OsBRI1 mutant (Nakamura et al., 2006). In the same context, lack of the brassinosteroid signal should stimulate down-regulation of brassinosteroid catabolic enzyme gene expression. Indeed, CYP734A2 transcripts decreased in d61-3. However, the expression levels of CYP734A4, CYP734A5, and CYP734A6 were up-regulated in d61-3. One possible explanation for this phenomenon is that accumulated bioactive CS in d61-3 sensed by OsBRI1 paralogs (OsBRL1 and OsBRL3) stimulates up-regulation of these CYP734As, because several lines of evidence indicate that OsBRL1 and OsBRL3 are partly involved in the detection of brassinosteroids in rice (Nakamura et al., 2006).
Seedlings grown in complete darkness showed the etiolation and unusual internode elongation known as skotomorphogenesis. In contrast, rice mutants with deficiencies in brassinosteroid biosynthesis or detection had a de-etiolated phenotype without internode elongation (Yamamuro et al., 2000; Hong et al., 2002, 2003). These findings indicate that the brassinosteroid-related rice mutants have a deficiency in skotomorphogenesis similar to that reported in A. thaliana. In the dark-grown seedlings, expression of OsDWARF and OsBRI1 was up-regulated (Figure 5b), suggesting that biosynthesis and sensitivity to brassinosteroids both increase in rice plants under complete darkness. This is a likely explanation for the up-regulation of CYP734A2, CYP734A4, and CYP734A5 expression. However, the level of CYP734A6 transcripts decreased in dark-grown seedlings even when they were treated with exogenous BL. Interestingly, the expression level of CYP734A1/BAS1 was slightly reduced in dark-grown A. thaliana seedlings (Turk et al., 2005). These results suggest that expression of CYP734A1/BAS1 and CYP734A6 is regulated in a light-dependent manner. Because phytochrome B acts as a negative regulator of brassinosteroid-regulated growth and development processes in rice (Jeong et al., 2007), regulation of CYP734A6 transcription may occur downstream of phytochrome B signal transduction.
Isolation of rice CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1 orthologs
We performed a blast search using the predicted amino acid sequences encoded by CYP734A1/BAS1 and CYP72C1/SOB7/CHI2/SHK1 as probes against the rice DNA databases, using the methods of Sakamoto et al. (2004). The entire coding region for the CYP734A genes was amplified by means of RT-PCR using total RNA extracted from whole seedlings (O. sativa L. ‘Nipponbare’). Reverse transcriptase PCR was performed with the oligonucleotide primers shown in Table S2. Amplified fragments were cloned into pBluescript II SK (Stratagene, http://www.stratagene.com/), and their sequences were determined.
A phylogenetic tree of CYP734As was reconstructed by the neighbor-joining (NJ) method (Saitou and Nei, 1987) on the basis of Kimura’s two-parameter distances (Kimura, 1980). phylip (Felsenstein, 1989) was used to perform the phylogenetic reconstruction. Bootstrap values were estimated (with 1000 replicates) to assess the relative support for each branch (Felsenstein, 1985). All positions containing alignment gaps were eliminated in pairwise sequence comparisons in the NJ analyses. The alignment used to produce the phylogeny is shown in Figure S4.
Plasmid constructs and plant transformation
The whole coding regions for CYP734A2, CYP734A4, and CYP734A6 were inserted between the rice actin1 promoter and the nopaline synthase polyadenylation signal of the hygromycin-resistant binary vector pAct-Hm2. This vector is modified from pBI-H1 (Ohta et al., 1990) so that it contains a rice actin promoter. The resulting construct was introduced into A. tumefaciens strain EHA105, and Agrobacterium-mediated transformation of rice (O. sativa L. ‘Nipponbare’) was performed as described by Hiei et al. (1994). Transgenic plants were selected on media containing 50 mg L−1 hygromycin.
Analysis of endogenous brassinosteroid levels
Shoots from control plants (created by means of transformation with an empty vector) and from transgenic rice that overexpressed CYP734A were harvested 4 weeks after regeneration. Brassinosteroids were then extracted, purified and quantified by the methods of Hong et al. (2003).
Heterologous expression in a baculovirus/insect cell system
Rice CYP734A cDNAs were cloned as BamHI–XhoI fragments in the pFastBac1 vector (Invitrogen, http://www.invitrogen.com/), and were then used to generate the corresponding recombinant Bacmid DNAs by transformation of Escherichia coli strain DH10Bac (Invitrogen). Preparation of the recombinant baculovirus DNAs that contained CYP734A cDNA and transfection of Sf9 (Spodoptera frugiperda 9) cells were carried out according to the instructions of the manufacturer (Invitrogen). Heterologous production of the CYP734A proteins in Sf9 cells and spectrophotometric analysis were carried out as described by Saito et al. (2004).
Microsomal fractions of the insect cells that expressed the CYP734As were obtained from the infected cells (300 ml of suspension-cultured cells). Infected cells were washed with phosphate-buffered saline buffer and suspended in buffer A, which consisted of 20 mm potassium phosphate (pH 7.25), 20% (v/v) glycerol, 1 mm EDTA, and 1 mm DTT. The cells were sonicated, and cell debris was removed by centrifugation at 10 000 g for 15 min. The supernatant was further centrifuged at 100 000 g for 1 h, and the pellet was homogenized with buffer A to provide the microsomal fractions. The microsomal fractions were stored at −80°C before the enzyme assay described in the next section.
Analysis of CYP734A activities
The activities of CYP734A were reconstituted by mixing each of the CYP734A-containing microsomes with purified A. thaliana NADPH-P450 reductase (Mizutani and Ohta, 1998). The reaction mixture consisted of 100 mm potassium phosphate (pH 7.25), 50 pmol ml−1 recombinant P450 protein, 0.1 unit per ml NADPH-P450 reductase, and 20 μm of brassinosteroids. Reactions were initiated by the addition of 1 mm NADPH, and were carried out at 30°C for 30 min. We also produced controls by using the same reaction mixture, but without NADPH (the ‘−NADPH’ treatment). The reaction products were extracted three times with half the original solution’s volume of ethyl acetate. The organic phase was collected and evaporated.
When brassinosteroids with a diol at the C-22 and C-23 positions of the side chain were applied to the assays, the reaction products were converted into 9-phenanthreneboronate derivatives and analyzed by means of high-performance liquid chromatography (HPLC) using the following procedure: The residue left behind after evaporation of the organic phase was treated with 1 mg ml−1 9-phenanthreneboronic acid in pyridine at 80°C for 30 min (Gamoh et al., 1989), and 20 μL of the sample was analyzed by using HPLC apparatus equipped with a SunFire C18 column (20 mm length × 4.6 mm internal diameter; Waters, http://www.waters.com/). Fluorescence detection was performed with excitation at a wavelength of 305 nm and emission at 350 nm. The column temperature was kept constant at 40°C, and the flow rate of the mobile phase was 1.0 ml per min. The binary gradient elution system consisted of distilled water (A) and acetonitrile (B) using the following gradient: 70% (B) from 0 to 2 min, 70 to 95% (B) from 2 to 5 min, 95% (B) from 5 to 15 min, and 70% (B) from 15 to 18 min. To determine the substrate specificities, we used the sum of the peak areas of the three catabolites to estimate the total activities of the three products. A JEOL JMS-700 was used to perform FAB-MS analysis of the reaction products.
For brassinosteroids without a diol side chain, the reaction products were converted into trimethylsilyl derivatives and analyzed by means of GC-MS, as follows. The residue left behind after evaporation of the organic phase was treated with 10 μl of N-methyl-N-trimethylsilyltrifluoroacetamide at 80°C for 30 min. The derivatized products were analyzed by means of GC-MS, as described by Fujita et al. (2006).
Substrate specificity of CYP734A2
The CYP734A2 assays (250 μl) were performed with various 22,23-hydroxylated brassinosteroids at a concentration of 20 μm. Reactions were initiated by the addition of NADPH and incubation at 30°C for 30 min. This incubation period afforded <20% consumption of each substrate. Reactions were terminated by the addition of 250 μl of ethyl acetate. The reaction products were extracted three times with 250 μl of ethyl acetate and converted to 9-phenanthreneboronate derivatives. The amounts of the reaction products were estimated by HPLC on the basis of the fluorescence of phenanthrene. Because CYP734A2 metabolized each of the 22,23-hydroxylated brassinosteroids into three oxidized products by successive oxidation, the value obtained from the sum of the peak areas of the three products was used to estimate the overall conversion rate of each substrate. Relative activity was calculated by comparing the value from each substrate with that from 6-deoxo-3-dehydroteasterone.
Gene expression analysis
To determine the organ specificity of CYP734A gene expression, we separately prepared total RNA from various organs of wild-type rice (O. sativa L. ‘Nipponbare’). Seeds of the wild-type, mutants, and transformants were sterilized in 1% NaClO and sown on MS medium. Seedlings were grown in a growth chamber for 2 weeks under continuous light. To investigate the effects of the brassinosteroids, seedlings of wild-type rice (2 weeks old) were treated with 100 nm BL. Total RNA was extracted from the whole seedlings. Single-strand cDNAs were synthesized by using an Advantage RT-for-PCR kit (Clontech, http://www.clontech.com/). Quantitative RT-PCR was performed with an iCycler iQ real-time PCR system (Bio-Rad Laboratories, http://www.bio-rad.com/). Expression levels were normalized against the values obtained for the rice ubiquitin gene, which was used as an internal reference. The primer sequences are listed in Table S3.
TS was supported by the Program for Improvement of the Research Environment for Young Researchers, from the Special Coordination Funds for Promoting Science and Technology, commissioned by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan; and by a Grant-in-Aid for Young Scientists (A) (no. 19688001) from MEXT. SF was supported by a Grant-in-Aid for Scientific Research (B) (no. 19380069) from MEXT. MM was supported by a Grant-in-Aid for Scientific Research (C) (no. 18580091) from MEXT; and by the Program for Promotion of Basic and Applied Researches for Innovations in Bio-oriented Industry (BRAIN).