Synthesis of 2RS,4RS-1-[2-Phenyl-4-[2-(2-trifluromethoxy-phenoxy)-ethyl]-1,3-dioxolan-2-yl-methyl]-1H-1,2,4-triazole Derivatives as Potent Inhibitors of Brassinosteroid Biosynthesis

Brassinosteroids are important phytohormones that affect many aspects of plant growth and development. In order to manipulate brassinosteroid levels in plant tissues by using specific biosynthesis inhibitors, we have carried out a systemic search for specific inhibitors of brassinosteroid biosynthesis. Synthesis of triazole derivatives based on the ketoconazole scaffold revealed a series of novel brassinosteroid biosynthesis inhibitors (the YCZ series). To explore the structure-activity relationships of this synthetic series, we now report the synthesis of new triazole derivatives with different aromatic structures at position 2 of 1,3-dioxolane skeleton. We found that the variation of aromatic substituent significantly affect the inhibitory potency. Structure-activity relationships studies indicated that 4-chlorophenyl analogue is the most potent inhibitor of BR biosynthesis with an IC50 value approximately 0.12 ± 0.04 µM, while a bulky biphenyl group exhibited a great negative effect on promoting the inhibitory potency with an IC50 larger than 10 µM.


Introduction
Brassinosteroids (BRs), a class of polyhydroxysteroids widely distributed in plant kingdom, are potent phytohormones that affect many aspects of plant growth and development [1]. Physiological studies have demonstrated that BR can induce diverse cellular responses such as stem elongation, OPEN ACCESS pollen tube growth, leaf bending, root inhibition, induction of ethylene biosynthesis and fruit ripening, and stress tolerance [1][2][3]. The identification of BR biosynthetic mutants of Arabidopsis, tomato, rice and pea established that BR is a new class of phytohormone with an essential role in plant growth and development [4][5][6][7]. Since then, efforts have been made to control the BR biosynthesis in plant tissues by genetic approaches and available evidence indicates that mutations in BR biosynthesis may be a means to improve biomass production [8,9]. Consequently, development novel methodology to manipulate BR levels in plant tissues attracts great research interests.
An alternative method to control the BR levels in plant tissues is the use of specific inhibitors that target the enzymes responsible for BR biosynthesis. BR biosynthesis inhibitors have consequently become highly viable candidates for plant growth regulators. Asami and Yoshida reported the discovery of brassinazole, the first synthetic BRs biosynthesis inhibitor (Brz series) [10]. Subsequent studies on the mode of actions of Brz provided important information about the functions of BRs [11,12]. Recent progress on molecular functional analysis enzymes of BR biosynthesis provided insight evidences that P450 enzymes play key roles in the process of hydroxylation of BR [13][14][15]. Accordingly, strategies for designing P450 inhibitors can be applied to the identification of BR synthesis inhibitors. Cytochrome P450 inhibition mechanisms have been studied in considerable detail [16], triazole derivatives have been demonstrated widespread utility as inhibitors of P450s, due to the intrinsic affinity of the nitrogen electron pair in heterocyclic molecules for the prosthetic heme iron [17].
The biological activity of this synthetic series (YCZ series) was evaluated by testing the ability of the compounds on retardation stem elongation of Arabidopsis seedlings. The compound-induced hypocotyls dwarfism were counteracted by the co-application of 10 nM brassinolide, the most potent BR, but not 1 M GA 3 , which produced seedlings that resembled BR-deficient mutants. In order to further study the structure-activity relationships of YCZ series, we report herein the synthesis of new triazole derivatives with different aromatic structure at the position 2 of the 1,3-dioxolane moiety to mimic the partial structure of 4-chlorophenyl moiety found in YCZ-14 (The general structure of target compounds 7a-h is shown in Figure 1). Structure-activity relationships of newly synthesized compounds were discussed.

Chemistry
Target compounds 7a-h were prepared according to a synthetic route (Scheme 1) as we previously described [22]. and (4) ketal formation to generate 6a-h. Compounds 2a-h were prepared by reacting different kinds of commercially available a-bromoketones 1 with triazole in DMF using a method that we described previously [23]. The alkylation of isopropylidene glycerol 3 was achieved using a method we described previously [22], and hydrolysis with 1 M HCl in MeOH yielded glyceryl tosylate 5. Ketal formation to generate 6a-h were carried out using three equivalents of trifluoromethanesulfonic acid (TfOH) in toluene at room temperature for 60 h, according to a method previously described [22]. All of the compounds synthesized in this work consist of four stereoisomers, and they were subjected to biological studies without further purification.

Bioassay Methods
The bioassay used for determination the activity of BR biosynthesis inhibitors was carried out by a method as we described previously [22]. Arabidopsis BR synthesis-deficient mutants such as dwarf 1 show remarkable dwarfism and the opening of the apical hook of cotyledons in the dark [4]. This unique de-etiolation in the dark phenotype has been used for screening for BR biosynthesis inhibitors [24]. In the present study, we adapted this assay method to determine the effects of test compounds on hypocotyls elongation of Arabidopsis seedlings grown in the dark, and we co-applied BL and GA with the test compounds to determine the reversibility of their effects. With this assay system, we evaluated the biological activities of synthesized compounds.

Biological Activities of Newly Synthesized Brassinosteroid Biosynthesis Inhibitors
The chemical structures of compounds applied for biological studies are shown in Table 1. To identify the aromatic chemical structure at position 2 of 1,3-dioxolane ring responsible for the retardation of Arabidopsis stem elongation, various aromatic substituent were introduced to the inhibitors 7a-h. We used YCZ-14 and Brz as positive controls. A phenyl analogue (compound 7a) was used as a baseline reference for structure-activity relationships discussions. The concentrations of all of the test compounds as well as Brz were assigned to be 0, 0.01, 0.05, 0.1, 0.5, 1 and 10 M, and the IC 50 values were calculated accordingly. As shown in Table 1, compound 7a exhibits inhibitory activity on retarding hypocotyls elongation of Arabidopsis seedling grown in the dark, with an IC 50 value of 0.46 ± 0.04 M, while the IC 50 of YCZ-14 was 0.12 ± 0.04 and Brz was 0.73 ± 0.13 M, respectively. This result indicates that the inhibitory potency of YCZ-14 (4-chlorophenyl analogue) is stronger than that of 7a and that a mono substituent at position 4 of the phenyl moiety may promote the inhibitory activity. To verify this possibility, we further introduce 4-methylphenyl, 4-fluorophenyl and 4-trifluoromethylphenyl moieties into the inhibitor (analogues 7b-d) to evaluate their effect on inhibitory activity.
We found that analogues with methyl and fluorine atom substituents at position 4 of the phenyl ring (compounds 7b,c) have a positive effect on promoting the inhibitory activity compared with that of 7a, with IC 50 values of 0.26 ± 0.05 and 0.21 ± 0.01 M, respectively. Interestingly, introducing a 4-trifluoromethylphenyl moiety (compound 7d) to the position 2 of 1,3-dioxolane, however, showed a significant negative effect on promoting inhibitory activity, with an IC 50 value approximately 0.73 ± 0.06 M. It is worthwhile to note that compound 7d shares the common 4-trifluoromethylphenyl moiety with Brz220, the most potent inhibitor of BR biosynthesis inhibitor reported by Asami et al. (the Brz series) [25]. Data obtained in this work suggests that the structure requirements for these two synthetic series (Brz and YCZ) on inhibition of BR biosynthesis are different. This observation implies the binding site of YCZ series may be different from Brz series. In order to further determine the structure-activity relationships of YCZ synthetic series, we next introduced a bulky aromatic moiety to the inhibitor. As shown in Table 1, the IC 50 value of the biphenyl analogue 7e and naphthalene analogue 7f are greater than 10 and 2.63 ± 0.39 M, respectively. This result indicates that a bulky aromatic moiety significantly reduces the inhibitory activity. Although the variation of aromatic structure greatly affects the inhibitory activity of this synthetic series, the 4-chlorophenyl analogue (YCZ-14) is the most potent inhibitor. This observation suggests that the substituent(s) of chlorine atom on the phenyl moiety enhance the inhibitory activity of this synthetic series. Thus, we next synthesized 2,4-dichlorophenyl analogue 7g and 3,4-dichlorophenyl analog 7h for further structure-activity relationships studies. We found the IC 50 values for 7g and 7h are 0.19 ± 0.05 and 2.40 ± 0.22 M, respectively. Although the structural difference of these two analogues is the position of the chlorine atom, the inhibition potencies of these two analogues are quite different. This result indicates that the position of the chlorine atom on the phenyl ring is sensitive to the binding site. Compared the inhibitory potency of analogues reported in this work with Brz, we found the potency of several YCZ analogues are greater than Brz, except 7d, 7e, 7f and 7g. Nevertheless, among the compounds studied in this work, YCZ-14 is the most potent inhibitor on retardation stem elongation of Arabidopsis seedlings.
It is known that GA biosynthesis inhibitors, such as paclobutrazol, retard the stem elongation of many plant species by blocking ent-kaurene oxidation and also mildly affect other cytochrome P450 mono-oxygenases [25]. This retardation can be rescued by the application of GA. In order to rule out the GA biosynthesis inhibitor among these analogues, we tested the effects of brassinolide, the most biologically active BR, and GA on the recovery of chemical induced dwarfism of Arabidopsis seedlings grown in the dark. Accordingly, we selected analogues with IC 50 values less than 1 M for studying the mode of actions. As shown in Table 2, compounds 7a to 7d and 7g were subjected to the bioassay at a concentration of 0.5 M and Arabidopsis seedlings were grown in the presence of BL (10 nM) or GA (1 M) for 5 days in the dark. Data were expressed in percentage by the comparison of none chemical treated control. As shown by the data, in the presence of BL (10 nM) or GA (1 M), the average hypocotyls length of Arabidopsis seedlings were approximately 114.3 ± 8.0 and 104.3 ± 5.4%, respectively. This result indicates that BL and GA stimulate hypocotyls elongation of Arabidopsis seedlings. We found all the test compounds exhibited high inhibitory activity on Arabidopsis seedlings elongation. The hypocotyls length Arabidopsis seedlings treated with 7a to 7d and 7g are approximately 45.9 ± 2.5, 37.6 ± 3.2, 32.5 ± 3.2, 56.7 ± 1.1 and 21.2 ± 2.3% of non-chemically treated seedlings, respectively, while the positive control of YCZ-14 was approximately 18.2 ± 2.0%, and that of Brz was 56.0 ± 3.3%, respectively. This result indicate that at a concentration of 0.5 M of chemicals, YCZ-14 and 7g strongly inhibited hypocotyls elongation of Arabidopsis seedlings, while Brz inhibited hypocotyls elongation at a degree about 56%. This result clearly indicates that the potency of YCZ-14 and 7g are greater than that of Brz on inhibition hypocotyls elongation in Arabidopsis seedlings. Co-application of BL (10 nM) showed different recovery to different test compounds. Among the compounds listed in Table 2, all the YCZ analogues showed good recovery to BL, the hypocotyls length were 99.8 ± 4.4, 95.4 ± 4.9, 101.2 ± 4.5, 103.3 ± 4.5 and 104.1 ± 3.3% of none chemical treated control, respectively, while the recovery degree of Brz is 63.0 ± 6.1%. Co-application of GA (1 M) did not show significant recovery for YCZ analogues (from 31 to 51%), while the recovery of Brz for GA treatment is 61.8 ± 2.3%. Interestingly, YCZ-14 not only inhibited hypocotyls elongation of Arabidopsis seedlings, it also induced morphological changes of Arabidopsis seedlings grown in the dark (As shown in Figure 2). These physiological changes can be counteracted by application of BL ( Figure 2C) but not GA ( Figure 2B).

General
Chemicals for synthesis were purchased from Kanto Chemicals Co. Ltd. (Tokyo, Japan) and Tokyo Kasei Co. Ltd. (Tokyo, Japan). Reagents are of the highest grade commercially available. Melting

YCZ-14+BL
points (mp) were determined with a Yanako melting point apparatus (Tokyo, Japan). 1 H-NMR spectra were recorded with a JEOL ECP-400 spectrometer (Tokyo, Japan), chemical shifts being expressed in ppm downfield from TMS as an internal standard. High resolution electrospray ionization Fourier transform ion cyclotron resonance mass spectra (ESI-FTICR) were recorded on an Exactive MS System (Thermo Fisher Scientific, Waltham, MA, USA).

Synthesis
Preparation of 1-phenyl-2-(1,2,4-triazol-1-yl)ethanone (2a). Compound 2a was prepared using phenacyl bromide as the starting material, as described previously [23]. To a vigorously stirred suspension of 1H-1,2,4-triazole (5.52 g, 0.08 mol) and phenacyl bromide 1a (9.95 g, 0.05 mol) in 30 mL acetone, was added triethylamine (8.1 g, 0.05 mol) dropwise over a period of 1 h with the temperature kept below 0 °C, and the reaction mixture was stirred for another 30 min at room temperature. The mixture was filtered to remove triethylamine hydrobromide salt precipitates, the precipitates was washed with 3 × 10 mL acetone, the combined washings and filtrate were evaporated under reduced pressure, and the residues were dissolved in 50 mL chloroform and then washed with 2 × 25 mL water. After evaporation of chloroform, the yellow solid was recrystallized with 2-propanol, and 1-phenyl-2-(1,2,4-triazol-1-yl) ethanone (2a) was obtained as a white solid with a yield of 22.7%.   (4). A dry 50 mL round-bottomed flask was charged with p-toluenesulfonyl chloride (7.1 g, 37 mmol), to which pyridine (15 mL) was added while stirring under nitrogen with a magnetic stir bar. The reaction flask was placed inside a container; an ice/water mixture could be added in the event that the reaction becomes too exothermic. 2,2-Dimethyl-1,3-dioxolane-4-methanol (3.29 g, 25 mmol) was added slowly and stirred for 1.5 h. The mixture was slowly added to a vigorously stirred mixture of de-ionized water (70 mL) and crushed ice (70 g) in a 300 mL Erlenmeyer flask and allowed to stir for an additional 0.5 h.

Bioassay Methods for Evaluation Brassinosteroid Biosynthesis Inhibitor
Seeds of Arabidopsis (Columbia ecotype) were purchased from Lehle Seeds (Round Rock, TX, USA). The seeds used for the assay were sterilized in 1% NaOCl for 20 min and washed with sterile distilled water. Seeds were sown on a 1% solidified agar medium containing half Murashige and Skoog salt in Petri dishes with or without chemicals. Plants were grown in 16 h light (240 µE/m 2 s) and 8 h dark conditions at 22 °C in a growth chamber with or without added chemicals. For the dark condition, Petri dishes were wrapped in four layers of aluminum foil. The biological activities of the test compounds were measured 5 days after sowing the seeds. The hypocotyl length of 15-20 Arabidopsis seedlings was measured with a ruler, and each experiment was repeated at least two times.

Conclusions
We have reported the synthesis and structure-activity relationships studies of 2RS,4RS-1-[2-phenyl-4-[2-(2-trifluromethoxyphenoxy)-ethyl]-1,3-dioxolan-2-yl-methyl]-1H-1,2,4-triazole derivatives as potent inhibitors of brassinosteroid biosynthesis. The analogues with different aromatic substituent at position 2 of dioxolane were successfully synthesized by a method we described previously [22]. The biological activity of the test compounds were evaluated by a method using Arabidopsis seedlings grown in the dark. We found that the variation of aromatic substituent significantly affect the inhibitory potency. Structure-activity relationships studies indicated that a 4-chlorophenyl analogue is the most potent inhibitor of BR biosynthesis, with an IC 50 value approximately 0.12 ± 0.04 M, while a bulky biphenyl group exhibited a great negative effect on promoting the inhibitory potency with an IC 50 larger than 10 M.