Structure-activity relationship of uniconazole, a potent inhibitor of ABA 8′-hydroxylase, with a focus on hydrophilic functional groups and conformation

Article abstract of DOI:10.1016/j.bmc.2007.12.019

The plant growth retardant S-(+)-uniconazole (UNI-OH) is a strong inhibitor of abscisic acid (ABA) 8′-hydroxylase, a key enzyme in the catabolism of ABA, a plant hormone involved in stress tolerance, stomatal closure, flowering, seed dormancy, and other physiological events. In the present study, we focused on the two polar sites of UNI-OH and synthesized 3- and 2″-modified analogs. Conformational analysis and an in vitro enzyme inhibition assay yielded new findings on the structure-activity relationship of UNI-OH: (1) by substituting imidazole for triazole, which increases affinity to heme iron, we identified a more potent compound, IMI-OH; (2) the polar group at the 3-position increases affinity for the active site by electrostatic or hydrogen-bonding interactions; (3) the conformer preference for a polar environment partially contributes to affinity for the active site. These findings should be useful for designing potent azole-containing specific inhibitors of ABA 8′-hydroxylase.

Full text of DOI:10.1016/j.bmc.2007.12.019

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 3141–3152
Structure–activity relationship of uniconazole, a potent inhibitor
of ABA 80-hydroxylase, with a focus on hydrophilic
functional groups and conformation
Yasushi Todoroki,a,* Kyotaro Kobayashi,a Hidetaka Yoneyama,a Saori Hiramatsu,a
Mei-Hong Jin,a Bunta Watanabe,b, Masaharu Mizutanib and Nobuhiro Hiraic
aDepartment of Applied Biological Chemistry, Faculty of Agriculture, Shizuoka University, Shizuoka 422-8529, Japan
bInstitute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
cDivision of Environmental Science and Technology, Graduate School of Agriculture, Kyoto University, Kyoto 606-8501, Japan
Received 21 November 2007; revised 11 December 2007; accepted 12 December 2007
Available online 4 March 2008
Abstract—The plant growth retardant S-(+)-uniconazole (UNI-OH) is a strong inhibitor of abscisic acid (ABA) 80-hydroxylase, a
key enzyme in the catabolism of ABA, a plant hormone involved in stress tolerance, stomatal closure, flowering, seed dormancy, and
other physiological events. In the present study, we focused on the two polar sites of UNI-OH and synthesized 3- and 200-modified
analogs. Conformational analysis and an in vitro enzyme inhibition assay yielded new findings on the structure–activity relationship
of UNI-OH: (1) by substituting imidazole for triazole, which increases affinity to heme iron, we identified a more potent compound,
IMI-OH; (2) the polar group at the 3-position increases affinity for the active site by electrostatic or hydrogen-bonding interactions;
(3) the conformer preference for a polar environment partially contributes to affinity for the active site. These findings should be
useful for designing potent azole-containing specific inhibitors of ABA 80-hydroxylase.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
revealed that UNI-OH strongly inhibits ABA 80-
hydroxylase, a key enzyme in ABA catabolism (Fig. 1).
Uniconazole [S-(+)-E-1-(4-chlorophenyl)-4,4-dimethyl-
2-(1,2,4-triazo-1-yl)-1-penten-3-ol, UNI-OH] is an
azole-containing P450 inhibitor developed as a plant
growth retardant in the 1980s.1,2 UNI-OH has since
been used as a plant growth regulator in agriculture
and horticulture. The main site of action of UNI-OH
is suggested to be ent-kaurene oxidase, which catalyzes
the three-step oxidation of ent-kaurene to ent-kaurenoic
acid,3 biosynthetic precursors of the plant hormone gib-
berellin (GA). This has prompted researchers to use
UNI-OH as a chemical tool inhibiting GA biosynthesis.
However, UNI-OH also inhibits brassinosteroid (BR)
biosynthesis,4,5 and alters the level of other plant hor-
mones, such as auxins, cytokinins, ethylene, and abscisic
acid (ABA).6 Recently, Kitahata et al.7 and Saito et al.8
ABA is the classical plant hormone involved in stress
tolerance, stomatal closure, flowering, seed dormancy,
and other physiological events.9–11 In addition, it was re-
ported recently that ABA functions as an endogenous
proinflammatory cytokine in humans.12 Proper endoge-
nous levels of ABA in plants are cooperatively con-
trolled by biosynthesis, transportation, and catabolic
inactivation in response to environmental changes.9–11
A natural or artificial chemical that perturbs this highly
controlled system is promising not only as a chemical
probe for the mechanism of ABA action,13 but also
for its potential agricultural, horticultural, and clinical
use. Catabolic inactivation of ABA is mainly controlled
by ABA 80-hydroxylase, which is the cytochrome P450
catalyzing the C80-hydroxylation of ABA into 80-hydro-
xy-ABA and its more stable tautomer phaseic acid,
which has much lower hormonal activity than ABA.14
ABA 80-hydroxylase was identified as CYP707A1-4 in
the model plant Arabidopsis thaliana in 2004,15,16 and
since then CYP707A isozymes have been found in vari-
ous plants.17–20 Gene knockdown and overexpression
Keywords: Uniconazole; Abscisic acid; ABA; P450; Azole.
*
Corresponding author. Tel./fax: +81 54 238 4871; e-mail:
aytodor@agr.shizuoka.ac.jp
Present address: Sagami Chemical Research Center, 2743-1 Hayak-
awa, Ayase, Kanagawa 252-1193, Japan.
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.12.019

3142
Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
Figure 1. Uniconazole (UNI-OH) is a potent inhibitor of ABA 80-hydroxylase and ent-kaurene oxidase.
1
studies suggest that ABA 80-hydroxylase is a key enzyme
controlling ABA concentration during water deficit
stress and dormancy.21,22 UNI-OH inhibits ABA 80-
hydroxylase at a very low concentration: its KI value
was 8 nM in an in vitro assay.8 UNI-OH-treated plants
have enhanced drought tolerance due to twofold higher
accumulation of endogenous ABA than untreated
plants.8 Thus, UNI-OH is a good inhibitor of ABA 80-
hydroxylase, though it has side effects such as inhibition
of GA and BR biosynthesis.
In fact, Katagi et al. reported, based on H NMR and
IR analyses, that UNI-OH adopts conformer A in non-
polar or weakly polar solvents (Fig. 2) by the formation
of an intramolecular hydrogen bond between the 3-hy-
droxy and N200.23 We designed UNI-X and IMI-X
(X = OH, OMe, F, and H) (Fig. 3) as probes to explore
the role of the 3-hydroxy and N200 of UNI-OH in bind-
ing to ABA 80-hydroxylase. The hydroxy group can act
as both a donor and an acceptor in a hydrogen bond,
whereas the methoxy group can act only as an acceptor.
Fluorine can act only as a weaker acceptor than hydroxy
and methoxy groups. The aliphatic hydrogen can be nei-
ther donor nor acceptor in a conventional hydrogen
bond. The introduction of a carbon at the 200-position
can eliminate the intramolecular hydrogen bond.
Although UNI-OMe,2 UNI-H,24 and IMI-OH25 are
known compounds, the detailed effects of all optically
pure UNI-X and IMI-X containing these known com-
pounds on ABA 80-hydroxylase have never been re-
ported. These analogs are expected to have different
conformational profiles and different interactions with
the active site; thus, a comparison between their struc-
ture and inhibitory activity against ABA 80-hydroxylase
can reveal the significance of the hydrophilic moieties of
UNI-OH in binding to the active site.
Azole-type inhibitors bind to P450 active sites by both
coordinating to the heme-iron atom and interacting with
surrounding protein residues. Because heme coordina-
tion is a common property of azole-containing inhibi-
tors, their affinity and specificity for individual P450
enzymes depends on structural properties other than
the azole group. To develop a specific inhibitor of
ABA 80-hydroxylase by structural modification of
UNI-OH, the structure–activity relationship of UNI-
OH should be revealed in detail. Kitahata et al. investi-
gated the structure–activity relationship of UNI-OH by
focusing on the two lipophilic substituents, the tert-butyl
at the 3-position and 4-chlorophenyl group at the 1-po-
sition,7 revealing that these functional groups are essen-
tial for inhibiting ABA 80-hydroxylase and should not
be modified in the lead optimization phase. Other func-
tional groups have not been examined.
2. Results and discussion
In the present study, we focused on the two hydrophilic
functional groups of UNI-OH, the aza-nitrogen (N200)
and the hydroxy group at the 3-position. The electro-
static or hydrogen-bonding interactions mediated by
these groups can affect not only electrostatic interactions
with the active site but also conformational properties.
2.1. Synthesis of UNI-OH analogs
Commercially available UNI-OH, uniconazole-P (Wako
Pure Chemical Industries, Ltd), was a mixture of S-(+)-
and R-(ꢀ)-UNI-OH at a ratio of 89:11, based on chiral
HPLC analysis; that is, the enantiomeric excess was

Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
3143
Figure 2. Four conformers A–D generated by bond rotations of C2–N100 and C2–C3 in UNI-X (R = N) and IMI-X (R = CH). Stick models of UNI-
OH: carbons, gray; hydrogens, white; oxygens, red; nitrogens, blue; chlorine, light green.
Figure 3. Structures of UNI-X and IMI-X.
79%. Optically pure S-(+)-UNI-OH (>99% ee) was ob-
tained by semi-preparatory HPLC with a chiral station-
ary phase. Only the S-enantiomers were used in this
study because S-UNI-OH and its derivatives (e.g., dinic-
Scheme 1. Synthesis and optical resolution of UNI-OH, UNI-OMe,
and UNI-F. Reagents: (i) Chiralpak AD-H; (ii) MeI, NaH; (iii)
Chiralpak AD-H; (iv) DAST; (v) Chiralcel OD-H.
onazole) are effective in plants, whereas the R-enantio-
mers are ineffective in plants but act as fungicides.26,27
UNI-OMe and UNI-F were synthesized according to
Scheme 1. Methylation of uniconazole-P with methyl io-
dide gave UNI-OMe, which was optically purified by
chiral HPLC to give S-(ꢀ)-UNI-OMe. Fluorination of
uniconazole-P with DAST gave UNI-F, which was a
1:1 mixture of (+)- and (ꢀ)-enantiomers, based on chiral
HPLC analysis. The reaction also gave the 1-fluorinated
compound as a by-product. This result suggests that the
fluorination was an SN1 reaction in spite of the observa-
tion that DAST fluorination favors an SN2 reaction in
most cases.28 This is probably because the carbocation
intermediate at the 3-position was stabilized by the
tert-butyl group and the conjugated system. The racemic
UNI-F was separated into each enantiomer by chiral
HPLC. The absolute configuration of optically pure
UNI-F was determined by X-ray crystalline structure
analysis; the (+)- and (ꢀ)-enantiomers have the R- and
S-configurations, respectively. UNI-H was synthesized
according to Scheme 2. Aldol condensation of pivalde-
hyde and 40-chloroacetophenone gave compound 1,
which was reduced and brominated to give compound
3. A coupling reaction of 3 and 1,2,4-1H-triazole gave
compound 4, which was reduced, brominated, and dehy-
drobrominated to give UNI-H.

3144
Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
Scheme 2. Synthesis of UNI-H and IMI-H. Reagents and condition: (i) NaOH; (ii) PtO2; (iii) Br2; (iv) 1H-1,2,4-triazole, Na2CO3; (v) NaBH4; (vi)
MsCl, Et3N, then LiBr, LiCO3; (vii) NaOH; (viii) 1H-imidazole, Na2CO3; (ix) NaBH4; (x) MsCl, Et3N, then LiBr, LiCO3; (xi) UV 365 nm.
IMI-OH, IMI-OMe, and IMI-F were synthesized
according to Scheme 3. A coupling reaction of imidaz-
ole and 1-bromo-3,3-dimethylbutan-2-one gave com-
was determined to be S. The absolute configuration
of optically pure IMI-F, which was not crystallized,
was estimated from the pattern of Cotton effects, which
was similar to that of optically pure IMI-OH. IMI-OH
and IMI-F have the same two chromophores, the imid-
azole and 4-chlorophenyl. Although these are linked by
a C@C double bond, the X-ray and theoretical struc-
tures of IMI-OH reveal that these are not on the same
plane owing to torsion. The 3D orientation of IMI-OH
and IMI-F with the same absolute configuration is
similar on the basis of the theoretical geometries
(Fig. 4). Thus, these two compounds with the same
absolute configuration should give similar circular
dichroism (CD) spectra. In CD analysis, the (ꢀ)-isomers
of IMI-OH and IMI-F gave similar CD spectra: nega-
tive first and positive second Cotton effects. Thus, we
determined the absolute configurations of (+)- and
(ꢀ)-IMI-F as R and S, respectively.
pound 7. Aldol condensation of
7
and 4-
chlorobenzaldehyde gave the Z-ketone 8Z, which was
converted into the E-ketone 8E by photoisomerization
using a single light source at 365 nm. The E-ketone 8E
was reduced to give racemic IMI-OH, which was meth-
ylated and fluorinated to give racemic IMI-OMe and
IMI-F, respectively. IMI-OH, IMI-OMe, and IMI-F
were resolved into each enantiomer by chiral HPLC.
IMI-H was synthesized in a similar manner to UNI-
H, according to Scheme 2. The absolute configuration
of IMI-OH was determined by X-ray crystalline struc-
ture analysis; the (+)- and (ꢀ)-enantiomers have the R-
and S-configurations, respectively. Because methylation
of S-(ꢀ)-IMI-OH gave the (ꢀ)-enantiomer of IMI-
OMe, the absolute configuration of (ꢀ)-IMI-OMe

Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
3145
Scheme 3. Synthesis and optical resolution of IMI-OH, IMI-OMe, and IMI-F. Reagents and condition: (i) K2CO3; (ii) p-chlorobenzaldehyde,
K2CO3, Ac2O; (iii) UV 365 nm; (iv) NaBH4; (v) Chiralpak AD-H; (vi) MeI, NaH; (vii) Chiralpak AD-H; (viii) DAST; (ix) Chiralpak AD-H.
C3 bond has a great effect on the molecular shape and
electrostatic distribution. C2–N100 bond rotation affects
the orientation of aza-nitrogens, N200 and N400 for
UNI-X and N300 for IMI-X. Because the aza-nitrogens
at the 4-position in the triazole ring and the 3-position
in the imidazole ring are coordinated with the heme iron
at the active site,29 the orientation of these nitrogens
determines the primary location of UNI-X and IMI-X
at the active site. The N200 of UNI-X should affect con-
formational preference by interaction with the 3-substi-
tuent. Other bond rotations will have only a small effect
on the molecular shape, electrostatic distribution, and
location within the active site, although they may give
many minor conformers. We constructed the four initial
geometries, conformers A–D (Fig. 2), on the basis of the
C2–N200 and C2–C3 bond rotations. The most stable
geometries of each conformer were obtained by
theoretical calculation using density functional theory.
Conformers A–D were fully optimized at the B3LYP/
Figure 4. Overlaid 3D structures of S-IMI-OH (blue) and S-IMI-F
(red) based on the theoretical geometries.
2.2. Theoretical conformational analysis
6-31G(d) level of the theory, and single point energies
were obtained at B3LYP/6-311++G(2df,2p) with zero-
point energy corrections. The geometries and energies
in the aqueous phase were calculated, respectively, using
Because UNI-X and IMI-X have tert-butyl and hydro-
philic groups at the 3-position, the rotation of the C2–

3146
Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
the Onsager and PCM models at the same level of the
theory as in the gas phase.
Because conformers A and B in UNI-OH were much
more stable than conformers C and D, the populations
of which were less than 5% in both the gas and aqueous
phases, we focused on only the A/B population ratio in
all the analogs (Table 1). UNI-OH prefers conformer A
in the gas phase but B in the aqueous phase. The result
in the gas phase agrees with the MNDO calculation of
Katagi et al.23 This conformational preference can be
explained by the intramolecular hydrogen bond between
the 3-OH and N200, which contributes to the conforma-
tional energy in the gas phase more than in the aqueous
phase. UNI-OMe and UNI-F prefer conformer B in
both phases, probably because UNI-OMe and UNI-F
have an electrostatic repulsion between the 3-OMe or
-F and N200. This repulsion seems to be weakened in
the aqueous phase. UNI-H and all the IMI-X types
favor A and B equally. Because these analogs have no
significant intramolecular hydrogen bond, the confor-
mational preference will not be affected by environmen-
tal conditions. For UNI-OH, the energy difference
between the most and least stable conformers was calcu-
lated to be less than 5.7 kcal molꢀ1 in the gas phase and
2.9 kcal molꢀ1 in the aqueous phase, and the energy
barriers for the C2–N200 and C2–C3 bond rotations were
estimated to be 8.2 and 11.2 kcal molꢀ1, respectively.
Figure 5. Selected NOESY correlations (arrows) for UNI-X and IMI-
X in CDCl3 or methanol-d4.
a through-space J-coupling between the 3-F and 500-H
in both solvents. These observations indicate that
UNI-OMe and UNI-F favor conformer B independent
of solvent polarity. However, UNI-H and all the IMI-
X types showed NOESY correlations between the 200
(and 500)-H and tert-butyl protons, and between 200
(and 500)-H and 1-H, indicating that they favor conform-
ers A and B equally. These results are consistent with the
theoretical results.
2.3. Experimental conformational analysis
The theoretical estimations for conformational profiles
of UNI-X and IMI-X were supported by NOESY corre-
lation experiments (Fig. 5). UNI-OH showed a NOESY
correlation between the 1-H and 500-H in CDCl3. This
was not observed in methanol-d4, whereas a NOESY
correlation between the tert-butyl protons and 500-H
was observed in this polar solvent. This indicates that
UNI-OH prefers conformer A in the less polar solvent
but B in the polar solvent. UNI-OMe and UNI-F
showed an NOE between the tert-butyl protons and
500-H in both CDCl3 and methanol-d4. UNI-F showed
The 3D structures of UNI-OH, UNI-F, and IMI-OH,
which were able to be crystallized in methanol–water,
were elucidated by X-ray crystallographic analysis
(Fig. 6). All three compounds adopted conformer B in
the crystals. For UNI-OH and UNI-F, this result agrees
with the theoretical and NMR-based experimental anal-
yses. IMI-OH, which was predicted to favor A and B
equally, was packed in the B form in the crystal. Because
an intermolecular hydrogen bond was observed between
the 3-OH and N400 between the adjacent molecules in the
crystal (CCDC 668251, see Section 4.7), this interaction
may drive the crystallization of the B-type IMI-OH.
Table 1. The calculated A/B conformer ratio of UNI-X and IMI-X
Compound
In gas phasea
In aqueous phaseb
c
dEAꢀB
A/B ratio dEAꢀB
A/B ratio
(kcal molꢀ1
)
(kcal molꢀ1
)
UNI-OH
UNI-OMe +2.71
ꢀ2.51
99/1
1/99
1/99
70/30
+0.87
+1.26
+0.94
ꢀ0.34
ꢀ0.05
+0.02
ꢀ0.09
ꢀ0.19
19/81
11/89
17/83
64/36
2.4. Inhibitory potency against ABA 80-hydroxylase
UNI-F
UNI-H
+3.14
ꢀ0.49
All the UNI-X and IMI-X types inhibited ABA 80-
hydroxylase, which was obtained from a microsomal
fraction of insect cells expressing Arabidopsis
CYP707A3.15 The S-enantiomers showed competitive
inhibition more strongly than the corresponding R-
enantiomers by a factor of more than 100 (data not
shown). The KI values of all the S-enantiomers, UNI-
H, and IMI-H are summarized in Table 2. In the
UNI-X types, UNI-OH, which has a hydroxy group at
the 3-position, was the most effective; the activity de-
creased in the order UNI-OH > UNI-OMe > UNI-
IMI-OH
IMI-OMe
IMI-F
ꢀ0.20
ꢀ0.09
ꢀ0.30
ꢀ0.17
58/42
54/46
62/38
57/43
52/48
49/51
54/46
58/42
IMI-H
a B3LYP/6-311++G(2df,2p)//B3LYP/6-31G(d).
b B3LYP/6-311++G(2df,2p)//B3LYP/6-31G(d). Geometry optimiza-
tion and frequency calculations were performed using the Onsager
method, and the single point energy calculations were performed
using the IEF-PCM method.
c dEAꢀB = EA ꢀ EB.

Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
3147
Figure 6. 3D structures of UNI-OH (top), UNI-F (middle), and IMI-OH (bottom). (a) X-ray ORTEP structures of the crystalline in MeOH–H2O.
All are R-enantiomers. R-UNI-F was obtained as mixed crystals. (b) Overlaid structures. Magenta: the S-enantiomers, constructed as mirror images
of the X-ray structures of the R-enantiomers. For UNI-F, the mirror image of another crystal structure is not shown. Cyan: theoretical models of the
S-enantiomers in MeOH.
Table 2. Inhibition constants for UNI-X and IMI-X of recombinant CYP707A3
X
UNI-X
KI (UNI-X)/KI (UNI-OH)
IMI-X
KI (IMI-X)/KI (IMI-OH)
KI (UNI-X)/KI (IMI-X)
KI (lM) (SD)
KI (lM) (SD)
OH
OMe
F
2.7 (0.4)
8.6 (1.4)
1
3
0.6 (0.2)
1.2 (0.4)
1.5 (0.7)
14 (3)
1
2
5
7
34 (5)
450 (99)
13
170
3
23
23
32
H
F > UNI-H. A similar tendency was observed in the
IMI-X types. The IMI-X types were always more effec-
tive than the corresponding UNI-X types. IMI-OH
(KI = 600 pM) was the most potent inhibitor of ABA
80-hydroxylase yet tested in an in vitro assay.
OH and IMI-OH were predicted to be more polar than
the other analogs. These two compounds were the most
effective inhibitors. Considering only this point, the
overall molecular polarity appears to be a significant
factor in enzyme inhibition. However, the trend in
the activity of other analogs cannot be explained by
only the molecular polarity; for example, in spite of
UNI-H and IMI-H, respectively, being as polar as
UNI-F and IMI-F, UNI-H and IMI-H were much less
active than UNI-F and IMI-F. This indicates that the
overall molecular polarity of UNI-X and IMI-X has
2.5. Structure–activity relationship
2.5.1. Overall molecular polarity. Based on both the cal-
culated partition coefficient30 and the retention factor
(k) in reverse phase HPLC analysis (Table 3), UNI-

3148
Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
Table 3. The calculated partition coefficient (logP) and retention
factor (k) from reverse phase HPLC analysis
Table 4. Binding energies between the heme model and azoles
Azole
Binding energya
(kcal molꢀ1
d
X
UNI-X
IMI-X
)
(kcal molꢀ1
)
logP
k
logP
k
1-Me-1,2,4-triazole
1-Me-imidazole
ꢀ11.13
ꢀ13.06
0
ꢀ1.93
OH
OMe
F
2.70
3.28
3.61
3.63
3.20
6.33
5.04
5.67
3.19
3.77
4.10
4.12
3.44
8.27
6.88
6.67
a Difference in energy between the azole complex and the sum of the
azole and heme model.
H
2.5.4. Contribution of the conformational preference in
the uncomplexed state. Although it is not necessary for
the ligand conformation in the protein–ligand complex
to be similar to the preferred conformation in the
uncomplexed situation, conformational consistency be-
tween the uncomplexed and complexed molecules
should be advantageous energetically. All the IMI-X
types prefer conformers A and B equally, whereas the
UNI-X types vary in conformational preference; UNI-
OH prefers A in a less polar environment and B in a po-
lar environment; UNI-OMe and UNI-F prefer B in both
environments; and UNI-H prefers A and B equally.
Therefore, the difference in activity between UNI-H
and IMI-H is expected to be a straightforward reflection
of the difference in the affinity to the active site between
the triazole and imidazole rings. In fact, in the above
theoretical analysis, imidazole bound to the heme more
strongly than triazole by a factor of ca. 30, which agrees
with the ratio of KI (UNI-H)/KI (IMI-H). The ratio of
KI(UNI-X)/KI(IMI-X) decreases in the order ꢀH(32) >
ꢀF(23) > ꢀOMe(7) > ꢀOH(5). The smaller influence
of ꢀF, ꢀOMe, and ꢀOH compared to that of ꢀH
may indicate that the conformer B preference in a polar
environment contributes energetically to the affinity to
the active site, although the theoretical energy difference
between conformers A and B in the uncomplexed state
cannot completely explain the order ꢀF(23) >
ꢀOMe(7) > ꢀOH(5).
little effect on binding affinity to the active site of
CYP707A3.
2.5.2. Triazole versus imidazole. Imidazole compounds
usually have a higher affinity than triazole compounds
for the heme of P450 in microsomes.31 The pKa value
of 1-methylimidazole is 7.12, which is larger than that
of 1-methyl-1,2,4-triazole (3.20)32; the N4 in imidazole
is more basic than that of 1,2,4-triazole. We calculated
binding energies between the heme model and 1-
methyl-1,2,4-triazole or 1-methylimidazole (Fig. 7 and
Table 4). The imidazole complex was estimated to be
more stable than the triazole complex by
1.93 kcal molꢀ1, meaning that the imidazole binds to
the heme more strongly than the triazole by a factor
of ca. 30, explaining why the IMI analogs are more po-
tent than the corresponding UNI analogs.
2.5.3. Effect of the 3-substituent. The polar functional
group at the 3-position contributes energetically to bind-
ing at the active site. In both UNI-X and IMI-X, the
apparent contribution increases in the order
ꢀF < ꢀOMe < ꢀOH. The hydroxy group can act as
both a donor and an acceptor in hydrogen bonds with
the active site, whereas the methoxy and fluorine can
act only as an acceptor. The hydrogen acceptor ability
of fluorine is weaker than that of oxygen.33 These prop-
erties may explain the contribution order of
ꢀF < ꢀOMe < ꢀOH. The effect of these polar groups
is apparently larger for the UNI-X types than for the
IMI-X types. The differences in structure between
UNI-X and IMI-X are the existence of N200 in the azole
ring and the conformational profile linked with N200.
The conformational preference affected by N200 may be
significant.
3. Conclusions
In the present study, we focused on the two polar sites of
S-(+)-uniconazole (UNI-OH), a strong inhibitor of
ABA 80-hydroxylase, and synthesized 3- and 200-modi-
fied analogs. Conformational analysis and in vitro en-
Figure 7. The complex between azoles (left: 1-methyl-1,2,4-triazole; right: 1-methylimidazole) and heme model (spin doublet state). The geometries
˚
were optimized at the B3LYP/(LanL2DZ, Fe; 6-31+G(d), CHNOS) level of theory. The distance between Fe and the azole N atom is 2.16 A for
˚
1-methyl-1,2,4-triazole and 2.15 A for 1-methylimidazole. Ball and stick models: carbons, gray; hydrogens, white; oxygens, red; nitrogens, blue;
sulfur, yellow; Fe, dark red.

Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
3149
zyme inhibition assays yielded new findings about the
structure–activity relationship of UNI-OH: (1) by
substituting imidazole for triazole, which increases affin-
ity to the heme iron, we identified a more potent com-
pound, IMI-OH; (2) the polar group at the 3-position
increases the affinity to the enzyme active site by electro-
static or hydrogen-bonding interactions; (3) the con-
former preference in a polar environment partially
contributes to the affinity to the active site. These find-
ings should be useful for designing potent azole-contain-
ing specific inhibitors of ABA 80-hydroxylase. The
conformation linked with the hydrophilic moieties may
be significant for designing the specific inhibitor.
(Hydrosphere C18, 150 · 20 mm, YMC; 75% MeOH;
7 mL minꢀ1; 230 nm) to obtain ( )-UNI-F (11 mg,
38 lmol, 22%) and 1 (23 mg, 78 lmol, 48%). ( )-UNI-
F: 1H NMR (500 MHz, CD3OD): d 0.73 (9H, d,
2
4JHF = 0.9 Hz, tert-butyl), 5.41 (1H, d, JHF = 44.3 Hz,
H-3), 7.31 (1H, s, H-1), 7.42 (2H, m, H-20 and H-60),
7.48 (2H, m, H-30 and H-50), 8.13 (1H, s, H-300), 8.67
5
(1H, d, JHF = 1.8 Hz, H-500); 13C NMR (125 MHz,
4
CD3OD): d 25.8 (d, JCF = 4 Hz, methyls of tert-butyl),
37.0 (d, 2JCF = 22 Hz, tertiary carbon of tert-butyl), 96.1
1
(d, JCF = 175 Hz, C3), 130.2 (C30 and C50), 131.5 (d,
4
5JCF = 2 Hz, C20 and C60), 133.4 (d, JCF = 2 Hz, C10),
135.0 (d, JCF = 18 Hz, C2), 135.2 (d, JCF = 7 Hz,
2
3
4
C1), 135.8 (C40), 146.0 (d, JCF = 8 Hz, C500), 152.7
(C300); 19F NMR (377 MHz, CD3OD): d ꢀ190.4 (d,
2JFH = 44.2 Hz, F-3); UV kmax (MeOH) nm (e): 254.0
(15,300); HRMS (EI): calcd for C15H17ClFN3 [M]+
293.1095, found 293.1091. The 1-fluorinated compound:
1H NMR (270 MHz, CD3OD): d 0.93 (9H, s, tert-butyl),
4. Experimental
4.1. General
2
Uniconazole-P was purchased from Wako Pure Chemi-
cal Industries, Ltd, Osaka, Japan. (+)-ABA was a gift
from Toray Industries Inc., Tokyo, Japan. 1H NMR
spectra were recorded with tetramethylsilane as the
5.97 (1H, d, JHF = 45.5 Hz, H-1), 6.07 (1H, d,
4JHF = 2.5 Hz and J = 0.6 Hz, H-3), 7.11 (2H, m, H-20
and H-60), 7.31 (2H, m, H-30 and H-50), 7.83 (1H, d,
6JHF = 0.9 Hz, H-300), 7.94 (1H, s, H-500); UV kmax
(MeOH) nm (e): 222.8 (14,800); HRMS (EI): calcd for
C15H17ClFN3 [M]+ 293.1095, found 293.1092. ( )-
UNI-F (7 mg) was subjected to chiral HPLC under the
following conditions: column, Chiralcel OD-H
(250 · 10 mm, Daicel); solvent, 3% 2-propanol in n-hex-
ane; flow rate, 3.0 mL minꢀ1; detection, 254 nm. The
materials at tR 15.1 and 19.3 min were collected to give
internal standard using
a
JEOL JNM-EX270
(270 MHz) and JNM-LA500 (500 MHz) NMR spec-
trometer. 13C NMR and 2D-correlation NMR experi-
ments were recorded using JNM-LA500 (500 MHz)
NMR spectrometer. 19F NMR spectra were recorded
with trifluorotoluene as the internal standard using a
Bruker Avance 400 (380 MHz) NMR spectrometer.
High resolution mass spectra (HRMS) were obtained
with a JEOL JMS-DX 303 HF mass spectrometer and
with a JEOL JMS-T100LC ‘AccuTOF’. Optical rota-
tions were recorded with a Jasco DIP-1000 digital polar-
imeter. Circular dichroism (CD) spectra were recorded
with a Jasco J-805 spectrophotometer. Column chroma-
tography was performed on silica gel (Wakogel C-200).
Purity in two solvent systems (H2O–MeOH and H2O–
MeCN) was determined using a Shimadzu LC-10ADVP
instrument, and all final compounds were >99% pure.
X-ray crystallography was performed with a Rigaku
AFC-5R diffractometer for synthetic procedure of the
known compounds (UNI-OMe, UNI-H, and IMI-
OH): see supplementary data.
(ꢀ)-UNI-F (3 mg) and (+)-UNI-F (4 mg), respectively,
26
D
with an optical purity of each 99.9%. (ꢀ)-UNI-F: ½aꢁ
ꢀ11.7 (MeOH, c 0.195); (+)-UNI-F: ½aꢁ26 +11.5 (MeOH,
D
c 0.221).
4.3. Synthesis of (S,E)-1-(1-(4-chlorophenyl)-3-methoxy-
4,4-dimethylpent-1-en-2-yl)-1H- imidazole (IMI-OMe)
4.3.1. Synthesis of ( )-IMI-OMe. ( )-IMI-OMe (40 mg,
131 lmol) was prepared from ( )-IMI-OH (42 mg,
144 lmol, 92%) in the similar manner to the preparation
1
of ( )-UNI-OMe (see supplementary data). H NMR
(500 MHz, CDCl3): d 0.74 (9H, s, tert-butyl), 3.39 (3H,
s, OMe), 3.91 (1H, s, H-3), 6.89 (1H, s, H-1), 7.08 (1H,
s, H-400 or H-500), 7.16 (1H, s, H-500 or H-400), 7.23 (2H,
m, H-20 and H-60), 7.39 (2H, m, H-30 and H-50), 7.77
(1H, s, H-200); 13C NMR (125 MHz, CD3OD): d 26.4
(methyls of tert-butyl), 35.4 (tertiary carbon of tert-bu-
tyl), 57.0 (OMe), 84.0 (C3), 120.4 (C500 or C400), 128.9
(C-30 and C50), 129.1 (C400 or C500), 129.9 (C1), 130.1 (C-
20 and C-60), 132.7 (C10 or C40), 133.9 (C40 or C10),
135.9 (C2), 137.3 (C200); UV kmax (MeOH) nm (e): 248.4
(15,600); HRMS (ESI-TOF, positive mode): calcd for
C17H22ClN2O [M+H]+ 305.1421, found 305.1419.
4.2. Synthesis and optical resolution of (E)-1-(1-(4-chlo-
rophenyl)-3-fluoro-4,4-dimethylpent-1-en-2-yl)-1H-1,2,4-
triazole (UNI-F)
Diethylaminosulfur trifluoride (60 lL, 458 lmol) was
added dropwise into a solution of uniconazole-P
(24 mg, 82 lmol) in dry CH2Cl2 (0.5 mL). The mixture
was stirred for 15 min at room temperature and then
quenched with sat. NH4Cl. The resulting mixture was
extracted with EtOAc (10 mL · 3). The organic layer
was washed with brine and H2O, dried over Na2SO4,
and concentrated in vacuo. The residual oil was purified
by silica gel column chromatography with 8–25%
EtOAc in hexane to obtain a mixture of UNI-F and
the 1-fluorinated compound (26 mg) as colorless oil.
The same reaction was repeated using 23 mg of unico-
nazole-P to obtain 11 mg of the same mixture. The com-
bined mixture was purified by ODS HPLC
4.3.2. Optical resolution of ( )-IMI-OMe. ( )-IMI-OMe
(32 mg) was subjected to chiral HPLC under the follow-
ing conditions: column, Chiralpak AD-H (250 · 10 mm,
Daicel); solvent, 2% 2-propanol in n-hexane; flow rate,
1.0 mL minꢀ1; detection, 254 nm. The materials at tR
26.0 and 35.4 min were collected to give (ꢀ)-IMI-OMe
(14.2 mg) and (+)-IMI-OMe (14.8 mg), respectively,
with an optical purity of each 99.9%. (ꢀ)-IMI-OMe:

3150
24
Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
24
D
½aꢁ ꢀ196.7 (MeOH, c 0.947); (+)-IMI-OMe: ½aꢁ
m, H-200), 7.89 (2H, m, H-30 and H-50); HRMS (ESI-
TOF, positive mode): calcd for C16H20N2OCl [M+H]+
291.1264, found 291.1234.
D
+203.8 (MeOH, c 0.967).
4.3.3. Determination of the absolute configuration of
optically pure IMI-OMe. S-(ꢀ)-IMI-OH (5.5 mg) was
methylated in the same manner as ( )-IMI-OH. After
purification, the methylated compound was subjected to
chiral HPLC under the following conditions: column,
Chiralpak AD-H (250 · 4.6 mm, Daicel); solvent, 2% 2-
propanol in n-hexane; flow rate, 1.0 mL minꢀ1; detection,
254 nm. The retention time of the methylated compound
agreed with that of (ꢀ)-IMI-OMe. Thus, the absolute
configuration of (ꢀ)-IMI-OMe was determined to be S.
4.5.2. 1-(4-Chlorophenyl)-2-(1H-imidazol-1-yl)-4,4-dim-
ethylpentan-1-ol (10). To a stirred solution of 9
(619 mg, 2.13 mmol) in MeOH (2 mL) was added
NaBH4 (120 mg, 3.16 mmol). The mixture was stirred
for 1.5 h at room temperature. After quenching with
1 M HCl, the resulting mixture was extracted with
EtOAc (50 mL · 3). The organic layer was washed with
brine, dried over Na2SO4, and concentrated in vacuo.
The residual oil was purified by silica gel column chro-
matography with 80% EtOAc in hexane to obtain 10
(diastereomeric mixture, 608 mg, 2.08 mmol, 96%) as a
colorless oil. 1H NMR (270 MHz, acetone-d6) of the ma-
jor diastereomer: d 0.76 (9H, s, tert-butyl), 1.79 (1H, dd,
J = 12.7 and 2.2 Hz, H-3), 2.12 (1H, dd, J = 12.7 and
10.5 Hz, H-3), 4.39 (1H, m, H-2), 4.95 (1H, m, H-1),
6.77 (1H, m, H-400 or H-500), 7.02 (1H, m, H-500 or H-
400), 7.16 (2H, m, H-20 and H-60), 7.24 (2H, m, H-30
and H-50), 7.33 (1H, m, H-200); HRMS (ESI-TOF, posi-
tive mode): calcd for C16H22ON2Cl [M+H]+ 293.1421,
found 293.1395.
4.4. Synthesis and optical purification of (S,E)-1-(1-(4-
chlorophenyl)-3-fluoro-4,4-dimethylpent-1-en-2-yl)-1H-
imidazole (IMI-F)
( )-IMI-F (34 mg, 116 lmol) was prepared from
( )-IMI-OH (625 mg, 2.15 mmol, 5%) in the similar manner
1
to the preparation of ( )-UNI-F. H NMR (500 MHz,
4
CD3OD): d 0.72 (9H, d, JHF = 0.6 Hz, tert-butyl),
5.26 (1H, d, JHF = 55.8 Hz, H-3), 7.06 (1H, dd,
2
5
J = 1.2 Hz, JHF = 1.2 Hz, H-500), 7.08 (1H, s, H-1),
7.37 (2H, m, H-20 and H-60), 7.38 (1H, overlapped with
H-20(60), H-400), 7.46 (2H, m, H-30 and H-50), 7.91 (1H,
d, 5JHF = 1.0 Hz, H-200); 13C NMR (125 MHz, CD3OD):
4.5.3. (Z)-1-(1-(4-Chlorophenyl)-4,4-dimethylpent-1-en-2-
yl)-1H-imidazole (11). To a stirred solution of 10
(568 mg, 1.94 mmol) in CH2Cl2 (20 mL) were added
Et3N (660 lL, 479 mg, 4.74 mmol) and MsCl (300 lL,
443 mg, 3.85 mmol) at 0 ꢁC. The mixture was stirred
for 5 h at room temperature under Ar. After quenched
with sat. NaHCO3, the organic layer was dried over
Na2SO4, and then concentrated in vacuo. To a stirred
solution of the residue in dry DMF (5 mL) were added
LiCO3 (986 mg, 13.3 mmol) and LiBr (monohydrate,
1010 mg, 9.62 mmol). The mixture was stirred for 3 h
at 120 ꢁC and then filtered at room temperature. Water
(50 mL) was added to the filtrate, and the resulting mix-
ture was extracted with Et2O (25 mL · 3). The combined
organic layer was washed with brine, dried over
Na2SO4, and concentrated in vacuo. The residual oil
was purified by silica gel column chromatography with
40% EtOAc in hexane to obtain 11 (111 mg,
0.404 mmol, 21%). 1H NMR (270 MHz, CDCl3): d
0.85 (9H, s, tert-butyl), 2.045 (2H, s, H-3), 6.27 (1H, s,
H-1), 6.70 (2H, m, H-2 and H-60), 6.84 (1H, dd,
J = 1.0 and 1.0 Hz, H-400 or H-500), 7.11 (1H, dd,
J = 1.0 and 1.0 Hz, H-500 or H-400), 7.14 (2H, m, H-30
and H-50), 7.37 (1H, dd, J = 1.0 and 1.0 Hz, H-200).
4
d 25.9 (d, JCF = 3 Hz, methyls of tert-butyl), 36.6 (d,
2JCF = 22 Hz, tertiary carbon of tert-butyl), 96.1 (d,
1JCF = 174 Hz, C3), 122.3 (C400), 129.2 (C500), 130.1
(C30 and C50), 131.4 (C20 and C60), 133.8 (d,
3
4JCF = 2 Hz, C10), 134.0 (d, JCF = 7 Hz, C1), 135.3 (d,
4
2JCF = 17 Hz, C2), 135.6 (C40), 138.7 (d, JCF = 4 Hz,
C200); UV kmax (MeOH) nm (e): 249.6 (14,300); HRMS
(ESI-TOF, positive mode): calcd for C16H19ClFN2
[M+H]+ 293.1221, found 293.1221. ( )-IMI-F (27 mg)
was subjected to chiral HPLC under the following con-
ditions: column, Chiralpak AD-H (250 · 10 mm, Dai-
cel); solvent, 4% 2-propanol in n-hexane; flow rate,
3.0 mL minꢀ1; detection, 254 nm. The materials at tR
28.3 and 44.8 min were collected to give (ꢀ)-IMI-F
(10 mg) and (+)-IMI-F (10 mg), respectively, with an
26
optical purity of each 99.9%. (ꢀ)-IMI-F: ½aꢁ ꢀ70.1
D
26
D
(MeOH, c 1.23); (+)-IMI-F: ½aꢁ +78.2 (MeOH, c 0.99).
4.5. Synthesis of IMI-H
4.5.1. 1-(4-Chlorophenyl)-2-(1H-imidazol-1-yl)-4,4-dim-
ethylpentan-1-one (9). The EtOH (4 mL) solution of 3
(1.50 g, 4.9 mmol, see supplementary data), 1H-imidaz-
ole (0.566 g, 8.3 mmol), and Na2CO3 (0.876 g,
8.3 mmol) were refluxed for 2 h. The mixture was
poured into water (50 mL) and extracted with EtOAc
(50 mL · 3). The combined organic layer was washed
with brine, dried over Na2SO4, and concentrated in va-
cuo. The residual oil was purified by silica gel column
chromatography with 60% EtOAc in hexane to obtain
4.5.4. (E)-1-(1-(4-Chlorophenyl)-4,4-dimethylpent-1-en-2-
yl)-1H-imidazole (IMI-H). The stirred solution of 11
(76 mg, 276 mmol) in acetone (7 mL) in a Pyrex round
bottom flask was irradiated with UV light (365 nm,
UVP B-100A) for 7 h. After concentration in vacuo,
the residue was purified by silica gel column chromatog-
raphy with 40% EtOAc in hexane to obtain IMI-H
(26 mg, 95 mmol, 34%) as a colorless oil. 1H NMR
(500 MHz, CDCl3): d 0.69 (9H, s, tert-butyl), 2.75
(2H, s, H2-3), 6.78 (1H, s, H-1), 7.05 (1H, s, H-400 or
H-500), 7.36 (2H, m, H-20 and -60), 7.39 (2H, m, H-30
and -50), 7.43 (1H, s, H-500 or H-400), 7.99 (1H, s, H-
200); 13C NMR (125 MHz, CDCl3): d 29.9 (methyls of
1
9 (0.634 g, 2.2 mmol, 45%) as a pale red oil. H NMR
(270 MHz, CDCl3): d 0.89 (9H, s, tert-butyl), 1.90
(1H, dd, J = 14.6 and 8.6 Hz, H-3), 2.16 (1H, dd,
J = 14.6 and 4.6 Hz, H-3), 5.73 (1H, dd, J = 8.6 and
4.6 Hz, H-2), 7.06 (1H, s, H-400 or H-500), 7.07 (1H, s,
H-500 or H-400), 7.47 (2H, m, H-20 and H-60), 7.62 (1H,

Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
3151
tert-butyl), 33.7 (tertiary carbon of tert-butyl), 42.8 (C3),
120.4 (C500 or C400), 125.7 (C1), 129.4 (C400 or C500), 129.7
(C30 and C50), 131.7 (C20 and C60), 134.3 (C10 or C40),
135.6 (C40 or C10), 137.5 (C200), 138.8 (C2); UV kmax
(MeOH) nm (e): 254.6 (13,300); HRMS (ESI-TOF, po-
sitive mode): calcd for C16H20N2Cl [M+H]+ 275.1315,
found 275.1295.
fore each set of measurements. Inhibition constants were
determined using the Enzyme Kinetics module of Sig-
maPlot 10 software35 after determining the mode of
inhibition by plotting the reaction velocities in the pres-
ence and absence of inhibitor on a double-reciprocal
plot. Values are reported as mean values with standard
errors of the entire datasets. For the non-inhibited enzy-
matic reaction, the KM for (+)-ABA was calculated to be
0.71 0.08 lM, based on 16 separate experiments. All
tests were conducted at least three times.
4.6. The CD analysis of IMI-OH and IMI-F
(ꢀ)-IMI-OH: CD kext (MeOH) nm (De):265 (ꢀ2.6), 249
(0), 231 (+2.4); (+)-IMI-OH: CD kext (MeOH) nm (De):
265 (+2.6), 249 (0), 231 (ꢀ2.3); (ꢀ)-IMI-F: CD kext
(MeOH) nm (De): 267 (ꢀ3.2), 242 (0), 229 (+1.5);
(+)-IMI-F: CD kext (MeOH) nm (De): 267 (+2.6), 242
(0), 230 (ꢀ1.2).
4.9. Computational method
4.9.1. UNI-X and IMI-X. All the minimum-energy con-
formers of UNI and IMI analogs were generated and min-
imized using MM3 combined with the molecular
dynamics simulation built into CAChe 3.11.36 These
MM3-minimized structures were fully optimized with
density functional theory, using the Becke three parame-
ter hybrid functional (B3LYP) method and the 6-31G(d)
basis set in Gaussian 03,37 followed by a calculation of the
harmonic vibrational frequencies at 298 K at the same le-
vel. The single point energies were calculated with
B3LYP/6-311++G(2df,2p) level of theory. The zero-
point energies were scaled by 0.9804.38 The geometries
and energies in aqueous solution were calculated using
the Onsager model and the PCM model in Gaussian 03.
4.7. X-ray crystallographic analysis of (ꢀ)-UNI-OH, (+)-
UNI-F, and (+)-IMI-OH
(ꢀ)-UNI-OH (3.6 mg) was dissolved in MeOH
(0.18 mL), followed by adding H2O. The solution was
left at room temperature to obtain colorless cubic. (+)-
UNI-F (115 mg) was dissolved in MeOH (10 mL), fol-
lowed by adding H2O (10 mL). The solution was left
at room temperature to obtain colorless prisms. (+)-
IMI-OH (6 mg) was dissolved in MeOH (0.3 mL), fol-
lowed by adding H2O. The solution was left at room
temperature to obtain colorless cubic. These structures
have been deposited in the Cambridge Crystallographic
Data Centre, reference number CCDC 668250 (UNI-
OH), CCDC 668251 (IMI-OH), and CCDC 668252
(UNI-F). These data can be obtained free of charge
via www.ccdc.cam.ac.uk/data_request/cif, or by email-
ing data_request@ccdc.cam.ac.uk, or by contacting
The Cambridge Crystallographic Data Centre, 12, Un-
ion Road, Cambridge CB2 1EZ, UK; fax: +44 1223
336033.
4.9.2. Azole–heme complex model. In the heme model,
ferric iron (FeIII) and methyl thiolate as the fifth ligand
were used. The propionate side chains of protoporphy-
rin were protonated in this model. In the initial struc-
tures, the methyl group of azoles was oriented to the
direction of the propionate side chain of the heme.
The dihedral angle Np–Fe–Naz–Caz and Fe–Naz distance
were set to 45ꢁ and 2.1 A.39 The geometry optimizations
˚
were performed at the B3LYP model with the 6-
31+G(d) basis set for C, H, O, N, and S, and LanL2DZ
basis set for Fe in the spin doublet state. For the azole
complex, BSSE corrections were applied during the
geometry optimizations.
4.8. Enzyme assays
Kinetic analysis was performed using the detailed proto-
cols described previously.15,34 A reaction mixture con-
taining 25 lg mLꢀ1 CYP707A3 microsomes, (+)-ABA
(final conc.: 0.2–8.0 lM), inhibitors (0 for control, 2–
500 nM, in 2 lL DMF), and 50 lM NADPH in
50 mM potassium phosphate buffer (pH 7.25) was incu-
bated for 10 min at 30 ꢁC. Reactions were initiated by
adding NADPH and stopped by addition of 50 lL of
1 M NaOH. The reaction mixtures were acidified with
100 lL of 1 M HCl. To extract the reaction products,
the mixtures were loaded onto Oasis HLB cartridges
(1 mL, 30 mg; Waters) and washed with 1 mL of 10%
MeOH in H2O containing 0.5% AcOH. The enzyme
products were then eluted with 1 mL of MeOH contain-
ing 0.5% AcOH, and the eluate was concentrated in va-
cuo. The dried sample was dissolved in 50 lL of MeOH,
and 10 lL was subjected to HPLC. HPLC conditions
were: ODS column, Hydrosphere C18 (150 · 6.0 mm,
YMC); solvent, 40% MeOH in H2O containing 0.1%
AcOH; flow rate, 1.0 mL minꢀ1; detection, 254 nm. En-
zyme activity was confirmed by determining the
amounts of phaseic acid (PA) in control experiments be-
4.10. The retention factor (k)
Column, Hydrosphere C18 (S-5 lm, 12 nm, 150 ·
6 mm); mobile phase, 70% MeOH; flow rate,
1.0 mL minꢀ1; detection, 254 nm. The retention factor
(k) can be calculated according to the following equa-
tion: k = (tR ꢀ tM)/tM, where tM is the dead time that
is taken as the first deviation of the baseline following
a 5 lL injection of 1% sodium nitrite solution and tR
is the retention time.
Acknowledgments
We thank Toray Industries Inc., Tokyo, Japan, for the
gift of (+)-ABA. Part of this research was carried out
using an instrument at the Center for Instrumental
Analysis of Shizuoka University. This research was sup-
ported by a Grant-in-Aid for Scientific Research (No.
18380192) from the Ministry of Education, Culture,
Sports, Science and Technology of Japan.

3152
Y. Todoroki et al. / Bioorg. Med. Chem. 16 (2008) 3141–3152
16. Kushiro, T.; Okamoto, M.; Nakabayashi, K.; Yamag-
Supplementary data
ishi, K.; Kitamura, S.; Asami, T.; Hirai, N.; Koshiba,
T.; Kamiya, Y.; Nambara, E. ENBO J. 2004, 23, 1647–
1656.
The synthetic procedures for all compounds, including
known compounds and their spectroscopic data, and
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmc.2007.12.019.
17. Yang, S. H.; Choi, D. Biochem. Biophys. Res. Commun.
2006, 350, 685–690.
18. Yang, S. H.; Zeevaart, J. A. D. Plant J. 2006, 47, 675–686.
19. Millar, A. A.; Jacobsen, J. V.; Ross, J. J.; Helliwell, C. A.;
Poole, A. T.; Scofield, G.; Reid, J. B.; Gubler, F. Plant J.
2006, 45, 942–954.
References and notes
20. Destefano-Beltran, L.; Knauber, D.; Huckle, L.; Suttle, J.
C. Plant Mol. Biol. 2006, 61, 687–697.
21. Umezawa, T.; Okamoto, M.; Kushiro, T.; Nambara, E.;
Oono, Y.; Seki, M.; Kobayashi, M.; Koshiba, T.; Kamiya,
Y.; Shinozaki, K. Plant J. 2006, 46, 171–182.
22. Okamoto, M.; Kuwahara, A.; Seo, M.; Kushiro, T.;
Asami, T.; Hirai, N.; Kamiya, Y.; Koshiba, T.; Nambara,
E. Plant Physiol. 2006, 141, 97–107.
1. Funaki, Y.; Ishiguri, T.; Kato, T.; Tanaka, S. J. Pesticide
Sci. 1984, 9, 229–236.
2. Funaki, Y.; Oshita, H.; Yamamoto, S.; Tanaka, S.; Kato, T.
(Sumitomo Chemical Co.). Ger. Offen. DE3010560, 1980.
3. Izumi, K.; Kamiya, Y.; Sakurai, A.; Oshio, H.; Takahashi,
N. Plant Cell Physiol. 1985, 26, 821–827.
23. Katagi, T.; Mikami, N.; Matsuda, T.; Miyamoto, J. J.
Pesticide Sci. 1987, 12, 627–633.
4. Iwasaki, T.; Shibaoka, H. Plant Cell Physiol. 1991, 32,
1007–1014.
24. Meyer, N.; Zeeh, B.; Buschmann, E.; Pommer, E. H. Eur.
Pat. Appl. EP 57365, 1982.
25. Funaki, Y.; Oshita, H.; Yamamoto, S.; Tanaka, S.; Kato,
T. (Sumitomo Chemical Co.). Jpn. Kokai Tokkyo Koho.
JP 54041875, 1979.
5. Asami, T.; Mizutani, M.; Fujioka, S.; Goda, H.; Min, Y.
K.; Shimada, Y.; Nakano, T.; Takatsuto, S.; Matsuyama,
T.; Nagata, N.; Sakata, K.; Yoshida, S. J. Biol. Chem.
2001, 276, 25687–25691.
6. Izumi, K.; Nakagawa, S.; Kobayashi, M.; Oshio, H.;
Sakurai, A.; Takahashi, N. Plant Cell Physiol. 1988, 29,
97–104.
26. Tanaka, S. Kagaku to Seibutsu 1986, 24, 180–182, in
Japanese.
27. Takano, H.; Oguri, Y.; Kato, T. J. Pesticide Sci. 1986, 11,
373–378.
7. Kitahata, N.; Saito, S.; Miyazawa, Y.; Umezawa, T.;
Shimada, Y.; Min, Y. K.; Mizutani, M.; Hirai, N.;
Shinozaki, K.; Yoshida, S.; Asami, T. Bioorg. Med. Chem.
2005, 13, 4491–4498.
8. Saito, S.; Okamoto, M.; Shinoda, S.; Kushiro, T.; Kos-
hiba, T.; Kamiya, Y.; Hirai, N.; Todoroki, Y.; Sakata, K.;
Nambara, E.; Mizutani, M. Biosci. Biotechnol. Biochem.
2006, 70, 1731–1739.
28. Middleton, W. J. Org. Chem. 1975, 40, 574–578.
29. Podust, L. M.; Poulos, T. L.; Waterman, M. R. Proc.
Natl. Acad. Sci. U.S.A. 2001, 98, 3068–3073.
30. Marvin, ChemAxon, http://www.chemaxon.com/; Csizm-
adia, F.; Tsantili-Kakoulidou, A.; Panderi, I.; Darvas, F.
J. Pharm. Sci. 1997, 86, 865.
31. Ballard, S. A.; Lodola, A.; Tarbit, M. H. Biochem.
Pharmacol. 1988, 37, 4643–4651.
32. Catalan, J.; Abboud, J. L. M.; Elguero, J. Adv. Heterocycl.
Chem. 1987, 41, 187–274.
33. Desiraju, G.; Steiner, T. The Weak Hydrogen Bond;
Oxford University Press: Oxford, 1999.
34. Ueno, K.; Yoneyama, H.; Mizutani, M.; Hirai, N.;
Todoroki, Y. Bioorg. Med. Chem. 2007, 15, 6311–
6322.
35. SigmaPlot 10; Systat Software, Inc.: Richmond, CA, 2006.
36. CAChe, 3.11; Oxford Molecular Ltd: London, 1998.
37. Gaussian 03 Revision D.01; Gaussian, Inc.: Wallingford,
CT, 2004.
38. Bauschlicher, C. W., Jr.; Partridge, H. J. Chem. Phys.
1995, 103, 1788–1791.
9. Hirai, N. Abscisic acid. In Comprehensive Natural Prod-
ucts Chemistry; Mori, K., Ed.; Elsevier: Amsterdam, 1999;
Vol. 8, pp 72–91.
10. Davies, W. J.; Jones, H. G. Abscisic Acid; BIOS Scientific
Publishers: Oxford, 1991.
11. Zeevaart, J. A. D.; Creelman, R. A. Annu. Rev. Plant
Physiol. Plant Mol. Biol. 1998, 39, 439–473.
12. Bruzzone, S.; Moreschi, I.; Usai, C.; Guida, L.; Damonte,
G.; Salis, A.; Scarfi, S.; Millo, E.; De Flora, A.; Zocchi, E.
Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5759–5764.
13. Huang, D.; Jaradat, MR.; Wu, W.; Ambrose, S. J.; Ross, A.
R.; Abrams, S. R.; Cutler, A. J. Plant J. 2007, 50, 414–428.
14. Nambara, E.; Marion-Poll, A. Annu. Rev. Plant. Biol.
2005, 56, 165–185.
15. Saito, S.; Hirai, N.; Matsumoto, C.; Ohigashi, H.; Ohta,
D.; Sakata, K.; Mizutani, M. Plant Physiol. 2004, 134,
1439–1449.
39. Rupp, B.; Raub, S.; Marian, C.; Ho¨ltje, H.-D. J. Comp.
Mol. Des. 2005, 19, 149–163.