Analysis of isomerization process of 8'-hydroxyabscisic acid and its 3'- fluorinated analog in aqueous solutions

Article abstract of DOI:10.1016/S0040-4020(00)00068-5

8'-Hydroxyabscisic acid (8'-HOABA), the first metabolite of abscisic acid (ABA) in plants, is spontaneously isomerized to low bioactive phaseic acid (PA). We investigated thermodynamic and kinetic properties of the isomerization process in aqueous solution buffered at the various pHs, along with the effects of 3'-fluorine. The 8'-HOABA/PA ratio at equilibrium was 2:98 at 25°C, while the 3'-fluoro-8'-HOABA/3'α-fluoro-PA ratio was 16:84, indicating that introduction of a fluorine at C-3' thermodynamically reduced isomerization of 8'-HOABA to PA. The isomerization became more rapid as pH increased; the rate constant at pH 10 was higher than that at pH 3 by a factor of 2000. Introduction of a fluorine at C-3' reduced the reaction rate by raising the activation enthalpy. This indicated that 8'-HOABA was also kinetically stabilized by the 3'-fluorine. These findings suggested that the control of pH and modification of the enone moiety of the ring would be useful to manipulate the catabolic inactivation rate of ABA. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00068-5

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 1649–1653
Analysis of Isomerization Process of 80-Hydroxyabscisic Acid and
its 30-Fluorinated Analog in Aqueous Solutions
*
Yasushi Todoroki, Nobuhiro Hirai and Hajime Ohigashi
Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan
Received 14 December 1999; accepted 21 January 2000
Abstract—80-Hydroxyabscisic acid (80-HOABA), the first metabolite of abscisic acid (ABA) in plants, is spontaneously isomerized to low
bioactive phaseic acid (PA). We investigated thermodynamic and kinetic properties of the isomerization process in aqueous solution buffered
at the various pHs, along with the effects of 30-fluorine. The 80-HOABA/PA ratio at equilibrium was 2:98 at 25ЊC, while the 30-fluoro-
80HOABA/30a-fluoro-PA ratio was 16:84, indicating that introduction of a fluorine at C-30 thermodynamically reduced isomerization of 80-
HOABA to PA. The isomerization became more rapid as pH increased; the rate constant at pH 10 was higher than that at pH 3 by a factor of
2000. Introduction of a fluorine at C-30 reduced the reaction rate by raising the activation enthalpy. This indicated that 80-HOABA was also
kinetically stabilized by the 30-fluorine. These findings suggested that the control of pH and modification of the enone moiety of the ring
would be useful to manipulate the catabolic inactivation rate of ABA. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
equivalent to those of 1 in some assays, but weaker than
those of 1 in others. Thus, it was difficult to precisely esti-
mate the net biological activities of 2, probably because the
stability of 2 in assay media was affected by factors such as
pH and the temperature. Despite such uncertainty, the
potency of 2 in some assays suggested that this compound
would have greater activity than 3. Another estimate of the
activity of 2 was afforded by the analog 30-fluoro-80-hydroxy-
ABA (6), which undergoes less isomerization than 2 (Fig.
1).8,9 The analog 6 was sufficiently stable to be isolated with-
out derivatization, and its activities were intermediate
between those of 30-fluoro-ABA (5) and 30a-fluorophaseic
acid (7).8 This suggested that the activity of 2 may be also
intermediate between those of 1 and 3. These findings showed
that 2 retains medium bioactivity, but loses it due to isomeri-
zation to 3. Thus, the isomerization of 2 to 3 is a key step in
the inactivation of 1, so characterizing the isomerization of 2
to 3 is indispensable to understand the inactivation process.
Abscisic acid (ABA, 1) is the primary hormone that induces
adaptive reactions to protect plants from desiccation and
low temperature,1,2 as well as for regulation of seed germi-
nation.3 In plants, the C-80 of 1 is hydroxylated by an
80-hydroxylase, which has been reported to be cytochrome
P450 monooxygenase,4 to give 80-hydroxy-ABA (2).1
Metabolite 2 isomerizes to phaseic acid (3) with a low
bioactivity before reduction to biologically inactive dihy-
drophaseic acid (4) (Fig. 1).1 Evaluation of the biological
activity of 2 is necessary to understand the inactivation
process of 1. Although the isomerization of 2 to 3 might
occur enzymatically in vivo,5 it does so spontaneously in
vitro, so it has not been easy to isolate 2 without 80-O-
protection, much less examine its bioactivity. Zou et al.6
and Walker-Simmons et al.7 reported that the activities of
the borate complex of 2 that dissociates in water were
Figure 1. Metabolic inactivation pathway of 1 and 5 in plants.
Keywords: isomerization; kinetics; metabolites; enones.
* Corresponding author. Fax: ϩ81-75-753-6284; e-mail: hirai@kais.kyoto-u.ac.jp
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00068-5

1650
Y. Todoroki et al. / Tetrahedron 56 (2000) 1649–1653
The isomerization of 2 to 3 is a cyclization reaction
triggered by intramolecular conjugated nucleophilic
addition of the oxygen of the 80-hydroxyl group to the elec-
tron-deficient C-20, which is present in the b-position of the
a,b-unsaturated carbonyl group. This reaction should be
promoted by deprotonation of the 80-hydroxyl group or
protonation on the 40-carbonyl oxygen. This means that
the reaction rate can be affected by pH of media. In fact,
2 was more stable under acidic than alkaline conditions.6
The electronic state of the enone moiety in the ring of 2
will also be a significant factor affecting the isomerization
rate. Previously, we found that the methyl ester of 6 slowly
isomerized to the methyl esters of 7 and 30b-fluorophaseic
acids (8).8,9 However, quantitative analysis of the isomeri-
zation process has not been reported. In the present study,
we examined the equilibrium ratios and the isomerization
rates of 2/3 and 6/7 in aqueous solution buffered at various
pHs and temperatures. The thermodynamic and kinetic
characteristics of the isomerization process are discussed
on the basis of computer-aided molecular orbital analysis
of model compounds in addition to experimental results.
It was necessary to determine the ratio of molar absorption
coefficient (e) between 2 and 3, and 6 and 7, to correct the
isomeric ratios based on peak areas in HPLC with the UV-
detector (254 nm). It would be difficult to precisely measure
the absolute ratio of e of these metabolites, especially that of
the labile metabolite 2, under these analytical conditions.
So, we obtained the relative value by comparing the ratio
of peak-area of the two isomers in the mixtures at the
various isomeric ratios (see Experimental). The obtained
relative ratio of e at 254 nm was 1.18 for 2/3, and 1.27
for 6/7. Peak areas of 2 and 3, and 6 and 7 were corrected
by the relative ratio of e to give the precise molar ratio of the
each isomeric pair.
Equilibrium ratios
The metabolites 2 and 6 were incubated at three tempera-
tures (5, 15 and 25ЊC; and 15, 25 and 35ЊC, respectively)
and in eight buffer solutions (pH 3–10). After a three-week
incubation, the ratio of isomers was measured by HPLC
with the above relative ratio of e. The 2/3 ratio was 2:98
under all pH and temperature conditions used; the free
energy difference (DGЊ) was Ϫ2.3 kcal/mol. The 6/7/8
ratio changed according to temperature, but not according
to pH. The isomer 8 with the axial fluorine was not found
under any conditions, probably because its axial fluorine is
1,3-diaxial to the side-chain and cannot be thermodynamic-
ally labilized. The 6/7 ratio was 14:86 (DGЊˆϪ1.03 kcal/
mol) at 15ЊC, 16:84 (DGЊˆϪ0.98 kcal/mol) at 25ЊC, and
18:82 (DGЊˆϪ0.92 kcal/mol) at 35ЊC; that is, the propor-
tion of 6 became larger as temperature increased. Thus, the
isomerization of 6 to 7 had negative entropy (DSЊϽ0),
which was estimated to be about Ϫ5.5 cal/mol deg if the
enthalpy (DHЊ) and DSЊ were little affected by alterations
in temperature. Isomerization formed the bicyclo[3.2.1]-
octane ring, so the negative entropy could be attributed to
a decrease in the flexibility. This must be true also of the
isomerization of 2 to 3, meaning that the 2/3 ratio should
also be dependent on temperature. However, we found no
such temperature dependency for the 2/3-equilibrium ratio.
If DSЊ for the isomerization of 2 to 3 was similar to that for
the isomerization of 6 to 7, then the population of 3 at 5ЊC
should be about 1.7%, which is smaller than that at 25ЊC by
only 0.3 points. Such a difference was too small to be
detected by our HPLC analysis.
Results and Discussion
Preparation of 2 and 6
1
Precise analysis of the H NMR spectrum of 3 showed the
presence of small signals corresponding to 2 at an intensity
of 1–2% of those of 3, confirming that a methanol solution
of 3 is actually a tautomeric mixture of 3 and a trace amount
of 2. Careful analysis of the methanol solution of 3 by HPLC
with an ODS column revealed the existence of a small peak
that eluted after 3.4 The compound corresponding to the
peak was collected, and it showed signals similar to those
of 1 on 1H NMR, except the absence of the methyl protons,
H3-80, and existence of an AB quartet at d 3.56 and 3.70,
which corresponds to the methylene protons adjacent to an
oxygen atom (Table 1). This compound was partially
converted to 3 on heating during concentration. Therefore,
this compound was identified as 2. The isolated 2 isomer-
ized slowly even in acidic solutions at Ϫ20ЊC, so it was used
for kinetic analysis of the isomerization as soon as possible
after isolation by HPLC, without concentration. Metabolite
6 was isolated from a rice cell suspension culture fed with
5.8
The above findings showed that the introduction of a
fluorine at C-30 stabilizes the 80-hydroxy compound
compared to the cyclized compound. We performed ab
initio molecular orbital calculations in the MP2/6-
311ϩG(d,p)//HF/6-31G(d) level, for the model reactions:
ethane!fluoroethane; and ethene!fluoroethene. This
substitution reaction by fluorine was more exoergic for
ethene than for ethane by ca. 2 kcal/mol. This agreed with
the observation that 2 was better stabilized than 3 by the
C-30 fluorine. This stabilization may have been due to
delocalization of electrons or less non-bonded repulsion in
sp2 hybridization compared to sp3.
Table 1. The 1H NMR signals of 1, 2, and 3 in methanol-d4 (values for the
chemical shifts are in d (ppm), and those for the coupling constants in
parentheses are in J (Hz)
H
1
2
3
2
4
5
6
30
5.74 (s)
5.84 (s)
5.79 (s)
7.77 (dd, 16.2, 0.6)
6.23 (dd, 16.2, 0.4)
2.03 (s)
7.56 (d, 16.1)
5.92 (d, 16.1)
1.94 (s)
8.10 (d, 16.0)
6.45 (d, 16.0)
1.94 (s)
2.70 (dd, 18.0, 2.4)
2.80 (d, 18.0)
2.39 (dd, 18.0, 2.4)
2.47 (dd, 18.0, 2.8)
1.22 (s)
5.92 (s)
5.90 (s)
50
2.18 (dd, 16.9, 0.5)
2.53 (d, 16.9)
1.93 (s)
2.37 (d, 17.2)
2.48 (d, 17.2)
1.91 (s)
3.56 (d, 11.1)
3.70 (d, 11.1)
1.06 (s)
70
80
Kinetic analysis
1.06 (s)
3.67 (d, 7.7)
3.95 (dd, 7.7, 2.8)
1.01 (s)
The isomerization rates for 2 and 6 were measured under
the same conditions as the equilibrium ratios. Under all
90
1.03 (s)

Y. Todoroki et al. / Tetrahedron 56 (2000) 1649–1653
Table 2. Rate constants (k) and activation energies of the isomerization at 25ЊC
1651
pH
2!3
k (/s)
5!6
k (/s)
DG‡ (kcal/mol)
DH‡ (kcal/mol)
DS‡ (cal/mol deg)
DG‡ (kcal/mol)
DH‡ (kcal/mol)
DS‡ (cal/mol deg)
3
4
5
6
7
8
9
10
6.8×10Ϫ6
8.7×10Ϫ6
1.1×10Ϫ5
1.8×10Ϫ5
5.5×10Ϫ5
1.7×10Ϫ4
2.5×10Ϫ3
1.1×10Ϫ2
24.5
24.4
24.2
23.9
23.3
22.6
21.0
19.9
10.6
11.2
12.2
12.2
12.4
17.0
18.9
16.7
Ϫ46.9
Ϫ44.3
Ϫ40.3
Ϫ39.3
Ϫ36.4
Ϫ18.9
Ϫ7.1
4.8×10Ϫ8
9.0×10Ϫ8
1.6×10Ϫ7
5.0×10Ϫ7
3.5×10Ϫ6
1.6×10Ϫ5
3.7×10Ϫ4
2.0×10Ϫ3
27.6
27.1
26.8
26.1
24.9
24.1
23.6
21.2
16.2
17.8
19.9
20.5
20.0
21.6
23.5
20.2
Ϫ39.8
Ϫ33.3
Ϫ25.3
Ϫ20.9
Ϫ18.6
Ϫ10.2
Ϫ2.4
Ϫ10.9
Ϫ5.3
conditions examined, isomerization was observed as a first-
order reaction in which the rate was proportional only to the
concentration of the 80-hydroxyl compound. As the pH and
temperature increased, the reaction proceeded more rapidly.
At 25ЊC, the half-life of 2 was 30 h at pH 3, 4 h at pH 7, and
shorter than 1 min at pH 10; that is, 2 was isomerized to 3
more rapidly at pH 10 than at pH 3 by a factor of 2000. The
temperature dependence of the rate was greater under alka-
line conditions than under acidic conditions. The Arrhenius
plots of the rate constants gave the activation energies of
Arrhenius and frequency factors, which were converted to
the kinetic parameters, i.e. the activation enthalpy (DH‡),
activation entropy (DS‡) and activation free energy (DG‡)
(Table 2). The DH‡ was higher under alkaline conditions
than under acidic conditions, whereas the DS‡ was negative
under all conditions and its absolute value was larger under
the acidic as compared to the alkaline conditions. This indi-
cated that the slow isomerization under acidic conditions
was dependent on the negative, large DS‡. The transition
structure under acidic conditions may be so solvated as to
cause a decrease in DH‡ and a rise in the absolute value of
DS‡. Alternatively, the reaction under acidic conditions may
be associated with a concerted mechanism.
Introducing a fluorine into the conjugated system should
have a large effect on the charge distribution. The fluorine
at C-30 can push the electrons of C-30 to C-20 by the electron-
donating effect, which is attributed to repulsion between the
p-electron at C-30 and the electrons of the outermost shells
of the fluorine atom, and also withdraw those of the 40-car-
bonyl by the inductive effect. In addition, the fluorine itself
will significantly affect polarity of the molecule. In the equi-
librium state, these effects of fluorine are observed experi-
mentally by the shifts of 13C NMR signals10 and
theoretically by the atomic charges calculated by Mulliken
population analysis.11 Our previous study revealed that the
13C signals corresponding to C-20 and C-40 in 5 were shifted
toward a higher field compared to those in 1, while C-30 in 5
was shifted toward a lower field compared to that in 1.9 This
indicated that introduction of a fluorine at C-30 in 1 or 2
increased the electron density at C-20 and C-40 and
decreased that of C-30. The charges of the atoms corre-
sponding to C-20, C-30, C-40 and O-40 in 2 and 6 were
estimated using the model compounds 11 and 12 at the
MP2/6-31G* level (Table 3). The results indicated that the
charges of C-20 and C-40 in 6 were shifted toward negative
compared to those of 2, while those of C-30 and O-40 in 6
were shifted toward positive compared to those of 2, in
agreement with the 13C shifts observed. The increase in
the negative charge at C-20 will result in the larger filled-
orbital-filled-orbital repulsion for a nucleophile. The repul-
sion between C-20 and O-80 can be represented by the small
overlap population which is calculated in the transition
state, so we simulated the nucleophilic addition of a meth-
oxide to models 11 and 12 (Fig. 2). The calculated overlap
population, 0.105, between the 20-carbon of 11 and the
oxygen of methoxide in the transition-state structure was
smaller than that for 12 (0.080), as expected. These results
suggested that the slow isomerization of 6 is dependent on
the electron-donating effect on C-20 by the 30-fluorine via C-30.
The isomerization of 6 was slower than that of 2 by a factor
of 10–100 under acidic conditions and by a factor of 3–10
under alkaline conditions. This indicated that the introduc-
tion of a fluorine at C-30 stabilized 2 kinetically in addition
to thermodynamically. The pH-dependence of the isomeri-
zation rate of 6 was similar to that of 2; at 25ЊC, the reaction
rate constant at pH 10 was 50000-fold larger than that at pH
3. At the same pH, the DH‡ of the isomerization of 6 was
higher than that of the isomerization of 2, while the absolute
value of the negative DS‡ of the isomerization of 6 was
smaller than that of 2. This meant that the slow reaction
of 6 was caused by an increase of the DH‡, unlike that in
low pH solution. This suggested a decrease in solvation at
the transition state of 6/7 isomerization. However, a better
explanation would be less interaction between the 20-carbon
and 80-oxygen of the transition state of 6/7 isomerization
compared to 2/3 isomerization. This was examined by
computer-assisted calculations for the model compounds.
Here, we demonstrated that introduction of a fluorine at C-30
of 2 can stabilize the 80-hydroxyl compound kinetically as
well as thermodynamically. The control of pH and the
Table 3. Mulliken charges of the heavy atoms in 11 and 12 (calculated in
the MP2/6-31Gء//HF/3-21Gء)
Compound
C-20
C-30
C-40
O-40
11
12
Ϫ0.142
Ϫ0.254
Ϫ0.278
0.383
0.533
0.484
Ϫ0.554
Ϫ0.518
Figure 2. Model compounds 11 (XˆH) and 12 (XˆF), and the nucleophilic
addition of a methoxide to those.

1652
Y. Todoroki et al. / Tetrahedron 56 (2000) 1649–1653
modification of the enone moiety of the ring will be useful to
manipulate the catabolic rate of ABA.
could be obtained by solving the following equation:
pˆꢀAxϪAy†=ꢀByϪBx†
where Ax and Bx are the areas of the isomer A and B in
sample x, respectively, and Ay and By are the areas of the
isomer A and B in sample y, respectively. The experiment
for obtaining the equilibrium ratio was repeated five times,
and the mean value was adopted. The standard errors were
less than 0.3 point in all cases.
Experimental
General procedures
1H NMR spectra were recorded with TMS as an internal
standard using a Bruker ARX500 (500 MHz) apparatus.
For clarity, the conventional ABA numbering system was
used to assign the peaks in the 1H NMR spectra. The carbon
corresponding to that of ABA in the model compounds 11
and 12 was represented by the same number as that of ABA.
Kinetics and activation energies
Values for the first-order rate constant (k) of overall reaction
were determined by fitting to the equation:
ln ‰ꢀAϪAe†=ꢀAoϪAe†ŠˆϪkt
Preparation of 2, 3, 6 and 7
where A is the concentration of 2 or 6 in sampled solution;
Ao is the initial concentration of 2 or 6; Ae is the equilibrium
concentration of 2 or 6; and t is the incubation time. The
correlation coefficient was higher than 0.996 in all cases.
The rate constants (k1) of the reaction 2!3 and 6!7 were
calculated according to
The natural metabolite 3 (3 mg) was prepared by hydrolysis
of the b-hydroxy-b-methylglutaryl ester of 80-hydroxy-
ABA isolated from immature seeds of Robinia
pseudacacia,12 and analyzed by HPLC with an ODS column
YMC AQ-311, 6 i.d.×100 mm (solvent: 45% MeOH in
0.1% aqueous AcOH at 1.0 ml minϪ1; detection: 254 nm).
After 3 eluted at tR 5.9 min, a compound corresponding to a
small peak of tR 7.2 min was collected, and concentrated for
identification to give a mixture (30 mg) of 2 and 3 in a ratio
k1ˆkK=ꢀKϩ1†
where K is the equilibrium constant. The Arrhenius plot of
k1 at the various temperatures determined the activation
energy (Ea) of Arrhenius. The correlation coefficients
were higher than 0.996 in all cases. The activation enthalpy
(DH‡) was calculated according to
1
of 70:30. Signals of 2 in the H NMR spectrum (500 MHz,
CD3OD) are shown in Table 1. Compounds 6 (0.6 mg) and 7
(0.2 mg) were prepared by feeding 5 (4.8 mg) to suspension
cultures of rice cells.9 For isomerization examination, 2 and
6 were collected as eluates by HPLC (40% MeOH for 2 and
45% MeOH for 6, respectively, in 0.1% aqueous AcOH) in a
flask cooled to Ϫ20ЊC, and the solution was preserved
below Ϫ20ЊC before incubation. The concentrations of 2
and 6 in the solutions were 35.4 and 49.1 mM, respectively.
The isomeric purity was always measured by HPLC
analysis just before incubation because 2 was converted to
its tautomeric isomer 3 even at Ϫ20ЊC, although 6 has never
isomerized under these conditions. The isomeric purity was
98.8–89.6% for the solution of 2 and 100% for that of 6.
DH‡ˆEaϪRT
where R is the gas constant; and T is the temperature. The
activation entropy (DS‡) was calculated according to
DS‡ˆR ln ꢀAh=ekBT†
where A is the frequency factor obtained by the Arrhenius
plot; h is the Planck constant; e is the base of natural loga-
rithm; and kB is the Boltzman constant. The activation free
energy (DG‡) was calculated according to
DG‡ˆDH‡ϪTDS‡
Incubation of 2 and 6 at various pHs
Computational methods
Buffer solutions used were 10 mM citric acid-20 mM
Na2HPO4 for pH 3–8, and 10 mM NH3–NH4Cl for pH 9
and 10. The initial pHs of buffers were adjusted by con-
sidering the addition of the sample solution containing
0.1% AcOH and the incubation temperature. The sample
solutions (20 ml each) were added to 180 ml of the buffer
solution. Incubation was performed at 5, 15 and 25ЊC for 2,
and at 15, 25 and 35ЊC for 6. Five-microliter aliquots were
analyzed by HPLC with the ODS column (solvent: 45%
MeOH in 0.1% aqueous AcOH at 1.0 ml minϪ1; detection:
254 nm) at appropriate intervals according to the isomeri-
zation rate, and areas of the peaks of 2 and 3, and of 6 and 7
were measured. The 2/3 and 6/7 ratios of areas obtained
were converted to the corresponding ratios of concen-
trations by considering the ratios of molar absorption
coefficients (e), which were obtained by comparing the
peak-areas of the two isomers in sample solutions of various
isomeric ratios as follows. The total amount of isomers was
constant regardless of their ratio, so the relative ratio (p) of e
The free energy difference of the model reactions,
ethane!fluoroethane and ethene!fluoroethene were
calculated using the Gaussian 98 (Revision A.3).13 The
lowest-energy geometries were obtained in the HF/6-
31G(d) level, and characterized in the MP2/6-311ϩG(d,p)
level.
The equilibrium geometries of models 11 and 12 were opti-
mized in the HF/3-21Gء level using PC SPARTAN Pro
(ver.1.01).14 The chemical properties for the optimized
geometries were calculated in the MP2/6-31Gء level using
the Gaussian 98. The transition state geometries of the
model reactions were determined and optimized employing
minimum energy path calculations in the HF/3-21Gء level
using PC SPARTAN Pro. A single imaginary frequency was
found for each calculation of transition states. The
geometries obtained were characterized in the MP2/6-
31Gء level using the Gaussian 98.

Y. Todoroki et al. / Tetrahedron 56 (2000) 1649–1653
1653
9. Todoroki, Y.; Hirai, N.; Ohigashi, H. Tetrahedron 1995, 51,
6911.
Acknowledgements
10. Breitmaier, E.; Voelter, W. Carbon-13 NMR Spectroscopy,
VHC Verlagsgesellschaft: Weinheim, 1987.
11. Mulliken, R. S. Journal of Chemical Physics 1955, 23, 1833.
See also 1841, 2338, 2343.
This research was supported by a Grant-in-Aid (Biocosmos
program) from the Ministry of Agriculture, Forestry and
Fisheries of Japan (BCP-99-I-A-1).
12. Hirai, N.; Fukui, H.; Koshimizu, K. Phytochemistry 1978, 17,
1625.
References
13. Gaussian 98, Revision A.3, Frisch, M. J.; Trucks, G. W.;
Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.;
Zakrzewski, V. G.; Montgomery, Jr., J. A.; Stratmann, R. E.;
Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.;
Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.;
Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A.; Gaussian, Inc.,
Pittsburgh PA, 1998.
1. Hirai, N. Abscisic acid; In Comprehensive Natural Products
Chemistry, Mori, K. Ed.; Pergamon Press: Oxford, 1999; 8, pp
72–91.
2. Leung, J. Annual Review of Plant Physiology and Plant
Molecular Biology 1998, 49, 199.
3. Yoshioka, T.; Endo, T.; Satoh, S. Plant and Cell Physiology
1998, 39, 307.
4. Krochko, J. E.; Abrams, G. D.; Loewen, M. K.; Abrams, S. R.;
Cutler, A. J. Plant Physiology 1998, 118, 849.
5. Milborrow, B. V.; Carrington, N. J.; Vanghan, G. T. Phyto-
chemistry 1988, 27, 757.
6. Zou, J.; Abrams, G. D.; Barton, D. L.; Taylor, D. C.; Pomeroy,
M. K.; Abrams, S. R. Plant Physiology 1995, 108, 563.
7. Walker-Simmons, M. K.; Holappa, L. D.; Abrams, G. D.;
Abrams, S. R. Physiologia Plantarum 1997, 100, 474.
8. Arai, S.; Todoroki, Y.; Ibaraki, S.; Naoe, Y.; Hirai, N.;
Ohigashi, H. Phytochemistry 1999, 52, 1185.
14. PC Spartan pro, version 1.01, Wavefunction, Inc., Irvine CA,
1999.