Structural refinement of inhibitors of urea-based soluble epoxide hydrolases

Article abstract of DOI:10.1016/S0006-2952(02)00952-8

The soluble epoxide hydrolase (sEH) is involved in the metabolism of arachidonic, linoleic, and other fatty acid epoxides, endogenous chemical mediators that play an important role in blood pressure regulation and inflammation. 1,3-Disubstituted ureas, carbamates, and amides are new potent and stable inhibitors of sEH. However, the poor solubility of the lead compounds limits their use. Inhibitor structure-activity relationships were investigated to better define the structural requirements for inhibition and to identify points in the molecular topography that could accept polar groups without diminishing inhibition potency. Results indicate that lipophilicity is an important factor controlling inhibitor potency. Polar groups could be incorporated into one of the alkyl groups without loss of activity if they were placed at a sufficient distance from the urea function. The resulting compounds had a 2-fold higher water solubility. These findings will facilitate the rational design and optimization of sEH inhibitors with better physical properties.

Full text of DOI:10.1016/S0006-2952(02)00952-8

Biochemical Pharmacology 63 02002) 1599±1608
Structural re®nement of inhibitors of urea-based soluble
epoxide hydrolases
Christophe Morisseaua,b, Marvin H. Goodrowa, John W. Newmana,b,
Craig E. Wheelocka,b, Deanna L. Dowdya, Bruce D. Hammocka,b,*
aDepartment of Entomology, University of California, Davis, CA 95616, USA
bU.C.D. Cancer Center, University of California, Davis, CA 95616, USA
Received 4 June 2001; accepted 14 November 2001
Abstract
The soluble epoxide hydrolase 0sEH) is involved in the metabolism of arachidonic, linoleic, and other fatty acid epoxides, endogenous
chemical mediators that play an important role in blood pressure regulation and in¯ammation. 1,3-Disubstituted ureas, carbamates, and
amides are new potent and stable inhibitors of sEH. However, the poor solubility of the lead compounds limits their use. Inhibitor
structure±activity relationships were investigated to better de®ne the structural requirements for inhibition and to identify points in the
molecular topography that could accept polar groups without diminishing inhibition potency. Results indicate that lipophilicity is an
important factor controlling inhibitor potency. Polar groups could be incorporated into one of the alkyl groups without loss of activity if
they were placed at a suf®cient distance from the urea function. The resulting compounds had a 2-fold higher water solubility. These
®ndings will facilitate the rational design and optimization of sEH inhibitors with better physical properties. # 2002 Published by
Elsevier Science Inc.
Keywords: Epoxide hydrolase; Blood pressure; Immune response; Urea inhibitors; Structure±activity relationships; Solubility
1. Introduction
Hydrolysis of the epoxides by sEH diminishes this activity
[7]. sEH hydrolysis of EETs also regulates their incorpora-
tion into coronary endothelial phospholipids, suggesting a
regulation of endothelial function by sEH [10]. In keeping
with this hypothesis, treatment of spontaneous hyperten-
sive rats 0SHRs) with selective sEH inhibitors signi®cantly
reduces their blood pressure [11]. Additionally, male
knockout sEH mice have signi®cantly lower blood pres-
sure than wild-type mice [12], further supporting the role
of sEH in blood pressure regulation.
The EETs have also demonstrated some anti-in¯amma-
tory properties in endothelial cells [13,14]. In contrast,
diols derived from epoxy-linoleate 0leukotoxin) perturb
membrane permeability and calcium homeostasis [5],
which results in in¯ammation that is modulated by nitric
oxide synthase and endothelin-1 [15,16]. Micromolar con-
centrations of leukotoxin reported in association with
in¯ammation and hypoxia [17] depress mitochondrial
respiration in vitro [18] and cause mammalian cardio-
pulmonary toxicity in vivo [15,19,20]. Leukotoxin toxicity
presents symptoms suggestive of multiple organ failure
and acute respiratory distress syndrome 0ARDS) [17]. In
EHs 0EC 3.3.2.3) catalyze the hydrolysis of epoxides or
arene oxides to their corresponding diols by the addition of
water [1]. In mammals, the soluble and microsomal EH
forms are known to complement each other in detoxifying
a wide array of mutagenic, toxic, and carcinogenic xeno-
biotic epoxides [2,3]. sEH is also involved in the metabo-
lism of arachidonic [4] and linoleic [5] acid epoxides,
which are endogenous chemical mediators [6]. Epoxides of
arachidonic acid 0epoxyeicosatrienoic acids or EETs) are
known effectors of blood pressure regulation [7], modulat-
ing vascular permeability [8]. The vasodilatory properties
of EETs are associated with an increased open-state prob-
ability of calcium-activated potassium channels leading to
hyperpolarization of the vascular smooth muscle [9].
* Corresponding author. Tel.: ‡1-530-752-7519; fax: ‡1-530-752-1537.
E-mail address: bdhammock@ucdavis.edu 0B.D. Hammock).
Abbreviations: EH, epoxide hydrolase; sEH, soluble epoxide hydro-
lase; EET, epoxyeicosatrienoic acid; DMF, N,N-dimethylformamide;
QSAR, quantitative structure±activity relationships; t-DPPO, trans-1,3-
diphenylpropene oxide.
0006-2952/02/$ ± see front matter # 2002 Published by Elsevier Science Inc.
PII: S 0 0 0 6 - 2 9 5 2 0 0 2 ) 0 0 9 5 2 - 8

1600
C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
inhibitor and residues of the sEH active site, mimicking
features encountered in the reaction coordinate of epoxide
ring opening by this enzyme [22,23]. These inhibitors
ef®ciently reduced epoxide hydrolysis in several in vitro
and in vivo models [11,21,24]. However, these dialkyl-
ureas have high crystal lattice energies as indicated by their
high melting points, and also have limited solubility in
water [11], which probably affects their in vivo ef®cacy.
Therefore, we investigated the effect of changes in the
structure of disubstituted ureas 0Fig. 1) in relation to their
potency as sEH inhibitors to design powerful inhibitors
with improved physical properties.
Fig. 1. Structures of sEH inhibitors.
both cellular and organismal models, leukotoxin-mediated
toxicity is dependent upon epoxide hydrolysis [5,21],
suggesting a role for sEH in the regulation of in¯ammation.
The bioactivity of these epoxy-fatty acids suggests that
inhibition of vicinal-dihydroxy-lipid biosynthesis may
have therapeutic value, making sEH a promising pharma-
cological target.
2. Materials and methods
We recently described 1,3-disubstituted ureas, carba-
mates, and amides 0Fig. 1) as new potent and stable
inhibitors of sEH. These compounds are competitive
tight-binding inhibitors with nanomolar KI values that
interact stoichiometrically with puri®ed recombinant
sEH [21]. From crystal structure determination, it was
shown that the urea inhibitors established hydrogen bond-
ing and salt bridges between the urea function of the
2.1. Reagents and apparatus
Compound 07) was obtained from Aldrich. The synth-
esis and characterization of compound 027) were described
in a report by Argiriadi et al. [23]. For the other compounds
described for this study, reaction yield, melting point, and
mass spectroscopy data are given in Table 1. Other che-
micals and reagents were purchased from Aldrich or Sigma
Table 1
Reaction yields, melting points 0m.p.), and mass spectraa data for the newly synthesized compounds
No.b Yield 0%) m.p. 08)
MS m/z 0Da)
No. Yield 0%) m.p. 08)
MS m/z 0Da)
tR 0min)c Mth
[M ‡ H] [2M ‡ H]
tR 0min) Mth
[M ‡ H] [2M ‡ H]
d
‡
‡
‡
‡
2
3
84
40
48
88
75
93
99
97
90
92
95
65
94
96
97
86
88
65
81
96
85±86
72±73
6.4
6.3
390.3 310.3
309.3 310.3
182.1 183.1
196.2 197.0
210.2 211.1
238.2 239.1
238.2 239.2
308.3 309.3
276.2 277.2
304.3 305.2
328.3 329.2
342.3 343.3
360.3 361.3
261.2 262.2
233.2 234.0
234.1 235.0
248.2 249.1
274.2 275.2
232.2 233.0
260.2 261.1
619.5 08)e
619.5 02)
24
25
26
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
67
97
84
37
90
70
85
90
82
92
95
74
81
86
71
84
86
72
39
204±205
183±184
223±224
240 dec.
110 dec.
185±186
201 dec.
86±87
1.8
1.8
2.4
2.5
2.7
3.1
2.0
2.3
1.2
1.1
1.3
1.3
1.4
1.9
2.0
3.6
1.6
2.8
4.0
236.1 237.1
473.1 024)
437.2 039)
505.1 019).
593.0 015)
645.3 038)
605.2 013)
218.1 219.0
252.1 253.0
296.1 297.0
322.2 323.2
302.1 303.1
263.1 264.0
226.2 227.1
242.2 243.1
256.2 257.2
237.2 238.1
270.2 271.2
284.2 285.1
298.2 299.2
326.3 327.2
340.3 341.3
325.3 326.3
340.3 341.3
392.3 393.3
4
123±124 1.3
176±177 1.4
223±224 1.6
224±225 1.9
149±150 2.1
180±182 4.7
253±254 3.2
208±210 3.9
365.2 048)
393.3 023)
421.3 031)
477.4 041)
477.3 025)
617.5 016)
553.3 022)
609.5 034)
657.4 08)
5
6
8
9
11
12
13
14
15
16
17
18
19
20
21
22
23
453.4 060)
485.3 010)
513.3 012)
475.3 010)
541.3 012)
569.4 014)
597.4 09)
98±99
140±141
108±109
137±138
130±131
89±90
>270
4.7
253 dec.f 4.0
245 dec. 5.8
190±191 1.0
115 dec. 1.0
212±213 1.3
182±183 1.7
226±227 3.7
203±204 1.9
196±197 3.1
685.4 013)
523.4 09)
467.3 016)
469.3 016)
497.3 019)
549.3 053)
465.3 021)
521.3 027)
133±134
110±111
135±139
94±95
653.4 013)
681.5 011)
651.5 010)
681.4 017)
142±143
a Mass spectra were measured by LC±MS: Waters 2790 liquid chromatograph equipped with a 30 Â 2:1 mm 3 mm C 18 XterraTM column 0Waters) and a
Micromass Quattro Ultima mass spectrometer Methanolic solutions 05 mL of 1±3 mg/mL) of the tested compounds were injected on the column equilibrated
with 60/40/0.1 H2O/CH3CN/HCOOH. Compounds were eluted at a flow rate of 300 mL/min with a linear solvent gradient from 60/40/0.1 H2O/CH3CN/
HCOOH to 100/0.1 CH3CN/HCOOH in 5 min and then maintained in this solvent for 8 min. Positive mode electrospray ionization was used with a capillary
voltage of 1 kV and a cone voltage of 80 V. Mass acquisition 0continuum 50±700 Da) was performed with a scan time of 5 sec and an interscan delay of
0.03 sec.
b Compound number.
c Retention time.
d Theoretical mono-isotopic mass.
‡
‡
e Dimeric ion [2M ‡ H] intensity as a percentage of the monomeric ion [M ‡ H] .
f Decomposition point indicated by the change of color or the failure to recrystallize upon cooling.

C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
1601
and were used without further puri®cation. Melting points
were determined with a Thomas-Hoover apparatus 0A.H.
Thomas Co.) and are uncorrected. Infrared 0IR) spectra were
recorded on a Mattson Galaxy Series FTIR 3000 spectro-
meter. Mass spectra were measured by LC±MS: a Waters
2790 liquid chromatograph equipped with a 30 Â 2:1 mm
3 mm C18 XterraTM column 0Waters) and a Micromass
Quattro Ultima mass spectrometer.
isocyanate was added. After stirring at room temperature
for 24 hr, a white crystalline solid was produced. The
mixture was cooled on chilled water. The solid product
was collected, and washed with cold hexanes providing
0.66 g 088%) of the title product as long white needles,
m.p. 194.0±194.58. IR 0KBr) 3344 0s, N±H), 3334 0s, NH),
1624 0vs, C=O), and 1564 0vs, amide II) cmÀ1. H-NMR
1
0CDCl3/TMS) d 4.23 0d, J ˆ 7:3 Hz, 1 H, NH), 4.14 0d,
J ˆ 7:4 Hz, 1 H, NH), 3.7 0m, 1 H, CH), 3.5 0m, 1 H, CH),
1.9 0m, 2 H), 1.8 0m, 2 H), 1.5 0m, 14 H), 1.3 0m, 3 H), 1.1
0m, 3 H) ppm. 13C-NMR 0CDCl3): d 157.2 0C=O), 49.9 0C-
10), 48.7 0C-1), 33.9 0C-2,6), 32.8 0C-20,80), 27.3 0C-30,70),
25.6 0C-4), 25.5 0C-50), 24.9 0C-3,5), 23.7 0C40,60) ppm.
2.2. Synthesis
Inhibitor structures are given in the tables, and boldface
numbers throughout the text refer to these compounds.
Compounds 01), and 04) to 043) were synthesized by con-
densation of the appropriate isocyanate and amine, whereas
compounds 02) and 03) were synthesized by condensation of
the appropriate acylchloride and ammonia. Reaction pro-
ducts were puri®ed by recrystallization and, in a few cases,
by ¯ash chromatography. In addition to sharp melting
points and single spots on silica gel TLC, the products
‡
LC±MS m/z 0relative intensity) 505.4 049, [2M ‡ H] ),
‡
253.2 0100, [M ‡ H] ).
2.5. Enzyme preparation
Recombinant mouse sEH and human sEH were pro-
duced in a baculovirus expression system [25,26] and
puri®ed by af®nity chromatography [27]. The preparations
were at least 97% pure as judged by SDS±PAGE and
scanning densitometry. No detectable esterase or glu-
tathione transferase activities were observed [28]. Glu-
tathione transferase activity can interfere with the
radiochemical assay used to generate the data reported
in this paper, while esterases, as well as glutathione
transferases, interfere with the high throughput screening
assay used to obtain the rank order of the compounds [21].
Protein concentration was quanti®ed by the Pierce BCA
assay 0Pierce) using BSA as the calibrating standard.
1
were characterized using H- and/or 13C-NMR 0General
Electric QE-300), infrared, and mass spectroscopy 0elec-
trospray mass spectrometry was run on every compound
and is reported in Table 1; for most compounds, FAB and
MALDI mass spectrometry also were performed). Based on
NMR and mass spectroscopy data, purities over 97% were
obtained for every compound described herein. As exam-
ples, the syntheses of compounds 01) and 010) are described
in the following section.
2.3. N-Cyclohexyl-N0-dodecylurea /1)
To a stirred cold solution of 0.38 g 03.0 mmol) of
dodecylamine in 5 mL of hexanes was added 0.56 g
03.0 mmol) of cyclohexyl isocyanate dissolved in 5 mL
of hexanes. After stirring at room temperature for 1 hr, a
white solid was obtained, which was recrystallized twice
from hexanes. The resulting white crystal 00.90 g; yield:
97%) had a melting point of 94.0±95.08. IR 0KBr) 3348 0m,
NH), 3324 0m, NH), 1623 0s, C=O), and 1577 0s, amide II)
cmÀ1. 1H-NMR 0CDCl3/TMS): d 5.48 0t, J ˆ 7:4 Hz, 1 H,
N0H), 5.32 0d, J ˆ 8:1 Hz, 1 H, NH), 3.5 0m, 1 H, cyclo-
hexyl), 3.09 0q, J ˆ 6:9 Hz, 2 H, CH2, C-10), 1.9 0m, 2 H,
cyclohexyl), 1.7 0m, 2 H, cyclohexyl), 1.6 0m, 1 H,
cyclohexyl), 1.4 0m, 2 H, CH2, C-20), 1.2 0bm, 21 H),
1.1 0m, 2 H, C-110), 0.85 0t, J ˆ 6:5 Hz, 3 H, CH3) ppm.
13C-NMR 0CDCl3): d 158.4 0C=O), 48.6 0C-1), 40.3 0C-10),
34.0 0C-2,6), 31.9 0C-20), 30.5 0C-30), 29.6 0C-40,50,60,70),
29.5 0C-80), 29.3 0C-90), 27.1 0C-100), 25.6 0C-4), 25.0 0C-
3,5), 22.6 0C-110), 14.1 0C-120) ppm. LC±MS m/z 0relative
2.6. IC Assay conditions
50
The rank order of compound potency was established
using a spectrophotometric substrate in a 96-well micro-
plate assay [29]. The precise IC values reported herein
50
were determined as described [21] using racemic [3H]t-
DPPO as substrate [28]. Enzymes 08 nM mouse sEH or
12 nM human sEH) were incubated with inhibitors for
5 min in sodium phosphate buffer 0pH 7.4) at 308 prior to
substrate introduction 0[S]®nal: 50 mM; ꢀ12,000 dpm/
assay). The enzymes were incubated at 308 for 5±
10 min, and the reaction was quenched by the addition
of 60 mL methanol and 200 mL isooctane, which extracts
the remaining epoxide from the aqueous phase. The activ-
ity was followed by measuring the quantity of radioactive
diol formed in the aqueous phase using a liquid scintilla-
tion counter 0Wallac model 1409). Under the conditions
used, rates were linear with both time and enzyme con-
centration and resulted in at least 5% but not more than
30% hydrolysis of the substrate. Assays were performed in
‡
‡
intensity): 621.5 019, [2M ‡ H] ), 311.3 0100, [M ‡ H] ).
2.4. N-Cyclohexyl-N0-cyclooctylurea /10)
triplicate. By de®nition, IC is the concentration of inhi-
50
bitor that reduces enzyme activity by 50%. The IC was
50
To 0.38 g 03.0 mmol) of cyclooctylamine in 20 mL
of hexanes, 0.39 mL 00.35 g, 3.0 mmol) of cyclohexyl
determined by regression of at least ®ve datum points with
a minimum of two points in the linear region of the curve

1602
C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
on either side of the IC . The curve was generated from at
50
While this assay allows rapid, accurate, and precise screen-
ing of numerous compounds in a 96-well plate format, it
does not permit the segregation of very potent compounds,
due to its low sensitivity. For example, as the effective
concentration of a candidate inhibitor approaches the
concentration of the target enzyme used in the assay, the
ability to distinguish among the relative potencies of
powerful inhibitors drops. Therefore, in this study, we
followed the screening assay with a more sensitive radio-
active-based assay to further optimize the inhibitor struc-
ture [28]. A 10±20-fold increase in sensitivity was
observed using the conditions reported here.
The effect of modifying the urea pharmacophore by
removing one nitrogen to give amides was studied ®rst
0Table 2). For both enzymes, 15±30-fold better inhibition
was obtained for the urea 01) than for the two correspond-
ing amides 02, 3). Interestingly, for the murine enzyme a
2-fold higher inhibition was obtained when the nitrogen
was on the cyclohexyl side of the central pharmacophore
02), while the reverse was observed for the human sEH.
Due to the increased effectiveness in inhibiting sEH, the
rest of this study was performed using compounds with a
central urea function.
least three separate runs, each in triplicate, to obtain the
standard deviation given in Section 3.
2.7. Structure±activity relationships
All para-substituent constants of the aromatic com-
pounds 017) to 031) came from the extensive compilation
of Hansch and Leo [30]. For the molar refractivity 0MR),
we used a base of 0.1 for protons. Multiple regression
analysis of the inhibition potency 0Àlog0IC )) of the para-
50
substituted ureas on the mouse and human sEH was
performed using Sigma-Plot v4.0 0Jandel Scienti®c) and
Excel 97 0Microsoft). Combinations of up to three free-
energy parameters were investigated. The F-test, which
uses the ratio of the regression sum of squares to the
residual sum of squares as a ®t criterion 0a ˆ 0:05), was
used to determine the con®dence of adding parameters to
multiple regression equations. Standardized regression
coef®cients were calculated to establish the rank order
of independent variable in¯uence on inhibitor potency.
2.8. Kinetic assay conditions
The effect of the size of the substitution on both sides of
the central urea pharmacophore in relation to inhibitor
potency was then investigated 0Table 3). For the murine
enzyme, when maintaining a cyclohexyl on one side, the
best inhibition was obtained if a methyl-cyclohexyl 09) or a
cyclooctyl 010) was present on the other side. The presence
of smaller rings 04±7) promoted a large drop in potency,
while a larger ring 011) reduced the inhibitor potency only
slightly. The bulky adamantyl group in compound 012)
showed a potency similar to that of the larger ring systems
08, 10, and 11), probably because these rings folded to give
a large lipophilic group. However, an order of magnitude of
potency was lost if the adamantyl group was placed one
carbon away from the urea nitrogen 013), and the carbon
alpha to the urea nitrogen was methylated. This a-methy-
lation may, however, slow P450-dependent N-dealkylation.
Comparing compounds 014) and 016) to 012) and 011),
respectively, the replacement of the cyclohexyl by an
adamantyl had a very different effect on the inhibitor
Dissociation constants were determined following the
method described by Dixon [31] for competitive tight-
binding inhibitors, using [3H]t-DPPO as the substrate [28].
Inhibitors at concentrations between 0 and 10 nM for 01)
and 042) were incubated in triplicate for 5 min in 0.1 M
sodium phosphate buffer 0pH 7.4) at 308 with 100 mL of
the enzyme 02 nM mouse sEH). Then substrate 02:5 <
‰SŠ
< 30 mM) was added. Velocity was measured as
final
described previously [28]. For each substrate concentra-
tion, the plots of the velocity as a function of the inhibitor
concentration allow the determination of an apparent
inhibition constant 0KIapp) [31]. The plot of these KIapp
values as a function of the substrate concentration allows
the determination of KI when ‰SŠ ˆ 0. Results are
means Æ SD of three separate determination of KI.
2.9. Solubility of inhibitor in the water phase
Serial dilutions of inhibitors were prepared in DMF 0[I]
from 0.5 to 50 mM). To 990 mL of sodium phosphate buffer
00.1 M, pH 7.4) 10 mL of the inhibitor solution was added
Table 2
Inhibition of mouse and human sEH by 1,3-disubstituted urea and amidesa
0‰DMFŠ
ˆ 1%; [I]®nal from 5 to 500 mM). Insolubility of
final
the inhibitor was indicated by the increase in turbidity of
the water solution. The turbidity was measured as the
absorbance at 600 nm using a blank containing only the
buffer and 1% of DMF at room temperature.
IC 0nM)
50
X
Y
No.b
Mouse sEH
Human sEH
NH
NH
CH2
NH
CH2
NH
1
2
3
9.8 Æ 0.4
190 Æ 5
85.2 Æ 0.5
3000 Æ 100
1400 Æ 100
270 Æ 10
3. Results
a Enzymes and inhibitor were incubated together for 5 min at 308 before
the addition of the substrate 0t-DPPO, 50 mM). Results are means Æ SD of
three separate experiments.
In previous studies [21,32], a spectrophotometric assay
[29] was used to determine the potency of sEH inhibitors.
b Compound number.

C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
1603
Table 3
Inhibition of mouse and human sEH by 1,3-disubstituted ureas: effect of cycloalkyl ring sizea
No.
IC 0nM)
50
Mouse sEH
Human sEH
R
R0
4
5
5100 Æ 100
640 Æ 40
9400 Æ 100
3600 Æ 200
6
7
109 Æ 2
1230 Æ 30
63 Æ 0.3
81.8 Æ 0.7
8
9
16.7 Æ 0.5
7.3 Æ 0.1
26.8 Æ 0.4
11.9 Æ 0.3
10
11
8.4 Æ 0.1
6.5 Æ 0.2
6.2 Æ 0.1
20 Æ 1
12
13
17 Æ 1
6.5 Æ 0.1
360 Æ 10
205 Æ 8
14
15
9.5 Æ 0.2
37 Æ 1
6.4 Æ 0.1
15 Æ 1
16
68 Æ 2
7.0 Æ 0.4
a Enzymes and inhibitor were incubated together for 5 min at 308 before the addition of the substrate 0t-DPPO, 50 mM). Results are means Æ SD of three
separate experiments.

1604
C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
potency. For the adamantyl derivatives 014 vs. 12) the IC
structural correlation of the selected urea derivatives with
mouse sEH inhibition potency produced the following
equation, which was signi®cant 0P < 0:05). Values in
parentheses are the standard deviations of each indepen-
dent variable coef®cient.
50
was reduced by 50%, while for the cyclododecyl deriva-
tives 016 vs. 11) the IC50 increased by 3-fold. As observed
for the cyclohexyl derivatives 012 and 13), compared to
014) a 4-fold loss of potency was observed if the adamantyl
group was placed one carbon away from the urea nitrogen
015). For the human enzyme, when maintaining a cyclo-
hexyl on one side, the best inhibition was obtained if a
cyclooctyl 010), as observed for the mouse sEH, or a
cyclododecyl 011) was present on the other side. Interest-
ingly, as observed for the mouse sEH, the presence of an
adamantyl maintained the inhibitor potency if placed close
to the urea nitrogen 012), but there was a two order of
magnitude decrease in potency if the adamantyl was placed
one carbon away from the urea nitrogen 013). Furthermore,
the replacement of the cyclohexyl by an adamantyl 011, 12
vs. 14, 16) maintained the high potency of the inhibitor.
After optimizing the size of the substitution around the
urea pharmacophore, we studied the effect of different
functions on one side of the inhibitors. Because of the
diversity of the starting materials available, we ®rst inves-
tigated the effect of 4-substituted phenyl derivatives
0Table 4). The results were analyzed using multiple regres-
sion QSAR relating the inhibitor potency expressed as
Àlogꢁic50† ˆ 0:55ꢁÆ0:06†p ‡ 0:6ꢁÆ0:1†sÀ
‡ 0:15ꢁÆ0:04†D-B5† À 2:54ꢁÆ0:08†;
N ˆ 15; r ˆ 0:97 ꢁr2 ˆ 0:94†; S ˆ 0:15; PꢁF† ˆ 0:92
01)
This equation explained 94% of the correlation among the
inhibitor potencies of N-cyclohexyl-N0-4-substituted-phe-
nylureas against the murine enzyme. However, as a pre-
cautionary note, since the cyclohexyl group in compounds
017±31) could act as either R or R0 in Fig. 1, the given
equation incorporates all binding conformations. It can be
seen that p, sÀ, and D-B5 have a positive in¯uence on the
Àlog0IC ), i.e. increase inhibitor potency. Therefore, large
50
hydrophobic groups with high sÀ values enhanced inhibi-
tion. Furthermore, inhibitor potency was affected more by
the hydrophobicity of the para-substituent than by its
electronic effect or width. The negative Hammet constant
0sÀ), which is well de®ned for anilines, represents the
capacity for substituents to accept a pair of electrons
through resonance [30]. In Eq. 01), the positive sign of
the parameter for sÀ indicates that inhibition is enhanced
by displacement of a pair of electrons from the phenyl ring
toward the para-substituents. It then displaces the loosely
held lone pair of electrons on the nitrogen through reso-
nance, therefore, promoting a positive charge on this atom.
For the human sEH, a signi®cant 0P < 0:05) structural
correlation with inhibition potency for selected urea deri-
vatives was also produced 0Eq. 02)).
Àlog0IC ) and molecular free energy changes to 4-sub-
50
stituent effects on electronic 0s, sÀ, s‡), hydrophobic
0p), and steric [MR, length 0D-L), width 0D-B5)] para-
meters. Structures were chosen such that free-energy para-
meters were noncolinear 0r2 < 0:4). The multiple
regression equations for each enzyme were developed in
a step-wise fashion as described in Section 2.5. The
Table 4
Inhibition of mouse and human sEH by 1-cyclohexyl-3-04-substituted-
phenyl) ureasa
À logꢁic50† ˆ 1:1ꢁÆ0:1†p ‡ 0:9ꢁÆ0:2†s À À2:9ꢁÆ0:1†;
N ˆ 15; r ˆ 0:93 ꢁr2 ˆ 0:86†; S ˆ 0:35; PꢁF† ˆ 0:90
02)
IC 0nM)
50
This equation explained 86% of the correlation in the
action of N-cyclohexyl-N0-4-substituted-phenylureas
against the human enzyme. A representation of the pre-
dicted inhibitor potency from Eq. 02) is shown in Fig. 2. It
can be seen that both p and sÀ have a positive in¯uence on
R
No.
Mouse sEH
Human sEH
±N0CH3)2
±NH2
±OH
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
89 Æ 2
1520 Æ 10
800 Æ 30
163 Æ 7
20.9 Æ 0.9
255 Æ 5
28.4 Æ 0.4
185 Æ 1
284 Æ 2
95 Æ 2
1210 Æ 20
13000 Æ 1000
11000 Æ 1000
390 Æ 1
±OCH3
±C0CH3)3
±CH3
35 Æ 1
the inhibitor potency, i.e. increase Àlog0IC ). Therefore,
50
262 Æ 4
hydrophobic groups with high sÀ values enhanced inhibi-
tion. Furthermore, as observed for the murine enzyme,
inhibitor potency was more affected by the hydrophobicity
of the para-substituent than by its electronic effect. As
observed for mouse sEH, inhibition was enhanced by the
formation of a positive charge on the nitrogen of the urea
through resonance. Contrary to the murine enzyme, the
addition of the width factor 0D-B5) did not improve the
relationship signi®cantly, underlining that the mouse is an
excellent but not perfect model for studying the activity of
inhibitors with the human enzyme. The benzophenone
±CH0CH3)2
±F
14.2 Æ 0.8
285 Æ 1
±H
1440 Æ 20
33 Æ 1
±Cl
±I
28.5 Æ 0.5
81 Æ 2
19.5 Æ 0.2
41 Æ 1
±Br
±COC6H5
±OCF3
±NO2
9.9 Æ 0.2
16.0 Æ 0.2
49 Æ 3
68 Æ 2
7.3 Æ 0.2
120 Æ 2
a Enzymes and inhibitor were incubated together for 5 min at 308
before the addition of the substrate 0t-DPPO, 50 mM). Results are
means Æ SD of three separate experiments.

C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
1605
than aromatic groups, we studied the effects of different
chemical functions at the end of aliphatic chains 0Table 5).
For both mouse and human sEH, increasing the chain
length from 6 to 12 carbons, which has little effect on
the overall hydrophobicity of the compounds, had a limited
effect on the inhibitor potency 0decreased the IC 5- and
50
3-fold, respectively) if a non-polar methyl group was
present at the end of the chain 01 and 32); a three order
of magnitude decrease was observed in the case of the very
polar carboxylic acid function 034 and 42). Interestingly,
for both enzymes the presence of any polar groups at the
end of a short chain 0<6 carbons, 33±35) resulted in a loss
of inhibition potency, while for a longer chain 0>11 car-
bons) the presence of polar groups 039±41) had little
in¯uence on the potency of the inhibitors. For all chain
lengths tested, the presence of a terminal carboxylic acid
function 034, 37, and 39) induced the largest decrease in
potency. The conversion of this acid function to an ester 038
and 40) or an amide 041) resulted in more potent inhibitors,
suggesting that the presence of a negative charge induced a
loss of inhibitory activity especially with short chain
compounds. As observed previously with cycloalkyl deri-
vatives 0see Table 3), the replacement of the cyclohexyl
042) by an adamantyl group 043) resulted in a slight
decrease of inhibitor potency toward the murine enzyme,
Fig. 2. HsEH QSAR results: inhibitor potency [Àlog0IC50)] as a function of
p and sÀ. The black circles represent the experimental values, while the
mesh represents the calculated inhibitor potencies from Eq. 02).
derivative 029) was quite active with the murine enzyme
and could be used as a photoaf®nity label, while the
tri¯uoromethoxy group 030) led to high activity with the
human enzyme.
Because inhibition potency was enhanced by hydropho-
bicity, and because aliphatic chains are more hydrophobic
while a 2-fold lower IC 0increase in potency) was
50
observed for the human enzyme.
To determine if the inhibitor mechanism of action is
retained upon structural modi®cation, the dissociation
constants of 01) and 042) were determined for mouse
Table 5
Inhibition of mouse and human sEH by 1-cyclohexyl-3-n-alkyl ureas: effect of alkyl chain length and functional group at the o positiona
No.
IC 0nM)
50
R
n
Z
Mouse sEH
Human sEH
5
CH3
32
33
34
35
55 Æ 3
181 Æ 2
221 Æ 2
CH2OH
COOH
CN
3050 Æ 60
17700 Æ 300
91000 Æ 1000
50000 Æ 2000
1320 Æ 30
6
7
COOH
36
1360 Æ 40
14700 Æ 200
COOH
37
38
134 Æ 2
1100 Æ 100
COOCH3
26.1 Æ 0.5
361 Æ 9
10
11
COOH
39
40
41
10.1 Æ 0.1
7.0 Æ 0.3
8.4 Æ 0.1
230 Æ 10
61 Æ 1
COOCH3
CONH2
167 Æ 3
CH3
1
42
9.8 Æ 0.4
85.2 Æ 0.5
COOH
11.1 Æ 0.1
112 Æ 1
11
COOH
43
18 Æ 1
69 Æ 2
a Enzymes and inhibitor were incubated together for 5 min at 308 before the addition of the substrate 0t-DPPO, 50 mM). Results are means Æ SD of three
separate experiments.

1606
C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
phase 0Table 6). As one would expect, the addition of a
carboxylic acid functionality to 01) to yield 042) doubled its
water solubility. Surprisingly, replacement of the cyclo-
hexyl group 042) by an adamantyl group 043) further
increased the solubility of the resulting compound.
4. Discussion
Over the past 30 years, sEH has been studied mainly for
its role in hepatic xenobiotic detoxi®cation [3]. We
recently showed that sEH present in kidneys and coronary
endothelial cells plays an important role in EET metabo-
lism [11,33]. EETs possess important vasodilating and
anti-in¯ammatory properties [6,7]. The bioactivity of these
epoxy-fatty acids suggests that inhibition of vicinal-dihy-
droxy-lipid biosynthesis may have therapeutic value, mak-
ing sEH a promising pharmacological target, and inhibition
of sEH has been proposed as a novel approach for the
treatment of hypertension [11,12]. Recently, we demon-
strated that ureas and carbamates inhibit sEH [21]. To
better de®ne the structural constraints on inhibitor potency,
we investigated the effect of altering the disubstituted-urea
inhibitor structure with the goal of increasing inhibitor
potency and improving its physical properties. Using a
more sensitive assay, it was possible to segregate very
potent compounds and, therefore, to further re®ne the
structural requirements for inhibition. Even if the overall
structures of the inhibitors for the murine and human sEH
are similar 0Fig. 1), differences between the two enzymes
were observed over the four series of compounds studied
0Tables 2±5). Again, these data caution that the mouse is an
excellent but not perfect model for studying the activity of
inhibitors with the human enzyme.
Fig. 3. Determination of the KI of compound 01) with the MsEH 02 nM)
using [3H]t-DPPO as substrate [28]. For each substrate concentration 02.5±
30.0 mM), the velocity is plotted as a function of the inhibitor concentration
00±10 nM) allowing the determination of an apparent inhibition constant
0KIapp) [31]. KIapp values are plotted as a function of the substrate
concentration 0inset). For ‰SŠ ˆ 0, a KI value of 3.8 nM was found. Similar
plots were obtained with the murine enzyme and 042) 0KI ˆ 3:8 nM).
sEH 0Fig. 3). For both chemicals, competitive tight-bind-
ing inhibition kinetics were obtained with r2 > 0:99.
Kinetics were obtained previously for mouse sEH and
compounds 07) and 027) [21,23], suggesting that the com-
petitive nature of the inhibition is retained over a wide
variety of structural modi®cations. KI values of 3:7 Æ 0:1
and 3:8 Æ 0:2 nM were obtained for 01) and 042), respec-
tively. The similar KI values obtained con®rm that the
presence of a carboxylic function ``far away'' from the
central function does not in¯uence the inhibitor potency of
the compound.
The structural features yielding potent inhibition for the
two enzymes studied are summarized as follows. As
observed before [21], sEHs are best inhibited by ureas
0X ˆ Y ˆ NH); however, the activity of the amides and
carbamates 0Table 2 and [21]) may present an opportunity
for inhibition. If substituents can be found leading to
To determine how structural changes affect availability,
the solubility of three inhibitors was measured in a water
Table 6
Solubility of 1,3-disubstituted ureas in water
No.
Solubilitya 0mM)
R
R0
±nC12H25
1
42
43
10 < S < 25
25 < S < 50
75 < S < 100
±nC11H22COOH
±nC11H22COOH
a Solubilities were measured in sodium phosphate buffer 0pH 7.4, 0.1 M) containing 1% of DMF. Results are the means of three separate experiments.

C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
1607
amides and carbamates as potent as the best ureas, the
resulting compounds may be much easier to formulate for
oral availability and may have superior absorption±distri-
bution±metabolism±excretion 0ADME) properties. Repla-
cement of one of the two nitrogens had the same effect on
inhibition potency 0reduced by one order of magnitude) for
both enzymes, but the orientation of this substitution to the
small 0R) or large 0R0) alkyl moieties affected potency
differently for the two mammalian orthologs tested. Spe-
ci®cally, substitution at the X or Y atom retained maximal
potency for the human and murine enzymes, respectively.
Inhibition of sEH is enhanced by the presence of two
hydrophobic groups 0R and R0) on both sides of the central
pharmacophore in which case one side 0R0) could be larger
than the other 0R) [21,32]. For both enzymes studied, the
best inhibition is obtained with a cyclohexyl or an ada-
mantyl appendage on the small side 0R). For the larger side
0R0), the enzymes have a different preference. For the
murine sEH, equally potent inhibition is obtained with
either cycloalkyl, alkyl, or aryl groups, while for the human
enzyme better inhibition is obtained with cycloalkyl or aryl
groups than with straight alkyl groups. In addition, for the
most potent inhibition, smaller R0 groups are permitted for
the murine sEH than for the human sEH. Interestingly, if
the R0 group is large enough, the presence of polar func-
tions 0Z) does not affect inhibitor potency. These data
suggest that there is a hydrogen-binding site where the
carboxylic acid ®ts in both enzymes. Therefore, improve-
ment of sEH inhibitor properties could be obtained with-
out potency loss by modi®cations at the Z position, and
such polar groups should improve the speci®city of the
inhibitor for the sEH. For example, as shown in Table 6, a
carboxylic acid function at the Z position increases the
water solubility of the resulting compounds. Such proper-
ties could be useful for the design of orally active sEH
inhibitors and the design of prodrugs with altered physical
properties.
enzymes studied, multi-regression analyses indicate that
groups that stabilize a positive charge on the nitrogen
through resonance improve inhibition. This ®nding sup-
ports the hypothesis of the formation of hydrogen bonds
between the amide function of the inhibitor and one or
more active site residues. In addition, hydrophobic groups
on both sides of the central urea function improve inhibi-
tion, agreeing with the establishment of numerous van der
Waals interactions with aliphatic and aromatic active-site
residues [22]. The hydrophobic nature of the sEH active
site is consistent with the large loss of inhibitor potency
observed when a polar group is added in close proximity to
the urea function. The X-ray structures indicate a long,
narrow, hydrophobic, `L'-shaped substrate pocket with an
opening at each end to the medium and the catalytic
residues at the bend of the `L' [22,23,35]. Current struc-
tures cannot distinguish between a tunnel where substrates
and inhibitors must thread their way in and a deep cleft that
is capable of opening. The X-ray structures also do not give
an idea as to how expandable the tunnel is with larger
inhibitors and substrates. Although we know that one side
of the `L' is larger than the other, we have not de®ned
groups on inhibitors that clearly orient them in only one
way in the active site.
In conclusion, we have further re®ned the structural
requirements for murine and human sEH inhibition by
1,3-disubstituted ureas. With long aliphatic chains for both
the murine and human enzymes, we showed that terminal
polar moieties could be present. Such groups increase
water solubility and decrease log P, while offering sites
for the attachment of groups to yield prodrugs. This work
will facilitate the design of sEH inhibitors with better
physical properties.
Acknowledgments
The authors thank Dr. A.D. Jones 0Pennsylvania State
University) for mass spectra analyses, and Dr. Yoshiaki
Nakagawa 0Kyoto University) for his helpful comments
and assistance in the QSAR multiple regression analysis.
This work was supported, in part, by NIEHS Grant R01-
ES02710, NIEHS Center for Environmental Health
Sciences 1P30-ES05707, and the NIH/NIEHS Superfund
Basic Research program P42-ES04699.
We hypothesized many years ago that epoxylipids might
be endogenous substrates for sEH [34]. The structure±
activity relationships shown in Table 5 support this hypoth-
esis, knowing that epoxy-fatty acids with a close distance
between the oxirane and the carboxylate functions, e.g.
5,6-epoxyeicosatrienoic acid, are poor sEH substrates,
while epoxy-fatty acids with a greater distance such as
11,12- or 14,15-epoxyeicosatrienoic acid, are very good
sEH substrates [4,34]. Possibly more potent and speci®c
inhibitors can be synthesized based on mimicking oxyli-
pins as hypothetical endogenous substrates.
References
In recent studies, we examined the mechanism of action
of urea inhibitors by analyzing the X-ray structure of the
mouse sEH with three urea inhibitors [22,23]. These
inhibitors acted by mimicking the activation of a substrate
epoxide ring for nucleophilic attack by the catalytic aspar-
tate residue. The carboxylate side chain of this nucleophilic
aspartate accepts a hydrogen bond from one of the urea NH
groups in each enzyme±inhibitor complex [23]. For both
[1] Oesch F. Mammalian epoxide hydrolases: inducible enzymes
catalyzing the inactivation of carcinogenic and cytotoxic metabolites
derived from aromatic and olefinic compounds. Xenobiotica 1973;3:
305±40.
[2] Hammock BD, Grant D, Storms D. Epoxide hydrolase. In: Sipes I,
McQueen C, Gandolfi A, editors. Comprehensive toxicology. Oxford:
Pergamon Press, 1997. p. 283±305.
[3] Fretland AJ, Omiecinski CJ. Epoxide hydrolases: biochemistry and
molecular biology. Chem Biol Interact 2000;129:41±59.

1608
C. Morisseau et al. / Biochemical Pharmacology 63 /2002) 1599±1608
[4] Zeldin DC, Kobayashi J, Falck JR, Winder BS, Hammock BD,
[20] Ishizaki T, Takahashi H, Ozawa T, Chang SW, Voelkel NF.
Leukotoxin, 9,10-epoxy-12-octadecenoate causes pulmonary vasodi-
latation in rats. Am J Physiol 1995;268:L123±8.
Snapper JR, Capdevila JH. Regio- and enantiofacial selectivity of
epoxyeicosatrienoic acid hydration by cytosolic epoxide hydrolase. J
Biol Chem 1993;268:6402±7.
[21] Morisseau C, Goodrow MH, Dowdy D, Zheng J, Greene JF, Sanborn
JR, Hammock BD. Potent urea and carbamate inhibitors of soluble
epoxide hydrolases. Proc Natl Acad Sci USA 1999;96:8849±54.
[22] Argiriadi MA, Morisseau C, Hammock BD, Christianson DW.
Detoxification of environmental mutagens and carcinogens: struc-
ture, mechanism, and evolution of liver epoxide hydrolase. Proc Natl
Acad Sci USA 1999;96:10637±42.
[5] Moghaddam MF, Grant DF, Cheek JM, Greene JF, Williamson KC,
Hammock BD. Bioactivation of leukotoxins to their toxic diols by
epoxide hydrolase. Nat Med 1997;3:562±6.
[6] Carroll MA, McGiff JC. A new class of lipid mediators: cytochrome
P450 arachidonate metabolites. Thorax 2000;55:S13±6.
[7] Capdevila JH, Falck JR, Harris RC. Cytochrome P450 and
arachidonic acid bioactivation: molecular and functional properties
on the arachidonate monooxygenase. J Lipid Res 2000;41:163±81.
[8] Oltman CL, Weintraub NL, VanRollins M, Dellsperger KC.
Epoxyeicosatrienoic acids and dihydroxyeicosatrienoic acids are
potent vasodilators in the canine coronary microcirculation. Circ Res
1998;83:932±9.
[23] Argiriadi MA, Morisseau C, Goodrow MH, Dowdy DL, Hammock
BD, Christianson DW. Binding of alkylurea inhibitors to epoxide
hydrolase implicates active site tyrosines in substrate activation. J
Biol Chem 2000;275:15265±70.
[24] Newman JW, Denton DL, Morisseau C, Koger CS, Wheelock CE,
Hinton DE, Hammock BD. Evaluation of fish models of soluble
epoxide hydrolase inhibition. Environ Health Perspect 2001;109:61±6.
[25] Grant DE, Storms DH, Hammock BD. Molecular cloning and
expression of murine liver soluble epoxide hydrolase. J Biol Chem
1993;268:17628±33.
[9] Fisslthaler B, Popp R, Kiss L, Potente M, Harder DR, Fleming I,
Busse R. Cytochrome P450 2C is an EDHF synthase in coronary
arteries. Nature 1999;401:493±7.
[10] Weintraub NL, Fang X, Kaduce TL, VanRollins M, Chatterjee P,
Spector AA. Epoxide hydrolases regulate epoxyeicosatrienoic acid
incorporation into coronary endothelial phospholipids. Am J Physiol
1999;277:H2098±108.
[26] Beetham JK, Tian T, Hammock BD. cDNA cloning and expression of
a soluble epoxide hydrolase from human liver. Arch Biochem
Biophys 1993;305:197±201.
[11] Yu Z, Xu F, Huse LM, Morisseau C, Draper AJ, Newman JW, Parker
C, Graham L, Engler MM, Hammock BD, Zeldin DC, Kroetz DL.
Soluble epoxide hydrolase regulates hydrolysis of vasoactive
epoxyeicosatrienoic acids. Circ Res 2000;87:992±8.
[27] Wixtrom RN, Silva MH, Hammock BD. Affinity purification of
cytosolic epoxide hydrolase using derivatized epoxy-activated
Sepharose gels. Anal Biochem 1988;169:71±80.
[28] Borhan B, Mebrahtu T, Nazarian S, Kurth MJ, Hammock BD.
Improved radiolabeled substrates for soluble epoxide hydrolase. Anal
Biochem 1995;231:188±200.
[12] Sinal CJ, Miyata M, Tohkin M, Nagata K, Bend JR, Gonzalez FJ.
Targeted disruption of soluble epoxide hydrolase reveals a role in
blood pressure regulation. J Biol Chem 2000;275:40504±10.
[13] Node K, Huo Y, Ruan X, Yang B, Spiecker M, Ley K, Zeldin DC,
Liao JK. Anti-inflammatory properties of cytochrome P450 epox-
ygenase-derived eicosanoids. Science 1999;285:1276±9.
[29] Dietze EC, Kuwano E, Hammock BD. Spectrophotometric substrates
for cytosolic epoxide hydrolase. Anal Biochem 1994;216:176±87.
[30] Hansch C, Leo A. Substituent constants for correlation analysis in
chemistry and biology. New York: Wiley, 1979.
[14] Campbell WB. New role for epoxyeicosatrienoic acids as anti-
inflammatory mediators. Trends Pharmacol Sci 2000;21:125±7.
[15] Ishizaki T, Shigemori K, Nakai T, Miyabo S, Ozawa T, Chang SW,
Voelkel NF. Leukotoxin, 9,10-epoxy-12-octadecenoate causes ede-
matous lung injury via activation of vascular nitric oxide synthase.
Am J Physiol 1995;269:L65±70.
[31] Dixon M. The graphical determination of Km and Ki. Biochem J
1972;129:197±202.
[32] Nakagawa Y, Wheelock CE, Morisseau C, Goodrow MH, Hammock
BG, Hammock BD. 3-D QSAR analysis of inhibition of murine
soluble epoxide hydrolase 0MsEH) by benzoylureas, arylureas, and
their analogues. Bioorg Med Chem 2000;8:2663±73.
[16] Ishizaki T, Shigemori K, Nakai T, Miyabo S, Hayakawa M, Ozawa T,
Voelkel NF, Chang SW. Endothelin-1 potentiates leukotoxin-induced
edematous lung injury. J Appl Physiol 1995;79:1106±11.
[17] Dudda A, Spiteller G, Kobelt F. Lipid oxidation products in ischemic
porcine heart tissue. Chem Phys Lipids 1996;82:39±51.
[33] Fang X, Kaduce TL, Weintraub NL, Harmon S, Teesch LM,
Morisseau C, Thompson DA, Hammock BD, Spector AA. Pathways
of epoxyeicosatrienoic acid metabolism in endothelial cells:
implication for the vascular effects of soluble epoxide hydrolase
inhibition. J Biol Chem 2001;276:14867±74.
[18] Sakai T, Ishizaki T, Ohnishi T, Sasaki F, Ameshima S, Nakai T,
Miyabo S, Matsukawa S, Hayakawa M, Ozawa T. Leukotoxin, 9,10-
epoxy-12-octadecenoate inhibits mitochondrial respiration of isolated
perfused rat lung. Am J Physiol 1995;269:L326±31.
[34] Chacos N, Capdevila J, Falck JR, Manna S, Martin-Wixtrom C, Gill
SS, Hammock BD, Estabrook RW. The reaction of arachidonic acid
epoxides 0epoxyeicosatrienoic acids) with
a cytosolic epoxide
hydrolase. Arch Biochem Biophys 1983;223:639±48.
[35] Yamada T, Morisseau C, Maxwell JE, Argiriadi MA, Christianson
DW, Hammock BD. Biochemical evidence for the involvement of
tyrosine in epoxide activation during the catalytic cycle of epoxide
hydrolase. J Biol Chem 2000;275:23082±8.
[19] Fukushima A, Hayakawa M, Sugiyama S, Ajioka M, Ito T, Satake T,
Ozawa T. Cardiovascular effects of leukotoxin 09,10-epoxy-12-
octadecenoate) and free fatty acids in dogs. Cardiovasc Res 1988;
22:213±8.