(5Z,11Z,15R)-15-Hydroxyeicosa-5,11-dien-13-ynoic acid: A stable isomer of 15(S)-HETE that retains key vasoconstrictive and antiproliferative activity

Article abstract of DOI:10.1016/j.prostaglandins.2016.04.001

15(S)-Hydroxyeicosa-(5Z,8Z,11Z,13E)-tetraenoic acid (15(S)-HETE) is a metabolite of arachidonic acid that elicits a number of biological effects including vasoconstriction and angiogenesis. (5Z,11Z,15R)-15-Hydroxyeicosa-5,11-dien-13-ynoic acid (HETE analog 1) is a synthetic isomer of 15(S)-HETE that is much more stable to autoxidation. Using isometric recording of isolated pulmonary arteries from male and female rabbits, HETE analog 1 and 15(S)-HETE were found to elicit concentration-dependent contractions that were slightly greater in females compared to males. The maximal response in females was greater with 15(S)-HETE. HETE analog 1 and 15(S)-HETE increased [³H]-thymidine incorporation in vascular smooth muscle cells cultured from male rabbit pulmonary arteries; both the maximal response and potency were greater with 15(S)-HETE. In contrast, HETE analog 1 produced a concentration-dependent inhibition in proliferation and migration of human hormone-independent prostate carcinoma PC-3 cells. The protocol for synthesis of HETE analog 1 is reported. The stability of this substance and its similar biological profile to 15(S)-HETE support future studies in eicosanoid research.

Full text of DOI:10.1016/j.prostaglandins.2016.04.001

Prostaglandins & other Lipid Mediators 123 (2016) 33–39
Contents lists available at ScienceDirect
Prostaglandins and Other Lipid Mediators
Original research article
(5Z,11Z,15R)-15-Hydroxyeicosa-5,11-dien-13-ynoic acid: A stable
isomer of 15(S)-HETE that retains key vasoconstrictive and
antiproliferative activity
Sandra L. Pfistera,∗, Peter G. Klimkob, Raymond E. Conrowb
a Department of Pharmacology and Toxicology, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI 53226, United States
b Alcon Laboratories, a Novartis Company, Fort Worth, TX, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
15(S)-Hydroxyeicosa-(5Z,8Z,11Z,13E)-tetraenoic acid (15(S)-HETE) is a metabolite of arachidonic acid
that elicits a number of biological effects including vasoconstriction and angiogenesis. (5Z,11Z,15R)-15-
Hydroxyeicosa-5,11-dien-13-ynoic acid (HETE analog 1) is a synthetic isomer of 15(S)-HETE that is much
more stable to autoxidation. Using isometric recording of isolated pulmonary arteries from male and
female rabbits, HETE analog 1 and 15(S)-HETE were found to elicit concentration-dependent contractions
that were slightly greater in females compared to males. The maximal response in females was greater
with 15(S)-HETE. HETE analog 1 and 15(S)-HETE increased [3H]-thymidine incorporation in vascular
smooth muscle cells cultured from male rabbit pulmonary arteries; both the maximal response and
potency were greater with 15(S)-HETE. In contrast, HETE analog 1 produced a concentration-dependent
inhibition in proliferation and migration of human hormone-independent prostate carcinoma PC-3 cells.
The protocol for synthesis of HETE analog 1 is reported. The stability of this substance and its similar
biological profile to 15(S)-HETE support future studies in eicosanoid research.
Received 28 December 2015
Received in revised form 14 March 2016
Accepted 18 April 2016
Available online 23 April 2016
Keywords:
Arachidonic acid
15-Lipoxygenase
15-HETE
Vasoconstriction
Proliferation
© 2016 Elsevier Inc. All rights reserved.
1. Introduction
compared to males, identifying for the first time sex-specific dif-
ferences in 15-HETE-induced contractions [3]. 15-HETE is also
Arachidonic acid (5Z,8Z,11Z,14Z-eicosatetraenoic acid) is a 20-
carbon polyunsaturated fatty acid stored in the phospholipids of
the cell membrane. A variety of physiological and pathological
stimuli release arachidonic acid from membrane phospholipids.
Free arachidonic acid is metabolized by at least three different
enzyme systems including cyclooxygenases, lipoxygenases and
cytochrome P450 epoxygenases to a variety of metabolites with a
range of biological actions [1]. Of particular interest are the vascular
effects of 15-lipoxygenase-derived arachidonic acid metabolites.
An oxygenation reaction catalyzed by a specific 15-lipoxygenase
converts arachidonic acid to 15-hydroperoxyeicosatetraenoic acid
(15-HPETE). 15-HPETE is a short-acting compound and is further
metabolized to 15-hydroxyeicosatetraenoic acid (HETE). 15-HPETE
and 15-HETE have distinct biological properties depending on the
target cell. For example, 15-HETE is a potent vasoconstrictor in vari-
ous vascular tissues [2]. In previous studies it was found that female
pulmonary arteries have enhanced 15-HETE-induced contractions
reported to be a pro-atherogenic metabolite because of its ability to
stimulate vascular cell proliferation [4]. However, these mitogenic
effects appear to be cell-specific because other studies in cancer
cells reported that 15-HETE reduced cell proliferation [5–9].
The natural product 15(S)-HETE has been shown to stimulate
ocular mucin secretion in vitro [10,11]. Research at Alcon demon-
strated that 15(S)-HETE protects the cornea in a rabbit model
of desiccation-induced injury [12]. These observations led to the
investigation at Alcon of 15(S)-HETE as a therapeutic agent for
treating dry eye disorders [13]. Early studies along these lines
were hampered by the susceptibility of 15(S)-HETE to autoxi-
dation; however, a surprising increase in stability was realized
by converting this carboxylic acid to the sodium or ammonium
salt [14,15]. In a quest to further improve stability, a program
of chemical synthesis was then initiated. Analogs were prepared
in which the labile skipped polyene array in 15(S)-HETE was
replaced with less reactive functional groups [16,17]. (5Z,11Z,15R)-
15-Hydroxyeicosa-5,11- dien-13-ynoic acid (1), its S enantiomer
and the racemic mixture were among the synthetic compounds
selected for study in various biological assays. In the present report,
we disclose the effects of 15(S)-HETE and HETE analog 1 on vascular
∗
Corresponding author.
E-mail address: spfister@mcw.edu (S.L. Pfister).
http://dx.doi.org/10.1016/j.prostaglandins.2016.04.001
1098-8823/© 2016 Elsevier Inc. All rights reserved.

34
S.L. Pfister et al. / Prostaglandins & other Lipid Mediators 123 (2016) 33–39
contractions in female and male rabbit pulmonary artery, and on
proliferation of both vascular smooth muscle and prostate cancer
cells.
Human hormone-independent prostate carcinoma PC-3 cells were
obtained from the American Type Culture Collection (Manassas,
VA). PC-3 cells were maintained in RPMI 1640 supplemented
with 10% fetal bovine serum, l-glutamine (2 mM), streptomycin
(100 ␮g/mL), and penicillin (100 U/mL) at 37 ◦C in 5% CO2. To
measure [3H]-thymidine incorporation, pulmonary artery vas-
cular smooth muscle cells or PC-3 cells were grown to 60%
confluency in 24-well plates. The cells were exposed to Dul-
becco’s modified Eagle’s medium (DMEM) containing antibiotics
(1% antibiotic/antimycotic solution) and 0.5% rabbit serum or fetal
bovine serum to arrest cell growth. After 48 h, new DMEM was
added containing 0.5% appropriate serum, and the cells were stimu-
lated with increasing concentrations of 15(S)-HETE or HETE analog
1 for 24 h. [3H]-Thymidine (1 ␮Ci/well) was added during the final
4 h. The medium was removed, and cells were washed with ice cold
PBS, followed by 0.3 mL of ice cold perchloric acid (0.3 M) and PBS.
In order to solubilize the cells, 0.5 mL of 0.1% SDS/0.1 N NaOH was
added to each well. Radioactivity in the solubilized monolayer was
determined by liquid scintillation spectrometry.
2. Materials and methods
2.1. Animals
Animal protocol was approved by the Institutional Animal Care
and Use Committee of the Medical College of Wisconsin, and pro-
cedures were performed in accordance with the National Institutes
of Health Guide for the Care and Use of Laboratory Animals (1996).
Six to eight week old female and male New Zealand White (NZW)
rabbits were obtained from Kuiper Rabbit Ranch (Gary, IN). The ani-
mals were housed in the Medical College of Wisconsin Biomedical
Research Facilities and maintained on a standard rabbit chow diet
and given tap water ad libitum.
2.2. Vascular reactivity
2.4. Cell migration
Pulmonary artery tissue was isolated as previously described
[18]. Briefly, female and male NZW rabbits were euthanized
(sodium pentobarbital, 120 mg/kg, iv), the heart and lungs removed
as a unit, and placed immediately in a Kreb’s bicarbonate buffer
of the following composition (mmol/L): NaCl 118, KCl 4, CaCl2
3.3, NaHCO3 24, KH2PO4 1.4, MgSO4 1.2 and glucose 11, pH 7.4.
The main pulmonary artery was identified at its origin from the
right ventricle, and both left and right pulmonary arteries were
dissected to their most distal end. The pulmonary artery distal
to the first branching of the left or right pulmonary artery was
used, and this is referred to as the intrapulmonary artery. After
dissection, the tissue was cleaned of adherent lung parenchyma
and connective tissue using care not to disturb the endothelial
layer. Rings of intrapulmonary artery were obtained (2–3 mm)
and mounted in a four-chamber myograph (model 610 M, Danish
Myo Technology) containing Kreb’s bicarbonate buffer (warmed
to 37◦ C) and continuously aerated with a 95% O2-5% CO2 mix-
ture. The resting tension was 1.0 g and was not different between
females and males. Vessels equilibrated at resting tension for 1 h.
Contractions were produced by raising the KCl concentration to
40 mmol/L. After vessels had reproducible, stable contractions to
KCl, then cumulative concentration-response curves to 15(S)-HETE
(10−9−5×10−6 mol/L) or HETE analog 1 (10−9–10−5 mol/L) were
determined. Experimental preparations were randomly assigned to
receive either 15(S)-HETE or HETE analog 1 in order to circumvent
any desensitization effects due to multiple concentration-response
curves in same vessel. To standardize for any minor differences in
tissue size, contractile responses were expressed as a percentage of
the maximal response to 40 mmol/L KCl. Maximal KCl contractions
were not different between females and males (1.06 0.07 g vs
0.92 0.09 g; female vs male).
For cell migration measurement by X-celligence system, PC3
cells (30,000 cells/well) were plated in the upper chamber (8.0 ␮m
filter pores without Matrigel) of the Migration CIM-Plate of the Real
Time Cell Analyzer DP instrument (Roche Applied Science) in the
serum-free media [20]. Media containing 10% fetal bovine serum
was placed in the bottom chamber. Varying concentrations of HETE
analog 1 were added in the upper chamber and the system was
equilibrated for 30 min in the incubator at 37 ◦C. Then cell migra-
tion was monitored and recorded every 15 min for up to 72 h. Four
wells were used and averaged for each treatment concentration.
2.5. Stability assay
To estimate relative storage stability, 15(S)-HETE and HETE ana-
log compounds were subjected to a heat-accelerated autoxidation
assay. Each HETE analog was dissolved in 100% ethanol (about
3300 ␮g/mL) as a stock standard (SS). Thirty microliters (30 ␮L) of
SS was added to 3.0 mL of diluent solution (DS), 50% EtOH/H2O, v/v,
to be used as reference standard. Another thirty microliters (30 ␮L)
of SS was dried with nitrogen gas at room temperature and then
stressed in an oven in air at 75◦ C for either 15 min, three h, one
week, two weeks or three weeks. When a sample was removed from
the oven, it was cooled in an ice/water bath, then the residue was
redissolved with 3.0 mL DS. Each heat-stressed sample was then
analyzed against its unstressed reference standard using HPLC.
2.6. HETE analog synthesis
A synthetic route suitable for producing HETE analog 1 on
multigram scale was developed. To summarize, 6-heptynoic acid
was converted [21] to iodo acid 2 and thence to amide 3 (70%
yield). Hydroboration [22] of 3 afforded cis iodo alkene 4 (96%).
Coupling [23] of 4 with (R)-1- octyn-3-ol [24] yielded enynol 5
(80%), which was silylated to give 6 (91%). Hydride reduction
of 6 gave aldehyde 7 (88%), which upon Wittig reaction with
4-carboxybutyltriphenylphosphane followed by desilylation pro-
vided 1 (89%). The synthetic scheme is shown in Fig. 1.
2.3. Measurement of [3H]-thymidine incorporation
Vascular smooth muscle cells were isolated and cultured from
male rabbit pulmonary arteries using previously described meth-
ods [19]. After collection of endothelial cells, strips of denuded
vessels were placed into gelatin coated flasks with M199 medium
containing 10% fetal calf serum with l-glutamine (1%) and antibi-
otics (1% antibiotic/antimycotic solution; 10,000 units penicillin,
10 mg streptomycin, 25 ng amphotericin B (Sigma)). Smooth mus-
cle cells migrated to the flasks within three to five days. Once
growth was established on the coated flasks, then the vessels
were removed and the cells resuspended in M199 medium con-
taining 20% fetal calf serum. Purity of smooth muscle cells was
confirmed by positive staining for smooth muscle cell ␣-actin.
2.6.1. 7-Iodo-6-heptynoic acid (2)
Potassium hydroxide (85%, 32 g, 0.49 mol) was dissolved in a
stirred solution of 6-heptynoic acid (18 g, 0.14 mol) in 145 mL of
MeOH and 72 mL of H2O under N2. The solution was cooled in ice.
Iodine (44 g, 0.17 mol) was added and the flask was wrapped with
Al foil to exclude light. The temperature was allowed to rise to 23 ◦C

S.L. Pfister et al. / Prostaglandins & other Lipid Mediators 123 (2016) 33–39
35
Fig. 1. Chemical synthesis of HETE analog 1.
over several hours. After 3 days the colorless mixture was poured
into excess 1 M aq NaHSO4 whereupon excess I2 was liberated. The
product was extracted into EtOAc and the brown color of I2 was
discharged with NaHSO3. The organic solution was extracted with
water (2×) and brine, dried (MgSO4), filtered and concentrated to
give 37.4 g of 2 as a tan solid. This material was used in the following
step without purification.
was allowed to warm to 23 ◦C whereupon a granular solid pre-
cipitated. Stirring was continued for 16 h. The mixture was eluted
through a pad of silica with Et2O and concentrated to give 2.76 g of
an oil. This material was purified by chromatography on 125 g of
silica, eluting with 33% → 40% → 50% EtOAc/hexane, to give 2.28 g
(96%) of 4 as an oil. 1H NMR (C6D6) ı 0.18 (s, 3H), 0.28 (s, 3H), 0.86
(t, J = 7, 3H), 1.03 (s, 9H), 1.27 (m, 4H), 1.42 (pent, J = 7, 2H), 1.55 (m,
2H), 1.81 (sept, J = 7, 4H), 2.32 (t, J = 7, 2H), 2.42 (q, J = 7, 2H), 2.88 (s,
3H), 3.05 (s, 3H), 4.60 (t, J = 6.5, 1H), 5.50 (d, J = 11, 1H), 5.67 (m, 1H).
LC/MS (+APCI) 427 (M+NH4)+, 410 (M+H)+, 278 (M + H-TBSOH)+.
2.6.2. N,O-Dimethyl-7-iodo-6-heptynohydroxamate (3)
N-3-(Dimethylamino)propyl-Nꢀ ethylcarbodiimide hydrochlo-
ride (55 g, 0.29 mol) was added to a stirred solution of compound
2 (37.4 g, ∼0.15 mol), N,O-dimethylhydroxylamine hydrochloride
(28.8 g, 0.295 mol), i-Pr2NEt (51 mL, 0.30 mol) and 4-DMAP (3.6 g,
0.030 mol) in 350 mL of dry DMF under N2. The mixture was stirred
for 16 h, then quenched with water and stirred for a further 20 min.
The product was extracted into EtOAc and the organic solution
was washed with water (2×), sat. aq. KH2PO4 (to pH 4.5), water
and brine, dried (MgSO4), filtered and concentrated. The residue
was purified by chromatography on 300 g of silica, eluting with
40% → 50% EtOAc/hexane, to give 30.4 g of a 94:6 mixture of 3 (70%
over two steps) and desiodo-3. 1H NMR (CDCl3) ı 1.57 (pent, J = 7,
2H), 1.74 (pent, J = 7, 2H), 2.40 (t, J = 7, 2H), 2.44 (br t, J = 7, 2H), 3.18
(s, 3H), 3.69 (s, 3H). LC/MS (+APCI) 296 (M+H)+.
2.6.4. N,O-Dimethyl-(6Z,10R)-10-hydroxypentadec-6-en-8-
ynohydroxamate
(5)
Under N2, iodo alkene 4 (1.96 g, 6.61 mmol) was dissolved in
45 mL of diethylamine (dried over 3A molecular sieves and deoxy-
genated with a stream of N2). (Ph3P)2PdCl2 (0.13 g, 0.18 mmol,
2.7 mol%) was added, followed by CuI (Alfa, 99.999%) (73 mg,
0.38 mmol, 5.8 mol%). After 10 min the mixture was purged with
a stream of H2 for 2 min and then placed under an atmosphere of
∼1:1 H2–N2 (balloon). R-(+)-1-Octyn-3-ol (Aldrich, 99%) (1.00 mL,
6.8 mmol) was then added in 0.1-mL portions over 2 min. After
1.7 h, the mixture was partitioned between diethyl ether and water.
The organic extract was washed with 1 M aqueous Na2S2O3 and
then eluted through a pad of Florisil with EtOAc. The eluate was
dried (Na2SO4), filtered and concentrated. The residue was puri-
fied by chromatography on 100 g of silica, eluting with 45% → 50%
EtOAc/hexane, to give 1.57 g (80%) of 5 as an oil. 1H NMR (C6D6) ı
0.86 (t, J = 7, 3H), 1.23 (m, 4H), 1.42 (pent, J = 7, 2H), 1.54 (br pent,
J = 7, 2H), 1.77 (sept, J = 7, 4H), 2.32 (br t, J = 7, 2H), 2.36 (q, J = 7, 2H),
2.84 (s, 3H), 2.95 (br d, J = 6.5, 1H, OH), 3.00 (s, 3H), 4.53 (q, J = 6.5,
1H), 5.53 (d, J = 11, 1H), 5.71 (m, 1H). LC/MS (+APCI) 296 (M + H)+,
278 (M+H–H2O)+.
2.6.3. N,O-Dimethyl-(Z)-7-iodo-6-heptenohydroxamate (4)
Borane·Me2S (1.0 mL, 10 mmol) was added to a stirred, ice-
cooled solution of cyclohexene (2.1 mL, 21 mmol) in 8.5 mL of dry
diethyl ether under N2. The mixture was allowed to warm to 23 ◦C,
resulting in the precipitation of dicyclohexylborane, and was then
re-cooled in ice. A solution of iodo alkyne 3 (2.45 g, 94% pure,
8.0 mmol) in 10 mL of dry diethyl ether was added over 3 min. The
mixture was allowed to warm to 23 ◦C whereupon the solid dis-
solved, then re-cooled in ice. Acetic acid (0.57 mL, 10 mmol) was
added, the solution was allowed to warm to 23 ◦C, then re-cooled
in ice. Ethanolamine (1.3 mL, 22 mmol) was added and the solution

36
S.L. Pfister et al. / Prostaglandins & other Lipid Mediators 123 (2016) 33–39
Table 1
2.6.5. N,O-Dimethyl-(6Z,10R)-10-(tert-
butyldimethylsiloxy)pentadec-6-en-8-ynohydroxamate
Percent decomposition at 75◦ C.
(6)
0.25 h
3.0 h
330 h
tert-Butyldimethylchlorosilane (1.6 g, 11 mmol) was added to a
stirred solution of enynol 5 (1.55 g, 5.25 mmol) and imidazole (1.4 g,
20 mmol) in 10 mL of dry DMF under N2. After 1.5 h, the flask and
contents were transferred to a cold room (5 ◦C) overnight. The mix-
ture was then quenched with water, allowed to warm to 23 ◦C, and
stirred for 2 h. Diethyl ether was added and the mixture was washed
with water (2×), brine, dried (Na2SO4), filtered and concentrated
to give 2.19 g of an oil. This material was purified by chromatog-
raphy on 130 g of silica, eluting with 20% → 25% EtOAc/hexane, to
give 1.96 g (91%) of 6 as an oil. 1H NMR (C6D6) ı 0.18 (s, 3H), 0.28
(s, 3H), 0.86 (t, J = 7, 3H), 1.03 (s, 9H), 1.27 (m, 4H), 1.42 (pent, J = 7,
2H), 1.55 (m, 2H), 1.81 (sept, J = 7, 4H), 2.32 (t, J = 7, 2H), 2.42 (q,
J = 7, 2H), 2.88 (s, 3H), 3.05 (s, 3H), 4.60 (t, J = 6.5, 1H), 5.50 (d, J = 11,
1H), 5.67 (m, 1H). LC/MS (+APCI) 427 (M + NH4)+, 410 (M + H)+, 278
(M + H-TBSOH)+.
HETE analog 1
15(S)-HETE
0.0 0.0
1.6 0.4
ND
94
ND
1
78
2
Data presented as mean SEM. ND, not determined.
60
50
40
30
20
10
0
60
50
40
30
20
10
0
FEMALE
MALE
FEMALE
MALE
***
**
*
*
-9
-8
-7
-6
-5
-9
-8
-7
-6
-5
HETE ANALOG 1 (Log M)
15(S)-HETE (LOG M)
Fig. 2. Contractile responses to 15(S)-HETE and HETE analog 1 in female and male
rabbit pulmonary arteries. Results are expressed as percent KCl (40 mM) con-
tractions, and data points are the mean SEM for n = 8–12. *p < 0.05; **p < 0.01;
***p < 0.001; female vs male.
2.6.6.
(6Z,10R)-10-(tert-Butyldimethylsiloxy)pentadec-6-en-8-ynal (7)
Diisobutylaluminum hydride (5.4 mL of a 1.0 M solution in
toluene) was added dropwise to a stirred, cooled (−78 ◦C bath)
solution of compound 6 (1.76 g, 4.30 mmol) in 38 mL of dry toluene
under N2. After 15 min, the reaction was quenched by dropwise
addition of MeOH (2 mL) and then allowed to warm to 23 ◦C. Diethyl
ether and sat. aq. Rochelle’s salt were added and the mixture was
stirred vigorously for 15 min, then separated. The organic solution
was washed with water (2×), brine, dried (MgSO4), filtered and
concentrated to give 1.50 g of an oil. This material was purified by
chromatography on 100 g of silica, eluting with 12% EtOAc/hexane,
to give 1.32 g (88%) of 7 as an oil which was used without delay in
the next step.
2.7. Statistical analysis
Data are expressed as the mean SEM. Statistical analysis of
the vascular reactivity data was performed by a two-way ANOVA
followed by Bonferroni’s posttest. The cell proliferation and cell
migration studies were analyzed using Student’s t-test. Values were
considered significant at P < 0.05.
2.8. Materials
15(S)-HETE was obtained from Cayman Chemical (Ann Arbor,
MI). [3H]-Thymidine was obtained from Du Pont NEN (Boston,
MA). Cell culture reagents were obtained from Invitrogen (Carlsbad,
CA). Flasks used in cell culture were obtained from Corning (Corn-
ing, NY). Chemicals used in the synthesis of HETE analog 1 were
obtained from Sigma-Aldrich (Milwaukee, WI). All other chemicals
were of reagent grade.
2.6.7. (5Z,11Z,15R)-15-Hydroxyeicosa-5,11-dien-13-ynoic acid
(1)
Potassium tert-butoxide (12 mL of a 1.0 M solution in THF) was
added to a stirred suspension of 4- carboxybutyltriphenylphospho-
nium bromide (2.5 g, 5.6 mmol, dried in vacuo at 65 ◦C) in 20 mL of
dry THF under N2. After 15 min the orange-red mixture was cooled
in a −78 ◦C bath. A solution of aldehyde 7 (1.32 g, 3.77 mmol) in
20 mL of dry toluene was added over 15 min, keeping the internal
temperature between −68 and −71 ◦C. The mixture was allowed to
warm to 23 ◦C, then quenched with sat. aq. KH2PO4 and stirred
for 16 h. The biphasic mixture (pH of aqueous phase = 4.5) was
extracted twice with diethyl ether. The combined organic extract
was dried (Na2SO4), eluted through a silica gel pad with diethyl
ether and concentrated to give 1.54 g of an oil. To a stirred solution
of this material in 20 mL of THF under N2 was added tetrabuty-
lammonium fluoride (1.0 M in THF, 20 mL) in two equal portions
16 h apart. After a further 4 h, the mixture was diluted with half-
saturated brine and extracted with three portions of EtOAc. The
combined organic extract was dried (Na2SO4), filtered and con-
centrated to give 1.7 g of an oil. This material was purified by
chromatography on 125 g of silica, eluting with 50% EtOAc/hexane,
to give 1.08 g (89%) of 1 as an oil. The ratio of 5Z to 5E isomers was
determined to be 94.4: 5.6 by 1H NMR. 1H NMR (C6D6) ı 0.87 (t,
J = 7, 3H), 0.93 (s, 1H, OH), 1.22 (m, 4H), 1.35 (m, 4H), 1.46 (br t,
J = 7, 2H), 1.57 (br t, J = 7, 2H), 1.71 (m, 2H), 2.00 (br q, J = 7, 4H), 2.16
(t, J = 7, 2H), 2.38 (br q, J = 6.5, 2H), 4.45 (br t, J = 6.5, 1H), 5.27 (m,
1H), 5.41 (m, 1H), 5.52 (d, J = 11, 1H), 5.70 (m, 1H). 13C NMR (C6D6)
ı 14.2, 22.9, 24.8, 25.3, 26.6, 27.2, 28.6, 29.5, 30.3, 31.8, 33.1, 38.3,
63.2, 81.8, 95.5, 109.4, 129.0, 131.2, 144.0, 178.8. LC/MS (+ESI) 303
(M+H–H2O)+; (−ESI) 319 (M−H)−.
3. Results
Stability assays at 75◦ C indicated that 15(S)-HETE decomposed
to the extent of 78 2% in 15 min, while HETE analog 1 (racemic
form) showed only 1.6 0.4% decomposition after 3 h (Table 1).
Thus HETE analog 1 proved suitable for extended storage and peri-
odic sampling in various forms (neat, solution, formulation) with
no special precautions other than routine refrigeration at 5 ◦C under
nitrogen.
The ability of 15(S)-HETE and HETE analog 1 to contract pul-
monary arteries from female and male rabbits is shown in Fig. 2.
Both 15(S)-HETE and HETE analog 1 produced a concentration-
dependent vasoconstriction that was significantly greater in
females compared to males. The EC50 values were not differ-
ent between 15(S)-HETE and HETE analog 1 (see Table 2). While
both compounds had a greater effect in females, the maximal
response was greater with 15(S)-HETE compared to HETE analog 1
(maximal contraction in females; 49.7 4.1% vs 27.3 4.4%; 15(S)-
HETE vs HETE analog 1). In males the maximal contraction to the
two compounds was not different (maximal contraction in males;
19.3 2.8% vs 24.1 2.3%; 15(S)-HETE vs HETE analog 1).
The effect of 15(S)-HETE and HETE analog 1 on vascular
smooth muscle cell proliferation is shown in Fig. 3. Both com-

S.L. Pfister et al. / Prostaglandins & other Lipid Mediators 123 (2016) 33–39
37
Table 2
HETE ANALOG 1
Comparisons of responses to 15(S)-HETE and HETE analog 1 in female and male
pulmonary artery.
1.0
0.8
0.6
0.4
0.2
0.0
15(S)-HETE
HETE analog 1
*
Female
Male
EC50 (␮M)
maximal contraction (% of KCl)
542.5 203.0
49.7 4.1
953.7 326.7
27.3 4.4
EC50 (␮M)
maximal contraction (% of KCl)
618.1 272.6
19.3 2.8
1250.0 406.1
24.1 2.3
Values are the mean SEM EC50, concentration of agonist yielding 50% of the max-
imal response. EC50 values were calculated using the nonlinear regression sigmoid
curve fitting program of PRISM.
*
15(S)-HETE
5
BASAL
10-8
10-7
10-6
*
Fig. 4. Effect of HETE analog 1 on [3H] thymidine incorporation in PC3 cells. The
values are normalized to [3H] thymidine uptake by control cells under basal con-
ditions. Results are the mean SEM from six independent experiments conducted
with quadruplicate samples in each experiment. Symbols above a column indicate a
statistical comparison between the indicated sample and the control cells (*p < 0.01).
4
3
*
2
1
0
Table 3
Comparisons of responses to 15(S)-HETE and HETE analog 1 in vascular smooth
muscle and PC3 cells.
15(S)-HETE
HETE analog 1
VSMC
PC3
EC50 (␮M)
EC50 (␮M)
5.6 3.5
ND
634 64.9*
1.5 0.9
BASAL
10-7
10-6
10-5
Values are the mean SEM EC50, concentration of agonist yielding 50% of the max-
imal response. EC50 values were calculated using the nonlinear regression sigmoid
curve fitting.
HETE ANALOG 1
*
P < 0.05 HETE analog 1 vs 15(S)-HETE. ND, not determined.
5
4
3
2
1
0
shows an example of cell migration through the filters in the PC3
cells and the bottom panel is the averaged migration results after
24 h.
*
4. Discussion
In this study, the biological characterization of HETE analog 1
showed that it contracted rabbit pulmonary artery. However, it was
less potent and efficacious than the naturally occurring metabo-
lite 15(S)-HETE. An interesting observation was that 15(S)-HETE
showed a greater sex-specific difference than did HETE analog 1.
While the vasoconstrictor mechanisms of many arachidonic acid
metabolites have been well defined, less is known about 15-HETE.
Early studies reported that 15-HETE constricted isolated blood ves-
sels, including pulmonary arteries [25–27]. This contrasts with
other work showing that 15-HETE relaxes preconstricted vessels
[28,29]. The exact cellular processes that contribute to 15-HETE
vasoconstriction were addressed primarily by the work of Zhu
and colleagues [30–32]. In pulmonary vascular smooth muscle
cells of neonatal rabbits exposed to hypoxia or normoxia, 15-HETE
decreased whole cell Kv currents. With rat pulmonary artery vas-
cular smooth muscle cells, 15-HETE acted at Kv 1.5, Kv 1.2 and
Kv 3.4 channels [33]. PKC inhibition also blocked 15-HETE-induced
vasoconstriction in the isolated rat aortas. It is not known whether
15-HETE activates different vasoconstrictor mechanisms in females
compared to males. It also remains unresolved whether 15-HETE
acts on a specific membrane receptor like many other biologically
active arachidonic acid metabolites. The relaxations to 15-HETE
noted by other studies happened in vessels preconstricted with
either the thromboxane mimetic U46619 or prostaglandin PGF2␣
raising the possibility that the mechanism of effect is due to non-
specific blockade of thromboxane or PGF receptors. HETE analog
1 will be an important tool for aiding in the delineation of cellular
BASAL
10-7
10-6
10-5
Fig. 3. 15(S)-HETE- (top panel) and HETE analog 1 (bottom panel) stimulated [3H]
thymidine incorporation in rabbit pulmonary artery vascular smooth muscle cells.
The values are normalized to [3H] thymidine uptake by control cells under basal con-
ditions. Results are the mean SEM from six independent experiments conducted
with quadruplicate samples in each experiment. Symbols above a column indicate a
statistical comparison between the indicated sample and the control cells (*p < 0.01).
pounds increased [3H]-thymidine incorporation but the effect was
greater with 15(S)-HETE (top panel) compared to HETE analog 1.
Results indicated that 15(S)-HETE was more potent than HETE ana-
log 1, as no effect on proliferation was seen with 10 ␮M HETE
analog 1 but the same concentration of 15(S)-HETE caused a 2-
fold increase compared to control cells (*p < 0.05). In the PC-3
cells, HETE analog 1 had opposite effects to those seen in the
vascular smooth cells. A concentration-dependent decrease in
[3H]-thymidine incorporation was observed (Fig. 4). Table 3 com-
pares the EC50 for compounds in vascular smooth muscle and
PC3 cells on cell proliferation. In a separate study, HETE analog
1 caused a concentration-dependent decrease in the migration of
cells through the filter pores of Boyden chamber as measured by
X-celligence and compared to control cells (Fig. 5). The top panel

38
S.L. Pfister et al. / Prostaglandins & other Lipid Mediators 123 (2016) 33–39
HETE ANALOG 1
while 15(S)-HETE is antiangiogenic [38,39]. Using PC3 cells, Shap-
0.5
0.4
0.3
0.2
0.1
0.0
pel and coworkers [6] showed that exogenous 15(S)-HETE caused
a concentration-dependent inhibition of proliferation with the soft
agar colony formation assay (IC50 of ∼ 30 ␮M). We tested HETE ana-
log 1 on [3H]-thymidine incorporation in PC3 cells and also found
it decreased cell proliferation in a concentration-dependent man-
ner with an IC50 of ∼2 ␮M. The ability of cancer cells to undergo
migration and invasion contribute to metastatic characteristic of
various cancers. A migration assay with the PC3 cells showed that
HETE analog 1 decreased cell migration over a 72 h period. These
results support the experimental evidence in studies that have
reported that overexpression of 15-lipoxygenase-2 in PC3 cells
causes these cells to stop proliferating and become senescent [39].
Because of the gaps in knowledge on the precise contribution of
15-lipoxygenase metabolites to cancer progression, this area of
investigation remains an active research focus and benefits from
the availability of a stable HETE analog.
Vehicle
10-7
10-6
10-5
0
12
24
36
48
60
72
Time (hours)
5. Conclusions
The present studies confirm that 15(S)-HETE has a role in
both pulmonary artery vasoconstriction and cellular proliferation.
These effects combined with previous work by others suggest
that 15(S)-HETE could potentially contribute to a variety of dis-
eases like hypertension, atherosclerosis and cancer. Furthermore,
15-lipoxygenase metabolites have been shown to contribute to
many other conditions, including obesity, diabetes and respira-
tory disease [40–42]. However, many details of 15(S)-HETE cell
signal transduction mechanisms are not understood. An important
contribution of the current work is the synthesis and biological
characterization of the stable HETE analog 1. This analog will be
an integral tool in future studies to further elucidate the mecha-
nism of endogenous 15(S)-HETE. Knowledge of these mechanisms
will advance eicosanoid research and may ultimately translate to
improved pharmacological treatments of disease.
100
75
50
25
0
*
*
vehicle
10-7
10-6
10-5
Acknowledgements
Fig. 5. Effect of HETE analog 1 on cell migration in PC3 cells. Top panel is repre-
sentative example of cell migration through filters over the 72 h collection period
and bottom panel is summarized results of cell migration at the 24 h time point.
Results are the mean SEM (n = 4). Symbols above a column indicate a statistical
comparison between the specified sample and the control cells (*p < 0.05).
The authors thank Dr. John Liao, Ms. Thivashnee Pillay, and Mr.
Adam Pfeiffer for their technical assistance, and Ms. Gretchen Barg
for her secretarial assistance. Studies at Medical College of Wis-
consin were supported by grants from the National Heart, Lung
and Blood Institute (HL093181) and American Heart Association
(0151421Z).
mechanisms that contribute to sex-specific differences in 15-HETE-
induced alterations in vascular tone.
Many studies have shown that 15-HETE increases vascular
smooth muscle cell proliferation and this was confirmed by our
studies using rabbit pulmonary artery smooth muscle cells. HETE
analog 1 similarly increased proliferation, although the analog
was less potent and maximal effect was reduced when compared
to 15(S)-HETE. This effect of 15-HETE and HETE analog 1 was
not unexpected, as 15-HETE has been shown to have mitogenic
effects in other vascular cells including endothelial cells and fibrob-
lasts. Atherosclerosis is associated with increased expression of
15-lipoxygenase and 15-HETE. Numerous studies have reported a
role for 15-HETE in angiogenesis that contributes to progression of
atherosclerosis [34–37].
In contrast to the vascular cells, the role of 15-lipoxygenase and
15(S)-HETE in cancer is unclear and controversial [8]. To review,
two mammalian forms of 15-lipoxygenase are known to exist,
15-lipoxygenase-1 and 15-lipoxygenase-2. 15-Lipoxygenase-1 uti-
lizes arachidonic acid to form 15(S)-HETE and to a lesser extent,
12(S)-HETE. 15-lipoxygenase-1 also utilizes linoleic acid to form
13(S)-hydroxyoctadecadienoic acid (13-HODE). The preferred sub-
strate for 15-lipoxygenase-2 is arachidonic acid which is converted
to 15(S)-HETE. In prostate cancer, 12(S)-HETE is proangiogenic
References
[1] A.D. Dobrian, D.C. Lieb, B.K. Cole, D.A. Taylor-Fishwick, S.K. Chakrabarti, J.L.
Nadler, Functional and pathological roles of the 12- and 15-lipoxygenases,
Prog. Lipid Res. 50 (2011) 115–131.
[2] D. Zhu, Y. Ran, Role of 15-lipoxygenase/15-hydroxyeicosatetraenoic acid in
hypoxia-induced pulmonary hypertension, J. Physiol. Sci. 63 (2012) 163–172.
[3] S.L. Pfister, Role of lipoxygenase metabolites of arachidonic acid in enhanced
pulmonary artery contractions of female rabbits, Hypertension 57 (2011)
825–832.
[4] J. Wittwer, M. Hersberger, The two faces of the 15-lipoxygenase in
atherosclerosis, Prostaglandins Leukot. Essent. Fatty Acids 77 (2007) 67–77.
[5] H.J. Brinkman, M.F. van Buul-Wortelboer, J.A. van Mourik, Selective
conversion and esterification of monohydroxyeicosatetraenoic acids by
human vascular smooth muscle cells: relevance to smooth muscle cell
proliferation, Exp. Cell Res. 192 (1991) 87–92.
[6] S.B. Shappell, R.A. Gupta, S. Manning, R. Whitehead, W.E. Boeglin, C.
Schneider, T. Case, J. Price, G.S. Jack, T.M. Wheeler, R.J. Matusik, A.R. Brash,
R.N. Dubois, 15S-Hydroxyeicosatetraenoic acid activates peroxisome
proliferator-activated receptor gamma and inhibits proliferation in PC3
prostate carcinoma cells, Cancer Res. 61 (2001) 497–503.
[7] E.J. Werner, R.W. Walenga, R.L. Dubowy, S. Boone, M.J. Stuart, Inhibition of
human malignant neuroblastoma cell DNA synthesis by lipoxygenase
metabolites of arachidonic acid, Cancer Res. 45 (1985) 561–563.
[8] J. Moreno, New aspects of the role of hydroxyeicosatetraenoic acids in cell
growth and cancer development, Biochem. Pharmacol. 77 (2009) 1–10.

S.L. Pfister et al. / Prostaglandins & other Lipid Mediators 123 (2016) 33–39
39
[9] S.E. Campbell, P.R. Musich, S.G. Whaley, J.B. Stimmel, L.M. Leesnitzer, S.
Dessus-Babus, M. Duffourc, W. Stone, R.A. Newman, P. Yang, K. Krishnan,
Gamma tocopherol upregulates the expression of 15-S-HETE and induces
growth arrest through a PPAR gamma-dependent mechanism in PC-3 human
prostate cancer cells, Nutr. Cancer 61 (2009) 649–662.
[10] R.S. Jackson 2nd, S.J. Van Dyken, M.D. McCartney, J.L. Ubels, The eicosanoid,
15-(S)-HETE, stimulates secretion of mucin-like glycoprotein by the corneal
epithelium, Cornea 20 (2001) 516–521.
[11] J.E. Jumblatt, L.T. Cunningham, Y. Li, M.M. Jumblatt, Characterization of human
ocular mucin secretion mediated by 15(S)-HETE, Cornea 21 (2002) 818–824.
[12] D.A. Gamache, Z.Y. Wei, L.K. Weimer, S.T. Miller, J.M. Spellman, J.M. Yanni,
Corneal protection by the ocular mucin secretagogue 15(S)-HETE in a rabbit
model of desiccation-induced corneal defect, J. Ocul. Pharmacol. Ther. 18
(2002) 349–361.
[13] F. Mantelli, M. Massaro-Giordano, I. Macchi, A. Lambiase, S. Bonini, The
cellular mechanisms of dry eye: from pathogenesis to treatment, J. Cell.
Physiol. 228 (2013) 2253–2256.
[27] B. Tesfamariam, M.L. Brown, R.A. Cohen, 15-Hydroxyeicosatetraenoic acid and
diabetic endothelial dysfunction in rabbit aorta, J. Cardiovasc. Pharmacol. 25
(1995) 748–755.
[28] T. Kriska, C. Cepura, L. Siangjong, T.C. Wan, J.A. Auchampach, A. Shaish, D.
Haratz, G. Kumar, J.R. Falck, K.M. Gauthier, W.B. Campbell, Effect of human
15-lipoxygenase-1 metabolites on vascular function in mouse mesenteric
arteries and hearts, Prostaglandins Other Lipid Mediat. 106 (2013) 8–15.
[29] M. Takahashi, N. Ozaki, S. Kawakita, M. Nozaki, Y. Saeki, Vascular effects of
15-hydroperoxyeicosatetranoic acid and 15-hydroxyeicosatetraenoic acid on
canine coronary arteries, Jpn. J. Pharmacol. 37 (1985) 325–329.
[30] L. Guo, Z. Qiu, L. Zhang, S. Chen, D. Zhu, Hypoxia suppresses Kv 2.1 channel
expression through endogenous 15-hydroxyeicosatetraenoic acid in rat
pulmonary artery, J. Physiol. Sci. 60 (2010) 373–381.
[31] L. Guo, X. Tang, X. Chu, L. Sun, L. Zhang, Z. Qiu, S. Chen, Y. Li, X. Zheng, D. Zhu,
Role of protein kinase C in 15-HETE-induced hypoxic pulmonary
vasoconstriction, Prostaglandins Leukot. Essent. Fatty Acids 80 (2009)
115–123.
[14] L.W. Schneider, R.E. Conrow, D.A. Gamache, T. Pasquine, J.M. Yanni, H.G.
Bhagat, Hydroxyeicosatetraenoate salts, compositions and methods of use in
treating dry eye disorders, US Patent 6,429,227 (2002).
[32] C.L. Lu, H. Ye, X.B. Tang, D.L. Zhu, ERK1/2 signaling pathway is involved in
15-hydroxyeicosatetraenoic acid-induced hypoxic pulmonary
vasoconstriction, Sheng Li Xue Bao 57 (2005) 605–611.
[15] R.E. Conrow, P. Harrison, M. Jackson, S. Jones, C. Kronig, I.C. Lennon, S.
Simmonds, Manufacture of
[33] Q. Li, R. Zhang, C.L. Lü, Y. Liu, Z. Wang, D.L. Zhu, The role of subtypes of
voltage-gated K+ channels in pulmonary vasoconstriction induced by
15-hydroeicosatetraenoic acid, Yao Xue Xue Bao 41 (2006) 412–417.
[34] T.M. Bocan, W.S. Rosebury, S.B. Mueller, S. Kuchera, K. Welch, A. Daugherty,
J.A. Cornicelli, A specific 15-lipoxygenase inhibitor limits the progression and
monocyte-macrophage enrichemnt of hypercholesterolemia induced
atherosclerosis in the rabbit, Atherosclerosis 136 (1998) 203–216.
[35] H. Kuhn, J. Belkner, S. Zaiss, T. Fahrenklemper, S. Wohlfeil, Involvement of
15-lipoxygenase in early stage atherosclerosis, J. Exp. Med. 179 (1994)
1903–1911.
(5Z,8Z,11Z,13E)(15S)-15-hydroxyeicosa-5,8,11,13-tetraenoic acid sodium salt
for clinical trials, Org. Process Res. Dev. 15 (2011) 301–304.
[16] R.E. Conrow, P.G. Klimko, Heteroatom-interrupted analogs of
15-hydroxyeicosatetraenoic acid and methods of use, US Patents 6,441,035
(2002); 6,436,994 (2002); 6,331,566 (2001).
[17] P.G. Klimko, M.R. Hellberg, J.R. Falck, R.E. Conrow, Hydroxyeicosatetraenoic
acid analogs and methods of their use in treating dry eye disorders. US
Patents 6,803,385 (2004); 6,552,084 (2003).
[18] S.L. Pfister, W.B. Campbell, Role of endothelium-derived metabolites of
arachidonic acid in enhanced pulmonary artery contractions in female
rabbits, Hypertension 27 (1996) 43–48.
[19] S.L. Pfister, D.A. Kotulock, W.B. Campbell, Vascular smooth muscle
thromboxane A2 receptors mediate arachidonic acid-induced sudden death
in rabbits, Hypertension 29 (part 2) (1997) 303–309.
[20] K. Nithipatikom, A.D. Gomez-Granados, A.T. Tang, A.W. Pfeiffer, C.L. Williams,
W.B. Campbell, Cannabinoid receptor type 1 (CB1) activation inhibits small
GTPase RhoA activity and regulates motility of prostate carcinoma cells,
Endocrinology 153 (2012) 29–41.
[21] S.E. Denmark, S.-M. Yang, Sequential ring-closing metathesis/Pd-catalyzed,
Si-assisted cross coupling reactions: general synthesis of highly substituted
unsaturated alcohols and medium-sized rings containing a 1,3-cis-cis diene
unit, Tetrahedron 60 (2004) 9695–9708.
[22] H.C. Brown, C.D. Blue, D.J. Nelson, N.G. Bhat, Vinylic organoboranes. 12.
Synthesis of (Z)-1-halo-1-alkenes via hydroboration of 1-halo-1-alkynes
followed by protonolysis, J. Org. Chem. 54 (1989) 6064–6067.
[23] A. Elangovan, Y.-H. Wang, T.-I. Ho, Sonogashira coupling reaction with
diminished homocoupling, Org. Lett. 5 (2003) 1841–1844.
[24] M.M. Midland, R.S. Graham, Asymmetric reduction of ␣,␤-acetylenic ketones
with B-3-pinanyl-9-borabicyclo[3.3.1]nonane: (R)-(+)-1-octyn-3-ol, Org.
Synth. Collect. 7 (1990) 402.
[25] K.E. Burhop, W.M. Selig, A.B. Malik, Monohydroxyeicosatetraenoic acids
(5-HETE and 15-HETE) induce pulmonary vasoconstriction and edema, Circ.
Res. 62 (1988) 687–698.
[26] H. Matsuda, K. Miyatake, S.E. Dahlen, Pharmacodynamics of
15(S)-hydroperoxyeicosatetraenoic (15-HPETE) and
[36] S. Rong, Q. Cao, M. Liu, J. Seo, L. Jia, E. Boudyguina, A.K. Gebre, P.L. Colvin, T.L.
Smith, R.C. Murphy, N. Mishra, J.S. Parks, Macrophage 12/15 lipoxygenase
expression increases plasma and hepatic lipid levels and exacerbates
atherosclerosis, J. Lipid Res. 53 (2012) 686–695.
[37] S. Sendobry, J.A. Cornicelli, K. Welch, T. Bocan, B. Tait, B.K. Trivedi, N. Colbry,
R.D. Dyer, S.J. Feinmark, A. Daugherty, Attenuation of diet-induced
atherosclerosis in rabbits with a highly selective 15-lipoxygenase inhibitor
lacking significant antioxidant properties, Br. J. Pharmacol. 120 (1997)
1199–1206.
[38] D. Nie, S. Krishnamoorthy, R. Jin, K. Tang, Y. Chen, Y. Qiao, A. Zacharek, Y. Guo,
J. Milanini, G. Pages, K.V. Honn, Mechanisms regulating tumor angiogenesis
by 12-lipoxygenase in prostate cancer cells, J. Biol. Chem. 281 (2006)
18601–18609.
[39] U.P. Kelavkar, W. Glasgow, S.J. Olson, B.A. Foster, S.B. Shappell, Overexpression
of 12/15-lipoxygenase, an ortholog of human 15-lipoxygenase-1, in the
prostate tumors of TRAMP mice, Neoplasia 6 (2004) 821–830.
[40] D.C. Lieb, J.J. Brotman, M.A. Hatcher, M.S. Aye, B.K. Cole, B.A. Haynes, S.D.
Wohlgemuth, M.A. Fontana, H. Beydoun, J.L. Nadler, A.D. Dobrian, Adipose
tissue 12/15 lipoxygenase pathway in human obesity and diabetes, J. Clin.
Endocrinol. Metab. 99 (2014) E1713–E1720.
[41] A. Brunnstrom, Y. Tryselius, S. Feltenmark, E. Andersson, H. Leksell, A. James,
B. Mannervik, B. Dahlen, H.E. Claesson, On the biosynthesis of 15-HETE and
eoxin C4 by human airway epithelial cells, Prostaglandins Other Lipid Mediat.
121 (2015) 83–90.
[42] Y.S. Song, M.S. Kim, D.H. Lee, D.K. Oh, D.Y. Yoon, 15-Hydroxyeicosatetraenoic
acid inhibits phorbol-12-myristate-13-acetate-induced MUC5AC expression
in NCI-H292 respiratory epithelial cells, J. Microbiol. Biotechnol. 25 (2015)
589–597.
15(S)-hydroxyeicosatetraenoic acid (15-HETE) in isolated arteries from guinea
pig, rabbit, rat and human, J. Pharmacol. Exp. Ther. 273 (1995) 1182–1189.