NMR and HPLC-MS/MS analysis of synthetically prepared linoleic acid diol glucuronides

Article abstract of DOI:10.1016/j.chemphyslip.2006.01.007

Hydroxylated fatty acids are important mediators of many physiological and pathophysiological processes in a variety of human tissues. Recent evidence shows that in humans many of these are ultimately excreted in the urine as the glucuronide conjugates. I

Full text of DOI:10.1016/j.chemphyslip.2006.01.007

Chemistry and Physics of Lipids 140 (2006) 75–87
NMR and HPLC-MS/MS analysis of synthetically prepared
linoleic acid diol glucuronides
Jie Yang, Martha D. Morton∗∗, Dennis W. Hill, David F. Grant∗
Department of Pharmaceutical Sciences (JY, DH, DFG) and Chemistry (MDM), University of Connecticut,
Storrs, CT 06269, United States
Received 30 September 2005; received in revised form 18 January 2006; accepted 19 January 2006
Available online 9 February 2006
Abstract
Hydroxylated fatty acids are important mediators of many physiological and pathophysiological processes in a variety of human
tissues. Recent evidence shows that in humans many of these are ultimately excreted in the urine as the glucuronide conju-
gates. In this paper we describe a general approach for the chemical synthesis of glucuronide conjugate derivatives of fatty
acids. The synthesis strategy employs three steps (epoxidation, hydrolysis and glucuronidation) using methyl linoleate as a model
non-hydroxylated starting compound. Hydroxylated starting compounds would require only the glucuronidation step. NMR and
HPLC-MS/MS experiments were used to help determine the structure of the synthesized glucuronide conjugates and to identify
fragmentation product ions useful for discriminating positional isomers in biological samples. This synthetic strategy should prove
useful for generating analytical standards in order to identify and quantify glucuronide metabolites of hydroxylated fatty acids
in humans.
© 2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Linoleic acid diols; Hydroxylated fatty acids; Glucuronides; NMR; HPLC-MS/MS
1. Introduction
nous compounds which are glucuronidated and excreted
in the urine. Many fatty acids and fatty acid metabo-
lites are known to be glucuronidated prior to excretion
and many of these have been shown to be associated
with various pathophysiological conditions. For exam-
ple, linoleic acid is metabolized in vivo and in vitro to a
variety of compounds, including some which have been
reported to be cytotoxic (Sevanian et al., 1979; Laethem
et al., 1992; Moghaddam et al., 1996). Glucuronidation
of linoleic acid and linoleic acid diols by human liver
microsomes was studied earlier to investigate the role of
glucuronide conjugation in the metabolism of linoleic
acid diols (Jude et al., 2000). Street et al. (1996a,b)
identified 16 C16 and C18 glucuronide-conjugated fatty
acid diols in the urine samples of patients with gen-
eralized peroxisomal disorders by using negative ion
Glucuronides are a pharmaceutically important class
of compounds (Rukhman et al., 2001). Some are antial-
lergenic (Takagaki et al., 1999), or have antimalarial
activity (Ramu and Baker, 1995), however, they are pri-
marily known for their importance in drug metabolism
and elimination. There are also a variety of endoge-
∗
Corresponding author. Tel.: +1 860 486 4265;
f∗a∗x: +1 860 486 4998.
Co-corresponding author. Present address: Department of Chem-
istry, University of Connecticut, Storrs, CT 06269, United States.
Tel.: +1 860 486 4069; fax: +1 860 486 2981.
E-mail addresses: martha.morton@uconn.edu (M.D. Morton),
david.grant@uconn.edu (D.F. Grant).
0009-3084/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.chemphyslip.2006.01.007

76
J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
fast atom bombardment-mass spectrometry. The major
compounds found were glucuronides of 9,10 and 12,13
linoleic acid diol. Glucuronides of arachidonic acid
metabolites are also excreted in human urine (Costa et
al., 1996; Watzer et al., 2000; Turgeon et al., 2003).
The in vivo metabolic pathway that is thought to
produce linoleic and arachidonic acid glucuronides
(epoxidation (Sevanian et al., 1979; Laethem et al.,
1992; Moghaddam et al., 1996), hydrolysis (Watabe
and Akamatsu, 1972; Westkaemper and Hanzlik, 1981;
Bellucci et al., 1989; Wistuba et al., 1989) and glu-
curonidation (Dutton, 1980; Kirkpatrick et al., 1984;
Radominska et al., 1997)) is very similar to the chemical
synthesis strategy employed here using methyl linoleate
as the starting compound.
semi-preparative silica column, analytical silica column
and/orreversephaseC8 columnfromVarianandAgilent,
respectively. HPLC-MS experiments were performed on
a Waters 616/626 HPLC system coupled to a Micromass
Quattro II Tandem Mass Spectrometer (Fisons Instru-
ments, Manchester, UK) using an electrospray (ES)
or atmospheric pressure chemical ionization sources
(APCI). All HPLC-MS/MS analysis was performed on
an HP1090 HPLC system coupled to a Micromass Q-
ToF2 mass spectrometer (Micromass, Manchester, UK).
All column chromatographic separations were done on
silica gel 60 (230–400 mesh), whereas all TLC (silica
gel) developments were performed on silica gel coated
on aluminum 250 ␮m layer sheets (Whatman, Maid-
stone, UK).
2. Experimental procedures
2.1. Synthesis procedures
Schemes 1 and 2 show the overall synthesis strategy
for synthesizing linoleic acid diol glucuronides (9) using
methyl linoleate (5) as the starting fatty acid.
Chemicals were purchased from either Sigma (St.
Louis, MO, USA) or Aldrich Chemical Ltd. (Milwau-
kee, WI, USA). Solvents used for chemical synthesis
were purchased from commercial sources were of ana-
lytical grade and were used without further purification
unless otherwise stated. All NMR spectra were recorded
on DRX-400 and Avance 500 MHz Bruker NMR spec-
trometers using tetramethylsilane as internal standard
and CDCl3 as solvent for linoleic derivatives. A solvent
mixture of CDCl3:CD3OD was used for glucuronide
derivatives of linoleic acid diol. HPLC purification was
performed with a Shimadzu HPLC 10A VP series using a
2.1.1. Methyl-d-glucuronate (2)
The procedure of Fehlhaber et al. (1987) was used to
prepare methyl-d-glucuronate.
2.1.2. 1,2,3,4-Tetra-O-(3-fluorobenzoyl)-d-methyl
glucuronate (3)
Methyl glucuronate (2) (1.6 g, 7.7 mmol) and 4-
dimethylamino pyridine (280 mg, 2.3 mmol) were dis-
Scheme 1.

J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
77
Scheme 2.
solved in 25 mL pyridine. The mixture was cooled in
an ice-water bath for 10 min. 3-Fluorobenzoyl chloride
(10 g, 53 mmol) was added drop wise to this mixture over
a 25 min period. The solution was stirred for 24 h and
then methanol (5 mL) was added and stirred for another
hour. The solution was diluted with CH2Cl2 (100 mL)
and washed with 1N HCl (100 mL) and water (100 mL).
TheorganicphasewasdriedoveranhydrousNa2SO4 and

78
J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
solvent evaporated under reduced pressure. The residual
oil was purified by column chromatography over silica
gel using hexane/ethyl acetate (3:1) as eluent to afford 3
(4.3 g, 81% yield). 1H NMR (400 MHz, CDCl3): δ 3.71
(s, 3H, COOCH3), 4.76 (d, 1H, J = 10.0 Hz, H-5), 5.68
(dd, 1H, J = 3.67, 10.0 Hz, H-2), 5.77 (dd, 1H, J = 9.9,
9.8 Hz, H-4), 6.23 (dd, 1H, J = 9.8, 9.8 Hz, H-3), 6.90
(d, 1H, J = 3.6 Hz, H-1), 7.49 (m, 16H, Ph). 13C NMR
(500 MHz, CDCl3): δ 171.3, 167.0, 164.7, 164.3, 164.2,
163.9, 163.8, 161.4, 130.6, 125.8, 121.0, 116.9, 90.0,
71.0, 70.2, 70.08, 70.03, 53.2.
pounds as a mixture. This mixture of epoxides was then
directly taken for hydrolysis.
2.1.5. 9,10- and 12,13-dihydroxy methyl linoleate
(7a and 7b)
A mixture of monoepoxides of methyl linoleate (6,
1.37g), acetonitrile (60 mL) and perchloric acid (70%,
0.7 mL) in water (55 mL) were stirred at room tempera-
tureovernight. Thereactionmixturewasthenneutralized
with sodium bicarbonate and concentrated under vac-
uum. The product was extracted in ethyl ether. The ether
layer was then washed with saturated sodium chloride
solution. The organic layer was dried over anhydrous
sodium sulphate and evaporated using a rotary evapora-
tor to afford an oil which contained both diols. The two
regioisomeric diols were then purified by normal phase
HPLC using isopropanol:hexane (3:97) as the mobile
phase to obtain pure 7a (450 mg, 31% yield) and 7b
(432 mg, 30% yield).
2.1.3. 2,3,4-Tri-O-(3-fluorobenzoyl)-α-d-methyl-
glucuronide bromide (4)
Hydrogen bromide in acetic acid (30 wt%,
15 mL) was added to a solution of 1,2,3,4-tetra-O-
(3-fluorobenzoyl)-d-methyl glucuronate (3) (4.3 g,
6.1 mmol) in 10 mL CH2Cl2, and the solution was
stirred overnight. It was then diluted with CH2Cl2
(100 mL) and washed with water (100 mL) followed
by saturated aqueous NaHCO3 (300 mL). The organic
phase was dried over anhydrous Na2SO4 and concen-
trated by evaporation. The residual oil was purified by
column chromatography on silica gel using hexane/ethyl
acetate (4:1) to give 4 (3.3 g, 85% yield). 1H NMR
(400 MHz, CDCl3) δ 3.71 (3H, s, COOCH3), 4.86
(1H, d, J = 10.2 Hz, H-5), 5.35 (1H, dd, J = 4.1, 9.9 Hz,
H-2), 5.73 (1H, dd, J = 10.0, 9.8 Hz, H-4), 6.22 (1H,
dd, J = 9.8, 9.8 Hz, H-3), 6.86 (1H, d, J = 4.0 Hz, H-1),
7.48 (12H, m, Ph). 13C NMR (100 MHz, CDCl3):
δ 166.5, 164.3, 164.3, 164.08, 164.05, 164.01, 163.7,
130.2, 125.6, 120.9, 116.7, 85.2, 72.2, 71.0, 70.3,
69.1, 53.2.
2.1.6. 9,10-Dihydroxy methyl linoleate (7a)
1HNMR(500 MHz, CDCl3): δ 0.88(t, 3H, J = 6.5 Hz,
CH3), 1.46 (m, 18H, H-3 to H-7, H-14 to H-17),
2.04 (m, 2H, H-2), 2.30 (m, 4H, H-8, H-11), 2.80
(bs, 2H, OH), 3.45 (bs, 2H, H-9, H-10), 3.66 (s,
3H, COOCH3), 5.48 (m, 2H, H-12, H-13). 13C NMR
(125 MHz, CDCl3): δ 174.4, 133.3, 124.9, 74.0, 73.8,
51.5, 34.1, 33.6, 31.9, 31.7, 31.6, 29.5, 28.9, 27.4, 25.7,
25.4, 24.9, 22.6, 14.1.
2.1.7. 12,13-Dihydroxy methyl linoleate (7b)
1HNMR(500 MHz, CDCl3): δ 0.87(t, 3H, J = 7.0 Hz,
CH3), 1.28 (bs, 12H, H-4 to H-7 and H-16 to H-
17), 1.32 (bs, 2H, H-15), 1.46 (bs, 2H, H-4), 1.60
(t, 2H, J = 7.0 Hz, H-3), 2.02 (q, 2H, J = 7.2 Hz, H-
8), 2.25 (bs, 2H, H-11), 2.27 (t, 2H, J = 7.3 Hz, H-
2), 2.62 (bs, 2H, OH), 3.42 (bs, 2H, H-12, H-
13), 3.64 (s, 3H, COOCH3), 5.42 (m, 1H, H-10),
5.52 (m, 1H, H-9). 13C NMR (125 MHz, CDCl3):
δ 174.2, 132.9, 124.8, 73.8, 73.7, 51.3, 33.9, 33.5,
31.7, 31.5, 29.3, 28.9, 27.2, 25.4, 25.2, 24.7, 22.4,
13.8.
2.1.4. 9,10- and 12,13-monoepoxide of methyl
linoleate (6)
Methyl linoleate (5, 3.0 g) was dissolved in
dichloromethane (13 mL), followed by the addition
of methyl trioxorhenium (VII, 15 mg) and pyridine
(0.1 mL). Hydrogen peroxide (30 wt%, 1.8 mL) was then
added drop wise over a 30 min period and the mixture
was stirred for 1–2 h. After completion of the reaction
(monitored by TLC), the aqueous phase was discarded.
Excess H2O2 in the yellow organic phase was decom-
posed by the addition of a catalytic amount of MnO2
until the yellow color totally disappeared. The solution
was filtered and dried over anhydrous Na2SO4. The sol-
vent was removed using a rotary evaporator. The crude
product which contained diepoxides and some unreacted
methyl linoleate was purified by column chromatogra-
phy over silica gel using ether:dichloromethane (2:98)
as eluent to afford 1.37 g (44% yield) of the title com-
2.1.8. 2,3,4-Tri-O-(3-fluorobenzoyl)-β-d-
glucuronides of 9,10- and 12,13-dihydroxy methyl
linoleate (8)
Pure 9,10- or 12,13-diol methyl linoleates (1.0 g
∼3 mmol) and CdCO3 (780 mg, 4.5 mmol) in toluene
(50 mL) were heated to reflux using a Dean–Stark water
separator while stirring under nitrogen atmosphere. A
solution of 4 (3.15 g, ∼4.5 mmol) in toluene (20 mL) was

J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
79
Table 1
13C NMR chemical shift assignment of the 2,3,4-tri-O-(3-fluorobenzoyl)-d-glucuronides of 9,10- and 12,13-dihydroxy methyl linoleates (in CDCl3)a
Assignment
8a
8b
8c
8d
8e
8f
8g
8h
8i
1ꢀ
2ꢀ
101.6
72.1
72.7
70.3
72.6
166.8
174.3
34.0
32.6
133.9
125.0
34.9
85.1
72.0
13.9
101.6
72.1
72.6
70.3
72.5
166.8
174.3
34.0
27.4
132.7
124.2
29.6
84.8
72.1
14.0
101.2
71.9
72.6
70.3
72.4
166.7
174.4
34.1
32.6
133.5
125.6
35.9
72.3
86.4
13.9
100.9
72.2
72.8
70.4
72.5
166.9
174.4
34.0
32.5
134.3
125.3
36.3
72.4
83.7
14.0
100.4
71.9
72.5
70.2
72.4
163.8
174.3
34.0
32.4
134.2
124.0
34.2
85.6
72.0
14.0
101.5
72.2
72.8
70.3
72.4
166.8
174.3
34.0
31.4
72.0
85.1
34.9
125.1
134.5
14.1
100.8
71.9
72.4
70.3
72.4
166.7
174.3
34.1
32.4
72.6
86.3
29.0
123.5
132.7
14.1
100.5
71.7
72.6
70.3
72.3
166.7
174.3
34.0
25
72.5
85.9
34.2
124.3
134.6
13.9
100.8
72.0
72.5
70.1
72.4
166.8
174.8
34.0
32.4
72.4
84.2
35.9
125.1
133.9
13.9
3ꢀ
4ꢀ
5ꢀ
6ꢀ
1
2
8
9
10
11
12
13
18
a
All FBz 13C resonate between 130.3 and 116.6 ppm. All methyl protecting groups at 6ꢀ resonate at 53.0 0.2 ppm. All methyl protecting groups
at 1 resonate at 51.4 0.1 ppm.
added drop wise over a 30 min period. The resulting mix-
ture was refluxed for about 24 h. After completion of the
reaction (monitored by TLC with the aid of orcinol ferric
chloride spray reagent), it was cooled to room tempera-
ture. The resulting solids were filtered and washed with
toluene. The combined organic solvents were washed
with water and dried over anhydrous sodium sulphate.
The solvent was removed by rotary evaporation. The
Fig. 1. 1H NMR spectra of deprotected final products. 9b is 13-glucuronide substituted dihydroxy linoleic acid. 9d is 12-glucuronide substituted
dihydroxy linoleic acid. 9h is 10-glucuronide substituted dihydroxy linoleic acid. 9i is 9-glucuronide substituted dihydroxy linoleic acid.

80
J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
Table 2
1H NMR chemical shift assignment and coupling-constants of the 2,3,4-tri-O-(3-fluorobenzoyl)-d-glucuronides of 9,10- and 12,13-dihydroxy methyl
linoleates (in CDCl3)a
Assignment
12Ss13S or 12Rs3S
12Ss13R or 12Rs13R
12S13Ss
12R13Ss
8a
8b
8c
8d
1ꢀ
2ꢀ
5.00 (d, 1H, J = 7.8)
5.55 (dd, 1H, J = 9.8, 7.8)
5.89 (t, 1H, J = 9.5)
5.70 (t, 1H, J = 9.8)
4.35 (d, 1H, J = 9.8)
2.32 (t, 2H, J = 7.3)
1.98 (q, 2H, J = 6.6)
5.50 (m, 1H)
5.01 (d, 1H, J = 7.8)
5.55 (dd, 1H, J = 9.5, 7.8)
5.90 (dd, 1H, J = 9.8, 9.5)
5.70 (t, 1H, J = 9.8)
4.35 (d, 1H, J = 9.8)
2.33 (t, 2H, J = 7.3)
2.06 (q, 2H, J = 7.0)
5.46 (m, 1H)
4.91 (d, 1H, J = 7.8)
5.54 (dd, 1H, J = 9.5, 7.8)
5.88 (dd, 1H, J = 9.7, 9.4)
5.66 (dd, 1H, J = 9.7, 9.4)
4.35 (d, 1H, J = 9.8)
2.30 (t, 2H, J = 7)
1.99 (m, 1H)
5.04 (d, 1H, J = 7.6)
5.55 (dd, 1H, J = 9.8, 7.6)
5.88 (dd, 1H, J = 9.8, 9.5)
5.70 (dd, 1H, J = 9.8, 9.5)
4.34 (d, 1H, J = 9.8)
2.32 (t, 2H, J = 7.6)
1.92 (q, 2H, J = 6.2)
5.33 (m, 1H)
3ꢀ
4ꢀ
5ꢀ
2
8
9
5.48 (m, 1H)
10
11a
11b
12
13
18
5.47 (m, 1H)
2.51 (m, 1H)
2.41 (m, 1H)
3.54 (q, 1H, J = 5.4)
3.48 (bs, 1H)
5.43 (m, 1H)
2.59 (m, 1H)
2.47 (m, 1H)
3.56 (m, 1H)
3.46 (m, 1H)
0.81 (t, 3H, J = 7.6)
5.48 (m, 1H)
2.09 (m, 1H)
2.30 (m, 1H)
3.61 (m, 1H)
3.52 (m, 1H)
0.68 (t, 3H, J = 7)
5.18 (m, 1H)
2.00 (m, 1H)
2.12 (m, 1H)
3.49 (m, 1H)
3.63 (m, 1H)
0.90 (t, 3H, J = 6.9)
0.81 (t, 3H, J = 7.3)
12R13Rs
8e
9R10Rs or 9R10Ss
8f
9R10Ss or 9S10Rs
8g
9R10Ss or 9S10Rs
8h
9s
8i
1ꢀ
2ꢀ
3ꢀ
4ꢀ
5ꢀ
2
4.96 (d, 1H, J = 7.6)
5.55 (dd, 1H, J = 9.6, 7.8)
5.90 (t, 1H, J = 9.6)
5.70 (t, 1H, J = 9.5)
4.37 (d, 1H, J = 9.8)
2.33 (t, 2H, J = 7.4)
1.74 (m, 2H)
4.98 (d, 1H, J = 7.5)
5.56 (dd, H, J = 9.8, 9.5)
5.89 (dd, 1H, J = 9.8,9.5)
5.69 (t, 1H, J = 9.5)
4.35 (d, 1H, J = 10)
2.32 (m, 2H)
4.98 (d, 1H, J = 7.8)
5.55 (dd, 1H, J = 9.5, 7.8)
5.89 (dd, 1H, J = 9.8, 9.5)
5.70 (t, 1H, J = 9.5)
4.37 (d, 1H, J = 9.8)
2.32 (t, 2H, J = 7.4)
1.40 (m, 2H)
4.95 (d, 1H, J = 7.7)
5.56 (t, 1H, J = 9.4)
5.9 (t, 1H, J = 9.2)
5.70 (t, 1H, J = 9.4)
4.36 (d, 1H, J = 9.6)
2.32 (t, 2H, J = 7.5)
1.48 (m, 2H)
5.02 (d, 1H, J = 7.7)
5.54 (t, 1H, J = 9.8, 7.9)
5.88 (t, 1H, J = 9.6)
5.68 (t, 1H, J = 9.4)
4.32 (d, 1H, J = 9.6)
2.32 (t, 2H, J = 7.7)
1.67 (m, 2H)
8
1.40 (m, 2H)
9
5.31 (m, 1H)
3.48 (m, 1H)
3.57 (m, 1H)
3.57 (m, 1H)
3.65 (m, 1H)
10
5.21 (m, 1H)
3.54 (m, 1H)
3.58 (m, 1H)
3.57 (m, 1H)
3.48 (m, 1H)
11a
11b
12
2.13 (m, 1H)
2.28 (m, 1H)
3.57 (m, 1H)
2.44 (m, 1H)
2.48 (m, 1H)
5.46 (m, 1H)
2.16 (m, 1H)
2.24 (m, 1H)
5.21 (m, 1H)
2.19 (m, 1H)
2.19 (m, 1H)
5.23 (m, 1H)
2.07(m, 1H)
1.97(m, 1H)
5.16 (m, 1H)
13
3.57 (m, 1H)
5.49 (m, 1H)
5.18 (m, 1H)
5.30 (m, 1H)
5.34 (m, 1H)
18
0.90 (t, 3H, J = 7.4)
0.91 (m, 3H)
0.90 (m, 3H)
0.89 (m, 3H)
0.89 (m, 3H)
a
All protons of FBz protect group resonate between 7.78 and 7.14 ppm. All protons on carbon 3–7 and 14–17 resonate from 1.68 to 1.14 ppm.
All protons in methyl protecting groups at 1 resonate at 3.68 0.01 ppm. All protons in methyl protecting groups at 6ꢀ resonate at 3.73 0.02 ppm.
residue was purified by column chromatography using
hexane:ethyl acetate (4:1) as an eluent for the silica
gel column. This was followed by normal phase HPLC
using 1% isopropanol in hexane for raw purification
and 0.6% isopropanol in hexane for re-purification to
under reduced pressure and 1 mL saturated sodium chlo-
ride solution was added. The pH of the solution was
adjusted to 3–4 by adding formic acid, and then the prod-
uct was extracted by ethyl ether. After removing ethyl
ether by rotary evaporator, the residue was purified by
reverse phase HPLC with Zorbax RX-C8 column, fit-
ted with a UV detector (202 nm) and using a mobile
phase of 33% acetonitrile, 0.01% formic acid in water to
isolate a pure final product. Reverse phase HPLC condi-
tions and peaks of the corresponding compounds were
confirmed by simultaneous detection using a photodi-
ode array detector (PAD) and mass spectrometer. To
ensure that peaks were being simultaneously detected
by the PAD and mass spectrometer, a splitter was used
and the tube length and speed of elution were adjusted
to a suitable value using the by-product of the reaction
(3-fluorobenzoic acid) as a standard. 13C NMR and 1H
obtain 8 as pure compounds where possible (8a–i). 13
C
1
NMR and H NMR data are listed in Tables 1 and 2,
respectively. The labels of the atoms are given in
Fig. 1.
2.1.9. β-Glucuronides of linoleic acid diol (9)
Each of the nine purified 2,3,4-tri-O-(3-fluoro-
benzoyl)-␤-d-glucuronides of 9,10- or 12,13-dihydroxy
methyl linoleate (8a–i) (50 mg) were hydrolyzed indi-
vidually with ∼3% aq. NaOH in 2 mL methanol while
stirring at room temperature for 5–8 h. After comple-
tion, as monitored by TLC, the methanol was evaporated

J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
81
Table 3
13C MR chemical shift assignment of the ␤-glucuronides of 9,10- and 12,13-dihydroxy linoleic acid (in d-methanol and CDCl3)
Assignment
9a
9b
9c
9d
9e
9f
9g
9h
9i
1ꢀ
2ꢀ
105.1
74.8
105.3
75.2
104.7
75.1
103.0
75.2
105.5
75.4
102.0
74.9
101.6
74.0
105.3
75.7
Limited isolate no
carbon data
3ꢀ
4ꢀ
5ꢀ
6ꢀ
1
76.9
72.4
75.8
165.9
178
76.7
73.4
75.4
169.9
177.6
35.6
77.1
73.0
76.1
77.7
73.8
73.3
77.5
73.6
76.4
77.5
73.4
76.2
169.0
174.2
35.3
76.7
72.6
75.3
168.7
177.5
34
77.9
73.9
74.4
177
35.0
171.0
36.3
173.4
35.3
178.8
36.1
2
34.8
3–7, 14–17 23.0–33.2
23.6–34.0
28.3
132.4
126.4
30.6
84.5
73.6
14.5
23.2–33.6
33.4
126.2
132.7
35.0
74.4
84.3
14.6
23.6–33.7
33.7
127.4
134.6
36.4
76.3
82.8
14.6
23.6–34.0
33.7
73.2
84.8
36.8
127.4
134.2
14.6
23.6–32.8
28.4
82.6
73.2
29.3
126.6
133.4
14.6
22.2–33.0
29.5
71.8
81.2
34.0
126.0
133.5
13.7
23.8–33.9
33.9
84.0
74.4
37.5
127.8
134.3
14.5
8
9
33.2
133.7
126.4
36.1
84.2
72.7
14.3
10
11
12
13
18
NMR data of 9 are listed in Tables 3 and 4, respectively.
The labels of the atoms are given in Fig. 1.
are theoretically obtained: two positional isomers (9,10-
and 12,13-diol) and four stereoisomers (RR/SS, RS/SR)
for each positional isomer. The RR and SS, as well as
the RS and SR are enantiomers that theoretically can-
not be separated without chiral column HPLC. However,
since there are additional chiral centers added, once the
glucuronide group is attached, some of these may be
separated by HPLC as the diastereomers (RR/SS with
RS/SR). This is consistent with our observation of peak
tailing when the 9,10- and 12,13-diols were analyzed by
normal phase HPLC (3% isopropanol in hexane). There-
fore, following the reaction of the 9,10- and 12,13-diol
with bromo sugar (4), theoretically 16 possible isomers
will be obtained for both the ␤ and ␣ glucuronide.
The final products we obtained agree with this pre-
diction since nine pure isomers were isolated as final
products.
The analysis of synthesized pure products or
mixtures by HPLC/MS was performed using a
3.5 ␮m Eclipse XDB-C18 (Zorbax) analytical column
(2.1 mm × 150 mm) with the flow rate of 0.1 mL/min.
The isocratic mobile phase consisted of 25% B (100%
acetonitrile) in A (0.01 mM ammonium acetate, pH 6).
For each injection, a 200 ␮L sample was loaded onto a
pre-column (10 ␮m Zorbax XDB-C8) with mobile phase
A, then washed at flow rate of 1.5 mL/min for 1 min.
The trap was then back flashed onto the analytical col-
umn using 25% B in order to introduce the sample onto
the Q-TOF mass spectrometer for analysis as described
above. Spectra were acquired at 3 s per scan using neg-
ative ion mode.
3. Results and discussion
3.2. Characterization of glucuronide derivatives of
dihydroxy methyl linoleate (8) and glucuronide
derivatives of dihydroxy linoleic acid (9)
3.1. Theoretical consideration of structural isomers
and stereoisomers
During the chemical synthesis described above
(Scheme 2), one of the cis double bonds in methyl
linoleate is epoxidized using aqueous hydrogen perox-
ide and catalytic methyl trioxorhenium/pyridine. This
epoxidation yields four different monoepoxide products
with 9R10S, 9S10R, 12S13R and 12R13S configura-
tions. The positional isomers were separated by HPLC.
When these monoepoxides are subsequently hydrolyzed
to their corresponding diols by perchloric acid, the stere-
oselectivity is lost and all the 4 × 2 possible isomers
The structure and relative stereochemistry of these
products were determined by extensive analysis of NMR
data (including 1H, 13C, gHMQC, gHMBC, gCOSY and
NOESY for each sample) as well as by molecular mod-
eling, except for compounds 8e, 8f, 8g, 8h and 8i where
COSYandNOESYdatawereusedfortheseassignments
and compared to minimized models. Absolute assign-
ments of stereochemistry are dependent on an X-ray
crystal structure, which is not possible for these com-
pounds. Circular dichroism would be useful, but beyond

82
J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
Table 4
1H NMR chemical shift assignment and coupling-constants of the ␤-glucuronides of 9,10- and 12,13-dihydroxy linoleic acid (in d-methanol and
CDCl3)
Assignment
12Ss13S and 12Rs13S
12Ss13R and 12Rs13R
12Ss13S
12R13Ss
9a
9b
9c
9d
1ꢀ
4.40 (d, 1H, J = 7.6)
3.26 (m, 1H)
3.41 (t, 1H, J = 9.0)
3.52 (m, 1H)
3.65 (d, 1H, J = 9.6)
7.79
4.40 (d, 1H, J = 7.6)
3.30 (m, 1H)
3.40 (t, 1H, J = 9.1)
3.46 (t, 1H, J = 9.1)
3.55 (d, 1H, J = 9.8)
8.52
4.30
4.42 (d, 1H, J = 7.6)
3.31 (m, 1H)
3.43 (t, 1H, J = 9.1)
3.51 (t, 1H, J = 7.3)
3.62 (m, 1H)
2ꢀ
3.18 (m, 1H)
3.41 (m, 1H)
3.45 (m, 1H)
3.52 (m, 1H)
8.51
3ꢀ
4ꢀ
5ꢀ
OH at 6ꢀ
8.41 (s, 1H)
OH at 1
4.4
2
2.26 (t, 2H, J = 7.2)
1.58, 1.48, 1.30
1.98 (m, 2H)
5.47 (m, 1H)
5.47 (m, 1H)
2.44 (m, 1H)
2.32 (m, 1H)
3.52 (m, 1H)
3.52 (m, 1H)
2.24 (t, 2H, J = 7.2)
1.58, 1.48, 1.30, . . .
2.03 (m, 2H)
5.42 (m, 1H)
5.50 (m, 1H)
2.40 (m, 1H)
2.52 (m, 1H)
3.57 (p, 1H, J = 5.0)
3.51 (p, 1H, J = 5.0)
0.88 (t, 3H, J = 7.3)
2.00 (m, 2H)
1.58, 1.44, 1.29, 1.24
2.00 (m, 1H)
5.48 (m, 1H)
5.48 (m, 1H)
2.25 (m, 2H)
2.25 (m, 2H)
3.52 (m, 1H)
3.52 (m, 1H)
0.84 (t, 3H)
2.28 (t, 2H, J = 7.9)
1.61, 1.43, 1.30
2.06 (m, 2H)
5.48 (m, 1H)
5.48 (m, 1H)
2.34 (m, 1H)
2.20 (m, 1H)
3.58 (dd, 1H, J = 8.5, 4.7)
3.58 (dd, 1H, J = 8.5, 4.7)
0.89 (t, 3H, J = 7.9)
3–7, 14–17
8
9
10
11a
11b
12
13
18
0.88 (t, 3H, J = 7.3)
12R13Rs
9R10Rs or 9R10Ss
9Ss10R
9e
9f
9g
9h
9i
1ꢀ
4.39 (d, 1H, J = 7.6)
3.24 (t, 1H, J = 8.5)
3.41 (t, 1H, J = 8.8)
3.45 (t, 1H, J = 8.5)
3.56 (d, 1H, J = 8.5)
8.44 and 4.59
4.40 (d, 1H, J = 7.9)
3.25 (m, 1H)
3.41 (t, 1H, J = 8.5)
3.48 (t, 1H, J = 9.8)
3.60 (d, 1H, J = 9.4)
8.44 and 4.52
4.39 (d, 1H, J = 7.2)
3.25 (t, 1H, J = 8.5)
3.41 (t, 1H, J = 9.1)
3.46 (t, 1H, J = 9.5)
3.56 (m, 1H)
4.42 (d, 1H, J = 7.3)
3.27 (t, 1H, J = 8.1)
3.43 (m, 1H)
3.48 (m, 1H)
3.59 (m, 1H)
4.41 (d, 1H, J = 7.9)
3.28 (t, 1H, J = 9.0)
3.41 (t, 1H, J = 9.0)
3.48 (t, 1H, J = 9.0)
3.56 (d, 1H, J = 9.0)
8.53
2ꢀ
3ꢀ
4ꢀ
5ꢀ
COOH at 6ꢀ and 1
8.44
8.53
2
2.24 (m, 2H)
2.27 (t, 2H, J = 6.6)
1.60, 1.50, 1.28
1.98 (dd, 2H, J = 9.8, 5.7)
3.54 (m, 1H)
3.54 (m, 1H)
2.47 (m, 1H)
2.37 (m, 1H)
5.49 (m, 1H)
5.48 (m, 1H)
0.89 (t, 3H, J = 6.3)
2.26 (t, 2H, J = 7.2)
1.60, 1.48, 1.32
2.06 (q, 2H, J = 7.1)
3.67 (m, 1H)
3.56 (m, 1H)
2.41 (m, 1H)
2.33 (m, 1H)
5.53 (m, 1H)
5.46 (m, 1H)
0.90 (m, 3H)
2.29 (t, 2H, J = 7.3)
1.62, 1.36
2.04 (q, 2H, J = 6.5)
3.69 (q, 1H, J = 6.5)
3.59 (m, 1H)
2.40 (m, 1H)
2.32 (m, 1H)
5.55 (m, 1H)
5.54 (m, 1H)
2.28 (t, 2H, J = 8.1)
1.62, 1.52, 1.35
2.04 (q, 2H, J = 6.5)
3.60 (m, 1H)
3.60 (m, 1H)
2.34 (m, 1H)
2.28 (m, 1H)
5.52 (m, 1H)
5.52 (m, 1H)
0.93 (t, 3H, J = 6.5)
3–7, 14–17
8
9
10
11a
11b
12
13
18
1.60, 1.39, 1.30
2.02 (q, 2H, J = 6.6)
5.51 (m, 1H)
5.53 (m, 1H)
2.38 (m, 1H)
2.25 (m, 1H)
3.67 (m, 1H)
3.56 (d, 1H, J = 8.5)
0.91 (t, 3H, J = 6.3)
0.92 (t, 3H, J = 7.3)
the scope of this work. Molecular mechanics optimiza-
tions were carried out to obtain low energy conformation
structures which further helped to identify isomers.
gesting that all the glucuronide substituted products are
␤. A small amount of inversion of the stereochemistry at
the anomeric carbon to give the ␣ anomer was seen, but
these products were not studied.
3.2.1. α-Glucuronide or β-glucuronide
The ␣- and ␤-glucuronide can usually be distin-
guished by the chemical shift and especially by the
coupling constant of proton H-1 on the sugar portion of
the molecule (Smith and Benet, 1986). The H-1 is a dou-
blet with J1–2 around 8 Hz for ␤-glucuronide while H-1
is a doublet with J1–2 around 4 Hz for the ␣-glucuronide.
For all products within the 8 and 9 series, the 1H NMR
spectra shows the H-1 proton as a doublet, therefore sug-
3.3. Positional isomer assignment and
stereochemistry for β-glucuronides of methyl
linoleate (8)
The analysis of stereochemistry in these epoxidized
methyl linoleate glucuronide metabolites is complicated
greatly by molecular flexibility. Molecular modeling
(Schrodinger’s Maestro) was used to build and mini-

J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
83
mize each enantiomer of the 9, 10, 12 and 13 substituted
derivatives. This gave the approximate low energy struc-
ture for each of the 16 possible derivatives. Four diastere-
omers of 9,10-dihydroxylinoleic acid were isolated. One
of these (8f) is either 9R10R substituted or 9R10S sub-
stituted. The minimized structure for these two show
the H9 orthogonal to H11 and H11ꢀ and therefore no
NOE should be observed between these protons, which
is supported by the NMR data. This gave us the ability
to establish stereochemistry for the free hydroxy carbon
of the linoleic derivative, however, molecular rotation
leading to transient NOEs reduced the chances of estab-
lishing stereochemistry for the glucuronide substituted
carbon. A strong NOE is observed from the anomeric
proton of the glucuronide to the proton of the substi-
tuted carbon in each isolated fraction. From modeling,
all S-substituted glucuronide linoleates have a mini-
mized structure with a small dihedral angle between the
anomeric proton and the hydrogen on the substituted car-
bon. This should lead to a large NOE between these two
protons. All R-substituted glucuronide linoleates have a
minimized structure with a large dihedral angle between
the anomeric proton and the hydrogen on the substituted
carbon. Rotation about this bond could lead to an NOE.
Rotation of Canomeric–Oanomeric–Clinoleic bond, leading to
a NOE between Hanomeric and H9, may be stabilized by
a hydrogen bond to the methoxyester of the glucuronide
or by the steric hindrance caused by the 3-fluorophenoxy
protecting groups on the glucuronide.
both modeling and NMR data better, but conformational
flexibility again can give transient NOE. However, a
largecoupling constant wouldbeexpected forH12–H13,
which is not present in the NMR data. Potential hydro-
gen bonding was not accounted for in the modeling data
and might help to determine the correct stereochemistry.
8b is either 12Ssub13R or 12Rsub13R. The 13R con-
former shows a potential NOE from H13 to H11 in the
modeling data. A small NOE from H13 (3.48 ppm) to
H11 (2.59 ppm) was observed in the NMR data. It is not
possible, with the current data, to determine the stereo-
chemistry at C12.
8d and 8c are diastereomers substituted at C-13 of
the linoleate. The assignment of 8d is likely 12R13Ssub.
There are NOEs from H13 to H11 and H11ꢀ. H12 also
has NOEs to H11 and H11ꢀ as well as to H10. There is
a probable hydrogen bond between the 13OH and the 2-
OPh3F. 8c is likely to be 12S13Ssub. H12 and H13 have
NOEs to H10. H13 does not have any NOE correlations
to H11 or H11ꢀ.
8e is a third diastereomer substituted at C13 of the
linoleate. This appears to be 12R13Rsub with weak
NOEs from H11 and H11ꢀ to the anomeric proton.
H13Has an NOE to H10. We believe we are seeing three
diastereomers due to restricted rotation. Rotation around
the anomeric oxygen is probably restricted by hydrogen
bonding. The steric bulk of the FBzO protecting groups
may also have an impact on an unexpected number of
diastereomers.
Two other compounds (8g and 8h), where the glu-
curonide is substituted onto carbon-10 were isolated.
Both of these show a high degree of proton overlap at
3.57 ppm of the H-9 and H-10. Neither J-couplings nor
NOEs can be used to identify these compounds unequiv-
ocally. There are differences in the shifts and couplings
of the H11, H11ꢀ and H8 protons. These are probably
9R10S substituted and 9S10R substituted. Again, hydro-
gen bonding may explain why these two are separable
by chromatography. Due to overlap, one of these com-
pounds may also be the second 9-substituted diastere-
omer.
8i is a 9-substituted diastereomer. Discerning the
stereochemistry, however, is problematic. Molecular
modeling showed NOE correlations from H10 to H11,
H11ꢀ and H12. These are expected due to the flexibil-
ity of the molecule. Each diastereomer has an energy
minimum with this possibility (Fig. 2).
3.4. HPLC-MS/MS investigation of synthesized
compounds
Electrospray ionization was used in the HPLC-
MS/MS analysis because of its compatibility with on-
line HPLC separation and its sensitivity and specificity
(Murphy, 1995). Mass spectrometer parameters were
optimized by direct injection of standard compounds.
The cone voltage was optimized to obtain the maxi-
mum signal for the pseudomolecular ion (M–H)− 489
for each compound. This was determined to be 70 V
for all isomers. The collision energy was optimized to
obtain unique fragment ions for each positional isomer.
Fig. 3 shows the collision induced (CID) mass spectra of
the positional isomers. In A, the cone voltage was 70 V
and collision energy was 25 eV, and the mass spectra of
these isomers are very similar. Fragmentation occurred
at the bond between the glucuronide and the diol oxy-
gen to give the m/z = 313 (linoleic acid fragment) and
m/z = 175 (glucuronide fragment) ions. When the cone
voltage was kept constant and the collision energy was
increased to 35 eV, isomer-specific fragmentation ions
8a and 8b are diastereomers, which are substi-
tuted at C-12 of the linoleate. 8a has two possibilities:
12Ssub13S or 12Rsub13S. H13Has medium NOE cor-
relations to H11 and H11ꢀ. From modeling data this
happens only in the 13S conformers. 12Ssub13S fits

84
J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
Fig. 2. NOESY data from 9i, 9h, 9d and 9b. The crosspeaks from the olefinic region to those peaks around 2 ppm and from 3.5 to 4.4 ppm and
2 ppm are used to determine stereochemistry.
were observed between the 9- or 10-glucuronides and
the 12- or 13-glucuronides (B–E in Fig. 3). Fig. 4 illus-
trates the proposed mechanism for the formation of the
m/z 313 ion in the CID mass spectrum of 9-glucuronide-
10-hydroxy-12-linoleic acid. A similar cleavage of the
glucuronide ether bond in the CID ionization of 10-
glucuronide-9-hydroxy-12-linoleic acid results in the
generation of the same ion structure. This fragmentation
Fig. 3. A typical mass spectrum of glucuronide positional isomers. (A) All isomers, the cone voltage is 70 V and the collision energy is 25 V, (B)
the 9-glucuronide, (C) the 10-glucuronide, (D) the 12-glucuronide and (E) the 13-glucuronide. The cone voltage and collision energy for B–E are
70 and 35 V, respectively.

J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
85
Fig. 4. Proposed fragmentation mechanism for the formation of the m/z 313 ion in the CID spectrum of 9-glucuronide-10-hydroxy-12-linoleic acid
and 12-glucuronide-13-hydroxy-9-linoleic acid.
mechanism in the CID ionization of 12-glucuronide-13-
hydroxy-9-linoleic acid (Fig. 4) and 13-glucuronide-12-
hydroxy-9-linoleic acid leads to a fragment ion of the
same mass but different structure. Subsequent fragmen-
tation of the respective m/z 313 ions (Figs. 5 and 6) thus
lead to isomer-specific ions that could be used to distin-
guishthe9/10glucuronidesfromthe12/13glucuronides.
Previous work has shown similar patterns of fragmenta-
tion for these types of fatty acids (Wheelan et al., 1996).
Preliminary analysis (data not shown) showed that
these positional isomer-specific fragment ions could
be used to identify isomers with close retention
times such as the 13-glucuronide 9e (retention time
11.35 min) and the 10-glucuronide 9g (retention time
11.21 min).
In conclusion, we have developed a general strategy
for the synthesis and purification of the glucuronides of
linoleic acid diols along with analytical techniques use-
ful for isomer-specific analysis. This general synthetic
scheme should also prove useful for the synthesis and
analysis of these and other fatty acid diols in biological
samples.
Fig. 5. Proposed fragmentation mechanism for the formation of the m/z 201 and m/z 171 ion from the m/z 313 ion fragment in the CID spectrum
of 9-glucuronide-10-hydroxy-12-linoleic acid and 10-glucuronide-9-hydroxy-12-linoleic acid. (Charge-remote allylic fragmentation as described
Wheelan et al., 1996.)

86
J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
Fig. 6. Proposed fragmentation mechanism for the formation of the m/z 201 and m/z 171 ion from the m/z 313 ion fragment in the CID spectrum
of 12-glucuronide-13-hydroxy-9-linoleic acid and 13-glucuronide-12-hydroxy-9-linoleic acid. (Charge-remote allylic fragmentation as described
Wheelan et al., 1996.)
Acknowledgments
Murphy, R.C., 1995. Lipid mediators, leukotrienes and mass-
spectrometry. J. Mass Spectrom. 30, 5–16.
Radominska, A., Little, J.M., Lehman, P.A., Samokyszyn, V., Rios,
G.R., King, C.D., Green, M.D., Tephly, T.R., 1997. Glucuronida-
tion of retinoids by rat recombinant UDP: glucuronosyltransferase
1.1 (bilirubin UGT). Drug Metab. Dispos. 25, 889–892.
Ramu, K., Baker, J.K., 1995. Synthesis, characterization, and anti-
malarial activity of the glucuronides of the hydroxylated metabo-
lites of arteether. J. Med. Chem. 38, 1911–1921.
Rukhman, I., Yudovich, L., Nisnevich, G., Gutman, A.L., 2001. Selec-
tive synthesis of both isomers of morphine 6-beta-d-glucuronide
and their analogs. Tetrahedron 57, 1083–1092.
Sevanian, A., Mead, J.F., Stein, R.A., 1979. Epoxides as products of
lipid autoxidation in rat lungs. Lipids 14, 634–643.
Smith, P.C., Benet, L.Z., 1986. Characterization of the isomeric esters
of zomepirac glucuronide by proton NMR. Drug Metab. Dispos.
14, 503–505.
This study was supported by the National Institutes
of Health (NIGMS 56708; NIEHS 011630) and by the
Department of Defense (N00014-99-1-0905; N00014-
99-1-06006).
References
Bellucci, G., Chiappe, C., Conti, L., Marioni, F., Pirini, G., 1989. Sub-
strate enantioselection in the microsomal epoxide hydrolase cat-
alyzed hydrolysis of monosubstituted oxiranes. Effects of branch-
ing of alkyl chains. J. Org. Chem. 54, 5978–5983.
Costa, C.C., Dorland, L., Kroon, M., Tavares de Almeida, I., Jakobs, C.,
Duran, M., 1996. 3-, 6- and 7-hydroxyoctanoic acids are metabo-
lites of medium-chain triglycerides and excreted in urine as glu-
curonides. J. Mass Spectrom. 31, 633–638.
Street, J.M., Evans, J.E., Natowicz, M.R., 1996a. Glucuronic acid-
conjugated dihydroxy fatty acids in the urine of patients with
generalized peroxisomal disorders. J. Biol. Chem. 271, 3507–
3516.
Dutton, G.J., 1980. Glucuronidation of Drugs and other Compounds.
CRC Press, Boca Raton, FL.
Fehlhaber, H.W., Snatzke, G., Vlahov, I., 1987. Synthese von 1.a-
Hydroperoxy-Zuckerderivaten. Liebigs Ann. Chem., 637–638.
Jude, A.R., Little, J.M., Freeman, J.P., Evans, J.E., Radominska-
Pandya, A., Grant, D.F., 2000. Linoleic acid diols are novel sub-
strates for human UDP-glucuronosyltransferases. Arch. Biochem.
Biophys. 380, 294–302.
Kirkpatrick, R.B., Falany, C.N., Tephly, T.R., 1984. Glucuronidation of
bile acids by rat liver 3-OH androgen UDP-glucuronyltransferase.
J. Biol. Chem. 259, 6176–6180.
Laethem, R.M., Laethem, C.L., Koop, D.R., 1992. Purification and
properties of a cytochrome P450 arachidonic acid epoxygenase
from rabbit renal cortex. J. Biol. Chem. 267, 5552–5559.
Moghaddam, M.F., Motoba, K., Borhan, B., Pinot, F., Hammock, B.D.,
1996. Novel metabolic pathways for linoleic and arachidonic acid
metabolism. Biochem. Biophys. Acta 1290, 327–339.
Street, J.M., Evans, J.E., Natowicz, M.R., Doroshevich, P., Ghosh,
A., 1996b. Identification of glucuronide-conjugated hydroxylated
fatty acids in the urine of children with generalized peroxisomal
disorder. Ann. N. Y. Acad. Sci. 804, 768–769.
Takagaki, H., Nakanishi, S., Kimura, N., Yamaguchi, S., Aoki, Y.,
Quinoline, G., 1999. Production process and anti-allergic agent,
Eur. Pat. Appl. EP 933378.
Turgeon, D., Chouinard, S., Belanger, P., Picard, S., Labbe,
J.F., Borgeat, P., Belanger, A., 2003. Glucuronidation of
arachidonic and linoleic acid metabolites by human UDP-
glucuronosyltransferases. J. Lipid Res. 44, 1182–1191.
Watabe, T., Akamatsu, K., 1972. Stereoselective hydrolysis of acyclic
olefin oxide to glycols by hepatic microsomal epoxide hydrolase.
Biochim. Biophys. Acta 279, 297–305.

J. Yang et al. / Chemistry and Physics of Lipids 140 (2006) 75–87
87
Watzer, B., Reinalter, S., Seyberth, H.W., Schweer, H., 2000. Determi-
nation of free and glucuronide conjugated 20-hydroxyarachidonic
acid (20-HETE) in urine by gas chromatography/negative ion
chemical ionization mass spectrometry. Prostaglandins Leukot.
Essent. Fatty Acids 62, 175–181.
Westkaemper, R.B., Hanzlik, R.P., 1981. Mechanistic studies of epox-
idehydrolaseutilizingacontinuousspectophotometricassay. Arch.
Biochem. Biophys. 208, 195–204.
Wheelan, P., Zirrolli, J.A., Murphy, R.C., 1996. Electrospray ioniza-
tion and low energy tandem mass spectrometry of polyhydroxy
unsaturated fatty acids. J. Am. Soc. Mass Spectrom. 7, 140–149.
Wistuba, D., Nowotny, H.P., Trager, O., Schurig, V., 1989. Cytochrome
P-450-catalyzed asymmetric epoxidation of simple prochiral and
chiral aliphatic alkenes: species dependence and effect of enzyme
induction on enantioselective oxirane formation. Chirality 1,
127–136.