9-Hydroxy-10,12-octadecadienoic acid (9-HODE) and 13-hydroxy-9,11-octadecadienoic acid (13-HODE): Excellent markers for lipid peroxidation

Article abstract of DOI:10.1016/S0009-3084(97)00070-4

Various conditions for conversion of (9S,10E,12Z)-9-hydroperoxy-10,12-octadecadienoic acid (9S-HPODE) and (13S,9Z,10E)-13-hydroperoxy-9,11-octadecadienoic acid (13S-HPODE) into the corresponding hydroxy acids, (9S,10E,12Z)-9-hydroxy-10,12-octadecadienoic acid (9S-HODE) and (13S,9Z,10E)-13-hydroxy-9,11-octadecadienoic acid (13S-HODE), were investigated in vitro. 9S-HODE and 13S-HODE were subjected to lipid peroxidation under various conditions: oxidation was carried out in air only, and in air/Fe²⁺/ascorbate, air/H₂O₂/Fe²⁺, air/Fe²⁺, and air/Fe³⁺. In contrast to the corresponding hydroperoxides (9S-HPODE and 13S-HPODE), 9-HODE and 13-HODE proved to be stable in all these oxidation experiments. Unexpectedly, hydroxy compounds obtained by reduction of hydroperoxides derived from arachidonic acid were not attacked by air/Fe²⁺/ascorbate or air/Fe²⁺. Thus, for instance, (15S,5Z,8Z,11Z,13E)-15-hydroxy-5,8,11,13-eicosatetraenoic acid (15-HETE) remained unchanged in spite of possessing the structural prerequisites for attack by radicals, i.e. a CH₂-group located between two double bonds. Consequently, metal-induced air oxidation reactions of these systems seem to be restricted to hydroperoxides of unsaturated acids (LOOH) and not to corresponding hydroxy compounds (LOH). The reported experiments explain why hydroxy derivatives of unsaturated acids, especially 9-HODE and 13-HODE, are enriched in naturally occurring lipid peroxidation (LPO) processes to a greater extent than any other LPO product and why they are nearly ideal markers for LPO.

Full text of DOI:10.1016/S0009-3084(97)00070-4

Chemistry and Physics of Lipids
89 (1997) 131–139
9-Hydroxy-10,12-octadecadienoic acid (9-HODE) and
13-hydroxy-9,11-octadecadienoic acid (13-HODE): excellent
markers for lipid peroxidation
Peter Spiteller, Gerhard Spiteller *
Lehrstuhl Organische Chemie I, Uni6ersita¨t Bayreuth, Uni6ersita¨tsstraße 30, 95440 Bayreuth, Germany
Received 30 April 1997; received in revised form 11 August 1997; accepted 20 August 1997
Abstract
Various conditions for conversion of (9S,10E,12Z)-9-hydroperoxy-10,12-octadecadienoic acid (9S-HPODE) and
(13S,9Z,10E)-13-hydroperoxy-9,11-octadecadienoic acid (13S-HPODE) into the corresponding hydroxy acids,
(9S,10E,12Z)-9-hydroxy-10,12-octadecadienoic acid (9S-HODE) and (13S,9Z,10E)-13-hydroxy-9,11-octadecadienoic
acid (13S-HODE), were investigated in vitro. 9S-HODE and 13S-HODE were subjected to lipid peroxidation under
various conditions: oxidation was carried out in air only, and in air/Fe2+/ascorbate, air/H2O2/Fe2+, air/Fe2+, and
air/Fe3+. In contrast to the corresponding hydroperoxides (9S-HPODE and 13S-HPODE), 9-HODE and 13-HODE
proved to be stable in all these oxidation experiments. Unexpectedly, hydroxy compounds obtained by reduction of
hydroperoxides derived from arachidonic acid were not attacked by air/Fe2+/ascorbate or air/Fe2+. Thus, for
instance, (15S,5Z,8Z,11Z,13E)-15-hydroxy-5,8,11,13-eicosatetraenoic acid (15-HETE) remained unchanged in spite of
possessing the structural prerequisites for attack by radicals, i.e. a CH2-group located between two double bonds.
Consequently, metal-induced air oxidation reactions of these systems seem to be restricted to hydroperoxides of
unsaturated acids (LOOH) and not to corresponding hydroxy compounds (LOH). The reported experiments explain
why hydroxy derivatives of unsaturated acids, especially 9-HODE and 13-HODE, are enriched in naturally occurring
lipid peroxidation (LPO) processes to a greater extent than any other LPO product and why they are nearly ideal
markers for LPO. © 1997 Elsevier Science Ireland Ltd.
Keywords: Lipid peroxidation; Linoleic acid; 9-HODE; 13-HODE; 9-HPODE; 13-HPODE
1. Introduction
The generation of lipid hydroperoxides
(LOOH) of unsaturated fatty acids is considered
to be involved in the development of numerous
* Corresponding author. Tel.: +49 921 552680; fax: +49
921 552671; e-mail:gerhard.spiteller@uni-bayreuth.de
0009-3084/97/$17.00 © 1997 Elsevier Science Ireland Ltd. All rights reserved.
PII S0009-3084(97)00070-4

132
P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
chronic diseases, such as atherosclerosis (Morel et
al., 1983, Steinbrecher et al., 1984), rheumatoid
arthritis (Wingard et al., 1993), diabetes (Nishi-
gaki et al., 1981), multiple sclerosis (Mickel, 1978)
and other nervous diseases (Braughler and Hall,
1989).
9,11-octadecadienoic acid (13-HODE)—were
detected in large amounts in atherosclerotic
plaques (Brooks et al., 1970, Harland et al., 1973,
Belkner et al., 1991, Ku¨hn et al., 1992) or the
arteries and blood of rats after feeding a diet rich
in cholesterol (Wang and Powell, 1991). [Note:
the expressions 9-HODE and 13-HODE are used
if the stereochemistry and E/Z configuration are
unknown. This is the case for products generated
in non-enzymic LPO processes.]
Recently we detected 9-HODE and 13-HODE
in unusually high amounts in heart tissue after
myocardial infarction (Dudda et al., 1996b), in
LDL of elderly people (Jira and Spiteller, 1996)
and in LDL of patients suffering from atheroscle-
rosis or rheumatic arthritis (Jira et al., 1997).
Thus, oxidation of linoleic acid seems to be a
common process in mammalians after tissue in-
jury, causing cell death.
LOOH is not stable. For example, it is cleaved
in the presence of Fe2+ in a Fenton-like reaction
’
to alkoxy (LO ) radicals (Gardner, 1989):
’
LOOH+Fe2+ “LO +Fe3+ +OH−
Otherwise, Fe3+ may produce lipid peroxyl
’
radicals (LOO ) (Esterbauer et al., 1989):
’
LOOH+Fe3+ “LOO +Fe2+ +H+
’
LO radicals undergo numerous reactions (Mar-
nett and Wilcox, 1995): they abstract hydrogens
from other molecules, especially from CH2-groups
activated by two adjacent double bonds to form
the corresponding hydroxy acids (LOH) (Frankel,
1982, Esterbauer, 1982); they are decomposed by
cleavage of adjacent C–C bonds by generation of
aldehydes (Frankel, 1985, Esterbauer et al., 1991);
they react with adjacent double bonds by forming
epoxy radicals which are transformed to epoxy
hydroxy acids (Hamberg, 1973, Gardner and
Jursinic, 1981); and they undergo a large number
of secondary oxidation reactions. Only a few of
the resulting degradation products are known: e.g.
malondialdehyde, 4-hydroxynonenal (Esterbauer
et al., 1991) and h-hydroxyaldehydes (Loidl-
Stahlhofen and Spiteller, 1994, Loidl-Stahlhofen
et al., 1994).
This pointed to the possibility that hydroxy
fatty acids (1) such as 9-HODE and 13-HODE
are stable to further oxidation. A priori this was
not expected since they possess the same conju-
gated system as the corresponding hydroperoxy
fatty acids (2), namely 9-hydroperoxy-10,12-oc-
tadecadienoic acid (9-HPODE) and 13-hydroper-
oxy-9,11-octadecadienoic
acid
(13-HPODE),
respectively. These easily undergo further oxida-
tive transformation reactions obviously caused by
hydrogen abstraction from C–H bonds activated
by the conjugated system (Scheme 1).
In this paper we report on oxidation experi-
ments which demonstrate that 9-HODE and 13-
HODE are in fact stable against oxidation under
usual LPO conditions. In addition, we demon-
strate that (15S,5Z,8Z,11Z,13E)-15-hydroxy-
5,8,11,13-eicosatetraenoic acid (15S-HETE) is
Research on the occurrence and physiological
activity of lipid peroxidation (LPO) products in
mammalians has previously been almost exclu-
sively restricted to those derived from arachidonic
acid, in spite of the fact that linoleic acid is much
more abundant than arachidonic acid (e.g. the
ratio of linoleic acid/arachidonic acid in low-den-
sity lipoprotein (LDL) was found to be about 7:1)
’
(Esterbauer et al., 1992) and that LO radicals
seem to be nearly equally able to abstract hydro-
gen atoms from activated CH2-groups of linoleic
acid as from those of arachidonic acid. Thus, for
instance, products obtained by reduction of
linoleic acid hydroperoxides—(9-hydroxy-10,12-
octadecadienoic acid (9-HODE) and 13-hydroxy-
Scheme 1. The conjugated systems of HODE (R=H) and
HPODE (R=OH) are the same. Hydrogens on carbons adja-
cent to the conjugated system are expected to be equally prone
to abstraction by radicals.

P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
133
also stable against oxidation, in contrast to its
hydroperoxide.
saturated straight chain alkanes (C10–C36) (Ko-
vats, 1958).
2.2. Thin-layer chromatography
2. Materials and methods
Preparative thin-layer chromatography was car-
ried out on glass plates 20 cm×20 cm covered
with a 0.75 mm layer of silica gel 60 PF254. The
plates were prechromatographed with methanol;
the 2 cm zone at the upper border containing
impurities was then discarded. Before use the
plates were activated for 2 h at 120°C.
2.1. Materials
N-Methyl-N-trimethylsilyltrifluoracetamide
(MSTFA) was obtained from Macherey and
Nagel (Du¨ren, Germany). Soybean lipoxygenase
(5.3 U), Fe2SO4’7H2O, linoleic acid, arachidonic
acid and sodium ascorbate were purchased from
Fluka (Neu Ulm, Germany). HCl and
FeCl3’6H2O were obtained from Merck (Darm-
stadt). Solvents were distilled before use.
2.3. High-performance liquid chromatography
(HPLC)
Gas chromatography (GC) was carried out
with a Carlo Erba HRGC 5160 Mega Series gas
chromatograph equipped with a flame ionization
detector using a DB-1 fused silica gel capillary
column (30 m×0.32 mm i.d.) (J&W Scientific,
Germany), covered with a 0.1 vm layer of liquid
phase. The temperature of the detector was kept
at 290°C, the injector temperature was 280°C.
Injector volume was 0.2–0.7 vl of a 1% (m/v)
solution. Temperature programme: 3 min isother-
mal at 80°C, 3°C/min from 80°C to 280°C, then
15 min isothermal at 280°C. The carrier gas was
hydrogen and the split ratio was 1:30. Peak inte-
gration and data recording were done with a
Merck Hitachi Chromatointegrator D-2500.
Gas chromatography/mass spectrometry (GC/
MS) was performed on a Finnigan MAT 95 dou-
ble focusing mass spectrometer with inverse
Nier-Johnson geometry, equipped with an EI ion
source operated at 70 eV. A Hewlett Packard
5890 Series II gas chromatograph with a fused
silica DB-05 glass capillary column (30 m×0.25
mm i.d., covered with a 0.1 vm layer of liquid
phase; J&W Scientific, Germany) was used for
sample separation. The injector temperature was
kept at 280°C; injection volumes were 0.6–1.5 vl
of a 1–2% (m/v) solution. Temperature pro-
gramme: 3 min isothermal at 50°C, increased
within 2 min to 100°C, then 3°C/min until 300°C,
finally 10 min isothermal at 300°C.
2.3.1. Analytical normal phase HPLC
(NP-HPLC)
The HPLC system comprised a Pumpsystem
Beckman System Gold with programmable sol-
vent module 126 and a Beckman System Gold
diode array detector 168, with a LKB Bromma
2210 channel recorder. The column was a Bischoff
Nucleosil 3 vm (25 cm×4.6 mm i.d.) column,
with a Bischoff Spherisorb 5 vm Si/NP (2 cm×4
mm) precolumn.
2.3.2. Preparati6e NP-HPLC
This was carried out with a Beckman System
Gold with programmable solvent module 125,
programmable detector module 166, a Kipp &
Zonen BD 41 recorder, a Bischoff Ultrasep FS
100 (6 vm) column (25 cm×20 mm i.d.) and a
Bischoff precolumn Ultrasep (2 cm×20 mm i.d.).
2.4. Synthesis of (13S,9Z,11E)-13-hydroperoxy-
9,11-octadecadienoic acid (13S-HPODE)
Synthesis of 13S-HPODE was carried out in a
slightly modified procedure as described by Ham-
berg and Samuelsson (1967). Briefly, 1 g of
linoleic acid (3.57 mmol) was suspended in 600 ml
of 0.1 M borate buffer (pH 9.0) and emulsified by
ultrasonic vibration. After removal of the ultra-
sonic bath the emulsion was vigorously magneti-
cally stirred at 20°C, and 100 mg soybean
lipoxygenase (5.3 U) were added. A few minutes
later the solution clarified. After 1 h oxidation
Kovats indices were determined by co-injection
of a 0.2 vl sample of a standard mixture of

134
P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
was stopped by adding 2 N HCl until the solution
became slightly cloudy. The solution was ex-
tracted three times with CHCl3. The CHCl3 layers
were dried by addition of Na2SO4 and filtered.
The solvent was then removed in vacuum at 20°C.
The residue was dissolved in 5 ml of hexane/iso-
propanol solution (10:1, v/v) and slowly passed
through a silica gel cartridge applying a slight
excess of pressure. The cartridge was rinsed with
another 5 ml of solvent. After solvent evapora-
tion, 0.96 mg (86% yield) of raw 13S-HPODE
was obtained as a colourless oil.
GC; if arachidonic acid methylate was still de-
tected, another 4 mg of lipoxygenase were added
and the reaction mixture was stirred for another
hour. Usually the reaction was complete after 2 h.
Preparation of (9S,10E,12Z)-9-hydroxy-10,12-
octadecadienoic acid (9S-HODE and (13S,9Z,
11E)-13-hydroxy,9,11-octadecadienoic acid (13S-
HODE) was done by reduction of 13-HPODE
and 9-HPODE, respectively, with SnCl2’2H2O in
a modified procedure of Hamberg (1971). First,
0.50 g (1.60 mmol) of 13S-HPODE (obtained by
soybean lipoxygenase oxidation) or 9S-HPODE
(obtained by tomato lipoxygenase) was dissolved
in a mixture of 100 ml of CHCl3, 200 ml of
methanol and 80 ml of H2O. The pH of the
solution was adjusted to 3 by addition of 2 N
HCl. Immediately afterwards 395 mg (1.75 mmol)
of SnCl2’2H2O were added. The solution was
stirred for 1 h at room temperature. Separation
into an aqueous and an organic layer was
achieved by addition of 100 ml each of H2O and
CHCl3. After centrifugation the chloroform layer
was separated. The aqueous layer was extracted
another time with 100 ml CHCl3. The combined
chloroform layers were dried with Na2SO4,
filtered and then the solvent was removed at 20°C
in vacuum.
The residue obtained after reduction of 9-
HPODE contained large amounts of linoleic acid.
This was removed by thin-layer chromatography:
0.47 g (1.59 mmol) of raw 9S-HODE was dis-
solved in 2.5 ml of a mixture of cyclohexane, ethyl
acetate and acetic acid 596:400:4 (v/v) and applied
to five thin-layer plates. The plates were devel-
oped with the solvent mixture given above. Imme-
diately after drying, identification of 9S-HODE
was achieved by irradiation of the plate with UV
light of 254 nm. 9S-HODE showed fluorescence
in a band with RF between 0.33 and 0.45. This
band was scraped off. The silica gel was eluted
three times each with 20 ml of dried ethyl acetate.
After filtration the solvent was removed at 30°C
using a rotor vap. Yield: 0.11 g of 9S-HODE
(25%).
2.5. Synthesis of (9S,10E,12Z)-9-hydroperoxy-
10,12-octadecanoic acid (9S-HPODE)
Synthesis of 9S-HPODE was carried out as
described by Galliard (Matthew et al., 1977).
Briefly 1.00 g of linoleic acid was suspended in
300 ml of 0.1 M phosphate buffer (pH 5.6) and 1
ml of Tween 20. Then 1 kg of fresh tomatoes,
homogenized in 500 ml of phosphate buffer (pH
5.6) using a Waring blender, were added. The
solution was vigorously magnetically stirred in air
for 5 h. The reaction was terminated by adjusting
the pH to 2–3 by addition of 2 N H2SO4. The
aqueous solution was extracted three times each
with 100 ml of diethyl ether. Separation of the
layers required centrifugation (1500 rpm, 10 min).
The ethereal solution was dried with Na2SO4.
After filtration the solvent was removed in vac-
uum at 20°C. Yield was 0.91 g of yellow oil which
still contained about two-thirds of linoleic acid.
2.6. Synthesis of (15S,5Z,8Z,11Z,13E)-15-
hydroperoxy-5,8,11,13-eicosatetraenoic acid
(15S-HPETE)
This acid was prepared in a similar way as
described for synthesis of 13S-HPODE following
the descriptions of Van Os et al. (1981) and Funk
et al. (1976). Instead of 1 g of linoleic acid, 100
mg of arachidonic acid were emulsified with 60 ml
of borate buffer pH 9.0 and only 4 mg lipoxyge-
nase were added by stirring. Completion of the
reaction was checked after stirring for 1 h. A
sample was withdrawn, reduced with P(OCH3)3
(see below) and trimethylsilylated and analysed by
Separation from isomeric byproducts of 9-
HODE and 13-HODE was accomplished by
purification with HPLC in a modified procedure
reported by Wu et al. (1995): 100 mg of 9S-

P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
135
HODE, prepurified by thin-layer chromatogra-
phy, or 13S-HODE obtained after reduction of
13S-HPODE, were dissolved in a mixture of 500
vl of hexane/isopropanol 95:5 (v/v). This mixture
was subjected to separation on a silica gel column
(25 cm×10 mm i.d.) with spherical beads (6 vm)
using a mixture of hexane/isopropanol/acetic acid
(99:1:0.1, v/v). Detection was achieved by measur-
ing the UV absorbance at 234 nm (diene system):
13S-HODE: elution between 10.8 and 12.0 min.
(13S,9E,11E)-HODE: elution between 17.6 and
18.6 min.
stopped by the addition of 10 ml of P(OMe)3.
After addition of 0.5 ml of CHCl3, the solution
was shaken at room temperature for 15 min and
then extracted three times with 0.5 ml of CHCl3.
The CHCl3 layer was separated and dried with
Na2SO4. Then the solvent was removed under a
flow of nitrogen at room temperature. The residue
was subjected to methylation and trimethylsilyla-
tion and investigated by GC and GC/MS as de-
scribed below.
Autoxidation of 15S-HETE was performed in a
similar manner, except that instead of 15 mg only
3 mg were used; the amounts of the other reagents
were equally reduced.
9S-HODE: elution between 22.6 and 24.2 min.
(9S,10E,12E)-HODE: elution between 25.8 and
27.4 min.
9S-HODE as its TMS methylate: RI 2309; MS
(70 eV), m/z (%): 382 (17, M+), 311 (15), 259
(2), 225 (100), 155 (44), 143 (23), 130 (35), 73
(82).
2.8. Esterification with diazomethane
The reduced aliquots (0.1–0.2 mg sample), dis-
solved in 10 vl of ethyl acetate, were transformed
to a top shaped vial and 200 vl of ethereal
diazomethane solution were added; after 10 min,
excess diazomethane and solvent was blown off
under a flow of nitrogen.
13S-HODE as its TMS methylate: RI 2297; MS
(70 eV), m/z (%): 382 (25, M+), 311 (100), 225
(40) 155 (17), 143 (23), 130 (50), 73 (80).
15S-HETE was obtained by reduction of 15S-
HPETE: 20 mg HPETE were dissolved in 0.5 ml
CHCl3 (top shaped vial), then 50 vl of P(OCH3)3
were added. The solution was shaken for 15 min
at room temperature. The solvent and excess of
reagent were blown off with nitrogen. The residue
was dissolved in 200 vl of ethyl acetate; 30 vl of
this solution (containing 3 mg of HETE pure
enough for use in oxidation experiments) were
transferred to a 250 ml round bottom flask for
further reaction.
2.9. Trimethylsilylation
MSTFA (10 vl) was added to the methylated
sample and shaken in a thermomixer for 60 min
at 40°C.
2.10. Compound identification
A 0.2–1.5 vl volume of the MSTFA solution
obtained after methylation and trimethylsilylation
was subjected to GC and GC/MS. Mass spectra
were recorded from all GC peaks.
2.7. Autoxidation experiments with 9S-HODE
and 13S-HODE
First, 15 mg (51 vmol) 9S-HODE and 13S-
HODE were emulsified in 22 ml of phosphate
buffer (pH 7.4) and 45 ml of 0.2 M KCl solution
were added. Autoxidation was started by addition
of: (1) 15 mg (85 vmol) of sodium ascorbate and
0.6 ml (1.0 mM) of FeSO4 solution; (2) 15 mg (53
vmol) of FeSO4’7H2O; (3) 15 mg (53 vmol) of
FeSO4’7H2O and 100 vl (50 vmol) of 3% H2O2
solution; (4) 15 mg (53 vmol) of FeCl3’6H2O.
Aliquots of 1 ml were removed after 0, 1, 12 and
24 h, 2 and 5 days, and 2 weeks. Oxidation was
3. Results
Since hydroperoxides in the presence of traces
of Fe2+ readily undergo further oxidation to
products which might also be generated from the
corresponding hydroxy acids (see Scheme 1), per-
fect sample purification was obligatory to study
oxidation of the latter.
13S-HPODE was prepared using soybean
lipoxygenase which develops its highest activity at

136
P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
pH 9.0 (Hamberg and Samuelsson, 1967).
Linoleic acid is of low solubility in borate buffer.
Therefore an emulsion of linoleic acid was pre-
pared applying vigorous stirring under ultrasonifi-
cation. After addition of soybean lipoxygenase the
solution clarified; thus obviously linoleic acid was
bound to lipoxygenase.
It turned out to be unnecessary to blow pure
oxygen through the solution, as recommended
earlier (Hamberg, 1983, Dolev et al., 1967), as this
did not increase the yield of 13S-HPODE. The
oxidation was terminated by addition of 2 N HCl.
HCl addition must be done carefully until slight
clouds become visible (pH 4.5). If more acid is
added, 13S-HPODE may decompose. In addition,
the voluminous precipitated enzyme is then trans-
ferred by following extraction with CHCl3 into
the organic layer. Excess enzyme was removed by
cartrige filtration.
The synthesis of 15S-HETE was performed
analogous to that of 13S-HPODE starting with
arachidonic acid and soybean lipoxygenase. The
product was only obtained in satisfactory yield if
the amount of soybean lipoxygenase was kept
low. The progress of the reaction was controlled
by sample withdrawal and product analysis. If
after 2 h of reaction time starting material was
still present, a second portion of soybean lipoxy-
genase was added and after another 2 h a second
control was performed. When the reaction was
done with the same or higher amounts of enzyme
compared to arachidonic acid, 14,15-epoxy-13-hy-
droxy-5,8,11-eicosatrienoic acid was obtained as
the main product, indicated by MS.
(Wu et al., 1995). We therefore decided to reduce
the raw hydroperoxides without further purifica-
tion, since the hydroxy derivatives produced are
stable to isomerization.
Either NaBH4 (Hamberg and Samuelsson,
1965, Graveland, 1970), SnCl2 (Hamberg, 1971),
P(Ph)3 (Porter et al., 1980, Neff et al., 1981),
P(OCH3)3 or P(CH3)3 (Shulman, 1977) is recom-
mended for reduction of hydroperoxides to the
corresponding alcohols: NaBH4 reduces not only
hydroperoxides, but also attacks oxo compounds.
Since hydroxyoxo acids are generated in the
course of further oxidation of LOOH (Loidl-
Stahlhofen et al., 1994), NaBH4 was not consid-
ered as a reducing agent.
Reduction with SnCl2 should be carried out in
aqueous acidic media at pH 3–4 to ensure solu-
tion of SnCl2. In contrast to previous reported
reductions (Hamberg, 1971), where SnCl2 in an
ethanolic emulsion was applied, we used a homo-
geneous solution for this purpose. Hydroperoxy
fatty acids are almost insoluble in water, therefore
the reduction was done in a mixture of chloro-
form, methanol and water (Bligh and Dyer, 1959).
In this single phase solution both SnCl2 and fatty
acid are well soluble. The reaction was terminated
after a few minutes. Separation of reaction prod-
ucts was achieved by adding chloroform and wa-
ter, forming two layers. This reduction proceeded
nearly completely and without any E/Z isomer-
ization. Therefore this method was used for re-
duction of sample amounts in milligram scale.
P(Ph)3 turned out not to be suitable for reduc-
tion of fatty acid hydroperoxides, since P(Ph)3
and P(Ph3)O (produced as oxidation product after
reduction of LOOH) are soluble in organic sol-
vents and therefore P(Ph3)O is extracted with
organic solvents. The GC retention times of
P(Ph3)O are nearly identical with those of the
main oxidation products of 9-HPODE and 13-
HPODE, causing severe peak overlapping.
9S-HPODE was prepared by a tomato lipoxy-
genase in borate buffer at pH 5.6 obtained by
homogenation of tomatoes. Tomato homogenate
is acidic, therefore the pH must be adjusted dur-
ing the reaction by careful addition of 1 N NaOH
if necessary. Unfortunately, only
a limited
amount of linoleic acid was converted to 9S-
HPODE.
In contrast, P(OCH3)3 was volatile enough for
removal in the course of solvent evaporation. The
reduction was achieved at pH 7.4, avoiding the
danger of acid-induced decomposition of LOOH,
the reaction was complete within minutes, but still
some E/Z isomerization was observed. Neverthe-
less, this method was always applied for reduction
Hydroperoxides readily undergo isomerization
(Hamberg and Samuelsson, 1967, Chan et al.,
1979, Frankel, 1982), even when stored in solu-
tion at 0°C. Decomposition is provoked in the
presence of traces of acid which is necessary to
exclude peak broadening during HPLC separation

P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
137
in miniscale. Reduction with P(CH3)3 was also
checked, but proved less sufficient than that with
P(OCH3)3 due to the high volatility of the reagent.
Samples of 9S-HODE had to be purified by
thin-layer chromatography to remove the large
amounts of linoleic acid by use of a solvent
mixture consisting of cyclohexane, ethyl acetate
and traces of acetic acid. Purification of all hy-
droxy derivatives was achieved with the same
solvent mixture (Wu et al., 1995).
15S-HETE were negative, only the starting mate-
rial was recovered.
In contrast, 9S-HODE and 13S-HODE are not
stable in an organic solvent. If a solution of
13S-HODE in ethyl acetate was stored for 4
months at room temperature in daylight and air,
nearly all the 13S-HODE disappeared.
4. Discussion
The air oxidation experiments with 9S-HODE
and 13S-HODE were carried out by addition of
amounts of iron ions greatly surpassing the Fe2+
concentration in biological samples. Conse-
quently, we deduce that the much smaller iron ion
amounts in biological material are not able to
induce oxidation of 9S-HODE and 13S-HODE.
In first experiments, 9S-HODE and 13S-HODE
as well as 15S-HETE were treated in air at pH 7.4
and room temperature with Fe2+/ascorbate, thus
using typical oxidation conditions applied also for
model oxidation of unsaturated acids (Thiele and
Huff, 1964, Yamagata et al., 1983, Esterbauer et
al., 1986). Even if the reaction temperature was
increased to 45–50°C, and the reaction time was
extended to 2 weeks, oxidation products were
only observed in traces by GC/MS, and 95% of
9S-HODE and 13S-HODE were recovered un-
changed. Under similar oxidation conditions 15S-
HETE also remained unattacked.
Hydroperoxides of linoleic acid (LOOH) and
the corresponding hydroxy acids (LOH) possess
an identical conjugated diene system (see Scheme
1). It was therefore assumed that LOOH and
LOH are equally prone to oxidation (Spiteller,
1996, Dudda et al., 1996a). This assumption was
checked by subjecting 9-HODE and 13-HODE to
air oxidation in the presence of catalytic amounts
of Fe2+ and ascorbate. The latter is a common
reagent to reduce Fe3+ generated in the course of
the reaction (Thiele and Huff, 1964, Wills, 1965).
9-HODE and 13-HODE proved to be stable in
this system. They proved stable also even after
prolonged reaction time (14 days): only traces of
oxidation products were detected. Similar results
were obtained by oxidation with equimolar
amounts of Fe2+/O2, while HPODEs are oxidized
nearly instantly with this reagent. Even addition
of H2O2 to a solution of 13S-HODE and Fe2+
’
Since preliminary experiments with 9S-HPODE
and 13S-HPODE revealed that addition of
equimolar amounts of Fe2+ induces oxidation
reactions much faster than Fe2+/ascorbate, 9S-
HODE, 13S-HODE and 15S-HETE were also
subjected to oxidation in air with Fe2+ alone.
Also in these cases oxidation was not observed.
(Fe2+ +H2O2 generates OH radicals) was not
successful: 95% of starting material was recovered
unchanged. Also after addition of Fe3+ only the
starting material was re-isolated.
This high stability of 9S-HODE and 13S-
HODE seems to indicate that either the hydro-
gens in positions 9 and 13, respectively (see
Scheme 1), are not equally activated as in the
corresponding hydroperoxides (since LOOH read-
ily undergo decomposition) or that the oxidizing
agent is peroxyl radicals.
Hydroxy acids derived by reduction of corre-
sponding hydroperoxides of arachidonic acid, e.g.
(15S,5Z,8Z,11Z,13E) - 15 - hydroxy - 5,8,11,13 -
eicosatetraenoic acid (15S-HETE), possess acti-
vated CH2-groups in positions 7 and 10. We
assumed therefore that these CH2-groups (Scheme
2) would suffer hydrogen abstraction as easily as
’
H2O2 and Fe2+ produce OH radicals in the
Fenton reaction. Therefore as well as Fe2+ we
also added H2O2 to samples of 9S-HODE and
13S-HODE. Even after 2 weeks’ reaction time,
GC revealed only the presence of unchanged
starting material. Also, autoxidation experiments
of 13S-HODE with air using Fe3+ as catalyst
showed that no reaction occurred: the starting
material was recovered unchanged. Addition of
Fe3+ caused also no oxidation. Similarly, the
oxidation experiments with 9S-HODE and with

138
P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
Scheme 2. CH2-groups in positions 7 and 10 are activated by two adjacent double bonds. In spite of this, 15S-HETE is not attacked
’
by LO , in contrast to 15S-HPETE.
Dudda, A., Herold, M., Ho¨lzel, C., Loidl-Stahlhofen, A., Jira,
the activated CH2-groups in arachidonic acid or
linoleic acid. Unexpectedly this assumption
W., Mlakar, A., Scheick, C., Spiteller, G., 1996a. Lipid
peroxidation, a consequence of cell injury?. S. Afr. J.
Chem. 49, 59–64.
proved not to be true. When we subjected 15S-
HETE to oxidation with air/Fe2+/ascorbate or
Dudda, A., Kobelt, F., Spiteller, G., 1996b. Lipid oxidation
Fe2+ even after 2 days nearly no loss of starting
products in ischemic porcine heart tissue. Chem. Phys.
Lipids 82, 39–51.
Esterbauer, H., 1982. Aldehydic products of lipid peroxida-
material was recognized. Thus we are forced to
conlcude that further degradation of lipid peroxi-
dation products requires the presence of a hy-
droperoxy group or a peroxyl radical.
tion. In: McBrien, D.C.H., Slater, T.F. (Eds.), Free Radi-
cals, Lipid Peroxidation and Cancer. Academic Press,
London, pp. 101–128.
Esterbauer, H., Benedetti, A., Lang, J., Fulceri, R., Fauler, G.,
Comporti, M., 1986. Studies on the mechanism of forma-
tion of 4-hydroxynonenal during microsomal lipid peroxi-
dation. Biochim. Biophys. Acta 876, 154–166.
Esterbauer, H., Rotheneder, M., Striegl, G., Waeg, G., Ashy,
A., Sattler, W., Ju¨rgens, G., 1989. Vitamin E and other
lipophilic antioxidants protect LDL against oxidation. Fat
Sci. Technol. 91, 316–324.
Esterbauer, H., Schaur, R.J., Zollner, H., 1991. Chemistry and
biochemistry of 4-hydroxynonenal, malonaldehyde and re-
lated aldehydes. Free Radic. Biol. Med. 11, 81–128.
Esterbauer, H., Gebicki, J., Puhl, H., Ju¨rgens, G., 1992. The
role of lipid peroxidation and antioxidants in oxidative
modification of LDL. Free Radic. Biol. Med. 13, 341–390.
Frankel, E.N., 1982. Volatile lipid oxidation products. Prog.
Lipid Res. 22, 1–33.
The high stability of 9S-HODE and 13S-
HODE to further oxidation explains why in-
creased amounts of these compounds compared
to other LPO products are detected in biological
samples after oxidative stress. Because linoleic
acid usually occurs in biological samples in much
higher amounts than arachidonic acid and due to
the high stability of 9S- and 13S-HODE, these
hydroxy acids are well suited to estimate oxidative
stress exerted in human LDL samples and tissue.
References
Frankel, E.N., 1985. Chemistry of free radicals and singlet
oxidation of lipids. Prog. Lipid Res. 23, 197–221.
Funk, M.O., Ramdas, I., Porter, P.M., 1976. Preparation and
purification of lipid hydroperoxides from arachidonic and
k-linolenic acid. Lipids 11, 113–117.
Gardner, H.W., 1989. Oxygen radical chemistry of polyunsat-
urated fatty acids. Free Radic. Biol. Med. 7, 65–86.
Gardner, H.W., Jursinic, P.A., 1981. Degradation of linoleic
acid hydroperoxides by a cysteine·FeCl3 catalyst as a
model for similar biochemical reactions. Biochim. Biophys.
Acta 665, 100–112.
Belkner, J., Wiesner, R., Ku¨hn, H., Lankin, V.Z., 1991. The
oxygenation of cholesterol esters by the reticulocyte lipoxy-
genase. FEBS Lett. 279, 110–114.
Bligh, E.G., Dyer, W.J., 1959. A rapid method of total lipid
extraction and purification. Can. J. Biochem. Physiol. 37,
911–917.
Braughler, J.M., Hall, E.D., 1989. Central nervous system
trauma and stroke. I. Biochemical considerations for oxy-
gen radical formation and lipid peroxidation. Free Radic.
Biol. Med. 6, 289–301.
Brooks, C.J.W., Harland, W.A., Steel, G., Gilbert, J.D., 1970.
Lipids of human atheroma: isolation of hydroxyoctadeca-
dienoic acids from advanced aortal lesions. Biochim. Bio-
phys. Acta 202, 563–566.
Chan, H.W.-S., Levett, G., Matthew, A., 1979. The mecha-
nism of the rearrangement of linoleate hydroperoxides.
Chem. Phys. Lipids 24, 245–256.
Dolev, A., Rohwedder, W.K., Dutton, H.J., 1967. Mechanism
of lipoxidase reaction. I. Specificity of hydroperoxidation
of linoleic acid. Lipids 2, 28–32.
Graveland, A., 1970. Enzymatic oxidations of linoleic acid and
glycerol-1-monolinoleate in doughs and flour–water sus-
pensions. J. Am. Oil Chem. Soc. 47, 352–361.
Hamberg, M., 1971. Steric analysis of hydroperoxides formed
by lipoxygenase oxygenation of linoleic acid. Anal.
Biochem. 43, 515–526.
Hamberg, M., 1973. New reaction of unsaturated fatty acid
hydroperoxides. Formation of 11-hydroxy-12,13-epoxy-9-
octa-decenoic acid from 13-hydroperoxy-9,11-octadeca-
dienoic acid. Lipids 8, 737–744.

P. Spiteller, G. Spiteller / Chemistry and Physics of Lipids 89 (1997) 131–139
139
Hamberg, M., 1983. Autoxidation of linoleic acid. Isolation
and structure of four dihydroxyoctadecadienoic acids.
Biochim. Biophys. Acta 752, 353–356.
Hamberg, M., Samuelsson, B., 1965. On the specifity of the
lipoxidase catalyzed oxygenation of unsaturated fatty
acids. Biochem. Biophys. Res. Commun. 21, 531–536.
Hamberg, M., Samuelsson, B., 1967. Mechanism of the
biosynthesis of prostaglandins E1 and F1h. J. Biol. Chem.
242, 5336–5343.
Harland, W.A., Gilbert, J.D., Brooks, C.J.W., 1973. Lipids of
human atheroma. VIII. Oxidized derivatives of cholesteryl
linoleate. Biochim. Biophys. Acta 316, 378–385.
Jira, W., Spiteller, G., 1996. Increase in hydroxy fatty acids in
human low density lipoproteins with age. Chem. Phys.
Lipids 84, 165–173.
Neff, W.E., Frankel, E.N., Weisleder, D., 1981. High-pressure
liquid chromatography of autoxidized lipids. II. Hydroper-
oxy-cyclic peroxides and other secondary products from
methyl linolenate. Lipids 16, 439–448.
Nishigaki, I., Hagihara, M., Tsunekawa, H., Maseki, M.,
Yagi, K., 1981. Lipid peroxide levels of serum lipoprotein
fractions of diabetic patients. Biochem. Med. 25, 373–378.
Porter, N.A., Wolf, R.A., Weenen, H., 1980. The free radical
oxidation of polyunsaturated lecithins. Lipids 15, 163–167.
Shulman, J.I., 1977. Reaction of organo-group 5A compounds
with tert-butyl hydroperoxide. J. Org. Chem. Soc. 42,
3970–3972.
Spiteller, G., 1996. Enzymic lipid peroxidation—a conse-
quence of cell injury?. Free Radic. Biol. Med. 21, 1003–
1007.
Jira, W., Spiteller, G., Richter, A., 1997. Increased levels of
lipid oxidation products in low density lipoproteins of
patients suffering from rheumatoid arthritis. Chem. Phys.
Lipids 87, 81–89.
Kovats, E., 1958. Gaschromatographische Charakterisierung
organischer Verbindungen. 1. Retentionsindices aliphatis-
cher Halogenide, Alkohole, Aldehyde und Ketone. Helv.
Chim. Acta 41, 1915–1932.
Steinbrecher, U.P., Parthasarathy, S., Leake, D.S., Witztum,
J.L., Steinberg, D., 1984. Modification of low-density lipo-
protein by endothelial cells involves lipid peroxidation and
degradation of low-density lipoprotein phospholipids.
Proc. Natl. Acad. Sci. U. S. A. 81, 3883–3887.
Thiele, E.H., Huff, J.W., 1964. Thiobarbituric reacting sub-
stance(s) produced in normal liver fractions. Arch.
Biochem. Biophys. 104, 468–472.
Ku¨hn, H., Belkner, J., Wiesner, R., Schewe, T., Lankin, V.Z.,
Tikhaze, A.K., 1992. Structure elucidation of oxygenated
lipids in human atherosclerotic lesions. Eicosanoids 5, 17–
22.
Loidl-Stahlhofen, A., Spiteller, G., 1994. Generation of h-hy-
droxy-aldehydic compounds in the course of lipid peroxi-
dation. Biochim. Biophys. Acta 1213, 140–148.
Loidl-Stahlhofen, A., Hannemann, K., Spiteller, G., 1994.
h-Hydroxyaldehydes, products of lipid peroxidation.
Biochim. Biophys. Acta 1211, 156–160.
Van Os, C.P.A., Rijke-Schilder, G.P.M., Vanhalbeek, H., Ver-
hagen, J., Vliegenthart, J.F.G., 1981. Double dioxygena-
tion of arachidonic acid by soybean lipoxygenase—1.
Kinetics and regiostereo specificities of the reaction steps.
Biochim. Biophys. Acta 663, 177–179.
Wang, T., Powell, W.S., 1991. Increased levels of mono-hy-
droxy metabolites of arachidonic acid and linoleic acid in
LDL and aorta from atherosclerotic rabbits. Biochim.
Biophys. Acta 1084, 129–138.
Wills, E.D., 1965. Mechanisms of lipid peroxide formation in
tissues. Role of metals and haematin proteins in the cataly-
sis of the oxidation of unsaturated fatty acids. Biochim.
Biophys. Acta 98, 238–251.
Marnett, L.J., Wilcox, A.L., 1995. The chemistry of lipid
alkoxyl radicals and their role in metal-amplified lipid
peroxidation. Biochem. Soc. Symp. 61, 65–72.
Matthew, J.A., Chan, H.W.-S., Galliard, T., 1977. A simple
method for the preparation of pure 9-D-hydroperoxide of
linoleic acid and methyl linoleate based on the positional
specificity of lipoxygenase in tomato fruit. Lipids 12, 324–
326.
Wingard, P.G., Tatzber, F., Esterbauer, H., Kus, M., Blake,
D.R., Morris, C.J., 1993. Presence of foam cells containing
oxidized low density lipoprotein in the synovial membrane
from patients with rheumatoid arthritis. Ann. Rheum. Dis.
52, 677–680.
Mickel, H.S., 1978. Peroxidized arachidonic acid effects on
human platelets and its proposed role in the induction of
damage to white matter. In: Kabara, J.J. (Ed.), Symp.
Pharmacol. Eff. Lipids. American Oil Chemists Society,
Champaign, IL, pp. 179–190.
Wu, Z., Robinson, D.S., Domoney, C., Casey, R., 1995.
High-performance liquid chromatographic analysis of the
products of linoleic acid oxidation catalyzed by pea (Pisum
sati6um) seed lipoxygenases. J. Agric. Food Chem. 43,
337–342.
Morel, D.W., Hessler, J.R., Chisolm, G.M., 1983. Low density
lipoprotein cytotoxicity induced by free radical peroxida-
tion of lipid. J. Lipid Res. 24, 1070–1076.
Yamagata, S., Murakami, H., Terao, J., Matsushita, S., 1983.
Nonenzymic oxidation products of methyl arachidonate.
Agric. Biol. Chem. 47, 2791–2799.
.