6

Assay of Phospholipid Hydraperoxides

by Chemiluminescence-Based High-Performance
Liquid Chromatography
Yorihiro Yamamoto, Yasuhiro Kambayashi, and Takako Ueda

1. Introduction
Lipid hydroperoxides (LOOH) are the primary products of lipid
peroxidation. Therefore, it would be desirable to detect and identify the lipid
hydroperoxides m tissues or blood. One of the most advanced methods for the
detection of LOOH is a chemiluminescence-based high-performance liquid
chromatography (HPLC) assay (1-5) as shown in Fig. 1. The reaction sequence
leading to the emission of light from isoluminol in the presence of LOOH and
microperoxidase, a heme catalyst, is assumed to be as follows:
LOOH + microperoxidase+ LO (1)
LO’ + isoluminol (QH-) + LOH + semiqumone radical (Q’-) (2)
Q’- + O2 + Q + O;- (3)
Q’- + 05 + isoluminol endoperoxide + light (A,,, = 430 nm) (4)
This assay has several advantages:

1 It is very sensitive, with the lower limit of detection being about 0.1 pmol or
less;
2. Interference by biological antioxidants is avoided because hydroperoxides and
antioxidants can be separated by HPLC; and
3. Information on which lipid class is oxidized can be obtained.

Usmg this assay, the presence of about 3 nM cholesteryl ester hydroperox-

ide (CE-OOH) in healthy human plasmas has been reported (6). Rats have
higher plasma CE-OOH levels (about 10 nM) than do humans (7). On the other
From Methods m Molecular Bology, vol 108 free Radtcal and Anboxrdanf Protocols
Edlted by D Armstrong 0 Humana Press Inc , Totowa, NJ

63
64 Yamamoto, Kambayashi, and Ueda

hand, phosphattdylcholme hydroperoxide (PC-OOH) IS undetectable in human

(2,3,6) and rat (7) plasma. The reason for the absence of PC-OOH in human
and rat plasma are as follows: Plasma glutathione peroxidase can reduce
PC-OOH to its alcohol, but not CE-OOH (8); PC-OOH m high density lipopro-
tein is converted to CE-OOH by 1ecithin:cholesterol acyltransferase (9).
Although phospholipid hydroperoxides (PL-OOH) are likely to be transient
products m vivo, it is a very useful marker of hprd peroxidation at least in vttro
and ex vivo systems. Here we describe the method for the detection of
PL-OOH at picomole levels and examples of its application.

2. Materials
2.1. Phospholipid Hydroperoxide Standards
1. Phosphatrdylcholme, soybean (Sigma, St. Lotus, MO).
2. a-Tocopherol.
3. 0.02% Triethylamme m methanol
4 Octadecylsrlyl column (Superiorex ODS, 20 x 250 mm, Shrserdo, Tokyo).
5 Phosphatrdylethanolamme: Soybean (Sigma)
6. Phosphatidylserme* Bovine brain (Sigma).
7. Phosphatrdylinosrtol. Soybean (Srgma)
8 Spectrometer.

2.2. Tissue Extraction

1. 100 pM 2,6-Dr-tert-butyl-4-methylphenol (BHT) in methanol.
2. 100 p&f BHT m chloroform/methanol (2/l [v/v]).
3 Microfuge (Eppendorf).

2.3. Reaction Solution

1. 100 m1I4 Borate buffer, pH 10.
2. 6-Amino-2,3-dihydro-1,4-phthalazine-drone (rsoluminol; Sigma).
3. Mrcroperoxrdase (MP- 11, Sigma)

2.4. Chemiluminescence Assay

1. 40 mM Sodium phosphate, monobasrc.
2. ter&Butyl alcohol
3 5 piV,4.6 x 250 mm Silica column (Supelco Japan).
4. 5 pm, 4.6 x 250 mm Ammopropyl column (Supelco).
5. 5 pm, 4.6 x 250 mm Silica gel guard column (Supelco).
6 Mixer: Model 2500-0.22 (Kratos, Westwood,NJ)
7 Chemrluminescence detector* Model 825-CL (Japan Spectroscoprcs, Tokyo).
Assay of LOOH HPLC 65

3. Method
3.1. Preparation of Phospholipids Hydroperoxides
Phosphatidylcholine was purified on a semipreparative octadecylsilyl col-
umn using methanol containing 0.02% triethylamine as the mobile phase at a
flow rate of 10 mL/min (I). PC-OOH were prepared by the aerobic autoxida-
tlon of purified phosphatidylcholme at room temperature for several d in the
presence of a-tocopherol (VE) and in the absence of radical imtlators. VE 1s
added to reduce the formation of truns,truns-hydroperoxldes, and a high con-
centration of VE accelerates the autoxidation of lipids (10). Synthesized PC-
OOH was purified with the same HPLC conditions previously described. The
concentration of PC-OOH was determmed by Its absorbance at 234 nm (E =
28,00O/M@cm) (II). Phosphatidylethanolamine hydroperoxide (PE-OOH),
phosphatidylserine hydroperoxide (PS-OOH), and phosphatidylmositol hydro-
peroxide (PI-OOH) were also prepared by the autoxidation of parent lipids.

3.2. Tissue Extraction

PC-OOH in plasma (or lipoprotein) can be extracted by shaking the sample
vigorously with 4 volumes of methanol contaimng 100 pjJ4 BHT (to prevent
oxidation). It is necessary to extract PC-OOH in tissue homogenate with two
volumes of chloroform/methanol (2/l, v/v) containing 100 pM BHT. After
centrifugation at 12,000 rpm for 3 mm, ahquots of the either methanol extract
from plasma or chloroform/methanol phase are injected onto HPLC (see Notes
1 and 2).

3.3. Chemiluminescence Assay

Equipment 1sarranged as outlined m Fig. 1. Antioxidants and hydroperox-
ides in the sample are separated by HPLC, using one of the chromatographic
conditions described below, and then mlxed with a reaction solution contain-
ing isoluminol and microperoxidase in a Kratos special mixer (see Note 3).
The reactions leading to the emission of light take place in a mixing coil made
from a piece of HPLC tubing (45 cm, inner volume -92 pL). The emitted light
is measured by a chemiluminescence detector.
The reaction solution for pump B is prepared as follows: 100 mM aqueous
borate buffer (38.14 g of sodium tetraborate decahydrate per/L) is prepared,
and the pH is adjusted to 10 with sodium hydroxide. Isoluminol (177.2 mg,
final concentration 1 rnM) is dissolved in 500 mL of methanol and 500 mL of
the aforementioned borate buffer, and then 5 mg of mlcroperoxidase is added.
The mobile phase for pump A is methanolltert-butyl alcohol/40 mM monoba-
66 Yamamoto, Kambayashi, and Ueda
sample

HPLC moblle phase pump A ---+ Injector

column

UV detector

4
reaction solution b pumpB ___) mixer
(k3olumlnol, mwoperox~dase)

mixing co11

chemlluminescence detector

Fig 1. Schematic diagram of the HPLC-isolummol chemllummescence assay.

sic sodium phosphate (6/3/l [v/v/v]). The flow rates of pump A and B are 1.0
and 1.5 mL/min, respectively (see Notes 4-6).

3.4. Silica Column for the Assay of PC-OOH

When a silica column is used, PC-OOH elutes at about 12 min. In this con-
dition, ascorbate, urate, and tocopherols elute within 5 mm and do not interfere
with the PC-OOH detection. Figure 2 demonstrates the absence of PC-OOH in
human plasma (Line 1). Lme 2 shows the PC-OOH peak spiked in methanol
extract. Lme 3 shows the absence of PC-OOH in the monophasic extract of
human plasma with four volumes of chloroform/methanol (l/2 [v/v]). Line 4
shows the formation of PC-OOH durmg the Folch extraction of human plasma.
Figure 3 shows the formation and the decay of PC-OOH during the incubation
of isolated hepatocytes with tert-butyl hydroperoxide (BOOH) at 37°C under
aerobic conditions, suggesting that the cell is capable of decomposmg
PC-OOH (12).
Assay of LOOH HPLC 67

I I I I I 1

0 2 4 6 8 10 12 14
Time (min)

Fig. 2. Chemiluminescence chromatogram of methanol extract from a healthy hu-

man plasma analyzed with a silica column using methanol/&t-butyl alcohol/40 mA4
monobasic sodium phosphate (6/3/l [v/v/v]) as the mobile phase. Line 1, injected
50 pL of methanol extract (corresponding to 10 pL of plasma), line 2, injected 50 pL
of the same methanol extract spiked with 0.8 pmol(80 nM) PC-OOH, line 3, injected
50 pL of chloroform/methanol (l/2 [v/v]) extract (correspondmg to 10 pl of plasma);
line 4, iqected Folch extract (corresponding to 40 PL of plasma).

3.5. Aminopropyl Column for the Assay of PL-OOH

Biological membranes consist of phospholipids, such as phosphatidylcho-
line, phosphatidylethanolamine, phosphatidylserine, and phosphatidylinositol.
The assayfor the measurement of hydroperoxides of these phospholipids would
be useful for studying the oxidation of biomembranes. Figure 4 demonstrates
68 Yamamoto, Kambayashi, and Ueda

-0 20 40 60 80
Time (min)
Fig. 3. Formatton and the decomposmon of PC-OOH durmg the mcubation of ISO-
lated hepatocytes with 0,O 25, 0.50, and 0.75 mA4 tert-butyl hydroperoxide (BOOH)
at 37°C under aerobic condttions.

the separation of PC-OOH, PE-OOH, PS-OOH, and PI-OOH using two setsof
silica-gel guard columns and an ammopropyl column m series. We have
observed the formatlon of PC-OOH and PE-OOH during the oxidation of liver
homogenate initiated with free-radical initiators or induced with BOOH (13).

4. Notes
1. Samples should be kept at -80°C for long storage.
2. Lipids are very susceptrble to oxtdatron when they are extracted from brological
samples because peroxrdases and other antioxidant enzymes are inactivated by
the extraction. a-Tocopherol can work as prooxidant m the organic extract (IO).
Evaporation of the extracts usually causes artifactual formatron of hptd hydrop-
eroxtdes as described m the text.
3. To mix an eluant and a chemrlummescence reagent well, the Kratos mixer IS
used, because rt 1s desrgned to accomplish a complete mixing of two soluttons
Even rf the Kratos mixer is replaced by a simple T~omt, only a small change was
observed in peak areas, suggesting that the rates of reacttons of hydroperoxtde-
induced oxrdatron of lsolummol are very rapld (5) Therefore, it IS not necessary
to use the Kratos mixer.
4. HPLC grade methanol should be used. When 2-propanol IS used instead of tert-
butyl alcohol, the baselme of chemrlummescence detector IS high and unstable
because of the high content of hydroperoxrdes in 2-propanol.
5. Methanol/40 mAI monobasic sodium phosphate (9/l [v/v]) can be used as a
mobile phase for a sthca column This shortens the retention ttme of PC-OOH
from 12 mm to 7 min.
Assay of LOOH HPLC 69

0 2 4 6 8 10 12
Time (min)
Frg. 4. Separation of 10 pmol of membrane phospholipids hydroperoxides by HPLC
monitored by chemrlummescence Two sets of sthca gel guard columns and
ammopropyl column m series were used; methanolltert-butyl alcohol/40 miU monoba-
sic sodium phosphate (6/3/l [v/v/v]) was used as the mobile phase.

6. Disappearance of chemrluminescence positive peak by the pretreatment with

triphenylphosphme assures that the peak is hydroperoxtde, because lipid hydrop-
eroxtdes are readily reduced by triphenylphosphine at room temperature

References
1. Yamamoto, Y , Brodsky, M. H., Baker, J. C., and Ames, B N. (1987) Detection
and characterization of lipid hydroperoxrdes at prcomole levels by hrgh-perfor-
mance liquid chromatography. Anal. Biochem. 160,7-13
2. Yamamoto, Y and Ames, B N. (1987) Detection of hprd hydroperoxtdes and
hydrogen peroxide at picomole levels by an HPLC and tsoluminol chemrlummes-
cence assay. Free Rad. Biol. Med.. 3,359-361.
3. Frer, B., Yamamoto, Y., Nrclas, D , and Ames, B. N (1988) Evaluatron of an
rsoluminol chemilummescence assay for the detection of hydroperoxtdes in
human blood plasma. Anal. Biochem. 175, 120-130.
4 Yamamoto, Y., Frer, B , and Ames, B. N. (1990) Assay of hprd hydroperoxtdes
using HPLC with post-column chemrlummescence detection Methods Enzymol
186,371-380.
5. Yamamoto, Y (1994) Chemdummescence-based high-performance hqmd chro-
matography assay of hprd hydroperoxrdes Methods Enzymol 233,3 19-324
70 Yamamoto, Kambayashi, and Uecfa

6 Yamamoto, Y and Ntkr, E. (1989) Presence of cholesteryl ester hydroperoxide m

human blood plasma. Blochem Bzophys. Res. Commun. 165,988-993.
7 Yamamoto, Y , Wakabayasht, K , Niki, E., and Nagao, M. (1992) Comparison of
plasma levels of lipid hydroperoxrdes and anttoxtdants m hyperhpidemtc Nagase
analbummemic rats, Sprague-Dawley rats, and humans. Biochem. Bzophys. Res.
Commun. 189,5 18-523.
8 Yamamoto, Y., Nagata, Y , Nrkr, E., Watanabe, K., and Yoshrmura, S. (1993)
Plasma glutathrone peroxrdase reduces phosphattdylcholme hydroperoxrde
Blochem. Blophys Res. Commun 193,133-138.
9. Nagata, Y., Yamamoto, Y., and Nrkt, E (1996) Reaction of phosphatidylcholme
hydroperoxrde m human plasma: The role of peroxtdase and 1ectthm:cholesterol
acyltransferase Arch Blochem. Blophys. 329,24-30.
10 Bowry, V. W and Stocker, R. (1993) Tocopherol-medrated peroxldatron. The
prooxidant effect of vitamin E on the radical-uutrated oxrdatron of human low-
density hpoprotem. J. Am. Chem. Sot. l&6029-6044.
11. Chan, H. W.-S and Levett, G. (1977) Autoxrdatton of methyl lmoleate Separa-
tion and analysis of tsomertc mixtures of methyl lmoleate hydroperoxtdes and
methyl hydroxylinoleates Lipids 12,99-104.
12. Nakamura, Y, Ohori, Y., and Yamamoto, Y. (1992) Formation of phosphatidyl-
choline hydroperoxtde m isolated rat hepatocytes treated with rert-butyl hydrop-
eroxrde, in Oxygen Radicals Proceedings of the 5th International Congress on
Oxygen Radicals (Yagi, K., Kondo, M., Nrkr, E., and Yoshtkawa, T., eds.), Excerta
Medrca, Amsterdam, pp 273-276.
13. Kambayashr, Y., Yamashita, S., Nrkt, E., and Yamamoto, Y. (1997) Oxtdatton of
rat liver phosphohptds: comparison of pathways m homogeneous solutron, m h-
posomal suspension and m whole tissue homogenates J Blochem. 121,425-43 1