Clinical Science (2001) 100, 343–355 (Printed in Great Britain)
Increased levels of high-density lipoprotein
cholesterol are ineffective in inhibiting the
development of immune responses to
oxidized low-density lipoprotein and
atherosclerosis in transgenic rabbits
expressing human apolipoprotein (apo)
A-I with severe hypercholesterolaemia
Agne' s BOULLIER*, Nathalie HENNUYER*, Anne TAILLEUX*, Christophe FURMAN*,
Nicolas DUVERGER†, Jean-Michel CAILLAUD†, Graciella CASTRO*,
Catherine FIEVET*, Jean-Charles FRUCHART* and Patrick DURIEZ*
*Department of Atherosclerosis, Pasteur Institute, INSERM U325 and University of Lille II, Lille, France, and †Aventis,
Center of Research, Vitry-Alfortville, Vitry sur Seine, France
A
B
S
T
R
A
C
T
High levels of high-density lipoprotein (HDL) cholesterol have been reported to protect against
the development of atherosclerosis in humans by increasing reverse cholesterol transport and
inhibiting the oxidation of low-density lipoprotein (LDL) due to the paraoxonase content of
HDL. The purpose of the present study was to assess if there are any relationships between in
vivo increases in serum levels of immunological LDL oxidation markers [autoantibodies against
oxidized LDL, autoantibodies against malondialdehyde-modified LDL, LDL immune complexes
and anti-cardiolipin autoantibodies], paraoxonase activity and the development of atherosclerosis in control rabbits and in transgenic rabbits expressing human apolipoprotein (apo) A-I.
A total of 13 apo A-I transgenic rabbits and 18 non-transgenic littermates were fed on a
cholesterol-rich diet (0.4 %, w/w) for 14 weeks, and were monitored at weeks 0, 2, 6, 10 and 14.
Aortic atherosclerotic lesions were measured at the end of this period. Human apo A-I
transgenic rabbits with high HDL cholesterol levels were not protected against the development
of atherosclerosis when they were fed on a cholesterol-rich diet which induced dramatic
hypercholesterolaemia. Immunological markers of LDL oxidation increased and serum
paraoxonase activity decreased similarly in control and transgenic rabbits. In conclusion, the
present study demonstrates that high HDL cholesterol levels are ineffective in inhibiting
increases in immunological markers of LDL oxidation and the development of atherosclerosis in
a mammal with severe hypercholesterolaemia.
Key words : anti-cardiolipin autoantibodies, anti-(oxidized LDL) autoantibodies, apolipoprotein A-I, atherosclerosis, IgG-bound
LDL, paraoxonase, rabbit, transgenesis.
Abbreviations : apo, apolipoprotein ; Ab-C, anti-cardiolipin autoantibodies ; LDL, low-density lipoprotein ; MDA, malondialdehyde ; Ab-LDL, autoantibodies against native LDL ; Ab-MDA-LDL, autoantibodies against MDA-modified LDL ;
Ab-OxLDL, autoantibodies against Cu#+-oxidized LDL ; HDL, high-density lipoprotein ; IgG-LDL, circulating immune complexes
measured as IgG-bound LDL ; MDA-LDL, MDA-modified LDL ; OxLDL, oxidized LDL ; PON-1, paraoxonase-1 ; TBARS,
thiobarbituric acid-reacting substances.
Correspondence : Dr Patrick Duriez, De! partement d’Athe! roscle! rose, INSERM U325, Institut Pasteur, 1, rue du Professeur
Calmette, BP 245, 59019 Lille cedex, France (e-mail Patrick.Duriez!pasteur-lille.fr).
# 2001 The Biochemical Society and the Medical Research Society
343
344
A. Boullier and others
INTRODUCTION
Substantial evidence indicates that oxidized low-density
lipoprotein (OxLDL) contributes to atherogenesis
through a number of mechanisms [1]. One of the critical
steps in the development and progression of atherosclerosis is the enhanced uptake of OxLDL by macrophages, which results in the formation of lipid-laden
foam cells, an early step in the development of atherosclerosis [1].
Even minor modifications to LDL render it immunogenic [2], and circulating autoantibodies that recognize
several forms of OxLDL, in particular that containing
malondialdehyde (MDA)-modified lysine, are prevalent
in humans and other species [3–5]. Plasma levels of
autoantibodies directed against MDA-modified LDL
(MDA-LDL) may be valuable in predicting the progression of atherosclerosis in humans [3]. Anti-OxLDL
autoantibodies are present in lesions, partly as immune
complexes with OxLDL [6], and immunoglobulinbound lipoprotein (circulating immune complexes) is a
marker of familial hypercholesterolaemia and atherosclerosis [7–9].
The anti-phospholipid antibody syndrome is characterized by the presence of circulating autoantibodies that
bind in vitro to moieties formed by negatively charged
phospholipids, such as cardiolipin (anti-cardiolipin autoantibodies ; Ab-C). The presence of a high level of Ab-C
has been reported to be an independent risk factor for
myocardial infarction or cardiac death [10]. A strong
cross-reaction has been reported between antibodies to
OxLDL and those to cardiolipin [11,12]. Ho$ rkko$ et al.
[13] demonstrated that Ab-C are directed against
epitopes of oxidized phospholipids, and cloned a monoclonal antibody that binds identically to OxLDL and to
oxidized cardiolipin.
Epidemiological studies in humans have suggested that
high levels of high-density lipoprotein (HDL) and
apolipoprotein (apo) A-I protect against the progression
of atherosclerosis [14]. This hypothesis was confirmed in
transgenic mice overexpressing human apo A-I [15].
However, most of the plasma cholesterol in mice is
carried in HDL and not in LDL, unlike in humans. On
the other hand, rabbits, like humans, are ‘ LDL mammals ’, in that the LDL fraction is the major carrier of
cholesterol in plasma [16]. Furthermore, rabbits develop
a human type of atherosclerotic lesion [17]. In these
animals, overexpression of human apo A-I induces a rise
in HDL cholesterol [18] and a decrease in the development of atherosclerosis [19]. This protection is probably related to an increase in so-called ‘ reverse cholesterol
transport ’ in transgenic rabbits, since we reported that
the first step (i.e. cellular cholesterol efflux) in this
metabolic pathway is increased in these animals [18].
Recent experimental evidence has suggested that
HDLs inhibit atherogenesis not only through their
# 2001 The Biochemical Society and the Medical Research Society
capacity to induce reverse cholesterol transport, but also
because they inhibit LDL oxidation [20–22] through
their paraoxonase-1 (PON-1) content [21]. PON-1 is a
calcium-dependent HDL-associated ester hydrolase [23].
Low levels of HDL-associated PON-1 are correlated
with hypercholesterolaemia [24] and susceptibility to
myocardial infarction [25,26]. Watson et al. [27]
suggested that PON-1 in HDL may protect against the
induction of inflammatory responses in artery wall cells
by destroying biologically active lipids in mildly oxidized
LDL. Recently it was reported that HDL isolated from
PON-1-deficient mice was unable to prevent LDL
oxidation [28,29], and that PON-1-null mice were more
susceptible to atherosclerosis than their wild-type littermates [28] or apo E-null littermates [29].
PON-1 is not the only protein of HDL with antioxidant activity, as human apo A-I possesses antioxidant
properties that might neutralize LDL lipid peroxidation
[30]. We have recently demonstrated that PON-1 activity
is reduced by a pro-atherosclerotic diet in control rabbits
and in transgenic rabbits expressing human apo A-I [31].
Therefore the present study was undertaken with the
following aims. (1) To determine if, in cholesterol-fed
rabbits, the concentrations of immunological LDL oxidation markers [autoantibodies against Cu#+-oxidized
LDL (Ab-OxLDL), autoantibodies against MDA-LDL
(Ab-MDA-LDL), Ab-C and circulating immune complexes measured as IgG-bound LDL (IgG-LDL)] are
correlated with the extent of development of aortic
atherosclerosis. (2) To evaluate if the transgenesis of
human apo A-I to the rabbit modifies the changes in the
concentrations of these markers in proportion to the
expected decrease in the development of atherosclerosis
during cholesterol feeding. (3) To determine possible
relationships between serum HDL levels, serum PON-1
activity, serum concentrations of LDL immunological
oxidation markers and the development of atherosclerosis during cholesterol feeding in control and human apo
A-I transgenic rabbits.
MATERIALS AND METHODS
Animal model
Groups of 13 New Zealand White rabbits transgenic for
human apo A-I (line 8 [19]) (four males, nine females) and
18 non-transgenic littermates (10 males, eight females)
(age 7 months) were used. Animals were housed individually in the Charles River Centre (Saint Aubin les
Elbeuf, France). Atherosclerosis was induced by feeding the animal on a cholesterol-rich (0.4 %, w\w) diet
for 14 weeks (120 g per day). The atherogenic diet was
prepared by spraying normal chow (NIH-09) with
cholesterol.
Blood was collected after 0, 2, 6, 10 and 14 weeks of the
high-cholesterol diet, from animals that had been starved
Oxidative markers in human apolipoprotein A-I transgenic rabbit
overnight. Serum was separated by low-speed centrifugation and kept either at 4 mC until lipid and lipoprotein
analysis ( 1 week) or frozen at k20 mC until measurement of anti-OxLDL markers and PON-1 activity ( 6
months).
The experimental procedures were carried out in
accordance with United States NIH guidelines [32].
Lipid measurements
Serum lipids (cholesterol, phospholipids and triacylglycerols) and HDL cholesterol were measured
colorimetrically using a microtitre plate reader and
commercially available reagents (Boehringer Mannheim).
Cholesterol was measured in the HDL-containing supernatant after precipitating apo B-containing lipoproteins
with sodium phosphotungstate\magnesium chloride
(Boehringer Mannheim). Non-HDL cholesterol values
were calculated by subtracting HDL cholesterol from
total cholesterol.
Measurement of immunological OxLDL
markers
LDL isolation and modification
LDL (1.019 d 1.063) was isolated by sequential
ultracentrifugation [33] from the pooled serum of five
control rabbits. The protein content of LDL was
measured [34]. LDLs were oxidized (OxLDL) by incubation for 24 h at 30 mC with CuSO (0.100 mg\ml
%
protein and 5 µmol\l CuSO ). MDA-LDL was produced
%
by adding freshly prepared MDA, as described by
Palinski et al. [2].
The extent of oxidation was determined by analysing
various markers of lipoprotein peroxidation : dienes were
measured at 234 nm in solution containing 0.1 mg\ml
protein, using ε l 2.95i10% M−":cm−" [35]. Thio#$%
barbituric acid-reacting substances (TBARS) were
measured using fluorimetric detection [36]. Lipid
peroxides were measured using the lipid-peroxidedependent oxidation of iodide to iodine [37]. The
fluorescence intensity of modified LDL, corresponding
to apolipoproteins modified by lipid peroxidation
products (in particular 4-hydroxynonenal) [38], were
measured at λexcitation l 360 nm\λemission l 430 nm and
at λexcitation l 400 nm\λemission l 470 nm. The electrophoretic mobility of modified LDL relative to native
LDL was measured using agarose gels and barbital buffer
(Sebia, Issy-les-Moulineaux, France). Free amino groups
were measured using trinitrobenzenesulphonic acid [39].
Measurement of autoantibodies against native LDL (Ab-LDL), AbOxLDL and Ab-MDA-LDL
Autoantibodies were measured by non-competitive
ELISA. A 96-well microtitre plate (Costar, Cambridge,
MA, U.S.A.) was divided into four quarters, and
one quarter was coated with 200 µl of 2 % (w\v) BSA, one
quarter with native LDL (100 µl ; 5 µg\ml), one quarter
with OxLDL (100 µl ; 5 µg\ml) and the final quarter with
MDA-LDL (100 µl ; 5 µg\ml) in PBS containing
2.7 mmol\l EDTA and 20 µmol\l butylated hydroxytoluene to prevent LDL oxidation. The plates were
incubated for 16 h at 4 mC. The plates were then washed
four times with 0.1 mol\l PBS containing 0.05 % Tween
20, 0.1 % BSA and 0.001 % aprotinin, and then blocked
with 200 µl of 2 % BSA in 0.1 mol\l PBS for 1 h at room
temperature. Plates were subsequently washed an additional four times with 0.1 mol\l PBS, and then serum
diluted 1 : 20 (v\v) in buffer A (0.1 mol\l PBS containing
0.05 % Tween 20 and 0.2 % BSA) was incubated for 2 h
at room temperature. Plates were then washed four times
with buffer A. Peroxidase-conjugated anti-(rabbit IgG)
prepared from goats (Sigma) and diluted 1 : 3000 in buffer
A was added (100 µl), and the mixture was left for 2 h
at 37 mC. After four successive washings with buffer
A, peroxidase substrate (o-phenylenediamine ; 100 µl,
1.5 g\l) in solution in phosphate citrate buffer (pH 5.5)
containing hydrogen peroxide (3.5 mmol\l) was added to
each well. The reaction, carried out in the dark, was
stopped with 100 µl HCl (1 mol\l) after 30 min ;
absorbance was measured at 492 nm. Autoantibody titres
were expressed in absorbance units. The within-run and
between-run coefficients of variation were 6 % and 9 %
respectively. All samples from one animal (i.e. weeks 0, 2,
6, 10 and 14) were analysed on the same plate in duplicate,
and samples from one control rabbit were paired with
samples from one transgenic rabbit on the same plate.
Measurement of Ab-C
A 96-well microtitre plate was coated with cardiolipin
(100 µl, 10 µg\ml ; Sigma E-5646 ; diphosphatidylglycerol) in solution in 0.1 mol\l PBS and incubated
overnight at 4 mC. After four washings with PBS containing 0.1 % fatty acid-free BSA (Sigma) and 0.05 %
Tween 20, non-specific binding sites were saturated by
addition of 200 µl of 2 % fatty acid-free BSA for 2 h at
room temperature. The plate was washed another four
times with 0.1 mol\l PBS containing 0.05 % Tween 20
and 0.1 % fatty acid-free BSA. Then serum diluted 1 :100
in buffer A (in which the BSA was fatty acid-free) was
incubated for 2 h at room temperature ; the blank
consisted of 100 µl of buffer A. The plate was washed
four times with buffer A, and peroxidase-conjugated goat
anti-(rabbit IgG) (Sigma), diluted 1 : 5000 in buffer A, was
added (100 µl), followed by incubation at 37 mC for 2 h.
Peroxidase-conjugated anti-(rabbit IgG) activity was
measured as described above. Ab-C titres were expressed
in absorbance units. All samples from one animal (weeks
0, 2, 6, 10 and 14) were analysed on the same plate in
duplicate, and samples from one control rabbit were
paired with samples from one transgenic rabbit on the
same plate.
# 2001 The Biochemical Society and the Medical Research Society
345
346
A. Boullier and others
Table 1
Basal levels of lipids and apo A-I in control and human apo A-I transgenic rabbits
All values are meanspS.D. Significance of differences compared with controls : *P
(Mann–Whitney U-test).
0.05, **P
0.01, *** P
0.001
Concentration (g/l)
Parameter
Controls (n l 18)
Transgenics (n l 13)
Total cholesterol
Triacylglycerols
Phospholipids
HDL cholesterol
Non-HDL cholesterol
Human apo A-I
0.57p0.17
0.83p0.20
0.83p0.20
0.22p0.09
0.39p0.28
–
1.08p0.26*
1.50p1.05**
1.85p0.30**
0.70p0.20***
0.38p0.08
1.50p0.31
Measurement of IgG-LDL
Goats were immunized with rabbit LDL, and immunoglobulins (IgG) were isolated by precipitation. IgG-LDL
was measured using a bi-site ELISA method. Microtitre
plates were coated overnight at room temperature with
goat immunoglobulins in solution (10 µg\ml ; 100 µl) in
0.1 mol\l PBS. These immunoglobulins contain antibodies that specifically recognize rabbit apo B that will
capture rabbit LDL and IgG-LDL. Each plate included
negative controls without capture antibodies. After
washing four times with 0.1 mol\l PBS containing 0.05 %
Tween 20, 0.1 % (w\v) BSA and 0.001 % aprotinin, the
solid-phase antibody was saturated with 5 % BSA\
0.1 mol\l PBS for 2 h at room temperature. Plates were
then washed four times with 0.1 mol\l PBS. After
washing, the solid-phase antibody was incubated in
duplicate with antigen (10 µl of serum diluted 1 : 500 in
buffer A). After 2 h of incubation at room temperature,
wells were washed four times with buffer A. Then
peroxidase-conjugated goat anti-(rabbit IgG) diluted
1 : 5000 in buffer A was added (100 µl), and the mixture
was left for 2 h at 37 mC. This antibody does not crossreact with the goat capture antibodies. After washing
four times with 0.1 mol\l PBS, enzymic substrate (ophenylenediamine) was added to develop the reaction as
described above, and absorbance was read at 492 nm. The
IgG-LDL titre was expressed in absorbance units. All
samples from one animal (weeks 0, 2, 6, 10 and 14) were
tested on the same plate, and samples from one control
rabbit were paired with samples from one transgenic rabbit on the same plate. The within-run and
between-run coefficients of variation were 7 % and
9 % respectively.
previously 1 : 10 (v\v) in 10 µmol\l eserine (Sigma) in
1 ml of substrate, and p-nitrophenol formation was
quantified by monitoring the absorbance at 405 nm at
25 mC for 5 min.
Evaluation of atherosclerosis
Animals were killed by an overdose of intravenous
ketamine (80 mg\kg) and carotid exsanguination. The
thoracic aorta from the aortic valves to the diaphragm,
above the coeliac artery, and the abdominal aorta from
the diaphragm to the iliac bifurcation were removed and
stripped of adventitial fat and tissue. Then aortas were cut
lengthwise into right and left halves by dorsal and ventral
incisions, and fixed in 10 % buffered formalin. The fixed
aortas were stained with Oil Red O to reveal fatty
deposits. Morphometric assessment of the percentage of
the total aorta covered by lipid deposits was determined
by computerized planimetry. For light microscopy
observations, segments of the Red O-stained aortas were
embedded in paraffin and then stained with
haematoxylin\eosin.
Statistical analysis
All data are expressed as meanspS.D. Serum data were
evaluated by the Mann–Whitney U-test, to compare each
parameter at every point of the protocol between the two
groups. Changes in each of these parameters, from week
0 through weeks 2, 6, 10 and 14, were tested by the paired
Wilcoxon t test. Differences in atheroma area between
the two groups were evaluated by Student’s t test.
Coefficients of correlation between the different variables
were calculated. P 0.05 was considered as significant.
RESULTS
Measurement of PON-1 activity
Lipoprotein profiles
PON-1 activity was measured using the method of
Mackness et al. [21]. Substrate (paraoxon ; Sigma) was
diluted in Tris buffer (100 mmol\l Tris and 2 mmol\l
CaCl , pH 8) at a concentration of 2 mmol\l. The
#
reaction was initiated by adding 10 µl of serum diluted
The basal (week 0) levels of total cholesterol, triacylglycerols and phospholipids were all significantly higher
in transgenic rabbits than in controls (Table 1). The
increase in total cholesterol corresponded to a 3-fold
higher level of HDL cholesterol (P 0.001), while
# 2001 The Biochemical Society and the Medical Research Society
Oxidative markers in human apolipoprotein A-I transgenic rabbit
(a)
(d)
(b)
(e)
(c)
Figure 1 Serum total cholesterol (a), non-HDL cholesterol (b), HDL cholesterol (c), triacylglycerols (d) and phospholipids (e)
in control ($) and human apo A-I transgenic (>) rabbits fed on a 0.4 % (w/w) cholesterol diet
All values (meanspS.D.) are in g/l. Significance of differences between transgenic rabbits and controls : *P 0.05, **P 0.01, ***P
(Mann–Whitney U-test). Significance of differences compared with week 0 : (o)P 0.05, (oo)P 0.01, (ooo)P 0.001.
Table 2
0.002, ****P
0.0001
Markers of modified LDL
λexc, excitation wavelength ; λem, emission wavelength ; a.u., arbitrary units.
Cu2+-oxidized LDL
MDA-LDL
Marker
Native LDL
Before dialysis
After dialysis
After dialysis
Diene conjugates (nmol/mg of protein)
Hydroperoxides (nmol/mg of protein)
TBARS (nmol/mg of protein)
Fluorescence (λexc 360/λem 430 nm) (a.u.)
Fluorescence (λexc 400/λem 470 nm) (a.u.)
Free NH2 groups (%)
Electrophoretic mobility (mm)
Electrophoretic mobility (mm) cf. native LDL
0.00
0.00
0.00
100
48
95
23
1
743
553
36
290
110
62
–
–
405
508
0.52
210
65
54
30
1.30
12
0.00
194
105
170
23
37
1.60
# 2001 The Biochemical Society and the Medical Research Society
347
348
A. Boullier and others
Table 3
Basal levels of immunological markers of LDL oxidation
All values (meanspS.D.) are expressed as 1000iabsorbance. Significance of difference compared with controls : *P
(Mann–Whitney U-test).
0.05
Level (1000iA)
Marker
Controls (n l 18)
Transgenics (n l 13)
Ab-LDL
Ab-OxLDL
Ab-MDA-LDL
IgG-LDL
Ab-C
158p62
221p125
522p162
427p80
164p71
199p81
290p155
710p309*
495p104
218p100
Figure 2 Serum levels of immunological markers of LDL oxidation in control ($) and human apo A-I transgenic (>) rabbits
fed on a 0.4 % (w/w) cholesterol diet
All values (meanspSD) are in 1000iabsorbance (A). Significance of differences between transgenic rabbits and controls : *P
differences compared with week 0 : (o)P 0.05, (oo)P 0.01, (ooo)P 0.001.
non-HDL cholesterol levels were similar between the
groups. The serum level of human apo A-I in the transgenic rabbits was 1.50p0.31 g\l.
During cholesterol feeding, total cholesterol levels
increased progressively and dramatically in both groups
# 2001 The Biochemical Society and the Medical Research Society
0.05, **P
0.01. Significance of
(Figure 1a), and the significant difference in levels
between the groups had disappeared at weeks 6 and 10,
but not at week 14. This increase in total cholesterol was
mediated by a rise in the non-HDL fraction (Figure 1b),
while HDL cholesterol levels were maintained at their
Oxidative markers in human apolipoprotein A-I transgenic rabbit
Table 4 Correlation coefficients between the different immunological markers of LDL
oxidation at week 0 (a) and at week 14 (b)
Significance : *P
(a) Week 0
0.05 ; **P
0.02 ; ***P
0.001. C, controls ; T, human apo A-I transgenic rabbits.
r
Ab-LDL
Ab-Ox-LDL
Ab-MDA-LDL
IgG-LDL
Ab-C
C
T
C
T
C
T
C
T
C
T
Ab-LDL
Ab-Ox-LDL
Ab-MDA-LDL
IgG-LDL
Ab-C
1
1
0.613***
0.642**
0.617***
0.643**
0.375
0.780***
0.592***
0.887***
1
1
0.627***
0.657**
0.493*
0.499
0.636*
0.631**
1
1
0.775***
0.636**
0.696***
0.680***
1
1
0.559**
0.759***
1
1
Ab-LDL
Ab-Ox-LDL
Ab-MDA-LDL
IgG-LDL
Ab-C
1
1
0.715***
k0.476
0.537**
k0.507
0.168
k0.449
0.471*
0.730***
1
1
0.880***
k0.325
0.493*
k0.298
0.470*
k0.370
1
1
k0.537**
k0.026
k0.396
k0.143
1
1
k0.044
0.640**
1
1
(b) Week 14
r
Ab-LDL
Ab-Ox-LDL
Ab-MDA-LDL
IgG-LDL
Ab-C
C
T
C
T
C
T
C
T
C
T
respective initial levels until week 10 in control rabbits
and week 6 in transgenic rabbits ; after this they fell
significantly (Figure 1c). HDL cholesterol levels
remained significantly higher in transgenic rabbits than in
controls during the entire cholesterol feeding period
(Figure 1c). Phospholipids and triacylglycerols increased
in both groups during cholesterol feeding (Figures 1d and
1e), and remained higher in transgenic than in control
rabbits at week 14.
Immunological markers of LDL oxidation
Chemical characteristics of rabbit LDLs used as antigens
Cu#+-oxidized LDL and MDA-LDL were used as
antigens after extensive dialysis for 24 h at 4 mC to remove
EDTA, metals and water-soluble aldehydes. Dialysed
OxLDL showed signs of lipid peroxidation, such as
increases in diene conjugates and hydroperoxide content
(Table 2). Levels of TBARS were very slightly increased
in dialysed OxLDL, indicating that only very small
amounts of MDA were tightly bound to apo B. On the
other hand, the fluorescence intensity of these lipoproteins was greatly increased, while levels of free NH
#
residues were decreased (k40 %), suggesting that other
lipid peroxidation products besides MDA (e.g. 4hydroxynonenal) were bound to apo B. The decrease in
the free NH residue content was concomitant with an
#
increase in electrophoretic mobility (1.30-fold compared
with that of native LDL). Lipids of dialysed MDA-LDL
were hardly oxidized at all (absence of diene conjugates
and of hydroperoxides), although they contained many
TBARS (MDA) molecules bound to apo B, which
dramatically increased their fluorescence intensity and
their electrophoretic mobility (1.60-fold compared with
native LDL), and concomitantly decreased their content
of free NH residues (k70 %).
#
Serum levels of immunological markers of LDL oxidation
These are reported in Table 3. Before cholesterol feeding
was started (week 0), serum levels of Ab-LDL, AbOxLDL, IgG-LDL and Ab-C were similar in the control
# 2001 The Biochemical Society and the Medical Research Society
349
350
A. Boullier and others
Table 5 Correlation coefficients between plasma lipids and immunological markers of LDL
oxidation at week 0 (a) and week 14 (b)
Significance : *P
(a) Week 0
0.05 ; **P
0.02 ; ***P
0.001. C, controls ; T, human apo A-I transgenic rabbits.
r
HDL cholesterol
C
T
Non-HDL cholesterol C
T
Ab-LDL
Ab-Ox-LDL
Ab-MDA-LDL
IgG-LDL
Ab-C
0.152
0.666***
k0.136
0.347
0.372
0.670***
0.106
0.444
0.041
0.326
k0.298
0.192
0.163
0.517
k0.110
0.142
0.203
0.619**
k0.006
0.264
Ab-LDL
Ab-Ox-LDL
Ab-MDA-LDL
IgG-LDL
Ab-C
k0.495*
k0.179
k0.081
k0.254
k0.105
0.006
k0.282
k0.166
0.158
k0.116
k0.255
0.201
k0.027
k0.392
k0.321
k0.284
0.167
0.042
0.440
k0.295
(b) Week 14
r
HDL cholesterol
C
T
Non-HDL cholesterol C
T
and the transgenic rabbit groups, while the level of AbMDA-LDL was significantly higher in the transgenic
group.
Figure 2 reports the time course of changes in serum
immunological OxLDL markers during cholesterol feeding. Ab-LDL levels were not modified during the entire
period in either group (Figure 2a), while Ab-OxLDL
levels increased significantly from week 10 in the control
group and non-significantly in the transgenic group
(Figure 2b). At all time points there were no significant
differences in serum levels of Ab-LDL or Ab-OxLDL
between control and transgenic rabbits.
Ab-MDA-LDL serum levels decreased significantly
from week 2 in both groups (Figure 2c). These levels were
significantly higher in the transgenic animals than in the
controls until week 10.
IgG-LDL serum levels increased significantly from
week 2 in the control group and from week 6 in the
transgenic group (Figure 2d), but there was never any
significant difference between the two groups.
Ab-C serum levels increased significantly from week 2
in controls and from week 6 in transgenic rabbits (Figure
2e). These levels were higher in the transgenic group than
in the control group, but the differences were not
significant.
Before cholesterol feeding was started (week 0), all
immunological OxLDL markers were correlated positively, except for IgG-LDL with Ab-LDL in the control
group and with Ab-OxLDL in the transgenic group
(Table 4a). At the end of the experiment (week 14), AbC levels were no longer correlated with Ab-MDA-LDL
or IgG-LDL levels in the control group (Table 4b). In the
transgenic group there was no longer any correlation
# 2001 The Biochemical Society and the Medical Research Society
between the immunological markers of LDL oxidation,
with the exception of positive correlations of Ab-C with
Ab-LDL and IgG-LDL (Table 4b).
Ab-LDL, Ab-OxLDL and Ab-C levels were highly
positively correlated with HDL cholesterol at week 0 in
the human apo A-I transgenic rabbit group (Table 5a).
There was a negative correlation at week 14 between AbLDL and HDL cholesterol in the control group (Table
5b).
Changes in serum PON-1 activity during
cholesterol feeding, and relationships with
HDL cholesterol and immunological OxLDL
markers
Before cholesterol feeding was started (week 0), serum
PON-1 activity was higher in the transgenic rabbits
than in the controls (3608p912 and 3097p
653 nmol:min−":ml−"), but, due to the high S.D.s,
the difference was not significant (P l 0.09) (Figure 3).
During cholesterol feeding, PON-1 activity decreased
progressively in both groups ; differences between the
groups remained non-significant. After 14 weeks on the
diet, PON-1 activity was decreased by 43p16 % and
50p11 % in control and transgenic rabbits respectively.
At weeks 0 and 14, PON-1 activity was significantly
and positively correlated with the HDL cholesterol level
in control rabbits, but not in transgenic animals (Table 6).
At week 0, PON-1 activity was positively correlated with
Ab-MDA-LDL in the transgenic group, while at week 14
there was no correlation between any immunological
marker of LDL oxidation and PON-1 activity in either
group (Table 6).
Oxidative markers in human apolipoprotein A-I transgenic rabbit
Relationship between immunological OxLDL
markers, lipoprotein concentrations, PON-1
activity and the development of
atherosclerosis
After 14 weeks of cholesterol feeding, the percentage area
of the aorta covered by lipid deposits was similar in
control and transgenic rabbits (23p14 % and 26p11 %
respectively ; P l 0.52). Table 7 shows that these surface
areas were not significantly correlated with the serum
levels of cholesterol, nor with the immunological markers
of LDL oxidation or PON activity.
Figure 3 Serum PON-1 activity in control ($) and human
apo A-I transgenic (>) rabbits fed on a 0.4 % (w/w) cholesterol diet
There were no significant differences between the transgenic and control groups.
Significance of differences compared with week 0 : (oo)P 0.01.
DISCUSSION
The present study does not confirm our preliminary
results showing that human apo A-I transgenic rabbits
(line 20) developed less atheroma when they were fed on
Table 6 Correlation coefficients between serum PON-1 activity and immunological markers of LDL oxidation or serum
cholesterol before and after 14 weeks of a 0.4 % cholesterol diet in rabbits
Significance compared with controls : *P
0.05, **P
0.01 (Mann–Whitney U-test).
Correlation with PON-1 activity (r)
Week 0
Week 14
Parameter
Control (n l 18)
Transgenics (n l 13)
Controls (n l 18)
Transgenics (n l 13)
Ab-LDL
Ab-OxLDL
Ab-MDA-LDL
IgG-LDL
Ab-C
HDL cholesterol
Non-HDL cholesterol
k0.114
0.115
0.125
0.208
k0.151
0.518*
0.293
0.436
0.413
0.576*
0.430
0.501
0.316
0.175
k0.382
k0.036
0.142
0.227
0.161
0.749**
0.447
0.111
0.217
0.476
k0.032
0.145
0.549
0.308
Table 7 Correlation coefficients between atherosclerosis area in the thoracic aorta and
immunological markers of LDL oxidation, serum cholesterol and serum PON-1 activity
There were no significant correlations.
Correlation with atherosclerosis area (r)
Ab-LDL
Ab-OxLDL
Ab-MDA-LDL
IgG-LDL
Ab-C
HDL cholesterol
Non-HDL cholesterol
PON-1 activity
Controls (n l 18)
Transgenics (n l 13)
0.458
0.167
0.148
0.424
0.090
0.447
0.200
0.470
0.288
0.032
0.435
0.216
0.032
0.063
0.449
0.421
# 2001 The Biochemical Society and the Medical Research Society
351
352
A. Boullier and others
a cholesterol-rich diet [19], but confirms our data from
human apo A-I transgenic rabbits of line 8 [31] showing
that these animals were not protected against atherosclerosis when they were fed a cholesterol-rich diet.
However, there were some differences between the
experimental protocols of these studies. In the earliest
study [19], transgenic rabbits belonged to line 20 of
transgenesis, and expressed five copies of the human apo
A-I transgene ; in the other studies they belonged to line
8, and expressed only two copies of this transgene.
Nevertheless, basal HDL cholesterol and human apo A-I
serum levels were similar in the transgenic groups of the
two lines. One major difference between the three studies
is that, in the earliest study [19], transgenic rabbits were
given a 0.3 % (w\w) cholesterol diet after an initial 2
weeks of supplementation with a 0.4 % (w\w) cholesterol
diet, while control rabbits remained on a diet
supplemented with 0.4 % (w\w) cholesterol during the
entire 14 weeks of treatment. In the present study, and in
the study by Mackness et al. [31], control and transgenic
rabbits were fed a 0.4 % (w\w) cholesterol diet during the
entire 14 weeks of the protocol. In the earliest study [19],
the goal of this adjustment was to produce similar levels
of non-HDL cholesterol in the two groups of rabbits.
These non-HDL cholesterol levels did not actually differ
between the two groups during the 14 weeks of treatment,
whereas in the present study non-HDL cholesterol levels
were significantly higher in the transgenic group than in
the control group at week 14.
As oxidation of LDL is the primary step in the
development of atherosclerosis [1] one goal of this study
was to determine the evolution in the serum concentrations of immunological markers of LDL oxidation
during a cholesterol feeding period that induces atheroma
formation in control and human apo A-I transgenic
rabbits. Ab-MDA-LDL serum levels were significantly
higher in apo A-I transgenic rabbits during the entire
protocol duration, even at week 0 (but with the exception
of week 14). At this time, the other immunological
markers of LDL oxidation had a tendency, but without
reaching significance, to be at higher levels in transgenic
than in control rabbits.
At week 0, animals were on a normolipidic diet, and
levels of non-HDL cholesterol in serum were similar in
the control and the apo A-I transgenic groups, while
triacylglycerol and phospholipid concentrations were
higher in the transgenic rabbits. This increase in phospholipid concentration was linked largely to the 3-fold higher
level of HDL in transgenic animals, but the increase
in triacylglycerols was due partly to increases in HDL
triacylglycerol and in non-HDL triacylglycerol (results
not shown).
The LDL of patients with diabetes, who have higher
serum triacylglycerol and phospholipid concentrations,
is more susceptible to lipid oxidation than the LDL of
control patients [38]. Therefore, by analogy with the
# 2001 The Biochemical Society and the Medical Research Society
human diabetic condition, we speculate that the degree of
LDL oxidation in vivo is higher in apo A-I transgenic
rabbits than in controls because of the higher triacylglycerol and phospholipid plasma concentrations. This
would lead to the formation of more oxidized immunogenic LDL and to the production of more Ab-MDALDL.
We have already shown that endothelial-derived vasorelaxation is impaired in human apo A-I transgenic
rabbits fed on a normal chow diet [40], and we have
suggested that increases in triacylglycerols and phospholipids in the lipoproteins of these animals would exacerbate lipoprotein oxidation, which would in turn
inhibit endothelial-dependent vasorelaxation [41–49].
Increases in LDL oxidation might not be the only cause
of lipid oxidation in human apo A-I transgenic rabbits
fed a standard chow diet (and of the resulting increase in
Ab-MDA-LDL plasma levels), because the possibility
cannot be excluded that high plasma levels of HDL
cholesterol (and HDL phospholipid) may induce
HDL oxidation [50] and the synthesis of autoantibodies
against oxidized HDL that would also recognize
OxLDL. The strong positive correlation between HDL
cholesterol and Ab-LDL, Ab-OxLDL and Ab-C at week
0 in the human apo A-I transgenic rabbit group supports the hypothesis that high plasma levels of HDL
cholesterol would stimulate the immunological response
against lipids.
We made the unexpected observation that Ab-MDALDL titres decreased progressively in both groups of
animals during the cholesterol feeding period, while the
titres of other immunological markers increased. AbMDA-LDL titres have been reported to be higher in
patients suffering from carotid atherosclerosis [3,51] and
in hypercholesterolaemic apo E-deficient mice which
develop atherosclerosis [52]. In order to explain our
observation, we speculated on the heterogeneity of antiOxLDL autoantibodies. Mironova et al. [4] have isolated
human anti-OxLDL autoantibodies. All purified IgG
antibodies showed some cross-reactivity with MDALDL, native LDL and cardiolipin, and had moderate-tolow affinity for OxLDL. We hypothesize that, in the
present study, Ab-MDA-LDL had high affinity, and AboxLDL and Ab-C had low affinity, for in vivo OxLDL.
We speculate that MDA-LDL is present at high concentrations in rabbits during the cholesterol feeding period
and that, due to their high affinity, Ab-MDA-LDL bind
to the MDA-LDL to constitute LDL immune complexes.
This mechanism would explain why IgG-LDL concentrations increased progressively while Ab-MDA-LDL
levels decreased progressively during the cholesterol
feeding period. Because of their putative low affinity for
OxLDL, Ab-OxLDL and Ab-C would not bind to the
circulating OxLDL and would remain free in the serum
and not associated with IgG-LDL, which accounts for
the standard increases reported in the serum concen-
Oxidative markers in human apolipoprotein A-I transgenic rabbit
trations of Ab-OxLDL and Ab-C during the development
of atherosclerosis. Recently, it has been reported that
autoantibodies against MDA-LDL are reduced after
acute feeding in patients with coronary artery disease
[53]. This suggests that binding of MDA-LDL epitopes
to circulating Ab-MDA-LDL would account for the
observed postprandial decrease in Ab-MDA-LDL.
We have shown in the present study that increases in
Ab-C levels occur before increases in Ab-OxLDL. We
may therefore conclude that, in human patients free of
anti-phospholipid syndrome, it would be more sensible
to measure Ab-C rather than anti-OxLDL titres in
order to evaluate atheroma development [10,54].
Changes in Ab-OxLDL, Ab-C and IgG-LDL concentrations during the cholesterol feeding period showed no
differences between apo A-I transgenic and control
rabbits, in accordance with the identical development of
atherosclerosis in the two groups. Furthermore, there
was no correlation between serum levels of these markers
of LDL oxidation and atherogenic lipoprotein concentrations (non-HDL cholesterol) in either group of
rabbits, as has been reported previously in humans
[54,55].
There was no correlation between titres of immunological markers of LDL oxidation and the extension of
aortic atheroma in rabbits. These data are in contrast with
the report of a correlation between the development of
atherosclerosis and titres of autoantibodies against
OxLDL in humans [3,54]. Nevertheless, the putative
correlation between plasma levels of immunological
markers of LDL oxidation and the development of
atherosclerosis in humans is highly controversial [55].
Duverger et al. [18,19] reported that the reduced
development of atherosclerosis in human apo A-I transgenic rabbits was due to their high HDL cholesterol
levels and to a subsequent increase in reverse cholesterol
transport. However, it has been proposed that an increase
in reverse cholesterol transport is not the only mechanism
of inhibition of atheroma development induced by HDL,
because HDL also inhibits the oxidation of LDL
[20–22,28,30]. PON-1 is located in HDL, and may be the
component of HDL responsible for decreasing the
accumulation of lipid peroxidation products in LDL [27].
As PON-1 is linked exclusively to HDL, its activity
tended to be higher (without the difference reaching
significance) in the apo A-I transgenic rabbits, which had
a 3-fold higher level of HDL cholesterol than the
controls. (In a subsequent protocol carried out with 99
controls and 92 transgenic rabbits, PON-1 activity was
significantly higher in the transgenic animals before
starting cholesterol feeding [31].) In the present study
there were positive correlations between HDL cholesterol levels and PON-1 activity in control rabbits at
week 0 and at week 14. During the cholesterol feeding
period, PON-1 activity decreased progressively and
similarly in both groups (k43p16 % and k50p11 % at
week 14 in comparison with week 0 in control and
transgenic rabbits respectively). During this period the
decrease in PON-1 activity was correlated with
the decrease in HDL cholesterol levels. There was no
negative correlation between serum levels of immunological markers of LDL oxidation and PON-1 activity in
either group before starting cholesterol feeding (week 0)
or after 14 weeks of treatment. We may speculate that a
negative correlation would have been apparent between
PON-1 activity and in vivo markers of LDL oxidation
if PON-1 had been effective in reducing LDL oxidation in vivo. In the present study we show that PON-1
activity decreased similarly in cholesterol-fed control and
human apo A-I transgenic rabbits, and that increases in
serum immunological markers of LDL oxidation and the
development of atherosclerosis were similar in the two
groups. These data suggest that, under these experimental
conditions, high serum HDL cholesterol levels are
ineffective in inhibiting the development of atherosclerosis in human apo A-I transgenic rabbits when compared
with control rabbits, perhaps because their associated
PON-1 activity does not induce greater inhibition of
LDL oxidation than does the HDL-associated PON-1
activity in control rabbits.
The high-cholesterol-fed rabbit is probably not a
convenient model in which to study the mechanisms of
atheroprotection in relation to the antioxidative properties of HDL, because of the extravagantly high levels of
atherogenic lipoproteins present in the serum of these
animals in comparison with the levels of anti-atherogenic
lipoproteins (control rabbits after 14 weeks of cholesterol
feeding : non-HDL cholesterol, 17.27 g\l ; HDL cholesterol, 0.14 g\l). In this case the inhibition of LDL
oxidation by HDL and the associated PON-1 activity is
a marginal phenomenon ; however, this does not exclude
its possible relevance in humans, where LDL cholesterol
levels are often below 1.90 g\l and HDL cholesterol levels are over 0.30 g\l. In normolipidaemic
patients, the molar ratio between HDL and LDL could
be instrumental in the inhibition of LDL oxidation by
HDL. Nevertheless, our study suggests that, in patients
with severe heterozygous or homozygous familial hypercholesterolaemia, as well as in those with severe mixed
dyslipidaemia, an isolated increase in HDL cholesterol
levels would not be effective in inhibiting the development of atherosclerosis, since the lack of a significant
increase in PON-1 activity means that there would be no
corresponding inhibition of LDL oxidation.
ACKNOWLEDGMENTS
We thank Alexandra Tavernier-Sommerville for linguistic assistance. This study was supported by the BIO\
AVENIR programme of Rho# ne-Poulenc Rorer SA.
# 2001 The Biochemical Society and the Medical Research Society
353
354
A. Boullier and others
REFERENCES
1 Chisolm, G. M. and Penn, M. S. (1996) Oxidized
lipoproteins and atherosclerosis. In Atherosclerosis and
Coronary Artery Disease (Fuster, V., Ross, R. and Topol,
E. J., eds.), pp. 129–149, Lippincott-Raven Publishers,
Philadelphia
2 Palinski, W., Yla$ -Herttuala, S., Rosenfeld, M. E. et al.
(1990) Antisera and monoclonal antibodies specific for
epitopes generated during oxidative modification of low
density lipoprotein. Arteriosclerosis 10, 325–333
3 Salonen, J. T., Yla$ -Herttuala, S., Yamamoto, R. et al.
(1992) Autoantibodies against oxidised LDL and
progression of carotid atherosclerosis. Lancet 339, 883–887
4 Mironova, M., Virella, G. and Lopes-Virella, M. F. (1996)
Isolation and characterization of human antioxidized LDL
autoantibodies. Arterioscler. Thromb. Vasc. Biol. 16,
222–229
5 Palinski, W., Miller, E. and Witztum, J. L. (1995)
Immunization of low density lipoprotein (LDL) receptordeficient rabbits with homologous malondialdehydemodified LDL reduces atherogenesis. Proc. Natl. Acad.
Sci. U.S.A. 92, 821–855
6 Yla-Herttua$ la, S., Palinski, W., Butler, S. W., Picard, S.,
Steinberg, D. and Witztum, J. L. (1994) Rabbit and human
atherosclerotic lesions contain IgG that recognize epitopes
of oxidized LDL. Arterioscler. Thromb. Vasc. Biol. 14,
32–40
7 Beaumont, J. L., Douced, F., Vivier, P. and Antonucci, M.
(1998) Immunoglobulin-bound lipoproteins (Ig-Lp) as
markers of familial hypercholesterolemia, xanthomatosis
and atherosclerosis. Atherosclerosis 74, 191–201
8 Tertov, V. V., Orekhov, A. N., Sayadyan, K. S. et al.
(1990) Correlation between cholesterol content in
circulating immune complexes and atherogenic properties
of CHD patients’ serum manifested in cell culture.
Atherosclerosis 81, 183–189
9 Tertov, V. V., Sobenin, I. A., Orekhov, A. N., Jaakkola,
O., Solakivi, T. and Nikkari, T. (1996) Characteristics of
low density lipoprotein isolated from circulating immune
complexes. Atherosclerosis 122, 191–199
10 Vaarala, O., Ma$ tta$ ri, M., Manninen, V. et al. (1995) Anticardiolipin antibodies and risk of myocardial infarction in
a prospective cohort of middle-aged men. Circulation 91,
23–27
11 Vaarala, O., Alfthan, G., Jauhiainen, M., Leirisalo-Repo,
M., Aho, K. and Palosuo, T. (1993) Crossreaction between
antibodies to oxidised low-density lipoprotein and to
cardiolipin in systemic lupus erythematosus. Lancet 341,
923–925
12 Vaarala, O., Puurunen, M., Lukka, M. et al. (1996)
Affinity-purified cardiolipin-binding antibodies show
heterogeneity in their binding to oxidized low-density
lipoprotein. Clin. Exp. Immunol. 104, 269–274
13 Ho$ rkko$ , S., Miller, E., Dudl, E. et al. (1996)
Antiphospholipid antibodies are directed against epitopes
of oxidized phospholipids. Recognition of cardiolipin by
monoclonal antibodies to epitopes of oxidized low density
lipoprotein. J. Clin. Invest. 98, 815–825
14 Gordon, D. J., Knoke, J., Probsfield, J. L., Superko, R. and
Tyroler, H. A. (1996) High-density lipoprotein cholesterol
and coronary heart disease in hypercholesterolemic men :
the Lipid Research Clinics Coronary Primary Prevention
Trial. Circulation 74, 1217–1225
15 Rubin, E. M., Krauss, R. M., Spangier, E. A., Verstuyft,
J. G. and Gift, S. M. (1991) Inhibition of early
atherogenesis in transgenic mice by human apolipoprotein
A-I. Nature (London) 353, 265–267
16 Chapman, J. M. (1986) Comparative analysis of
mammalian plasma lipoproteins. Methods Enzymol. 128,
71–147
17 Hunt, C. E. and Duncan, L. A. (1985)
Hyperlipoproteinemia and atherosclerosis in rabbits fed
low-level cholesterol and lecithin. Br. J. Exp. Pathol. 66,
35–46
18 Duverger, N., Viglietta, C., Berthou, L. et al. (1996)
Transgenic rabbits expressing human apolipoprotein A-I in
the liver. Arterioscler. Thromb. Vasc. Biol. 16, 1424–1429
# 2001 The Biochemical Society and the Medical Research Society
19 Duverger, N., Kruth, H., Emmanuel, F. et al. (1996)
Inhibition of atherosclerosis development in cholesterolfed human apolipoprotein A-I-transgenic rabbits.
Circulation 94, 713–717
20 Ohta, T., Takata, S., Morino, Y. and Matsuda, I. (1989)
Protective effect of lipoproteins containing apolipoprotein
A-I on Cu2j-catalyzed oxidation of human low density
lipoprotein. FEBS Lett. 257, 435–442
21 Mackness, M. I., Arrol, S. and Durrington, P. N. (1991)
Paraoxonase prevents accumulation of lipoperoxides in
low-density lipoprotein. FEBS Lett. 286, 152–154
22 Tribble, D. L., Chu, B. M., Gong, E. L., Van Venrooij, F.
and Nichols, A. V. (1995) HDL antioxidant effects as
assessed using a nonexchangeable probe to monitor
particle-specific peroxidative stress in LDL-HDL
mixtures. J. Lipid Res. 36, 2580–2589
23 La Du, B. N. (1992) Human serum
paraoxonase\arylesterase. In Pharmacogenetics of Drug
Metabolism (Halow, W., ed.), pp. 51–91, Pergamon Press
Inc., New York
24 Mackness, M. I., Harty, D., Bhatnagar, D. et al. (1991)
Serum paraoxonase activity in familial
hypercholesterolaemia and insulin-dependent diabetes
mellitus. Atherosclerosis 86, 193–199
25 Mackness, M. I., Arrol, S., Abbott, C. A. and Durrington,
P. N. (1993) Is paraoxonase related to atherosclerosis ?
Chem. Biol. Interact. 87, 167–171
26 Mackness, M. I. (1989) Possible medical significance of
human serum ‘ A ’ esterases. In Enzymes Hydrolyzing
Organophosphorus Compounds (Reiner, E., Aldridge,
W. N. and Hoshin, F. C. G., eds.), pp. 202–213, EllisHorwood, Chichester
27 Watson, A. D., Berliner, J. A., Hama, S. Y. et al. (1995)
Protective effect of high density lipoprotein associated
paraoxonase. Inhibition of the biological activity of
minimally oxidized low density lipoprotein. J. Clin. Invest.
96, 2882–2891
28 Shih, D. M., Gu, L., Xia, Y. R. et al. (1998) Mice lacking
serum paraoxonase are susceptible to organophosphate
toxicity and atherosclerosis. Nature (London) 394,
284–287
29 Shih, D. M., Xia, Y. R., Wang, X. P. et al. (2000)
Combined serum paraoxonase knockout\apolipoprotein E
knockout mice exhibit increased lipoprotein oxidation and
atherosclerosis. J. Biol. Chem. 275, 17527–17535
30 Hayek, T., Oiknine, J., Dankner, G., Brook, J. G. and
Aviram, M. (1995) HDL apolipoprotein A-I attenuates
oxidative modification of low density lipoprotein : studies
in transgenic mice. Eur. J. Clin. Chem. Clin. Biochem. 33,
721–725
31 Mackness, M., Boullier, A., Hennuyer, N. et al. (2000)
Paraoxonase activity is reduced by a pro-atherogenic diet
in rabbits. Biochem. Biophys. Res. Commun. 269, 232–236
32 NIH (1985) Guide for the Care and Use of Laboratory
Animals, DHEW Publication no. (NIH) 85–23, Office of
the Science and Health Reports, DRR\NIH, Bethesda,
MD
33 Havel, R. J., Eder, H. A. and Bragdon, J. H. (1955) The
distribution and chemical composition of ultracentrifugally
separated lipoproteins in human serum. J. Clin. Invest. 34,
1345–1353
34 Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall,
R. J. (1951) Protein measurement with the folin phenol
reagent. J. Biol. Chem. 193, 265–275
35 Esterbauer, H., Dieber-Rotheneder, M., Waeg, G., Striegl,
G. and Ju$ rgens, G. (1990) Biochemical, structural, and
functional properties of oxidized low density lipoprotein.
Chem. Res. Toxicol. 3, 77–92
36 Ohkawa, H., Ohishi, N. and Yagi, K. (1979) Assay for
lipid peroxides in animal tissues by TBA reaction. Anal.
Biochem. 95, 351–358
37 El-Saadani, M., Esterbauer, H., El-Sayed, M., Goher, M.,
Nassar, A. Y. and Ju$ rgens, G. (1989) A spectrophotometric
assay for lipid peroxides in serum lipoproteins using a
commercially available reagent. J. Lipid Res. 30, 627–630
38 Esterbauer, H., Gebicki, J., Phal, H. and Ju$ rgens, G. (1992)
The role of lipid peroxidation and antioxidants in oxidative
modification of low density lipoprotein. Free Radical Biol.
Med. 13, 341–390
Oxidative markers in human apolipoprotein A-I transgenic rabbit
39 Steinbrecher, U. P., Witzum, J. L., Parthasarathy, S. and
Steinberg, D. (1987) Decrease in reactive amino groups
during oxidation or endothelial cell modification of low
density lipoprotein : correlation with changes in receptor
mediated catabolism. Atherosclerosis 7, 135–143
40 Lebuffe, G., Boullier, A., Tailleux, A. et al. (1997)
Endothelial derived vasorelaxation is impaired in human
apo A-I transgenic rabbits. Biochem. Biophys. Res.
Commun. 241, 205–211
41 Creager, M. A., Cooke, J. P., Mendelsohn, M. E. et al.
(1990) Impaired vasodilation of forearm resistance vessels
in hypercholesterolemic humans. J. Clin. Invest. 86,
228–234
42 Rabini, R. A., Fumelli, P., Galassi, R. et al. (1994)
Increased susceptibility to lipid oxidation of low-density
lipoproteins and erythrocyte membranes from diabetic
patients. Metab. Clin. Exp. 12, 1470–1474
43 Sorensen, K. E., Celermajer, D. S., Georgakopoulos, J. R.,
Hatcher, G., Betteridge, D. J. and Deanfield, J. E. (1994)
Impairment of endothelium-dependent dilatation is an
early event in children with familial hypercholesterolemia
and is related to the lipoprotein (a) level. J. Clin. Invest.
93, 50–55
44 Kusterer, K., Pohl, T., Fortmeyer, H. P. et al. (1999)
Chronic selective hypertriglyceridemia impairs
endothelium dependent vasodilatation in rats. Cardiovasc.
Res. 42, 783–793
45 Lewis, T. V., Dart, A. M. and Chin-Dusting, J. P. (1999)
Endothelium-dependent relaxation by acetylcholine is
impaired in hypertriglyceridemic humans with normal
levels of plasma LDL cholesterol. J. Am. Coll. Cardiol. 33,
805–812
46 Cooke, J. P., Singer, A. H., Tsao, P., Zera, P., Rowan, R.
A. and Billigham, M. E. (1992) Antiatherogenic effects of
L-arginine in the hypercholesterolemic rabbits. J. Clin.
Invest. 90, 1168–1172
47 Wang, B. Y., Singer, A. H., Tsao, P. S., Drexler, H., Kosek,
J. and Cooke, J. P. (1994) Dietary arginine prevents
atherogenesis in the coronary artery of the
hypercholesterolemic rabbit. J. Am. Coll. Cardiol. 23,
452–458
48 Kugiyama, K., Kerns, S. A., Morrisett, J. D., Roberts, R.
and Henry, P. D. (1990) Impairment of endotheliumdependent arterial relaxation by lysolecithine modifiedlow-density lipoproteins. Nature (London) 344,
160–162
49 Yokoyama, M., Hirata, K. I., Miyake, R., Akita, H.,
Ishikawa, Y. and Fukuzaki, H. (1990)
Lysophosphatidylcholine : essential role in the inhibition
of endothelium-dependent vasorelaxation by oxidized low
density lipoprotein. Biochem. Biophys. Res. Commun.
168, 301–308
50 Nakajima, T., Origuchi, N., Matsunaga, T. et al. (2000)
Localization of oxidized HDL in atheromatous plaques
and oxidized HDL binding sites on human aortic
endothelial cells. Ann. Clin. Biochem. 37, 179–186
51 Holvoet, P., Perez, G., Zhao, Z., Brouwers, E., Bernar, H.
and Collen, D. (1995) Malondialdehyde-modified low
density lipoproteins in patients with atherosclerotic
disease. J Clin. Invest. 95, 2611–2619
52 Palinski, W., Ord, V. A., Plump, A. S., Breslow, J. L.,
Steinberg, D. and Witztum, J. L. (1994) Apo E-deficient
mice are a model of lipoprotein oxidation in atherogenesis. Demonstration of oxidation-specific epitopes in
lesions and high titers of autoantibodies to malondialdehyde-lysine in serum. Arterioscler. Thromb. 14,
605–616
53 Le, N. A., Kyung, S. and Brown, W. V. (2000) Evidence
for the in vivo generation of oxidatively modified epitopes
in patients with atherosclerotic endothelium. Metab. Clin.
Exp. 49, 1271–1277
54 Erikka, A. T., Narvanen, O., Lehto, S., Uusitupa, M. I. and
Yla-Herttuala, S. (2000) Autoantibodies against oxidized
low-density lipoprotein and cardiolipin in patients with
coronary heart disease. Arterioscler. Thromb. Vasc. Biol.
20, 204–209
55 Boullier, A., Hamon, M., Walters-Laporte, E. et al. (1995)
Detection of autoantibodies against oxidized low-density
lipoproteins and IgG-bound low density lipoproteins in
patients with coronary artery disease. Clin. Chim. Acta
30, 1–10
Received 3 August 2000/16 October 2000; accepted 11 December 2000
# 2001 The Biochemical Society and the Medical Research Society
355