Free Radical Biology & Medicine, Vol. 37, No. 4, pp.

557 570, 2004

Copyright D 2004 Elsevier Inc.
Printed in the USA. All rights reserved
0891-5849/$-see front matter

doi:10.1016/j.freeradbiomed.2004.04.035

Original Contribution
A COMPREHENSIVE STUDY OF OXIDATIVE STRESS AND ANTIOXIDANT
STATUS IN PREECLAMPSIA AND NORMAL PREGNANCY

ELISA LLURBA,*,y EDUARD GRATACos, y PILAR MARTN-GALLAN,* LLUIS CABERO,y and CARMEN DOMINGUEZ *
*Biochemistry and Molecular Biology Center and y Fetal Medicine Unit, Department of Obstetrics,
Hospital Universitari Vall dHebron, 08035 Barcelona, Spain

(Received 19 February 2004; Revised 26 April 2004; Accepted 28 April 2004)

Available online 31 May 2004

AbstractOxidative stress has been increasingly postulated as a major contributor to endothelial dysfunction in
preeclampsia (PE), although evidence supporting this hypothesis remains inconsistent. This study aimed to analyze in
depth the potential role of oxidative stress as a mechanism underlying endothelial damage in PE and the pregnant
womans susceptibility to the disease. To this end, indicative markers of lipoperoxidation and protein oxidation and
changes in antioxidant defense systems were measured in blood samples from 53 women with PE and 30 healthy
pregnant controls. Results, analyzed in relation to disease severity, showed PE women, compared with women with
normal pregnancy, to have: (1) significantly enhanced antioxidant enzyme SOD and GPx activities in erythrocytes; (2)
similar plasma a-tocopherol levels and significantly increased a-tocopherol/lipids in both mild and severe PE; (3)
significantly decreased plasma vitamin C and protein thiol levels; (4) similar erythrocyte glutathione content, total plasma
antioxidant capacity, and whole plasma oxidizability values; (5) significantly elevated plasma total lipid hydroperoxides,
the major initial reaction products of lipid peroxidation, in severe PE; (6) no intracellular or extracellular increases in any
of the secondary end-products of lipid peroxidation, malondialdehyde or lipoperoxides; (7) similar oxidative damage to
proteins quantified by plasma carbonyl levels, immunoblot analysis, and advanced oxidation protein products
assessment; and (8) significantly elevated and severity-related soluble vascular cell adhesion molecule-1 serum levels
reflecting endothelial dysfunction. No correlations were found among plasma levels of circulating adhesion molecules
with regard to lipid and protein oxidation markers. Globally, these data reflect mild oxidative stress in blood of
preeclamptic women, as oxidative processes seem to be counteracted by the physiologic activation of antioxidant
enzymes and by the high plasma vitamin E levels that would prevent further oxidative damage. These results do not
permit us to conclude that oxidative stress might be a pathogenetically relevant process causally contributing to the
disease. D 2004 Elsevier Inc. All rights reserved.

KeywordsAntioxidants, Endothelial dysfunction, Glutathione peroxidase, Lipid peroxidation, Malondialdehyde,

Oxidative stress, Preeclampsia, Protein oxidation, Superoxide dismutase, Vitamin C, Vitamin E, Free radicals

INTRODUCTION maternal death in developed countries and a major

Preeclampsia (PE), a multisystem, multifactorial disease
targeted by endothelial damage which precedes the
clinical diagnosis, is estimated to affect 2% of all
pregnancies [1]. Despite being the leading cause of
Address correspondence to: Carmen Dominguez, Centre dInves-
tigacions Bioquimica & B. Molecular, Hospital Materno-Infantil
Vall dHebron (Pl-14), 08035 Barcelona, Spain; Fax: +3493 4894064;
E-mail: mcdominguez@vhebron.net.
contributor to maternal and perinatal morbidity, the
etiology and mechanisms responsible for the pathogen-
esis of PE remain to be elucidated. The placenta plays a
major role in the pathogenesis of PE, characterized by
abnormal placentation and reduced placental perfusion
[2]. Placental ischemia is thought to lead to widespread
activation and dysfunction of the maternal vascular
endothelium, which result in enhanced formation of
endothelin and thromboxane, increased vascular sensi-
tivity to angiotensin II, and decreased formation of

557
558 E. Llurba et al.
vasodilators such as nitric oxide and prostacyclin [3].
Thus, the endothelial dysfunction underlying the clini-
cal syndrome seems to be the consequence of a pro-
found imbalance in the physiological adaptations to
pregnancy.
Impaired placentation is a feature of PE, but alone
does not suffice to explain the endothelial dysfunction
and maternal syndrome, which points to an interaction
between reduced perfusion and genetically determined
maternal constitutional factors further modified by preg-
nancy-specific changes [4]. This concept has lent coher-
ence to the notion that multiple possible pathways exist
in the preclinical phase of the disease, which might
account for the clinical variability observed in PE pre-
sentation. Several hypotheses invoke oxidative stress as a
cellular process contributing to endothelial dysfunction
in PE and a plausible convergence point for interaction of
the fetoplacental unit and maternal predisposing factors
involved in the disorder [4,5]. Oxidative stress, resulting
from the imbalance between increased reactive oxygen
species (ROS) formation and defects in antioxidant
defense mechanisms, has been implicated in the patho-
genesis of many vascular disorders, including atheroscle-
rosis, diabetes, and hypertension [6,7]. Cells of the
human body are continually attacked by ROS which
arise as natural by-products of normal cellular energy
production or are generated in large amounts in special
situations. ROS are involved in a variety of physiologic
and pathophysiologic processes in which increased oxi-
dative stress may play an important role in disease
mechanisms; however, augmented oxidative stress may
result from the pathologic process [6].
A state of oxidative stress could be generated in
placenta, perhaps through a variant of the hypoxia
reperfusion phenomenon in the developing intervillus
spaces [5,8]. There is indirect evidence of increased
ROS production in placental tissue of preeclamptic
women [9] and direct evidence of local oxidative stress
in which induction of xanthine oxidase activity and
increased formation of nitrotyrosines have been demon-
strated [10,11]. Moreover, antioxidant activities of super-
oxide dismutase and glutathione peroxidase and tissue
levels of vitamin E are lower in preeclamptic placenta
compared with normal pregnancies [12].
The hypothesis that free radical generation contributes
to endothelial dysfunction in PE is supported by studies
showing an increase in circulating lipid peroxides and a
decrease in antioxidant patterns in PE patients [13 18].
However, other studies failed to provide evidence of
elevated circulating secondary products of lipid perox-
idation in PE [19 21]. Recently, determination of the
major urinary isoprostane, 8,12iso-iPF2a-VI, using highly
specific and sensitive methods for the quantitation of
lipid peroxidation in vivo, showed no differences be-
tween severe PE patients and healthy pregnant women
[22]. Despite evidence supporting peroxidation processes
in preeclamptic placentas [9 12,23], conclusive evi-
dence of systemic oxidative stress in PE is lacking. A
clear demonstration of systemic oxidative damage asso-
ciated with a decrease in plasmatic antioxidants is of
clinical relevance and has formed the basis of preventive
strategies currently under evaluation that are based on
antioxidant supplements [24,25].
Dyslipidemia is common in PE and, via oxidation
of susceptible lipids, may contribute to endothelial
activation; we previously reported that triglyceride
and free fatty acids were already elevated in the first
and second trimesters in women who subsequently
developed PE [26]. Moreover, other studies demon-
strated that there is a predominance of the atherogenic
small low-density lipoproteins (LDL) and that vascular
cell adhesion molecules are increased in association
with hyperlipidemia in preeclampsia [27,28]. Some
questions arise: does increased triglyceride and choles-
terol content in plasma of preeclamptic women imply
greater susceptibility to oxidation, is it related to
antioxidant activity or the presence of other additional
factors, and if it occurs, is it reflected in such a
complex environment as plasma with high concentra-
tions of antioxidants.
For this study, our interest focused on the potential
role of oxidative stress in the pathophysiology of PE,
because the mechanisms underlying the pregnant wom-
ans susceptibility to this endothelial pathology are
unknown. To this end, we assessed a wide range of
indicative parameters of lipid peroxidation and protein
oxidation together with changes in the status of antiox-
idant defense systems measured in blood samples from
53 women with PE and compared them with those of 30
healthy pregnant controls; the results were also analyzed
in relation to severity of PE, with the altered metabolic
parameters or with endothelial activation circulating
markers. A further aim was to examine whether a
different susceptibility of pregnant women to developing
PE may be a function of their own endogenous antio-
xidant status.
MATERIAL AND METHODS
Subjects and blood sample collection
The study protocol was approved by the Ethics
Committee for Clinical Research of the Vall dHebron
Hospital. Subjects were selected from women attending
the Obstetrics and Gynecology Department from June
2001 to June 2003. Eligible cases were singleton preg-
nancies with the diagnosis of PE. Criteria for the defini-
tion of PE were those of the International Society for the
 Oxidative stress and preeclampsia
Study of Hypertension in Pregnancy. Preeclampsia was
diagnosed in previously normotensive women with two
repeat (4 h apart) diastolic blood pressure measurements
of 90 mmHg or greater after the 20th week of gestation,
plus proteinuria of more than 300 mg/l in 24 h as
measured quantitatively or, in its defect, >2+ protein by
dipstick in two repeat measurements (4 h apart). Cases of
PE were further subclassified as mild or severe. The
definition of severe PE was established if any of the
following were present: persistent diastolic blood pres-
sure greater than 110 mmHg, platelet count <100,000/ml,
elevated transaminase levels (GOT > 40, GPT > 40),
hemolysis (haptoglobin level <0.06 g/l), or neurological
involvement. Women in labor, with ruptured membranes,
multiple pregnancy, or any concurrent medical compli-
cations before or developing during pregnancy, such as
diabetes mellitus or inflammatory diseases, were not
considered eligible for the study. Smoking was also
considered an exclusion criterion. The control group
comprised a consecutive sample of pregnant women
followed up at our institution and undergoing routine
third-trimester blood analysis and with none of the
exclusion criteria. No cases or controls had received
vitamin supplements or aspirin in the month before
enrollment.
Sera from two pregnant women, as possible positive
controls, were included in the Western blot immunoassay
for analysis of oxidatively modified proteins because one
of them suffered gestational diabetes and the other
combined hyperlipidemia without hypertension, condi-
tions that increase the risk of vascular disease and are
linked to oxidative stress.
Venous blood samples were drawn after 8 h fast and
processed within 1 h. Blood was collected into tubes
containing EDTA or heparin that were centrifuged at
1500  g for 5 min at 4jC and erythrocytes were washed
three times with NaCl 0.9% (wt/vol). Analyses of gluta-
thione (GSH) were performed within 6 h after blood
collection; glutathione peroxidase (GPx), superoxide
dismutase (SOD), and total lipid hydroperoxides (HPx)
were analyzed within 1 week. Aliquots of EDTA plasma
and washed erythrocytes were stored at 85jC until
analysis (in less than 1 month).
Reagents
All chemicals were analytical grade. Butylated
hydroxytoluene, 2-thiobarbituric acid (TBA), and
1,1,3,3-tetraethoxypropane were purchased from Sigma
Chemical Co. (St. Louis, MO, USA), high-performance
liquid chromatography (HPLC) grade acetonitrile was
from Romil (Cambridge, UK), and n-butanol, methanol,
ethanol, and n-hexane were from Merck (Darmstadt,
Germany). Highly purified water (resistivity = 18 M?
cm) obtained through a Milli-Q water purification system
559
(Millipore, Bedford, MA, USA) was used for solution
preparation and for the mobile phase. a-Tocopherol,
chloramine-T, potassium iodide, hydrogen peroxide
(H2O2), 2,4-dinitrophenylhydrazine (DNPH), and phos-
photungstic acid were purchased from Sigma Chemical
Co.. Tocopherol acetate and serum calibration standard
for vitamin E analysis were obtained from Chromsys-
tems (Munich, Germany). Triphenylphosphine 99%
(TPP) was purchased from Aldrich (Gillingham, Dorset,
UK). Ammonium ferrous sulfate in aqueous sulfuric acid
and xylenol orange in methanol were obtained as the
commercial preparation (PeroXOquant Quantitative Per-
oxide Assay, lipid-compatible formulation) from Pierce
(Rockford, IL, USA). The commercially available syn-
thetic hydroperoxides 5-HPETE and 13-HPODE were
purchased from ICN Biomedicals, Inc. (Aurora, OH,
USA). 2,2-Azobis (2-amidinopropane)hydrochloride
(AAPH) was obtained from Polysciences (Warrington,
PA, USA). 6-Hydroxy-2,5,7,8-tetramethylchroman 2-
carboxylic acid (Trolox) was obtained from Aldrich
Chemical Co.(Milwaukee, WI, USA) and h-phycoery-
thrin from Porphirium cruentum was obtained from
Sigma. Polyvinylidene fluoride membranes, Immobi-
lon-P, were purchased from Millipore and the Trans-Blot
SD semidry system from Bio-Rad (Hercules, CA, USA).
Rabbit anti-dinitrophenylhydrazone (DNP) antiserum
was obtained from Dako (Carpenteria, CA, USA) and
the chemiluminescence kit from Tropix (Bedford, MA,
USA). All other general chemicals used were of the
highest purity available.
Analytical methods
Determination of lipid and protein peroxidation
parameters. Plasma malondialdehyde (MDA) concentra-
tion was determined as its diethylthiobarbituric acid
adduct (TBA-MDA), after reverse-phase isocratic HPLC
separation of the MDA-TBA complex based essentially
on the method of Fukunaga et al. [29] with the modifi-
cations previously described in detail [30]. Erythrocyte
MDA content was also determined by reversed-phase
HPLC [31] with fluorescence detection.
Plasma lipid peroxide (LPO) levels were measured
spectrofluorometrically, according to the optimized
method of Yagi [32]. In this method, lipids and proteins
are precipitated by use of a phosphotungstic acid sulfu-
ric acid system from other TBA-reactive substances
(TBARS) such as glucose and water-soluble aldehydes
in plasma. Modifications introduced by Wasowitz et al.
[33] were incorporated for acid pH control of the reaction
and extractions. Lipid peroxide concentrations were
expressed in terms of MDA with 1,1,3,3-tetraethoxypro-
pane as a standard (Amol/l).
Total lipid hydroperoxide plasma levels were quan-
tified by PeroXOquant Quantitative Peroxide Assay
560 E. Llurba et al.
FOX (ferrous oxidation/xylenol orange technique),
according to the method of Nourooz-Zadeh et al.
[34]. Plasma HPx concentration in the samples was
determined as a function of the mean absorbance
difference of samples with and without HPx elimination
by TPP at 560 nm using an H2O2 standard curve in the
concentration range of 0 20 Amol/l. Sample assays
were made in triplicate and the acceptable CV had to
be under 5%.
Plasma protein sulfhydryl (SH) were measured by the
method designed by Hu [35].
Analysis of advanced oxidation protein products
(AOPP) was made using the method described by
Witko-Sarsat et al. [36]. AOPP concentrations were
expressed as Amol/l chloramine-T equivalents. The
within-run and between-run CVs were 2.7 and 4.1%,
respectively.
Plasma protein carbonyl content (PCC) levels were
evaluated after the DNPH assay [37] with slight mod-
ifications as detailed previously [30]. Protein carbonyl
groups were expressed as nmol/mg protein.
Determination of oxidized carbonylated protein levels by
Western blot analysis. The carbonyl content of plasma
proteins was also measured by SDS PAGE and immu-
noblot detection of protein-bound 2,4-dinitrophenylhy-
drazones as described elsewhere [37] and detailed
previously by us [38]. The gel system was calibrated
for molecular weight determination by measuring the
migration of standard proteins (7.2 218 kDa). The
protein concentration was determined by means of the
Bio-Rad protein assay reagent (Bio-Rad, Munich, Ger-
many) using bovine serum albumin as standard, accord-
ing to the Bradford method [39]. Plasma protein
concentration was not altered by PE. However, plasma
protein determination was done previously and calcu-
lated in each case, in order that equal amount of proteins
were applied on each lane (22.5 Ag/lane). Densitometric
analysis of the spots on the autoradiography was per-
formed using the Molecular Analyst 2.1.2 software
(Bio-Rad).
Erythrocyte assay of antioxidant enzymes and glutathione.
SOD activity was measured in erythrocytes by a com-
mercially available kit (Ransod; Randox Laboratories,
Grumlin, UK) based on the method developed by
McCord and Fridovich [40] and expressed as U/g Hb.
Hemoglobin concentration was determined by the cyan-
methemoglobin method [41], using Sigma Diagnostic
reagents. Erythrocyte glutathione peroxidase activity
(er-GPx) was determined using a commercial kit (Ransel,
Randox Laboratories) and expressed as U/g Hb. This
method is based on Paglia and Valentine [42]. GSH
content was measured in erythrocytes using the enzymatic
method described by Anderson [43] and expressed as
Amol/g Hb.
Plasma antioxidants. Plasma a-tocopherol concentra-
tion was determined by reversed-phase HPLC [44]
with ultraviolet detection at 280 nm (Waters Model
486 tunable absorbance detector) and peak automatic
integration as detailed previously [30]. Because of the
clear metabolic relationship between plasma a-tocoph-
erol and plasma lipid parameters [45], we also express
a-tocopherol levels as a ratio of a-tocopherol/total
lipids (total cholesterol + triglycerides). Plasma lipid
profile (total cholesterol and triglycerides) and uric acid
were measured by quantitative enzymatic assays (Sig-
ma Chemical Co.) according to the manufacturers
protocols, and the results were standardized using
standard solutions for cholesterol, glycerol, and uric
acid calibrators.
Plasma ascorbic acid (vitamin C) was measured by
reversed-phase HPLC with ultraviolet detection, and the
method used was essentially that described by Levine et
al. [46].
Total radical-trapping antioxidant parameter (TRAP)
analyses. Plasma antioxidant activity was measured
using a modification of Glazers fluorometric assay
[47]. The water-soluble vitamin E analog Trolox was
used as a standard and AAPH as generator of peroxyl
radicals. Fluorescence was measured at the emission at
565 nm and excitation at 540 nm. The results were
calculated in the following way: %TRAP = 1  (D
abs. sample or Trolox/D abs. blank)  100. TRAP levels
were expressed as arbitrary units equivalent to Trolox
antioxidant activity and calculated from a calibration
curve determined daily between the TRAP percentage
and the concentrations of the three standards. The linear
regression coefficient of the calibration curve was never
below 0.999. Intra- and interday coefficients of variation
were 2.9 and 6.6%, respectively.
Measurement of blood plasma oxidizability was car-
ried out according to Kontush and Beisiegel [48].
Soluble adhesion molecule assays. Vascular cell adhesion
molecule-1 (VCAM-1) circulating levels were analyzed
with commercially available enzyme immunoassay (EIA)
purchased from Roche Diagnostics (Mannheim, Ger-
many) and plasma levels of intercellular adhesion mole-
cule-1 (ICAM-1) with an EIA from HyCult biotechnology
(Uden, Netherlands).
Statistical analysis
Results were statistically analyzed by analysis of
variance (StatView 4.01, statistical package). Correla-
tions between variables were studied by linear regression.
 Oxidative stress and preeclampsia 561

Table 1. Clinical Characteristics of Preeclamptic and Normal Pregnant

Women

Characteristic Preeclamptic women Normal pregnant

(n = 53) women (n = 30)

Age at delivery (years) 31.12 F 5.6 29.4 F 5.8

Parity 1.3 F 0.2 1.4 F 0.2
BMI at sampling 35.1 F 2.8 34.03 F 3.3
GA at sampling (weeks) 32.5 F (24 39) 33.7 F (26 40)
GA at delivery (weeks) 32.5 F 3.2** 39.4 F 1.6
Systolic blood pressure 157.8 F 17.9*** 120.6 F 12.6
(mmHg)
Diastolic blood pressure 99.1 F 12.3*** 66.8 F 8
(mmHg)
Mean birth weight (g) 2216 F 690* 3177 F 305

Values are given as means F SD.

***p < .0001, **p < .001, *p < .05 compared with normal
pregnancy (Students t test).

Data are given as means F SEM and differences were

considered significant when p < .05.

RESULTS

Description of the clinical characteristics of the study
groups is shown in Table 1. The control women did not
differ from those in the PE groups regarding age, parity,
BMI, and gestational age. Women with PE delivered on
average 5 weeks early and consequently mean birth
weight was lower than for healthy pregnant women. As
expected, blood pressure and urine protein were signif-
icantly raised in preeclamptic patients. Biochemical
parameters are given in Table 2. Triglyceride concen-
trations were significantly increased in women with PE
compared with normal pregnant controls matched for
gestational age at sampling. Cholesterol levels showed
no differences between the study groups. Plasma uric
acid level was significantly higher in women with PE
than in normal pregnant controls.
Plasma total HPx, the major initial reaction products
of lipid peroxidation, were 130% higher in the overall
group of PE patients compared with controls (Fig. 1A)
and were significantly elevated in women with severe PE
Fig. 1. (A) Plasma lipid hydroperoxides in preeclamptic (PE) and
normal pregnant women and (B) lipid hydroperoxides in mild PE,
severe PE (Sev PE), and controls. Values represented are means F
SEM. *p < .05 versus control subjects.
compared with controls (Fig. 1B). Levels of HPx seemed
to be related to severity of the disease.
Plasma MDA and LPO concentrations did not differ
significantly between the overall PE group and the
controls and neither did erythrocyte MDA content
(Table 3). When data were analyzed according to
severity of the disease, plasma MDA, erythrocyte
MDA, and LPO values were similar among PE groups,
regardless of clinical presentation. In short, neither of
the secondary end-products of lipid peroxidation, MDA
or LPO, showed increases intracellularly or extracellu-
larly in our PE study group compared with control
pregnancies. Whole plasma oxidation-induced assay,
which could serve as a measure of the oxidizability of

Table 3. Plasma MDA, er-MDA, and LPO and Oxidizability Values in

Table 2. Serum Levels of Cholesterol, Triglycerides, and Uric Acid in Severe, Mild, and Total Groups of Preeclamptic Women and Normal
Preeclamptic and Normal Pregnant Women Pregnant Women

Preeclampsia Normal pregnancy Severe PE Mild PE Total PE Normal

(n = 53) (n = 30) (n = 38) (n = 15) (n = 53) pregnancy
(n = 30)
Triglycerides (mM) 3.04 F 1.2* 2.50 F 0.9
Total cholesterol (mM) 6.97 F 1.8 7.6 F 1.5 MDA (AM) 0.62 F 0.38 0.55 F 0.29 0.60 F 0.36 0.66 F 0.17
Uric acid (AM) 341.5 F 96.7** 249.6 F 55.2 er-MDA (AM) 0.65 F 0.27 0.66 F 0.2 0.66 F 0.25 0.65 F 0.15
LPO (AM) 2.07 F 0.62 2.27 F 0.69 2.12 F 0.64 2.1 F 0.63
Values expressed as means F SD. Oxidizability 6.4 F 1.7 6.4 F 1.6 6.4 F 1.7 6.9 F 1.1
* p < .05 vs. normal pregnant controls.
** p < .0001 vs. normal pregnant controls. Values expressed as means F SD.
562 E. Llurba et al.
Fig. 2. (A) Plasma protein-SH groups in preeclamptic (PE) and normal
pregnant women and (B) protein-SH groups in mild PE, severe PE (Sev
PE), and controls. Values represented are means F SD. ***p < .0001
versus control subjects.
plasma lipoproteins, showed no differences between PE
patients and controls with normal pregnancies (Table 3).
In women with PE, plasma thiol, a parameter of
nonoxidation of SH groups, was significantly lower ( p
< .001) than in normal pregnant controls (Fig. 2A).
Moreover, the decrease was constant in all PE patients,
although the more marked loss of SH groups occurred in
patients with the severe form of PE (Fig. 2B).
The values of some protein oxidation markers are
summarized in Table 4. Plasma protein carbonyl content,
a marker of protein oxidative damage, decreased signif-
icantly in all women with PE, regardless of severity.
Plasma AOPP showed no significant differences either
between PE forms or between the total PE group and the
controls.
Table 4. Plasma Protein Carbonyl Groups (PCG) and Advanced
Oxidation Protein Products (AOPP) in Women with Severe PE or Mild
PE and Normal Pregnant Controls
Severe PE Mild PE Total PE Control
(n = 38) (n = 15) (n = 53) group
(n = 30)
PCG (nM/ 0.59 F 0.3* 0.58 F 0.26* 0.58 F 0.3* 0.89 F 0.29
mg protein)
AOPP (AM) 168.56 F 88 171.8 F 82.6 169.5 F 86 143.2 F 66
Values expressed as means F SD.
* p < .01 vs. normal pregnant controls.
A representative Western blot immunoassay for de-
tection of oxidatively modified plasma proteins is shown
in Fig. 3. Visual inspection of Coomassie brilliant blue
staining for total protein on Western transfer blots
indicated equal protein loading of plasma samples from
preeclamptic women and controls (image not shown).
Our results show a certain basal protein oxidation level
in both controls and PE group and densitometric analysis
of all stained plasma protein bands from PE patients and
controls supported the conclusion of similar detectable
amounts of carbonyl groups. Densitometric intensity
(DI) of carbonyl plasma proteins, immunodetected after
DNP derivatization, was 21% higher in a plasma sample
from a pregnant woman with combined hyperlipidemia
than in controls or preeclamptic women. The most
striking differences were observed in a plasma sample
from a gestational diabetic woman, in which relative DI
of oxidized protein spots was higher (45%) than in
controls.
Our results showed activity of the antioxidant enzyme
SOD to be significantly higher in red blood cells of PE
patients compared with controls (Fig. 4A). The increase
in SOD activity was significantly raised in both severe
and mild PE, although the differences were more marked
in severe PE (Fig. 4B). Erythrocyte GPx activity showed
no significant differences when PE patients were com-
pared (as a whole) to controls (Fig. 4C). However, when
analyzed for disease severity, women with severe PE had
a significant increase in GPx activity compared to
healthy pregnant women (Fig. 4D).
Fig. 3. Immunochemical detection of carbonylated proteins in plasma of
control pregnant women (C), preeclamptic women (P), a gestational
diabetes patient (D), and a pregnant woman with familial hyper-
triglyceridemia (T). Western blot analysis (protein concentrations are
22.5 Ag/lane in the sample) is presented. Molecular weight markers are
shown on the left. A typical immunoblot representative of 10
independent experiments is shown.
 Oxidative stress and preeclampsia 563

Fig. 4. Erythrocyte antioxidant enzyme activities in preeclamptic (PE) and normal pregnant women. (A) Superoxide dismutase (SOD)
activities in PE and normal pregnant women and (B) SOD in mild PE, severe PE (Sev PE), and controls. (C) Glutathione peroxidase
(GPx) activities in PE and normal pregnant women and (D) GPx in mild PE, severe PE, and controls. Values represented are means F
SD. *p < .05 versus control subjects and vs. mild PE patients.

Concentrations of some antioxidative defense systems
are shown in Table 5; plasma levels of the liposoluble
antioxidant a-tocopherol were similar in PE patients and
controls. However, a-tocopherol levels corrected by total
lipids (cholesterol and triglycerides) were significantly
increased in both mild and severe PE compared with
controls. Plasma concentrations of ascorbic acid (vitamin
C) were found to be significantly decreased in PE
women compared with healthy pregnant women, al-
though low vitamin C values were not observed in any
case. In women with PE, plasma TRAP values were
comparable to values in normotensive pregnant women
(Table 5). No significant differences in erythrocyte GSH
content or ratios of GSH to hemoglobin were found
between normal pregnant women and those with PE
(Table 5).
The serum level of soluble VCAM-1, which reflects
endothelial cell activation, was found to be significantly
higher in women with PE compared with normal preg-
nant women (Fig. 5A). Moreover, the severe form of PE
showed the highest levels of VCAM-1 (Fig. 5C). Serum
levels of ICAM-1 were not significantly elevated in PE

Table 5. Plasma a-Tocopherol, a-Tocopherol/Total Lipids (TL), Total Reactive Antioxidant Potential (TRAP), Ascorbic Acid, and Erythrocyte GSH
Content in PE Groups and Normal Pregnant Women

Severe PE (n = 26) Mild PE (n = 16) Total PE (n = 46) Normal pregnancy (n = 35)

a-Tocopherol (AM) 38.5 F 13.1 39.3 F 12.4 38.7 F 12.7 37.4 F 7.1
a-Tocopherol/TL (AM/mM) 8.4 F 2.1* 8.8 F 2.2* 8.5 F 2.1* 7.4 F 1.3
TRAP (AM) 2070 F 337 2090 F 198 2077 F 293 2025 F 289
Ascorbic acid (AM) 40.2 F 35.7* 35.87 F 26.5** 38.66 F 32.4* 56.93 F 24.7
Erythrocyte GSH 12.2 F 2.9 11.7 F 4.4 2.1 F 3.4 12.9 F 2.3
(AM/g Hb)

Values expressed as means F SD.

*p < .05, **p < .01 vs. normal pregnant controls.
564 E. Llurba et al.

Fig. 5. (A) Plasma soluble concentrations of vascular cell adhesion molecule-1 (VCAM) and (B) intercellular adhesion molecule 1
(ICAM) in preeclamptic and normal pregnant controls. (C) VCAM and (D) ICAM in the preeclampsia subgroups and in normal pregnant
controls. Values are expressed as means F SD. *p < .05, **p < .001, and ***p < .0005 versus normal pregnant controls.

patients and showed no differences when analyzed for
disease severity (Figs. 5B and 5D).
DISCUSSION
Oxidative stress has been proposed as an underlying
mechanism that contributes to endothelial dysfunction
associated with PE. The clinical significance of this
attractive pathogenic pathway remains to be substantiat-
ed because no general consensus on the existence of
systemic oxidant stress in PE has yet been attained
[4,5,20 22]. The root of the controversy may lie in the
fact that definitive evidence for the involvement of free
radicals in disease processes is hampered by the lack of
comparative methods and, therefore, the difficulty of
comparing published data to assess absolute levels of
oxidative stress in vivo. Small and heterogeneous patient
groups might also have been leading reasons for dis-
agreement. The present work was undertaken to gain
further insight into the role of oxidative damage in PE; to
this end, and to better characterize oxidative stress status,
the study was approached from several angles and
conducted in a large cohort of healthy and preeclamptic
pregnant women. The PE study group was well charac-
terized as demonstrated by clinical and analytic data at
enrollment, perinatal outcome, and the PE-selective and
severity-related elevation in cell adhesion molecule
VCAM-1, a marker of endothelial dysfunction.
Lipid peroxidation, a central feature of oxidant stress,
may alter membrane integrity and/or membrane-associ-
ated functions and can be assessed by a number of
methods that include the quantification of either primary
or secondary peroxidation end products. Our results
showed total lipid hydroperoxides, the major initial
reaction products of lipid peroxidation, to be elevated
in plasma of the preeclamptic women studied; further-
more, the highest values of hydroperoxides were found
in patients with the severe form of the disease. However,
on evaluating lipid hydroperoxides using the highly
sensitive HPLC chemiluminescence assay, no primary
lipid peroxidation products could be detected in samples
from PE patients in any of the groups in either plasma or
red blood cell homogenates [21]. Comparison of results
between the latter study and ours is not appropriate
owing to the differences in sensitivity and specificity of
the methods employed.
A salient finding of our study was that in this group of
PE patients, plasma levels of the most widely used
markers of secondary end products of lipid peroxidation,
such us MDA and lipoperoxides, did not increase com-
pared with those of healthy pregnant women. Further-
more, oxidative damage to lipids does not seem to be
related to severity of the disease because no differences
were found when preeclamptic women were classified as
having mild or severe PE. Our data are consistent with
other previously published studies using methods com-
 Oxidative stress and preeclampsia
parable to ours, which also failed to find evidence that
circulating markers of oxidative stress are significantly
elevated in plasma of PE women [19 21,49]. Quantifi-
cation of F2-isoprostanes seems to be a significantly
more accurate marker of oxidative stress in vivo in
humans and animals than other compounds [50]; with
the use of this precise and sensitive method for evaluat-
ing this highly abundant product of lipid peroxidation in
urine, Regan et al. [22] reported no evidence for en-
hanced lipid peroxidation in severe PE patients. Contrary
to the above studies, various authors have reported
significant increases in secondary products of lipoperox-
idation in plasma of PE women [13 17,51]. However,
comparison of the results of the present study with the
latter published data is not possible because some of
these studies estimated lipid peroxidation as TBARS
measurement [15 17,51], the interlaboratory conditions
of which vary, and the technique is relatively nonspecific
and nonsensitive for assessing lipid peroxides in biolog-
ical specimens. The MDA results presented herein were
obtained by HPLC, a highly specific, sensitive, and
reproducible method, which measures MDA-TBA ad-
duct without interference from other substances present
in plasma in variable degrees [52]. Our findings differ
from those obtained in other studies in which plasma
MDA was also determined by HPLC, yielding serum
concentrations of MDA 20 50% higher in women with
PE compared to controls [13,14]. The differences could
probably lie in the heterogeneity of this pathology, which
is often overlooked, together with the small number of
cases included in some studies.
Similarly, plasma levels of lipoperoxides, another
characteristic marker of lipid peroxidation, did not un-
dergo significant changes with PE. Thus, the validity of
our findings is supported by the consistency of the results
obtained by using two indices of oxidative damage to
lipids: MDA and LPO. However, plasma concentrations
of these secondary products of lipid peroxidation were
significantly raised in women with PE or normal preg-
nancy (43 and 40%, respectively) compared with non-
pregnant women. This comparison should be stressed
and is in line with the previously reported findings which
concluded that circulating markers of oxidative stress are
raised in normal pregnancy and preeclampsia [20].
On the other hand, the process of lipid peroxidation of
membranes has been implicated as one of the primary
events in oxidative cellular damage and has been shown
to be associated with fine structure disturbance and
subsequent function loss of biological membranes. As
mature human erythrocytes have no nucleus or other
organelles, the plasma membrane in these cells is the
critical target. In this study, erythrocytic MDA, evaluated
by a highly sensitive and accurate method, showed no
evident increase in preeclamptic women. This finding
565
has not been previously reported, but is wholly consistent
with the above results of plasma MDA because by
separate analysis of extracellular and erythrocyte MDA,
our data demonstrate that, in PE, red blood cells and
surrounding blood plasma are highly resistant to oxida-
tive damage, which indicates that antioxidant protective
mechanisms are highly efficient against peroxidative
stress in PE-complicated pregnancies.
With regard to intracellular antioxidant defense status
in PE, no changes were observed in total content of
tripeptide GSH, the most abundant antioxidant at cellular
level, in erythrocytes of PE women. GSH has several
antioxidant functions; it scavenges singlet oxygen and
hydroxyl radicals and also acts as a cosubstrate of GPx in
the elimination of both H2O2 and other organic perox-
ides. Interestingly, antioxidant activity of the cytosolic
GPx rose in patients with severe PE and, moreover, an
early activation was produced in Cu,Zn-SOD which
resulted in a clear increase in antioxidant activity in
erythrocytes of the whole PE group, consistent with a
concerted adaptive response to elevated levels of oxidiz-
ing species that elicit a specific compensatory increase in
antioxidant enzyme activities. Thus, PE would induce an
initial ROS overgeneration promptly followed by induc-
tion of endogenous antioxidant enzymes to reinforce
resistance to oxidative injury. GPx is the only antioxidant
enzyme known to directly reduce phospholipid hydro-
peroxides within membranes and lipoproteins, acting in
conjunction with a-tocopherol to inhibit lipid peroxida-
tion. Overexpression of the antioxidant complex enzyme
SOD has been postulated as one of the major mecha-
nisms by which cells counteract the deleterious effects of
ROS and protect themselves from oxidative damage and,
in PE, could play a protective role against initial perox-
idation processes. Nevertheless, analysis of molecular
mechanisms linked to antioxidant response, which did
not form part of the present study, may provide mecha-
nistic clues to the activation of antioxidant defense
systems in normal and PE pregnancies.
Proteins, which constitute major components of living
cells, are among the main targets of ROS-induced oxi-
dation that provokes specific chemical oxidative modifi-
cations. Considering that oxidative stress yields a range
of products, rather than a single product, various indica-
tors were used in this study to characterize the oxidative
damage to proteins; therefore, protein oxidation products
were evaluated through quantification of both plasma
protein carbonyl content and AOPP levels and visualized
by Western blot analysis of oxidatively modified plasma
proteins. In general terms, no clear changes in protein
oxidation markers in plasma of preeclamptic women nor
an apparent correlation with severity of the disease was
observed. If we focus on the most general indicator and
by far the most commonly used marker of protein
566 E. Llurba et al.
oxidation, protein carbonyl content, not only was no
increase observed in plasma PCC in PE but rather we
saw a decrease, which might indicate that the catabolic
rate of damaged serum proteins would be increased in
this condition. A further possible explanation for this
apparently rare result would lie in the turnover time of
protein-bound carbonyls, which varies from hours to
days, unlike reactive aldehydes derived from lipid per-
oxidation that turn over within minutes [53]. On the other
hand, the decrease in plasma PCC levels during PE might
also reflect the efficient activation of antioxidant reduc-
ing systems and removal mechanisms which partially
remove circulating oxidatively modified proteins that are
metabolized at a greater rate by the liver [6]. Thus, if a
rise in oxygen radical activity is produced in PE, then
synergistic and consecutive actions of the complex
defense systems against free radical attack would be
triggered to control ROS generation. It is known that
the accumulation of ROS-mediated protein damage in
living organisms is limited by the preferential degrada-
tion of oxidized proteins by various proteases; however,
excessive oxidation and/or conversion of oxidized pro-
teins to cross-linked derivatives renders them resistant to
proteolytic degradation by the proteasome [54]. With the
use of different analytical methods, higher PCC levels
have been reported in the placenta and maternal blood of
PE-associated pregnancies [55,56].
Western blot analysis, performed with anti-DPNH
antibodies, which provides substantial improvements in
the sensitive and specific detection of protein-bound
carbonyl groups, revealed that the pattern of oxidized
proteins follows the same trend as that observed for
plasma PCC in healthy and PE women. A noteworthy
finding of this study were the intense spots of oxidized
proteins that appeared along the lane corresponding to a
plasma sample from a gestational diabetic woman, clear-
ly demonstrating that oxidative damage to proteins does
occur to a great extent in gestational diabetes. Previous
works from this laboratory [30,38] and numerous other
studies showed that hyperglycemia induces oxidative
stress and decreases antioxidant defenses and that lipids
are an important source of chemical modification of
proteins in diabetes [57]. Considering that plasma tri-
glyceride levels in the patient with gestational diabetes
were similar to those found in the preeclamptic group
and that oxidatively modified proteins were not found in
plasma of PE women leads us to think that, in PE,
antioxidant defense mechanisms are activated to protect
lipid oxidizable substrates from lipid peroxidation reac-
tions and efficiently inhibit the formation of advanced
lipoxidation end-products on proteins. Pregnant women
who develop PE may share many risk factors with
patients with metabolic syndromes, such as insulin resis-
tance, dyslipidemia, and obesity, in whom activation of
endothelial cell function antedates the clinical diagnosis
and has features in common with atherosclerosis. Thus,
the other case we considered of interest to study was the
pregnant woman in whom plasma triglyceride levels
were far higher than those of preeclamptic women due
to familial hypertriglyceridemia, which, being conducive
to vascular pathology, could in part be etiologically
related to PE, because both are characterized by an
excess or imbalance of lipids in blood. The serum sample
from this pregnant woman contained oxidized proteins to
a greater degree than those from PE women, which
points to the excess of oxidizable substrate in plasma
as the etiologic factor involved in oxidative injury in this
case. The results of the parallel study conducted in the
above pregnant women with metabolic disorders to
compare the oxidative stress markers with those of the
PE group clearly demonstrated that gestational diabetes
and, to a lesser extent, familial hypertriglyceridemia
induced evident oxidative damage to proteins, thereby
indicating oxygen radical activity.
Vitamin E, mainly a-tocopherol, is the major peroxyl
radical scavenger in biological lipid phases such as
membranes or LDL. Its antioxidant action has been
ascribed to its ability to chemically act as a lipid-based
free radical chain-breaking molecule, thereby inhibiting
lipid peroxidation and oxidized LDL formation and
protecting the organism against oxidative damage. The
role of vitamin E in diseases involving lipid oxidation
and endothelial damage, such as atherosclerosis or PE,
has been widely studied, although contradictory results
exist, i.e., normal or increased vitamin E levels in
preeclamptic women. Some [14,16,18,58], but not all
[15,19,20,25], studies found vitamin E to be decreased in
severe PE. In this study we found concentrations of
plasma a-tocopherol to be similar in PE patients and
healthy pregnant women; furthermore, a-tocopherol lev-
els corrected by total lipids (cholesterol and triglycerides)
were significantly increased in both mild and severe PE
compared with normal pregnancy. It should be men-
tioned here that a-tocopherol levels were significantly
higher in pregnant than in nonpregnant women (data not
shown). Different dietary vitamin E intake and the non-
lipid adjustment of plasma vitamin E levels may explain
the paucity of consistent results reported in the literature.
It has been demonstrated that vitamin E levels rise
progressively throughout normal pregnancy, probably
due to the parallel increase in circulating lipids [2,7],
which can influence fat-soluble antioxidant concentra-
tions and may confound their interpretation as indicators
of antioxidant status. Consequently, as performed in the
present study, the simultaneous lipid standardization of
serum lipid-soluble antioxidant concentrations, which
better reflects the relative levels of vitamin E, is recom-
mended [59]. Our results suggest that the lipid perox-
 Oxidative stress and preeclampsia
idation process, if it were induced in PE, would be
strongly suppressed by the presence of high plasma
levels of the lipophilic antioxidant a-tocopherol. How-
ever, circulating vitamin E levels in PE may be insuffi-
cient to protect the increased small, dense LDL particles
which result from hypertriglyceridemia and which are
particularly susceptible to oxidation [60]. Along this line,
in a recent study, we showed that women with a history
of PE had significantly increased susceptibility to lipo-
protein oxidation 1 to 3 years after delivery compared
with women who had had a normal pregnancy [61].
This study investigated a broad spectrum of blood
antioxidants in PE and revealed that the group of PE
women had significantly lower levels of the water-soluble
antioxidant molecule ascorbic acid than controls. This is
in keeping with data from various observational studies
which suggested that, in patients with PE, ascorbate may
be utilized to a greater extent to counteract free radical-
mediated cell disturbances [4,5,17,18,62] and with stud-
ies by others who advocate vitamin C and E supplemen-
tation for the prevention of PE in women at increased risk
of the disease [24,25]. The results of this study would
suggest that water-soluble antioxidant nutrients may ini-
tially be consumed, resulting in a reduction in vitamin C
plasma levels followed, if required, by lipid-soluble
antioxidants, mainly tocopherols. Vitamin C, present in
aqueous compartmentscytosol, plasma, and other body
fluidsis able to trap most ROS present therein and
functions as a first-line defense mechanism against free
radicals. The interaction between vitamins C and E in the
antioxidant defense of biochemical systems is well estab-
lished because ascorbic acid can reduce tocopheroxyl
radicals directly or indirectly and thus support the anti-
oxidant activity of vitamin E. Upon oxidation, these
micronutrients need to be regenerated in the biological
setting, hence the need for further coupling to nonradical
reducing systems such as GSH, dihydrolipoate, or
NADPH and NADH [63]. It is noteworthy that in our
group of pregnant preeclamptic women, vitamin C levels
declined, albeit moderately, while maintaining adequate
vitamin C status (>23 AM); however, plasma vitamin E
levels remained high, resulting in a vitamin C/vitamin E
ratio 37% lower in PE women than in controls, which
could prove unfavorable for the physiologic balance
between both vitamins and their antioxidant function.
Along these lines, it has been proposed that low ascorbate
in PE plasma may result in decreased rates of decompo-
sition of S-nitrosothiols and that it may be a possible
mechanism for impaired NO group release in PE [64]. It
should be pointed out that both the intake and the plasma
levels of vitamins C and E are generally high in our
population (on a Mediterranean diet), which may con-
tribute to the lower incidence of PE in these countries.
Considering that this group of pregnant PE women does
567
not have abnormally low levels of antioxidant vitamins, it
seems unlikely that, in our environment, vitamin C and E
supplementation would play a beneficial role in PE
prevention. Alternatively, a diet deficient in antioxidants
and/or poor antioxidant status would lower the protective
effect and thereby contribute to oxidative stress.
This study showed plasma thiol groups, biological
markers of protein protection against oxidative stress, to
be susceptible to oxidative damage because levels in PE
patients were low. It is worth noting that partial losses of
thiol compounds in this PE group were related to severity
of the disease. These compounds, among other redox-
sensitive molecules, play a significant role in the cell by
minimizing the deleterious consequences of oxygen
activation processes. Essentially, all plasma SH groups
are protein associated and, as albumin is the most
abundant plasma protein, it would lose in part its extra-
cellular antioxidant power; therefore, decreases found in
thiols could also be considered very early products of
protein oxidation [35]. Recent results indicated that
protein thiol/disulfide oxidoreductases, such as thiore-
doxin, glutaredoxin, and protein disulfide isomerase,
were adaptively induced in preeclamptic placentas,
which suggests possible roles for these specific antiox-
idants in regenerating oxidatively damaged proteins and
protecting placental functions against oxidative stress
[65,66]. Because the metabolism of the reactive species
should be regulated selectively at or near the sites of their
generation, increases in the expression of these reducing
systems in preeclamptic placentas would act as a protec-
tive mechanism by which placental cells minimize tissu-
lar oxidative damage, but would not restore or prevent
the oxidation of protein sulfhydryls in plasma, after low-
level oxidative stress.
Human plasma protection against free radical injury is
offered by a wide spectrum of antioxidants with synergic
action so that individual measurements of antioxidant
concentrations in blood do not always reflect the level of
antioxidant status. We showed that global plasma anti-
oxidant capacity was not decreased in sera of PE women,
despite low levels of vitamin C and thiol groups, which
demonstrates an adequate capacity of plasma to protect
its environment from free radical aggression. As the
TRAP activity of human plasma is mainly attributable
to uric acid, protein thiol groups, ascorbate, a-tocopher-
ol, and bilirubin [6], the low plasma levels of ascorbate
and thiols and the increase in uric acid and a-tocopherol
concentrations found in plasma of PE women might
account for the maintenance of the overall redox network
in plasma of these patients. Uric acid has been proposed
as an important antioxidant molecule and, being present
in plasma of PE women at concentrations almost 40%
higher than in plasma of healthy pregnancies, should
confer protection against free radical activity. High uric
568 E. Llurba et al.
acid concentrations have been reported to bear a potential
protective physiological function against oxidative dam-
age by scavenging of peroxyl radicals in vivo; however,
the consequences of increased plasma urate levels should
not necessarily be regarded as beneficial because hyper-
uricemia has been associated with higher vascular event
rates and renal morbidity.
Finally, with respect to the queries initially raised in
this work, i.e., the potential role of increased plasma
lipids of PE women in producing oxidants, the findings
of this extensive study show that lipid susceptibility to
oxidation measured directly in serum samples was not
altered, and consequently, the more usual markers of
lipid peroxidation did not rise. Therefore, from these
cumulative data it may be deduced that no clear associ-
ation exists between lipid oxidation products in the
maternal bloodstream and endothelial dysfunction in
PE. Results of this study reflect a mild oxidative stress
level in blood of preeclamptic women, which does not
permit us to conclude that oxidant stress might be a
pathogenetically relevant process causally contributing to
the disease. It seems plausible that the observed findings
represent reactive changes in the context of an endothe-
lial or inflammatory disease. Nevertheless, if reactive
species are generated in preeclamptic placentae, localized
tissue-specific increases in oxidative stress might be
promoted. It should be borne in mind that, in vivo,
increased lipid or protein oxidation will always be
restricted to a small percentage of tissue and consequent-
ly, changes in concentration of peroxidation markers in
blood, if they occur at all, will be low. Moreover, many
of the peroxidation products are rapidly detoxified and/or
metabolized and therefore increased plasma levels often
exist for only a short time after an oxidative challenge.
In conclusion, this study provides the most compre-
hensive report on biochemical indices of oxidative dam-
age and antioxidant status in preeclamptic women in
whom biomarkers of lipid and protein oxidation process-
es were measured at different stages in both plasma and
erythrocytes. The findings of this in-depth study do not
support the concept of a generalized state of oxidative
stress in PE women and raise speculation that in PE free
radicals would react with lipids, although this process
might be interrupted promptly by the physiologic acti-
vation of antioxidant enzymes and by the high plasma
vitamin E levels that protect against free radical damage.
Similarly, the adaptive activation of secondary defense
systems and removal mechanisms against early protein
oxidative damage would prevent further oxidative mod-
ification of macromolecular cell components. As one
issue critical for evaluating the mechanisms underlying
the damage is its source, and this study was conducted in
plasma and erythrocytes, the results do not preclude the
existence of oxidative stress processes restricted to other
biological compartments, such as the placenta or the small
dense subfraction of low-density lipoproteins. Therefore,
potential disease mechanisms involving oxidative stress
in the genesis of endothelial dysfunction in PE are not
confirmed by the results of this study. Further studies are
required to gain insight into the molecular mechanisms
underlying endothelial damage in preeclamptic pregnan-
cies and seek initial pathophysiologic changes which
could lead to early diagnosis in preeclampsia.
Acknowledgments This work was supported by grants from the Fondo
de Investigacion Sanitaria (FIS 99/1103, FIS 01/1397), Centre Network
RCMN (C03/08), and Group Network RGDM (G03/212) financed by
the Carlos III Institute of Health. We are grateful to Miss Christine
OHara for her help with the English version of the manuscript.
REFERENCES
[1] Roberts, J. M.; Taylor, R. N.; Musci, T. J.; Rodgers, G. M.; Hubel,
C. A.; McLaughlin, M. K. Preeclampsia: an endothelial cell dis-
order. Am. J. Obstet. Gynecol. 161:1200 1204; 1989.
[2] Roberts, J. M. Preeclampsia: what we know and what we do not
know. Semin. Perinatol. 24:24 28; 2000.
[3] Granger, J. P.; Alexander, B. T.; Bennett, W. A.; Khalil, R. A.
Pathophysiology of pregnancy-induced hypertension. Am. J.
Hypertens. 14:178S 185S; 2001.
[4] Roberts, J. M.; Lain, K. Y. Recent insights into the pathogenesis
of pre-eclampsia. Placenta 23:359 372; 2002.
[5] Hubel, C. A. Oxidative stress in the pathogenesis of preeclampsia.
Proc. Soc. Exp. Biol. Med. 222:222 235; 1999.
[6] Halliwell, B.; Gutteridge, J. M. C., eds. Free radicals in biology
and medicine, 3rd ed. Oxford: Clarendon Press; 1999.
[7] Chen, K.; Thomas, S. R.; Keaney, J. F. Beyond LDL oxidation:
ROS in vascular signal transduction. Free Radic. Biol. Med. 35:
117 132; 2003.
[8] Caniggia, I.; Winter, J.; Lye, S. J.; Post, M. Oxygen and placental
development during the first trimester: implications for the patho-
physiology of pre-eclampsia. Placenta 21(Suppl. A):S25 S30;
2000.
[9] Wang, Y.; Walsh, S. W.; Kay, H. H. Placental lipid peroxides and
thromboxane are increased and prostacyclin is decreased in women
with preeclampsia. Am. J. Obstet. Gynecol. 167:946 949; 1992.
[10] Many, A.; Hubel, C. A.; Fisher, S. J.; Roberts, J. M.; Zhou, Y.
Invasive cytotrophoblasts manifest evidence of oxidative stress in
preeclampsia. Am. J. Pathol. 156:321 331; 2000.
[11] Myatt, L.; Rosenfield, R. B.; Eis, A. L.; Brockman, D. E.; Greer,
I.; Lyall, F. Nitrotyrosine residues in placenta: evidence of per-
oxynitrite formation and action. Hypertension 28:488 493;
1996.
[12] Wang, Y.; Walsh, S. W. Antioxidant activities and mRNA expres-
sion of superoxide dismutase, catalase, and glutathione peroxidase
in normal and preeclamptic placentas. J. Soc. Gynecol. Invest.
3:179 184; 1996.
[13] Hubel, C. A.; McLaughlin, M. K.; Evans, R. W.; Hauth, B. A.;
Sims, C. J.; Roberts, J. M. Fasting serum triglycerides, free fatty
acids, and malondialdehyde are increased in preeclampsia, are
positively correlated, and decrease within 48 hours post partum.
Am. J. Obstet. Gynecol. 174:975 982; 1996.
[14] Gratacos, E.; Casals, E.; Deulofeu, R.; Cararach, V.; Alonso, P. L.;
Fortuny, A. Lipid peroxide and vitamin E patterns in women with
different types of hypertension in pregnancy. Am. J. Obstet.
Gynecol. 178:1072 1076; 1998.
[15] Uotila, J.; Tuimala, R.; Aarnio, P.; Pykko, K.; Ahotupa, M. Find-
ings on lipid peroxidation and antioxidant function in hyperten-
sive complications of pregnancy. Br. J. Obstet. Gynecol. 100:
270 276; 1993.
[16] Wang, Y.; Walsh, S. W.; Guo, J.; Zhang, J. The imbalance be-
tween thromboxane and prostacyclin in preeclampsia is associated
 Oxidative stress and preeclampsia
with an imbalance between lipid peroxides and vitamin E in ma-
ternal blood. Am. J. Obstet. Gynecol. 165:1695 1700; 1991.
[17] Mutlu-Turkoglu, U.; Ademoglu, E.; Ibrahimoglu, L.; Aykac-
Toker, G.; Uysal, M. Imbalance between lipid peroxidation and
antioxidant status in preeclampsia. Gynecol. Obstet. Invest. 46:
37 40; 1998.
[18] Mikhail, M. S.; Anyaegbunam, A.; Garfinkel, D.; Palan, R. P.;
Basu, J.; Romney, S. L. Preeclampsia and antioxidant nutrients:
decreased plasma levels of reduced ascorbic acid, a-tocopherol,
and h-carotene in women with preeclampsia. Am. J. Obstet. Gy-
necol. 171:150 157; 1994.
[19] Davidge, S. T.; Hubel, C. A.; Brayden, R. D.; Capeless, E. C.;
McLaughlin, M. K. Sera antioxidant activity in uncomplicated
and preeclamptic pregnancies. Obstet. Gynecol. 79:897 901;
1992.
[20] Morris, J. M.; Gopaul, N. K.; Endresen, M. J.; Knight, M.;
Linton, E. A.; Dhir, S.; Anggard, E. E.; Redman, C. W. Circu-
lating markers of oxidative stress are raised in normal pregnancy
and pre-eclampsia. Br. J. Obstet. Gynaecol. 105:1195 1199;
1998.
[21] Diedrich, F.; Renner, A.; Rath, W.; Kuhn, W.; Wieland, E. Lipid
hydroperoxides and free radical scavenging enzyme activities in
preeclampsia and HELLP (hemolysis, elevated liver enzymes,
and low platelet count) syndrome: no evidence for circulating
primary products of lipid peroxidation. Am. J. Obstet. Gynecol.
185:166 172; 2001.
[22] Regan, C. L.; Levine, R. J.; Baird, D. D.; Ewell, M. G.; Martz,
K. L.; Sibai, B. M.; Rokach, J.; Lawson, J. A.; FitzGerald, G. A.
No evidence of lipid peroxidation in severe preeclampsia. Am. J.
Obstet. Gynecol. 185:572 578; 2001.
[23] Wang, Y.; Walsh, S. W. Placental mitochondria as a source of
oxidative stress in pre-eclampsia. Placenta 19:581 586; 1998.
[24] Chappell, L. C.; Seed, P. T.; Briley, A. L.; Kelly, F. J.; Lee, R.;
Hunt, B. J.; Parmar, K.; Bewley, S. J.; Shennan, A. H.; Steer, P. J.;
Poston, L. Effect of antioxidants on the occurrence of pre-eclamp-
sia in women at increased risk: a randomised trial. Lancet
354:810 816; 1999.
[25] Chappell, L. C.; Seed, P. T.; Kelly, F. J.; Briley, A.; Hunt, B. J.;
Charnock-Jones, D. S.; Mallet, A.; Poston, L. Vitamin C and E
supplementation in women at risk of preeclampsia is associated
with changes in indices of oxidative stress and placental function.
Am. J. Obstet. Gynecol. 187:777 784; 2002.
[26] Gratacos, E.; Casals, E.; Sanllehy, C.; Cararach, V.; Alonso, P. N.;
Fortuny, A. Variation in lipid levels during pregnancy in women
with different types of hypertension. Acta Obstet. Gynecol. Scand.
75:896 901; 1996.
[27] Hubel, C. A.; Lyall, F.; Weissfeld, L.; Gandley, R. E.; Roberts, J.
M. Small low-density lipoproteins and vascular cell adhesion
molecule-1 are increased in association with hyperlipidemia in
preeclampsia. Metabolism 47:1281 1288; 1998.
[28] Sattar, N.; Bendomir, A.; Berry, C.; Shepherd, J.; Greer, I. A.;
Packard, C. J. Lipoprotein subfraction concentrations in pree-
clampsia: pathogenic parallels to atherosclerosis. Obstet. Gynecol.
89:403 408; 1997.
[29] Fukunaga, K.; Suzuki, T.; Takama, K. Highly sensitive high-per-
formance liquid chromatography for the measurement of malon-
dialdehyde in biological samples. J. Chromatogr. 621:77 81;
1993.
[30] Domnguez, C.; Ruiz, E.; Gussinye, M.; Carrascosa, A. Oxidative
stress at onset and in early stages of type 1 diabetes mellitus in
children and adolescents. Diabetes Care 21:1736 1742; 1998.
[31] Richard, M. J.; Guiraud, P.; Meo, J.; Favier, A. High-performance
liquid chromatographic separation of malondialdehyde thiobar-
bituric acid adduct in biological materials (plasma and human
cells) using a commercially available reagent. J. Chromatogr.
577:9 18; 1992.
[32] Yagi, K. Assay for blood plasma or serum. Methods Enzymol.
105:328 331; 1984.
[33] Wasowicz, W.; Ne`ve, J.; Peretz, A. Optimized steps in fluoromet-
ric determination of thiobarbituric acid-reactive substances in se-
rum: importance of extraction pH and influence of sample
preservation and storage. Clin. Chem. 39:2522 2526; 1993.
569
[34] Nourooz-Zadeh, J.; Tajaddini-Sarmandi, J.; McCarthy, S.; Better-
idge, J.; Wolff, S. P. Elevated levels of authentic plasma hydro-
peroxides in NIDDM. Diabetes 44:1054 1058; 1995.
[35] Hu, M. L. Measurement of protein thiol groups and glutathione in
plasma. Methods Enzymol. 233:380 385; 1994.
[36] Witko-Sarsat, V.; Friedlander, M.; Cape`illere-Blandin, C.;
Nguyen-Khoa, T.; Nguyen, A. T.; Zingraff, J.; Jungers, P.;
Descamps-Latscha, B. Advanced oxidation protein products as
a novel marker of oxidative stress in uremia. Kidney Int. 49:
1304 1313; 1996.
[37] Levine, R. L.; Williams, J. A.; Stadtman, E. R.; Shacter, E. Car-
bonyl assays for determination of oxidatively modified proteins.
Methods Enzymol. 233:346 363; 1994.
[38] Martin-Gallan, P.; Gussinye, M.; Carrascosa, A.; Domnguez, C.
Biomarkers of diabetes-associated oxidative stress and antioxidant
status in young diabetic patients with or without subclinical com-
plications. Free Radic. Biol. Med. 34:1563 1574; 2003.
[39] Bradford, M. M. A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of pro-
tein-dye binding. Anal. Biochem. 72:248 254; 1976.
[40] McCord, J. M.; Fridovich, I. Superoxide dismutase. An enzymatic
function for erythrocuprein (hemocuprein). J. Biol. Chem. 244:
6049 6055; 1969.
[41] Mahoney, J. J.; Vreman, H. J.; Stevenson, D. K.; Van Vessel, A. L.
Measurements of carboxyhaemoglobin by five spectrophotome-
ters (co-oximeters) in comparison with reference methods. Clin.
Chem. 39:1693 1700; 1993.
[42] Paglia, D. E.; Valentine, W. N. Studies on the quantitative and
qualitative characterization of erythrocyte glutathione peroxidase.
J. Lab. Clin. Med. 70:158 169; 1967.
[43] Anderson, M. E. Determination of glutathione and glutathione di-
sulfide in biological samples. Methods Enzymol. 113:548 555;
1985.
[44] Bieri, J. G.; Tolliver, T. J.; Catignari, G. L. Simultaneous determi-
nation of a-tocopherol and retinol in plasma or red cells by high
pressure liquid chromatography. Am. J. Clin. Nutr. 32:2143 2149;
1979.
[45] Jordan, P.; Brubacher, D.; Moser, U.; Stahelin, B.; Gey, K. F.
Vitamin E and vitamin A concentrations in plasma adjusted for
cholesterol and triglycerides by multiple regression. Clin. Chem.
41:924 927; 1995.
[46] Levine, M.; Wang, Y.; Rumsey, S. C. Analysis of ascorbic acid
and dehydroascorbic acid in biological samples. Methods Enzy-
mol. 299:65 76; 1999.
[47] Glazer, A. N. Phycoerythrin fluorescence-based assay for reactive
oxygen species. Methods Enzymol. 186:161 168; 1990.
[48] Kontush, A.; Beisiegel, U. Measurement of oxidizability of blood
plasma. Methods Enzymol. 299:35 49; 1999.
[49] Bowen, R. S.; Moodley, J.; Dutton, M. F.; Theron, A. J. Oxidative
stress in pre-eclampsia. Acta Obstet. Gynecol. Scand. 80:719 725;
2001.
[50] Morrow, J. D.; Roberts, L. J. Mass spectrometric quantification of
F2-isoprostanes in biological fluids and tissues as measure of
oxidant stress. Methods Enzymol. 300:3 12; 1999.
[51] Jain, S. K.; Wise, R. Relationship between elevated lipid perox-
ides, vitamin E deficiency and hypertension in preeclampsia. Mol.
Cell. Biochem. 151:33 38; 1995.
[52] Esterbauer, H.; Lang, J.; Zadravec, S.; Slater, T. F. Detection of
malonaldehyde by high-performance liquid chromatography.
Methods Enzymol. 105:319 328; 1984.
[53] Reinheckel, T.; Noack, H.; Lorenz, S.; Wiswedel, I.; Augustin, W.
Comparison of protein oxidation and aldehyde formation during
oxidative stress in isolated mitochondria. Free Radic. Res. 29:
297 305; 1998.
[54] Stadtman, E. R. Importance of individuality in oxidative stress
and aging. Free Radic. Biol. Med. 33:597 604; 2002.
[55] Zusterzeel, P. L.; Mulder, T. P.; Peters, W. H.; Wiseman, S. A.;
Steegers, E. A. Plasma protein carbonyls in nonpregnant, healthy
pregnant and preeclamptic women. Free Radic. Res. 33:471 476;
2000.
[56] Zusterzeel, P. L.; Rutten, H.; Roelofs, H. M.; Peters, W. H.;
Steegers, E. A. Protein carbonyls in decidua and placenta of
570 E. Llurba et al.

pre-eclamptic women as markers for oxidative stress. Placenta M.; Kashimura, M. Enhanced protein levels of protein thiol/disul-
22:213 219; 2001. phide oxidoreductases in placentae from pre-eclamptic subjects.
[57] Baynes, J. W. Chemical modification of proteins by lipids in Placenta 22:566 572; 2001.
diabetes. Clin. Chem. Lab. Med. 41:1159 1165; 2003. [66] Sawicki, G.; Dakour, J.; Morrish, D. W. Functional proteomics of
[58] Kharb, S.; Gulati, N.; Singh, V.; Singh, G. P. Lipid peroxidation neurokinin B in the placenta indicates a novel role in regulating
and vitamin E levels in preeclampsia. Gynecol. Obstet. Invest. cytotrophoblast antioxidant defences. Proteomics 3:2044 2051;
46:238 240; 1998. 2003.
[59] Gross, M.; Yu, X.; Hannan, P.; Prouty, C.; Jacobs, D. R., Jr.
Lipid standardization of serum fat-soluble antioxidant concentra-
tions: the YALTA study. Am. J. Clin. Nutr. 77:458 466; 2003. ABBREVIATIONS
[60] Tribble, D. L.; Thiel, P. M.; van den Berg, J. J.; Krauss, R. M.
Differing alpha-tocopherol oxidative lability and ascorbic acid AOPP advanced oxidation protein product
sparing effects in buoyant and dense LDL. Arterioscler. Thromb.
Vasc. Biol. 15:2025 2031; 1995. DNPH dinitrophenylhydrazine
[61] Gratacos, E.; Casals, E.; Gomez, O.; Llurba, E.; Mercader, I.; GPx glutathione peroxidase
Cararach, V.; Cabero, L. Increased susceptibility to low density GSH glutathione
lipoprotein oxidation in women with a history of pre-eclampsia.
Brit. J. Obstet. Gynaecol. 110:400 404; 2003. HPx lipid hydroperoxides
[62] Hubel, C. A.; Kagan, V. E.; Kisin, E. R.; McLaughlin, M. K.; ICAM-1 intercellular adhesion molecule-1
Roberts, J. M. Increased ascorbate radical formation and ascor- LPO lipid peroxides
bate depletion in plasma from women with preeclampsia: impli-
cations for oxidative stress. Free Radic. Biol. Med. 23:597 609; MDA malondialdehyde
1997. PCC protein carbonyl content
[63] Sies, H.; Stahl, W.; Sundquist, A. R. Antioxidant functions of PE preeclampsia
vitamins: vitamins E and C, beta carotene, and other carotenoids.
Ann. N. Y. Acad. Sci. 669:7 20; 1992. ROS reactive oxygen species
[64] Tyurin, V. A.; Liu, S. X.; Tyurina, Y. Y.; Sussman, N. B.; Hubel, SOD superoxide dismutase
C. A.; Roberts, J. M.; Taylor, R. N.; Kagan, V. E. Elevated levels TBA thiobarbituric acid
of S-nitrosoalbumin in preeclampsia plasma. Circ. Res. 88:
1210 1215; 2001. TBARS thiobarbituric-acid-reactive substances
[65] Shibata, E.; Ejima, K.; Nanri, H.; Toki, N.; Koyama, C.; Ikeda, VCAM-1 vascular cell adhesion molecule-1