Cell Tissue Res (2001) 306:409–416
DOI 10.1007/s004410100424
REGULAR ARTICLE
A. Burlacu · V. Jinga · A. V. Gafencu · M. Simionescu
Severity of oxidative stress generates different mechanisms
of endothelial cell death
Received: 1 March 2001 / Accepted: 11 May 2001 / Published online: 23 October 2001
© Springer-Verlag 2001
Abstract The role of reactive oxygen species (ROS) in
the pathogenesis of vascular diseases is well established,
but few data exist on the mechanisms by which ROS in-
duce endothelial cell (EC) death. We examined the con-
ditions and the mechanisms by which oxidative stress in-
duces EC death, using cultured confluent bovine aortic
ECs exposed for 30 min to different concentrations of
hydroxyl radicals (HO•) generated by hydrogen peroxide
(H2O2) in the presence of 100 µM ferrous sulfate
(FeSO4). Cell viability assays, Hoechst DNA staining,
TUNEL (TDT-mediated dUTP-biotin nick end-labeling)
analysis, agarose gel electrophoresis and annexin V as-
say were used to determine the effect of HO• on the via-
bility of ECs, and to distinguish between apoptosis and
necrosis. The results showed that at concentrations of up
to 0.1 mM H2O2/FeSO4, the large majority of cells are
viable, except for ~12.5% death, which occurs by apop-
tosis. At a concentration of 0.2 mM H2O2, the cell viabil-
ity is reduced to 66%, while EC apoptosis remained at
comparable values (14%). At high oxidative stress
(0.5 mM H2O2), the cell viability was drastically reduced
(~39%), and the prevalent form of death was necrosis;
apoptosis accounted for only ~17%. Together, these data
indicate that: (1) HO• induce EC death either by apopto-
sis or necrosis and (2) the mechanisms of EC death differ
as a function of the concentration of HO. Thus, the same
insult can cause apoptosis and/or necrosis, as a function
of the intensity rather than the nature of the insult.
Keywords Endothelial cells · Apoptosis · Necrosis ·
Reactive oxygen species · Atherosclerosis · Cell culture ·
Bovine
A. Burlacu · V. Jinga · A.V. Gafencu · M. Simionescu (✉)
Institute of Cellular Biology and Pathology “N. Simionescu”,
8 BP Hasdeu Street, Bucharest 79691, Romania
e-mail: simionescum@simionescu.instcellbiopath.ro
Tel.: +40-1-4110860, Fax: +40-1-4111143
Introduction
Cell death is a natural or induced process. Apoptosis is
considered an active and progressive response to physio-
logical or pathological stimuli (Herget and Kock 1999).
The term was introduced by Kerr et al. (1972) to distin-
guish this special form of cell death from necrosis. Cel-
lular necrosis and apoptosis differ in mechanism, mor-
phology and biochemistry (Schwartzman and Cidlowski
1993).
Apoptosis is characterized by early prominent con-
densation of nuclear chromatin, loss of plasma mem-
brane phospholipid asymmetry, activation of specific
proteases and endonucleases, enzymatic cleavage of the
DNA into oligonucleosomal fragments, segmentation of
the cell into membrane-bound “apoptotic bodies” and fi-
nally engulfment and degradation of the apoptotic bodies
by neighboring cells. Consequently, the apoptotic bodies
are degraded intracellularly and thus the cell dying by
apoptosis causes minimal disturbance to the surrounding
tissue (Kerr et al. 1972; Umansky 1996).
By contrast, necrosis is induced by severe environ-
mental disturbances and is characterized by swelling of
the cell cytoplasm, organelles and nuclei, clumping of
the chromatin and early disruption of the plasma mem-
brane (Kerr et al. 1972; Fesus et al. 1991). The rupture
of necrotic cells and the release of lysosomal and other
enzymes into the surrounding media causes further tis-
sue destruction and inflammation (Lee and Shacter
1999). Because the two forms of cell death elicit distinct
mechanisms and opposite consequences on the patho-
physiology of the diseases, it is relevant to establish and
clearly distinguish between necrotic and apoptotic cell
death at the level of a lesion, such as the atherosclerotic
plaque.
The endothelium constitutes the inner lining of the
entire cardiovascular system, having a large array of
functions, among which are control of the vascular tone,
provision of an antithrombotic surface, a regulatory role
in transcytosis of plasma proteins, and maintenance of
plasma homeostasis (Simionescu and Simionescu 1991).
410
Therefore, it is safe to assume that death or injury of the
vascular endothelial cells (accompanied by loss of EC
regulatory functions) is a critical event in complicated
vascular diseases.
Recent data suggest that vascular endothelium is sen-
sitive to damage by reactive oxygen species (ROS) that
ultimately leads to cell death (Lee and Shacter 1999).
ROS include a group of molecules such as singlet oxy-
gen (•O2), hydrogen peroxide (H2O2), superoxide anion
(O2–•) and hydroxyl radicals (HO•), which are generated
under a variety of conditions in vivo such as acute and
chronic inflammation (Finkel 1998), atherosclerosis, and
diabetes (Simionescu et al. 1997). Possible sources of
ROS include cigarette smoke, oxidized lipoproteins and
activated polymorphonuclear leukocytes or monocytes
(de Bono and Yang 1995). Reactivity of either H2O2 or
O2–• with other molecules is rather low, but in the pres-
ence of trace amounts of a transition metal they are con-
verted to HO• via the Fenton reaction (Kocha et al.
1997). Clinical studies as well as experimental evidence
suggest a causal pathophysiologic role of increased oxi-
dative stress in the progression of atherosclerosis
(Dimmeler et al. 1998).
Since ROS may induce EC death by either apoptosis
or necrosis, and these two processes have different con-
sequences, the elucidation of the mechanisms of EC
death may be relevant for late atherosclerotic events such
as plaque rupture and thrombosis.
We present here evidence that oxidative stress (HO•
generated by H2O2 in the presence of FeSO4) affects en-
dothelial cell viability and that the mechanism of cell
death – apoptosis and necrosis – varies according to the
intensity of the insult.
Materials and methods
Materials
All chemicals were from Sigma Chemical Co. (St. Louis, MO),
unless otherwise stated. Hydrogen peroxide was purchased from
Riedel-deHaën (Deisenhofen, Germany), Hoechst 33258 was from
Molecular Probes Inc. (Eugene, OR), the apoptosis detection
system was from Promega (Mannheim, Germany) and the TACS
annexin V-biotin apoptosis detection kit consisting of binding
buffer, annexin V-biotin and propidium iodide was from R&D
Systems (Minneapolis, USA).
Endothelial cell culture
ECs were isolated from bovine aorta by collagenase digestion as
described (Jinga et al. 1986). Cells were maintained in culture me-
dium (Dulbecco’s modified Eagle’s medium supplemented with
10% fetal calf serum, 100 U/ml penicillin, 100 µg/ml streptomycin
and 50 µg/ml neomycin). Confluent cells displayed the character-
istic cobblestone morphology, expressed von Willebrand factor
and had the capacity to take up acetylated low-density lipoprotein.
The cells were used at the 4th to 9th passages.
Exposure of EC to oxidative stress
Quiescent confluent cells obtained by incubation for 24 h in se-
rum-free medium were incubated in Hanks’ balanced saline solu-
tion (HBSS) containing various concentrations of H2O2 and a
fixed concentration of 100 µM ferrous sulfate (H2O2/FeSO4). Dif-
ferent dilutions of H2O2 were obtained from a 30% stock solution,
the concentration of which was determined by using the potassium
permanganate titration technique, bearing in mind the report
that 1 ml 0.1 N KMnO4 is equivalent to 1.7 mg H2O2 (Bernt and
Bergmeyer 1974). Briefly, 50 ml ~0.5% H2O2 was mixed with
5 ml 1 N H2SO4 and titrated with 0.1 N KMnO4 solution to a per-
manent pink color. After incubation for 30 min at 37°C with vari-
ous concentrations of H2O2 (0.02–0.5 mM H2O2/FeSO4), ECs
were washed with HBSS and further incubated in freshly prepared
culture medium for 24 h. Controls consisted of incubation of the
cells in the same conditions except that H2O2 was omitted.
Cell viability assays
Confluent ECs grown in 96-well tissue culture plates were incu-
bated with 0.1, 0.2, and 0.5 mM H2O2/FeSO4 as described above.
After 24 h, the EC viability was assessed using MTT (dimethylthi-
azol-diphenyltetrazolium bromide), a water-soluble tetrazolium
salt, which is converted to an insoluble purple formazan by cleav-
age of the tetrazolium ring by mitochondria dehydrogenase en-
zymes from the live cells. This water-insoluble formazan (solubi-
lized using isopropanol) is measured spectrophotometrically,
yielding absorbency as a function of its concentration. The absorp-
tion at 540 nm with a background subtraction at 690 nm was mea-
sured and the results were expressed as a percentage of the con-
trol. MTT assay was applied for each concentration, and the
probes were used in triplicate. For comparison of the data, cell vi-
ability was determined by trypan blue exclusion assay.
Hoechst-DNA staining
ECs grown on slides were incubated with 0.1, 0.2 and 0.5 mM
H2O2/FeSO4 as described above. After 24 h, the cells were ex-
posed to 0.1 µg/ml Hoechst 33258 in phosphate-buffered saline
(PBS) for 15 min at 25°C, washed and then examined with an epi-
fluorescence microscope (Nikon Microphot-SA) to distinguish the
apoptotic cells by their fragmented and condensed nuclei.
Agarose gel electrophoresis
Cultured ECs exposed to H2O2/FeSO4 were mechanically de-
tached from the culture plates with a rubber policeman and mixed
with non-adherent cells collected by centrifugation of the culture
medium. After two washes with cold PBS, cellular DNA was ex-
tracted (as described by Lizard et al. 1996). Briefly, after 18 h of
cell lysis at 37°C in lysis buffer (10 mM TRIS-HCl, pH 8.2, with
35 mM SDS 10 mM EDTA, 40 mM NaCl, and 1 mg/ml proteinase
K), DNA was precipitated with 2 vol 100% ethanol for 24 h at
–20°C, resuspended in TE buffer (10 mM TRIS-HCl, 0.2 mM
EDTA, pH 7.5) and quantified by spectrophotometry at 260 nm.
The DNA samples were applied on 1.8% agarose gel prepared in
TBE buffer (80 mM TRIS borate, pH 8, 2 mM EDTA) containing
0.1 µg/ml ethidium bromide for 18 h at 20 V. DNA standards
ranged from 100 to 2072 bp. Gels were examined under UV light
and photographed.
TUNEL analysis
The nuclei containing abundant 3′-HO• termini were detected us-
ing the TUNEL (TdT-mediated dUTP-biotin nick end-labeling) kit
according to the manufacturer’s instructions. Briefly, the cells
grown on slides and treated with H2O2/FeSO4 as above were incu-
 411
bated in 4% formaldehyde solution for 25 min on ice, and then
permeabilized in 0.2% Triton X-100 in PBS for 5 min at 4°C.
TUNEL reaction mixture containing both terminal transferase
(TdT) and fluorescein isothiocyanate (FITC)-conjugated dUTP
was added to cells and kept for 60 min at 37°C in a humidified
chamber. Immersing the slides in sodium saline citrate (SSC) solu-
tion stopped the reaction; propidium iodide was used to counter-
stain the cells. The negative controls were done in the absence of
the TdT enzyme in the reaction mixture. The positive controls
were obtained by covering the cells with 1 µg/ml DNase I for
10 min at 25°C after Triton X-100 treatment. The probes were ex-
amined using epifluorescence microscopy and at least 1000 cells
from randomly selected fields were counted to determine the per-
centage of apoptotic cells.

Annexin V-FITC assay

The early stage of apoptosis was detected using the TACS annexin Fig. 1 Concentration-dependent effect of H2O2/FeSO4-generated
V-biotin apoptosis detection assay (according to the manufactur- HO• on the viability of cultured aortic ECs. Cells were incubated
er’s instructions). Briefly, the ECs grown on slides were treated for 30 min with various concentrations of H2O2 in the presence of
with H2O2/FeSO4 (as above) and 24 h later were washed with cold 100 µM FeSO4 and after 24 h trypan blue exclusion and MTT as-
PBS containing Ca2+ and Mg2+. Then the cells were incubated for says were performed. The values obtained for each condition (done
15 min at 22–25°C with annexin V incubation reagent (containing in triplicate) indicate the percentage of viable cells as compared to
1 µg/ml annexin V-biotin and 5 µg/ml propidium iodide). After the baseline condition (cells incubated in the absence of H2O2)
three washes with the binding buffer (10 mM HEPES, pH 7.4,
150 mM NaCl, 5 mM KCl, 1 mM MgCl2, 1.8 mM CaCl2), the
cells were further incubated with extravidin-FITC for 15 min at
22°C. Mounted cells were examined using epifluorescence mi-
croscopy.

Results
Effect of HO• on viability of EC

The viability of EC treated with H2O2 in the presence of

100 µM FeSO4 was assessed by two methods, the MTT
and trypan blue exclusion assays, and the results ob-
tained were correlated (Fig. 1). The first procedure is
based on the conversion of MTT to formazan by living
cells only, due to the action of the mitochondrial dehy-
drogenases. The trypan blue method used as an alterna-
tive for evaluation of cell viability is based on the uptake
of trypan blue by dead cells only; this staining also facil-
itates the visualization of cell morphology.
The data obtained with the two methods, recorded as
a function of the H2O2 concentration, are illustrated in
Fig. 1. The viability of EC exposed to 0.1 mM H2O2 was
~90% (compared with untreated cells), suggesting that
this concentration is not cytotoxic to cells. Cell viability
decreased by increasing concentration of H2O2, being
particularly significant at 0.5 mM H2O2 (Fig. 1).
To evaluate by an alternative method the effect of
various concentrations of H2O2 (0.1–0.5 mM)/100 µM Fig. 2 Hoechst-DNA staining followed by fluorescence microsco-
FeSO4 on the viability of the adherent remaining cells, py of nuclei of ECs exposed in culture to 0 (control), 0.1, 0.2 and
Hoechst nuclear staining followed by fluorescence mi- 0.5 mM H2O2/FeSO4 (a–d, respectively) and the corresponding
phase-contrast microscopy (e–h). Compared to control (a, e), as a
croscopy was employed. In this method, the intensity of function of H2O2 concentration, the number of attached ECs de-
nuclear staining depends on the condensation level of the creases (right panel); moreover, of the remaining cells, the number
DNA and/or the membrane permeability. Control cells of fluorescent nuclei increases (b–d). Note that at a concentration
exhibited a weak nuclear fluorescence indicating the of 0.5 mM H2O2/FeSO4, the number of adherent cells is drastically
reduced (h) and all exhibit fluorescent nuclei (d). The nuclear
presence of normal, uncondensed DNA and intact cell staining may be due either to the presence of condensed chromatin
membrane (Fig. 2a). The viability decreased with the in- (indicative of apoptosis) or to the enhanced membrane permeabili-
creased concentrations of H2O2; in consequence, the per- ty (characteristic of necrosis). Bar 40 µm
412
Fig. 3 Fragmentation of genomic DNA (DNA laddering) of cul-
tured ECs exposed to H2O2/FeSO4. Gel electrophoresis of DNA
isolated from cells after exposure to increasing concentrations of
H2O2 (0.02, 0.05, 0.1, 0.2 and 0.5 mM, lanes 2–6, respectively) in
the presence of 100 µM FeSO4. As compared to control (cells ex-
posed to FeSO4, only, lane 1), concentrations up to 0.1 mM
H2O2/FeSO4 induce a gradual increase in DNA laddering (lanes
2–4). At concentrations higher than 0.1 mM H2O2, the DNA
laddering was almost absent (lanes 5, 6). The gel pattern is repre-
sentative of four separate experiments. Lane 7 depicts the 100 to
2072 bp standard DNA
centage of Hoechst-stained nuclei in the remaining ad-
herent cells was enhanced (Fig. 2b–d). Phase-contrast
microscopy also showed the diminishing number of ad-
herent ECs with the increased H2O2 concentration
(Fig. 2e–h). These results correlated well with those ob-
tained by the above two assays, indicating a dose-depen-
dent effect of H2O2/FeSO4 on cellular viability and nu-
clear condensation.
Detection of the DNA fragmentation
A characteristic biochemical feature of apoptosis is the
increased endogenous endonuclease activity that is asso-
ciated with DNA cleavage. DNA staining assay records
indiscriminately both necrotic cells (disrupted cell mem-
brane) and apoptotic cells (increased DNA condensa-
tion).
In order to distinguish particularly apoptosis of EC
exposed to oxidative stress, two specific techniques,
DNA laddering and the TUNEL assay, were employed.
The appearance as a ladder of nucleosomal DNA frag-
ments in agarose gel is an accepted hallmark of apopto-
sis. Upon isolation from ECs incubated with 0.02, 0.05,
and 0.1 mM H2O2/FeSO4, DNA appeared on agarose
gels as a ladder, which was more pronounced with the
increasing concentration of H2O2 (Fig. 3, lanes 2–4, re-
spectively). This indicated a rise in the number of apo-
ptotic cells as a function of the severity of the oxidative
conditions; control ECs (maintained in normoxic condi-
tions) exhibited only a basal level of DNA laddering
(Fig. 3, lane 1). Interestingly, internucleosomal DNA
fragmentation appeared to be dose dependent only in the
Fig. 4 Quantitative determination by TUNEL assay of apoptotic
endothelial cells still adherent after exposure to increasing concen-
trations of H2O2/FeSO4. After TUNEL assay, the cells were count-
erstained with propidium iodide (red staining) and counted in ran-
dom fields. The diagram presents the percentage of apoptotic cells
determined by dividing the number of positive-TUNEL cells
(green fluorescent nuclei) by the total number of cells (stained
red); the average (mean ± SEM) of three different experiments is
presented. A minimum of 1000 cells per assay was counted. The
upper panel shows fluorescence microscopy of TUNEL reaction:
left the positive control, center the negative control, right cells ex-
posed to mild concentration (0.05 mM) H2O2/FeSO4. Note that in
these mild oxidative conditions only two apoptotic cells are dis-
played (arrows)
range of 0.02–0.1 mM H2O2/FeSO4; less or no DNA
laddering was observed when ECs were treated with
higher H2O2 concentrations, such as 0.2 and 0.5 mM
(Fig. 3, lanes 5, 6, respectively). This suggests that high
doses of H2O2/FeSO4 generated an increase in cell death
by necrosis rather than apoptosis since, during necrosis,
nucleases and proteases are activated simultaneously
and, thus, the DNA degradation products are usually
mono- and oligonucleotides.
Quantification of DNA fragmentation was performed
by TUNEL assay of control and H2O2/FeSO4 exposed
cells. The data obtained show that in normoxic condi-
tions, i.e., H2O2-untreated ECs, ~2% of nuclei are
stained by TUNEL reaction (Fig. 4), indicating basal ap-
optosis due to culture manipulation. As compared with
untreated cells, exposure of ECs to H2O2 (up to 0.1 mM)
in the presence of FeSO4 for 30 min and further incuba-
tion for 24 h in normal culture medium was followed by
a proportionate increase in the percentage of apoptosis.
At 0.05 mM H2O2, the TUNEL-positive cells illustrated
in Fig. 4, upper level, right (arrows), represented
7.5±0.5% of the remaining adherent cells, whereas at
0.1 mM H2O2 this figure increased to 12.8±1.2%. In

Fig. 5a–m Fluorescence microscopy of annexin V binding to ECs
at 4 h, and 24 h after exposure to H2O2. a–d Cells were exposed
for 30 min to H2O2 (0.05, 0.1, and 0.2 mM in b–d, respectively) in
the presence of 100 µM FeSO4, and after 4 h, stained with annexin
V (as described under “Materials and methods”). In the control,
cells were incubated with FeSO4 only (a). Note that even in this
short time binding of annexin V occurs, and increases as a func-
tion of H2O2 concentration (b–d). e–m Cells were treated as above
except that binding of annexin V was assessed after 24 h. The con-
centration of H2O2 (in mM) was: 0.02 (g), 0.05 (h, i), 0.1 (j), 0.2
(k, l), 0.5 (m). The staining reveals increased binding of annexin
V with the increase in H2O2 concentration. Double staining of EC
with propidium iodide (e, h, k) and annexin V (f, i, l) illustrates
the presence of necrotic cells at concentrations starting with
0.2 mM H2O2 (k, l). Control experiments are shown in e and f.
Bar 40 µm
Fig. 4, the pictures illustrate: the positive control for
TUNEL reaction (DNase-treated EC), where all cell nu-
clei exhibited a green fluorescence as a result of DNA
fragmentation (upper, left), the negative control for
TUNEL reaction (central picture), and the few apoptotic
cells that result from the mild oxidative condition (right
picture). Although there is an apparent increase in
TUNEL-positive cells at 0.5 mM H2O2 as compared to
0.1 H2O2 (Fig. 4), the figures are not significant when
considering that the concentration of H2O2 was 5 times
higher. As already reported (Kockx and Herman 2000),
the TUNEL assay is not without pitfalls and needs to be
combined with additional techniques. In our experi-
ments, the data obtained by the laddering technique
(Fig. 3) and double staining with annexin V and prop-
idium iodide (Fig. 5) suggest that the decrease in EC via-
413
bility in strong oxidant conditions was the result of ne-
crotic rather than apoptotic cell death.
Translocation of phosphatidylserine from
the inner leaflet to the outer face of the plasmalemma
It was demonstrated that, in most cell types, the initiation
of apoptosis is accompanied by the translocation of
phosphatidylserine (PS) from the inner leaflet to the out-
er surface of the plasmalemma; typically, PS externaliza-
tion precedes formation of membrane blebs and DNA
fragmentation (van Engeland et al. 1998; Zhang et al.
1997; Koopman et al. 1994).
Since PS can be reliably detected with annexin V
(a 38-kDa protein that naturally binds PS), we employed
a biotin conjugate of annexin V that was applied to
ECs exposed to oxidative stress to determine whether
PS is exposed to the surface of EC plasmalemma. The
cells were incubated with various concentrations (0.02–
0.5 mM) of H2O2/FeSO4 and the annexin V binding was
monitored by fluorescence microscopy at different time
intervals (as described under “Materials and methods”).
After 4 h, the cell fluorescence was almost absent in con-
trol cells; in the case of the ECs incubated in oxidative
conditions, the fluorescence started to be noticeable and
increased slightly as a function of H2O2 concentration
(Fig. 5a–d).
Stronger staining, indicating enhanced binding of an-
nexin V to PS translocated to the plasma membrane out-
er leaflet, was observed 24 h after cell treatment; as ob-
414
served at 4 h, the binding of annexin V occurs in a dose-
dependent manner (Fig. 5h–m). When the cells were
double stained concomitantly with fluorescent annexin
and propidium iodide, at 0.05 mM H2O2 the ECs exhibit-
ed green fluorescence, indicative of apoptosis and no
staining of propidium iodide (red), denoting the absence
of necrotic cells (Fig. 5f, i). Interestingly, the same pro-
cedure applied on cells exposed to a concentration of
H2O2 above 0.1 mM, and in particular at 0.2 mM,
showed that the fluorescence of annexin was accompa-
nied by significant propidium iodide staining (Fig. 5g, j).
These data corroborate well with the results obtained
with the above techniques and indicate that high oxida-
tive stress (beginning with 0.2 mM H2O2) induces EC
necrosis as the predominant form of cell death.
Discussion
It is generally accepted that, in normal conditions, endo-
thelial cells are remarkably quiescent; their low turnover
can be safely associated with reduced apoptosis. In nor-
mal conditions, apoptosis may contribute to the stability
of blood vessel wall. However, in lesion-prone areas of
the arterial segment of the vasculature, ECs are charac-
terized by an increased turnover, suggesting a mechanis-
tic link between the cell turnover and the susceptibility
to atherosclerotic plaque development (Caplan and
Schartz 1973).
In these locations, increased EC turnover may be a
consequence of accelerated apoptosis that may alter the
normal functions of the endothelium, thus promoting
atherosclerosis. Apoptotic ECs were detected on the lu-
minal surface of atherosclerotic coronary vessels, but
not in normal vessels (Kockx and Herman 1998). The
level of apoptotic cell death was found to be strongly re-
lated to the stage of the atherosclerotic plaque develop-
ment: intimal thickening and fatty streaks exhibited
small apoptotic cells, whereas advanced atherosclerotic
plaques showed foci of apoptosis (Kockx and Herman
1998).
Endothelial cell death can be induced by numerous
insults to the cells of the vascular wall and plays a criti-
cal role in myocardial infarction, heart failure and
plaque rupture in atherosclerosis (Rubanyi 1993). Inju-
ry or loss of the endothelium implies the deprivation of
the numerous regulatory functions fulfilled by these
cells.
The factors that induce EC death and whether apop-
tosis is beneficial or dangerous for the development of
atherosclerosis are still controversial (Kockx and
Herman 2000). It has been reported that high glucose
concentrations mimicking the diabetic situation trigger
EC apoptosis (Baumgartner-Parzer et al. 1995). Oxi-
dized low-density lipoprotein, which plays a key role in
atherosclerosis, has been shown to stimulate the endoge-
nous suicide cell death program of EC via activation of
caspases (Dimmeler et al. 1997; Harada-Shiba et al.
1998).
Growing evidence indicates an important role for
ROS in hypertension, restenosis after balloon angioplas-
ty and atherosclerosis (Diaz et al. 1997). However, little
and contradictory information exists about the mecha-
nisms by which ROS elicit their effect on the structure
and function of the cells of the blood vessels. It was re-
ported that exposure of smooth muscle cells (SMC) to
superoxide anion leads to their proliferation, an impor-
tant event in plaque formation and hypertension (Baas
and Berk 1995). Also, it was demonstrated that H2O2 in-
duces vascular SMC death, which may occur by apopto-
sis (Li et al. 1997, 1999; Fiorani et al. 1995). Although
for many years necrotic vascular SMC has been detected
in atherosclerotic plaques, it has recently become evident
that apoptosis of SMC occurs as well. Fewer data exist
on the correlation between the ROS status and endotheli-
al cell death.
We questioned whether the hydroxyl radicals affect
directly the ECs, whether their effect is dependent on
concentration and whether this insult may promote cell
death by apoptosis and/or necrosis. To this purpose, cul-
tured ECs were exposed to oxidative stress for different
time intervals and concentrations, and various techniques
that differentiate between apoptosis and necrosis were
used.
For mimicking oxidative stress, we chose hydrogen
peroxide, which, in the presence of ferrous sulfate, gen-
erates HO• via the Fenton reaction:
M(n–1) + H2O2 → Mn + HO– + HO• (in which M is iron).
The results showed that exposure of cultured ECs to
H2O2 in the presence of 100 µM FeSO4 led to a dose-
dependent decrease in cell viability (as assessed by MTT
and trypan blue exclusion assays). Concentrations up to
0.1 mM H2O2 did not affect dramatically the cell viabili-
ty whereas higher concentrations decreased the cell via-
bility, reaching ~50% of the control value at 0.5 mM
H 2O 2.
In experiments in which the cells were exposed to
H2O2/FeSO4 and the unfixed remaining adherent cells
were stained with Hoechst, it was found that as a func-
tion of severity of the oxidative treatment the number
of adhering cells decreased, whereas the percentage of
Hoechst-stained cells increased. Since this method
indicates indiscriminately either high membrane perme-
ability or increased nuclear condensation, the intense
Hoechst-staining denotes cells undergoing either a ne-
crotic (disrupted cell membrane) or an apoptotic process
(increased DNA condensation).
Cell apoptosis can be assessed chiefly by the presence
of nuclear DNA fragmentation, a commonly accepted
key marker of this process. Using TUNEL assay it was
found that at high H2O2 concentrations such as 0.2 and
0.5 mM cell death by apoptosis represents 14% and
18%, respectively. Since in strong oxidative conditions
(i.e., 0.5 mM H2O2), the decrease in EC viability is 50%
(MTT assay) and the apoptosis represents ~18%, one can
safely assume that the predominant form of cell death is
necrosis. This assumption is validated by the DNA

laddering assay that showed that the DNA fragmentation
was significantly diminished or even absent for the cells
treated with 0.2 and 0.5 mM H2O2 as compared with
those exposed to mild oxidative conditions (0.02–
0.1 mM H2O2).
As an additional control, we applied the same assays
on ECs incubated with H2O2 in the absence of ferrous
sulfate; in these conditions, cellular apoptosis (by
TUNEL and DNA laddering assays) was not detected
except for concentrations above 0.2 mM H2O2 (data not
shown). Moreover, in experiments in which catalase
(500 U/ml) was added to cells exposed to H2O2/FeSO4,
we found that the former abolished completely the effect
of 0.1 mM H2O2/FeSO4 on the caspase-3 activation (un-
published data). This result indicated that hydrolyses of
H2O2 by catalase prevented the effect of H2O2 or the
Fenton reaction from occurring.
These results are backed and corroborate well with
the data obtained by alternative methods used to detect
simultaneously translocation of phosphatidylserine from
the inner to the outer leaflet of the plasma membrane (an
early event in apoptosis) and the cellular incorporation
of propidium iodide (indicative of necrosis). Staining
with fluorescent annexin V and propidium iodide indi-
cated that up to 0.1 mM H2O2 the EC death occurs pre-
dominantly by apoptosis in a concentration-dependent
manner; above this concentration, the necrosis becomes
the predominant form of cell death.
Although there is some variation within each assay
used, when the results of the three methods are taken to-
gether (TUNEL, DNA laddering and double staining
with annexin V and propidium iodide), they indicate that
the intensity of the oxidative stress determines the path-
way of cell death.
It has been postulated that the EC response to hyperli-
pemia and increased oxidative stress is the successive
and gradual adaptation of cellular constitutive functions
(modulation), followed by dysfunction, injury and cell
death (Simionescu et al. 1997). Based on the data report-
ed here, one can hypothesize that, as a function of the se-
verity of the oxidative stress, ECs are initially affected
by apoptosis after which, gradually, necrosis becomes
the prevalent process of cell death contributing to athero-
sclerosis complications, such as vascular thrombosis and
plaque rupture.
Together, these data indicate that (1) as a function of
the concentration, hydroxyl radicals may induce EC
death either by apoptosis or necrosis and (2) the same in-
sult can cause different mechanisms of cell death, de-
pending on the intensity rather than its nature.
References
Baas AS, Berk BC (1995) Differential activation of mitogen-acti-
vated protein kinases by H2O2 and –O2• in vascular smooth
muscle cells. Circ Res 77:29–36
Baumgartner-Parzer S, Wagner L, Pettermann M, Grillari J, Gessl
A, Waldhausl W (1995) High-glucose-triggered apoptosis in
cultured endothelial cells. Diabetes 44:1323–1327
415
Bernt E, Bergmeyer HU (1974) Inorganic peroxides. In: Bergmey-
er H (eds) Methods in enzymatic analysis, 4. Academic Press,
New York, pp 2247–2249
Caplan BA, Schartz CJ (1973) Increased endothelial cell turnover
in areas of in vivo Evans blue uptake in the pig aorta. Athero-
sclerosis 17:401–410
de Bono DP, Yang WD (1995) Exposure to low concentrations of
hydrogen peroxide causes delayed endothelial cell death and
inhibits proliferation of surviving cells. Atherosclerosis 114:
235–245
Diaz MN, Frei B, Vita JA, Keaney J Jr (1997) Antioxidants and
atherosclerotic heart disease. N Engl J Med 95:1062–1071
Dimmeler S, Haendeler J, Galle J, Zeiher AM (1997) Oxidized
low-density lipoprotein induces apoptosis of human endotheli-
al cells by activation of CPP32-like proteases: a mechanistic
clue to the response to injury hypothesis. Circulation 95:1760–
1768
Dimmeler S, Hermann C, Zeiher AM (1998) Apoptosis of endo-
thelial cells. Contribution to the pathophysiology of athero-
sclerosis? Eur Cytokine Netw 9:697–698
Fesus L, Davies PJA, Piacentini M (1991) Apoptosis: molecular
mechanisms in programmed cell death. Eur J Cell Biol 56:
170–177
Finkel T (1998) Oxygen radicals and signaling. Curr Opin Cell
Biol 10:248–253
Fiorani M, Cantoni O, Tasinato A, Boscoboinok D, Azzi A (1995)
Hydrogen peroxide- and fetal bovine serum-induced DNA
synthesis in vascular smooth muscle cells: positive and nega-
tive regulation by protein kinase C isoforms. Biochem Bio-
phys Acta 1269:98–104
Harada-Shiba M, Kinoshita M, Kamido H, Shimokado K (1998)
Oxidized low density lipoprotein induces apoptosis in cultured
human vein endothelial cells by common and unique mecha-
nisms. J Biol Chem 273:9681–9687
Herget T, Kock M (1999) Apoptosis – a double-edged sword. BIF
Futura 14:173–185
Jinga V, Bogdan I, Fruchter J (1986) Experimental model for the
quantitative estimation of transendothelial transport in vitro; a
two compartment system. Rev Roum Morphol Physiol 23:
227–230
Kerr JF, Wyllie AH, Currie AR (1972) Apoptosis: a basic biologi-
cal phenomenon with wide range implications in tissue kinet-
ics. Br J Cancer 26:239–527
Kocha T, Yamaguchi M, Ohtaki H, Fukuda T, Aoyagi T (1997)
Hydrogen peroxide-mediated degradation of protein: different
oxidation modes of copper- and iron-dependent hydroxyl radi-
cals on the degradation of albumin. Biochim Biophys Acta
1337:319–326
Kockx MM, Herman AG (1998) Apoptosis in atherogenesis: im-
plications for plaque destabilization. Eur Heart J 19(Suppl
G):G23–G28
Kockx MM, Herman AG (2000) Apoptosis in atherosclerosis:
beneficial or detrimental? Cardiovascular Res 45:736–746
Koopman G, Reutelingsperger CP, Kuijten GA, Keehnen RM,
Pals ST, van Oers MH (1994) Annexin V for flow cytometric
detection of phosphatidylserine expression on B cells undergo-
ing apoptosis. Blood 84:1415–1420
Lee Y-J, Shacter E (1999) Oxidative stress inhibits apoptosis in
lymphoma cells. J Biol Chem 274:19792–19798
Li PF, Dietz R, von Harsdorf R (1997) Reactive oxygen species
induce apoptosis of vascular smooth muscle cell. FEBS Lett
404:249–252
Li PF, Maasch C, Haller H, Dietz R, von Harsdorf R (1999) Re-
quirement for protein kinase C in reactive oxygen species-
induced apoptosis of vascular smooth muscle cells. Circula-
tion 100:967–973
Lizard G, Deckert V, Dubrez L, Moisant M, Gambert P, Lagrost L
(1996) Induction of apoptosis in endothelial cells treated with
cholesterol oxides. Am J Pathol 148:1625–1638
Rubanyi GM (1993) The role of endothelium in cardiovascular
homeostasis and diseases. J Cardiovasc Pharmacol 22
(Suppl 4):S1–S14
416
Schwartzman RA, Cidlowski JA (1993) Apoptosis: the biochemis-
try and molecular biology of programmed cell death. Endo-
crine Rev 14:133–151
Simionescu M, Simionescu N (1991) Endothelial transport of
macromolecules; transcytosis and endocytosis. Cell Biol Rev
25:1–80
Simionescu M, Sima A, Popov D (1997) The vascular endotheli-
um in diabetes-induced accelerated atherosclerosis. In: Born
GVR, Schwartz CJ (eds) Vascular endothelium physiology,
pathology, and therapeutic opportunities. Schattauer, Stuttgart,
pp 329–344
Umansky SR (1996) Apoptosis: molecular and cellular mecha-
nisms (a review). Mol Biol 30:285–295
Van Engeland M, Nieland LJ, Ramaekers FC, Schutte B,
Reutelingsperger CP (1998) Annexin V-affinity assay: a re-
view on an apoptosis detection system based on phosphatidyl-
serine exposure. Cytometry 31:1–9
Zhang G, Gurtu V, Kain SR, Yan G (1997) Early detection of ap-
optosis using a fluorescent conjugate of annexin V. Biotech-
niques 23:525–531