Am J Physiol Heart Circ Physiol 284: H549H558, 2003.
First published October 10, 2002; 10.1152/ajpheart.00708.2002.
Cell death during ischemia: relationship to mitochondrial
depolarization and ROS generation
JACQUES LEVRAUT, HIROTARO IWASE, Z.-H. SHAO,
TERRY L. VANDEN HOEK, AND PAUL T. SCHUMACKER
Pulmonary and Critical Care Medicine, University of Chicago, Chicago, Illinois 60637
Submitted 22 August 2002; accepted in final form 7 October 2002
Levraut, Jacques, Hirotaro Iwase, Z.-H. Shao, Terry
L. Vanden Hoek, and Paul T. Schumacker. Cell death
during ischemia: relationship to mitochondrial depolariza-
tion and ROS generation. Am J Physiol Heart Circ Physiol
284: H549H558, 2003. First published October 10, 2002;
10.1152/ajpheart.00708.2002.Ischemia-reperfusion injury
induces cell death, but the responsible mechanisms are not
understood. This study examined mitochondrial depolariza-
tion and cell death during ischemia and reperfusion. Con-
tracting cardiomyocytes were subjected to 60-min ischemia
followed by 3-h reperfusion. Mitochondrial membrane poten-
tial (m) was assessed with tetramethylrhodamine methyl
ester. During ischemia, m decreased to 24 5.5% of
baseline, but no recovery was evident during reperfusion.
Cell death assessed by Sytox Green was minimal during
ischemia but averaged 66 7% after 3-h reperfusion. Cyclo-
sporin A, an inhibitor of mitochondrial permeability transi-
tion, was not protective. However, pharmacological antioxi-
dants attenuated the fall in m during ischemia and cell
death after reperfusion and decreased lipid peroxidation as
assessed with C11-BODIPY. Cell death was also attenuated
when residual O2 was scavenged from the perfusate, creating
anoxic ischemia. These results suggested that reactive oxy-
gen species (ROS) were important for the decrease in m
during ischemia. Finally, 143B-0 osteosarcoma cells lacking
a mitochondrial electron transport chain failed to demon-
strate a depletion of m during ischemia and were sig-
nificantly protected against cell death during reperfusion.
Collectively, these studies identify a central role for mito-
chondrial ROS generation during ischemia in the mitochon-
drial depolarization and subsequent cell death induced by
ischemia and reperfusion in this model.
reactive oxygen species; hypoxia; oxidants
REACTIVE OXYGEN SPECIES (ROS) have been implicated as
participants in the myocardial damage induced by is-
chemia-reperfusion (I/R) (1, 4, 9, 17, 21, 25). Most
studies have focused on the importance of oxidant
stress generated during reperfusion, when a burst of
ROS is generated after oxygen is reintroduced into the
system after a prolonged period of ischemia (32, 39).
However, growing evidence suggests that oxidant
stress begins during ischemia before reperfusion. For
example, in cardiomyocytes subjected to simulated I/R,
Address for reprint requests and other correspondence: P. T. Schu-
macker, Dept. of Medicine MC6026, The Univ. of Chicago, 5841
South Maryland Ave., Chicago, IL 60637 (E-mail: pschumac
@medicine.bsd.uchicago.edu).
http://www.ajpheart.org 0363-6135/03 $5.00 Copyright 2003 the American Physiological Society
we observed (32, 33) an increase in ROS generation
during ischemia followed by a large burst of oxidant
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016
production during the first few minutes after reoxygen-
ation. In that model, antioxidants were more protective
when given throughout the experiment than when
given only at reperfusion, which supports the idea that
oxidants generated during the ischemic phase contrib-
ute to cell injury and are important determinants of
cell survival and recovery of function (2, 35).
ROS generation cannot occur during ischemia unless
some residual O2 is still present. Previous studies us-
ing cardiomyocytes revealed that trace levels of O2 are
still detectable during simulated ischemia (PO2 57
mmHg). During ischemia, indexes of oxidant stress
were attenuated by mitochondrial electron transport
inhibitors, suggesting that the ROS are generated by
mitochondria (2, 33, 34). Collectively, these observa-
tions support the notion that superoxide is generated
during ischemia despite the conditions of low O2 con-
centration ([O2]) (11), and they suggest that these
oxidants may play an important role in determining
cell survival during I/R.
Although previous studies indicate that oxidants
generated during ischemia may contribute to cell dam-
age, the specific mechanism by which these ROS dis-
rupt cellular function is not known. The present study
sought to clarify the physiological consequences of ox-
idants generated during ischemia before reperfusion.
We hypothesized that oxidant stress generated at the
mitochondria during ischemia could contribute to a
loss of mitochondrial membrane potential (m),
which in turn could contribute to the overall cellular
injury and survival.
MATERIALS AND METHODS
Cell culture and perfusion system. Embryonic chick cardi-
omyocytes were prepared as previously described (35) and
were grown on glass coverslips in a humidified incubator.
Experiments were performed on spontaneously contracting
cells at 35 days after isolation, under controlled O2-CO2
conditions at 37C on an inverted microscope. A flow-through
chamber was created by clamping a stainless steel spacer
ring between two coverslips, allowing perfusion of the space
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked advertisement in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
H549
H550 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA
between with buffered salt solutions (BSS; 0.5 ml/min) equil-
ibrated with O2-CO2 gas mixtures in a water-jacketed col-
umn. Stainless steel tubing connecting the column to the
chamber prevented diffusive entry of ambient O2 through the
tubing wall.
I/R model. Cells were equilibrated for 30 min by superfu-
sion with BSS (in mM: 120 NaCl, 18 NaHCO3, 4 KCl, 1
MgSO4, 0.8 NaH2PO4, 1.4 CaCl2, and 5.6 glucose; 5% CO2,
pH 7.35). During simulated ischemia, cells were superfused
with a variation of BSS containing 20 mM 2-deoxyglucose
(2-DOG) to inhibit glycolysis, zero glucose, 8 mM K, 5 mM
lactate, and low [O2] and hypercarbia (pH 6.8) obtained by
bubbling with 80% N2-20% CO2. During hypoxic acidosis,
cells were superfused with BSS bubbled with 80% N2-20%
CO2. In other experiments, complete anoxia was achieved in
the ischemic or hypoxic medium by adding EC-Oxyrase (10
l/ml), an oxidase mixture that reduces O2 to H2O. Anoxia
was confirmed with an optical phosphorescence quenching
method using a porphryin probe in solution to measure PO2
within the perfusion chamber (Oxyspot) (24). After ischemia
or hypoxia/anoxia (1 h), reperfusion was carried out with
normoxic BSS (3 h). Tetramethylrhodamine methyl ester
(TMRE) measurements were obtained during baseline,
throughout ischemia, and during the first 90 min of reperfu-
sion.
Mitochondrial membrane potential. m was assessed
with TMRE. This cationic dye enters the cells and accumu-
lates within mitochondria according to the Nernst equation
(12). Cells were loaded with TMRE (100 nM) at 37C for 45
min; studies were carried out in the continued presence of the
dye (10 nM). Mean fluorescence intensity was measured
every minute (excitation 535 nm, emission 610 nm) for a field
of cells. Under these nonquenching conditions, mitochondrial
depolarization with FCCP causes an immediate decrease in
TMRE fluorescence, whereas oligomycin causes an immedi-
ate increase (6). Fluorescence intensity is expressed as the
percentage of initial brightness after background subtrac-
tion.
Cell viability. Viability was measured in the same field of
cells used to assess m. Dead cells were identified with
Sytox Green, a membrane-impermeant dye that is excluded
from cells when the plasma membrane is intact. In dying
cells with increased plasma membrane permeability, nuclear
fluorescence becomes apparent. To assess cell death in a field
of cells, a fluorescent image (10 objective) was acquired and
the number of fluorescent nuclei was counted (Metamorph;
Universal Imaging). Digitonin (300 M) was added to per-
meabilize all cells in the field. Cell counts were then re-
peated, and the previous counts were normalized to that
value.
Lipid peroxidation assay. Cardiomyocytes were loaded
with C11-BODIPY (10 M), and fluorescence was measured
during baseline, ischemia, and reperfusion. Because of its
lipophilic nature, this fluorophore localizes to cell mem-
branes and can be used to assess oxidative stress in that
environment (27). On oxidation, the fluorescence (excitation
480 nm, emission 525 nm) of C11-BODIPY increases.
Generation of respiration-deficient 0-cells. To clarify the
significance of mitochondrial ROS generation for membrane
damage and cell death during ischemia, mitochondria-defi-
cient cells (0-cells) were generated from wild-type 143B
osteosarcoma cells (American Type Culture Collection) (22).
The 0-cells were generated by incubating rapidly dividing
wild-type cells with ethidium bromide (50 ng/ml), which
inhibits replication of mitochondrial DNA (8). The mitochon-
drial DNA encodes specific subunits that are critical for
electron transport, so 0-cells do not possess a functional
AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003
electron transport system and cannot generate ATP or ROS
in mitochondria. Cells were maintained in medium supple-
mented with pyruvate (2 mM), uridine (50 g/ml), and 5-bro-
mouridine (0.015 mg/ml). Loss of mitochondrial DNA was
confirmed by semiquantitative PCR and by loss of cell viabil-
ity when uridine supplements to the media were withdrawn.
Reagents and analysis. Reagents were obtained from
Sigma, and fluorophores were obtained from Molecular
Probes. Replicate experiments were carried out with sepa-
rate coverslips. Treatment and control group were matched
by using cells isolated on the same day.
RESULTS
To assess changes in m, contracting cardiomyo-
cytes on coverslips were loaded with TMRE, placed in
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016
a flow-through chamber on an inverted microscope,
and superfused with BSS (37C) equilibrated with nor-
moxic (21% O2, 5% CO2) gas. TMRE loading properties
in cardiomyocytes, and the nonquenching characteris-
tics of this fluorophore in assessing membrane poten-
tial under nonquenching conditions were reported pre-
viously (6). Stable levels of fluorescence were typically
observed under baseline conditions (Fig. 1A). This flu-
orescence was rapidly dissipated on addition of the
protonophore FCCP, consistent with the expected loss
of m induced by this uncoupling agent.
To assess the effects of ischemia on m, cells loaded
with TMRE were studied for 30 min under baseline
normoxic conditions followed by 60 min of simulated
ischemia. During ischemia, a progressive decrease in
TMRE fluorescence was observed, reaching 24.0
5.5% of the initial intensity after 1 h. Minimal recovery
of TMRE fluorescence was seen after return to nor-
moxia (reperfusion; Fig. 1B). Cell death within the
same field of cells was minimal (5%) at the end of
ischemia but increased significantly during 3-h reper-
fusion (Fig. 1C). Inspection of TMRE fluorescence im-
ages revealed that the majority of cells lost all fluores-
cence during ischemia and some cells retained some
fluorescence at an attenuated level. Reperfusion was
not associated with significant recovery of fluorescence
in either case.
Mitochondrial depolarization could conceivably be
caused by activation of the mitochondrial permeability
transition (MPT) pore, a high-conductance putative
channel in the inner mitochondrial membrane (10).
The opening of the MPT pore has been suggested to
contribute to the decrease in m during I/R injury
(18). Cyclosporin A inhibits the opening of this pore
and was therefore used to evaluate its contribution to
membrane depolarization and cell death during I/R.
Cyclosporin A (0.2 M) had no significant effect on
membrane depolarization during simulated ischemia
(Fig. 2A) and had no significant effect on cell death
after 3-h reperfusion (Fig. 2B). Additional studies us-
ing a higher concentration (0.5 M) also failed to abol-
ish the fall in TMRE fluorescence (data not shown).
These results suggest that opening of the MPT pore
does not contribute significantly to the depolarization
and cell death in this model.
www.ajpheart.org
 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA H551

Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016

Fig. 2. A: effects of cyclosporin A (CyA) on mitochondrial potential in
cardiomyocytes as assessed with TMRE fluorescence during base-
line, ischemia, and reperfusion in a flow-through chamber. Cyclo-
sporin A was added to the buffer perfusate throughout the experi-
ment. B: cell death in the same experiments as assessed with the
membrane-impermeable probe Sytox Green.

subsequent cell death in this model, pharmacological

antioxidants were added to the perfusate throughout
the experiment and the effects on depolarization and
cell death were assessed (Table 1). The thiol reductants
2-mercaptopropionyl glycine (2-MPG; 400 M) and pyr-
rolidine dithiocarbamate (PDTC, 10 M) significantly

Table 1. Effect of antioxidants on mitochondrial

Fig. 1. A: representative experiment showing tetramethylrhodam- membrane potential at end of ischemia and
ine methyl ester (TMRE) fluorescence under nonquenching condi- cell death after reperfusion
tions to assess mitochondrial membrane potential (m). B: m in
cardiomyocytes as assessed with TMRE fluorescence during base- Fall in m (end Cell Death (end
line, ischemia, and reperfusion in a flow-through chamber. C: cell n of ischemia), % of reperfusion), %
death in the same population of cells assessed with the membrane-
impermeant probe Sytox Green during baseline, simulated ischemia, Control 9 76 5.5 66 7
and reperfusion. (mean values SE, n 9). 2-MPG (0.4 mM) 4 31 13 18 15
NAC (0.5 mM) 4 34 15 15 5
PDTC (10 M) 4 31 10 17 5
We previously found (33) evidence of mitochondrial 1,10-Phenanthroline (10 M) 4 20 10 61
oxidant stress during ischemia, before the start of Values are means SE for n fields of cells. m, mitochondrial
reperfusion. To determine whether ROS generation potential; 2-MPG, 2-mercaptopropionyl glycine; NAC, N-acetylcys-
during ischemia contributed to the fall in m and tome; PDTC, pyrrolidine dithiocarbamate.

AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003 www.ajpheart.org

H552 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA

attenuated the fall in m at the end of ischemia, as did

N-acetyl-L-cysteine (NAC; 0.5 mM). Likewise, the metal
chelator 1,10-phenanthroline significantly attenuated
the decrease in m during ischemia. These antioxidant
compounds also significantly lessened cell death after
3-h reperfusion. Collectively, these studies suggested
that oxidant stress during ischemia contributes to the
fall in m.
ROS generated from the mitochondrial electron
transport chain can induce cardiolipin oxidation within
the inner mitochondrial membrane (29). This oxidative
damage could contribute to the observed membrane
depolarization by compromising the integrity of the
inner membrane. To detect lipid peroxidation, cardio-
myocytes loaded with the lipophilic probe C11-BODIPY

Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016

were subjected to simulated I/R. During ischemia, a sig-
nificant increase in fluorescence was observed, consistent
with an increase in the oxidation of this fluorophore. If
H2O2 is required for the oxidative damage during ische-
mia, then lipid peroxidation should be attenuated if SOD
is inhibited. Accordingly, the Cu,Zn-SOD inhibitor dieth-
yldithiocarbamate (DDC; 1 mM) was used to attenuate
SOD activity. This caused a significant attenuation in
C11-BODIPY fluorescence (Fig. 3A). In replicate experi-
ments, the rate of increase in fluorescence during ische-
mia (0.77 0.19 arbitrary units (a.u./min); n 6) was
significantly greater (P 0.01) than during baseline
(0.04 0.09 a.u./min; n 6). Administration of DDC
significantly decreased the slope of this relationship
(0.43 0.13, n 3; P 0.01), suggesting that H2O2
contributes to lipid peroxidation. If free iron in the cell
contributes to hydroxyl radical generation by the Fenton
reaction, then iron chelation should also attenuate lipid
peroxidation. Addition of the chelator 1,10-phenanthro-
line (10 M) during ischemia caused a significant de-
crease in the fluorescence signal (Fig. 3B). In replicate
experiments, 1,10-phenanthroline administration signif-
icantly decreased the rate of fluorescence increase
(0.63 0.13 a.u./min, n 3; P 0.01) compared with
ischemia. To determine whether mitochondria are the
source of these oxidants, the electron transport inhibitor
myxothiazol was added during ischemia to inhibit ROS
generation by complex III (30). Myxothiazol blocks elec-
tron transfer from ubiquinol to the Rieske iron-sulfur
center in complex III, thereby preventing the generation
of ubisemiquinone, which is a major source of superoxide
generation. Myxothiazol (2 M) significantly attenuated
the rate of increase in fluorescence during ischemia
(0.026 0.106 a.u./min, n 3). Addition of DMSO, the
solvent used in myxothiazol experiments, produced no
detectable effect.
ROS generation during ischemia most likely begins
Fig. 3. Representative tracing of cell lipid peroxidation as assessed
with univalent electron transfer to O2, thereby gener- with the lipophilic probe C11-BODIPY in cardiomyocytes during
ating superoxide. This process requires that some re- baseline and simulated ischemia. A: diethyldithiocarbamate (DDC),
sidual O2 must still be present to provide substrate for an inhibitor of Cu,Zn-SOD, was added during ischemia to attenuate
that reaction. In previous studies (33) we found that the production of hydrogen peroxide. B: 1,10-phenanthroline, an iron
low levels of O2 were present during ischemia in this chelator, was added during ischemia to attenuate the availability of
free iron. C: myxothiazol, a mitochondrial electron transport inhibi-
model. In the present study, O2 tension within the tor, was added during ischemia to block reactive oxygen species
flow-through chamber was assessed with a phospho- production from complex III. a.u., Arbitrary units.
rescence quenching technique previously shown to be
accurate at low [O2] (24, 37). During simulated ische-
AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003 www.ajpheart.org
 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA H553

effects on membrane potential and cell death (data not

shown). These findings suggested that residual O2 con-
tributes to cell death and the irreversible decline in
m during I/R.
During reperfusion after anoxic ischemia, TMRE flu-
orescence increased progressively, indicating a resto-

Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016

Fig. 4. Representative tracing showing PO2 in buffer perfusate
within the flow-through chamber during standard hypoxic ischemia.
Note that anoxic conditions were not achieved despite vigorous
bubbling with N2-CO2 gas mixtures in the equilibration column.

mia the O2 tension decreased progressively, reaching a

value of 7 mmHg within 10 min (Fig. 4). Therefore,
it is possible that superoxide could be generated using
residual O2 that is present during ischemia.
If residual O2 contributes to the generation of super-
oxide during ischemia, and if these oxidants contribute
to mitochondrial depolarization and cell death, then
significant protection should ensue if residual O2 is
scavenged from the system during ischemia. To test
this, an enzymatic O2 scavenger was added to the
perfusate to create anoxic conditions during ischemia,
thereby limiting the availability of O2 as an electron
acceptor. Anoxic conditions (PO2 0 mmHg) were
created by adding EC-Oxyrase to the ischemia buffer
after it had been equilibrated with 80% N2-20% CO2.
Measurements confirmed that this decreased the PO2
within the chamber from 7 to 0.1 mmHg during
ischemia. Heat inactivation of EC-Oxyrase (100C for
15 min) abolished its O2-scavenging properties (data
not shown). During anoxic ischemia (PO2 0 mmHg),
the decrease in m was attenuated (43 7% de-
crease; Fig. 5A) and cell death after reperfusion was
lessened (12 6%; P 0.001) compared with standard
hypoxic ischemia (66 7%; PO2 7 mmHg) (Fig. 5C).
Heat treatment of EC-Oxyrase abolished its protective

Fig. 5. A: m in cardiomyocytes as assessed with TMRE fluores-

cence during baseline, ischemia, and reperfusion in a flow-through
chamber. Standard hypoxic ischemia produced PO2 of 7 mmHg in the
flow-through chamber. Anoxic ischemia, generated by adding the
enzymatic O2 scavenger EC-Oxyrase to the ischemic perfusate, pro-
duced an PO2 of 0.1 mmHg in the flow-through chamber (mean
values SE, n 4 each). B: m assessed with TMRE fluorescence
during anoxic acidosis. Anoxic acidosis was generated by adding
EC-Oxyrase to standard buffered salt solution (BSS, containing
glucose) bubbled with 80% N2-20% CO2. In 1 group, oligomycin was
added to the perfusate only during anoxic acidosis. (mean values
SE, n 4 each) C: cell death after 3-h reperfusion with normoxic BSS
in the same populations of cells represented in A and B.

AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003 www.ajpheart.org

H554 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA
ration of m (Fig. 5A). By contrast, minimal evidence
of recovery was seen during reperfusion after standard
hypoxic ischemia. This suggested that the mechanism
responsible for loss of m may be different between
hypoxic and anoxic ischemia. Normally, m reflects a
balance between the rate of proton extrusion from the
mitochondrial matrix (a function of the rate of electron
transport) and the rate at which protons reenter the
matrix (a function of ATP synthase activity and/or ion
leaks). During anoxia, electron flux should cease.
Therefore, m should decrease unless glycolytic ATP
is available to maintain m through reverse opera-
tion of the ATP synthase. During ischemia in our
model, inhibition of glycolysis by 2-DOG may have
prevented reverse operation of the ATP synthase. To
explore the mechanism responsible for the fall in m
during anoxia, cells were subjected to anoxia under the
same acidic conditions used for ischemia (20% CO2),
except that glucose was added to the perfusate and
2-DOG was omitted to permit glycolysis to continue.
During anoxic acidosis, no decrease in m was ob-
served (Fig. 5B) and minimal cell was evident after
reperfusion (Fig. 5C), which suggested that m was
maintained during anoxia by reverse operation of the
ATP synthase. To confirm that reverse operation of the
ATP synthase was responsible for sustaining m dur-
ing anoxia when glycolysis remained functional, oligo-
mycin (10 M) was added to inhibit the ATP synthase
during anoxic acidosis. Under those conditions, m
decreased significantly (Fig. 5B). During reperfusion
(21% O2, 5% CO2) without oligomycin, clear evidence of
mitochondrial repolarization was evident because pro-
ton pumping was restored when electron transport
resumed and the mitochondria membrane integrity
was not compromised. The decrease in m caused by
anoxic acidosis with oligomycin was associated with
relatively low cell death (Fig. 5C). These findings indi-
cate that low levels of residual O2 during ischemia are
injurious because they contribute to an irreversible
decline in m. By contrast, the decline in m caused
by anoxia plus glucose deprivation is reversible and
associated with minimal cell death.
Preliminary studies of standard hypoxic ischemia
suggested that a correlation may exist between the
magnitude of the decrease in m and subsequent cell
death. To determine whether such a dose-response
relationship exists, we experimentally varied the se-
verity of ischemia by adjusting the residual level of O2
during ischemia without changing its duration (1 h). In
these experiments the PO2 during ischemia was in-
creased from 7 mmHg (hypoxic ischemia) to 15
mmHg. Subsequent cell death was measured after-3 h
reperfusion. As shown in Fig. 6, the extent of cell death
during reperfusion was significantly attenuated when
the fall in m during ischemia was less severe. This
suggested that the magnitude of mitochondrial depo-
larization during ischemia might contribute mechanis-
tically to the cell death measured at the end of reper-
fusion.
To determine the significance of mitochondrial ROS
generation for the depolarization and cell death during
AJP-Heart Circ Physiol VOL
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016
Fig. 6. Cell death during reperfusion as a function of the decrease in
m during ischemia in cardiomyocytes studied in a flow-through
chamber. The decrease in m was adjusted by lessening the sever-
ity of hypoxia during the ischemia. Each data point represents an
independent experiment.
I/R, 0-cells were loaded with TMRE and studied dur-
ing simulated ischemia. Despite a lack of electron
transport, 0-cells maintain m by ATP/ADP ex-
change via the adenine nucleotide translocator in the
inner membrane (5). Wild-type 143B osteosarcoma
cells demonstrated a marked depletion of m during
ischemia that was qualitatively similar to that seen in
cardiomyocytes (Fig. 7). However, ischemia failed to
produce a similar depletion of TMRE fluorescence in
0-cells. Cell death in wild-type cells averaged 82.2
9.9% vs. 28.7 7.5% in the mitochondria-deficient cells
(P 0.001). Thus the 0-cells were protected against
mitochondrial depolarization and subsequent cell
death compared with wild-type cells.
DISCUSSION
These studies demonstrate that mitochondria un-
dergo a significant and irreversible decrease in poten-
tial during ischemia. The degree of depolarization cor-
relates with the extent of cell death during reperfusion.
However, mitochondrial depolarization by itself does
not cause cell death, because administration of anoxia
plus oligomycin caused a reversible mitochondrial de-
polarization without causing significant cell death. Mi-
tochondrial depolarization during ischemia was trig-
gered by ROS generated from the mitochondrial
electron transport chain despite the low [O2] condi-
tions. These oxidants appear to initiate a cascade of
lipid peroxidation that disrupts the integrity of the
inner mitochondrial membrane, thereby preventing re-
polarization during reperfusion. Activation of the MPT
pore apparently did not contribute to this process,
because attempts to inhibit the activation of that pore
failed to prevent depolarization or cell death. By con-
trast, a variety of pharmacological antioxidant com-
pounds attenuated both the fall in m and subse-
quent cell death. Furthermore, scavenging of residual
O2 during ischemia prevented the depletion of m
284 FEBRUARY 2003 www.ajpheart.org
 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA H555

ischemic preconditioning (13, 26, 31). The activation of

preconditioning confers significant protection against
subsequent lethal ischemia. In contrast, higher levels
of oxidant stress are observed at the start of reperfu-
sion, when a transient burst of ROS generation is
observed that correlates with subsequent cell death
(32). During ischemia before reperfusion, ROS are gen-
erated by the electron transport chain of mitochondria
(33, 34), although the significance of these oxidants in
cell injury is not fully understood. The present study
focused on the relationship between oxidant stress
during ischemia, cell survival, and the fall in m in a
cardiomyocyte model.
Relationship between ischemic ROS and m. m
is normally maintained by proton pumping, which is

Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016

linked to the rate of electron transport. If the O2 ten-
sion in the cell falls below a critical level of 14 mmHg,
m should decrease because electron transport be-
comes limited by the availability of O2 at cytochrome
oxidase. In our study, large decreases in TMRE fluo-
rescence were observed during ischemia, indicating
that a significant fall in m must have occurred.
However, the fall in m could not be explained by a
lack of O2, because the O2 tension did not fall below the
critical level during the ischemic exposure (19). More-
over, m failed to show significant recovery during
reperfusion, which suggests that the mitochondria sus-
tained an irreversible injury during the ischemic expo-
sure. The extent of cell death after 3-h reperfusion
correlated significantly with the extent of the loss in
m, which suggests that the mitochondrial injury
sustained during ischemia could contribute to cell
death during reperfusion.
Several observations support the conclusion that
ROS contribute importantly to the decline in m

Fig. 7. A: m in wild-type 143B osteosarcoma cells as assessed with

TMRE fluorescence during baseline, ischemia, and reperfusion in a
flow-through chamber (n 5). B: mitochondrial potential in mutant
0-143B osteosarcoma cells as assessed with TMRE fluorescence
during baseline, ischemia, and reperfusion in a flow-through cham-
ber (n 5).

and significantly protected cells. Finally, 143B cells

lacking a mitochondrial electron transport chain failed
to demonstrate a depletion of m during ischemia
and were significantly protected against cell death
during reperfusion. Collectively these studies identify
a central role for mitochondrial oxidant generation
during ischemia in the irreversible mitochondrial de-
polarization and subsequent cell death induced by is-
chemia and reperfusion in this model (Fig. 8).
Role of ROS during I/R. ROS have long been associ-
ated with I/R injury. It is increasingly evident that
ROS play diverse roles in I/R, ranging from protective
effects at one extreme to damage-inducing effects at
the other. For example, low levels of oxidants appear to Fig. 8. Proposed scheme of cellular injury during ischemia before
function as signaling agents during the induction of reperfusion. ROS, reactive oxygen species.

AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003 www.ajpheart.org

H556 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA
during ischemia. First, a variety of chemically dissim-
ilar antioxidant compounds given during ischemia
were able to attenuate the fall in m and to signifi-
cantly lessen cell death after reperfusion. Second, the
scavenging of residual O2 during anoxic ischemia ab-
rogated both the fall in m and later cell death. Both
the enzymatic activity of EC-Oxyrase and its protective
effects were abolished by heat denaturation, which
indicates that the protection it provided was due to its
O2 scavenging properties. The studies with anoxic is-
chemia indicate that low residual levels of O2 during
ischemia are important for mitochondrial depolariza-
tion and cell death because they act as a substrate for
the generation of ROS. Third, the results with C11-
BODIPY indicate that lipid peroxidation occurs during
ischemia and that ROS originate from the mito-
chondrial electron transport chain. Fourth, the 0-cells
lacking an electron transport chain failed to exhibit
mitochondrial depolarization and were significantly
protected against cell death during reperfusion. We
conclude that mitochondrial ROS generation during
ischemia contributes importantly to the fall in m.
This response is likely due to the formation of lipid
peroxides, which could undermine m by destabiliz-
ing the inner membrane. A loss of membrane integrity
could also explain why m failed to recover when
normal O2 levels were restored during reperfusion.
This conclusion is consistent with the findings of
Lesnefsky et al. (23), who demonstrated that cardio-
lipin levels decrease during ischemia in subsarcolem-
mal mitochondria. Cardiolipin is a membrane phospho-
lipid found in high abundance within mitochondria
that interacts with electron transport proteins (14, 15)
and may also be important for maintaining the integ-
rity of the inner membrane in terms of its ability to
support the transmembrane potential. Mitochondrial
oxidant generation during ischemia may explain the
loss of cardiolipin, which could contribute to the mito-
chondrial dysfunction associated with I/R (28). Alter-
natively, oxidants could contribute to mitochondrial
damage through direct oxidation of other lipids and
proteins or by promoting the opening of the MPT pore
(3). Each of these mechanisms could contribute to mi-
tochondrial depolarization because of their effects on
electron transport and/or mitochondrial membrane in-
tegrity. However, our findings suggest that the de-
crease in m was not a result of MPT pore opening,
because cyclosporin A treatment had no significant
effect on the decrease in membrane potential during
ischemia or the extent of cell death. This conclusion is
consistent with previous studies suggesting that MPT
pore opening is unlikely to occur under the low pH
conditions of ischemia and is more likely to occur after
reperfusion (18, 20); indeed, we observed a fall in m
during ischemia before reperfusion. However, other
investigators have used higher (10, 16) or lower (38)
concentrations of cyclosporin A to inhibit the opening
of that pore, so it is not clear whether protection would
have been observed at different concentrations or with
other inhibitors of the MPT pore.
AJP-Heart Circ Physiol VOL
The data suggest that H2O2 and hydroxyl radicals,
rather than superoxide, are responsible for the oxida-
tive damage to mitochondria. Normally, superoxide
degradation by SOD is an important step in preventing
oxidant injury by that radical, so it is surprising that
SOD inhibition is protective during ischemia. One ex-
planation is that H2O2 may pose an unusual threat to
the cell under conditions of ischemia, when release of
iron from sites where it is normally chelated could
facilitate hydroxyl radical generation via the Fenton
reaction. Relative to hydroxyl radical, superoxide is far
less reactive and may be less injurious during rela-
tively short periods of ischemia.
Relationship between mitochondrial depolarization
and cell death. We observed a significant correlation
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016
between the fall in TMRE fluorescence and cell death.
When the magnitude of the fall in m was manipu-
lated by adjusting the severity of the ischemia, cell
survival was found to be worse in experiments where
the fall in m was larger. Previous studies in cardio-
myocytes demonstrated that ROS generation tends to
increase as O2 tension is lowered from 35 to 7 mmHg
(6). We therefore suggest that milder ischemia was
protective in the present study because of the lesser
oxidant stress it generated and the associated decrease
in oxidant damage to membranes. An oxidant-medi-
ated disruption of mitochondrial inner membrane in-
tegrity during ischemia could conceivably lead to cell
death by promoting matrix swelling, release of cyto-
chrome c to the cytosol, and activation of the cell
apoptotic machinery (36). However, mitochondrial de-
polarization by itself was not lethal to these cells, as
evidenced by the minimal extent of death observed
when m was depleted with anoxic acidosis plus
oligomycin. For a given degree of depolarization, it is
conceivable that matrix swelling caused by lipid per-
oxidation and loss of membrane integrity is more se-
vere than when caused simply by electron transport
inhibition. In the former case, reoxygenation would not
promote recovery of m because the loss of membrane
integrity would defeat the effects of proton pumping. In
the latter case, a restoration of electron transport dur-
ing reoxygenation would allow m to recover.
Interestingly, hypoxia alone (PO2 7 mmHg) was
not sufficient to induce significant cell death in this
study. Likewise, hypercapnic acidosis (20% CO2) was
well tolerated under normoxic conditions. However,
when the two conditions were combined, 60% cell
death resulted. Acidosis may increase cell death by
causing the release of Fe2 from intracellular sites
where it is normally chelated. The ROS released in
response to hypoxia alone appear to be well tolerated
by cells (7). However, in a setting where acidosis causes
release of iron ions, these ROS may lead to generation
of hydroxyl radicals as a consequence of Fenton inter-
actions. The resulting loss of membrane integrity could
explain the observed decrease in m and subsequent
failure to recover during reoxygenation. This interpre-
tation is consistent with our observation that an iron
chelator was most protective of m and cell viability.
Although the importance of the Fenton reaction in I/R
284 FEBRUARY 2003 www.ajpheart.org
 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA
injury is not new (1, 4, 9, 17, 21, 25), its potential
involvement in cellular injury during ischemia before
reperfusion has not been explored previously, to our
knowledge.
In summary, these studies reveal that ROS gener-
ated during ischemia contribute to the irreversible loss
of m. The extent of this damage correlates with the
extent of cell death during reperfusion, and interven-
tions that minimize the oxidative stress during ische-
mia also attenuate the loss of m. These findings
therefore suggest that oxidant generation during ische-
mia before reperfusion plays a significant role in deter-
mining cell death during reperfusion.
This study was supported by National Heart, Lung, and Blood
Institute Grants HL-32646 and HL-35440. J. Levraut was supported
by a grant from the Socie te de Re animation de Langue Franc aise.
REFERENCES
1. Ambrosio G, Zweier JL, Duilio C, Kuppusamy P, Santoro
G, Elia PP, Tritto I, Cirillo P, Condorelli M, and Chiariello
M. Evidence that mitochondrial respiration is a source of poten-
tially toxic oxygen free radicals in intact rabbit hearts subjected
to ischemia and reflow. J Biol Chem 268: 1853218541, 1993.
2. Becker LB, Vanden Hoek TL, Shao ZH, Li CQ, and Schu-
macker PT. Generation of superoxide in cardiomyocytes during
ischemia before reperfusion. Am J Physiol Heart Circ Physiol
277: H2240H2246, 1999.
3. Bernardi P. Mitochondrial transport of cations: channels, ex-
changers, and permeability transition. Physiol Rev 79: 1127
1155, 1999.
4. Bolli R, Jeroudi MO, Patel BS, DuBose CM, Lai EK, Rob-
erts R, and McCay PB. Direct evidence that oxygen-derived
free radicals contribute to postischemic myocardial dysfunction
in the intact dog. Proc Natl Acad Sci USA 86: 46954699, 1989.
5. Buchet K and Godinot C. Functional F1-ATPase essential in
maintaining growth and membrane potential of human mito-
chondrial DNA-depleted rho cells. J Biol Chem 273: 22983
22989, 1998.
6. Budinger GRS, Duranteau J, Chandel NS, and Schu-
macker PT. Hibernation during hypoxia in cardiomyocytes: role
of mitochondria as the O2 sensor. J Biol Chem 273: 33203326,
1998.
7. Chandel NS, McClintock DS, Feliciano CE, Wood TM, Me-
lendez JA, Rodriguez AM, and Schumacker PT. Reactive
oxygen species generated at mitochondrial Complex III stabilize
HIF-1- during hypoxia: a mechanism of O2 sensing. J Biol
Chem 275: 2513025138, 2000.
8. Chandel NS and Schumacker PT. Cells depleted of mitochon-
drial DNA (0) yield insight into physiological mechanisms.
FEBS Lett 454: 173176, 1999.
9. Cordis GA, Maulik G, Bagchi D, Riedel W, and Das DK.
Detection of oxidative DNA damage to ischemic reperfused rat
hearts by 8-hydroxydeoxyguanosine formation. J Mol Cell Car-
diol 30: 19391944, 1998.
10. Di Lisa F, Menabo R, Canton M, Barile M, and Bernardi P.
Opening of the mitochondrial permeability transition pore
causes depletion of mitochondrial and cytosolic NAD and is a
causative event in the death of myocytes in postischemic reper-
fusion of the heart. J Biol Chem 276: 25712575, 2001.
11. Duranteau J, Chandel NS, Kulisz A, Shao Z, and Schu-
macker PT. Intracellular signaling by reactive oxygen species
during hypoxia in cardiomyocytes. J Biol Chem 273: 11619
11624, 1998.
12. Ehrenberg B, Montana V, Wei MD, Wuskell JP, and Loew
LM. Membrane potential can be determined in individual cells
from the Nernstian distribution of cationic dyes. Biophys J 53:
785794, 1988.
13. Forbes RA, Steenbergen C, and Murphy E. Diazoxide-in-
duced cardioprotection requires signaling through a redox-sen-
sitive mechanism. Circ Res 88: 802809, 2001.
AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003
H557
14. Fry M, Blondin GA, and Green DE. The localization of tightly
bound cardiolipin in cytochrome oxidase. J Biol Chem 255:
99679970, 1980.
15. Fry M and Green DE. Cardiolipin requirement for electron
transfer in complex I and III of the mitochondrial respiratory
chain. J Biol Chem 256: 18741880, 1981.
16. Gibson EM, Henson ES, Villanueva J, and Gibson SB. MEK
kinase 1 induces mitochondrial permeability transition leading
to apoptosis independent of cytochrome c release. J Biol Chem
277: 1057310580, 2002.
17. Granger DN and Korthuis RJ. Physiologic mechanisms of
postischemic tissue injury. Annu Rev Physiol 57: 311332, 1995.
18. Halestrap AP, Kerr PM, Javadov S, and Woodfield KY.
Elucidating the molecular mechanism of the permeability tran-
sition pore and its role in reperfusion injury of the heart. Bio-
chim Biophys Acta 1366: 7994, 1998.
19. Kennedy FG and Jones DP. Oxygen dependence of mitochon-
drial function in isolated rat cardiac myocytes. Am J Physiol Cell
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016
Physiol 250: C374C383, 1986.
20. Kerr PM, Suleiman MS, and Halestrap AP. Reversal of
permeability transition during recovery of hearts from ischemia
and its enhancement by pyruvate. Am J Physiol Heart Circ
Physiol 276: H496H502, 1999.
21. Kim MS and Akera T. O2 free radicals: cause of ischemia-
reperfusion injury to cardiac Na K ATPase. Am J Physiol
Heart Circ Physiol 252: H252H257, 1987.
22. King MP and Attardi G. Human cells lacking mtDNA: repopu-
lation with exogenous mitochondria by complementation. Sci-
ence 246: 500503, 1989.
23. Lesnefsky EJ, Slabe TJ, Stoll MSK, Minkler PE, and Hop-
pel CL. Myocardial ischemia selectively depletes cardiolipin in
rabbit heart subsarcolemmal mitochondria. Am J Physiol Heart
Circ Physiol 280: H2770H2778, 2001.
24. Lo LW, Koch CJ, and Wilson DF. Calibration of oxygen-
dependent quenching of the phosphorescence of Pd-meso-tetra
(4-carboxyphenyl) porphine: a phosphor with general application
for measuring oxygen concentration in biological systems. Anal
Biochem 236: 153160, 1996.
25. Opie LH. Reperfusion injury and its pharmacologic modifica-
tion. Circulation 80: 10491062, 1989.
26. Pain T, Yang XM, Critz SD, Yue Y, Nakano A, Liu GS,
Heusch G, Cohen MV, and Downey JM. Opening of mito-
chondrial KATP channels triggers the preconditioned state by
generating free radicals. Circ Res 87: 460466, 2000.
27. Pap EHW, Drummen GPC, Post JA, Rijken PJ, and Wirtz
KWA. Fluorescent fatty acid to monitor reactive oxygen in single
cells. Methods Enzymol 319: 603612, 2000.
28. Paradies G, Petrosillo G, Pistolese M, Di Venosa N, Serena
D, and Ruggiero FM. Lipid peroxidation and alterations to
oxidative metabolism in mitochondria isolated from rat heart
subjected to ischemia and reperfusion. Free Radic Biol Med 27:
4250, 1999.
29. Petrosillo G, Ruggiero FM, Pistolese M, and Paradies G.
Reactive oxygen species generated from the mitochondrial elec-
tron transport chain induce cytochrome c dissociation from beef-
heart submitochondrial particles via cardiolipin peroxidation.
Possible role in the apoptosis. FEBS Lett 509: 435438, 2001.
30. Turrens JF, Alexandre A, and Lehninger AL. Ubisemiqui-
none is the electron donor for superoxide formation by complex
III of heart mitochondria. Arch Biochem Biophys 237: 408414,
1985.
31. Vanden Hoek TL, Becker LB, Shao Z, Li C, and Schu-
macker PT. Reactive oxygen species released from mitochon-
dria during brief hypoxia induce preconditioning in cardiomyo-
cytes. J Biol Chem 273: 1809218098, 1998.
32. Vanden Hoek TL, Becker LB, Shao ZH, Li CQ, and Schu-
macker PT. Preconditioning in cardiomyocytes protects by at-
tenuating oxidant stress at reperfusion. Circ Res 86: 541548,
2000.
33. Vanden Hoek TL, Li C, Shao Z, Schumacker PT, and
Becker LB. Significant levels of oxidants are generated by
isolated cardiomyocytes during ischemia prior to reperfusion. J
Mol Cell Cardiol 29: 25712583, 1997.
www.ajpheart.org
H558 MITOCHONDRIAL DEPOLARIZATION DURING ISCHEMIA

34. Vanden Hoek TL, Shao Z, Li C, Schumacker PT, and
Becker LB. Mitochondrial electron transport can become a
significant source of oxidative injury in cardiomyocytes. J Mol
Cell Cardiol 29: 24412450, 1997.
35. Vanden Hoek TL, Shao Z, Li C, Zak R, Schumacker PT,
and Becker LB. Reperfusion injury in cardiac myocytes after
simulated ischemia. Am J Physiol Heart Circ Physiol 270:
H1334H1341, 1996.
36. Vander Heiden MG, Chandel NS, Li XX, Schumacker PT,
Colombini M, and Thompson CB. Outer mitochondrial mem-
brane permeability can regulate coupled respiration and cell
survival. Proc Natl Acad Sci USA 97: 46664671, 2000.
37. Wilson DF, Rumsey WL, Green TJ, and Vanderkooi JM.
The oxygen dependence of mitochondrial oxidative phosphoryla-
tion measured by a new optical method for measuring oxygen
concentration. J Biol Chem 263: 27122718, 1988.
38. Woodfield K, Rueck A, Brdiczka D, and Halestrap AP.
Direct demonstration of a specific interaction between cyclophi-
lin-D and the adenine nucleotide translocase confirms their role
in the mitochondrial permeability transition. Biochem J 336:
287290, 1998.
39. Zweier JL, Flaherty JT, and Weisfeldt ML. Direct measure-
ment of free radical generation following reperfusion of ischemic
myocardium. Proc Natl Acad Sci USA 84: 14041407, 1987.

Downloaded from http://ajpheart.physiology.org/ by 10.220.33.2 on September 18, 2016

AJP-Heart Circ Physiol VOL 284 FEBRUARY 2003 www.ajpheart.org