Tài liệu Báo cáo khoa học: Pyruvate reduces DNA damage during hypoxia and after reoxygenation in hepatocellular carcinoma cells pptx

Pyruvate reduces DNA damage during hypoxia and after
reoxygenation in hepatocellular carcinoma cells
3
Emilie Roudier*, Christine Bachelet and Anne Perrin
Unite
´
de Biophysique Cellulaire et Mole
´
culaire, IFR ‘RMN biome
´
dicale: de la cellule a
`
l’homme’, CRSSA, BP 87, La Tronche, France
Pyruvate, as well as lactate, is an end-product of gly-
colysis. Its production is enhanced in tumor cells,
where high rates of aerobic glycolysis, historically
known as the Warburg effect, are observed [1]. It is
only lately that pyruvate has been described as playing
an important role in cancer progression. First of all,
alterations in components of pyruvate metabolism
have been reported in tumor cells, and appeared to
increase cancer cell proliferation [2,3]. Moreover,
recent evidence supports a novel role of pyruvate in
metabolic signaling in tumors. Pyruvate has been
reported to promote hypoxia-inducible factor (HIF-1)
stability and activate HIF-1-inducible gene expression.
This can promote the malignant transformation and
survival of cancer cells [4,5]. Pyruvate also exhibits
strong angiogenic activity in vitro and in vivo and
positively affects angiogenic processes [6]. As the
angiogenic switch is a crucial event in tumorigenesis,

pyruvate may be important for cancer progression. All
together, these findings suggest that pyruvate could
induce the molecular signaling usually caused by
hypoxia.
Chronic or transient hypoxia in the tumor is induced
by heterogeneous bloodflow resulting from impaired
vascularization [7]. Tumor cells are often exposed to
shorter or longer periods of hypoxia or ischemia fol-
lowed by reoxygenation or recirculation. An adaptive
response of cancer cells takes place through multi-
faceted changes [8], which are mainly coordinated by
HIF-1 [9]. The outcome is clonal selection of the
tumor cells that are most resistant and well adapted to
hypoxia [10]. Many alterations occur that induce
Keywords
DNA damage; glutathione; hypoxia;
pyruvate; reoxygenation
Correspondence
A. Perrin, CRSSA ⁄ RBP, Unite
´
de
biophysique cellulaire et mole
´
culaire, 24
Avenue des Maquis du Gre
´
sivaudan, BP 87,
38702 La Tronche Cedex, France
2
Fax ⁄ Tel: + 33 4 76 63 68 79

E-mail:
*Present address
De
´
partement de kine
´
siologie, Universite
´
de
Montre
´
al, Canada
(Received 5 July 2007, revised 10 August
2007, accepted 14 August 2007)
doi:10.1111/j.1742-4658.2007.06044.x
Pyruvate is located at a crucial crossroad of cellular metabolism between
the aerobic and anaerobic pathways. Modulation of the fate of pyruvate,
in one direction or another, can be important for adaptative response to
hypoxia followed by reoxygenation. This could alter functioning of the
antioxidant system and have protective effects against DNA damage
induced by such stress. Transient hypoxia and alterations of pyruvate
metabolism are observed in tumors. This could be advantageous for cancer
cells in such stressful conditions. However, the effect of pyruvate in tumor
cells is poorly documented during hypoxia⁄ reoxygenation. In this study, we
showed that cells had a greater need for pyruvate during hypoxia. Pyruvate
decreased the number of DNA breaks, and might favor DNA repair. We
demonstrated that pyruvate was a precursor for the biosynthesis of gluta-
thione through oxidative metabolism in HepG2 cells. Therefore, gluta-
thione decreased during hypoxia, but was restored after reoxygenation.
Pyruvate had beneficial effects on glutathione depletion and DNA breaks

induced after reoxygenation. Our results provide more evidence that the
a-keto acid promotes the adaptive response to hypoxia followed by reoxy-
genation. Pyruvate might thus help to protect cancer cells under such
stressful conditions, which might be harmful for patients with tumors.
Abbreviations
GSH(c-glutamyl), c-glutamyl glutathione; HIF-1, hypoxia-inducible factor; PCA, perchloric acid; ROS, reactive oxygen species.
5188 FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
deleterious effects, such as resistance to anticancer
treatment, tumor growth, and metastasis development
[11,12]. Moreover, they allow survival under low par-
tial pressure of oxygen.
One of the alterations induced by hypoxia involves
modifications of glutathione metabolism. Glutathione
is an important intracellular antioxidant and redox
potential regulator that plays a vital role in drug
detoxification and in cellular protection against
damage by free radicals, peroxides, and toxins [13].
Hypoxia enhances the expression of c-glutamylcysteine
synthetase [14] and glutathione transferase [15] in can-
cer cells. Such alterations of the glutathione system
can enhance survival of cancer cells.
In general, hypoxia and reoxygenation may induce
important modifications to the functioning of the
antioxidative system. Multilevel adaptation to oxidative
damage could then follow, tending to enhance protec-
tive mechanisms. This could have beneficial effects on
protection against DNA damage [16]. Pyruvate may
play a role in this adaptive response. This has been
intensively studied in normal cells and tissues, such as

heart, liver, and brain. Generally, this a-keto acid is
associated with protective effects against hypoxia and
reoxygenation. This is mainly ascribed to its ability to
maintain redox status [17], intervening in the DNA
repair system [18–20] and restoring antioxidant capaci-
ties [21–26]. This adaptive response, beneficial in the
case of normal cells and tissues, could become deleteri-
ous for tumor carriers when it takes place in malignant
cells [27,28]. However, the role of pyruvate during
hypoxia is poorly documented in cancer cells.
We previously showed that pyruvate could favor the
glycolytic pathway from glucose to lactate in glial and
hepatic cells underhypoxic conditions [29]. We now
investigated whether such an effect might affect the
adaptive response of tumor cells to hypoxia and reoxy-
genation. We particularly focused on the antioxidant
system, in particular glutathione metabolism, and on
studying DNA damage. The metabolism of glutathione
and DNA breaks were investigated in hepatocellular
carcinoma HepG2 cells cultivated with or without
pyruvate during and after hypoxia.
In the present study, we showed that cells had a
greater need for pyruvate during hypoxia. Pyruvate
decreased DNA breaks and might favor DNA repair.
We demonstrated that pyruvate was a precursor for
the biosynthesis of glutathione through oxidative
metabolism in HepG2 cells. Therefore, glutathione
decreased during hypoxia, but was restored after reox-
ygenation. Pyruvate had beneficial effects on glutathi-
one depletion and DNA breaks induced after hypoxia.

Our results provide more evidence that a-keto acids
promote adaptative responses to hypoxia followed by
reoxygenation. Pyruvate might thus help to protect of
cancer cells during such stressful conditions, which
might be harmful for patients with tumors.
Results
HepG2 cells have greater pyruvate requirements
under hypoxic conditions
Pyruvate content was quantified in cell culture medium
under normoxic and hypoxic conditions, and in the
absence and presence of pyruvate (0.8 mm) in the cul-
ture medium (Fig. 1). Total oxygen depletion was com-
plete after 1 h of incubation in the anaerobic jar (see
Experimental procedures). The detection limits of the
assay did not allow measurement of intracellular pyru-
vate.
Under normoxic conditions and in the absence of
pyruvate, HepG2 cells produced and secreted pyruvate.
The extracellular level rose linearly up to 2.2 lmolÆ
mg
)1
protein after 6 h, corresponding to 0.4 mm in the
culture medium. When added to the culture medium,
the cells consumed pyruvate, with its concentration
decreasing slightly and linearly from 0.75 mm to
0.63 mm between 0 and 6 h of incubation.
Under hypoxic conditions, and in the absence of
exogenous pyruvate, pyruvate excretion by the cells
remained identical to that seen under normoxia for the
first hour and slowed down during the second hour.

After 1–2 h of incubation (depending on the experi-
ments, data not shown), HepG2 cells stopped secreting
pyruvate and consumed it, as shown by the decrease in
pyruvate concentration. After reoxygenation (data not
shown), the production of pyruvate started again until
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Extracellular pyruvate (mM)
0 1 2 3 4 5 6
Time (hours)
Normoxia with pyruvate
Normoxia without pyruvate
Hypoxia with pyruvate
Hypoxia without pyruvate
Fig. 1. Extracellular pyruvate content in culture medium of HepG2
cells under normoxia (in dark) and hypoxia (in white) when exoge-
nous pyruvate (0.8 m
M) is present (circles) or not present (square).
The level of extracellular pyruvate was quantified every hour. Val-
ues are means ± SD of three independent experiments.
E. Roudier et al. Pyruvate reduces DNA damage during ⁄ after hypoxia
FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works

1
5189
a level of approximately 0.4 mm was reached. When
added to the culture medium, exogenous pyruvate
disappeared linearly at a higher rate than under norm-
oxia up to 3 h, and then more rapidly from 3 to 5 h of
incubation. The concentration stabilized at 0.4 mm.
The cells seemed to adjust between production and
consumption of pyruvate in order to maintain the
extracellular pyruvate level at about 0.4 mm in the cul-
ture medium. This level was maintained after 6 h (data
not shown). This could not be achieved under hypoxia
when no exogenous pyruvate was added. Moreover,
these data show that HepG2 cells have a higher uptake
of pyruvate under hypoxic conditions.
Pyruvate decreases oxidative stress during
hydrogen peroxide exposure
Antioxidative properties have been widely ascribed to
pyruvate. We therefore investigated whether pyruvate
might have those antioxidant capacities in our experi-
mental conditions.
HepG2 cells were cultivated in medium enriched
with 5.5 mm glucose in the absence and presence of
pyruvate (0.8 mm). Cells were exposed to increasing
doses of hydrogen peroxide from 25 to 300 lm.
Figure 2 shows the level of reactive oxygen species
(ROS) in HepG2 cells with and without pyruvate.
The ROS level was lowered in the presence of pyru-
vate for the range of concentrations from 75 to
300 lm hydrogen peroxide. At 240 lm exposure and

after normalization to a control without any cells
(100%), the levels of ROS were 71% and 54% for cells
incubated without and with pyruvate, respectively.
In our experimental conditions, pyruvate could thus
decrease the generation of oxidative stress induced by
hydrogen peroxide in HepG2 cells and may also have
played a direct antioxidant role in HepG2 cells.
Exogenous pyruvate protects DNA under hypoxia
and after reoxygenation
We wondered whether the increased pyruvate uptake
might be related to a protective effect against the con-
sequences of hypoxic stress. As both hypoxia and reox-
ygenation have been reported to induce DNA damage
[30–32], we examined the effects of pyruvate addition
during hypoxia and after reoxygenation on DNA
damage.
HepG2 cells were incubated for 6 h under normoxic
and hypoxic conditions without and with pyruvate
(0.8 mm). DNA fragmentation was estimated with the
comet assay. The assay was carried out immediately
after the 6 h incubation period, and then 1 and 2 h
after reoxygenation of the cell cultures (Fig. 3).
Under normoxia, DNA fragmentation remained
unchanged irrespective of the condition (with or with-
out pyruvate) or incubation time.
After 6 h under hypoxia and in the absence of pyru-
vate, an increase in DNA fragmentation was observed.
This increase was not observed in the presence of
pyruvate. One hour after reoxygenation, DNA frag-
mentation reached a maximum in both the absence

and the presence of pyruvate, and thereafter decreased.
The increase induced by 1 h of reoxygenation was
lower in the presence of 0.8 mm pyruvate, although
10000
12000
0 50 100 150 200 250 300 350
0
2000
4000
6000
8000
with pyruvate
without pyruvate
H
2
O
2
(µM)
Radical oxygen species
(fluorescence units / s)
Fig. 2. Pyruvate decreases the level of ROS in the medium of
HepG2 cells exposed to hydrogen peroxide. HepG2 cells were incu-
bated with 5.5 m
M glucose in the presence of 0.8 mM pyruvate,
and exposed to increasing doses of hydrogen peroxide (from 0 to
300 l
M). Levels of ROS were estimated by spectrofluorometry
using 2¢,7¢-dichlorodihydrofluoresceine bi-acetate. The data are from
one representative experiment.
0

5
10
15
20
25
30
35
40
6 h
6 h + 1 h reoxygenation
6 h + 2 h reoxygenation
Normoxia without
pyruvate
Normoxia with
pyruvate
Hypoxia without
pyruvate
Hypoxia with
pyruvate
*
*
*
†
†
Conditions
DNA fragmentation
(Tail extent moment)
Fig. 3. Effect of pyruvate on DNA fragmentation induced by
hypoxia and reoxygenation in HepG2 cells. Cells were incubated for
6 h with glucose (5.5 m

M) in the absence and presence of pyruvate
(0.8 m
M) under normoxic or hypoxic conditions. Thereafter, cells
exposed to hypoxia were incubated under normoxic conditions
(reoxygenation) for 1 or 2 h. Other conditions and statistical
calculations are described under Experimental procedures. Values
are means ± SD, and P-values ¼ 0.05 are compared in the
following way: *hypoxia versus normoxia;

with versus without
pyruvate
13
.
Pyruvate reduces DNA damage during ⁄ after hypoxia E. Roudier et al.
5190 FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
nonsignificant. After 2 h of reoxygenation, DNA dam-
age returned to normoxic levels in the cells treated
with 1 mm pyruvate, whereas it remained higher in
nontreated cells (47%).
These results show that the presence of pyruvate
significantly reduces DNA fragmentation induced by
hypoxia and reoxygenation. Moreover, the comet
assay has recently been described as an efficient
method for detecting DNA repair [33]. We therefore
concluded that, after reoxygenation, HepG2 cells culti-
vated with pyruvate repair DNA damage sooner.
Exogenous pyruvate restores glutathione levels
after reoxygenation but is ineffective at
maintaining glutathione levels during hypoxia

Glutathione is one of the main antioxidant compounds
in the cell, and it is also essential for DNA synthesis
and repair [34]. We wondered whether the beneficial
effect of pyruvate on DNA damage could be mediated
by glutathione. To answer this question, we analyzed
the effect of pyruvate on the glutathione content of
HepG2 cells during hypoxia and after reoxygenation.
HepG2 cells were incubated for 6 h under normoxic
and hypoxic conditions in the absence and presence of
pyruvate (0.8 mm). Thereafter, hypoxic cells were re-
incubated under normoxic conditions (reoxygenation).
The total intracellular glutathione content (oxidized
and reduced forms) was assayed in the cells immedi-
ately after incubation (6 h), and 1 and 2 h after reoxy-
genation (Fig. 4).
Under normoxia, intracellular reduced and oxidized
glutathione levels remained unchanged irrespective of
the presence of pyruvate and incubation time. After
6 h under hypoxia, the glutathione content decreased
by approximately 50%, independently of the presence
of pyruvate. This profile remained unchanged after 1 h
of reoxygenation. After 2 h of reoxygenation, a clear
difference was observed between cells incubated with
and without pyruvate. When pyruvate was lacking, the
glutathione levels remained significantly low, whereas
in the presence of the a-keto acid, the glutathione level
was restored.
Pyruvate is a precursor of glutathione under
normoxia
To further investigate the effects of pyruvate on gluta-

thione content during hypoxia, we examined pyruvate
metabolism using
13
C-NMR. The distribution of
13
C
labeling was analyzed after incubation of HepG2 cells
in a culture medium containing [
13
C
3
]pyruvate, under
normoxia and hypoxia, for 6 h.
[
13
C
3
]Pyruvate mainly led to the production of
13
C-enriched lactate, alanine, glutamate, glutamine and
glutathione as previously described (data not shown).
Under hypoxia, the decreased abundance of
13
C-enriched glutamine and glutamate was concomi-
tant with the increased abundance of enriched lactate
and alanine. This is due to the blockade of the tricar-
boxylic acid cycle induced by lack of oxygen. We
examined the peaks corresponding to the glutathione
carbons, and quantified the relative abundance of the
corresponding molecules under hypoxia and normoxia

(Fig. 5A). Incubation with [
13
C
3
]pyruvate resulted in
labeling of [
13
C
2
]c-glutamyl glutathione [GSH(c-glut-
amyl)], [
13
C
4
]c-glutamyl glutathione and [
13
C
3
]c-
glutamyl glutathione. Peaks corresponding to
[
13
C
3
]glutamine and [
13
C
3
]GSH(c-glutamyl) were indis-
tinguishable, as were those corresponding to [

13
C
2
]glu-
tamine and [
13
C
2
]GSH(c-glutamyl). The presence of
[
13
C
4
]GSH(c-glutamyl) indicated that HepG2 cells con-
sumed pyruvate to form glutathione under normoxia.
Hypoxia induced decreases, respectively, of 92% and
82% in
13
C
2
and
13
C
3
glutamine ⁄ glutathione peak
intensities. The [
13
C
4
]glutathione peak was reduced by

85%. This dramatic decrease indicated that hypoxia
induced strong inhibition of the glutathione synthesis
pathway from pyruvate.
We also analyzed intracellular glutathione (oxidized
and reduced form) by enzymatic assay during normoxia
0
20
40
60
80
100
120
140
160
Conditions
normoxia without
pyruvate
normoxia with
pyruvate
hypoxia without
pyruvate
hypoxia with
pyruvate
6 h
6 h + 1 h reoxygenation
6 h + 2 h reoxygenation
*
*
*
*

†
Intracellular glutathione
(reduced and oxidized nmol/mg of protein)
Fig. 4. Effect of pyruvate on glutathione content during
hypoxia ⁄ reoxygenation in HepG2 cells. Cells were incubated for 6 h
with glucose (5.5 m
M) in the absence and presence of pyruvate
(0.8 m
M) under normoxic or hypoxic conditions. Thereafter, cells
exposed to hypoxia were incubated under normoxic conditions
(reoxygenation) for 1 or 2 h. Levels of oxidized and reduced gluta-
thione were measured in the cell extracts. Other conditions and
statistical calculations are described under Experimental proce-
dures. Values are means ± SD, and P-values ¼ 0.05 are compared
in the following way: *hypoxia versus normoxia;

with versus
without pyruvate
14
.
E. Roudier et al. Pyruvate reduces DNA damage during ⁄ after hypoxia
FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
5191
and hypoxia (from 1 to 6 h, Fig. 5B). In both the pres-
ence and the absence of pyruvate, a regular time-depen-
dent increase in glutathione content occurred under
normoxia, whereas no variation was observed under
hypoxia. After 3 h of incubation, the glutathione level in
cells exposed to hypoxia was significantly lower than in

cells exposed to normoxia, confirming that addition
of exogenous pyruvate failed to restore glutathione
synthesis.
To confirm that the unchanged intracellular glutathi-
one was due to inhibition of synthesis and not simply
excretion from the cell during hypoxia, we also analyzed
extracellular levels after 6 h of incubation (Fig. 5C
shows intracellular and extracellular glutathione levels
and the sum of both). Even though the extracellular
content increased in response to hypoxia, indicating
release by the cells, the sum of both contents showed an
overall decrease in the glutathione level. This effect was
independent of pyruvate supplementation.
Together, these data indicate that pyruvate is a
precursor of glutathione under normoxic conditions
through glutamate generated by oxidative metabolism
(Fig. 6). However, pyruvate cannot be used as a gluta-
thione precursor under hypoxia, because of the lack of
oxygen.
Discussion
Modulation of the fate of pyruvate fate in one direction
or another can be important for adaptive responses to
hypoxia followed by reoxygenation [26,35]. Repression
of pyruvate dehydrogenase (EC 1.2.4.1) and switching
between the highly active tetrameric and the inactive
dimeric forms of pyruvate kinase (EC 2.7.1.40) are
observed in tumors [2,3]. Such alterations of pyruvate
metabolism could be advantageous for cancer cells
under such stressful conditions [36]. Our present work
provides new insights into the role of pyruvate in tumor

cells during hypoxia. We show here that tumor HepG2
cells are inclined to maintain extracellular pyruvate at a
constant level (0.4 mm). Hypoxia inhibits such regula-
tion, whereas pyruvate supplementation restores it. We
also observed that hypoxic cells increase their consump-
tion of exogenous pyruvate and stop releasing it when
the a-keto acid is absent from the medium. Our data
indicate that the endogenous need for pyruvate increases
under hypoxia. Pyruvate protects cells from DNA
breaks induced by both hypoxia and reoxygenation.
0.0
[4-
13
C]-GSH
[3-
13
C]- GSH
and/or -Gln
0.5
1.0
1.5
2.0
2.5
3.0
3.5
2.11
2.07
2.52
0.38
0.31

0.20
normoxia
hypoxia
[2-
13
C]-GSH
and/or - Gln
13
C-enriched metabolites
Relative peak area (arbitrary unit)
175
0
25
50
75
100
125
150
0 1 2 3 4 5 6
Time (hours)
*
*
*
*
Normoxia without pyruvate
Normoxia with pyruvate
Hypoxia without pyruvate
Hypoxia with pyruvate
Intracellular glutathione
(reduced and oxidized nmol/mg of protein)

0
25
50
75
100
125
150
175
Extracellular Intracellular Extracellular and
intracellular
Normoxia
with pyruvate
Hypoxia
without pyruvate
Normoxia with
pyruvate
Hypoxia with
pyruvate
*
*
*
*
*
*
Localization
Glutathione (reduced and
oxidized nmol/mg of protein)
A
B
C

Fig. 5. Effect of hypoxia and pyruvate on glutathione synthesis by
HepG2 cells. (A) Glutathione
13
C labeling following incubation of the
cells with
13
C-enriched pyruvate: cells were incubated for 6 h with
[
13
C
3
]pyruvate and 5.5 mM glucose under normoxic and hypoxic
conditions. After NMR analysis, the peaks corresponding to [
13
C
3
]glu-
tamine (Gln) and ⁄ or [
13
C
3
]GSH (c-glutamyl), [
13
C
4
]GSH(c-glutamyl),
and [
13
C
2

]glutamine and ⁄ or [
13
C
2
]GSH(c-glutamyl) were identified.
The relative amount was quantified by integration of the peak area.
(B) Intracellular kinetics of intracellular glutathione (oxidized and
reduced) concentration: biochemical measurement of total glutathi-
one was performed every hour. HepG2 cells were cultivated in the
absence (squares) and presence (circles) of pyruvate under normoxic
(in black) or hypoxic (in white) conditions. Other conditions and
statistical calculations are described under Experimental procedures.
Values are means ± SD; *P ¼ 0.05 as compared with corresponding
normoxic condition. (C) Distribution of total glutathione in the intra-
cellular and extracellular compartments: HepG2 cells were incubated
for 6 h under normoxic and hypoxic conditions in the absence and
presence of pyruvate (1 m
M). Intracellular and extracellular total glu-
tathione contents were assayed separately. The sum of both is also
shown. Values are means ± SD; *P ¼ 0.05 as compared with the
corresponding normoxic condition.
Pyruvate reduces DNA damage during ⁄ after hypoxia E. Roudier et al.
5192 FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
That is not well correlated with the beneficial effect on
the antioxidant glutathione. Actually, pyruvate restores
glutathione levels only during reoxygenation, and has
no effect under hypoxia. We demonstrate here that the
a-keto acid is a precursor of glutathione through the tri-
carboxylic acid cycle, and that hypoxia severely impairs

its biosynthesis.
Regulation of extracellular pyruvate content has pre-
viously been reported in tumor cells as well as in non-
tumor cells [37]. This mechanism is assumed to be a
way for cells to lower oxidative stress. Pyruvate has
antioxidant properties [21,22] that could possibly also
be manifested under our experimental conditions. The
phenomenon described by O’Donnell-Tormey might
well take place in HepG2 cells.
Impairment of the regulation of the extracellular
content and the increase of its uptake induced by
hypoxia show that alteration of pyruvate metabolism
takes place. The greater uptake might be due to an
increased need to maintain antioxidant capacity in the
cells. However, it might also result from a metabolic
requirement. Pyruvate increases the ratio of lactate
production to glucose consumption (+ 19%, data not
shown). This is in line with our previous work [29]
indicating enhancement of glycolysis activity in the
presence of pyruvate. Increased glycolysis might thus
maintain the ATP supply during oxygen deprivation
while oxidative phosphorylation is seriously impaired.
Such an increase has been reported to have protective
effects against hypoxic stress [38,39].
Hypoxia, known to alter trhe antioxidant system,
affects glutathione metabolism as well. In our experi-
mental conditions, we observed a decrease of intracel-
lular content together with release of the glutathione.
In rat primary hepatocytes in vitro, hypoxia induced
activation of the glutathione transporters, resulting in

increased glutathione export [40]. A similar mechanism
might occur in HepG2 cells, where glutathione trans-
port is functional [41]. Furthermore, our results also
show that hypoxia induces a decrease in total glutathi-
one. Pyruvate has no effect on the hypoxia-induced
decrease of glutathione content, whereas it allows
complete restoration of total glutathione after reoxy-
genation. Hypoxia-induced inhibition of glutathione
synthesis from pyruvate might be the main mechanism
responsible for such an effect. Under normoxia, we
demonstrate that pyruvate supplies glutamate through
oxidative metabolism in tumor cells. Inhibition of the
Glutamate
Mitochondrion
2 ×FAD
+
Glucose
2× Pyruvate
2× Lactate
Alanine
2
×
NADH, H
+
2×NAD
+
Glycolysis
2NAD
+
+

2ADP + Pi
2NADH, H
+
+ 2ATP
Pyruvate
2×NADH, H
+
+ 2×CO
2
2×NAD
+
Krebs Cycle
2×
2×Acetyl-CoA
CoA
CoA
4×CO
2
6
×NADH, H
+
6×NAD
+
2ADP + Pi
2ATP
2
×FADH, H
+
Cytosol
Glutamate

5-Oxoproline
γ-Glutamyl-cysteine
Cysteine
Glycine
ADP + Pi
+ ATP
+ ATP
ADP + Pi
Glutamine
GS-X
(conjugate)
Amino acids (AA)
γ-Glutamyl
-AA
Cysteinyl-
glycine
γ-Glutamyl
-AA
Cysteinylglycine
conjugate
Extracellular are
a
2×
Glutathione
GSH
Fig. 6. Pathway of glutathione synthesis from pyruvate, showing how pyruvate, through the tricarboxylic acid cycle, is involved in glutamate
synthesis and then in the formation of glutathione (GSH).
E. Roudier et al. Pyruvate reduces DNA damage during ⁄ after hypoxia
FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1

5193
tricarboxylic acid cycle thus results in significant inhi-
bition of glutathione synthesis. Usually, cysteine is
assumed to be the limiting precursor of the two-step
reaction leading to glutathione synthesis. However,
glutamate plays an important role in glutathione syn-
thesis [3]. Negative feedback from glutathione itself on
the step catalyzed by glutamate-cysteine ligase can be
prevented by glutamate [1,2]. It might even become
limiting under conditions of mitochondrion blockage
in muscle [4] and during hypoxia in glial cells [5]. Pyru-
vate enhances glycolysis activity during hypoxia in
HepG2 cells [29]. By providing more substrates for the
tricarboxylic acid cycle and through indirect effects on
redox status, pyruvate might thus favor restoration of
the glutamate pool and then the glutathione pool after
reoxygenation (Fig. 6).
Induction of DNA damage by hypoxia and reoxygen-
ation is a well-known phenomenon mainly caused by
ROS [30–32]. A functioning antioxidant system is essen-
tial to reduce such damage. Despite the absence of an
effect on glutathione, pyruvate has beneficial effects on
DNA breaks under hypoxia. Many studies have
reported that pyruvate improves antioxidant capacities
and protects normal tissues against damage induced
by hypoxia ⁄ reoxygenation and ischemia ⁄ reperfusion
[23,26]. The increased uptake of pyruvate might allow
HepG2 cells to maintain their antioxidative capacities
in a way independent of glutathione during hypoxia.
However, after reoxygenation, the beneficial effect of

pyruvate against DNA breaks is well correlated with
glutathione restoration. DNA breaks decrease faster in
cells treated with pyruvate, suggesting stimulation of
DNA repair. As glutathione is essential for DNA syn-
thesis in general, as well as for DNA repair [34], the
action of pyruvate on glutathione might also favor
DNA repair. During hemorrhagic shock and ischemia
followed by reperfusion, pyruvate interferes with poly-
(ADP-ribose)-polymerase (EC 2.4.2.30) activity by pre-
venting loss of total NAD
+
content [19,20,42]. This
might enhance DNA repair. Pyruvate thus plays a role
in DNA protection and repair in tumor HepG2 cells
during hypoxia and after reoxygenation. The mecha-
nism might involve both antioxidant properties and
metabolic activity. The exact mechanism underlying the
pyruvate effects has yet to be further examined.
In light of the data in the literature and our work
with HepG2 cells, we concluded that pyruvate might
act similarly in tumor cells and nontumor cells. It could
then act as a protective agent against DNA damage
under conditions of tumor hypoxia. Many strategies
for cancer therapy are based on increases in the number
of DNA strand breaks (e.g. radiotherapy and alkylat-
ing agents). Pyruvate might thus reduce the efficiency
of treatment by limiting DNA damage and favoring
repair. Pyruvate also stimulates angiogenesis [4,6]; this
might favor tumor reoxygenation and progression. Fur-
thermore, a high intracellular glutathione level corre-

lates with a high level of proliferation of tumor cells
[43] and with resistance to anticancer treatment [44,45].
Pyruvate might support a high glutathione level in
reoxygenated tumors. All of these effects might favor
tumor development and lower efficacy of some thera-
pies. They remain to be verified in vivo in tumors. If
they were verified, it would confirm that no matter how
pyruvate acted, it would be deleterious in cancer.
In conclusion, this study confirms the importance
and the multilevel and very complex implications of
pyruvate for cell responses to hypoxia ⁄ reoxygenation.
Our results are in agreement with current literature
identifying pyruvate as a protector of normal and
tumor cells subjected to hypoxia. Furthermore, this
study provides additional data confirming that what-
ever the underlying mechanisms at work, the presence
of pyruvate is undesirable in tumor cells. Indeed, con-
trolling pyruvate levels might be an advantageous way
of modulating tumor resistance and improving the
efficiency of certain cancer therapies.
Experimental procedures
Cell culture
A human hepatocellular carcinoma HepG2 cell line was
purchased from the American Type Cell Collection. It
derives originally from a hepatocellular carcinoma biopsy
and synthesizes nearly all human plasma proteins [46]. The
cell line is not tumorigenic in immunosuppressed mice.
Cells, used between passages 76 and 82, were grown in
Petri dishes coated with type 1 collagen (10 lgÆmL
)1

,
60 lLÆcm
)2
, 30 min at 37 °C; Sigma-Aldrich, Saint-Quentin
Fallavier, France) in DMEM (Sigma-Aldrich) supple-
mented with 10% fetal bovine serum (Biowhittaker,
Cambrex Corporate, East Rutherford, NJ), antibiotics
(100 000 UÆL
)1
penicillin, 100 mgÆL
)1
streptomycin; Boeh-
ringer Ingelheim, Paris, France), 4 mm glutamine (Jacques
Boy Institute, Reims, France), 1% nonessential amino acids
and 1 mm pyruvate (Seromed Biochrom KG, Berlin, Ger-
many) in a 95% air ⁄ 5% CO
2
humidified atmosphere. Half
of the medium was changed every 2 days. The cell cultures
were split at confluence with 0.25% (w ⁄ v) trypsin (Jacques
Boy Institute) and seeded at a density of 4 · 10
4
cellsÆcm
)2
.
For the experiments, cells were used at subconfluency
(2 days after passage) as determined by the growth curve
based on cell numbers and protein quantification.
To rule out the presence of mycoplasma contamination,
tests were performed using a commercially available

detection kit (Polylabo ⁄ VWR International, Fontenay-
Pyruvate reduces DNA damage during ⁄ after hypoxia E. Roudier et al.
5194 FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
sous-Bois, France). Cell viability was routinely determined
using Trypan Blue exclusion.
Incubation of cells
Cells were incubated as previously described [29]. Briefly,
subconfluent HepG2 cells were incubated for 6 h with
DMEM base (Sigma-Aldrich) containing 5.5 mm glucose,
10 mm Hepes (Sigma-Aldrich), 10% fetal bovine serum,
antibiotics (100 000 UÆL
)1
penicillin, 100 mgÆL
)1
strepto-
mycin), 1% nonessential amino acids and 4 mm glutamine,
both without and with pyruvate (1% of a 100 mm stock
solution, 0.8 mm final), in a normoxic atmosphere with
5% CO
2
and under low oxygen pressure in anaerobic jars.
The oxygen content in the jars was monitored with an oxy-
gen electrode (O
2
sensor; Mettler-Toledo, Viroflay, France).
Similar results were obtained by incubating the cells in an
oxygen-depleted atmosphere using a cell culture incubator
with N
2

,O
2
and CO
2
control (Queue Systems Inc., Ashe-
ville, NC). For the reoxygenation experiments, cells incu-
bated under hypoxia were removed from the jar and placed
in a normoxic atmosphere with 5% CO
2
.
Extraction and quantification of extracellular
pyruvate
Perchloric acid (PCA) (8%, v ⁄ v) was added to the culture
medium (2 : 1, v ⁄ v). After homogenization, the mixture
was centrifuged
4
(4000 g, 10 min, 4 °C, CR3i centrifuge and
swing-out rotor; Jouan, Saint-Herblain, France) and the
supernatant was stored at ) 80 °C prior to assay.
Pyruvate was quantified by the method previously
described by Marbach & Weil [47], with slight modifica-
tions. Briefly, the assay was carried out in a 96-well plate
with 160 lL of sample (PCA extract), 40 lL of Tris ⁄ HCl
buffer (1.5 m Trizma base with 0.05% w ⁄ v
5
sodium azide,
pH 10.5; Sigma-Aldrich) and 40 lL of a mixture (0.3 : 2,
v ⁄ v) of NADH-Na
2
(4.55 mgÆmL

)1
) and Trizma base buffer
per well. First, the absorbance was measured at 340 nm:
10 lL of lactate dehydrogenase (EC 1.1.1.27) (400 UÆmL
)1
in 3.2 m ammonium sulfate, pH 6.5) was added to each
well, and the absorbance was measured after complete sta-
bilization. The absorbance at 340 nm resulting from the
oxidation of NADH to NAD reflected the amount of pyru-
vate originally present in the sample. The pyruvate sample
concentration was determined according to a standard
curve established between 0 and 0.5 mm pyruvate.
Comet assay
The comet assay was performed according to the method
described by Singh et al. [48], using alkaline electrophoresis,
which allows detection of single-strand and double-strand
breaks.
Cells cultivated in Petri dishes (25 mm in diameter)
were suspended in 0.5 mL of NaCl ⁄ P
i
, and cell density
was estimated with a Malassez slide. An aliquot of the
suspension was added to low molecular weight agarose
(0.8%, p ⁄ v in NaCl ⁄ P
i
) to obtain a final concentration of
25 · 10
4
cellsÆmL
)1

. Eighty microliters of agarose-sus-
pended cells was placed on an agarose-coated slide (high
molecular weight, 1% p ⁄ v in NaCl ⁄ P
i
) and covered with
a coverslip. After 5 min on ice, the coverslip was gently
removed. The slides were incubated (1 h, 4 °C in the
dark) in lysis buffer (45 mL of buffer containing 2.5 m
NaCl, 0.1 m EDTA, 10 mm Tris with 5 mL of dimethyl-
sulfoxide and 0.5 mL of Triton X-100, pH 10). The slides
were rinsed twice with electrophoresis buffer (300 mm
NaOH, 1 mm EDTA) for 10 min, and then for 25 min.
Electrophoresis was carried out at 25 V and 300 mA for
37 min at room temperature. Finally, the slides were
washed with Tris ⁄ HCl buffer (pH 7.5). DNA was stained
using an ethidium bromide solution (0.1 mgÆmL
)1
,50lL
per slide, k
ex
¼ 525 nm and k
em
¼ 650 nm). The slides
were read with an epifluorescence microscope
6
equipped
with a CDD camera (Zeiss Axioskop20, Carl Zeiss,
Microscope Division, Oberkochen, Germany)
7
. The images

were analyzed with the Komet 3.0 image analysis system
(formerly by Kinetic Imaging; now Andor Technology,
Belfast, UK). Fragmentation was expressed in Tail Extent
Moment, taking into account tail length and the percent-
age of DNA in the comet tail. Images of 500 randomly
selected cells were analyzed from each sample.
Extraction and quantification of glutathione
After incubation, 0.5 mL of water was added to the Petri
dishes (100 mm in diameter). At a low temperature, the cell
monolayer was removed by scraping. Cells were collected in
a 5 mL tube, and the volume was adjusted to 1 mL before
addition of 0.2 mL of metaphosphoric acid (6%, p ⁄ v).
After shaking (30 s), the mixture was centrifuged
8
(4000 g,
10 min, 4 °C, CR3i centrifuge and swing-out rotor). The
pellet was used for protein quantification by the Folin–
Lowry method [49]. The supernatant was kept at ) 80 °C
prior to glutathione assay.
In the case of medium samples, 0.2 mL of metaphos-
phoric acid (6%, p ⁄ v) was added to 1 mL of culture med-
ium and treated as described above.
Total glutathione (oxidized and reduced forms) was
quantified as previously described by Tietze [50]. The assay
was performed in a 96-well plate. Twenty microliters
of sample was placed in each well with 150 l L of Mops ⁄
EDTA buffer containing 0.165 UIÆmL
)1
glutathione reduc-
tase. Then, 0.267 mgÆmL

)1
b-NADPH and 75 lLof
5,5¢-dithiobis(2-nitrobenzoic acid) solution (0.04 mg mL
)1
in 0.4 m Mops ⁄ 2mm EDTA, pH 6.75) were added succes-
sively. After shaking, the absorbance was measured at
E. Roudier et al. Pyruvate reduces DNA damage during ⁄ after hypoxia
FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
5195
412 nm using a plate reader spectrophotometer (Thermo
Clinical Labsystems France, Cergy Pontoise, France).
A standard curve was obtained with a commercial gluta-
thione solution (20 mm, Sigma-Aldrich). Standard samples
were treated for extraction using the protocol described for
biological samples.
PCA extraction and
13
C-NMR analysis of cell
extracts
For
13
C-NMR, the cell line was incubated in the presence
of 5.5 mm glucose (unenriched) with 1 mm [
13
C
3
]pyruvate
(Euriso-Top, Saint Aubin, France). Cells were exposed to
both normoxic and hypoxic conditions.

After 6 h of incubation, the medium was discarded and
the cultures were washed twice with cold NaCl ⁄ P
i
, immedi-
ately frozen in liquid nitrogen, and stored at ) 80 °C until
further treatment. PCA extraction was performed following
standard procedures. Briefly, 0.3 mL of 12% PCA was
added to the Petri dishes (100 mm in diameter), the cell
monolayer was removed by scraping with a spatula, and the
cell suspensions obtained from seven individual Petri dishes
were pooled. After homogenization, the final cell suspension
was centrifuged
9
at 8000 g for 10 min (CR3i centrifuge and
swing-out rotor), and the supernatant was adjusted to
pH 7.4 with KOH. The samples were again centrifuged
10
at
8000 g for 10 min (CR3i centrifuge and swing-out rotor),
the supernatant was lyophilized, and the dry residue was
dissolved in 2.2 mL of H
2
O ⁄ 20% D
2
O for NMR analysis.
Proton-decoupled spectra of PCA extracts were recorded
on a Bruker AM400
11
narrow-bore spectrometer equipped
with a 10 mm

31
P ⁄
13
C probe (Bruker, Wissembourg,
France).
13
C-NMR spectra were recorded at 100.62 MHz,
and each spectrum was the sum of 15 000 free induction
decays. A 90° pulse was applied, with a repetition time of
3 s and an acquisition time of 0.819 s. The temperature was
maintained at 20 °C, and the
13
C chemical shifts were refer-
enced to the resonance of tetramethylsilane at 0 p.p.m.
Peak intensities were normalized to the peak intensity of
ethylene glycol, used as an internal reference. Relative areas
of the peaks were normalized to the area of the reference
peak, arbitrarily fixed at 100. The NMR analysis of the
PCA extracts was reproducibly repeated four times for both
normoxic and hypoxic conditions. The data shown are
from a representative experiment.
ROS measurement
The determination of intracellular oxidant production is
based on the oxidation of 2¢,7¢-dichlorodihydrofluorescein
(Sigma-Aldrich) to the fluorescent 2¢,7¢-dichlorofluorescein.
HepG2 cells were incubated in the absence and presence of
pyruvate (1 mm) in DMEM base containing 5 mm glucose
in the fluorescence spectrometer. After addition of
2¢,7¢-dichlorodihydrofluorescein (10 lgÆmL
)1

), fluorescence
emission was measured continuously at 520 nm after excita-
tion at 499 nm. The value of the slope was proportional to
the intracellular ROS levels.
Statistical analysis of results
The experiments were reproducibly repeated four times.
Values are means ± standard deviation (n ¼ 6 separate
Petri dishes). Statistical analysis of the data was done by
anova. The Newman–Keuls unpaired t-test was used to
determine statistical significance.
Acknowledgements
The present study was partially funded by the Institut
Fe
´
de
´
ratif de Recherche (IFR-1) ‘Biomedical NMR:
From Cell to Man’ (Grenoble). In particular, we owe
thanks to Professor J F. Le Bas.
References
1 Warburg O (1956) On respiratory impairment in cancer
cells. Science 124, 269–270.
2 Koukourakis MI, Giatromanolaki A, Sivridis E, Gatter
KC & Harris AL (2005) Pyruvate dehydrogenase and
pyruvate dehydrogenase kinase expression in non small
cell lung cancer and tumor-associated stroma. Neoplasia
7, 1–6.
3 Mazurek S, Boschek CB, Hugo F & Eigenbrodt E
(2005) Pyruvate kinase type M2 and its role in tumor
growth and spreading. Semin Cancer Biol 15, 300–308.

4 Lu H, Forbes RA & Verma A (2002) Hypoxia-inducible
factor 1 activation by aerobic glycolysis implicates the
Warburg effect in carcinogenesis. J Biol Chem 277,
23111–23115.
5 Lu H, Dalgard CL, Mohyeldin A, McFate T, Tait AS
& Verma A (2005) Reversible inactivation of HIF-1
prolyl hydroxylases allows cell metabolism to control
basal HIF-1. J Biol Chem 280, 41928–41939.
6 Lee MS, Moon EJ, Lee SW, Kim MS, Kim KW & Kim
YJ (2001) Angiogenic activity of pyruvic acid in in vivo
and in vitro angiogenesis models. Cancer Res 61, 3290–
3293.
7 Minchinton AI & Fryer KH (1996) Fluctuations in the
oxygenation of experimental tumours. Adv Exp Med
Biol 388, 353–356.
8 Liu L & Simon MC (2004) Regulation of transcription
and translation by hypoxia. Cancer Biol Ther 3, 492–
497.
9 Semenza GL (2003) Targeting HIF-1 for cancer ther-
apy. Nat Rev Cancer 3, 721–732.
10 Semenza GL (2002) HIF-1 and tumor progression:
pathophysiology and therapeutics. Trends Mol Med 8,
S62–S67.
Pyruvate reduces DNA damage during ⁄ after hypoxia E. Roudier et al.
5196 FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
11 Bustamante E, Morris HP & Pedersen PL (1981) Energy
metabolism of tumor cells. Requirement for a form of
hexokinase with a propensity for mitochondrial binding.
J Biol Chem 256, 8699–8704.

12 Mazurek S, Eigenbrodt E, Failing K & Steinberg P
(1999) Alterations in the glycolytic and glutaminolytic
pathways after malignant transformation of rat liver
oval cells. J Cell Physiol 181, 136–146.
13 Meister A & Anderson ME (1983) Glutathione. Annu
Rev Biochem 52, 711–760.
14 O’Dwyer PJ, Yao KS, Ford P, Godwin AK & Clayton
M (1994) Effects of hypoxia on detoxicating enzyme
activity and expression in HT29 colon adenocarcinoma
cells. Cancer Res 54, 3082–3087.
15 Koomagi R, Mattern J & Volm M (1999) Glucose-related
protein (GRP78) and its relationship to the drug-
resistance proteins P170, GST-pi, LRP56 and angio-
genesis in non-small cell lung carcinomas. Anticancer Res
19, 4333–4336.
16 Stankov K & Kovacevic Z (1999) Sensitivity of the pool
of adenine nucleotides to oxidative stress and protective
effect of glutamine. Arch Oncol 7, 141–144.
17 Bassenge E, Sommer O, Schwemmer M & Bunger R
(2000) Antioxidant pyruvate inhibits cardiac formation
of reactive oxygen species through changes in redox
state. Am J Physiol Heart Circ Physiol 279, H2431–
H2438.
18 Mongan PD, Karaian J, Van Der Schuur BM, Via
DK & Sharma P (2003) Pyruvate prevents poly-ADP
ribose polymerase (PARP) activation, oxidative
damage, and pyruvate dehydrogenase deactivation
during hemorrhagic shock in swine. J Surg Res 112,
180–188.
19 Sharma P, Karian J, Sharma S, Liu S & Mongan PD

(2003) Pyruvate ameliorates post ischemic injury of rat
astrocytes and protects them against PARP mediated
cell death. Brain Res 992, 104–113.
20 Ying W, Alano CC, Garnier P & Swanson RA (2005)
NAD+ as a metabolic link between DNA damage and
cell death. J Neurosci Res 79, 216–223.
21 Salahudeen AK, Clark EC & Nath KA (1991) Hydro-
gen peroxide-induced renal injury. A protective role for
pyruvate in vitro and in vivo. J Clin Invest 88, 1886–
1893.
22 Schwartz C, Morgan RL, Way LM & Way JL (1979)
Antagonism of cyanide intoxication with sodium pyru-
vate. Toxicol Appl Pharmacol 50, 437–441.
23 DeBoer LW, Bekx PA, Han L & Steinke L (1993) Pyru-
vate enhances recovery of rat hearts after ischemia and
reperfusion by preventing free radical generation. Am J
Physiol 265, H1571–H1576.
24 Dobsak P, Courderot-Masuyer C, Zeller M, Vergely
C, Laubriet A, Assem M, Eicher JC, Teyssier JR,
Wolf JE, Rochette L et al. (1999)
12
Antioxidative prop-
erties of pyruvate and protection of the ischemic rat
heart during cardioplegia. J Cardiovasc Pharmacol 34,
651–659.
25 Kerr PM, Suleiman MS & Halestrap AP (1999) Rever-
sal of permeability transition during recovery of hearts
from ischemia and its enhancement by pyruvate. Am J
Physiol 276, H496–H502.
26 Borle AB & Stanko RT (1996) Pyruvate reduces anoxic

injury and free radical formation in perfused rat hepato-
cytes. Am J Physiol 270, G535–G540.
27 Halliwell B (2007) Oxidative stress and cancer: have we
moved forward? Biochem J 401, 1–11.
28 Fruehauf JP & Meyskens FLJ (2007) Reactive oxygen
species: a breath of life or death. Clin Cancer Res 13,
789–794.
29 Perrin A, Roudier E, Duborjal H, Bachelet C, Riva-
Lavieille C, Leverve X & Massarelli R (2002) Pyruvate
reverses metabolic effects produced by hypoxia in gli-
oma and hepatoma cell cultures. Biochimie 84, 1003–
1011.
30 Englander EW, Greeley GH Jr, Wang G, Perez-Polo JR
& Lee HM (1999) Hypoxia-induced mitochondrial and
nuclear DNA damage in the rat brain. J Neurosci Res
58, 262–269.
31 Kamiike W, Shimizu S, Hatanaka N, Morimoto Y,
Miyata M, Yoshida Y & Matsuda H (1994) Hypoxia–
reoxygenation causes DNA damage in hepatocytes.
Transplant Proc 26, 907–909.
32 Moller P, Loft S, Lundby C & Olsen NV (2001) Acute
hypoxia and hypoxic exercise induce DNA strand
breaks and oxidative DNA damage in humans. FASEB
J 15, 1181–1186.
33 Speit G & Hartmann A (2006) The comet assay: a sensi-
tive genotoxicity test for the detection of DNA damage
and repair. Methods Mol Biol 314, 275–286.
34 Lertratanangkoon K, Savaraj N, Scimeca JM & Tho-
mas ML (1997) Glutathione depletion-induced thymi-
dylate insufficiency for DNA repair synthesis. Biochem

Biophys Res Commun 234, 470–475.
35 Rigobello MP & Bindoli A (1993) Effect of pyruvate on
rat heart thiol status during ischemia and hypoxia fol-
lowed by reperfusion. Mol Cell Biochem 122, 93–100.
36 Gatenby RA & Gillies RJ (2004) Why do cancers have
high aerobic glycolysis? Nat Rev Cancer 4, 891–899.
37 O’Donnell-Tormey J, Nathan CF, Lanks K, DeBoer CJ
& de la Harpe J (1987) Secretion of pyruvate. An anti-
oxidant defense of mammalian cells. J Exp Med 165 ,
500–514.
38 Van Rooyen J, McCarthy J & Opie LH (2002)
Increased glycolysis during ischaemia mediates the
protective effect of glucose and insulin in the iso-
lated rat heart despite the presence of cardiodepres-
sant exogenous substrates. Cardiovasc J S Afr 13,
103–109.
39 Wheeler TJ, Wiegand CB & Chien S (2005) Fructose-
1,6-bisphosphate enhances hypothermic preservation of
E. Roudier et al. Pyruvate reduces DNA damage during ⁄ after hypoxia
FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1
5197
cardiac myocytes. J Heart Lung Transplant 24, 1378–
1384.
40 Khan S & O’Brien PJ (1997) Rapid and specific efflux
of glutathione before hepatocyte injury induced by
hypoxia. Biochem Biophys Res Commun 238,
320–322.
41 Sze G, Kaplowitz N, Ookhtens M & Lu SC (1993)
Bidirectional membrane transport of intact

glutathione in HepG2 cells. Am J Physiol 265,
G1128–G1134.
42 Mongan PD, Capacchione J, West S, Karaian J,
Dubois D, Keneally R & Sharma P (2002)
Pyruvate improves redox status and decreases indica-
tors of hepatic apoptosis during hemorrhagic shock in
swine. Am J Physiol Heart Circ Physiol 283, H1634–
H1644.
43 Huang ZZ, Chen C, Zeng Z, Yang H, Oh J, Chen L &
Lu SC (2001) Mechanism and significance of increased
glutathione level in human hepatocellular carcinoma
and liver regeneration. FASEB J 15, 19–21.
44 Bump EA, Cerce BA, al-Sarraf R, Pierce SM & Koch
CJ (1992) Radioprotection of DNA in isolated nuclei by
naturally occurring thiols at intermediate oxygen ten-
sion. Radiat Res 132, 94–104.
45 Godwin AK, Meister A, O’Dwyer PJ, Huang CS, Ham-
ilton TC & Anderson ME (1992) High resistance to cis-
platin in human ovarian cancer cell lines is associated
with marked increase of glutathione synthesis. Proc Natl
Acad Sci USA 89, 3070–3074.
46 Aden DP, Fogel A, Plotkin S, Damjanov I & Knowles
BB (1979) Controlled synthesis of HBsAg in a differen-
tiated human liver carcinoma-derived cell line. Nature
282, 615–616.
47 Marbach EP & Weil MH (1967) Rapid enzymatic mea-
surement of blood lactate and pyruvate. Use and signifi-
cance of metaphosphoric acid as a common precipitant.
Clin Chem 13, 314–325.
48 Singh NP, McCoy MT, Tice RR & Schneider EL (1988)

A simple technique for quantitation of low levels of
DNA damage in individual cells. Exp Cell Res 175,
184–191.
49 Lowry OH, Rosebrough NJ, Farr AL & Randall RJ
(1951) Protein measurement with the Folin phenol
reagent. J Biol Chem 193, 265–275.
50 Tietze F (1969) Enzymic method for quantitative deter-
mination of nanogram amounts of total and oxidized
glutathione: applications to mammalian blood and other
tissues. Anal Biochem 27, 502–522.
Pyruvate reduces DNA damage during ⁄ after hypoxia E. Roudier et al.
5198 FEBS Journal 274 (2007) 5188–5198 Journal compilation ª 2007 FEBS. No claim to original French government works
1