ORIGINAL RESEARCH ARTICLE

Journal of

Cellular
Physiology

Permanent Culture of
Macrophages at Physiological
Oxygen Attenuates the
Antioxidant and Immunomodulatory
Properties of Dimethyl Fumarate
BENJAMIN HAAS,1,2 SANDRA CHRUSCIEL,1,3 SARAH FAYAD-KOBEISSI,1,2
JEAN-LUC DUBOIS-RAND,4 FRANCISCO AZUAJE,5 JORGE BOCZKOWSKI,1,3
ROBERTO MOTTERLINI,1,2* AND ROBERTA FORESTI1,2*
1

Universit
e Paris-Est, UMR_S955, UPEC, France

INSERM, U955, Equipe 03, France

INSERM, U955, Equipe 04, France

AP-HP, H^opital Henri Mondor, Service de Cardiologie 1, Cr
eteil, France

CRP-Sant
e, Norlux Neuro-Oncology Laboratory, Luxembourg, Luxembourg

We hypothesized that O2 tension inuences the redox state and the immunomodulatory responses of inammatory cells to dimethyl
fumarate (DMF), an activator of the nuclear factor Nrf2 that controls antioxidant genes expression. This concept was investigated in
macrophages permanently cultured at either physiological (5% O2) or atmospheric (20% O2) oxygen levels and then treated with DMF or
challenged with lipopolysaccharide (LPS) to induce inammation. RAW 264.7 macrophages cultured at 20% O2 exhibited a pro-oxidant
phenotype, reected by a lower content of reduced glutathione, higher oxidized glutathione and increased production of reactive oxygen
species when compared to macrophages continuously grown at 5% O2. At 20% O2, DMF induced a stronger antioxidant response
compared to 5% O2 as evidenced by a higher expression of heme oxygenase-1, NAD(P)H:quinone oxydoreductase-1 and superoxide
dismutase-2. After challenge of macrophages with LPS, several pro-inammatory (iNOS, TNF-a, MMP-2, MMP-9), anti-inammatory
(arginase-1, IL-10) and pro-angiogenic (VEGF-A) mediators were evaluated in the presence or absence of DMF. All markers, with few
interesting exceptions, were signicantly reduced at 5% O2. This study brings new insights on the effects of O2 in the cellular adaptation to
oxidative and inammatory stimuli and highlights the importance of characterizing the effects of chemicals and drugs at physiologically
relevant O2 tension. Our results demonstrate that the common practice of culturing cells at atmospheric O2 drives the endogenous
cellular environment towards an oxidative stress phenotype, affecting inammation and the expression of antioxidant pathways by
exogenous modulators.
J. Cell. Physiol. 230: 11281138, 2015. 2014 Wiley Periodicals, Inc.

Common practice work in vitro relies on culture of cells in

incubators where the oxygen (O2) tension corresponds to
atmospheric levels (20% O2). However, the endogenous
physiological O2 tension found in tissues is in the range of 25%
(Tsai et al., 2003; Carreau et al., 2011) suggesting that this
discrepancy may lead to misinterpretation of results and may
explain why effects observed in vitro cannot always be

reproduced in vivo and vice versa. Indeed, culturing cells at low

O2 tension dramatically affects their phenotype through
modications of primary cellular function, like proliferation
rate in lymphocytes (Atkuri et al., 2005), myoblasts (Duguez
et al., 2012) as well as the phagocytic capacity of macrophages
(Grodzki et al., 2013). Cell culture O2 tension also signicantly
impacts on cell metabolism and gene expression (Parrinello

Abbreviations: ARG1, arginase 1; DCF-DA, 20 ,70 -dicholorouorescein diacetate; DMEM, Dulbeccos Modied Eagle Medium; DMF,
dimethyl fumarate; DTNB, 5,50 -dithiobis(2-nitrobenzoic acid); FBS,
foetal bovine serum; GSH, reduced glutathione; GSSG, oxidized
glutathione; HO-1, heme oxygenase-1; HPRT1, hypoxanthine
phosphoribosyltransferase 1; IL-10, interleukin-10; iNOS, inducible
nitric oxide synthase; LPS, lipopolysaccharide; MMP, matrix
metalloproteinase; NADPH, nicotinamide adenine dinucleotide
phosphate; NQO-1, NAD(P)H;quinone oxydoreductase-1; Nrf2,
nuclear factor (erythroid-derived 2)-like 2; PBS, phosphate buffered
saline; PTGES, prostaglandin E synthase; ROS, reactive oxygen
species; SOD, superoxide dismutase; TNF-a, tumor necrosis factor
alpha; VEGF, vascular endothelial growth factor.

from the Agence National de la Recherche;

Contract grant number: 12-ISV50001.
Contract grant sponsor: Fonds National de la Recherche
Luxembourg;
Contract grant number: AFR 3979613.

Conict of interest: Authors have nothing to declare.

Accepted manuscript online in Wiley Online Library

(wileyonlinelibrary.com): 9 October 2014.
DOI: 10.1002/jcp.24844

Contract grant sponsor: AREMCAR foundation;

Contract grant number: HEMOXY-2.
Contract grant sponsor: Blanc II International Grant MITO-CO
2 0 1 4 W I L E Y P E R I O D I C A L S , I N C .

*Correspondence to: Roberto Motterlini or Dr. Roberta Foresti,

INSERM U955 Equipe 3, Faculty of Medicine, University of Paris Est
Creteil, 8 rue du G
en
eral Sarrail, 94010 Cr
eteil, France.
E-mail: roberta.foresti@inserm.fr (R.F),
roberto.motterlini@inserm.fr (R.M)
Manuscript Received: 29 July 2014
Manuscript Accepted: 1 October 2014

1128

RESPONSES OF MACROPHAGES AT PHYSIOLOGICAL OXYGEN

et al., 2003; Wood et al., 2011). Interestingly, cells grown at

physiological O2 are also characterized by a redox status that is
more similar to that found in vivo, with a lower reactive oxygen
species (ROS) production (Halliwell, 2003). These fundamental
observations should prompt us to regularly adopt physiological
O2 levels in investigations carried out in cultured cells and it is
surprising to note that only few studies in vitro are conducted
by growing cells constantly at reduced O2 oxygen tensions. The
implementation of these revised cell culture conditions could
become a robust complementary approach to in vivo
experimentations enabling an improved characterization of
cellular biological responses to external and internal stimuli or
stresses.
One cellular response that may be directly dependent on O2
levels is the induction of antioxidant enzymes as a consequence
of oxidative stress. The nuclear factor (erythroid-derived 2)like 2 (Nrf2) transcription factor, which controls the
expression of a wide array of antioxidant genes such as heme
oxygenase-1 (HO-1) and NAD(P)H:quinone oxydoreductase-1
(NQO1), is considered an essential regulator of cellular
stress adaptation (Prestera and Talalay, 1995; Alam et al., 1999;
Ramos-Gomez et al., 2001; Dinkova-Kostova and Talalay,
2008). Interestingly, Nrf2 activation and HO-1 induction are
promoted by a series of molecules possessing electrophilic
properties and similar structures (Satoh et al., 2006; Satoh
et al., 2013; Motterlini and Foresti, 2014). Thus, naturallyderived substances, such as sulforaphane, carnosol, caffeic acid
phenylester, and curcumin, but also endogenously generated
compounds including fumarate, exert protective effects by
allowing the nuclear translocation of Nrf2 and the transcription
of defensive proteins (Ishii et al., 2000; Dinkova-Kostova et al.,
2002; Balogun et al., 2003; Yates and Kensler, 2007; Ashraan
et al., 2012; Foresti et al., 2013). Some of these compounds are
investigated for their potential therapeutic use and dimethyl
fumarate (DMF), a derivative of fumarate, is currently
employed for the treatment of psoriasis (Ghoreschi et al.,
2011) and has been reported to elicit cardioprotective
(Ashraan et al., 2012) and neuroprotective effects (Linker
et al., 2011; Albrecht et al., 2012). It is possible that the
benecial properties of DMF stem from a modulation of
anti-oxidant and inammatory responses (Meissner et al.,
2012) and, considering that inammation is tightly linked to the
redox state of the cells (Hu et al., 2010; Brune et al., 2013),
drugs that target both processes may be of therapeutic
interest.
The aim of this study was to investigate how an O2
tension that more closely reects a physiological situation
affects the antioxidant and immunomodulatory properties of
the Nrf2 activator DMF. To achieve this, antioxidant and
inammatory factors were evaluated in murine macrophages
permanently growing at 20% or 5% O2 following challenge
with lipopolysaccharide (LPS) in the presence or absence of
DMF.
Materials and Methods
Reagents
Dimethyl fumarate (DMF), N-acetylcysteine (NAC) and all other
chemicals were obtained from Sigma Aldrich (St. Louis, MO) unless
otherwise specied.
Cell culture
RAW 264.7 murine macrophages (ATCC, Teddington, Middlesex, UK) were grown in DMEM containing 4.5 g/L glucose and
Glutamax1 (Life Technologies, Saint-Aubin, France) supplemented with 10% ftal bovine serum (Lonza) and penicillin
(100 U/ml) streptomycin (100 mg/ml) (Life Technologies)
following the recommendations of ATCC for cell culture.
JOURNAL OF CELLULAR PHYSIOLOGY

Macrophages were either continuously cultured under standard

culture conditions, that is, at atmospheric oxygen tension (20%
O2), 5% CO2 and 37C in a normal incubator, or at 5% O2, an
oxygen tension that more closely resembles the physiological
situation (Atkuri et al., 2007; Carreau et al., 2011). The permanent
culture of cells at 5% O2 was achieved by seeding, growing and
sub-culturing cells in a glove box chamber (Coy Laboratories,
Grass Lake, MI) with 5% CO2 at 37C. Seeded cells were allowed
at least three sub-cultures under 5% O2 before starting any
treatment. Cells at 90% conuence were treated with dimethyl
fumarate (550 mM), LPS (from E. coli 026:B6 10 ng/ml) (a
prototypic molecule classically used to induce inammation in
macrophages (Abuarqoub et al., 2006)), menadione (25 mM),
N-acetylcysteine (1 mM), or a combination of treatments as
indicated. Cells and cell culture supernatants were collected and
stored at 80C until analysis.
Pericellular detection of oxygen and measurement of
oxygen consumption
The importance of continuous non-invasive monitoring of O2 at
the pericellular levels in adherent cells have been previously
emphasized (Bambrick et al., 2011). Pericellular measurements of
O2 implies placement of the sensor as close as possible to the
growing cells. This was achieved by detection of the O2
concentration in cultured cells in real time using a Sensor Dish
Reader (PreSens, Regensburg, Germany). Briey, cells were
seeded at 5 or 20% O2 in a 24-well plates coated with an oxygen
detection probe (Oxodish) and changes in oxygen concentrations
were recorded continuously for 24 h. From these values, oxygen
consumption was calculated after assessing the O2 diffusion rate
into the medium at 5 and 20% O2 according to a previously
published report (Degn et al., 1973). Oxygen consumption rate
was expressed as picomoles O2/min.
Determination of reduced and oxidized glutathione
Reduced (GSH) and oxidized glutathione (GSSG) were determined in untreated RAW macrophages 24 h after sub-culturing
them at 5 or 20% O2. GSH was also measured in cells cultured at
the two O2 tensions and exposed for 0, 3, and 6 h to 25 mM DMF.
At the end of the incubation, GSH was measured in deproteinized
cell lysates (2% w/v sulfosalicylic acid in distilled water) incubated
for 5 min with 5,50 -dithiobis(2-nitrobenzoic acid) (DTNB 0.2 mM
in 0,3 M sodium phosphate, 10 mM EDTA, pH 7,5) in the dark.
Samples were read at 412 nm and GSH content was calculated
based on a GSH standard curve as previously described (Motterlini
et al., 2000). Results were expressed as mM of GSH over mg of
protein. GSSG content was determined according to the methods
by Rahman and coworkers (Rahman et al., 2006) and detailed in the
Supplementary Information.
ROS production
Basal ROS production was measured in untreated cells cultured at
5 or 20% O2. ROS were also determined in cells treated with
10 mM DMF for 6 h prior to exposure for 45 min to 25 mM
menadione, a generator of superoxide anion. Cells were treated
with 40 mM 20 ,70 -dicholorouorescein diacetate (DCF-DA) for
30 min at 37C and ROS generation was then measured by a
uorescence plate reader (Tecan, Grdig, Austria) set at 480 nm
for excitation and 520 nm for emission. Results were expressed as
mean uorescence intensity.
ATP measurements
ATP production was measured using an ATPLiteTM bioluminescent assay kit (Perkin Elmer, Courtaboeuf, France) following the
manufacturers instructions.

1129

1130

HAAS ET AL.

Heme oxygenase activity assay

Cells growing at 5 or 20% O2 were exposed for 6 h to increasing
concentrations of DMF (550 mM) or pretreated with 1 mM NAC
for 18 h prior to exposure to DMF for 6 h. Heme oxygenase
activity assay was performed using a procedure previously
described (Vesely et al., 1998; Sawle et al., 2005; Abuarqoub et al.,
2006) and detailed in the Supplementary Information. Results were
expressed as pmol bilirubin/mg of protein/h.
Western immunoblotting
HO-1 protein expression was determined in cell samples treated
as for the heme oxygenase activity assay (Sawle et al., 2005; Sawle
et al., 2008). The expression of inducible nitric oxide synthase
(iNOS) protein was also measured in samples of cells growing at 5
or 20% O2 and treated for 6 h with 10 ng/ml LPS in the presence or
absence of DMF (1050 mM). Detailed procedures are reported in
the Supplementary Information.
Subcellular fractionation and detection of Nrf2
RAW 264.7 macrophages growing at 5 or 20% O2 were stimulated
with 25 mM DMF for 1 and 3 h to assess nuclear translocation of
Nrf2. Nuclear enrichment was performed using a commercial
nuclear isolation kit from Active Motif (La Hulpe, Belgium)
according to the manufacturers instructions. Protein concentration of the nuclear fractions was determined using the BCA
protein assay kit. Nrf2 antibody (clone C-20 rabbit polyclonal,
Santa Cruz Biotechnology, Santa Cruz, CA) was used to
determine the nuclear presence of Nrf2 in response to treatments.

Lamin A/C (clone N-18, goat polyclonal, Santacruz Biotechnology)

was used as a nuclear control.
Real-time quantitative polymerase chain reaction
Cells at 5 or 20% O2 were challenged for 6 h with 10 ng/ml LPS in
the presence or absence 25 mM DMF and total RNA was extracted
using the RNeasy Protect Mini kit (Qiagen, Courtboeuf Cedex,
France) according to the manufacturers instructions. Contamination by genomic DNA was prevented by a DNase I treatment
(NucleoSpin1, Macherey Nagel, Germany). RNA concentrations
were measured by NanoDrop spectrophotometer (Thermo
Scientic, Wilmington, DE) and the RNA Integrity Number (RIN)
by Bioanalyzer (Agilent, Les Ulis, France). cDNA synthesis was
carried out with 5 mg of the total RNA using a reverse
transcription (RT) kit (Promega, Lyon, France) and a Biometra
TGradient Thermocycler (Biometra GmBH, Goettingen, Germany). The oligonucleotide primers (Sigma Aldrich, St. Louis, MO)
for RT-PCR used in this study are reported in Supplementary
Table 1. Equal amounts of cDNA, an equivalent of 10 ng of RNA,
were used for quantitative RT-PCR reactions carried out in a ABI
PRISM 7000 Sequence Detection System (Applied Biosystems, Life
Technologies) using Platinum SYBRGreen qPCR SuperMix-UDG
with ROX (Life Technologies). Hypoxanthine phosphoribosyltransferase 1 (HPRT1) was used as gene of reference, since its
mRNA levels were stably expressed at 20% and 5% oxygen.
Nitrite production
Quantication of nitrite production was performed using the
Griess method following challenge of cells with 10 ng/ml LPS in the

Fig. 1. Oxygen consumption, metabolic and oxidative stress profiles of RAW 264.7 macrophages permanently grown at 5% and 20% O2.
(A) Oxygen consumption rate and (B) oxygen concentration were measured at the pericellular level over a 24 h period in macrophages that
have been permanently growing at 20% O2 or 5% O2. (C) Cells grown at 20% and 5% O2 were seeded with fresh medium and ATP levels
measured after 6 and 24 h. (D and E) Cells grown at 20% and 5% O2 were seeded with fresh medium and GSH and GSSG content assessed
after 24 h. (F) Cells grown at 20% and 5% O2 were seeded with fresh medium and incubated in the presence of the fluorescent probe
20 ,70 -dicholorofluorescein diacetate. Basal ROS production was measured after 6 h and data expressed as mean fluorescence intensity (MFI).
Data are expressed as mean  SEM. *P < 0.05 vs. 6 h (C) or 20% O2 (D,E,F), yP < 0.05 vs. 20% O2 (C).

JOURNAL OF CELLULAR PHYSIOLOGY

RESPONSES OF MACROPHAGES AT PHYSIOLOGICAL OXYGEN

presence or absence DMF (10, 25, 50 mM) (Sawle et al., 2005;

Motterlini et al., 2013). Briey, cell culture supernatants were
equally mixed with Griess reagent (2 mM N-(1-naphtyl) ethylenediamide, 30 mM sulfanilamide, 140 mM o-phosphoric acid),
incubated for 10 min at room temperature and samples read at
550 nm. Concentrations of nitrite were calculated based on a
standard curve using sodium nitrite and expressed in mM.
Zymography assay
RAW macrophages were stimulated for 24 h with 10 ng/ml LPS in
the presence or absence of DMF (1050 mM). Cell culture
supernatants were harvested, treated with a protease inhibitor
cocktail without EDTA and stored at 80C before measurement
of MMP-9 activity. Briey, equal volumes of supernatants were
resolved on 8% acrylamide gel containing 20 mg of gelatin. Gels
were washed for 3  20 min in a renaturing solution (0,25% Triton
X-1001) to recover MMP-9 integrity and incubated overnight at
37C in a developing solution to recover gelatinolytic activity

(0,5 M Tris, 2 M NaCl, 50 mM CaCl2, pH 7.6). Gels were then

stained with coomassie blue for 15 min, and destained in 30%
methanol/10% acetic acid solution for 5 min. Gels were placed in a
contrasting solution (5% glycerol, 10% acetic acid) until image
capture. Images were captured with GeneSys software in a G:Box
F3 imagery station (Syngene, Cambridge, UK).
Isolation and culture of murine peritoneal
macrophages
Studies were conducted in compliance with INSERM guidelines
regarding the fair treatment of animals, under a license from the
French administration to conduct animal research. C57Bl/6J mice
(Janvier, Le Genest Saint-Isle, France) were injected intraperitoneally with 1 ml of sterile 4% thioglycolate solution. After
24 h, peritoneal macrophages were harvested via three successive
peritoneal lavages using phosphate-buffered saline (PBS). The
exudates were centrifuged at 1500 rpm for 10 min at 4C and
pellets were suspended and pooled in Dulbeccos Minimum Eagle
Medium (DMEM) containing 4.5 g/L glucose and Glutamax1
supplemented with 10% fetal bovine serum and penicillin (100 U/
ml) streptomycin (100 mg/ml). Cells were resuspended at a nal
concentration of 2  106 cells per well and plated in a 6-well
microplate. After 2 h of sedimentation, the culture medium was
changed and cells allowed to rest overnight at either 5 or 20% O2.
The following day macrophages were stimulated with LPS in the
presence or absence of 25 mM DMF and both cells and culture
supernatants were collected and stored at 80C for detection of
HO-1 protein expression and Tumor Necrosis Factor-a (TNF-a)
analysis.
Enzyme-linked immunosorbent assay
TNF-a production was assessed in supernatants of primary
macrophages collected 24 h after different treatments (see above).
An ELISA TNF-a mouse colorimetric kit (Enzo Life Sciences,
Villeurbanne, France) was used according to the manufacturers
instructions. Results were expressed in pg/ml.
Statistical analysis
All data are expressed as mean  S.E.M. Students t-test or, when
required, one-way ANOVA with Bonferroni post hoc tests were
used for comparison of experimental data. Analyses were
performed using GraphPad Prism software v 5.0 (San Diego, CA).
A P-value <0.05 was considered signicant.
Results
Metabolic and oxidative stress parameters
in RAW macrophages permanently cultured
at 5 or 20% O2

Fig. 2. Differential expression of HO-1 by dimethyl fumarate in

RAW 264.7 macrophages cultured at 20% and 5% O2. (A) HO
activity and (B) HO-1 protein expression in macrophages grown at
20% O2 or 5% O2 were measured after 6 h treatment with 550 mM
dimethyl fumarate (DMF). (C) Nuclear translocation of Nrf2 was
measured in macrophages grown at 20% O2 or 5% O2 after
treatment for 1 and 3 h with 25 mM DMF. Data are presented as
mean  SEM. *P < 0.05 vs. control (0 mM DMF); #P < 0.05 vs. 20% O2.
HO-1 and Nrf2 immunoblottings are representative of three
independent experiments.

JOURNAL OF CELLULAR PHYSIOLOGY

In order to characterize the phenotype of RAW 264.7

macrophages grown under 5% or 20% O2, we rst measured
the O2 consumption rate (OCR) and the changes in the
concentration of O2 at the pericellular level over a period of
24 h after cell seeding. It is important to emphasize that 5% O2
was selected to reect the average O2 concentration found in
tissues as previously described (Atkuri et al., 2005; Atkuri et al.,
2007). Our results show that OCR was similar between 20%
and 5% O2 (Fig. 1A), even though the quantity of O2 dissolved
into the medium (Fig. 1B) was markedly higher in cells cultured
at 20% compared to 5% O2. We also measured the ATP
content in cells cultured under both O2 tensions for 6 and 15 h
after seeding and observed an increase of ATP at 15 h in both
conditions (Fig. 1C), which may be a consequence of cell
growth and progressive nutrient consumption over time to
maintain cellular function. Interestingly, ATP production in cells
at 5% O2 in the presence of 2-deoxy-D-glucose (10 mM), an

1131

1132

HAAS ET AL.

inhibitor of glycolysis, was similar to that of control cells after

6 h of treatment (70.0  7.5 vs. 67.8  2.3 mM/well, respectively) suggesting that cells grown at 5% O2 do not rely
primarily on glycolysis for energy production. We then wanted
to determine the redox status of the cells by evaluating their
glutathione content and basal ROS levels at the two O2
tensions examined. We observed that macrophages grown at
20% O2 contained a lower amount of reduced glutathione

(GSH) and higher levels of oxidized glutathione (GSSG)

compared to 5% O2 (Fig. 1D and 1E, respectively). In addition,
basal ROS production in cells cultured at 20% O2 was nearly
twice than that detected at 5% O2 (Fig. 1F). Together, these
results support the hypothesis that cells cultured at
atmospheric oxygen tension in standard incubators drives the
endogenous cellular environment towards an oxidative stress
phenotype, at least when compared to 5% O2.

Fig. 3. Modulation of the antioxidant response to dimethyl fumarate (DMF) by oxygen tension in RAW 264.7 macrophages. (A) RAW 264.7
macrophages permanently growing at 20% O2 or 5% O2 were treated with 25 mM DMF and GSH levels were assessed at 0, 3, and 6 h.
(B) Macrophages grown at 20% O2 or 5% O2 were pre-treated for 6 h with 10 m DMF. ROS production was measured in response to menadione
(25 mM for 45 min) using 20 ,70 -dichlorofluorescein diacetate. (C and D) Macrophages cultured at 20% were pre-treated for 18 h with or without
N-acetylcysteine (1 mM) and then incubated for 6 h in the presence of increasing concentrations of DMF. Heme oxygenase activity and HO-1
protein expression were measured in whole cell lysates at the end of the incubation as reported in Materials and Methods. (E and F)
Macrophages cultured at 5% were pre-treated for 18 h with or without N-acetylcysteine (1 mM) and then incubated for 6 h in the presence of
increasing concentrations of DMF. Heme oxygenase activity and HO-1 protein expression were measured at the end of the incubation. Data
are presented as mean  SEM. HO-1 immunoblottings are representative of three independent experiments. *P < 0.05 vs. control (0 mM
DMF); #P < 0.05 vs. 20% O2 or vs. DMF without NAC.

JOURNAL OF CELLULAR PHYSIOLOGY

RESPONSES OF MACROPHAGES AT PHYSIOLOGICAL OXYGEN

Impact of oxygen tension on the induction of the cellular

antioxidant response

Having conrmed substantial differences in the redox status of

cells growing at the two O2 tensions, we tested whether the
inducible antioxidant response is also affected by O2 levels. For
these experiments, we incubated cells for 6 h with DMF, an
activator of Nrf2 and inducer of HO-1 potentially possessing
anti-inammatory characteristics. As shown in Figure 2A and
2B, DMF (550 mM) increased heme oxygenase activity and
HO-1 protein expression in a concentration-dependent
manner at both O2 tensions; however, the extent of this
induction was signicantly less pronounced in cells cultured at
5% O2 (see also quantied bands in Supplementary Figure 1).
Interestingly, the nuclear translocation of Nrf2 elicited by DMF
was evident after 1 and 3 h treatment at both O2 tensions
(Fig. 2C). Since Nrf2 is a master regulator of the cellular
antioxidant response and controls a variety of genes, including
two key enzymes involved in glutathione synthesis (Lehmann
et al., 2007), we measured the effect of 25 mM DMF on GSH
synthesis at 0, 3, and 6 h in RAW 264.7 macrophages cultured
at 20% and 5% O2. We rst observed signicantly higher levels
of GSH at basal conditions in RAW macrophages permanently
cultured at 5% O2. In addition, we found that DMF signicantly
increased GSH 6 h after treatment at both O2 tensions, with a
higher potency at 20% O2. (Fig. 3A) We also measured ROS
production after challenge of RAW macrophages with
menadione, a generator of superoxide anion, in the presence
or absence of 25 mM DMF at the two O2 tensions. Firstly, we
observed that DMF did not signicantly alter basal ROS levels
after 6 h incubation at both O2 tensions (Fig. 3B). Secondly,
ROS production stimulated by menadione was higher at 20%
O2 than at 5% O2 and DMF partially reduced this effect at 20%
O2. Notably, when RAW macrophages were pre-treated for
18 h with 1 mM N-acetylcysteine (NAC), an antioxidant agent
used as a precursor of glutathione synthesis, the increase in
heme oxygenase activity and HO-1 expression caused by DMF
was signicantly reduced at 20% O2 (Fig. 3C and D), but not at
5% O2 (Fig. 3E and F). Finally, we also evaluated the mRNA
expression of HO-1 and two other antioxidant enzymes,
NQO-1 and SOD2, in response to DMF in macrophages
cultured at both O2 tensions. As shown in Figure 4A and B,
HO-1 and NQO-1 mRNAs were signicantly increased 6 h
after DMF treatment, but this increase was less prominent at 5
than 20% O2. In contrast, SOD2 expression remained
unaffected by DMF (Fig. 3C), although the basal mRNA of this
enzyme was signicantly lower at 5% O2 compared to 20% O2.
Taken together, these results indicate that 20% O2 tension
promotes an endogenous pro-oxidative phenotype that is
evident both at basal conditions and in response to exogenous
oxidant stimulation. Furthermore, the antioxidant response
stimulated by DMF is also clearly affected by the different O2
levels as evidenced by a stronger HO-1 induction at 20%
compared to 5%.
Impact of oxygen tension on the inammatory response
induced by lipopolysaccharide in RAW 264.7
macrophages

DMF is recognized as an activator of the antioxidant response

by its modulation of the Nrf2/HO-1 axis (Lehmann et al., 2007).
In view of the interesting ndings related to the antioxidant
response obtained when comparing cells grown at atmospheric
and physiological O2 tensions, we also examined whether the
inammatory response, typically mounted in macrophages
challenged with infectious agents, could be modulated by O2.
To guide us on the selection of biologically relevant potential
targets of DMF related to inammation, we established a
computational prediction model on chemical-protein interJOURNAL OF CELLULAR PHYSIOLOGY

Fig. 4. Differential induction of antioxidant genes by DMF in RAW

264.7 macrophages cultured at 20% and 5% O2. HO-1 (A), NQO-1
(B) and SOD2 (C) mRNA expression in RAW 264.7 macrophages
permanently grown at 20% O2 or 5% O2. Control cells (untreated)
and cells treated for 6 h with 25 mM DMF were analyzed. Total RNA
was extracted from cells at the end of the incubation and gene
expression was measured by real-time quantitative polymerase
chain reaction as reported in Materials and Methods. Hprt1 was
used as a housekeeping gene. Data are presented as mean  SEM.
*P < 0.05 vs. control (CON); #P < 0.05 vs. 20% O2.

action (STICH 3.1) centered on DMF (Supplementary

Information and Supplementary Fig. 2A) and then expanded to
other Nrf2 activators with similar chemical structure such as
curcumin, CAPE and carnosol (Supplementary Fig. 2B). This
approach is based on the computational identication of drugtarget interactions and their characterization at a systems level
(Azuaje, 2013). By cross-comparing the lists of potential
inammatory targets of these compounds, we selected iNOS,
TNF-a, MMP-2, MMP-9, PTGES, and VEGF-A (Supplementary
Table 2) and measured their modulation in response to DMF at
20 and 5% O2. It is important to note that HO-1, NQO-1 and
Nrf2 appeared in every network generated and SOD-2 also
appeared in curcumin and CAPE networks (Supplementary
Table 2). We found that RAW 264.7 macrophages exposed to
LPS at 20% O2 exhibited an increased iNOS mRNA (Fig. 5A)

1133

1134

HAAS ET AL.

conducted in RAW macrophages cultured in medium

containing high glucose (4.5 g/L), following cell culture
recommendations. This medium is largely used in in vitro
studies although the presence of high amount of glucose will
facilitate inammatory response (Freemerman et al., 2014).
Indeed, our group recently found that even at 20% oxygen a
reduction of glucose in culture to 5 mM leads to a signicantly
lower production of nitrite in LPS-stimulated macrophages
(unpublished observations). We performed qPCR to measure
the mRNA expression of TNF-a PTGES, MMP-2 and MMP-9
(pro-inammatory mediators), VEGF-A (pro-angiogenic) and
the anti-inammatory markers arginase-1 (ARG1) and IL-10.
ARG1 and IL-10 were not suggested by the prediction model
but measuring their expressions was of particular interest as
ARG1 competes with iNOS for the substrate arginine and IL-10
is described as an anti-inammatory and anti-angiogenic
cytokine (Wu et al., 2010). Interestingly, our results shows that
TNF-a, PTGES and MMP-2 were signicantly increased by LPS
at 20% O2 while their production was markedly attenuated in
macrophages cultured at 5% O2 (Figs. 6A, B, 7A and C). In the
case of the anti-inammatory response, we found that IL-10
was more expressed at 20% than 5% O2, whereas ARG1
mRNA was slightly increased by LPS at 20% O2 but was
extremely highly expressed at 5% O2 (Fig. 6C and D,
respectively). Treatment of cells with DMF resulted, in the
majority of the cases, in a down-regulation of both pro- and
anti-inammatory mediators after LPS challenge at both O2
tensions (Fig. 6AD and Fig. 7A and C). Particular attention was
given to MMP-9 and VEGF-A, regulators of cellular migration
and angiogenesis, respectively. First of all, in line with the other
pro-inammatory cytokines, we found that MMP-9 mRNA
levels and activity (zymography assay) were higher in cells
cultured at 20% compared to 5% O2 (Fig. 7C and D). Secondly,
DMF down-regulated MMP-9 at both O2 tension but was
accompanied by an increase in the synthesis of VEGF-A mRNA
only at 5% O2 (Fig. 7B).
Impact of oxygen tension on the oxidative stress and antiinammatory response in murine primary macrophages

Fig. 5. Modulation of the iNOS pathway in response to LPS in

RAW 264.7 macrophages cultured at 20% and 5% O2. iNOS mRNA
(A), protein expression (B) and nitrite production (C) were
measured in response to lipopolysaccharide (LPS) in RAW 264.7
macrophages permanently grown at 20% O2 or 5% O2. Cells were
challenged with 10 ng/ml LPS in the presence or absence of 25 mM
DMF before assessement of iNOS mRNA and protein expression.
Nitrite production was measured in cell culture supernatants 24 h
treatment with LPS in combination with increasing concentrations
of DMF (10 to 50 mM). Data are expressed as mean  S.E.M.
iNOS immunoblottings are representative of three independent
experiments. *P < 0.05 vs. control; yP < 0.05 vs. LPS alone;
#
P < 0.05 vs. 20% O2.

and protein expression (Fig. 5B). At 5% O2, iNOS mRNA was

still induced by LPS but to a signicantly lower extent compared
to 20% O2 (Fig. 5A), while iNOS protein expression was slightly
higher at 5% O2 but not signicantly different from 20% (as
detected by densitometric analysis, not shown) (Fig. 5B).
Nitrite production following LPS stimulation was also higher at
20% than at 5% O2 (Fig. 5C), reecting the pattern observed for
iNOS mRNA expression at the two O2 tensions. Figure 5B and
C also reveals that treatment of cells with increasing
concentrations of DMF (1050 mM) signicantly decreased
iNOS protein expression as well as nitrite production at both
O2 tensions. We note that all our experiments were
JOURNAL OF CELLULAR PHYSIOLOGY

To conrm that our ndings on the differential oxidative stress

and anti-inammatory response at 20 and 5% O2 is not
restricted to a macrophage cell line, we performed additional
experiments using primary murine macrophages collected
from the peritoneal cavity of mice after stimulation with
thioglycolate. Our results show that a lower induction of HO-1
was observed in cells treated with 25 mM DMF at 5% compared
to 20% O2 (Fig. 8A). We also measured the production of TNFa protein in murine primary macrophages treated with LPS and
DMF at 20% and 5% O2. Interestingly, also primary macrophages challenged with LPS produced TNF-a only at 20% O2
and this effect was markedly attenuated in the presence of
DMF. TNF-a production was virtually undetectable at 5% O2
(Fig. 8B) and we were unable to detect any nitrite generation in
these macrophages at either O2 tensions (data not shown).
These ndings support the idea that macrophages growing at
20% O2 are primed to stimulate an enhanced LPS-inammatory
response compared to that elicited at 5% O2, in parallel to the
pro-oxidant status observed at 20% O2.
Discussion

O2 concentrations in tissues are maintained between 25%

while cell culture studies are normally conducted in incubators
at atmospheric O2 tension, which is around 20%. Thus, the
concentration of O2 used in cells grown in vitro effectively
mimic hyperoxic conditions. From a biochemical perspective,
O2 levels are expected to greatly inuence several physiological
and pathophysiological processes (Davies, 2000; Finkel and

RESPONSES OF MACROPHAGES AT PHYSIOLOGICAL OXYGEN

Fig. 6. Differential induction of pro-inflammatory and anti-inflammatory mediators in response to LPS in RAW 264.7 macrophages cultured
at 20% and 5% O2. RAW 264.7 macrophages were permanently grown at 20% O2 or 5% O2 and then challenged with 10 ng/ml LPS in the
presence or absence of 25 mM DMF. After 6 h incubation, total RNA was extracted and gene expression was measured by real-time
quantitative polymerase chain reaction as reported in Materials and Methods. The pro-inflammatory mediators analyzed were TNF-a (A),
PTGES (B), while the anti-inflammatory genes were IL-10 (C) and Arg1 (D). Hprt1 was used as a housekeeping gene. Data are expressed as
mean  S.E.M. *P < 0.05 vs. control; yP < 0.05 vs. LPS alone; #P < 0.05 vs. 20% O2.

Holbrook, 2000; Trachootham et al., 2008) and we have

attempted here to characterize within a well-known system of
cultured RAW 264.7 macrophages how O2 affects both the
antioxidant and inammatory responses. Crucial to this
approach was the adoption in our study of a dedicated
chamber, where cells are not temporarily exposed to
physiological O2 (5%) for short period of times as previously
described (Grodzki et al., 2013), but permanently grown over
several passages and then treated for the experiments at 5% O2
without any further exposure to atmospheric O2 levels.
Although at rst we could not predict how well cells would
survive in these conditions, we were somewhat surprised to
nd that macrophages happily grow for several weeks at 5% O2
and that this behavior is not restricted to this cell type but
applies also to cardiomyocytes (H9c2), human stem and
endothelial cells and primary murine astrocytes (unpublished
data). In this article we report that the pro-oxidant status and
the antioxidant response induced by the Nrf2 activator DMF in
RAW macrophages permanently cultured at 5% O2 are
considerably diminished compared to cells growing at 20% O2.
In addition, the inammatory response stimulated by LPS was
differentially affected by O2 concentrations, with a higher
production of inammatory mediators at 20% O2 that was
markedly reduced by DMF. These data indicate that cells
cultured at 20% O2 are strongly affected by an underlying prooxidant phenotype, which inuences inammation and the
cellular adaptation to DMF, an immunomodulatory drug
(Zhang, 2013).
Although we showed that RAW 264.7 macrophages grown
at 5% O2 contained less O2 in the culture medium, they
displayed a similar O2 consumption and a slightly lower ATP
production compared to cells cultured at 20% O2 suggesting
that 5% O2 does not severely limit energy production. We also
JOURNAL OF CELLULAR PHYSIOLOGY

excluded the possibility that such energy production was

dependent on glycolysis, since ATP levels at 5% O2 were
comparable between cells cultured in the presence or absence
of 2-deoxy-D-glucose to inhibit glycolysis. We therefore
concluded that macrophages at 5% O2 grew and replicated by
relying on mitochondrial respiration as the main source of
energy, which seems to be efcient and fully functional at O2
levels well below the ones found in standard cell culture
incubators. However, the higher concentrations of O2 in
normal incubators affected the redox status of the cells, as
macrophages grown at 20% O2 consistently displayed: (1)
increased ROS production and GSSG, with lower GSH
content; and (2) augmented ROS generation upon challenge
with menadione, a compound that stimulates superoxide
formation based on cellular processing by NADPH oxidases
and mitochondrial complex I (Thor et al., 1982; Criddle et al.,
2006) compared to cells at 5% O2. These data are in line with
ndings by Atkuri and co-workers, who demonstrated that the
GSH content in T lymphocytes cultured for 3 days at 5% O2
was higher than at 20% O2, and, importantly, was equivalent to
that of freshly isolated T lymphocytes (Atkuri et al., 2007).
Thus, it appears that cells grown under O2 concentrations that
exceed the basic requirements for energy production will
redirect the excess of O2 towards formation of ROS, which will
then inuence a series of endogenous processes and create an
imbalance in the cellular redox status. It is important to note
that, unlike the cells in Atkuris experiments, macrophages in
our study were continuously maintained at 5% O2 for a period
of 10 days before starting the experiments. In addition, we
extended the initial interesting nding of Atkuri and coworkers by extensively characterizing the metabolic prole
and closely examining the inammatory and anti-oxidant
proles of cells permanently grown at 5% and 20% O2.

1135

1136

HAAS ET AL.

Fig. 7. Differential induction of metalloproteinases and pro-angiogenic factor VEGF-A in response to LPS in RAW 264.7 macrophages
cultured at 20% and 5% O2. RAW 264.7 macrophages were permanently grown at 20% O2 or 5% O2 and then challenged with 10 ng/ml LPS in
the presence or absence of DMF (25 mM). After 6 h incubation, total RNA was extracted and gene expression was measured by real-time
quantitative polymerase chain reaction as reported in Materials and Methods. MMP-2, VEGF-A and MMP-9 mRNA level were analyzed (A, B
and C respectively). Hprt1 was used as a housekeeping gene. MMP-9 secretion and activity were measured by zymography on conditioned cell
culture medium after 24 h of incubation with 10 ng/ml LPS in presence or absence of 1050 mM DMF (D). Data are expressed as mean  S.E.M.
*P < 0.05 vs. control; yP < 0.05 vs. LPS alone; #P < 0.05 vs. 20% O2. MMP-9 zymogram is representative of three independent experiments.

One interesting process that we investigated was the cellular

response elicited by DMF, a positive regulator of the
expression of antioxidant genes through its action on the Nrf2
pathway (Linker et al., 2011; Takaya et al., 2012; Foresti et al.,
2013). Our results reveal that the antioxidant mediator HO-1
was induced by DMF in RAW 264.7 and primary murine
macrophages but this effect was signicantly lower at 5% O2.
Enhanced oxidative stress at 20% O2 contributed to higher
HO-1 expression, as co-incubation of DMF with NAC
signicantly lowered HO-1 activity and protein at 20% but not
at 5% O2. Stimulation of GSH production and expression of
NQO-1 by DMF exhibited similar trends, indicating a general
mechanism of antioxidant responses that are more pronounced at 20% O2. Although the basal level of SOD2 mRNA
was higher at 20% O2, no changes were observed in response
to DMF at both O2 tensions, suggesting that expression of
SOD2 is not dependent on Nrf2 activation by DMF.
Interestingly, and in line with the HO-1 expression data, Nrf2
activation was higher in cells at 20% O2.
We also observed that O2 tensions caused notable
differential changes in the prole of inammatory markers
elicited by LPS in RAW and primary macrophages. In essence,
all the pro-inammatory molecules measured, together with
the anti-inammatory IL-10, were signicantly more expressed
at 20% O2, supporting the hypothesis that the basal pro-oxidant
state of macrophages cultured at 20% O2 favors or contributes
to the activation of the inammatory response. As observed for
some markers in previous studies and as predicted with the use
of a computational approach in this study (Lehmann et al., 2007;
Yamazoe et al., 2009; Wilms et al., 2010; Lin et al., 2011), DMF
markedly decreased the synthesis of inammatory mediators at
20% O2; the effect of DMF was less appreciable at 5% O2 since
JOURNAL OF CELLULAR PHYSIOLOGY

the synthesis of these markers at this O2 concentration was

only mildly affected by LPS.
Several sets of data in relation to inammation merit
attention. First, although nitrite production at 5% O2 was half of
that detected at 20% O2, iNOS protein expression was slightly
higher at 5% O2 but not signicantly different from 20%. There
is a possibility that at low oxygen tension, iNOS protein
stability is enhanced and/or the protein is less prone to
degradation. Moreover, it has been reported that nitric oxide
synthesis is highly dependent on O2 tension in LPS-challenged
RAW 264.7 macrophages (Robinson et al., 2011) and this could
partially explain why we observed lower nitrite generation at
5% O2. However, a second explanation is also likely based on
the additional results of ARG1. This enzyme uses arginine and
water to produce urea and ornitine, effectively competing with
NO synthase, which instead converts this aminoacid to NO
and citrulline (Bansal and Ochoa, 2003). Our data suggest that
this competition probably occurred to a signicant extent at 5%
O2, limiting arginine for NO production since ARG1
expression at this O2 tension was several folds greater than
that stimulated by LPS at 20% O2. Of note, ARG1 is a M2-like
marker involved in the pro-healing phase of inammation (Mills
et al., 2000; Campbell et al., 2013) and its lower expression at
20% O2, combined with marked synthesis of pro-inammatory
molecules, suggests that the inammatory response at 20% O2
is skewed towards pro-inammatory activities of macrophages.
Third, the signicant down-regulation of MMP-9 at 5%
compared to 20% O2 after LPS challenge could again be
explained by lower production of ROS in macrophages
cultured at 5% O2, as MMP-9 has been shown to be upregulated in inammatory cells not only through the canonical
NFkB pathway, but also through the increased ROS that trigger

RESPONSES OF MACROPHAGES AT PHYSIOLOGICAL OXYGEN

Fig. 8. HO-1 induction and TNF-a production in primary murine

peritoneal macrophages cultured at 20% and 5% O2. (A) HO-1
protein expression was measured in primary macrophages growing
at 20% or 5% O2. Peritoneal macrophages were collected from
C57Bl/6J mice after treatment with thioglycolate solution as
reported in Materials and Methods. Cells were seeded immediately
at 20% or 5% O2 overnight and then treated with 25 mM DMF for 6 h
before analysis. HO-1 immunoblottings are representative of three
independent experiments. (B) TNF-a production was assessed in
the supernatants of peritoneal macrophages cultured as above.
Cells grown at 20% or 5% O2 were treated with 10 ng/ml LPS
in the presence or absence of 25 mM DMF for 24 h. Data are
expressed as mean  S.E.M. Data are expressed as mean  S.E.M.
*P < 0.05 vs. control; yP < 0.05 vs. LPS alone; #P < 0.05 vs. 20% O2.

p38/AP1 expression (Woo et al., 2004). In addition, we showed

that DMF is capable of suppressing the increase in MMP-9
expression and activity at 20% O2, as previously reported for
other Nrf2 activators such as curcumin, CAPE and carnosol
(Huang et al., 2005; Jin et al., 2005; Chae et al., 2012; Kohli et al.,
2013). Finally, the expression of the pro-angiogenic factor
VEGF and its regulation by LPS in different O2 tension is of
particular interest. In our model, VEGF mRNA expression was
equivalent in cells cultured at 20% and 5% O2 and this result,
together with the similar O2 consumption rate and ATP
content, further indicates that RAW macrophages cultured at
5% O2 are not characterized by a hypoxic phenotype. Indeed,
VEGF has been reported to be strongly increased by hypoxia in
different cell types, including macrophages (Xiong et al., 1998).
In contrast, VEGF mRNA expression was highly up-regulated in
LPS-activated macrophages only at 5% O2 and treatment with
DMF signicantly amplied this response. As this effect was
associated with a marked suppression of MMP-9 expression
and increased HO-1, our results suggest the establishment of a
pro-angiogenic phenotype where inammatory cell migration
is down-regulated to promote an angiogenic and antiinammatory response at low O2 tension. There are also
reports showing a role of HO-1 in regulating VEGF synthesis
and an anti-inammatory VEGF-driven angiogenesis that downregulates leukocyte inltration in vivo (Bussolati et al., 2003;
Dulak et al., 2004) but we did not investigate this link in our
model. In addition, recent ndings by Wu and collaborators
described a role of IL-10 in dampening VEGF synthesis only in
JOURNAL OF CELLULAR PHYSIOLOGY

M1-macrophages (pro-inammatory) but not in M2-macrophages possessing anti-inammatory phenotypes (Wu et al.,
2010). Whether this switch in phenotype is affected by O2
tension requires further investigations. At the beginning of this
project we measured HIF-1 alpha in cells but could not detect
any expression at the two O2 tensions. We attributed this
result to the fact that our experiments are conducted when
cells have already adapted to the 5% O2. It is possible that
during the adaptation period HIF pathways may uctuate in
response to O2, together with many other metabolic changes
that could be investigated in the future.
In conclusion, our study demonstrates the important role of
O2 tension on the cellular functions of macrophages, highlighting essential differences in the redox status, inammation
and DMF-dependent responses between 5 and 20% O2 levels.
This study represents a step forward in the understanding of
the redox biology of macrophages and how to set up
experimental conditions in vitro that mimic as closely as
possible the physiological O2 tension found in vivo, as originally
promoted by the group of Herzenberg and colleagues (Atkuri
et al., 2005, 2007). To this end, the differential response to the
immunomodulatory drug DMF found in cells grown at the two
O2 conditions also warrants the importance of screening
pharmacologically active compounds for therapeutic purposes
at appropriate concentrations of O2. Although a large scale
modication of standard culture approaches will be a challenge
for most laboratories worldwide, we believe that efforts to
reproduce the physiological cell environment will lead to
signicantly improved research outcome with less artifactual
results.
Acknowledgements

This work was supported by AREMCAR foundation (RM and

RF) and a Blanc II International Grant MITO-CO from the
Agence National de la Recherche (RM). BH is recipient of a
postdoctoral grant from Fonds National de la Recherche
Luxembourg (AFR project ID 3979613). SC and SK are
recipients of doctoral fellowships from Coeur & Poumon
Foundation and ClinServ International Lebanon, respectively.
The authors thank the imaging platform of INSERM U955 for
the use of their imaging station.
Literature Cited
Abuarqoub H, Foresti R, Green CJ, Motterlini R. 2006. Heme oxygenase-1 mediates the antiinammatory actions of 20 -hydroxychalcone in RAW 264.7 murine macrophages. Am J
Physiol Cell Physiol 290:C1092C1099.
Alam J, Stewart D, Touchard C, Boinapally S, Choi AM, Cook JL. 1999. Nrf2, a CapnCollar
transcription factor, regulates induction of the heme oxygenase-1 gene. J Biol Chem
274:2607126078.
Albrecht P, Bouchachia I, Goebels N, Henke N, Hofstetter HH, Issberner A, Kovacs Z,
Lewerenz J, Lisak D, Maher P, Mausberg AK, Quasthoff K, Zimmermann C, Hartung HP,
Methner A. 2012. Effects of dimethyl fumarate on neuroprotection and immunomodulation. J Neuroinammation 9:163.
Ashraan H, Czibik G, Bellahcene M, Aksentijevic D, Smith AC, Mitchell SJ, Dodd MS,
Kirwan J, Byrne JJ, Ludwig C, Isackson H, Yavari A, Stottrup NB, Contractor H, Cahill TJ,
Sahgal N, Ball DR, Birkler RI, Hargreaves I, Tennant DA, Land J, Lygate CA, Johannsen M,
Kharbanda RK, Neubauer S, Redwood C, de Cabo R, Ahmet I, Talan M, Gunther UL,
Robinson AJ, Viant MR, Pollard PJ, Tyler DJ, Watkins H. 2012. Fumarate is
cardioprotective via activation of the Nrf2 antioxidant pathway. Cell Metab 15:361371.
Atkuri KR, Herzenberg LA, Herzenberg LA. 2005. Culturing at atmospheric oxygen levels
impacts lymphocyte function. Proc Natl Acad Sci USA 102:37563759.
Atkuri KR, Herzenberg LA, Niemi AK, Cowan T, Herzenberg LA. 2007. Importance of
culturing primary lymphocytes at physiological oxygen levels. Proc Natl Acad Sci USA
104:45474552.
Azuaje F. 2013. Drug interaction networks: An introduction to translational and clinical
applications. Cardiovasc Res 97:631641.
Balogun E, Hoque M, Gong P, Killeen E, Green CJ, Foresti R, Alam J, Motterlini R. 2003.
Curcumin activates the heme oxygenase-1 gene via regulation of Nrf2 and the antioxidant
responsive element. Biochem J 371:887895.
Bambrick LL, Kostov Y, Rao G. 2011. In vitro cell culture pO(2) is signicantly different from
incubator pO(2). Biotechnol Prog 27:11851189.
Bansal V, Ochoa JB. 2003. Arginine availability, arginase, and the immune response. Curr
Opin Clin Nutr Metab Care 6:223228.
Brune B, Dehne N, Grossmann N, Jung M, Namgaladze D, Schmid T, von Knethen A,
Weigert A. 2013. Redox control of inammation in macrophages. Antioxid Redox Signal
19:595637.

1137

1138

HAAS ET AL.

Bussolati B, Ahmed A, Pemberton H, Landis RC, Di Carlo F, Haskard DO, Mason JC. 2003.
Bifunctional role for VEGF-induced heme oxygenase-1 in vivo: induction of angiogenesis
and inhibition of leukocytic inltration. Blood 103:761766.
Campbell L, Saville CR, Murray PJ, Cruickshank SM, Hardman MJ. 2013. Local arginase 1
activity is required for cutaneous wound healing. J Invest Dermatol 133:24612470.
Carreau A, Hafny-Rahbi B, Matejuk A, Grillon C, Kieda C. 2011. Why is the partial oxygen
pressure of human tissues a crucial parameter. Small molecules and hypoxia 15:1239
1253.
Chae IG, Yu MH, Im NK, Jung YT, Lee J, Chun KS, Lee IS. 2012. Effect of Rosemarinus
ofcinalis L. on MMP-9, MCP-1 levels, and cell migration in RAW 264.7 and smooth muscle
cells. J Med Food 15:879886.
Criddle DN, Gillies S, Baumgartner-Wilson HK, Jaffar M, Chinje EC, Passmore S, Chvanov M,
Barrow S, Gerasimenko OV, Tepikin AV, Sutton R, Petersen OH. 2006. Menadioneinduced reactive oxygen species generation via redox cycling promotes apoptosis of
murine pancreatic acinar cells. J Biol Chem 281:4048540492.
Davies KJ. 2000. Oxidative stress, antioxidant defenses, and damage removal, repair, and
replacement systems. IUBMB Life 50:279289.
Degn H, Lilleor M, Iversen JJ. 1973. The occurrence of a stepwise-decreasing respiration rate
during oxidative assimilation of different substrates by resting Klebsiella aerogenes in a
system open to oxygen. Biochem J 136:10971104.
Dinkova-Kostova AT, Holtzclaw WD, Cole RN, Itoh K, Wakabayashi N, Katoh Y,
Yamamoto M, Talalay P. 2002. Direct evidence that sulfhydryl groups of Keap1 are the
sensors regulating induction of phase 2 enzymes that protect against carcinogens and
oxidants. Proc Natl Acad Sci USA 99:1190891311.
Dinkova-Kostova AT, Talalay P. 2008. Direct and indirect antioxidant properties of inducers
of cytoprotective proteins. Mol Nutr Food Res 52:S128S138.
Duguez S, Duddy WJ, Gnocchi V, Bowe J, Dadgar S, Partridge TA. 2012. Atmospheric oxygen
tension slows myoblast proliferation via mitochondrial activation. PLoS ONE 7:e43853.
Dulak J, Zagorska A, Jozkowicz A. 2004. Complex role of heme oxygenase-1 in angiogenesis.
Antioxid Redox Signal 6:858866.
Finkel T, Holbrook NJ. 2000. Oxidants, oxidative stress and the biology of ageing. Nature
408:239247.
Foresti R, Bains SK, Pitchumony TS, de Castro Bras, Drago LE, Dubois-Rande F, Bucolo JL,
Motterlini C. 2013. Small molecule activators of the Nrf2-HO-1 antioxidant axis modulate
heme metabolism and inammation in BV2 microglia cells. Pharmacol Res 76:132148.
Freemerman AJ, Johnson AR, Sacks GN, Milner JJ, Kirk EL, Troester MA, Macintyre AN,
Goraksha-Hicks P, Rathmell JC, Makowski L. 2014. Metabolic reprogramming of
macrophages: glucose transporter (GLUT1)-mediated glucose metabolism drives a proinammatory phenotype. J Biol Chem 289:78847896.
Ghoreschi K, Bruck J, Kellerer C, Deng C, Peng H, Rothfuss O, Hussain RZ, Gocke AR,
Respa A, Glocova I, Valtcheva N, Alexander E, Feil S, Feil R, Schulze-Osthoff K, Rupec RA,
Lovett-Racke AE, Dringen R, Racke MK, Rocken M. 2011. Fumarates improve psoriasis
and multiple sclerosis by inducing type II dendritic cells. J Exp Med 208:22912303.
Grodzki AC, Giulivi C, Lein PJ. 2013. Oxygen tension modulates differentiation and primary
macrophage functions in the human monocytic THP-1 cell line. PLoS ONE 8:e54926.
Halliwell B. 2003. Oxidative stress in cell culture: an under-appreciated problem? FEBS Lett
540:36.
Hu R, Saw CL, Yu R, Kong AN. 2010. Regulation of NF-E2-Related Factor 2 signaling for
cancer chemoprevention: Antioxidant coupled with antiinammatory. Antioxid Redox
Signal 13:16791698.
Huang SC, Ho CT, Lin-Shiau SY, Lin JK. 2005. Carnosol inhibits the invasion of B16/F10
mouse melanoma cells by suppressing metalloproteinase-9 through down-regulating
nuclear factor-kappa B and c-Jun. Biochem Pharmacol 69:221232.
Ishii T, Itoh K, Takahashi S, Sato H, Yanagawa T, Katoh Y, Bannai S, Yamamoto M. 2000.
Transcription factor Nrf2 coordinately regulates a group of oxidative stress-inducible
genes in macrophages. J Biol Chem 275:1602316029.
Jin UH, Chung TW, Kang SK, Suh SJ, Kim JK, Chung KH, Gu YH, Suzuki I, Kim CH. 2005.
Caffeic acid phenyl ester in propolis is a strong inhibitor of matrix metalloproteinase-9 and
invasion inhibitor: Isolation and identication. Clin Chim Acta 362:5764.
Kohli S, Chhabra A, Jaiswal A, Rustagi Y, Sharma M, Rani V. 2013. Curcumin suppresses
gelatinase B mediated norepinephrine induced stress in H9c2 cardiomyocytes. PLoS ONE
8:e76519.
Lehmann JC, Listopad JJ, Rentzsch CU, Igney FH, von Bonin A, Hennekes HH, Asadullah K,
Docke WD. 2007. Dimethylfumarate induces immunosuppression via glutathione
depletion and subsequent induction of heme oxygenase 1. J Invest Dermatol 127:835845.
Lin SX, Lisi L, Dello RC, Polak PE, Sharp A, Weinberg G, Kalinin S, Feinstein DL. 2011. The
anti-inammatory effects of dimethyl fumarate in astrocytes involve glutathione and haem
oxygenase-1. ASN Neuro 3:e00055.
Linker RA, Lee DH, Ryan S, van Dam AM, Conrad R, Bista P, Zeng W, Hronowsky X, Buko A,
Chollate S, Ellrichmann G, Bruck W, Dawson K, Goelz S, Wiese S, Scannevin RH,
Lukashev M, Gold R. 2011. Fumaric acid esters exert neuroprotective effects in
neuroinammation via activation of the Nrf2 antioxidant pathway. Brain 134:678692.
Meissner M, Valesky EM, Kippenberger S, Kaufmann R. 2012. Dimethyl fumarate - only an
anti-psoriatic medication. J Dtsch Dermatol Ges 10:793801.
Mills CD, Kincaid K, Alt JM, Heilman MJ, Hill AM. 2000. M-1/M-2 macrophages and the Th1/
Th2 paradigm. J Immunol 164:61666173.

JOURNAL OF CELLULAR PHYSIOLOGY

Motterlini R, Foresti R. 2014. Heme oxygenase-1 as a target for drug discovery. Antioxid
Redox Signal 20:18101826.
Motterlini R, Foresti R, Bassi R, Calabrese V, Clark JE, Green CJ. 2000. Endothelial heme
oxygenase-1 induction by hypoxia: modulation by inducible nitric oxide synthase (iNOS)
and S-nitrosothiols. J Biol Chem 275:1361313620.
Motterlini R, Sawle P, Hammad J, Mann BE, Johnson TR, Green CJ, Foresti R. 2013.
Vasorelaxing effects and inhibition of nitric oxide in macrophages by new iron-containing
carbon monoxide-releasing molecules (CO-RMs). Pharmacol Res 68:108117.
Parrinello S, Samper E, Krtolica A, Goldstein J, Melov S, Campisi J. 2003. Oxygen sensitivity
severely limits the replicative lifespan of murine broblasts. Nat Cell Biol 5:741747.
Prestera T, Talalay P. 1995. Electrophile and antioxidant regulation of enzymes that detoxify
carcinogens. Proc Natl Acad Sci USA 92:89658969.
Rahman I, Kode A, Biswas SK. 2006. Assay for quantitative determination of glutathione and
glutathione disulde levels using enzymatic recycling method. Nat Protoc 1:31593165.
Ramos-Gomez M, Kwak MK, Dolan PM, Itoh K, Yamamoto M, Talalay P, Kensler TW. 2001.
Sensitivity to carcinogenesis is increased and chemoprotective efcacy of enzyme
inducers is lost in nrf2 transcription factor-decient mice. Proc Natl Acad Sci USA
98:34103415.
Robinson MA, Baumgardner JE, Otto CM. 2011. Oxygen-dependent regulation of nitric
oxide production by inducible nitric oxide synthase. Free Radic Biol Med 51:19521965.
Satoh T, McKercher SR, Lipton SA. 2013. Nrf2/ARE-mediated antioxidant actions of proelectrophilic drugs. Free Radic Biol Med 65:645657.
Satoh T, Okamoto SI, Cui J, Watanabe Y, Furuta K, Suzuki M, Tohyama K, Lipton SA. 2006.
Activation of the Keap1/Nrf2 pathway for neuroprotection by electrophilic phase II
inducers. Proc Natl Acad Sci USA 103:768773.
Sawle P, Foresti R, Mann BE, Johnson TR, Green CJ, Motterlini R. 2005. Carbon monoxidereleasing molecules (CO-RMs) attenuate the inammatory response elicited by
lipopolysaccharide in RAW264.7 murine macrophages. Br J Pharmacol 145:800810.
Sawle P, Moulton BE, Jarzykowska M, Green CJ, Foresti R, Fairlamb IJ, Motterlini R. 2008.
Structure-activity relationships of Methoxychalcones as inducers of Heme Oxygenase-1.
Chem Res Toxicol 21:14841494.
Takaya K, Suzuki T, Motohashi H, Onodera K, Satomi S, Kensler TW, Yamamoto M. 2012.
Validation of the multiple sensor mechanism of the Keap1-Nrf2 system. Free Radic Biol
Med 53:817827.
Thor H, Smith MT, Hartzell P, Bellomo G, Jewell SA, Orrenius S. 1982. The metabolism of
menadione (2-methyl-1,4-naphthoquinone) by isolated hepatocytes. A study of the
implications of oxidative stress in intact cells. J Biol Chem 257:1241912425.
Trachootham D, Lu W, Ogasawara MA, Nilsa RD, Huang P. 2008. Redox regulation of cell
survival. Antioxid Redox Signal 10:13431374.
Tsai AG, Johnson PC, Intaglietta M. 2003. Oxygen gradients in the microcirculation. Physiol
Rev 83:933963.
Vesely MJJ, Exon DJ, Clark JE, Foresti R, Green CJ, Motterlini R. 1998. Heme oxygenase-1
induction in skeletal muscle cells: hemin and sodium nitroprusside are regulators in vitro.
Am J Physiol 275:C1087C1094.
Wilms H, Sievers J, Rickert U, Rostami-Yazdi M, Mrowietz U, Lucius R. 2010.
Dimethylfumarate inhibits microglial and astrocytic inammation by suppressing the
synthesis of nitric oxide, IL-1beta, TNF-alpha and IL-6 in an in-vitro model of brain
inammation. J Neuroinammation 7:30.
Woo CH, Lim JH, Kim JH. 2004. Lipopolysaccharide induces matrix metalloproteinase-9
expression via a mitochondrial reactive oxygen species-p38 kinase-activator protein-1
pathway in Raw 264.7cells. J Immunol 173:69736980.
Wood IS, Stezhka T, Trayhurn P. 2011. Modulation of adipokine production, glucose uptake
and lactate release in human adipocytes by small changes in oxygen tension. Pugers Arch
462:469477.
Wu WK, Llewellyn OP, Bates DO, Nicholson LB, Dick AD. 2010. IL-10 regulation of
macrophage VEGF production is dependent on macrophage polarisation and hypoxia.
Immunobiology 215:796803.
Xiong M, Elson G, Legarda D, Leibovich SJ. 1998. Production of vascular endothelial growth
factor by murine macrophages: regulation by hypoxia, lactate, and the inducible nitric
oxide synthase pathway. Am J Pathol 153:587598.
Yamazoe Y, Tsubaki M, Matsuoka H, Satou T, Itoh T, Kusunoki T, Kidera Y, Tanimori Y,
Shoji K, Nakamura H, Ogaki M, Nishiura S, Nishida S. 2009. Dimethylfumarate inhibits
tumor cell invasion and metastasis by suppressing the expression and activities of matrix
metalloproteinases in melanoma cells. Cell Biol Int 33:10871094.
Yates MS, Kensler TW. 2007. Chemopreventive promise of targeting the Nrf2 pathway.
Drug News Perspect 20:109117.
Zhang DD. 2013. Bardoxolone Brings Nrf2-Based Therapies to Light. Antioxid Redox Signal
19:517518.

Supporting Information

Additional supporting information may be found in the online

version of this article at the publishers web-site.