Original Contribution

Cell-type-specific downregulation of heme oxygenase-1by lipopolysaccharide via Bach1 in primary human mononuclear cells

Mirrin J. Dorresteijn a,b,c,1, Ananta Paine d,1, Eva Zilian d, Maaike G.E. Fenten a,b,c,Eileen Frenzel e, Sabina Janciauskiene e, Constanca Figueiredo d, Britta Eiz-Vesper d,Rainer Blasczyk d, Douwe Dekker b,c, Bas Pennings b,c,f, Alwin Scharstuhl b,c, Paul Smits b,c,Jan Larmann g, Gregor Theilmeier g, Johannes G. van der Hoeven a,h,Frank A.D.T.G. Wagener c,f, Peter Pickkers a,h,2, Stephan Immenschuh d,n,2Q1

a Department of Intensive Care Medicineb Department of Pharmacology and Toxicologyc Radboud Institute for Molecular Life Sciencesd Institute for Transfusion Medicinee Department of Internal Medicine–Respiratory Medicinef Department of Orthodontics and Craniofacial Biologyg Department of Anesthesiology and Intensive Care Medicine, Hannover Medical School, 30625 Hannover, Germanyh Nijmegen Center for Infectious Diseases, Radboud University Medical Center, 6500 HB Nijmegen, The Netherlands

a r t i c l e i n f o

Article history:Received 20 June 2014Received in revised form24 October 2014Accepted 29 October 2014

Keywords:Bach1Heme oxygenase-1InflammationLipopolysaccharideMononuclear cellsToll-like receptorFree radicals

a b s t r a c t

Heme oxygenase (HO)-1 is the inducible isoform of heme degradation, which is upregulated by multiplestress stimuli. HO-1 has major immunomodulatory and anti-inflammatory effects via its cell-type-specific functions in mononuclear cells. Contradictory findings have been reported on HO-1 generegulation by the Toll-like receptor (TLR) 4 ligand lipopolysaccharide (LPS) in these cells. Therefore, wereinvestigated the effects of LPS on HO-1 gene expression in human and murine mononuclear cellsin vitro and in vivo. Remarkably, LPS downregulated HO-1 mRNA levels in primary human peripheralblood mononuclear cells (PBMCs), CD14þ monocytes, macrophages, dendritic cells, and granulocytes,but upregulated HO-1 mRNA levels in murine macrophages and human monocytic leukemia cell lines.Furthermore, experiments with human CD14þ monocytes revealed that activation of other TLRsincluding TLR1, -2, -5, -6, -8, and -9 decreased HO-1 mRNA expression. LPS-dependent downregulationof HO-1 was specific, because expression of cyclooxygenase-2, NADP(H)-quinone oxidoreductase-1, andperoxiredoxin-1 was increased under the same experimental conditions. Notably, LPS upregulatedexpression of Bach1, a critical transcriptional repressor of HO-1. Moreover, knockdown of this nuclearfactor enhanced basal and LPS-dependent HO-1 expression in mononuclear cells. Finally, downregula-tion of HO-1 in response to LPS was confirmed in PBMCs from human individuals subjected toexperimental endotoxemia. In summary, LPS downregulates HO-1 expression in primary humanmononuclear cells via a Bach1-mediated pathway. Because LPS-dependent HO-1 regulation is cell-type-and species-specific, experimental findings in cell lines and animal models need careful interpretation.

& 2014 Elsevier Inc. All rights reserved.

Introduction

The heme oxygenase (HO)3 system has powerful antioxidant andanti-inflammatory effects [1–4]. HO is the rate-limiting enzyme ofheme degradation and produces equimolar amounts of carbon mon-oxide (CO), iron, and biliverdin [5]. HO-1, the inducible isoform of thisenzyme, is upregulated by multiple pro-oxidant and proinflammatorystimuli via the interplay of various transcription factors and signalingcascades [6,7]. By contrast, the constitutively expressed isoform, HO-2,

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

67686970717273747576777879808182

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/freeradbiomed

Free Radical Biology and Medicine

http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.5790891-5849/& 2014 Elsevier Inc. All rights reserved.

Abbreviations: AD, actinomycin D; CHX, cycloheximide; COX, cyclooxygenase; DC,dendritic cell; HO, heme oxygenase; LPS, lipopolysaccharide; MM-6, MonoMac-6;NOS, nitric oxide synthase; NQO1, NADP(H)-quinone oxidoreductase-1; PBMC,peripheral blood mononuclear cell; PBS, phosphate-buffered saline; Prx, peroxir-edoxin; RT-PCR, reverse transcription polymerase chain reaction; TLR, Toll-likereceptor

n Corresponding author. Fax: þ49 511 532 2079.E-mail address: Immenschuh.Stephan@mh-hannover.de (S. Immenschuh).1 These authors contributed equally to this work.2 These authors contributed equally to this work.

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎

is not regulated by such stimuli [3,8]. Anti-inflammatory and immu-nomodulatory effects of HO-1 have been shown in murine and humangenetic HO-1 deficiency [9,10]. Moreover, targeted induction of HO-1or administration of its effector molecules has been associated withbeneficial effects in animal models of sepsis and organ transplantation[11–15]. In particular, the immunomodulatory effects of HO-1 havebeen linked with its cell-type-specific functions in mononuclear cells[16,17].

Lipopolysaccharide (LPS) is a cell wall component of gram-negative bacteria and a ligand for the cell surface receptor Toll-likereceptor (TLR) 4, which is a classical activator of the innateimmune response [18]. Moreover, LPS is known to be a centralmediator of sepsis and septic shock and is frequently used as animmunostimulator in in vitro and in vivo models [19]. Activationof mononuclear cells with LPS leads to increased production ofreactive oxygen species, nitric oxide (NO), and proinflammatorycytokines [20]. In animal models of sepsis, increased HO-1 geneexpression has been found in macrophages of various organs[3,21,22]. Moreover, independent groups have demonstrated thatLPS markedly upregulated HO-1 in various models of human androdent monocytes or leukemic cell lines in vitro [23–26]. Conflict-ingly, HO-1 has also been shown to be markedly downregulated byLPS in primary human monocytes [27] and during LPS-dependentmaturation of dendritic cells (DCs) [28]. To reconcile these contra-dictory findings on HO-1 expression, we reinvestigated the reg-ulation of HO-1 by LPS in different types of mononuclear cellsin vitro and in vivo. Moreover, we addressed the role of the nuclearrepressor Bach1 [29,30], which has recently been shown to becrucial for the immunological functions of HO-1 in mononuclearcells [31].

Materials and methods

All experiments were performed after approval of the localethics committee and are in agreement with the Declaration ofHelsinki and Good Clinical Practice guidelines.

Cell preparation and culture

Mouse macrophages were prepared as described previously [32].Human PBMCs were isolated from buffy coats of healthy blood donorsby discontinuous Ficoll–Paque density gradient centrifugation andresuspended at a concentration of 1�106 cells/ml in standard M'medium (Lonza, Verviers, Belgium) supplemented with 10% humanAB serum (C.C.pro, Neustadt/W., Germany), as described before [33].Primary monocytes were isolated from the PBMCs using the MACSmagnetic separation system with a Human Monocyte Isolation Kit II(negative isolation; Miltenyi Biotec, Bergisch Gladbach, Germany).CD14þ cells were >95% pure as determined by flow cytometry.CD14þ cells were cultured in six-well flat-bottom plates in 1 ml M'medium. For the generation of human macrophages, isolated CD14þ

cells were treated with 800 U/ml M-CSF (PeproTech, Hamburg,Germany) for 4 days. The generation of immature human DCs wasperformed as described previously [33]. Human neutrophils wereisolated from freshly obtained peripheral blood of healthy volunteersas previously described [34]. In brief, purified neutrophils werewashed in PBS and then resuspended in RPMI 1640. Neutrophilpurity was typically 90% as determined by cytospin, and cell viabilitywas more than 95% according to 0.4% trypan blue staining. Monocytichuman leukemia cell lines, THP-1 and MonoMac-6 (MM-6) (ATCC,Manassas, VA, USA), were also maintained in standard M' medium.Cell cultures were kept under air/CO2 (19/1) at 100% humidity. Afterovernight culture, the mediumwas replaced with medium containingTLR ligands and incubated for 18 h. The transcriptional inhibitoractinomycin D (AD; 5 μg/ml) and protein synthesis inhibitor

cycloheximide (CHX; 10 μg/ml) were purchased from Sigma–Aldrich(St. Louis, MO, USA). Cell cultures were pretreated with these twocompounds for 30 min before treatment with LPS. The various TLRagonists (Invivogen, Toulouse, France) were used at the followingconcentrations: 1 mg/ml TLR1/2 agonist Pam3CSK4, 108 freeze-driedcells of heat-killed Listeria monocytogenes (HKLM)/ml for TLR2 activa-tion, 5 mg/ml TLR2 agonist lipoteichoic acid, 10 mg/ml TLR3 agonistpoly(I:C), 1 mg/ml TLR4 agonist LPS (Escherichia coli 0111:B4; MerckBiosciences, La Jolla, CA, USA), 1 mg/ml TLR5 agonist flagellin, 1 mg/mlTLR2/6 agonist fibroblast-stimulating ligand 1, 1 mg/ml TLR7 agonistimiquimod, 5 mg/ml TLR8 ligand ssRNA40, 10 mg/ml TLR9 ligandoligodeoxynucleotide 2006. Heme (Frontier Scientific, Logan, UT,USA) was applied at a final concentration of 10 mmol/L. All otherchemicals were purchased from Sigma–Aldrich and Roche AppliedScience (Mannheim, Germany), unless otherwise indicated.

Transfection of CD14þ cells with Bach1 small interfering RNA (siRNA)

CD14þ cells (1�106 cells/well) were transfected using theHuman Monocyte Nucleofector Kit C (Amaxa Biosystems, Gaithers-burg, MD, USA) according to the manufacturer's protocol. Equivalentmolar concentrations of human Bach1 siRNA and nontargetingnegative control siRNA (Thermo Fisher Scientific, Waltham, MA,USA) were used to yield a final concentration of 50 nmol/L. Trans-fected cells were incubated for 48 h before treatment. To evaluatethe efficiency of Bach1 knockdown, cells were lysed and subjected toreal-time reverse transcription (RT)-PCR and Western blot analysis.

Experimental human endotoxemia

All healthy volunteers were participants of a pharmacologicalintervention study (NCT00785018) [35,36], which was approvedby the local ethics committee (CMO Arnhem-Nijmwegen) underNo. 2008-197. The study was also formally announced at thecentral Dutch medical ethics commission and received approval(NL 24017.091.08). The experiments were conducted after receiv-ing written informed consent and according to the rules of GoodClinical Practice. Ten nonsmoking, healthy male volunteers wereincluded, who were not taking any medication and did not have arelevant medical history. No abnormalities were present duringphysical examination, laboratory tests, and electrocardiography.Volunteers tested negative for HIV and hepatitis B and did notsuffer from a febrile illness in the 14 days before the study.Subjects refrained from food 10 h before the start of the experi-ment. All experiments were performed at the Radboud UniversityMedical Center intensive care research unit. During experiments,heart rate (electrocardiogram) and blood pressure (intra-arterially)were monitored continuously. Before LPS infusion, prehydrationwith 1500 ml of 2.5%/0.45% glucose/NaCl was administered intra-venously. Thereafter, 150 ml/h was infused to maintain an optimalhydration status. At t¼0 h all subjects received a dose of 2 ng/kgLPS (purified LPS dissolved in 0.9% NaCl) prepared from E. coliO:113 (Lot Ec-5, Center for Biological Evaluation and Research,FDA, Bethesda, MD, USA) intravenously over 1 min. The subjectswere observed for 10 h after LPS infusion with serial bloodsampling.

Real-time RT-PCR

For our in vitro experiments, total RNAwas isolated from cells withthe RNeasy kit (Qiagen, Hilden, Germany) and cDNA was synthesizedby reverse transcription using the High-Capacity cDNA reversetranscription kit (Applied Biosystems, Foster City, CA, USA). Inventor-ied Q2TaqMan assay mixes (Applied Biosystems) were used for quanti-fication as previously described [37]. Human TaqMan assays usedwere Hs00157965_m1 for HO-1, Hs01573472_g1 for COX-2 (PTGS2),

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎2

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

Hs00168547_m1 for NAD(P)H-quinone oxidoreductase-1 (NQO1),Hs00602020_mH for peroxiredoxin-1 (Prx1), Hs00230917_m1 forBach1, and 4333764 F/4352934E for glyceraldehyde-3-phosphatedehydrogenase (GAPDH). The mouse TaqMan assays Mm005160-07_m1 for HO-1 and 4352932E for GAPDH were used. Amplificationwas performed using the TaqMan Gene Expression Master Mix on aStepOnePlus real-time PCR system (Applied Biosystems) according tothe manufacturer's instructions.

For our in vivo endotoxemia experiments, PAXGene blood RNAtubes (PreAnalytiX, Qiagen, Venlo, The Netherlands) were used formRNA isolation at baseline and 2 and 4 h after LPS administrationand stored at �80 1C until further processing. RNA was extractedusing the whole blood RNA kit (Qiagen) according to the manufac-turer's instructions. cDNA synthesis was performed using an Omnis-cript reverse transcriptase kit (Qiagen). cDNA was stored at �20 1Cuntil further analysis. Amplification was performed with the ABIPrism 7000 sequence detection system. The master mix solutionconsisted of the TaqMan Universal PCR Mastermix, sterile water, andprobe. The primer–probe sets used were Hs99999905_m1 forGAPDH, Hs00157965_m1 for HO-1, Hs00157969_m1 for HO-2, andHs01573471_m1 for COX-2 (all from Applied Biosystems). The con-stitutively expressed gene GAPDH was used for normalization ofcDNA levels in all experiments. The ΔΔCT method was used toquantify mRNA levels, according to the manufacturer's protocol.

Western blot

For in vitro experiments, cells obtained from cell cultures werewashed twice with 0.9% NaCl. Then 300 μl of lysis buffer (2% SDS,10% glycerol, bromophenol blue, 0.4 M dithiothreitol, 4% proteaseinhibitors) was added, and the cells were scraped from the culturedishes and then homogenized by passage through a 25-gaugeneedle. The homogenate was incubated for 3 min at 95 1C and theprotein content was determined in the supernatant by theBradford method. Total protein (50 to 100 mg) was loaded onto aSDS–polyacrylamide gel (NuPAGE 4–12% Bis–Tris gel, Invitrogen)and electrophoretically separated under reducing conditions andthen transferred to polyvinylidene difluoride membranes (Milli-pore, Bedford, MA, USA) essentially as described previously [38].Membranes were blocked with PBS containing 5% skim milk orCandor blocking solution (Candor, Wangen im Allgäu, Germany),for 1 h at room temperature. The primary antibodies against HO-1(Enzo Life Sciences, Lörrach, Germany) and GAPDH (Santa CruzBiotechnology, Heidelberg, Germany) were used at a 1:1000dilution. The primary antibody against Bach1 was from Santa CruzBiotechnology and was used at a 1:750 dilution. Secondaryantibodies were goat anti-rabbit IgG–horseradish peroxidase(HRP; Bio-Rad, München, Germany) and anti-mouse IgG–HRP(Dako, Glostrup, Denmark) and were used at 1:20,000 and1:2000, respectively. The ECL chemiluminescence detection sys-tem (Thermo Fisher Scientific, Schwerte, Germany) was used fordetection according to the manufacturer's instructions. The che-miluminescent autoradiographic signals were visualized andquantitated with the Fluorchem FC2 gel documentation systemor using GE-X ray films.

For in vivo experiments, blood was drawn at baseline and 2 and4 h after LPS administration using CPT tubes (Becton–Dickinson,Alphen a/d Rijn, The Netherlands) for isolation of PBMCs accordingto the instructions of the manufacturer. Cells were stored at�80 1C after snap-freezing until analysis. Then, the PBMC pelletswere defrosted and lysed for 30 min on ice using lysis buffer(1 mmol/L EDTA, 1 mmol/L phenylmethanesulfonyl fluoride,0.5% Triton X-100, and protease inhibitor (Complete Mini, RocheApplied Science, Almere, The Netherlands) in PBS 1:100). Celllysates were centrifuged for 2 min at 20,000g at 4 1C and 10 mllysate was separated by SDS–PAGE using a 12.5% gel. Proteins were

blotted onto a nitrocellulose membrane using the Bio-Rad iBlotsystem and the membrane was blocked for 1 h at room tempera-ture using 5% ELK (milk powder, Albert Heijn, The Netherlands) inPBS. The blot was incubated for 1 h at room temperature withrabbit anti-HO-1 antibody (1:5000, Stressgen/ITK, Uithoorn, TheNetherlands). Mouse anti-β-actin antibody was simultaneouslyincubated to serve as a protein loading control (1:100,000,Sigma–Aldrich). Antibodies were diluted in 2.5% ELK containing0.1% Tween 20. Afterward, the membrane was rinsed three timeswith PBS and washed three times for 10 min with PBS containing0.1% Tween 20. The secondary antibodies, IRDye800-conjugatedanti-mouse IgG (1:10,000, Rockland, Heerhugowaard, The Nether-lands) and Alexa Fluor 680 goat anti-rabbit IgG (1:10,000, Mole-cular Probes), were incubated for 45 min at room temperature inOdyssey blocking buffer containing 0.1% Tween 20 and 0.01% SDS.After being rinsed with PBS three times and washed three timeswith PBS containing 0.1% Tween 20, the membrane was scannedusing the Odyssey infrared imaging system (Li-Cor Biosciences).Expression of HO-1 was assessed using channel 700 and β-actinexpression at channel 800. Intensity of the bands was determinedusing the Odyssey application software.

Data analysis, calculations, and statistics

All data were tested using either paired or unpaired Student's ttest as appropriate (Prism 5, GraphPad Software, La Jolla, CA, USA).All data are expressed as the mean 7 SEM and po0.05 wasconsidered statistically significant.

Results

LPS-dependent downregulation of HO-1 gene expression in cellcultures of primary human mononuclear cells

We determined the expression of HO-1 in various types ofprimary human mononuclear cells isolated from healthy volunteerdonors. In CD14þ monocytes LPS markedly downregulated HO-1protein and mRNA expression levels (Fig. 1A), whereas the proto-typical HO-1 inducer heme caused a prominent upregulation ofHO-1 (Fig. 1A). In time-response studies we found that primingwith LPS led to a fast (within 3 h) and persistent (50% lower after24 h) downregulation of HO-1 compared to controls (Fig. 1B).Moreover, this effect of LPS was dose-dependent (Fig. 1C). Supple-mentation of LPS-treated CD14þ cells with heme restored HO-1expression to mRNA levels of untreated control cells, indicatingthat heme can compensate for TLR4-mediated downregulation ofHO-1 (data not shown).

HO-1 mRNA regulation by LPS was also determined in variousother primary human cells of mononuclear origin, including totalPBMCs, monocytes, macrophages, DCs, and granulocytes. Althoughto different extents, LPS priming of mononuclear cells decreasedHO-1 gene expression (Fig. 1D). In contrast to human CD14þ

monocytes, HO-1 gene expression was markedly increased by LPSin mouse macrophages (Fig. 1E). Notably, in both human andmurine mononuclear cells, LPS upregulated production of tumornecrosis factor-α, independent of the differences in HO-1 regula-tion (data not shown). HO-1 gene expression was increased by LPSin two human mononuclear cell lines, THP-1 and MM-6, both ofleukemic origin (Fig. 1F). Finally, we also investigated the effects ofTLR ligands, other than LPS, on HO-1 gene expression in humanCD14þ monocytes. All TLR ligands, except the TLR7 ligand imi-quimod, significantly downregulated HO-1 mRNA levels (Table 1).

Taken together, the findings show that HO-1 expression isdownregulated by LPS in primary human mononuclear cells, butnot in primary mouse macrophages or human leukemia cell lines.

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 3

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 1. Effects of LPS and heme on HO-1 gene expression in primary human mononuclear cells and human leukemia cell lines. (A) CD14þ monocytes from human PBMCswere isolated and cultured as described under Materials and methods. The effects of LPS (1 mg/ml) and heme (10 mmol/L) on HO-1 expression were examined byWestern blotanalysis after 6 h (left) and RT-PCR after 3 h (right). Values 7 SEM represent the fold induction of HO-1 mRNA levels normalized to GAPDH from 10 independentexperiments. Statistics: Student's t test for paired values, *pr0.002, significant difference for treatment versus control. (B, C) Human CD14þ monocytes were cultured(B) with LPS (1 mg/ml) for the times indicated and (C) at increasing LPS concentrations (0.1 to 10 mg/ml) for 3 h, after which HO-1 mRNA expression was determined by RTPCR. Values 7 SEM represent relative HO-1 mRNA levels normalized to GAPDH from three independent experiments. Statistics: Student's t test for paired values, *pr0.007,significant difference for treatment versus control. (D) Human peripheral blood mononuclear cells (PBMC), monocytes (Mono), macrophages (Mac), dendritic cells (DC), andgranulocytes (Gran) were cultured as described under Materials and methods. Cells were treated with LPS (1 mg/ml) for 3 h and HO-1 mRNA expression was determined byRT-PCR. Values 7 SEM represent relative HO-1 mRNA levels normalized to GAPDH from three independent experiments. Statistics: Student's t test for paired values,*pr0.007, significant difference for treatment versus control. (E) Mouse alveolar macrophages were isolated and cultured as described under Materials and methods. Theeffects of LPS (1 mg/ml) and heme (10 mmol/L) on HO-1 expression were examined by Western blot analysis after 6 h (left) and RT-PCR after 3 h (right). Values 7 SEMrepresent the fold induction of HO-1 mRNA levels normalized to GAPDH from three independent experiments. Statistics: Student's t test for paired values, *pr0.002,significant difference for treatment versus control. (F) Human MM-6 and THP-1 leukemia cell lines were treated for 3 h with LPS (1 mg/ml) or heme (10 mmol/L) and HO-1mRNA expression was determined by RT-PCR (n¼3). Values 7 SEM represent the fold induction of HO-1 mRNA levels normalized to GAPDH from three independentexperiments. Statistics: Student's t test for paired values, *pr0.0001, significant difference for treatment versus control.

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎4

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

Downregulation of HO-1 by LPS but not other stress-inducible genesin human CD14þ monocytes

To determine the specificity of LPS-dependent HO-1 down-regulation in human mononuclear cells, we also studied theregulation of the inducible genes COX-2, NQO1, and Prx1, all ofwhich have previously been shown to be coordinately upregulatedto HO-1 in mononuclear cells [26,32,39,40]. In contrast to HO-1,treatment of human CD14þ cells with LPS, but not with heme,markedly upregulated COX-2 expression (Fig. 2) and to a lowerextent that of the stress-inducible antioxidant genes NQO1 andPrx1. Remarkably, COX-2 mRNA expression was also increased byother TLR ligands (Table 2). These findings suggest that down-regulation of HO-1 by LPS is specific in mononuclear cells.

Role of Bach1 in LPS-dependent HO-1 downregulation

Bach1 is a transcriptional repressor of HO-1, the activity ofwhich is mainly regulated via intracellular levels of free heme

[29,41]. Whereas priming of CD14þ monocytes with LPS increased,exposure to heme decreased Bach1 protein and mRNA expressionlevels (Fig. 3A). We also determined the effect of the transcrip-tional inhibitor AD and that of the protein synthesis inhibitor CHXon HO-1 and Bach1 mRNA expression in CD14þ monocytes.Whereas preincubation with AD only partially restored LPS-dependent downregulation of HO-1 mRNA levels, CHX almostcompletely blunted this effect (Fig. 3B). Moreover, AD reducedboth basal and LPS-dependent upregulation of Bach1 mRNA levels.In contrast, LPS-mediated induction of Bach1 mRNA was blockedby CHX (Fig. 3C), which suggests that a pathway involving de novoprotein synthesis is involved in this regulatory mechanism. Toconfirm the role of Bach1 in HO-1 regulation by LPS, we alsoperformed knockdown experiments with siRNA directed againstBach1, which decreased the levels of Bach1 mRNA and protein(Figs. 4A and B). In comparison to control cells, Bach1-knockdowncells exhibited markedly higher basal and LPS-mediated HO-1mRNA levels (Fig. 4C). In summary, these findings suggest that

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Table 1Fold change in HO-1 mRNA levels after 3 h incubation of human CD14þ monocyteswith various TLR ligands.

HO-1 SEM p value

Control 1.00 0.00 NATLR1/TLR2 (Pam3CSK4) 0.14 0.01 o0.0001TLR2 (HKLM) 0.15 0.02 o0.0001TLR2 (LTA) 0.12 0.01 o0.0001TLR3 (poly(I:C)) 0.73 0.07 0.02TLR4 (LPS) 0.07 0.02 o0.0001TLR5 (flagellin) 0.08 0.02 o0.0001TLR2/TLR6 (FSL1) 0.12 0.02 o0.0001TLR7 (imiquimod) 0.95 0.12 NSTLR8 (ssRNA40) 0.28 0.03 o0.0001TLR9 (ODN2006) 0.60 0.09 0.01

Concentrations used for incubation were 1 mg/ml Pam3CSK4, 108 freeze-dried cells ofheat-killed L. monocytogenes/ml (HKLM) for TLR2 activation, 5 mg/ml TLR2 agonistlipoteichoic acid (LTA), 10 mg/ml TLR3 agonist poly(I:C), 1 mg/ml TLR4 agonist LPS (E.coli 0111:B4), 1 mg/ml TLR5 agonist flagellin, 1 mg/ml TLR2/6 agonist fibroblast-stimulating ligand 1 (FSL1), 1 mg/ml TLR7 agonist imiquimod, 5 mg/ml TLR8 ligandssRNA40, 10 mg/ml TLR9 ligand oligodeoxynucleotide (ODN). Data are expressed as themean and SEM. The p values were determined using a paired Student t test of baselinecompared to the values after 3 h. NA, not applicable; NS, not significant.

Fig. 2. Regulation of COX-2, NQO1, and Prx1 by LPS and heme in comparison to HO-1 in CD14þ monocytes. CD14þ monocytes from human PBMCs were incubated with LPS(1 mg/ml) and heme (10 mmol/L) for 3 h, as indicated, and mRNA expression levels of HO-1, COX-2, NQO1, and Prx1 were determined. Values 7 SEM represent the foldinduction of mRNA levels normalized to GAPDH from at least three independent experiments. Statistics: Student's t test for paired values, *pr0.05, significant difference fortreatment versus control.

Table 2Fold change in COX-2 mRNA levels after 3 h incubation of human CD14þ

monocytes with various TLR ligands.

COX-2 SEM p value

Control 1.00 0.05 NATLR1/TLR2 (Pam3CSK4) 11.25 1.95 o0.007TLR2 (HKLM) 121.42 14.72 o0.002TLR2 (LTA) 89.67 25.22 o0.02TLR3 (poly(I:C)) 0.92 0.19 NSTLR4 (LPS) 657.23 162.25 o0.02TLR5 (flagellin) 1398.49 175.27 o0.003TLR2/TLR6 (FSL1) 20.28 3.04 o0.004TLR7 (imiquimod) 1.22 0.11 NSTLR8 (ssRNA40) 902.93 106.36 o0.002TLR9 (ODN2006) 1.11 0.11 NS

Concentrations used for incubation were 1 mg/ml Pam3CSK4, 108 freeze-dried cellsof heat-killed L. monocytogenes/ml (HKLM) for TLR2 activation, 5 mg/ml TLR2agonist lipoteichoic acid (LTA), 10 mg/ml TLR3 agonist poly(I:C), 1 mg/ml TLR4agonist LPS (E. coli 0111:B4), 1 mg/ml TLR5 agonist flagellin, 1 mg/ml TLR2/6 agonistfibroblast-stimulating ligand 1 (FSL1), 1 mg/ml TLR7 agonist imiquimod, 5 mg/mlTLR8 ligand ssRNA40, 10 mg/ml TLR9 ligand oligodeoxynucleotide (ODN). Data areexpressed as the mean and SEM. The p values were determined using a pairedStudent t test of baseline compared to the values after 3 h. NA, not applicable; NS,not significant.

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 5

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

LPS-mediated downregulation of HO-1 by LPS is mediated via aBach1-dependent mechanism.

HO-1 regulation in mononuclear cells of human individuals subjectedto experimental endotoxemia

We next determined HO-1 expression in PBMCs isolated from10 healthy male individuals with endotoxemia induced by infusionof LPS (2 ng/kg body wt). To confirm the proinflammatory effectsof LPS, various clinical and laboratory parameters were analyzed.Clinical signs of inflammation including headache, muscle pain,nausea, an increase in body temperature of 1.970.5 1C (po0.001),decreased mean arterial blood pressure (10178 to 7978 mm Hg;po0.001), and rise in heart rate (6676 bpm at baseline to 9377bpm; po0.001) after LPS challenge were observed. In addition, weobserved a marked increase in leukocyte counts from4.771.9�109 to 11.372.5�109 cells/L 6 h after LPS administra-tion (po0.001) (data not shown) and increased C-reactive protein(Table 3). Moreover, a number of inducible pro- and anti-inflammatory cytokines were upregulated (Table 3). Similar toour in vitro findings, HO-1 mRNA expression was downregulated,whereas COX-2 mRNA was upregulated in PBMCs, which wereisolated 2 and 4 h after onset of LPS treatment. Accordingly, HO-1protein expression in human PBMCs was markedly reduced at2 and 4 h after LPS infusion (Fig. 5). Taken together, the data fromthis in vivo model of human endotoxemia confirm our in vitrofindings on LPS-dependent downregulation of HO-1 gene expres-sion in mononuclear cells.

Discussion

HO-1 plays a major immunomodulatory role via its cell-type-specific anti-inflammatory functions in mononuclear cells[16,17,42,43]. The importance of HO-1 in inflammation has beenhighlighted in studies on murine and human genetic HO-1 defi-ciency, in both of which proinflammatory phenotypical featureshave been observed [9,10,43]. In the present study, we demonstratethat LPS downregulates HO-1 gene expression in various primaryhuman mononuclear cells in vitro and in vivo. Furthermore, weprovide evidence that LPS-dependent downregulation of HO-1 ismediated via the transcriptional repressor Bach1.

Reportedly, HO-1 expression in mononuclear phagocytes is upre-gulated by a large variety of stress stimuli, including LPS [7,44].Specifically, LPS has been shown to increase HO-1 gene expression inrodent peripheral blood mononuclear cells and in the mouse mono-cyte cell lines RAW264.7 and J766.4 [24,25,45,46]. In opposition tothese earlier findings, other reports have demonstrated that LPSdownregulates HO-1 expression in human mononuclear cells[27,28]. Hence, we hypothesized that these discrepancies are relatedto the cell type used in the various experimental models. In support ofthis hypothesis, we found that LPS markedly downregulates HO-1gene expression in primary human mononuclear cells, includingCD14þ monocytes, macrophages, DCs, and granulocytes (Fig. 1).Moreover, LPS-dependent downregulation of HO-1 gene expressionwas also observed in mononuclear cells obtained from healthy humanindividuals exposed to experimental endotoxemia (Fig. 5). These dataare in agreement with previous findings on human monocytes [27]and DCs [28]. In contrast to primary human mononuclear cells, HO-1was upregulated in LPS-treated rodent macrophages (Fig. 1E). Thus,the current observations together with those of earlier reports on HO-1 regulation by LPS suggest that species-specific differences betweenrodent and human mononuclear cells may be responsible for thesediscrepancies, which may have various explanations. Rodents not

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 3. Regulation of Bach1 by LPS in CD14þ monocytes. Effects of actinomycin D(AD) and cycloheximide (CHX) on LPS-mediated regulation of HO-1 and Bach1mRNA levels. (A) CD14þ monocytes were exposed to LPS (1 mg/ml) and heme(10 mmol/L) for 3 h, as indicated, and protein (left) and mRNA (right) expressionlevels of Bach1 were determined as described under Materials and methods. Values7 SEM represent the fold induction of Bach1 levels normalized to GAPDH fromthree independent experiments. Statistics: Student's t test for paired values,*po0.05, **po0.005, significant difference for treatment versus control.(B) Effects of pretreatment with AD (5 μg/ml, left) and CHX (10 μg/ml, right) onLPS-dependent regulation of HO-1 mRNA expression levels in CD14þ monocytes.Values 7 SEM represent the fold induction of HO-1 mRNA levels normalized toGAPDH from three independent experiments. Statistics: Student's t test for pairedvalues, *pr0.001, **pr0.02, significant difference for treatment versus control.(C) Effects of AD (left) and CHX (right) on LPS-mediated induction of Bach1 mRNAexpression in CD14þ monocytes. Values 7 SEM represent the fold induction ofBach1 mRNA levels normalized to GAPDH from three independent experiments.Statistics: Student's t test for paired values, *pr0.05, significant difference fortreatment versus control.

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎6

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

only live in an environment different from that of humans, but alsoare of much smaller size and have significantly shorter life spans. Ithas also been demonstrated that rodents are more resistant to LPS incomparison to humans, indicating that their immune systems mayhave evolved in subtly different ways. For example, inducible NOsynthase (NOS), which is synonymous with NOS-2, is readily inducedby interferon-γ and LPS in murine, but not in human, macrophages[47]. Notably, HO-1 induction in mononuclear cells by LPS priming iswell documented in animal experiments. For example, HO-1 expres-sion has been shown to be markedly increased in heart [21], kidney,and, to a lesser extent, liver, lung, and brain of LPS-treated rats [22].Moreover, our results demonstrate differences in LPS-dependentregulation of HO-1 gene expression between primary human mono-nuclear cells and human leukemia cell lines (Fig. 1), which arefrequently used for in vitro experiments. Taken together, currentfindings along with those of previous reports indicate that experi-mental observations on HO-1 regulation in murine cell culturemodels and in transformed monocyte cell lines cannot be directlytranslated to primary human mononuclear cells.

HO-1 downregulation by LPS in human mononuclear cellsinvolves the transcriptional repressor Bach1, which controls HO-1gene expression via interactionwith a Maf recognition site in the HO-1 gene promoter [29,48]. A critical role of Bach1 in this regulatorypathway is supported by the inverse pattern of HO-1 and Bach1protein and mRNA levels in LPS-treated CD14þ cells (Figs. 1 and 3)and by HO-1 expression in Bach1-knockdown cells (Fig. 4). Inaccordance with our current findings, Miyazaki and colleagues haveshown an upregulation of Bach1 by LPS in human mononuclearcells [27]. The mechanism of Bach1-dependent HO-1 regulation inresponse to LPS seems to involve a protein with a short half-life,which regulates de novo synthesis of Bach1, as indicated by experi-ments with the protein synthesis inhibitor CHX (Fig. 3). Bach1 is aheme-binding protein and intracellular levels of free heme regulatethis nuclear repressor [29,30]. Therefore, it is conceivable that freeheme levels are involved in LPS-mediated HO-1 regulation via Bach1in mononuclear cells. The marked LPS-dependent upregulation of theheme-containing protein COX-2 in parallel to HO-1 downregulation(Fig. 2, Table 2) might suggest that incorporation of heme into COX-2can decrease intracellular free heme levels, which in turn leads toBach1-dependent repression of HO-1. In addition, Nrf2, a masterregulator of the inducible antioxidant cytoprotective response [49],has been shown to be involved in HO-1 regulation by LPS [7].However, the current data show a differential LPS-dependent upre-gulation of the Nrf2-inducible genes NQO1 and Prx1 in comparisonto that of HO-1 (Fig. 2), suggesting a key role for Bach1 in LPS-dependent downregulation of HO-1 in primary human mononuclearcells (Figs. 3 and 4). Therefore, these findings are similar to those of arecent report on the Bach1-specific regulation of HO-1 by electro-philic stimuli in human keratinocytes [50].

The Bach1/HO-1 pathway plays a central role in the differentia-tion and immunological functions of mononuclear cells, as recentlydemonstrated in Bach1/HO-1 double-knockout mice [31,51]. Inde-pendently, myeloid HO-1 has been shown to have critical immuno-modulatory functions in a mouse conditional knockout model inwhich HO-1 was specifically deleted in mononuclear cells [16].Importantly, the opposite regulation of HO-1 by LPS and other TLRligands in murine and human macrophages has clinical implications,

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 4. Role of Bach1 in LPS-dependent HO-1 downregulation. Knockdown of Bach1 in human CD14þ cells (siBach1) was performed as described under Materials andmethods. (A) Values 7 SEM represent relative Bach1 mRNA levels normalized to GAPDH from three independent experiments. Statistics: Student's t test for paired values,*pr0.05, significant difference for treatment versus control. (B) A representative Western blot is presented and values 7 SEM with relative levels of Bach1 proteinnormalized to GAPDH from three independent experiments are shown. Statistics: Student's t test for paired values, *pr0.05, significant difference for treatment versuscontrol. (C) LPS-dependent regulation of HO-1 mRNA levels in human CD14þ cells after knockdown of Bach1 (siBach1). Values 7 SEM represent the fold induction of Bach1mRNA levels normalized to GAPDH from four independent experiments. Statistics: Student's t test for paired values, *pr0.002, significant difference for treatment versuscontrol; **pr0.002, significant difference for Bach1 siRNA versus control siRNA.

Table 3Inflammatory markers after infusion of LPS (2 ng/kg) into healthy human indivi-duals (n¼10).

Marker Time after LPS (h) p value

0 2 4 24

IL-6 (pg/ml) 670 1,5217209 6507237 670 o0.001TNF-α (pg/ml) 2070 1,2137187 821783 2070 o0.001IL-1β (pg/ml) 1070 3273 3475 1070 o0.001MCP-1 (pg/ml) 4474 3,6057267 6,16171,302 6275 o0.01IL-10 (pg/ml) 670 73711 47711 771 o0.001IL-1RA (pg/ml) 6177 2,5787481 31,54072,583 82715 o0.001CRP (mg/L) 570 — — 3974 o0.001

Data are expressed as the mean and SEM. The p values were determined using apaired Student t test of baseline compared to the peak value. IL, interleukin; IL-1RA,IL-1 receptor antagonist; TNF-α, tumor necrosis factor-α; MCP-1, monocyte che-motactic protein-1; CRP, C-reactive protein.

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 7

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

because myeloid HO-1 is a potential target for therapeutic interven-tions in various inflammatory disorders such as sepsis [52] andmalaria [53].

Conclusions

In conclusion, this study demonstrates that the TLR4 ligand LPSspecifically downregulates HO-1 in human mononuclear cells viathe transcriptional repressor Bach1. These results imply thatexperimental findings on HO-1 gene expression in various speciesand cell lines need careful interpretation, because HO-1 regulationin primary human mononuclear cells is different from that inwidely used experimental animal models. Further studies arewarranted to understand the regulation of HO-1 in mononuclearcells of different species. These studies may be important for

clinical applications in which myeloid HO-1 serves as a therapeutictarget for the treatment of inflammation.

Acknowledgments

We thank the Intensive Care Research Unit under the super-vision of Tijn Bouw for their indispensable help during the humanendotoxemia experiments. M.J.D. is the recipient of a ZonMWAGIKO stipendium (No. 92003507). F.A.D.T.G.W. is supported bygrants from the Dutch Burns Foundation (No. 09.110) and theDutch Diabetes Research Foundation (No. 2006.00.055). Work in S.I.'s laboratory is supported by the Deutsche Forschungsge-meinschaft (SFB547, A8) and the Bundesministerium für Bildungund Forschung (No. 0316045B). S.J. is supported by the GermanLung Research Center.

References Q3

[1] Maines, M. D. The heme oxygenase system: a regulator of second messengergases. Annu. Rev. Pharmacol. Toxicol. 37:517–554; 1997.

[2] Immenschuh, S.; Ramadori, G. Gene regulation of heme oxygenase-1 as atherapeutic target. Biochem. Pharmacol. 60:1121–1128; 2000.

[3] Ryter, S. W.; Alam, J.; Choi, A. M. Heme oxygenase-1/carbon monoxide: frombasic science to therapeutic applications. Physiol. Rev. 86:583–650; 2006.

[4] Lundvig, D. M.; Immenschuh, S.; Wagener, F. A. Heme oxygenase, inflamma-tion, and fibrosis: the good, the bad, and the ugly? Front. Pharmacol. 3:81;2012.

[5] Tenhunen, R.; Marver, H. S.; Schmid, R. The enzymatic conversion of heme tobilirubin by microsomal heme oxygenase. Proc. Natl. Acad. Sci. USA 61:748–755;1968.

[6] Alam, J.; Cook, J. L. How many transcription factors does it take to turn on theheme oxygenase-1 gene? Am. J. Respir. Cell Mol. Biol. 36:166–174; 2007.

[7] Paine, A.; Eiz-Vesper, B.; Blasczyk, R.; Immenschuh, S. Signaling to hemeoxygenase-1 and its anti-inflammatory therapeutic potential. Biochem. Phar-macol. 80:1895–1903; 2010.

[8] Trakshel, G. M.; Kutty, R. K.; Maines, M. D. Purification and characterization ofthe major constitutive form of testicular heme oxygenase: the noninducibleisoform. J. Biol. Chem. 261:11131–11137; 1986.

[9] Poss, K. D.; Tonegawa, S. Reduced stress defense in heme oxygenase 1-deficient cells. Proc. Natl. Acad. Sci. USA 94:10925–10930; 1997.

[10] Yachie, A.; Niida, Y.; Wada, T.; Igarashi, N.; Kaneda, H.; Toma, T.; Ohta, K.;Kasahara, Y.; Koizumi, S. Oxidative stress causes enhanced endothelial cellinjury in human heme oxygenase-1 deficiency. J. Clin. Invest. 103:129–135;1999.

[11] Chung, S. W.; Liu, X.; Macias, A. A.; Baron, R. M.; Perrella, M. A. Hemeoxygenase-1-derived carbon monoxide enhances the host defense response tomicrobial sepsis in mice. J. Clin. Invest. 118:239–247; 2008.

[12] Tsoyi, K.; Lee, T. Y.; Lee, Y. S.; Kim, H. J.; Seo, H. G.; Lee, J. H.; Chang, K. C. Heme-oxygenase-1 induction and carbon monoxide-releasing molecule inhibitlipopolysaccharide (LPS)-induced high-mobility group box 1 release in vitroand improve survival of mice in LPS- and cecal ligation and puncture-inducedsepsis model in vivo. Mol. Pharmacol. 76:173–182; 2009.

[13] Akamatsu, Y.; Haga, M.; Tyagi, S.; Yamashita, K.; Graca-Souza, A. V.; Ollinger, R.;Czismadia, E.; May, G. A.; Ifedigbo, E.; Otterbein, L. E.; Bach, F. H.; Soares, M. P.Heme oxygenase-1-derived carbon monoxide protects hearts from transplantassociated ischemia reperfusion injury. FASEB J 18:771–772; 2004.

[14] Sato, K.; Balla, J.; Otterbein, L.; Smith, R. N.; Brouard, S.; Lin, Y.; Csizmadia, E.;Sevigny, J.; Robson, S. C.; Vercellotti, G.; Choi, A. M.; Bach, F. H.; Soares, M. P.Carbon monoxide generated by heme oxygenase-1 suppresses the rejection ofmouse-to-rat cardiac transplants. J. Immunol. 166:4185–4194; 2001.

[15] Vulapalli, S. R.; Chen, Z.; Chua, B. H.; Wang, T.; Liang, C. S. Cardioselectiveoverexpression of HO-1 prevents I/R-induced cardiac dysfunction and apop-tosis. Am. J. Physiol. Heart Circ. Physiol 283:H688–694; 2002.

[16] Tzima, S.; Victoratos, P.; Kranidioti, K.; Alexiou, M.; Kollias, G. Myeloid hemeoxygenase-1 regulates innate immunity and autoimmunity by modulatingIFN-beta production. J. Exp. Med. 206:1167–1179; 2009.

[17] Hull, T. D.; Agarwal, A.; George, J. F. The mononuclear phagocyte system inhomeostasis and disease: a role for heme oxygenase-1. Antioxid. RedoxSignaling 20:1770–1788; 2014.

[18] Beutler, B. TLR4 as the mammalian endotoxin sensor. Curr. Top. Microbiol.Immunol. 270:109–120; 2002.

[19] Bahador, M.; Cross, A. S. From therapy to experimental model: a hundredyears of endotoxin administration to human subjects. J. Endotoxin Res.13:251–279; 2007.

[20] Guha, M.; Mackman, N. LPS induction of gene expression in human mono-cytes. Cell. Signalling 13:85–94; 2001.

[21] Tamion, F.; Richard, V.; Renet, S.; Thuillez, C. Protective effects of heme-oxygenase expression against endotoxic shock: inhibition of tumor necrosis

123456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566

676869707172737475767778798081828384858687888990919293949596979899

100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132

Fig. 5. HO-1 mRNA and protein expression in PBMCs from human individuals afterinfusion of LPS. Human PBMCs were isolated before and 2 or 4 h after LPS infusioninto human individuals, as described under Materials and methods. (A) Expressionof HO-1 and COX-2 mRNA levels at baseline compared to 2 and 4 h after LPSinfusion was determined by RT-PCR. Values 7 SEM represent the relative HO-1and the fold induction of COX-2 mRNA levels normalized to GAPDH from 10individuals. Statistics: Student's t test for paired values, *po0.001, significantdifference for treatment versus baseline control. (B) HO-1 protein expression atbaseline compared to 2 and 4 h after LPS infusion was determined by Westernblotting. Values 7 SEM represent the relative HO-1 mRNA protein levels normal-ized to β-actin from 10 individuals. Statistics: Student's t test for paired values,*po0.005, significant difference for treatment versus baseline control.

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎8

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i

factor-alpha and augmentation of interleukin-10. J. Trauma 61:1078–1084;2006.

[22] Furuichi, M.; Yokozuka, M.; Takemori, K.; Yamanashi, Y.; Sakamoto, A. Thereciprocal relationship between heme oxygenase and nitric oxide synthase inthe organs of lipopolysaccharide-treated rodents. Biomed. Res. 30:235–243;2009.

[23] Camhi, S. L.; Alam, J.; Otterbein, L.; Sylvester, S. L.; Choi, A. M. K. Induction ofheme oxygenase-1 gene expression by lipopolysaccharide is mediated by AP-1activation. Am. J. Respir. Cell Mol. Biol. 13:387–398; 1995.

[24] Immenschuh, S.; Tan, M.; Ramadori, G. Nitric oxide mediates the lipopoly-saccharide dependent upregulation of the heme oxygenase-1 gene expressionin cultured rat Kupffer cells. J. Hepatol. 30:61–69; 1999.

[25] Wijayanti, N.; Huber, S.; Samoylenko, A.; Kietzmann, T.; Immenschuh, S. Roleof NF-kappaB and p38 MAP kinase signaling pathways in thelipopolysaccharide-dependent activation of heme oxygenase-1 gene expres-sion. Antioxid. Redox Signaling 6:802–810; 2004.

[26] Rushworth, S. A.; Chen, X. L.; Mackman, N.; Ogborne, R. M.; O'Connell, M. A.Lipopolysaccharide-induced heme oxygenase-1 expression in human monocyticcells is mediated via Nrf2 and protein kinase C. J. Immunol. 175:4408–4415; 2005.

[27] Miyazaki, T.; Kirino, Y.; Takeno, M.; Samukawa, S.; Hama, M.; Tanaka, M.;Yamaji, S.; Ueda, A.; Tomita, N.; Fujita, H.; Ishigatsubo, Y. Expression of hemeoxygenase-1 in human leukemic cells and its regulation by transcriptionalrepressor Bach1. Cancer Sci. 101:1409–1416; 2010.

[28] Chauveau, C.; Remy, S.; Royer, P. J.; Hill, M.; Tanguy-Royer, S.; Hubert, F. X.;Tesson, L.; Brion, R.; Beriou, G.; Gregoire, M.; Josien, R.; Cuturi, M. C.; Anegon, I.Heme oxygenase-1 expression inhibits dendritic cell maturation and proinflam-matory function but conserves IL-10 expression. Blood 106:1694–1702; 2005.

[29] Ogawa, K.; Sun, J.; Taketani, S.; Nakajima, O.; Nishitani, C.; Sassa, S.; Hayashi,N.; Yamamoto, M.; Shibahara, S.; Fujita, H.; Igarashi, K. Heme mediatesderepression of Maf recognition element through direct binding to transcrip-tion repressor Bach1. EMBO J. 20:2835–2843; 2001.

[30] Igarashi, K.; Sun, J. The heme–Bach1 pathway in the regulation of oxidative stressresponse and erythroid differentiation. Antioxid. Redox Signaling 8:107–118; 2006.

[31] So, A. Y.; Garcia-Flores, Y.; Minisandram, A.; Martin, A.; Taganov, K.; Boldin, M.;Baltimore, D. Regulation of APC development, immune response, and auto-immunity by Bach1/HO-1 pathway in mice. Blood 120:2428–2437; 2012.

[32] Vijayan, V.; Baumgart-Vogt, E.; Naidu, S.; Qian, G.; Immenschuh, S. Bruton'styrosine kinase is required for TLR-dependent heme oxygenase-1 geneactivation via Nrf2 in macrophages. J. Immunol. 187:817–827; 2011.

[33] Paine, A.; Kirchner, H.; Immenschuh, S.; Oelke, M.; Blasczyk, R.; Eiz-Vesper, B.IL-2 upregulates CD86 expression on human CD4(þ) and CD8(þ) T cells. J.Immunol. 188:1620–1629; 2012.

[34] Al-Omari, M.; Korenbaum, E.; Ballmaier, M.; Lehmann, U.; Jonigk, D.;Manstein, D. J.; Welte, T.; Mahadeva, R.; Janciauskiene, S. Acute-phase proteinalpha1-antitrypsin inhibits neutrophil calpain I and induces random migra-tion. Mol. Med. 17:865–874; 2011.

[35] Dorresteijn, M. J.; Visser, T.; Cox, L. A.; Bouw, M. P.; Pillay, J.; Koenderman, A. H.;Strengers, P. F.; Leenen, L. P.; van der Hoeven, J. G.; Koenderman, L.; Pickkers, P. C1-esterase inhibitor attenuates the inflammatory response during human endotox-emia. Crit. Care Med. 38:2139–2145; 2010.

[36] van Eijk, L. T.; Dorresteijn, M. J.; Smits, P.; van der Hoeven, J. G.; Netea, M. G.;Pickkers, P. Gender differences in the innate immune response and vascularreactivity following the administration of endotoxin to human volunteers. Crit.Care Med. 35:1464–1469; 2007.

[37] Saragih, H.; Zilian, E.; Jaimes, Y.; Paine, A.; Figueiredo, C.; Eiz-Vesper, B.; Blasczyk, R.;Larmann, J.; Theilmeier, G.; Burg-Roderfeld, M.; Andrei-Selmer, L. C.; Becker, J. U.;Santoso, S.; Immenschuh, S. PECAM-1-dependent heme oxygenase-1 regulation viaan Nrf2-mediated pathway in endothelial cells. Thromb. Haemostasis 111:1077–1088;2014.

[38] Wijayanti, N.; Kietzmann, T.; Immenschuh, S. Heme oxygenase-1 gene activa-tion by the NAD(P)H oxidase inhibitor 4-(2-aminoethyl) benzenesulfonylfluoride via a protein kinase B, p38-dependent signaling pathway in mono-cytes. J. Biol. Chem. 280:21820–21829; 2005.

[39] Immenschuh, S.; Stritzke, J.; Iwahara, S. -I.; Ramadori, G. Up-regulation ofHeme-Binding Protein 23 (HBP23) gene expression by lipopolysaccharide ismediated via a nitric oxide-dependent signaling pathway in rat Kupffer cells.Hepatology 30:118–127; 1999.

[40] Naidu, S.; Vijayan, V.; Santoso, S.; Kietzmann, T.; Immenschuh, S. Inhibitionand genetic deficiency of p38 MAPK up-regulates heme oxygenase-1 geneexpression via Nrf2. J. Immunol. 182:7048–7057; 2009.

[41] Sun, J.; Brand, M.; Zenke, Y.; Tashiro, S.; Groudine, M.; Igarashi, K. Hemeregulates the dynamic exchange of Bach1 and NF-E2-related factors in the Maftranscription factor network. Proc. Natl. Acad. Sci. USA 101:1461–1466; 2004.

[42] Willis, D.; Moore, A. R.; Frederick, R.; Willoughby, D. A. Heme oxygenase: anovel target for the modulation of the inflammatory response. Nat. Med2:87–90; 1996.

[43] Kapturczak, M. H.; Wasserfall, C.; Brusko, T.; Campbell-Thompson, M.; Ellis, T. M.;Atkinson, M. A.; Agarwal, A. Heme oxygenase-1 modulates early inflammatoryresponses: evidence from the heme oxygenase-1-deficient mouse. Am. J. Pathol.165:1045–1053; 2004.

[44] Choi, A. M.; Alam, J. Heme oxygenase-1: function, regulation, and implicationof a novel stress-inducible protein in oxidant-induced lung injury. Am. J.Respir. Cell Mol. Biol. 15:9–19; 1996.

[45] Olszanecki, R.; Marcinkiewicz, J. Taurine chloramine and taurine bromamineinduce heme oxygenase-1 in resting and LPS-stimulated J774.2 macrophages.Amino Acids 27:29–35; 2004.

[46] Waltz, P.; Carchman, E. H.; Young, A. C.; Rao, J.; Rosengart, M. R.; Kaczorowski, D.;Zuckerbraun, B. S. Lipopolysaccaride induces autophagic signaling in macrophagesvia a TLR4, heme oxygenase-1 dependent pathway. Autophagy 7:315–320; 2011.

[47] Bogdan, C. Nitric oxide and the immune response. Nat. Immunol. 2:907–916;2001.

[48] Sun, J.; Hoshino, H.; Takaku, K.; Nakajima, O.; Muto, A.; Suzuki, H.; Tashiro, S.;Takahashi, S.; Shibahara, S.; Alam, J.; Taketo, M. M.; Yamamoto, M.; Igarashi, K.Hemoprotein Bach1 regulates enhancer availability of heme oxygenase-1gene. EMBO J. 21:5216–5224; 2002.

[49] Kensler, T. W.; Wakabayashi, N.; Biswal, S. Cell survival responses to environ-mental stresses via the Keap1–Nrf2–ARE pathway. Annu. Rev. Pharmacol.Toxicol. 47:89–116; 2007.

[50] MacLeod, A. K.; McMahon, M.; Plummer, S. M.; Higgins, L. G.; Penning, T. M.;Igarashi, K.; Hayes, J. D. Characterization of the cancer chemopreventive NRF2-dependent gene battery in human keratinocytes: demonstration that theKEAP1–NRF2 pathway, and not the BACH1–NRF2 pathway, controls cytopro-tection against electrophiles as well as redox-cycling compounds. Carcinogen-esis 30:1571–1580; 2009.

[51] Haldar, M.; Kohyama, M.; So, A. Y.; Kc, W.; Wu, X.; Briseno, C. G.; Satpathy, A. T.;Kretzer, N. M.; Arase, H.; Rajasekaran, N. S.; Wang, L.; Egawa, T.; Igarashi, K.;Baltimore, D.; Murphy, T. L.; Murphy, K. M. Heme-mediated SPI-C inductionpromotes monocyte differentiation into iron-recycling macrophages. Cell156:1223–1234; 2014.

[52] Larsen, R.; Gozzelino, R.; Jeney, V.; Tokaji, L.; Bozza, F. A.; Japiassu, A. M.;Bonaparte, D.; Cavalcante, M. M.; Chora, A.; Ferreira, A.; Marguti, I.; Cardoso,S.; Sepulveda, N.; Smith, A.; Soares, M. P. A central role for free heme in thepathogenesis of severe sepsis. Sci. Transl. Med. 2:51ra71; 2010.

[53] Cunnington, A. J.; de Souza, J. B.; Walther, M.; Riley, E. M. Malaria impairsresistance to Salmonella through heme- and heme oxygenase-dependentdysfunctional granulocyte mobilization. Nat. Med 18:120–127; 2012.

123456789

1011121314151617181920212223242526272829303132333435363738394041424344454647

48495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293

M.J. Dorresteijn et al. / Free Radical Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎ 9

Please cite this article as: Dorresteijn, MJ; et al. Cell-type-specific downregulation of heme oxygenase-1 by lipopolysaccharide via Bach1in primary human mononuclear cells. Free Radic. Biol. Med. (2014), http://dx.doi.org/10.1016/j.freeradbiomed.2014.10.579i