Tumor necrosis factor ╨Ю┬▒ acceleration of inflammatory responses by down-regulating heme oxygenase 1 in human peripheral monocytes.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 56, No. 2, February 2007, pp 464–475 DOI 10.1002/art.22370 © 2007, American College of Rheumatology Tumor Necrosis Factor ␣ Acceleration of Inflammatory Responses by Down-Regulating Heme Oxygenase 1 in Human Peripheral Monocytes Yohei Kirino, Mitsuhiro Takeno, Shuji Murakami, Masayoshi Kobayashi, Hideo Kobayashi, Kenji Miura, Haruko Ideguchi, Shigeru Ohno, Atsuhisa Ueda, and Yoshiaki Ishigatsubo Objective. To examine the interaction between heme oxygenase 1 (HO-1), a stress-induced antiinflammatory protein, and tumor necrosis factor ␣ (TNF␣) in human peripheral blood monocytes. Methods. Peripheral blood mononuclear cells (PBMCs) were obtained from healthy donors or from patients with rheumatoid arthritis (RA) receiving the anti–tumor necrosis factor ␣ (anti-TNF␣) monoclonal antibody infliximab. CD14ⴙ cells were isolated by magnetic cell sorting, cultured with TNF␣ or auranofin, and transfected with a plasmid encoding HO-1 or an HO-1– specific small interfering RNA vector. Protein and messenger RNA (mRNA) levels were examined by immunoblotting and real-time polymerase chain reaction. Cytokine levels in culture supernatants were measured by enzyme-linked immunosorbent assay. HO-1 gene transcription was evaluated using a luciferase reporter gene assay. Actinomycin D and cycloheximide were used to monitor the stability of mRNA and protein. Results. HO-1 is constitutively expressed by CD14ⴙ PBMCs from healthy donors. TNF␣ suppressed HO-1 expression by accelerating the decay of mRNA without affecting gene transcription or protein stability. Forced expression or selective knock-down of the HO-1 gene expression resulted in down-regulation or upregulation, respectively, of proinflammatory cytokine synthesis by monocytes. Treatment with infliximab significantly increased HO-1 mRNA levels and reduced TNF␣ synthesis by PBMCs from RA patients. Conclusion. TNF␣ accelerated inflammatory responses by down-regulating HO-1 expression in human monocytes. TNF antagonists may block this TNFdependent suppression of HO-1 expression, resulting in an amelioration of inflammation. Dr. Kirino’s work was supported by a 2005 grant from the Yokohama Foundation for Advancement of Medical Science. Dr. Takeno’s work was supported by a 2004–2005 grant-in-aid for scientific research (project 16590991) from the Ministry of Education, Culture, Sports, and Technology of Japan and a 2006 grant from the Yokohama Foundation for Advancement of Medical Science. Dr. Ishigatsubo’s work was supported by grants from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (Yokohama City University Center of Excellence Program), the Ministry of Health, Labor, and Welfare (Health Science Research on Specific Disease), and Yokohama City University (2006 Strategic Research project K18006). Yohei Kirino, MD, Mitsuhiro Takeno, MD, PhD, Shuji Murakami, MD, Masayoshi Kobayashi, MD, Hideo Kobayashi, MD, PhD, Kenji Miura, MD, Haruko Ideguchi, MD, PhD, Shigeru Ohno, MD, PhD, Atsuhisa Ueda, MD, PhD, Yoshiaki Ishigatsubo, MD, PhD: Yokohama City University, Graduate School of Medicine, Yokohama, Japan. Address correspondence and reprint requests to Yoshiaki Ishigatsubo, MD, PhD, Department of Internal Medicine and Clinical Immunology, Yokohama City University, Graduate School of Medicine, 3-9 Fukuura, Kanazawa-ku, Yokohama, Kanagawa, 236-0004, Japan. E-mail: firstname.lastname@example.org. Submitted for publication May 19, 2006; accepted in revised form October 23, 2006. Heme oxygenase (HO) is an enzyme that converts heme into carbon monoxide, Fe2⫹, and biliverdin (1,2). HO-1, an inducible isozyme of HO, is a 32-kd heat-shock protein that is expressed in response to a variety of noxious stimuli, including heavy metals, hyperoxia, hypoxia, endotoxins, and hydrogen peroxide (1,2). Evidence suggests that increased expression of HO-1 is beneficial in a variety of pathologic conditions (1,2). For example, HO-1 gene therapy has been successfully used to treat various lung diseases in animals (3–7), and chemical induction of HO-1 expression has been shown to improve lupus nephritis in MRL/lpr mice (8). In contrast, a deficiency of HO-1 expression is associated with severe chronic inflammation, as shown in studies of HO-1–knockout mice (9) and in a patient with HO-1 deficiency (10). These findings are consistent with a physiologic role of HO-1 in protection against inflammation. 464 TNF␣ DOWN-REGULATION OF HO-1 IN HUMAN PERIPHERAL MONOCYTES Investigators in our laboratory have been studying the regulation of HO-1 expression in humans with inflammatory and rheumatic diseases. Previous studies have shown that serum HO-1 levels are increased in patients with active hemophagocytic syndrome and adult-onset Still’s disease and that HO-1 levels are correlated with serum ferritin levels (11). Recent studies indicate that HO-1 is abundantly expressed in the synovial tissues of patients with rheumatoid arthritis (RA), although serum HO-1 levels are not elevated (11,12). Up-regulating HO-1 expression in RA-derived synovial cell lines resulted in suppressed inflammatory responses, suggesting that endogenously expressed HO-1 plays a regulatory role in the development of synovial inflammation and that induction of HO-1 may be of therapeutic utility (12). Inflammation is mediated, at least in part, by various cytokines that can alter HO-1 expression. For example, there is a close association between interleukin-10 (IL-10) and HO-1 (3,13,14). IL-10 induces the expression of HO-1, while HO-1 augments the production of IL-10 and reduces the production of proinflammatory cytokines. The interaction between HO-1 and tumor necrosis factor ␣ (TNF␣), which plays a pivotal role in the pathogenesis of chronic inflammatory diseases (15), is less clear. HO-1 expression is reduced by TNF␣ in chondrocytes derived from osteoarthritis patients (16), whereas TNF␣-dependent enhancement of HO-1 expression was observed in human endothelial cells (17,18), human monocytic tumorderived cell lines (19), and retinal pigment epithelial cells (20). The interplay between TNF␣ and HO-1 may be clarified by studying monocyte/macrophage cell linage, which might elucidate the mechanism underlying certain inflammatory disorders. This is particularly relevant, since TNF␣ antagonists are critical therapeutic agents used for the treatment of chronic inflammatory diseases, including RA (15,21,22). In the present study, we investigated the effects of TNF␣ on messenger RNA (mRNA) expression, gene transcription, mRNA stability, and posttranslational regulation of HO-1 in human monocytes. The inflammatory responses of monocytes transfected with HO-1 complementary DNA (cDNA) or HO-1–specific small interfering RNA (siRNA) were also examined. The results indicated that TNF␣ significantly reduces HO-1 expression in peripheral monocytes, thereby augmenting inflammatory responses. In addition, a single injection of infliximab resulted in the up-regulation of HO-1 mRNA 465 expression in peripheral blood mononuclear cells (PBMCs) from RA patients. PATIENTS AND METHODS Patients and healthy donors. Twelve patients with RA (10 women and 2 men), who met the American College of Rheumatology (formerly, the American Rheumatism Association) 1987 criteria (23), were enrolled in the study (Table 1). The mean ⫾ SD age of the patients was 53.6 ⫾ 13.1 years, and they had a mean ⫾ SD disease duration of 8.8 ⫾ 7.6 years, a mean ⫾ SD Steinbrocker stage of 2.4 ⫾ 1.2, and a mean ⫾ SD global functional status of 1.8 ⫾ 0.6. All of the patients had been treated with methotrexate (mean ⫾ SD dosage 7.3 ⫾ 1.0 mg/week) and/or a combination of the following agents: corticosteroids (10 patients), nonsteroidal antiinflammatory drugs (5 patients), and diseasemodifying antirheumatic drugs (DMARDs) (sulfasalazine in 1 patient, bucillamine in 1 patient, mizoribine in 1 patient, and actarit in 1 patient). None of the patients received other cytotoxic agents or gold preparations. RA disease activity was evaluated with the Disease Activity Score in 28 joints (DAS28) (24). Blood was drawn into heparinized tubes before, 14 days after, and ⬎6 months after the patients received the initial injection of infliximab. PBMCs and monocytes were also derived from blood samples obtained from healthy donors. All experiments were performed after obtaining written informed consent. The study was approved by the local Institutional Review Board. Reagents. Recombinant human TNF␣ was obtained from R&D Systems (Minneapolis, MN). Auranofin was obtained from Wako (Osaka, Japan), and IgG1 was obtained from Serotec (Oxford, UK). Actinomycin D and cycloheximide were obtained from Sigma-Aldrich (St. Louis, MO), and infliximab was kindly provided by Tanabe Seiyaku (Osaka, Japan). Cell preparation and culture. PBMCs were isolated by centrifugation over Ficoll-Hypaque (ICN, Aurora, OH). In some experiments, PBMCs were further fractionated into CD14⫹ and CD14– cells by magnetic-activated cell sorting (Miltenyi Biotec, Gladbach, Germany) using human CD14 MicroBeads (Miltenyi Biotec). Flow cytometric analysis showed that CD14⫹ cells were ⬎95% pure, while the negatively selected population contained ⬍1% CD14⫹ cells. The resultant CD14⫹ cells were quiescent, as previously described (25). These cells were incubated in HEPES-modified RPMI 1640 (Sigma-Aldrich) containing 10% fetal calf serum (Equitech-Bio, Kerrville, TX), 2 mM L-glutamine (SigmaAldrich), 100 units/ml of penicillin plus 100 g/ml of streptomycin (Sigma-Aldrich) at 37°C in an atmosphere of 5% CO2 in air. To determine HO-1 expression at the mRNA and protein levels, cells were cultured in the presence or absence of recombinant human TNF␣ (0.1–10 ng/ml) and/or auranofin (0.1 g/ml) for 6–24 hours. To evaluate the stability of HO-1 mRNA and protein, cells were incubated in the presence of 5 g/ml of actinomycin D and 10 g/ml cycloheximide, respectively. 5.4 1.5 5.7 3.3 1.5 0.1 4.4 0.1 5.2 0.8 0.4 0.1 2.1 0.1 4.7 0.3 0.3 0.1 0.3 0.1 0.8 0.5 0.7 0 2.6 ⫾ 2.3 0.6 ⫾ 1.0‡ 2 weeks 0.2 2.4 0.2 0.9 3.1 0.8 1.0 1.4 0.4 0.4 0.2 0 0.9 ⫾ 1.0‡ ⬎6 months 2 weeks ⬎6 months 0 2 weeks DAS28 score ⬎6 months 0 2 weeks 3.6 4.2 3.4 3.8 6.4 4.7 4.8 3.8 3.4 3.1 3.3 3.5 4.0 ⫾ 0.9‡ G G M G M† N G M N M M G – EULAR ⬎6 months response HO-1:CD14 ratio 52 29 10 6.47 3.9 1.78 1.4 3.6 69 52 37 6.57 6.33 2.65 1.1 0.8 29 19 19 5.66 4.88 3.85 1.0 1.1 64 24 39 5.43 4.31 2.79 0.1 4.1 88 40 ND 7.82 5.35 ND 1.7 3.1 51 37 35 6 5.21 5.12 0.6 1.2 41 15 25 5.9 4.04 2.55 0.5 1.2 68 40 99 8.22 5.97 4.53 0.8 1.0 49 19 45 4.79 3.81 5.32 1.2 1.6 12 2 3 3.43 1.58 2.38 0.7 0.9 40 12 13 4.27 3.57 3.48 0.6 1.8 7 3 5 3.93 2.79 2.2 0.5 0.6 48 ⫾ 24 24.3 ⫾ 15.6‡ 30 ⫾ 27.1§ 5.71 ⫾ 1.46 4.31 ⫾ 1.34‡ 3.33 ⫾ 1.22‡ 0.8 ⫾ 0.5 1.8 ⫾ 1.2‡ 0 ESR, mm/hour * The mean ⫾ SD age of the patients was 53.6 ⫾ 13.1 years. CRP ⫽ C-reactive protein; ESR ⫽ erythrocyte sedimentation rate; DAS28 ⫽ Disease Activity Score in 28 joints; HO-1 ⫽ heme oxygenase 1; EULAR ⫽ European League Against Rheumatism; G ⫽ good response; M ⫽ moderate response; ND ⫽ no data; N ⫽ no response. † Evaluated by the DAS28 using the CRP level. ‡ P ⬍ 0.01 versus pretreatment level, by Wilcoxon’s signed rank test. § P ⬍ 0.05 versus pretreatment level, by Wilcoxon’s signed rank test. 1/45/F 2/73/F 3/59/F 4/68/F 5/53/M 6/55/F 7/37/F 8/50/F 9/71/F 10/53/M 11/50/F 12/29/F Mean ⫾ SD 0 CRP, mg/dl Characteristics of the rheumatoid arthritis study patients* Patient/age/sex Table 1. 466 KIRINO ET AL TNF␣ DOWN-REGULATION OF HO-1 IN HUMAN PERIPHERAL MONOCYTES Immunoblot analysis. The expression of HO-1 protein was determined by immunoblotting as described previously (12). Briefly, cells were treated for 30 minutes on ice with lysis buffer (137 mM NaCl, 20 mM Tris HCl, 50 mM NaF, 1 mM EDTA, and Triton X-100) supplemented with a protease inhibitor (Sigma-Aldrich), and the supernatants were recovered by centrifugation for 30 minutes at 15,000 revolutions per minute. The samples were resolved electrophoretically on a 4–20% gradient of polyacrylamide gel (Daiichi Kagaku, Tokyo, Japan) and transferred onto a polyvinylidene difluoride membrane (Millipore, Billerica, MA). After blocking overnight at 4°C with 5% skim milk–Tris buffered saline, the membrane was incubated for 1 hour at room temperature or overnight at 4°C with optimally diluted anti–HO-1 murine monoclonal antibody (Stressgen, Victoria, British Columbia, Canada), anti–inducible nitric oxide synthase (anti-iNOS) rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), or antiactin goat polyclonal antibody (Santa Cruz Biotechnology), and subsequently for 45 minutes with horseradish peroxidase (HRP)–conjugated anti-mouse secondary antibody (Amersham Life Sciences, Piscataway, NJ), HRPconjugated donkey anti-rabbit IgG (Amersham Life Sciences), or rabbit anti-goat IgG HRP conjugate (Zymed, South San Francisco, CA). The signals were developed with the enhanced chemiluminescence detection system (Amersham Life Sciences). The amount of blotted protein was measured densitometrically by using Scion image analysis software (Scion, Frederick, MD) and an image processing software (National Institutes of Health Image Engineering, Bethesda, MD). Reverse-transcription–polymerase chain reaction (RT-PCR) and real-time PCR. Total RNA was isolated from cells with TRIzol reagent (Invitrogen, Carlsbad, CA) (8,11,12). One microgram of total RNA served as template for singlestrand cDNA synthesis in a reaction using oligo(dT) primers and SuperScript II (Invitrogen). For the PCR, 1 l of cDNA was incubated with 9.375 l of deionized distilled water, 2 l of dNTP, 2.5 l of 10⫻ PCR buffer, and 0.125 l of Taq polymerase (Takara, Otsu, Japan) and a primer pair for HO-1 (sense 5⬘-CAGGCAGAGAATGCTGAG-3⬘ and antisense 5⬘-GCTTCACATAGCGCTGCA-3⬘), CD14 (sense 5⬘-CGGCCGAAGAGTTCACAAGT-3⬘ and antisense 5⬘AGTGCAGTCCTGTGGCTTC-3⬘), GAPDH (sense 5⬘ACAGTCAGCCGCATC-3⬘ and antisense 5⬘-AGGTGCGGCTCCCTA-3⬘), and ␤-actin (sense 5⬘-TCCTGTGGCATCCACGAAACT-3⬘ and antisense 5⬘-GAAGCATTTGCGGTGGACGAT-3⬘). Cycling conditions included 27 cycles of amplification for 30 seconds at 94°C, 30 seconds at 55°C, 1 minute at 72°C, and a final extension phase consisting of 1 cycle of 10 minutes at 72°C. PCR products were run on a 1.5% agarose gel stained with ethidium bromide. The primers and probes for human HO-1, CD14, TNF␣, and GAPDH used in the real-time PCR were purchased from PE Applied Biosystems (Foster City, CA). Realtime PCR was performed using a BD QTaq DNA Polymerase (BD Biosciences Clontech, Mountain View, CA), and the data were analyzed with the ABI Prism 7700 sequence detection system (PE Applied Biosystems, Foster City, CA). Briefly, one-fiftieth of the cDNA derived from 1 g of total RNA, 200 nmoles/liter of probe, and 800 nmoles/liter of primers were 467 incubated in 25 l of total reaction buffer including 0.75 units of the QTaq DNA Polymerase at 50°C for 2 minutes and 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. The analysis system (PE Applied Biosystems) determined the number of cycles at which the amplified DNA in the sample exceeded the threshold (Ct) during the PCR. Gene expression levels in the individual samples were calculated on standard curves of each cDNA generated by serial dilutions of the PCR amplified products. The data on HO-1, CD14, and TNF␣ were standardized to the expression of GAPDH in the same samples, using a multiplex PCR technique. The level of HO-1 mRNA expression in each sample was reported as arbitrary units. The ⌬⌬Ct method was used to semiquantify TNF␣ mRNA levels, according to manufacturer’s protocol (PE Applied Biosystems). Plasmid construction. The human HO-1 cDNA– expressing plasmid pcDNAHO-1 was constructed as previously described (12). We amplified 4.5 kb and 4.0 kb of the HO-1 promoter regions by using KOD plus DNA polymerase (Takara) from human genomic DNA with panels of the following primers: for 4.5 kb sense, 5⬘-TTGGGCTTGTCTTCCTTGCT-3⬘; for 4.0 kb sense, 5⬘-CCTCAGCTTCTCTTTAGGTG-3⬘, and for the common antisense, 5⬘CATCCGGCCGGTGCTGGGCTCGT-3⬘. The PCR products were cloned using pcR-Blunt II-TOPO (Invitrogen) and then subcloned into the pGL3 Basic Vector (Promega, Madison, WI) at the Kpn I and Xho I restriction sites. The resultant constructs were designated as pHO-1(⫺4.5k) and pHO-1 (–4.0k), respectively. We used pSilencer neo (Ambion, Austin, TX) as the siRNA expression vector. As previously described (12), the sequences of human HO-1–specific siRNA were determined according to the AA(N19) rule (where N represents any nucleotide) (26). Two complementary oligonucleotides were synthesized (Takara): 5⬘-GATCCGTGCTGAGTTCATGAGGAACTTCAAGAGAGTTCCTCATGAACTCAGCATTTTTTGGAAA-3⬘ (sense) and 5⬘-AGCTTTTCCAAAAAATGCTGAGTTCATGAGGAACTCTCTTGAAGTTCCTCATGAACTCAGCACG-3⬘ (antisense). These oligonucleotides were annealed followed by ligation into the linearized plasmid at Bam HI and Hind III restriction sites. The pSilencer neo Negative Control (Ambion), encoding a hairpin siRNA whose sequence is not found in humans, was used as the negative control. The HO-1 siRNA expression vector was named psHO-1, and the scrambled siRNA expression vector was named psCont. Transfections. CD14⫹ cells (2 ⫻ 105 cells/well) were transfected for 24 hours with 1 g of plasmid in the presence of FuGene 6 transfection reagent (Roche Diagnostics, Indianapolis, IN) in a 6-well plate (Sumitomo, Tokyo, Japan), according to manufacturer’s protocol. Thereafter, transfected cells were used for experiments. The transfection efficacy of the green fluorescent protein–expressing plasmid pMAX-GFP (Amaxa, Cologne, Germany) to human primary monocytes in the same procedures was evaluated by flow cytometric analysis. The results showed that the efficacy was consistently ⬎40% (data not shown). Reporter gene assay. CD14⫹ cells (2 ⫻ 105 /well) were transfected with 1 g of pHO-1(–4.0k) or pHO-1(–4.5k) in the presence of 50 ng/ml of pRL-CMV (Promega) expressing Renilla luciferase. After 16 hours, cells were divided into 3 468 KIRINO ET AL Figure 1. Endogenous expression of heme oxygenase 1 (HO-1) protein and mRNA in peripheral blood mononuclear cells (PBMCs) from healthy donors. A, Levels of HO-1 and actin protein in freshly isolated PBMCs, CD14⫹, and CD14– cells were examined by immunoblotting (top) and by densitometric scanning (bottom). Densitometric values are the mean and SEM level of HO-1 protein relative to actin in 7 independent experiments. B and C, Expression of mRNA for HO-1, CD14, and ␤-actin in freshly isolated CD14⫹ and CD14– cells was determined by reverse transcription–polymerase chain reaction (RT-PCR) (B) and was estimated semiquantitatively by real-time PCR (C). Semiquantitative values are the mean and SEM expression of mRNA for HO-1 and CD14 relative to GAPDH in 5 healthy donors. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.001, by paired t-test. separate wells of a 24-well plate (Sumitomo) and stimulated with auranofin (1 g/ml) or TNF␣ (10 ng/ml) for another 12 hours. Firefly and Renilla luciferase activities were measured with a TD20/20 luminometer (Turner Designs, Sunnyvale, CA) using a Dual Luciferase Reporter Assay system (Promega). Firefly luciferase activities were standardized against Renilla luciferase activity and were expressed as the relative luciferase activity. Enzyme-linked immunosorbent assay (ELISA). Concentrations of IL-6, IL-8, and TNF␣ in the culture supernatants were determined by ELISA using optimal pairs of capture and detection biotinylated antibodies. Antibodies against IL-6 and TNF␣ were obtained from R&D Systems, and IL-8 was obtained from BD Biosciences. Recombinant human IL-6, IL-8, and TNF␣ (R&D Systems) were used for standards. Statistical analysis. Paired t-test, Wilcoxon’s signed rank test, and Spearman’s rank correlation were used to test for differences. P values less than 0.05 were considered statistically significant. RESULTS Preferential expression of HO-1 in CD14-positive cells in PBMCs from healthy donors. HO-1 expression by PBMCs from normal healthy donors was examined. Freshly isolated PBMCs expressed low, but detectable, amounts of HO-1 protein (Figure 1A). When the cells were fractionated according to CD14 expression, the CD14⫹ population preferentially expressed HO-1 protein and mRNA (Figures 1A–C). Thus, HO-1 was constitutively expressed by circulating CD14⫹ monocytes, but was rarely expressed by lymphocytes from healthy donors. TNF␣ suppression of HO-1 in peripheral monocytes. Proinflammatory cytokines affect HO-1 expression (16–20). When PBMCs were treated with TNF␣, HO-1 protein and mRNA levels decreased in a dosedependent manner (Figures 2A and B). Auranofin induced HO-1 expression in PBMCs, a finding consistent with earlier results involving RA-derived synovial and monocytic cell lines (12,27) (Figure 2A). This auranofin-induced HO-1 expression was also suppressed by TNF␣, whereas the gold compound partly restored the TNF-dependent down-regulation of HO-1 (Figure 2A). Irrespective of the presence or absence of auranofin, the TNF␣-dependent suppression of HO-1 was abrogated by infliximab (Figure 2C). However, infliximab did not alter HO-1 expression in the absence of TNF␣, which indicates that it did not directly modulate HO-1 expression through binding to membrane-type TNF␣ (Figure 2C). Since PBMCs consist of mixed populations, it is uncertain whether TNF␣ directly acts on CD14⫹ monocytes, which preferentially express HO-1. Otherwise, the effect can be mediated by CD14– cells in response to TNF␣. To this end, we examined the effects of TNF␣ on HO-1 in fractionated CD14⫹ and CD14– cells (Figure TNF␣ DOWN-REGULATION OF HO-1 IN HUMAN PERIPHERAL MONOCYTES 469 Figure 2. Effects of recombinant human tumor necrosis factor ␣ (TNF␣) on the expression of heme oxygenase 1 (HO-1) protein and mRNA in peripheral blood mononuclear cells (PBMCs) from healthy donors. A, Effect of TNF␣ on endogenous or auranofin (0.1 g/ml)–induced HO-1 protein expression in PBMCs was examined by immunoblotting (top) and by densitometric scanning (bottom). Densitometric values are the mean and SEM relative change in 7 healthy controls; a value of 1 was assigned to the untreated sample. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01; # ⫽ P ⬍ 0.001; § ⫽ P ⬍ 0.0001, by paired t-test. B, Effect of TNF␣ (1 ng/ml) on HO-1 mRNA expression in PBMCs was determined by reverse transcription– polymerase chain reaction (RT-PCR) (top) and was estimated semiquantitatively by real-time PCR (bottom). Semiquantitative values are the mean and SEM relative change in 9 healthy controls. # ⫽ P ⬍ 0.001, by paired t-test. C, Effect of infliximab (10 g/ml) or isotype-matched IgG1 on TNF␣ (1 ng/ml)–induced suppression of HO-1 in PBMCs. Results are representative of 7 independent experiments. D, Effect of TNF␣ (1 ng/ml) on freshly isolated CD14⫹ and CD14– cells. Results are representative of 7 independent experiments. 2D). HO-1 expression was decreased by treatment with TNF␣ in CD14⫹ cells, whereas HO-1 protein was undetected in CD14– cells (Figure 2D). These data indicate that TNF␣ directly attenuates HO-1 expression at the mRNA and protein levels in CD14⫹ monocytes. No effect of TNF␣ on transcription of the HO-1 gene by human monocytes. Expression of HO-1 protein is regulated by NF-E2–related factor 2 (Nrf2) and other transcription factors, such as NF-B, activator protein 1 (AP-1), AP-2, and hypoxia inducible factor 1␣ (2,19,28–31). To determine whether TNF is involved in HO-1 gene transcription, a luciferase-expressing plasmid containing 4.5 kb of the human HO-1 promoter region, designated pHO-1(⫺4.5k), was used. In addition, a pHO-1(⫺4.0k) plasmid lacking the antioxidant response element where the activator Nrf2 and the BACH-1 repressor are involved in HO-1 gene regulation through competition for interactions with small Maf proteins (32), was used (Figure 3A). Auranofin increased luciferase activity ⬃10-fold in monocytes transfected with pHO-1(⫺4.5k), but not pHO-1 (⫺4.0k), suggesting that the effect of auranofin depends on the antioxidant response element (Figure 3B). In contrast, TNF␣ did not affect the luciferase activity of monocytes transfected with either pHO-1(⫺4.5k) or pHO-1(⫺4.0k) (Figure 3B). The data indicate that TNF␣ suppresses HO-1 mRNA without affecting HO-1 gene transcription in human peripheral monocytes. 470 KIRINO ET AL Figure 3. Effects of tumor necrosis factor ␣ (TNF␣) on gene transcription, mRNA decay, and protein degradation of heme oxygenase 1 (HO-1) in human peripheral monocytes. A, Map of the HO-1 promoter region, showing the positions of ⫺4.5 kb and ⫺4.0 kb from which plasmids were constructed, as well as the antioxidant response element (ARE) and a luciferase (Luc)–coding region. B, Monocytes were transfected with plasmid pHO-1(⫺4.0k) or pHO-1(⫺4.5k) for 16 hours and then stimulated with TNF␣ (10 ng/ml) or auranofin (1 g/ml) for another 12 hours. Values are the mean and SEM relative change in luciferase activity in 5 healthy donors; a value of 10 was assigned to auranofin-treated samples from pHO-1(⫺4.5k)–transfected cells. # ⫽ P ⬍ 0.001; NS ⫽ not significant, by paired t-test. C, Effect of TNF␣ on HO-1 mRNA stability. Peripheral blood mononuclear cells were treated with TNF␣ (1 ng/ml) in the presence of the transcription inhibitor actinomycin D (Act D; 5 g/ml). HO-1 mRNA levels were corrected to those of GAPDH based on real-time polymerase chain reaction results. Values are the mean and SEM relative change in 5 independent experiments; a value of 1 was assigned to the control sample at 0 hours. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, by paired t-test. D, Effect of the TNF␣ on HO-1 protein stability. CD14⫹ cells were treated with TNF␣ (10 ng/ml) in the presence of the translation inhibitor cycloheximide (CHX; 10 g/ml). Shown are the results of HO-1 and actin protein expression (from 1 of 7 experiments) (top) and of densitometric scans (all 7 experiments) (bottom). Densitometric values are the mean and SEM relative change in cycloheximide activity. TNF␣ acceleration of HO-1 mRNA decay in PBMCs. TNF can modulate mRNA stability (33,34). The absence of TNF␣ activity in HO-1 gene transcription suggested that the observed effects involved posttranscriptional regulation. To monitor mRNA degradation, PBMCs were incubated with TNF␣ plus actinomycin D, which blocks gene transcription. Realtime PCR analysis showed that TNF␣ treatment significantly shortened the half-life of HO-1 mRNA (Figure 3C). No effect of TNF␣ on HO-1 protein stability. TNF␣ can also be involved in posttranslational regulation, since TNF␣ down-regulates the inhibitor of DNA binding 1 (ID-1) protein through activation of the ubiquitin/proteasome degradation pathway (35). Moreover, HO-1 degradation is partially suppressed by MG132, an ubiquitin/proteasome inhibitor, indicating that this pathway may affect HO-1 protein stability (36). To examine the effects of TNF␣ at the posttranslational level, HO-1 protein levels in CD14⫹ monocytes were evaluated in the presence of 10 g/ml of cycloheximide, a protein translation inhibitor. TNF␣ had no effect on the degradation of endogenously expressed HO-1 protein (Figure 3D). These results indicate that TNF␣ suppresses HO-1 expression by enhancing mRNA decay, rather than by affecting gene transcription and protein stability. HO-1 regulation of inflammatory responses in human monocytes. Previous studies established that the absence of HO-1 is associated with vigorous inflammatory responses (9,10). Thus, TNF␣-dependent reduction in HO-1 expression may further enhance the development of inflammation. To examine this possibility, pcDNAHO-1 was transfected into CD14⫹ monocytes (Figures 4A and B). These transfected cells secreted significantly lower amounts of TNF␣, IL-6, and IL-8 than did the mock-transfected controls (Figure 4C). TNF␣ DOWN-REGULATION OF HO-1 IN HUMAN PERIPHERAL MONOCYTES Figure 4. Effects of transfecting monocytes with a heme oxygenase 1 (HO-1)–encoding plasmid. A, HO-1 protein levels were monitored in mock-transfected monocytes or in monocytes transfected with plasmid pcDNAHO-1. B, Densitometric analysis of HO-1 expression levels. Values are mean and SEM of 5 independent experiments; a value of 1 was assigned to the mock-transfected cells. ⴱ ⫽ P ⬍ 0.05, by paired t-test. C, Culture supernatants from mock-transfected or pcDNAHO1–transfected monocytes were harvested, and levels of tumor necrosis factor ␣ (TNF␣), interleukin-6 (IL-6), and IL-8 were determined by enzyme-linked immunosorbent assay. Values are the mean and SEM of 5 independent experiments; a value of 1 was assigned to the mock-transfected cells. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, by paired t-test. HO-1 expression by CD14⫹ cells was then reduced using an HO-1–specific siRNA-expressing plasmid (Figures 5A and B). In contrast to the HO-1 transfectants, cytokine production was significantly elevated in psHO-1–transfected cells (Figure 5C). Moreover, much higher amounts of the cytokine and iNOS were produced by psHO-1–transfected cells in the presence of exogenous TNF␣ (Figures 5A and C). These associations indicate an inverse relationship between HO-1 expression and inflammatory responses in monocytes. Since TNF␣ suppresses HO-1 expression, inflammation is augmented. Elevated levels of HO-1 mRNA in PBMCs from patients treated with infliximab. Based on the observation that TNF␣-dependent suppression of HO-1 protein accelerates inflammation, the effect of the TNF␣ antagonist infliximab on HO-1 mRNA expression was monitored in RA patients. All patients had active disease, based on DAS28 scores ⬎3.2 (Table 1). Consistent with previous findings (11), baseline HO-1 mRNA levels in 471 PBMCs from RA patients and normal controls were similar (data not shown). In the RA patients, C-reactive protein (CRP) levels and erythrocyte sedimentation rates (ESRs) fell significantly after infliximab treatment (Table 1). In contrast, although the levels of HO-1 mRNA rose in PBMCs from 8 of the 12 patients, this difference did not achieve statistical significance as compared with pretreatment levels (P ⫽ 0.12) (Figure 6A). Infusion of infliximab has been reported to cause a reduction in the number of monocytes (21,37). This was confirmed in the current study, where the number of monocytes, as determined by morphology, declined significantly after treatment (P ⬍ 0.01) (data not shown). In parallel, levels of CD14 mRNA in PBMCs were also decreased (Figure 6B). Since CD14⫹ monocytes are a major source of HO-1 in PBMCs (Figure 1), levels of HO-1 mRNA were reevaluated by adjusting for the CD14 content. The results indicated that HO-1 mRNA levels adjusted to reflect the frequency of monocytes, which had increased after therapy (P ⬍ 0.05) (data not shown). The HO-1:CD14 mRNA ratio also significantly increased after administration of infliximab (Figure 6C and Table 1). Moreover, the change in the HO-1:CD14 mRNA ratio correlated inversely with the serum CRP level, the ESR, and the DAS28 score (P ⬍ 0.05 by Spearman’s rank correlation) (data not shown). The level of TNF␣ mRNA fell significantly by day 14 (Figure 6D) and was inversely correlated with the HO-1:CD14 mRNA ratio (P ⬍ 0.05) (data not shown). Although the reduction in the number of CD14⫹ cells may partly explain this decline in TNF␣, no significant correlation between CD14 and TNF␣ mRNA levels was observed. Taken together, these data suggest that infliximab restores TNF␣-dependent deficiency of HO-1 in PBMCs, which leads to an attenuation of inflammatory responses, including the synthesis of TNF␣. DISCUSSION The present study demonstrates that TNF␣ down-regulates HO-1 expression by human peripheral blood monocytes and that reduced HO-1 expression facilitates the development of inflammatory responses, including the synthesis of TNF ␣ . Thus, a selfperpetuating cycle linking TNF␣ production and low HO-1 expression may exist. Pharmacologic blockade of this interaction represents a promising strategy for the treatment of inflammatory disorders. Indeed, the clinical efficacy of infliximab, a TNF antagonist, was associ- 472 KIRINO ET AL Figure 5. Effects of transfecting monocytes with the short interfering RNA (siRNA) expression vector psHO-1. A, Levels of heme oxygenase 1 (HO-1) and inducible nitric oxide synthase (iNOS) were determined by immunoblotting of CD14⫹ cells that had been transfected with the siRNA expression vector psHO-1 or with the scrambled siRNA expression vector psCont, with or without tumor necrosis factor ␣ (TNF␣; 10 ng/ml). B, Densitometric analysis of HO-1 expression levels. Values are the mean and SEM of 5 independent experiments; a value of 1 was assigned to the psCont-transfected cells. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, by paired t-test. C, Culture supernatants from psHO-1–transfected or psCont-transfected monocytes were harvested, and levels of TNF␣, interleukin-6 (IL-6), and IL-8 were determined by enzyme-linked immunosorbent assay. Production of IL-6 and IL-8 was also evaluated in the presence of exogenous TNF␣ (10 ng/ml). Values are the mean and SEM of 5 independent experiments; a value of 1 was assigned to the psCont-transfected cells. ⴱ ⫽ P ⬍ 0.05; § ⫽ P ⬍ 0.0001, by paired t-test. ated with increased HO-1 mRNA expression and decreased TNF␣ synthesis in RA patients. Although HO-1 is categorized as an inducible enzyme in response to various stimuli (1), the present study indicates that substantial amounts of this protein are present in freshly isolated peripheral blood monocytes from healthy donors. Since HO-1–deficient cells and individuals are susceptible to inflammatory stimuli and oxidative stresses (9,10), the constitutive expression of HO-1 protein may contribute to the maintenance of homeostasis. The present study focuses on the interaction between TNF␣ and HO-1 expression in circulating monocytes, since TNF␣ plays a pivotal role in the pathogenesis of chronic inflammatory diseases such as RA (15). The HO-1 promoter contains both AP-1 and NF-B binding sites, which can be activated by TNF, suggesting that this cytokine may act as an HO-1 inducer. Indeed, previous studies indicate that HO-1 expression is up-regulated in human endothelial and monocytic cell lines, such as U937 and THP-1, by TNF␣ (17–20), a finding confirmed in preliminary studies in our laboratory (data not shown). Nevertheless, the present study revealed that unlike tumor-derived mono- TNF␣ DOWN-REGULATION OF HO-1 IN HUMAN PERIPHERAL MONOCYTES Figure 6. Changes in heme oxygenase 1 (HO-1) mRNA levels following infliximab treatment of rheumatoid arthritis (RA) patients. The mRNA was extracted from peripheral blood mononuclear cells obtained from 12 RA patients before and 2 weeks after their first infliximab treatment. Levels of mRNA for HO-1, CD14, and GAPDH were determined by real-time polymerase chain reaction. A, HO-1 and B, CD14 mRNA levels in the individual patients were standardized against GAPDH. The ratio of HO-1 mRNA to that of GAPDH was expressed in arbitrary units (AU). C, Ratio of HO-1 to CD14 mRNA in individual patients. D, Relative change in tumor necrosis factor ␣ (TNF␣) mRNA expression. A value of 1 was assigned to the pretreatment levels. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⫽ P ⬍ 0.01, by Wilcoxon’s signed rank test. cytic cell lines (19), TNF␣ had suppressive effects on HO-1 expression in human peripheral monocytes, a finding consistent with that reported for chondrocytes (16). It is important to determine what causes the discrepancies among the cells. Despite the suppressive effect of TNF␣ on HO-1 mRNA expression, the reporter gene assay showed no effect of this cytokine on HO-1 gene transcription by monocytes. Rather, TNF␣ promoted the degradation of HO-1 mRNA. A similar acceleration of mRNA degra- 473 dation by TNF has been reported for ␤5 integrin in osteoclasts, although the molecular detail of the relevant mechanisms remains unclear (34). Since multiple cytoplasmic enzymes and factors are involved in mRNA turnover (38), further study will be needed to identify the mechanism by which TNF␣ accelerates the degradation of HO-1 mRNA. Accumulating evidence shows that HO-1 protein helps to protect the host from pathologic inflammatory reactions (1). Results of the present study indicate that forced expression of HO-1 reduces the production of proinflammatory cytokines by monocytes. Elimination of the HO-1 protein increased inflammatory cytokine production. These findings are consistent with the conclusion that HO-1 has a beneficial effect on inflammatory disorders. HO-1 is present in the joint lesions of patients with RA (12,39) and in the joints of animals with adjuvant-induced or collagen-induced arthritis (40,41). Our previous study established that HO-1 reduced the production of proinflammatory cytokines by RA synovial cell lines (12). HO-1 is also implicated in the bone destruction of RA joints, since HO-1 attenuates inflammation-induced osteoclastogenesis and bone loss in vivo and in vitro (39). HO-1–expressing cells were primarily located in the lining and sublining layer, while few were detected at the cartilage–pannus junction, where TNF␣–producing cells are abundant (39,42). Thus, the localization of HO-1–deficient cells is intimately related to the destruction of bone and cartilage in RA joint lesions (39). These data suggest that HO-1 plays a role in the bone destruction as well as inflammation in RA patients. Previous studies suggest that the antirheumatic effects of gold agents are mediated, at least in part, by the induction of HO-1 (12,27). Yet, it is clear that corticosteroids and other DMARDs, including sulfasalazine and D-penicillamine, do not directly modulate the expression of HO-1 in synovial cells from RA patients (12). However, as shown in Figure 2, the effects of auranofin may be modulated by the level of TNF␣ in the joint. Infliximab is an established therapy for RA (43,44). While its pharmacologic targets are membranebound or soluble TNF␣ molecules, infliximab has unexpectedly wide immunomodulatory activity (45). For example, the clinical efficacy of this agent persists far longer than its half-life. We previously demonstrated that the number of TNF␣-secreting PBMCs decreased in patients with Behçet’s disease receiving infliximab (46). This observation suggested that the TNF agonist 474 KIRINO ET AL not only neutralized soluble TNF␣, but also modulated macrophage function by binding to membraneassociated TNF␣ (45). The present study shows that TNF␣ synthesis decreases when HO-1 expression increases in patients receiving infliximab. However, the monoclonal antibody did not directly effect HO-1 expression in monocytes in the absence of TNF␣. It is likely that infliximab blocked the TNF␣-dependent suppression of HO-1 expression in monocytes, resulting in reduced TNF␣ synthesis. In summary, the present study demonstrates that TNF-dependent suppression of HO-1 expression in monocytes accelerates inflammatory responses and promotes further TNF synthesis. TNF antagonists serve to disrupt this vicious circle, contributing to the remission of inflammation. Additional strategies designed to interfere with the interaction between TNF and HO-1 may also be of benefit in the treatment of inflammatory disorders. ACKNOWLEDGMENTS The authors are greatly indebted to Dr. Dennis M. Klinman (Center for Biologics Evaluations and Research, Food and Drug Administration, Bethesda, MD) for his review and invaluable suggestions in preparing the manuscript. The authors also thank Mr. Tom Kiper (Yokosuka, Japan) for his review of the manuscript. AUTHOR CONTRIBUTIONS Dr. Ishigatsubo had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Dr. Ishigatsubo. Acquisition of data. Drs. Kirino, Takeno, and Murakami. Analysis and interpretation of data. Drs. Kirino, Takeno, Murakami, M. Kobayashi, H. Kobayashi, Miura, Ideguchi, Ohno, Ueda, and Ishigatsubo. Manuscript preparation. Drs. Kirino, Takeno, and Ishigatsubo. Statistical analysis. Drs. Kirino, Takeno, and Ishigatsubo. REFERENCES 1. Otterbein LE, Soares MP, Yamashita K, Bach FH. Heme oxygenase-1: unleashing the protective properties of heme. Trends Immunol 2003;24:449–55. 2. Morse D, Choi AM. Heme oxygenase-1: the “emerging molecule” has arrived. Am J Respir Cell Mol Biol 2002;27:8–16. 3. Inoue S, Suzuki M, Nagashima Y, Suzuki S, Hashiba T, Tsuburai T, et al. Transfer of heme oxygenase 1 cDNA by a replicationdeficient adenovirus enhances interleukin 10 production from alveolar macrophages that attenuates lipopolysaccharide-induced acute lung injury in mice. Hum Gene Ther 2001;12:967–79. 4. Hashiba T, Suzuki M, Nagashima Y, Suzuki S, Inoue S, Tsuburai T, et al. Adenovirus-mediated transfer of heme oxygenase-1 cDNA attenuates severe lung injury induced by the influenza virus in mice. Gene Ther 2001;8:1499–507. 5. Tsuburai T, Suzuki M, Nagashima Y, Suzuki S, Inoue S, Hasiba T, et al. Adenovirus-mediated transfer and overexpression of heme oxygenase 1 cDNA in lung prevents bleomycin-induced pulmonary fibrosis via a Fas-Fas ligand-independent pathway. Hum Gene Ther 2002;13:1945–60. 6. Tsuburai T, Kaneko T, Nagashima Y, Ueda A, Tagawa A, Shinohara T, et al. Pseudomonas aeruginosa-induced neutrophilic lung inflammation is attenuated by adenovirus-mediated transfer of the heme oxygenase 1 cDNA in mice. Hum Gene Ther 2004;15:273–85. 7. Shinohara T, Kaneko T, Nagashima Y, Ueda A, Tagawa A, Ishigatsubo Y. Adenovirus-mediated transfer and overexpression of heme oxygenase 1 cDNA in lungs attenuates elastase-induced pulmonary emphysema in mice. Hum Gene Ther 2005;16:318–27. 8. Takeda Y, Takeno M, Iwasaki M, Kobayashi H, Kirino Y, Ueda A, et al. Chemical induction of HO-1 suppresses lupus nephritis by reducing local iNOS expression and synthesis of anti-dsDNA antibody. Clin Exp Immunol 2004;138:237–44. 9. Poss KD, Tonegawa S. Reduced stress defense in heme oxygenase 1-deficient cells. Proc Natl Acad Sci U S A 1997;94:10925–30. 10. Yachie A, Niida Y, Wada T, Igarashi N, Kaneda H, Toma T, et al. Oxidative stress causes enhanced endothelial cell injury in human heme oxygenase-1 deficiency. J Clin Invest 1999;103:129–35. 11. Kirino Y, Takeno M, Iwasaki M, Ueda A, Ohno S, Shirai A, et al. Increased serum HO-1 in hemophagocytic syndrome and adultonset Still’s disease: use in the differential diagnosis of hyperferritinemia. Arthritis Res Ther 2005;7:R616–24. 12. Kobayashi H, Takeno M, Saito T, Takeda Y, Kirino Y, Noyori K, et al. Regulatory role of heme oxygenase 1 in inflammation of rheumatoid arthritis. Arthritis Rheum 2006;54:1132–42. 13. Lee TS, Chau LY. Heme oxygenase-1 mediates the antiinflammatory effect of interleukin-10 in mice. Nat Med 2002;8: 240–6. 14. Ricchetti GA, Williams LM, Foxwell BM. Heme oxygenase 1 expression induced by IL-10 requires STAT-3 and phosphoinositol-3 kinase and is inhibited by lipopolysaccharide. J Leukoc Biol 2004;76:719–26. 15. Choy EH, Panayi GS. Cytokine pathways and joint inflammation in rheumatoid arthritis. N Engl J Med 2001;344:907–16. 16. Fernandez P, Guillen MI, Gomar F, Alcaraz MJ. Expression of heme oxygenase-1 and regulation by cytokines in human osteoarthritic chondrocytes. Biochem Pharmacol 2003;66:2049–52. 17. Terry CM, Clikeman JA, Hoidal JR, Callahan KS. Effect of tumor necrosis factor-␣ and interleukin-1␣ on heme oxygenase-1 expression in human endothelial cells. Am J Physiol Heart Circ Physiol 1998;274:H883–91. 18. Terry CM, Clikeman JA, Hoidal JR, Callahan KS. TNF-␣ and IL-1␣ induce heme oxygenase-1 via protein kinase C, Ca2⫹, and phospholipase A2 in endothelial cells. Am J Physiol Heart Circ Physiol 1999;276:H1493–501. 19. Muraosa Y, Shibahara S. Identification of a cis-regulatory element and putative trans-acting factors responsible for 12-O-tetradecanoylphorbol-13-acetate (TPA)-mediated induction of heme oxygenase expression in myelomonocytic cell lines. Mol Cell Biol 1993;13:7881–91. 20. Udono-Fujimori R, Takahashi K, Takeda K, Furuyama K, Kaneko K, Takahashi S, et al. Expression of heme oxygenase-1 is repressed by interferon-␥ and induced by hypoxia in human retinal pigment epithelial cells. Eur J Biochem 2004;271:3076–84. 21. Ziolkowska M, Maslinski W. Laboratory changes on anti-tumor necrosis factor treatment in rheumatoid arthritis. Curr Opin Rheumatol 2003;15:267–73. 22. Olsen NJ, Stein CM. New drugs for rheumatoid arthritis. N Engl J Med 2004;350:2167–79. 23. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 TNF␣ DOWN-REGULATION OF HO-1 IN HUMAN PERIPHERAL MONOCYTES 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. Prevoo ML, van ‘t Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified disease activity scores that include twenty-eight-joint counts: development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 1995;38:44–8. Kaufmann A, Musset B, Limberg SH, Renigunta V, Sus R, Dalpke AH, et al. “Host tissue damage” signal ATP promotes nondirectional migration and negatively regulates toll-like receptor signaling in human monocytes. J Biol Chem 2005;280:32459–67. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschl T. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 2001;411:494–8. Kataoka K, Handa H, Nishizawa M. Induction of cellular antioxidative stress genes through heterodimeric transcription factor Nrf2/small Maf by antirheumatic gold(I) compounds. J Biol Chem 2001;276:34074–81. Alam J, Stewart D, Touchard C, Boinapally S, Choi AM, Cook JL. Nrf2, a Cap’n’Collar transcription factor, regulates induction of the heme oxygenase-1 gene. J Biol Chem 1999;274:26071–8. Sun J, Hoshino H, Takaku K, Nakajima O, Muto A, Suzuki H, et al. Hemoprotein Bach1 regulates enhancer availability of heme oxygenase-1 gene. EMBO J 2002;21:5216–24. Lavrovsky Y, Schwartzman ML, Levere RD, Kappas A, Abraham NG. Identification of binding sites for transcription factors NF-B and AP-2 in the promoter region of the human heme oxygenase 1 gene. Proc Natl Acad Sci U S A 1994;91:5987–91. Lee PJ, Jiang BH, Chin BY, Iyer NV, Alam J, Semenza GL, et al. Hypoxia-inducible factor-1 mediates transcriptional activation of the heme oxygenase-1 gene in response to hypoxia. J Biol Chem 1997;272:5375–81. Sun J, Brand M, Zenke Y, Tashiro S, Groudine M, Igarashi K. Heme regulates the dynamic exchange of Bach1 and NF-E2related factors in the Maf transcription factor network. Proc Natl Acad Sci U S A 2004;101:1461–6. Esnault S, Malter JS. Hyaluronic acid or TNF-␣ plus fibronectin triggers granulocyte macrophage-colony-stimulating factor mRNA stabilization in eosinophils yet engages differential intracellular pathways and mRNA binding proteins. J Immunol 2003;171: 6780–7. Inoue M, Ross FP, Erdmann JM, Abu-Amer Y, Wei S, Teitelbaum SL. Tumor necrosis factor ␣ regulates ␣v␤5 integrin expression by osteoclast precursors in vitro and in vivo. Endocrinology 2000;141:284–90. Ling MT, Kwok WK, Fung MK, Xianghong W, Wong YC. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 475 Proteasome mediated degradation of Id-1 is associated with TNF␣-induced apoptosis in prostate cancer cells. Carcinogenesis 2006;27:205–15. Wu WT, Chi KH, Ho FM, Tsao WC, Lin WW. Proteasome inhibitors up-regulate haeme oxygenase-1 gene expression: requirement of p38 MAPK (mitogen-activated protein kinase) activation but not of NF-B (nuclear factor B) inhibition. Biochem J 2004;379(Pt 3):587–93. Lorenz HM, Grunke M, Hieronymus T, Antoni C, Nusslein H, Schaible TF, et al. In vivo blockade of tumor necrosis factor-␣ in patients with rheumatoid arthritis: longterm effects after repeated infusion of chimeric monoclonal antibody cA2. J Rheumatol 2000;27:304–10. Wilusz CJ, Wilusz J. Bringing the role of mRNA decay in the control of gene expression into focus. Trends Genet 2004;20: 491–7. Zwerina J, Tzima S, Hayer S, Redlich K, Hoffmann O, HanslikSchnabel B, et al. Heme oxygenase 1 (HO-1) regulates osteoclastogenesis and bone resorption. FASEB J 2005;19:2011–3. Devesa I, Ferrandiz ML, Guillen I, Cerda JM, Alcaraz MJ. Potential role of heme oxygenase-1 in the progression of rat adjuvant arthritis. Lab Invest 2005;85:34–44. Devesa I, Ferrandiz ML, Terencio MC, Joosten LA, van den Berg WB, Alcaraz MJ. Influence of heme oxygenase 1 modulation on the progression of murine collagen-induced arthritis. Arthritis Rheum 2005;52:3230–8. Chu CQ, Field M, Feldmann M, Maini RN. Localization of tumor necrosis factor ␣ in synovial tissues and at the cartilage–pannus junction in patients with rheumatoid arthritis. Arthritis Rheum 1991;34:1125–32. Lipsky PE, van der Heijde DM, St.Clair EW, Furst DE, Breedveld FC, Kalden JR, et al. Infliximab and methotrexate in the treatment of rheumatoid arthritis: Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. N Engl J Med 2000;343:1594–602. Maini RN, Breedveld FC, Kalden JR, Smolen JS, Davis D, Macfarlane JD, et al. Therapeutic efficacy of multiple intravenous infusions of anti–tumor necrosis factor ␣ monoclonal antibody combined with low-dose weekly methotrexate in rheumatoid arthritis. Arthritis Rheum 1998;41:1552–63. Choo-Kang BS, Hutchison S, Nickdel MB, Bundick RV, Leishman AJ, Brewer JM, et al. TNF-blocking therapies: an alternative mode of action? Trends Immunol 2005;26:518–22. Misumi M, Hagiwara E, Takeno M, Takeda Y, Inoue Y, Tsuji T, et al. Cytokine production profile in patients with Behçet’s disease treated with infliximab. Cytokine 2003;24:210–8.