Inhibition and Genetic Deficiency of p38 MAPK Up-Regulates Heme Oxygenase-1 Gene Expression via Nrf2¹

Heme oxygenase (HO)-1³ is the inducible isoform of the first and rate-controlling enzyme of heme degradation (1). HO-1 is induced by a variety of oxidative stress stimuli and exerts potent antioxidant and anti-inflammatory effects via its products bilirubin and carbon monoxide (CO) (2, 3). The anti-inflammatory functions of HO-1 have been described in HO-1 knockout mice (4, 5) and have essentially been confirmed in a human case of genetic HO-1 deficiency (6). HO-1-deficient mice exhibit a disproportional activation of monocytes and are highly vulnerable to endotoxin-mediated toxicity (4, 7). The potential significance of HO-1 in the adaptive immune system has been implied by a recent report, in which genetic deficiency of HO-1 decreased the suppressive activity of regulatory T cells (8). Importantly, overexpression of HO-1 either by gene transfer or by pharmacological inducers has beneficial effects in various animal models of inflammatory disease (3, 9, 10, 11, 12, 13).

The up-regulation of HO-1 gene expression and that of other genes during inflammation and in response to pharmacological stimuli is mediated by a network of signaling pathways, among which MAPK play a central role (14). Three major subfamilies of MAPK are known: ERK, JNK, and p38. Although ERK mainly mediates cellular responses to hormones and growth factors, JNK and p38 are primarily activated by stress-related stimuli (14). Immunologically active cells such as monocytes and T cells display a high content of p38 MAPK (15) and it has initially been shown that p38 is activated by LPS stimulation of monocytes (16, 17). Subsequently, p38 has been shown to play a critical role for the regulation of proinflammatory genes such as cytokines (18) and cyclooxygenase 2 (Cox-2) (19). Because targeted deletion of the p38α gene in mice causes embryonic lethality (20), p38α appears to be of major physiological significance among the known p38 isoforms. Although numerous studies have demonstrated that activation of p38 mediates HO-1 induction by a host of stimuli (2, 3, 9), an inhibitory role of p38 on HO-1 gene expression has previously been reported (21, 22). Thus, it was the primary goal of the present study to evaluate the role of p38 for HO-1 gene expression in further detail.

In the current report, it is shown that small molecule inhibitors of p38 induced HO-1 expression in monocytic cells and that basal HO-1 gene expression was constitutively high in MEF from p38α^−/− mice. Moreover, p38 inhibition-dependent HO-1 gene induction was regulated via the transcription factor (TF) Nrf2, which is a central regulator of the cellular stress response. Activation of ERK and reactive oxygen species (ROS) were involved in this up-regulation of HO-1 gene expression.

DMEM and RPMI 1640 were obtained from PAA Laboratories, FCS was from Biochrom, Ficoll-Paque was from Pharmacia, CD14⁺ immunomagnetic microbeads were from Miltenyi Biotec, and polyvinylidene difluoride membranes were from Millipore. All other chemicals were purchased from Sigma-Aldrich and Roche Applied Science, unless otherwise indicated.

RAW264.7 cells were from American Type Culture Collection, MEF from p38α^−/− mice were from Dr. A. R. Nebreda (Spanish National Cancer Center, Madrid, Spain) (23) and MEF from Nrf2^−/− mice were from Dr. L. Higgins (University of Dundee, Dundee, U.K.) (24) and were grown in DMEM supplemented with 10% FCS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Human PBMC were isolated and maintained in culture as described previously (25). All cell cultures were kept under air/CO₂ (19:1) at 100% humidity. Treatment of cells with PMA (0.5 μM), LPS (Escherichia coli 0111:B4; 1 μg/ml), inhibitors of p38 MAPK (SB202190 and SB203580; Calbiochem), ERK (PD 98059; Calbiochem), or JNK (SP600125; Calbiochem) was performed with serum-free medium. N-acetylcysteine (NAC) was added to the culture medium 30 min before treatment with PMA as indicated.

p38α^−/− and p38α^+/+ MEF were cultured overnight in complete medium and cytoplasmic and nuclear proteins were extracted using a ProteoJET cytoplasmic and nuclear protein extraction kit (Fermentas). All fractions were brought to equal volumes and protein content was determined for the cytosol and nuclear extract fractions.

Cells were washed with 0.9% NaCl and thereafter lysis and protein quantitation were essentially performed as described previously (26). Fifty micrograms of total protein was separated on a 12% SDS-polyacrylamide gel and was blotted onto polyvinylidene difluoride membranes. Membranes were blocked with TBS containing 5% skim milk or 5% BSA, 50 mM Tris-HCl (pH 7.6), 150 mM NaCl, and 0.1% Tween 20 for 1 h at room temperature. The primary Abs against HO-1 (Stressgen), Cox-2 (Alexis Biochemicals), β-actin (BioZol), and GAPDH (Hytest) were used at 1/1000 dilutions. The primary Abs against Nrf2 and lamin B1 were from Santa Cruz Biotechnology and were used at 1/2000 dilutions. The primary Ab against Bach1 was a gift from Dr. K. Igarashi (Tohoku University School of Medicine, Sendai, Japan) and was used at 1/1000 dilution (27). Primary Abs for detection of phosphorylated and total MAPK were from Cell Signaling and were applied at the concentrations recommended by the manufacturer, respectively. Secondary Abs were goat anti-rabbit IgG HRP and anti-mouse IgG HRP (DPC Biermann) and were used at 1/20,000 and 1/100,000, respectively. The ECL chemiluminescent detection system (Amersham Biosciences) was used for detection according to the manufacturer’s instructions. The chemiluminescent autoradiographic signals were visualized and quantitated with the Fluorchem FC2 gel documentation system.

Two siRNA sequences, GAACGUUGUUUCCUGGUACTT (α1) and GAUGCUCGUUUUGGACUCAG (α2), targeting mouse p38α mRNA were the same as previously published (28) and HPLC purified (MWG Biotech) with 2-nt 3′ terminal dTdT overhangs at each strand. A negative control siRNA was obtained from Dharmacon. For siRNA transfection studies, RAW 264.7 cells were cultured overnight in the absence of penicillin and streptomycin. The cells were transfected with the siRNAs at the indicated concentrations using Lipofectamine RNAiMAX transfection reagent (Invitrogen). The transfection procedure was followed according to the instructions given by the manufacturer. Forth-eight hours after transfection, cells were lysed and subjected to Western blot analysis.

The luciferase reporter gene constructs pHO-4045-luc was a gift from Dr. M. A. Perella (Harvard Medical School, Boston, MA) (29) and pE2-luc was a gift from Dr. J. Alam (Alton Ochsner Medical Center, New Orleans, LA) (30). Plasmid pFA-CHOP with the transactivation domain of the TF CHOP, plasmid pFA-Elk with the transactivation domain of the TF Elk each fused with the DNA-binding domain of yeast Gal4, and the empty control vector pFC2-dbd were purchased from Stratagene. The plasmid pAP-1-luc with three AP-1 repeats in front of a minimal fos promoter was a gift from Dr. C. A. Hauser (The Burnham Institute, La Jolla, CA) (31).

After growth for 24 h, transfection of plasmid DNA into RAW264.7 cells and MEF was performed by using FuGENE (Roche Applied Science) as described previously (25). Unless otherwise indicated, cells were transfected with 0.5–1 μg of the reporter plasmid and in cotransfection experiments with 0.1–1 μg of the indicated expression vectors. Transfection efficiency was controlled using 0.1 μg of Renilla luciferase expression vector pRL-SV40 (Promega) as described previously (25). Cells were lysed with luciferase lysis reagent (Promega) and luciferase activity was determined with the commercial Dual Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. Cells were either harvested 24 h after transfection or treated for another 18 h with the indicated reagents, respectively. Relative light units of firefly luciferase activity were normalized with Renilla luciferase activity.

NE were prepared as described previously (25). The sequences of the biotin-labeled oligonucleotides (MWG Biotech) used for the EMSA are as follows: E2, (5′-GTGCCTTTTCTGCTGAGTCAAGGTCCGGGG-3′); E2 mutant (E2 mut), (5′-GTGCCTTTTCTGCTTATTTAAGGTCCGGGG-3′); and NF-κB consensus oligonucleotide with sequence (5′-AGTTGAGGGGACTTTCCCAGGC-3′) with respective oligonucleotides of the noncoding strand. For competition assays, an excess of unlabeled oligonucleotides was added, as indicated. After preincubation for 10 min at room temperature, the biotin-labeled probe was added and incubation was continued for another 20 min. For immunodepletion analysis, 1–2 μl of Ab either directed against Nrf2 (Santa Cruz Biotechnology) or antisera raised against mouse NB1 protein (from Dr. D. Stroncek, National Institutes of Health, Bethesda, MD) were added to the EMSA reaction as indicated. The reaction mixture was loaded onto a 6% native polyacrylamide gel in 0.5% Tris-borate-EDTA and was blotted onto nylon membranes (Pierce). After UV cross-linking, a LightShift Chemiluminescent EMSA kit (Pierce) was used to detect interaction between biotin end-labeled DNA and protein with a streptavidin-HRP conjugate and the chemiluminescent substrate.

Intracellular generation of ROS was detected with dihydroethidium (DHE) dye (Invitrogen). The fluorescent by-product ethidium, which is produced after DHE cleavage by ROS, was detected with a flow cytometer. Cells were incubated with or without SB202190 (10 μM) for 1 h, after which they were washed and further incubated for 20 min with 5 μM DHE. Cells were then washed with PBS and the levels of ROS were determined with a FACSCalibur flow cytometer (BD Biosciences). DHE-detectable superoxide anion (FL2-H) fluorescent signals are displayed as histograms. Ratios of signals vs control data were calculated by mean fluorescence intensity.

In agreement with earlier work, the phorbol ester PMA and the proinflammatory mediator LPS induced HO-1 (25, 32) and Cox-2 (33) gene expression in cell cultures of RAW264.7 monocytes (Fig. 1,A). Previous reports have shown that p38 MAPK contributes to the PMA-dependent induction of these two genes (25, 32, 33, 34). Although pretreatment of RAW264.7 cells with SB202190, which pharmacologically blocks p38α and p38β MAPK activity, reduced PMA-dependent induction of HO-1 and Cox-2, treatment with this compound alone markedly induced gene expression of HO-1 (Fig. 1,B), but not that of Cox-2 (data not shown). This unexpected observation was confirmed in a second series of experiments, in which HO-1 gene expression in RAW264.7 cells was time-dependently up-regulated by SB202190 and also by a second p38α inhibitor, SB203580 (Fig. 1,C). In addition, treatment with both inhibitors induced HO-1 gene expression in a dose-dependent manner (Fig. 1,D). To further investigate the specificity of p38 inhibition-dependent induction of HO-1 gene expression, we specifically knocked down the expression of p38α in RAW264.7 cells by two different variants of siRNA. As determined by Western blot analysis, both p38α siRNAs knocked down p38α expression and caused a marked induction of HO-1 gene expression after 48 h. Treatment with control siRNA did not have an effect on p38α or HO-1 gene expression (Fig. 1,E). To further specify these results, we investigated PMA-dependent regulation of HO-1 and Cox-2 gene expression in cell cultures of MEF that were genetically deficient for p38α (p38α^−/− MEF) in comparison to that in wild-type MEF (p38α^+/+ MEF). Although PMA induced HO-1 and Cox-2 expression in p38α^+/+ MEF, this compound had only minor effects in p38α^−/− MEF. Importantly, the basal level of HO-1 protein expression was significantly higher in p38α^−/− MEF, if compared with that in p38α^+/+ MEF (Fig. 1,F). In contrast, Cox-2 protein expression was hardly detectable in both p38α^−/− and p38α^+/+ MEF (Fig. 1 F). In addition, HO-1 gene expression was also augmented in human PBMC in the presence of the p38 inhibitor SB203580 (data not shown). Thus, our data indicate that pharmacological inhibition and genetic deficiency of p38α causes an increased HO-1 gene expression.

It is generally accepted that numerous stimuli can activate p38 MAPK (14). Since pharmacological inhibition of p38 in RAW264.7 cells has an effect on HO-1 gene expression, which is similar to that in p38^−/− MEF (Fig. 1), we examined p38 activity in untreated RAW264.7 cells. Interestingly, p38 phosphorylation could be detected under basal conditions and exposure to SB202190 and SB203580 markedly attenuated this phosphorylation (Fig. 2,A). In addition, we verified the functionality of p38 under basal conditions and also the effect of SB202190 and SB203580 on its functionality in RAW264.7 cells. A luciferase reporter construct with five binding sites for the yeast transcription factor Gal4 was cotransfected with a construct, allowing expression of a fusion protein consisting of the Gal4 DNA-binding domain and the transactivation domain of the TF CHOP (pFA-CHOP). Transactivation of CHOP is regulated by p38-dependent phosphorylation of regulatory serine residues within the CHOP transactivation domain (35). Similar to the data on p38 phosphorylation in Fig. 2,A, high luciferase activity of pFA-CHOP was observed under basal conditions and exposure of RAW264.7 cells to SB202190 and SB203580 reduced reporter gene activity (Fig. 2,B). When p38^−/− and p38^+/+ MEF were transfected with pFA-CHOP, luciferase activity was low in p38^−/− MEF, whereas it was significantly higher in p38^+/+ cells. In this study, reporter gene activity was inhibited by SB202190 and SB203580 (Fig. 2 C). The data demonstrate that the basal activity of p38 MAPK is high in RAW264.7 monocytic cells and was inhibited by the treatment with SB202190 and SB203580.

To investigate whether inhibition of MAPK other than p38 would affect HO-1 gene expression, RAW264.7 cells were treated with specific inhibitors of the JNK (SP600125) and ERK (PD98059) pathways, respectively. For a comparison, cells were also treated with PMA. In contrast to p38 inhibitors, treatment with these inhibitors did not induce HO-1 gene expression (Fig. 3), indicating that inhibition of the JNK and ERK pathways did not affect basal HO-1 gene expression in RAW264.7 cells.

Nrf2, a member of the cap’n’collar family of basic leucine zipper proteins (36, 37), plays a major regulatory role for the stress-dependent transcriptional induction of HO-1 gene expression (38, 39). Although p38 has been shown to be involved in Nrf2 activation, the exact regulatory role of p38 for Nrf2 activation is not well understood (40). To further investigate this pathway, we determined the regulation of luciferase reporter gene constructs pHO-4045-luc and pE2-luc, both of which carry the proximal (E1) and distal (E2) enhancer sequences of the mouse HO-1 promoter (38). Both enhancer sequences contain multiple copies of stress response element (StRE) sites that are targeted by Nrf2 (3). For a comparison, we also transfected p38^−/− and p38^+/+ MEF with the pAP-1-luc construct, which carries three copies of the consensus recognition sequence of AP-1 (31). Basal luciferase activity of the reporter gene constructs pHO-4045-luc and pE2-luc, but not that of pAP-1-luc, was markedly higher in p38^−/− MEF compared with that in p38^+/+ MEF (Fig. 4,A). Furthermore, treatment with SB202190 up-regulated reporter gene activity of these constructs in RAW264.7 cells (Fig. 4,B). Luciferase activity of the pE2-luc construct was slightly higher in comparison to that of the pHO-4045-luc in p38^−/− MEF and in RAW264.7 cells (Fig. 4). Taken together, the data suggest that an Nrf2-dependent transcriptional mechanism is involved in HO-1 gene activation via inhibition of p38.

Nrf2 regulates HO-1 gene expression via shuttling between the nucleus and cytoplasm in response to various stress stimuli (36, 37). To further investigate whether p38 may regulate Nrf2 localization, nuclear and cytoplasmic Nrf2 levels were determined in p38^−/− and p38^+/+ MEF. Nrf2 levels were higher in nuclear fractions of p38^−/− MEF compared with those in p38^+/+ MEF and Nrf2 was hardly detectable in the cytoplasmic fractions (Fig. 5). However, only minor differences were observed for total Nrf2 levels between p38^−/− MEF and p38^+/+ MEF. These findings may suggest that genetic deficiency of p38 may affect nuclear translocation of Nrf2, but may only have a minor effect on total Nrf2 levels. For a comparison, we also determined protein levels of Bach1, which is a counterregulator of Nrf2 and represses HO-1 gene expression (27). Remarkably, cytosolic and nuclear levels of Bach1 were inverse to those of Nrf2, with higher Bach1 levels in cytoplasmic fractions of p38^−/− MEF compared with those in p38^+/+ MEF (data not shown).

To further examine the role of p38 on Nrf2-dependent HO-1 gene regulation, Nrf2^−/− and Nrf2^+/+ MEF were treated with small molecule inhibitors of p38 and PMA. Treatment with SB202190 and SB203580 significantly induced HO-1 gene expression in Nrf2^+/+ MEF, but not in Nrf2^−/− MEF (Fig. 6,A) and luciferase activity of transfected pHO-4045-luc and pE2-luc constructs was markedly higher in Nrf2^+/+ MEF compared with that in Nrf2^−/− MEF (Fig. 6,B). Furthermore, treatment with SB202190 up-regulated luciferase activity of two HO-1 gene constructs to a magnitude similar to that by PMA in Nrf2^+/+ MEF, but not in Nrf2^−/− MEF (Fig. 6,B). In contrast, reporter gene activity of the pAP-1-luc construct was not affected by p38 inhibitors (Fig. 6 B). Taken together, the data suggest that inhibition of p38 induces HO-1 gene expression in a Nrf2-dependent manner.

We also examined the binding activity of nuclear proteins from RAW264.7 cells to a biotin-labeled E2 oligonucleotide with a StRE site of the mouse HO-1 gene promoter (38) in EMSA studies. After treatment with p38 inhibitors, NE from RAW264.7 cells formed an inducible DNA-protein complex, which was similar to that observed in response to treatment with LPS (Fig. 7,A). The intensity of the DNA-protein complex from SB202190-treated cells was decreased by an unlabeled E2 oligonucleotide, but not by an oligonucleotide with a mutation of the StRE site (E2 mut, Fig. 7,B) or a NF-κB oligonucleotide (Fig. 7,C). Incubation of the DNA-protein-binding reaction with an Ab against Nrf2 abolished the formation of a DNA-protein complex. Likewise, NE from Nrf2^−/− MEF treated with SB202190 failed to form a DNA-protein complex (Fig. 7 C). Taken together, the data indicate that the StRE site of the E2 region is a nuclear binding site for Nrf2 and may be involved in p38 regulation of HO-1 gene expression.

ERK is involved in HO-1 gene induction by numerous stimuli (3, 9). Since p38 MAPK has previously been shown to have a negative regulatory effect on ERK activity (23), we investigated the effect of p38 MAPK inhibition on ERK phosphorylation and its potential effect on HO-1 gene expression in p38^−/− and p38^+/+ MEF. As demonstrated in Fig. 8,A, ERK was strongly phosphorylated in p38^−/− MEF, but not in p38^+/+ MEF. Serum starvation did not affect ERK phosphorylation in either cell type. Treatment with a specific inhibitor of ERK attenuated the increased levels of HO-1 in p38^−/− MEF, suggesting a regulatory cross-talk of these protein kinases for HO-1 gene regulation (Fig. 8,B). Subsequently, we determined the effect of p38 inhibition on ERK phosphorylation in RAW264.7 cells. Treatment with p38 inhibitors induced ERK phosphorylation in these cells similar to that by the known ERK activator PMA (Fig. 8,C). Finally, the reporter gene construct pFA-Elk, which is specifically regulated via ERK, was markedly induced by p38 inhibitors in RAW264.7 cells. As a control, treatment with PMA strongly induced luciferase activity of pFA-Elk (Fig. 8 D). Taken together, the data suggest that genetic deficiency or pharmacological inhibition of p38 activates ERK and may be involved in the induction of HO-1 gene expression.

Genetic deficiency of p38α MAPK has previously been shown to lead to an increased accumulation of ROS (41). Accordingly, p38^−/− MEF exhibited a 2-fold increase of intracellular ROS levels in comparison to p38^+/+ MEF (Fig. 9, A and B). Moreover, treatment of p38^+/+ MEF with SB202190 augmented intracellular levels of ROS (Fig. 9, A and B) and the generation of ROS was also enhanced in SB202190-treated RAW264.7 cells (data not shown). ROS have previously been demonstrated to up-regulate HO-1 gene expression in immunologically active cells and have been proposed to function as putative second messengers for the induction of this gene (2, 9, 25, 42). To investigate whether ROS would directly affect HO-1 gene expression, we treated p38^+/+ and p38^−/− MEF with the antioxidant NAC. NAC reduced basal HO-1 expression in p38^−/− MEF, but had no effect in p38^+/+ MEF (Fig. 9,C). Similarly, treatment with NAC markedly attenuated HO-1 gene expression in RAW264.7 cells. In line, induction of HO-1 expression by PMA, a known inducer of ROS production in monocytes (43), was abrogated by NAC (Fig. 9 D). These data suggest that genetic deficiency and pharmacological inhibition of p38 increases the intracellular generation of ROS, which are involved in the up-regulation of HO-1 gene expression.

The enzyme HO-1, which catalyzes the first and rate-limiting step of heme degradation, has attracted major attention in recent years, because its products CO and biliverdin have potent antioxidant and anti-inflammatory functions. Activation of the p38 MAPK pathway has previously been shown to be involved in HO-1 induction by a wide variety of stimuli. In the present study, it is demonstrated that: 1) pharmacological inhibition of p38 in monocytic cells and genetic deficiency of p38α in MEF cause an up-regulation of HO-1 gene expression, 2) the increased HO-1 expression by inhibition of p38 is mediated via the TF Nrf2, and 3) activation of ERK and ROS are involved in this HO-1 gene regulation.

Inhibition of p38 MAPK by small molecule inhibitors in monocytic cells and genetic deficiency of p38α in MEF led to increased gene expression of HO-1, but not that of Cox-2 (Fig. 1). It is remarkable that HO-1 levels were constitutively high in p38^−/− MEF, because basal HO-1 expression is known to be barely detectable in most cells and tissues under nonstimulated conditions (2, 3). The present findings are in keeping with a recent report demonstrating that basal expression of several genes was increased without external stress stimuli in immortalized p38α^−/− cardiomyocytes (44). Up-regulation of HO-1 gene expression by inhibition of p38, however, was unexpected, because this protein kinase is considered to play a major role for mediating HO-1 induction by multiple stress stimuli (2, 3, 9). Thus, the present observations suggest that p38α may have a dual function for HO-1 gene regulation.

p38 MAPK modulates gene expression via transcriptional and posttranscriptional mechanisms (45). The present study demonstrates that HO-1 gene expression by inhibition or genetic deficiency of p38 occurs on the transcriptional level via the TF Nrf2 (Figs. 4–7), which is a central regulator of the cellular oxidative stress response (36, 37) and plays a major role for the induction of HO-1 (3, 38, 39). Nrf2 targets to StRE, several of which are localized in the 5′ flanking sequence of the HO-1 gene promoter (3). StRE, also referred to as antioxidant response elements or electrophile response elements, have been identified in a large number of phase II detoxifying enzymes and antioxidant stress proteins (40). An inhibitory role of p38 on Nrf2-mediated gene regulation is also supported by the finding that expression of peroxiredoxin 1, which is coordinately regulated with HO-1 (26) via an Nrf2-dependent mechanism (39), was induced by inhibition of p38 (data not shown). Similar to the findings of our present report, chemical blockage of p38 has previously been shown to enhance StRE-mediated induction of phase II-detoxifying enzymes (21) and the sulforaphane-dependent HO-1 activation via Nrf2 in human hepatoma HepG2 cells (22). Our observations also implicate that Bach1, which is a counterregulator of Nrf2 and a repressor of HO-1 gene expression (27), may be involved in the transcriptional regulation of HO-1 via p38 MAPK (Fig. 5). Finally, we cannot exclude the possibility that Nrf2-independent regulatory mechanisms are involved in the up-regulation of HO-1 gene expression via p38 MAPK.

In accordance with an earlier report (23), the present study demonstrates that genetic deficiency of p38α caused an increased ERK activity (Fig. 8). Since HO-1 gene expression in p38^−/− MEF was blocked by the ERK inhibitor PD98059, it is conceivable that the MAPK p38 and ERK have opposite regulatory roles for governing HO-1 gene expression. This assumption would correspond with a previous study, in which pharmacological inhibition of p38 induced low-density lipoprotein receptor expression via activation of ERK in HepG2 cells (46). Independently, others have demonstrated that blockage of p38 increased LPS-dependent inducible NO synthase gene expression and NO production via activation of JNK in macrophages (47). Considerable variations have been reported for the regulatory role of various MAPK on HO-1 gene expression. Elbirt et al. (48) have reported that activation of ERK and p38 MAPK were required for the arsenite-dependent induction of HO-1 gene expression in avian hepatocytes. By contrast, both JNK and p38 MAPK were necessary for HO-1 induction by this compound in rat hepatocytes (49). Thus, a complex network of MAPK appears to modulate HO-1 gene expression in a stimulus-, cell-, and species-specific manner.

p38 could be involved in the sensing of cellular oxidative stress, because transformed p38α-deficient cells have been reported to accumulate much larger amounts of ROS than wild- type cells (41). This observation corresponds with our present findings (Fig. 9) and ROS are likely mediating ERK activation in cells lacking p38α (23). The latter assumption would be supported by recent reports, which demonstrated that ROS mediate a sustained activity of ERK in human hepatoma cells (50) and in a model of cerebral ischemia (51). In other studies it has been shown that ROS-dependent ERK induction regulates Nrf2 activity and subsequent regulation of StRE-dependent gene expression in different experimental models (52, 53). Since ROS appear to be involved in the p38-dependent increase of HO-1 gene expression (Fig. 9), it is conceivable that ERK-mediated Nrf2 activation may cause HO-1 gene induction in a ROS-dependent manner. Alternatively, ROS are assumed to modify interactions between the cytosolic inhibitor of Nrf2, Kelch-like ECH-associated protein 1, and Nrf2, allowing stabilization and nuclear translocation of Nrf2 (54). Therefore, it is plausible that the accumulation of ROS which is caused by inhibition of p38 may have a direct effect on nuclear localization of Nrf2 and induction of HO-1 gene expression. The intimate connections among ROS, ERK, and Nrf2, however, are currently under intense investigation.

Since HO-1 has potent antiapoptotic functions (3), the increase of HO-1 expression levels in p38α^−/− MEF corresponds with the observation that p38α deficiency makes these cells less sensitive to apoptosis via up-regulation of antiapoptotic genes (23). It is also important to note that antiapoptotic effects of HO-1 have recently been associated with the degradation of p38α in endothelial cells (55). Taken together, these findings may suggest an intimate interaction of HO-1 and p38α regulation. HO-1 has attracted major attention in recent years, because it has antioxidant cytoprotective and anti-inflammatory and immunomodulatory functions (3, 9, 10, 11, 12, 13). In particular, the anti-inflammatory role of HO-1 has initially been shown in HO-1-deficient mice that exhibited a proinflammatory phenotype and were highly sensitive to LPS-mediated toxicity. Remarkably, these observations in HO-1^−/− mice have essentially been confirmed in a human case of HO-1 genetic deficiency (6). More recently, it has been demonstrated that loss of HO-1 activity in APC is associated with impaired suppressor function of regulatory T cells (8).

In summary, the present study demonstrates that the HO-1 gene is activated via pharmacological inhibition and genetic deficiency of p38 MAPK. These findings not only give new insights into the regulatory mechanisms of HO-1 gene expression, but also show the complex regulatory network that modulates the inflammatory response in monocytic cells. These findings may ultimately help to develop novel therapeutic approaches for the treatment of inflammatory diseases.

We thank Karnati Srikanth and Dr. Barbara Ahlemeyer for guidance in subcellular fractionation and the DHE assay and Silke Werth and Annette Zeyer for excellent technical assistance. We also thank Dr. Angel Nebreda for the supply of p38^−/− MEF, Dr. Larry Higgins for the supply of Nrf2^−/− MEF, Dr. Kazuhiko Igarashi for the supply of Bach1 Ab, and Dr. Mark A. Perella, Dr. Jawed Alam, and Dr. Craig Hauser for the supply of plasmids.

The authors have no financial conflict of interest.