Enterohemorrhagic Escherichia coli (EHEC) are the causative agent of hemolytic-uremic syndrome. In the first stage of the infection, EHEC interact with human enterocytes to modulate the innate immune response. Inducible NO synthase (iNOS)-derived NO is a critical mediator of the inflammatory response of the infected intestinal mucosa. We therefore aimed to analyze the role of EHEC on iNOS induction in human epithelial cell lines. In this regard, we show that EHEC down-regulate IFN-γ-induced iNOS mRNA expression and NO production in Hct-8, Caco-2, and T84 cells. This inhibitory effect occurs through the decrease of STAT-1 activation. In parallel, we demonstrate that EHEC stimulate the rapid inducible expression of the gene hmox-1 that encodes for the enzyme heme oxygenase-1 (HO-1). Knock-down of hmox-1 gene expression by small interfering RNA or the blockade of HO-1 activity by zinc protoporphyrin IX abrogated the EHEC-dependent inhibition of STAT-1 activation and iNOS mRNA expression in activated human enterocytes. These results highlight a new strategy elaborated by EHEC to control the host innate immune response.

Infection with enterohemorrhagic Escherichia coli (EHEC)4 may lead to diseases ranging from watery diarrhea to life-threatening complications such as hemolytic-uremic syndrome, the main cause of acute renal failure in children worldwide. Human infection occurs by ingestion of contaminated food derived from rearing animals, including meat, milk, or soiled vegetables (1). EHEC reach and colonize the lower intestine, initiating the formation of attaching and effacing lesions to the epithelium and the development of a mucosal innate immune response. Clinical data have established that type 1 cytokines and chemokines are over-expressed in EHEC-infected patients (2, 3). More particularly, an elevated ratio between pro- and anti-inflammatory cytokines is a reliable marker of the severity of EHEC infection and development of hemolytic-uremic syndrome (4, 5). Therefore, the capacity of bacteria to modulate the enterocyte innate immune response may be a critical step in pathogenesis.

Enterocytes are the first cells in contact with EHEC. This interaction yields to the locus of enterocyte effacement (LEE)-dependent structural modifications of the epithelium, i.e., attaching/effacing lesions, and to the development of a nonspecific defense program. In vitro studies have highlighted that EHEC induce the synthesis of chemokines and cytokines in human epithelial cell lines (6, 7), but may also subvert the innate immune response by inhibiting various signal transduction pathways and suppressing the activation of transcription factors such as NF-κB or STAT-1 (8, 9, 10). Other proinflammatory mediators induced by enteropathogenic bacteria may play a critical role in the pathophysiology of enteritis and related diseases. Among them, NO is a potent effector of the innate immune system, which plays a major role in the development of inflammation (11) and in the fight against pathogens (12). Under pathophysiologic conditions, NO is synthesized in large amounts by the inducible NO synthase (iNOS). The transcription of the gene encoding iNOS can be induced by bacteria (12, 13) or by proinflammatory cytokines (13) and requires the activation of diverse transcription factors including NF-κB or STAT-1 (14). We recently established that chemical and cellular sources of NO suppress the EHEC production of Shiga-toxin (Stx), the main virulence factor responsible for hemorrhagic symptoms, without affecting the growth and the survival of the bacteria (15). However, in the context of EHEC/NO interaction study, the effect of these bacteria on iNOS induction has not yet been investigated.

The aim of this study was to determine whether EHEC modulate iNOS expression and activity in human epithelial cell lines, stimulated or not by proinflammatory cytokines. We found that EHEC suppress the STAT-1-dependent transcription of the gene encoding iNOS in activated cells. This inhibition requires the activation of the enzyme heme oxygenase (HO)-1 by EHEC.

The NF-κB inhibitors, namely Z-Leu-Leu-Leu-CHO (20 μM) and (E)3-[(4-methylphenyl)sulfonyl]-2-propenenitrile (Bay 11–7082; 20 μM), and the JAK-2 inhibitor 1,2,3,4,5,6-hexabromocyclohexane (20 μM) were purchased from Calbiochem. The STAT-1 inhibitor (2-fluoroadenine-9-b-D-arabinofuranoside also named fludarabine; 20 μM), the HO-1 inhibitor zinc protoporphyrin IX (ZnPP; 10 μM), bilirubin, and the carbon monoxide (CO) donor tricarbonyldichlororuthenium-(II)-dimer (CORM-2) were obtained from Sigma-Aldrich.

Bacteria used throughout this study are listed in Table I. Six O157:H7 and O113:H21 EHEC strains were isolated from patients with hemorrhagic colitis or hemolytic and uremic syndrome; these strains were selected because they express the major virulence factors usually carried by the majority of the bacteria belonging to both serotypes. The eae, espA, espD, sepL, and stx1/stx2 isogenic mutants in the strain EDL933 were used and have been previously described (10, 16, 17, 18, 19). Commensal E. coli (isolates EcG1 and EcG2) were obtained as follows: Feces from healthy human were 104-fold-diluted and plated overnight at 37°C on sorbitol MacConkey agar plates; purity of isolated colonies was monitored by direct microscopic observation after Gram staining, and phylum was determined by API gallery and 16S rRNA gene sequencing as previously described (20). We determined by PCR that the strains EcG1 and EcG2 do not possess the genes stx2, espA, and eae (data not shown).

Table I.

List of bacteria

StrainOriginaSerotypeLEEstx1/stx2Source
EDL933 HC/O O157:H7 +/+ ATCC # 700927 
Sakaï HUS/O O157:H7 +/+ STEC Centerb 
8624 HC/O O157:H7 −/+ STEC Centerb 
CL3 HUS O113:H21 − −/+ STEC Centerb 
CL15 HUS O113:H21 − −/+ STEC Centerb 
87307 HUS O113:H21 − −/+ STEC Centerb 
EDL933 Δeae NA O157:H7 +/+ Ref. 19  
EDL933 ΔespA NA O157:H7 +/+ Ref. 16  
EDL933 ΔespD NA O157:H7 +/+ Ref. 17  
EDL933 ΔsepL NA O157:H7 +/+ Ref. 18  
EDL933 Δstx1/stx2 NA O157:H7 −/− Ref. 10  
DH5α ND ND − −/− Stratagene 
EcG1 ND − −/− This work 
EcG2 ND − −/− This work 
StrainOriginaSerotypeLEEstx1/stx2Source
EDL933 HC/O O157:H7 +/+ ATCC # 700927 
Sakaï HUS/O O157:H7 +/+ STEC Centerb 
8624 HC/O O157:H7 −/+ STEC Centerb 
CL3 HUS O113:H21 − −/+ STEC Centerb 
CL15 HUS O113:H21 − −/+ STEC Centerb 
87307 HUS O113:H21 − −/+ STEC Centerb 
EDL933 Δeae NA O157:H7 +/+ Ref. 19  
EDL933 ΔespA NA O157:H7 +/+ Ref. 16  
EDL933 ΔespD NA O157:H7 +/+ Ref. 17  
EDL933 ΔsepL NA O157:H7 +/+ Ref. 18  
EDL933 Δstx1/stx2 NA O157:H7 −/− Ref. 10  
DH5α ND ND − −/− Stratagene 
EcG1 ND − −/− This work 
EcG2 ND − −/− This work 
a

HC, hemorrhagic colitis; HUS, hemolytic-uremic syndrome; O, outbreak; NA, not applicable; ND, not determined; F, healthy human feces.

b

National Food Safety and Toxicology Center, Michigan State University, East Lansing, MI; STEC center: http://shigatox.net.

Bacteria were isolated on Luria-Bertani agar plates. One clone was grown overnight at 37°C in Luria-Bertani broth then diluted and grown in cell culture medium for 2 h to the exponential growth phase (A600 ∼ 0.2–0.4). These bacteria were used to infect epithelial cells.

The human colonic epithelial cell lines Hct-8 and Caco-2 were maintained in DMEM, 10% FBS, 1% sodium pyruvate, 10 mM HEPES, 100 U/ml penicillin, and 100 μg/ml streptomycin. T84 cells were cultured in supplemented DMEM/Ham’s F12 (1/1). Cells (106/ml) were cultured for 7–10 days at 37°C under 5% CO2. Confluent cells were stimulated with 50 ng/ml IFN-γ, 20 ng/ml TNF-α, and/or 5 ng/ml IL-1β (each obtained from Pierce), in the presence or absence of CORM-2, bilirubin, or bacteria at a multiplicity of infection of 1–100 in complete medium without antibiotics; cytokines and bacteria were added at the same time. Pharmacological inhibitors were added to cell 30 min before infection. Following infection, cocultures were washed and 1) RNA or proteins were extracted from fixed cells, or 2) a fresh complete medium containing 2 mM l-arginine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 50 μg/ml gentamicin was added to determine NO production after 24 h.

Cells cultured in 6-well plates were transfected overnight with 200 nM small interfering RNA (siRNA) for human HO-1 (5′-AAGCCACACAGCACUAUGUAAdTdT-3′ and 5′-UUACAUAGUGCUGUGUGGCUUdTdT-3′; Ref. 21) or lamin as control (Dharmacon), using siLentFect Lipid Reagent (Bio-Rad). After 18 h, cells were washed and stimulated with cytokines and/or infected with EHEC.

Total RNA of epithelial cells was isolated using TRIzol (Molecular Research Center) and was reverse-transcribed (2 μg) using oligo(dT) primers and 5 U/μl Superscript II reverse transcriptase (Invitrogen). PCR was conducted using 1 μl of cDNA, 0.25 U Platinum TaqDNA polymerase (Invitrogen), 0.12 pmol/μl each of 5′ and 3′ human iNOS (5′-ACCTCAGCAAGCAGCAGAAT-3′ and 5′-ATCTGGAGGGGTAGGCTTGT-3′, 269 bp) or hmox-1 (the gene encoding HO-1) primers (5′-ACATCTATGTGGCCCTGGAG-3′ and 5′-GGCAATCTTTTTGAGCACCT-3′, 287 bp), and 0.03 pmol/μl each of β-actin primers (436 bp; 10). One PCR cycle consisted of the following: 94°C for 30 s, 58°C for 30 s, and 72°C for 1.5 min; the total cycle number was 30. Real-time PCR was performed using the same primers and the SYBR Premix Ex Taq kit (Takara); reactions were conducted in the Mastercycler ep Realplex Silver apparatus (Eppendorf). One PCR cycle consisted of the following: 94°C for 30 s, 58°C for 30 s, and 72°C for 45 s. Results were calculated using the comparative cycle threshold method and are expressed as relative mRNA expression compared with uninfected cells.

Nuclear proteins of cultured epithelial cells were purified using the NE-PER Nuclear and Cytoplasmic Extract kit (Pierce) and protein concentration was measured with the Protein Assay kit (Bio-Rad). NF-κB and STAT-1 DNA-binding activities were analyzed as previously described (10) using 2 μg proteins. The double-stranded oligonucleotide probes (40 fmoles) containing the NF-κB (5′-TCCAAGGGGACTTTCCATG-3′) and the STAT-1 (5′-GACATTTCCCGTAAATC-3′) binding consensus sequence (italicized letters) were used.

Immunocytochemistry and confocal microscopy were performed as previously described (10). The anti-NF-κB p65 rabbit polyclonal Ab (1/500; Calbiochem), the anti-STAT-1 rabbit polyclonal Ab (1/100; Epitomics), and the goat anti-rabbit IgG DyLight™ 647 conjugated (1/1000; Pierce) were used.

Immunoblots were performed as previously described using a rabbit anti-human HO-1 Ab (StressGen; Ref. 21). Equal transfer of proteins was confirmed by staining the nitrocellulose membrane with Ponceau Red. Proteins from chondrocytes of Coprotoporphyrin-treated mice were used as positive control (21).

In culture supernatants, the concentrations of the stable oxidized products of NO, NO3 and NO2, were determined using the Nitrite/Nitrate Assay kit (Sigma-Aldrich), followed by spectrophotometric measure of absorbency at 540 nm.

Quantitative data are presented as the mean ± SEM. Student’s t test was used to determine significant differences between two groups. ANOVA with the Student-Newman-Keuls test was used to analyze significant differences among multiple groups; p = 0.05 was considered significant.

Hct-8 cells were infected with different strains of EHEC in the presence or absence of IFN-γ. The six EHEC strains tested, the commensal E. coli isolates, or E. coli DH5α were not able to induce iNOS mRNA expression in the epithelial cell lines after 6 h of coculture (Fig. 1,A). We then sought to determine the effect of bacteria on activated cells. As shown in Fig. 1,B, iNOS mRNA expression was increased in Hct-8 cells treated 6 h with IFN-γ. When IFN-γ and EHEC isolates belonging to O157:H7 or O113:H21 serotypes were added together with the cells, the levels of iNOS transcripts were dramatically decreased in comparison to cells treated with IFN-γ; the inhibitory effect was observed with a multiplicity of infection of 10 or 100, but not of 1 (data not shown). However, iNOS mRNA expression in activated epithelial cells was not inhibited by the commensal E. coli isolates or by the strain DH5α (Fig. 1,B). These results were confirmed by real-time PCR (Fig. 1,C). The expression of the iNOS gene was up-regulated by ∼20-fold under IFN-γ stimulation when compared with control cells. This increase was significantly inhibited by 97–105% by the use of EHEC, according to the strain tested. Similarly, IFN-γ-induced NO synthesis in Hct-8 cells was completely inhibited using the six EHEC strains (Fig. 1,D); NO generation was not inhibited by commensal E. coli isolates or by the strain DH5α (Fig. 1 D).

FIGURE 1.

Effect of EHEC on iNOS induction in human enterocytes. Hct-8 cells were infected with bacteria for 6 h. RNA was purified, reverse transcribed, and the expression level of the gene encoding iNOS was analyzed by RT-PCR (A). Cells treated with IFN-γ were used as positive control for iNOS mRNA expression. Cells were treated with IFN-γ in the presence or absence of bacteria. Subsequently, iNOS gene transcription was analyzed by RT-PCR (B) and real-time PCR (C). The 6-h stimulated cells were washed and incubated for 18 h in complete medium containing antibiotics. [NO2] plus [NO3] were then measured in culture supernatants (D). For C and D, ∗∗, p < 0.01 vs control cells and cells treated with IFN-γ and infected with the EHEC strains (n = 4). Ctrl, Untreated cells; EDL, EDL933.

FIGURE 1.

Effect of EHEC on iNOS induction in human enterocytes. Hct-8 cells were infected with bacteria for 6 h. RNA was purified, reverse transcribed, and the expression level of the gene encoding iNOS was analyzed by RT-PCR (A). Cells treated with IFN-γ were used as positive control for iNOS mRNA expression. Cells were treated with IFN-γ in the presence or absence of bacteria. Subsequently, iNOS gene transcription was analyzed by RT-PCR (B) and real-time PCR (C). The 6-h stimulated cells were washed and incubated for 18 h in complete medium containing antibiotics. [NO2] plus [NO3] were then measured in culture supernatants (D). For C and D, ∗∗, p < 0.01 vs control cells and cells treated with IFN-γ and infected with the EHEC strains (n = 4). Ctrl, Untreated cells; EDL, EDL933.

Close modal

Similar results were obtained with Caco-2 and T84 cells (Table II). Moreover, iNOS mRNA expression and NO production increased by ∼20% when the cells were stimulated with a cytokine mix (IFN-γ, TNF-α, and IL-1β), in comparison to the cells that were only stimulated with IFN-γ; in this condition, the EHEC inhibitory effect was also observed.

Table II.

Induction of iNOS in human epithelial cell linesa

CtrlIFN-γIFN-γ + EDL933
Caco-2    
 iNOS mRNAb 18.4 ± 1.9 0.5 ± 0.2 
 [NOx] (μM) 0.8 ± 0.3 20.2 ± 3.7 1.2 ± 0.7 
T84    
 iNOS mRNAb 17.3 ± 0.8 0.8 ± 0.2 
 [NOx] (μM) 0.6 ± 0.1 21.2 ± 0.6 1.1 ± 0.5 
CtrlIFN-γIFN-γ + EDL933
Caco-2    
 iNOS mRNAb 18.4 ± 1.9 0.5 ± 0.2 
 [NOx] (μM) 0.8 ± 0.3 20.2 ± 3.7 1.2 ± 0.7 
T84    
 iNOS mRNAb 17.3 ± 0.8 0.8 ± 0.2 
 [NOx] (μM) 0.6 ± 0.1 21.2 ± 0.6 1.1 ± 0.5 
a

Caco-2 or T84 cells were stimulated with IFN-γ, in the presence or absence of the strain EDL933, for 6 h. Subsequently, iNOS mRNA expression was analyzed by real-time PCR. Cells activated for 6 h were washed and incubated with fresh medium containing antibiotics for 18 h. NO concentrations were measured in the culture supernatants. Each data is the mean ± SEM of n = 3–4 independent experiments.

b

Express as a fold increase/Ctrl.

Together these data demonstrate that O157:H7 and O113:H21 EHEC repress iNOS mRNA expression in human enterocytes activated by IFN-γ.

We next aimed to identify the molecular mechanism by which EHEC suppress IFN-γ-induced iNOS gene expression. We first demonstrated that iNOS mRNA expression (Fig. 2,A) and NO production (Fig. 2 B) in response to IFN-γ stimulation was completely inhibited using a STAT-1 or a JAK-2 inhibitor, but not by the NF-κB inhibitors Z-Leu-Leu-Leu-CHO and Bay 11–7082. Similar results were obtained when TNF-α and IL-1β were used to stimulate the cells together with IFN-γ (data not shown). Thus, these data suggest that iNOS transcription in activated epithelial cells involves the JAK-2/STAT-1 transduction pathway.

FIGURE 2.

STAT-1 and NF-κB activation. A and B, Identification of the transcription factor implicated in iNOS induction. Hct-8 cells were treated for 30 min with pharmacological inhibitors (PI), namely the NF-κB inhibitors MG132 (MG) and Bay 11–7082 (Bay), the STAT-1 inhibitor (SI), or the JAK-2 inhibitor (JI); cells were then stimulated for 6 h with IFN-γ. The expression of the iNOS gene was analyzed by RT-PCR (A). Activated cells were incubated for 18 h in complete medium containing antibiotics and NO generation was measured in culture supernatants (B). C, Nuclear proteins from cells activated with IFN-γ ± EDL933 (EDL) for 3 h were extracted and analyzed by EMSA. Similar results were obtained in three independent experiments. D, Cellular localization of STAT-1 and NF-κB. Hct-8 cells were activated for 3 h and fixed on Lab-Tek slides. STAT-1 and NF-κB were immunodetected using an anti-p65 Ab (red) and with an anti-STAT-1 Ab (red), respectively. Nuclear outline was defined by DAPI staining (blue). Merged images are shown and overlay is presented in pink. The same results have been obtained in three independent experiments.

FIGURE 2.

STAT-1 and NF-κB activation. A and B, Identification of the transcription factor implicated in iNOS induction. Hct-8 cells were treated for 30 min with pharmacological inhibitors (PI), namely the NF-κB inhibitors MG132 (MG) and Bay 11–7082 (Bay), the STAT-1 inhibitor (SI), or the JAK-2 inhibitor (JI); cells were then stimulated for 6 h with IFN-γ. The expression of the iNOS gene was analyzed by RT-PCR (A). Activated cells were incubated for 18 h in complete medium containing antibiotics and NO generation was measured in culture supernatants (B). C, Nuclear proteins from cells activated with IFN-γ ± EDL933 (EDL) for 3 h were extracted and analyzed by EMSA. Similar results were obtained in three independent experiments. D, Cellular localization of STAT-1 and NF-κB. Hct-8 cells were activated for 3 h and fixed on Lab-Tek slides. STAT-1 and NF-κB were immunodetected using an anti-p65 Ab (red) and with an anti-STAT-1 Ab (red), respectively. Nuclear outline was defined by DAPI staining (blue). Merged images are shown and overlay is presented in pink. The same results have been obtained in three independent experiments.

Close modal

STAT-1 and NF-κB activation were then assessed by EMSA. As shown in Fig. 2 C, we observed an increase in STAT-1 DNA-binding activity upon stimulation with IFN-γ in comparison to control cells. This enhancement was fully inhibited when EDL933 was added to the activated cells. Moreover, the basal STAT-1 activation observed in control cells was suppressed after infection with the bacteria. Oppositely, NF-κB DNA binding activity was not modulated by IFN-γ or by the strain EDL933.

These results were confirmed by immunofluorescence and confocal microscopy (Fig. 2 D). The nuclear translocation of STAT-1 was evidenced in IFN-γ-treated Hct-8 cells, and was suppressed by the addition of the strain EDL933. NF-κB was detected only in the cytoplasm of control or IFN-γ-activated cells; its nuclear translocation was observed in ∼10% of EDL933-infected cells, as previously described (10).

We conclude that the inhibition of STAT-1 activation by EHEC blocks the induction of iNOS gene transcription.

HO-1 activity has been previously described to inhibit iNOS induction (21, 22, 23, 24, 25). We therefore hypothesized that EHEC suppress iNOS mRNA expression through the activation of HO-1. We first analyzed hmox-1 mRNA expression levels in Hct-8 cells in response to EHEC by RT-PCR (Fig. 3,A) and semiquantitative PCR (Fig. 3,B). The gene hmox-1 was up-regulated 0.5 h after the beginning of the infection, reached a peak after 1 h, and decreased from this level at 2 and 3 h. Compared with uninfected cells, the level of HO-1 protein in human enterocytes was enhanced in response to a 3-h EHEC infection (Fig. 3,C). IFN-γ or cytokines did not induce HO-1 in Hct-8 cells (Fig. 3 C). Additionally, hmox-1 was not up-regulated when the cells were infected with EcG1, EcG2, or DH5α (data not shown).

FIGURE 3.

Induction of HO-1 by EHEC. A and B, Expression of hmox-1 mRNA in response to EHEC. Hct-8 cells were infected with EDL933 in the presence or absence of IFN-γ and/or TNF-α and IL-1β. At indicated times, total RNA was extracted and the expression of hmox-1 was analyzed by RT-PCR (A) and semiquantitative PCR (B; n = 4). C, The presence of HO-1 protein was monitored by Western blotting. Equal amounts of protein (25 μg) were loaded on the gel. P, Positive control.

FIGURE 3.

Induction of HO-1 by EHEC. A and B, Expression of hmox-1 mRNA in response to EHEC. Hct-8 cells were infected with EDL933 in the presence or absence of IFN-γ and/or TNF-α and IL-1β. At indicated times, total RNA was extracted and the expression of hmox-1 was analyzed by RT-PCR (A) and semiquantitative PCR (B; n = 4). C, The presence of HO-1 protein was monitored by Western blotting. Equal amounts of protein (25 μg) were loaded on the gel. P, Positive control.

Close modal

To investigate the role of HO-1 in EHEC-mediated iNOS induction inhibition in human enterocytes, we first used a siRNA-based strategy. Cells were first exposed to HO-1 or lamin siRNA, and were then were stimulated by IFN-γ and/or infected with EDL933; HO-1 and iNOS mRNA expression were subsequently assessed by real-time PCR (Fig. 4,A). The HO-1 mRNA was expressed upon EDL933 infection when compared with uninfected cells, but was not induced by IFN-γ. Inversely, iNOS mRNA expression was up-regulated in the presence of IFN-γ but was not stimulated by the bacteria. The expression of hmox-1 in EDL933-infected cells was completely inhibited using HO-1 siRNA; this inhibition was concomitantly associated with the partial, but significant, restoration of iNOS mRNA expression. The lamin siRNA had no effect on hmox-1 and iNOS mRNA expression in Hct-8 cells stimulated with IFN-γ and infected by EDL933. To further verify the association of HO-1 induction with the inhibition of iNOS mRNA expression, we used the HO inhibitor ZnPP in Hct-8 cells/EDL933 cocultures. In the presence of ZnPP, the EHEC-induced inhibition of iNOS mRNA expression was completely suppressed (Fig. 4,B). Concomitantly, STAT-1 activation was partially restored using HO-1 siRNA, but not lamin siRNA, on cells activated by IFN-γ and infected with EDL933 (Fig. 4 C).

FIGURE 4.

Inhibition of iNOS gene transcription by HO-1 in activated cells. A, hmox-1 interference. Cells were transfected or not with HO-1 or lamin siRNA before a stimulation with IFN-γ, in the presence or absence of EDL933. Subsequently, the hmox-1 (open bars) and iNOS (plain bars) mRNA expression levels were analyzed by real-time PCR after 2 and 6 h of infection, respectively. For both genes, ∗, p < 0.05 vs control cells; §, p < 0.05 vs cytokine-treated cells; #, p < 0.05 vs HO-1 siRNA-treated cells infected with EDL933 and stimulated with IFN-γ (n = 4). B, HO activity inhibition. Cells were treated with IFN-γ, EDL933, and/or ZnPP for 6 h. The level of iNOS mRNA was measured by real-time PCR. ∗∗, p < 0.01 vs control cells and EDL933-infected cells treated with IFN-γ (n = 3). C, Effect of HO-1 on STAT-1 activation. Nuclear proteins from cells transfected or not with HO-1 or lamin siRNA and activated for 3 h with IFN-γ and/or EDL933 were extracted and analyzed by EMSA. Similar results were obtained in two independent experiments.

FIGURE 4.

Inhibition of iNOS gene transcription by HO-1 in activated cells. A, hmox-1 interference. Cells were transfected or not with HO-1 or lamin siRNA before a stimulation with IFN-γ, in the presence or absence of EDL933. Subsequently, the hmox-1 (open bars) and iNOS (plain bars) mRNA expression levels were analyzed by real-time PCR after 2 and 6 h of infection, respectively. For both genes, ∗, p < 0.05 vs control cells; §, p < 0.05 vs cytokine-treated cells; #, p < 0.05 vs HO-1 siRNA-treated cells infected with EDL933 and stimulated with IFN-γ (n = 4). B, HO activity inhibition. Cells were treated with IFN-γ, EDL933, and/or ZnPP for 6 h. The level of iNOS mRNA was measured by real-time PCR. ∗∗, p < 0.01 vs control cells and EDL933-infected cells treated with IFN-γ (n = 3). C, Effect of HO-1 on STAT-1 activation. Nuclear proteins from cells transfected or not with HO-1 or lamin siRNA and activated for 3 h with IFN-γ and/or EDL933 were extracted and analyzed by EMSA. Similar results were obtained in two independent experiments.

Close modal

To gain further insight into the mechanism by which HO-1 activity inhibits iNOS mRNA expression, we treated activated Hct-8 cells with the two main products of heme metabolism, namely bilirubin and CO. The IFN-γ-dependent iNOS gene expression was not modulated by various concentrations of bilirubin (Fig. 5). Inversely, a concentration-dependent inhibition of iNOS mRNA expression was observed using the CO donor CORM-2 (Fig. 5).

FIGURE 5.

Hct-8 cells were stimulated for 6 h with IFN-γ in the presence or absence of various concentrations of bilirubin or CORM-2. Subsequently, iNOS gene transcription was analyzed by real-time PCR. ∗∗, p < 0.01 when compared with cells treated with IFN-γ (n = 3–4).

FIGURE 5.

Hct-8 cells were stimulated for 6 h with IFN-γ in the presence or absence of various concentrations of bilirubin or CORM-2. Subsequently, iNOS gene transcription was analyzed by real-time PCR. ∗∗, p < 0.01 when compared with cells treated with IFN-γ (n = 3–4).

Close modal

We next reasoned that EHEC-induced HO-1 activity may suppress iNOS mRNA expression in unstimulated epithelial cells. HO-1 expression was up-regulated by 11 ± 3.1-fold in EDL933-infected cells when compared with control cells; this increase was completely inhibited using HO-1 siRNA, but not lamin siRNA. However, iNOS mRNA expression was not induced in cells infected with EHEC in the presence or absence of HO-1 or lamin siRNA (Fig. 6).

FIGURE 6.

Cells were transfected or not with HO-1 or lamin siRNA before an infection with the strain EDL933. The expression of hmox-1 (open bars) and iNOS (plain bars) genes were analyzed by real-time PCR after 2 and 6 h of infection, respectively. ∗, p < 0.05 vs control cells and infected cells treated with HO-1 siRNA (n = 3).

FIGURE 6.

Cells were transfected or not with HO-1 or lamin siRNA before an infection with the strain EDL933. The expression of hmox-1 (open bars) and iNOS (plain bars) genes were analyzed by real-time PCR after 2 and 6 h of infection, respectively. ∗, p < 0.05 vs control cells and infected cells treated with HO-1 siRNA (n = 3).

Close modal

These results indicate that HO-1 activity is required for EHEC-mediated inhibition of STAT-1 activation and iNOS mRNA expression in IFN-γ-activated human enterocytes, but does not modulate iNOS transcription in untreated cells.

We analyzed the effect on hmox-1 induction and on iNOS inhibition of the main bacterial factors implicated in the EHEC pathogenesis: intimin is encoded by the LEE gene eae and allows the intimate attachment of the bacteria to epithelial cells; EspA forms the needle of the type III secretion system; EspD protein is essential for the formation of surface appendages and is integrated in the cytoplasmic membranes of eukaryotic cells; SepL is involved in the translocation of bacterial factors; the main virulence factor is Stx encoded by the genes stx1 and stx2. The gene hmox-1 was up-regulated by 15–30-fold in Hct-8 cells cocultured with EDL933, eae-, espA-, espD-, sepL-, or stx1-/stx2- strains when compared with uninfected cells (Fig. 7). Concomitantly, there was no significant difference in the degree of down-regulation of iNOS mRNA in IFN-γ-treated Hct-8 cells between the wild-type strain and isogenic mutants (Fig. 7).

FIGURE 7.

Effect of different mutant strains on hmox-1 induction and iNOS transcription inhibition. Hct-8 cells were infected with EDL933 (EDL) or with isogenic mutants lacking eae, espA, espD, sepL, or stx1/stx2 (stx) genes, in the presence of IFN-γ. The relative expression of hmox-1 (open bars) and iNOS (plain bars) genes was measured by semiquantitative PCR after 2 and 6 h of infection, respectively. ∗, p < 0.05 vs the others conditions (n = 3–5).

FIGURE 7.

Effect of different mutant strains on hmox-1 induction and iNOS transcription inhibition. Hct-8 cells were infected with EDL933 (EDL) or with isogenic mutants lacking eae, espA, espD, sepL, or stx1/stx2 (stx) genes, in the presence of IFN-γ. The relative expression of hmox-1 (open bars) and iNOS (plain bars) genes was measured by semiquantitative PCR after 2 and 6 h of infection, respectively. ∗, p < 0.05 vs the others conditions (n = 3–5).

Close modal

Our results show that EHEC-induced HO-1 in human epithelial cells is a critical modulator of the innate immune response. Although HO-1 has been previously implicated in attenuating the inflammatory response (26, 27) and in the inhibition of iNOS gene expression (22, 23, 24, 25), the present study is the first to demonstrate that a pathogenic bacterium suppresses NO production by inducing HO-1 in activated human colonic epithelial cells.

In our experimental conditions, activation of the transcription factor STAT-1 was required for induction of iNOS gene transcription in activated Hct-8 cells because a STAT-1 inhibitor or a JAK-2 inhibitor, but not NF-κB inhibitors, suppressed the IFN-γ-induced iNOS mRNA expression, as previously described with other cell types (14, 28). Similar to previous findings obtained with Hep-2 and T84 cells (9, 29), we found that EHEC interferes with STAT-1 activation in Hct-8 and Caco-2 cells activated with IFN-γ or with the three cytokines, without any effect on NF-κB. Nonetheless, we establish herein that EHEC inhibit the activation of STAT-1 by inducing HO-1. The enzyme HO-1 catabolizes heme into CO, biliverdin that is rapidly converted to bilirubin, and iron (30). These products have antioxidant properties and may mediate the anti-inflammatory response of HO-1 (26, 31, 32, 33). It has been described (34) that the HO-1-dependent inhibition of the IFN-γ-STAT-1 signaling pathway in murine endothelial cells is dependent on the formation of bilirubin. Nonetheless, our experiments further indicate that bilirubin has no effect on iNOS gene expression in cytokine-treated Hct-8 cells; rather, iNOS mRNA expression was inhibited by the CO donor CORM-2. Therefore, we propose that CO is the key element of iNOS gene regulation, as recently established in Caco-2 cells (35). Aside from its role in heme degradation, HO-1 has a critical function in maintaining cellular homeostasis. HO-1-deficient mice do not normally survive and the animals that do survive die within a year while demonstrating signs of chronic inflammation in numerous organs (36). Thus, it is proposed that HO-1 serves as an adaptive mechanism to protect cells from oxidative stress (37). In this context, an increased expression of HO-1 has also been shown in pathophysiological states such as atherosclerosis (38), sepsis (39), Helicobacter pylori-positive and -negative gastritis, and ulcerative colitis (40); in these two last inflammatory diseases, HO-1 was immunodetected in superficial and cryptic epithelial cells. Additionally, the protective role of HO-1 has been demonstrated in experimental colitis in trinitrobenzene sulfonic acid and dextran sodium sulfate models (41, 42), and in IL-10-deficient mice (43). The preventive effect of HO-1 was correlated with decreased free radical production and the inhibition of iNOS expression in colonic tissues (41). Together with our findings, these reports suggest that HO-1 could be over-expressed in the human colonic epithelium during EHEC infection and may modulate the host inflammatory response and the development of the hemorrhagic symptoms.

When hmox-1 expression was inhibited by siRNA, a partial restoration of STAT-1 activation and iNOS mRNA expression was observed. However, when HO-1 activity was inhibited by ZnPP, the transcription of the gene encoding iNOS was completely restored. According to this result, we hypothesize that the constitutive isoforms of HO, namely HO-2 and HO-3, could be implicated in part in iNOS gene down-regulation. Nonetheless, we might speculate EHEC suppress STAT-1 activation by others mechanisms than HO-1 induction or inhibit iNOS transcription by acting on different transcription factors. Intriguingly, O157:H7 EHEC express an HO encoded by the gene chuS (44). This gene is shared by other pathogenic bacteria such as Corynebacterium diphtheriae (45) or Neisseria meningitides (46), but is not present in nonpathogenic E. coli laboratory strains. It has been previously demonstrated that a bacterial enzyme, namely the arginase of Helicobacter pylori, can act as a potent regulator of iNOS activity in eukaryotic cells (12). Thus, we constructed a chuS mutant in the EDL933 background to determine the implication of bacterial HO on host iNOS mRNA expression. We did not find any differences in the levels of iNOS induction in cells infected with EDL933 or with the isogenic mutant chuS, in the presence or absence of cytokines (data not shown). These results demonstrate that EHEC HO does not modulate iNOS gene expression in human enterocytes.

High-output NO formation in pathological conditions may determine pathogen behavior and host susceptibility. Therefore, pathogenic bacteria have elaborated different strategies to modulate iNOS expression and NO production (12). The iNOS gene is up-regulated in human epithelial cells infected with enteroinvasive bacteria such as E. coli or Salmonella dublin (47, 48), or in IFN-γ-primed enterocytes infected with S. dublin (13). Conversely, we and others have observed that iNOS is not directly induced in colonic epithelial cells by EHEC (this work) and enteropathogenic E. coli (EPEC; Ref. 49) that develop extracellularly. Furthermore, these bacteria inhibit the IFN-γ-induced iNOS mRNA expression (this work and Ref. 49). However, NF-κB activation is the main target for EPEC (49), whereas our data highlight that EHEC disrupt the STAT-1 signaling. We suggest that this discrepancy can be explained by the different technical approaches to stimulate iNOS mRNA expression (cytokine concentration, time of stimulation). Alternatively, EPEC and EHEC may have elaborated different strategies to disrupt cellular signalization and limit NO production. In this respect, it has been described that EHEC but not EPEC disrupt the IFN-γ-STAT-1 signaling in human epithelial cells (29). Moreover, the EPEC inhibitory effect on iNOS induction is linked to the LEE-encoded type three secretion system because a sep-2 mutant strain that lacks a functional secretion machinery fails to repress the cytokine-induced NO production (49), whereas we show that 1) LEE positive and LEE negative EHEC strains are potent inhibitors of iNOS mRNA expression, and 2) LEE gene isogenic mutants exhibit the same inhibitory effect as the parental LEE-positive strain EDL933. Thus, we suggest that adhesion and the type III secretion system are not involved in inhibition of NO production. Moreover, our experimental data indicate that Stx is not involved in HO-1 induction and iNOS inhibition. The identification of the bacterial factor(s) responsible for hmox-1 expression and NO production inhibition deserves additional investigation. Such research is currently underway in our laboratory.

The immune function of NO refers to its ability to modulate the inflammatory process and to kill pathogens. Moreover, we recently demonstrated that chemical or cellular sources of NO inhibit the expression of the gene stx2 and the synthesis of Stx2, without affecting the growth and the survival of the bacteria (15). It should be noted that, in our previous report, we first stimulated the cells with cytokines for 18 h to induce iNOS, before adding EHEC (15); in the present study, Hct-8 cells were stimulated with cytokines and EHEC at the same time. Based on these both findings, it is tempting to speculate that EHEC suppress iNOS mRNA expression to decrease the NO-dependent Stx production inhibition. Therefore, limiting host NO production in human enterocytes may represent a strategy elaborated by EHEC to favor their own virulence.

We thank T. S. Whittam (Michigan State University, East Lansing, MI) and C. A. Guzman (Helmholtz Centre for Infection Research, Braunschweig, Germany) for the gift of strains. Isolation and identification of human commensal E. coli isolates has been performed with the support of C. Del’Homme and E. Delmas (Institut National de la Recherche Agronomique, Saint-Genès-Champanelle, France). C. Besson and C. Leyrat are thanked for technical assistance. We are grateful to M. T. Corvol (Université Paris Descartes, Paris, France) for data analysis.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by grants from Institut National de la Recherche Agronomique and Région Auvergne (to M.V.), from Université Paris Descartes, Arthritis Fondation Courtin, Institut National de la Santé et de la Recherche Médicale (to F.R.), and from Société Française de Rhumatologie (to N.T.).

4

Abbreviations used in this paper: EHEC, enterohemorrhagic Escherichia coli; LEE, locus of enterocyte effacement; iNOS, inducible NO synthase; Stx, Shiga-toxin; HO, heme oxygenase; Bay 11–7082, (E)3-[(4-methylphenyl)sulfonyl]-2-propenenitrile; ZnPP, zinc protoporphyrin IX; CO, carbon monoxide; CORM-2, tricarbonyldichlororuthenium-(II)-dimer; siRNA, small interfering RNA; EPEC, enteropathogenic E. coli.

1
Tarr, P. I., C. A. Gordon, W. L. Chandler.
2005
. Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome.
Lancet
365
:
1073
-1086.
2
Murata, A., T. Shimazu, T. Yamamoto, N. Taenaka, K. Nagayama, T. Honda, H. Sugimoto, M. Monden, N. Matsuura, S. Okada.
1998
. Profiles of circulating inflammatory- and anti-inflammatory cytokines in patients with hemolytic uremic syndrome due to E. coli O157 infection.
Cytokine
10
:
544
-548.
3
Proulx, F., J. P. Turgeon, C. Litalien, M. M. Mariscalco, P. Robitaille, E. Seidman.
1998
. Inflammatory mediators in Escherichia coli O157:H7 hemorrhagic colitis and hemolytic-uremic syndrome.
Pediatr. Infect. Dis. J.
17
:
899
-904.
4
Litalien, C., F. Proulx, M. M. Mariscalco, P. Robitaille, J. P. Turgeon, E. Orrbine, P. C. Rowe, P. N. McLaine, E. Seidman.
1999
. Circulating inflammatory cytokine levels in hemolytic uremic syndrome.
Pediatr. Nephrol.
13
:
840
-845.
5
Westerholt, S., T. Hartung, M. Tollens, A. Gustrau, M. Oberhoffer, H. Karch, B. Klare, K. Pfeffer, P. Emmrich, R. Oberhoffer.
2000
. Inflammatory and immunological parameters in children with haemolytic uremic syndrome (HUS) and gastroenteritis-pathophysiological and diagnostic clues.
Cytokine
12
:
822
-827.
6
Dahan, S., V. Busuttil, V. Imbert, J. F. Peyron, P. Rampal, D. Czerucka.
2002
. Enterohemorrhagic Escherichia coli infection induces interleukin-8 production via activation of mitogen-activated protein kinases and the transcription factors NF-κB and AP-1 in T84 cells.
Infect. Immun.
70
:
2304
-2310.
7
Dalmasso, G., A. Loubat, S. Dahan, G. Calle, P. Rampal, D. Czerucka.
2006
. Saccharomyces boulardii prevents TNF-α-induced apoptosis in EHEC-infected T84 cells.
Res. Microbiol.
157
:
456
-465.
8
Hauf, N., T. Chakraborty.
2003
. Suppression of NF-κB activation and proinflammatory cytokine expression by Shiga toxin-producing Escherichia coli.
J. Immunol.
170
:
2074
-2082.
9
Jandu, N., P. J. Ceponis, S. Kato, J. D. Riff, D. M. McKay, P. M. Sherman.
2006
. Conditioned medium from enterohemorrhagic Escherichia coli-infected T84 cells inhibits signal transducer and activator of transcription 1 activation by γ interferon.
Infect. Immun.
74
:
1809
-1818.
10
Gobert, A. P., M. Vareille, A. L. Glasser, T. Hindre, T. de Sablet, C. Martin.
2007
. Shiga toxin produced by enterohemorrhagic Escherichia coli inhibits PI3K/NF-κB signaling pathway in globotriaosylceramide-3-negative human intestinal epithelial cells.
J. Immunol.
178
:
8168
-8174.
11
Gobert, A. P., Y. Cheng, M. Akhtar, B. D. Mersey, D. R. Blumberg, R. K. Cross, R. Chaturvedi, C. B. Drachenberg, J. L. Boucher, A. Hacker, et al
2004
. Protective role of arginase in a mouse model of colitis.
J. Immunol.
173
:
2109
-2117.
12
Gobert, A. P., D. J. McGee, M. Akhtar, G. L. Mendz, J. C. Newton, Y. Cheng, H. L. Mobley, K. T. Wilson.
2001
. Helicobacter pylori arginase inhibits nitric oxide production by eukaryotic cells: a strategy for bacterial survival.
Proc. Natl. Acad. Sci. USA
98
:
13844
-13849.
13
Salzman, A. L., T. Eaves-Pyles, S. C. Linn, A. G. Denenberg, C. Szabo.
1998
. Bacterial induction of inducible nitric oxide synthase in cultured human intestinal epithelial cells.
Gastroenterology
114
:
93
-102.
14
Ganster, R. W., B. S. Taylor, L. Shao, D. A. Geller.
2001
. Complex regulation of human inducible nitric oxide synthase gene transcription by Stat 1 and NF-κB.
Proc. Natl. Acad. Sci. USA
98
:
8638
-8643.
15
Vareille, M., T. de Sablet, T. Hindre, C. Martin, A. P. Gobert.
2007
. Nitric oxide inhibits Shiga-toxin synthesis by enterohemorrhagic Escherichia coli.
Proc. Natl. Acad. Sci. USA
104
:
10199
-10204.
16
Beltrametti, F., A. U. Kresse, C. A. Guzman.
1999
. Transcriptional regulation of the esp genes of enterohemorrhagic Escherichia coli.
J. Bacteriol.
181
:
3409
-3418.
17
Kresse, A. U., M. Rohde, C. A. Guzman.
1999
. The EspD protein of enterohemorrhagic Escherichia coli is required for the formation of bacterial surface appendages and is incorporated in the cytoplasmic membranes of target cells.
Infect. Immun.
67
:
4834
-4842.
18
Kresse, A. U., F. Beltrametti, A. Muller, F. Ebel, C. A. Guzman.
2000
. Characterization of SepL of enterohemorrhagic Escherichia coli.
J. Bacteriol.
182
:
6490
-6498.
19
Gobert, A. P., A. Coste, C. A. Guzman, T. Hindré, T. de Sablet, J. P. Girardeau, C. Martin.
2008
. Modulation of chemokine gene expression by Shiga-toxin producing Escherichia coli belonging to various origins and serotypes.
Microbes Infect.
10
:
159
-165.
20
Robert, C., C. Chassard, P. A. Lawson, A. Bernalier-Donadille.
2007
. Bacteroides cellulosilyticus sp. nov., a cellulolytic bacterium from the human gut microbial community.
Int. J. Syst. Evol. Microbiol.
57
:
1516
-1520.
21
Benallaoua, M., M. Francois, F. Batteux, N. Thelier, J. Y. Shyy, C. Fitting, L. Tsagris, J. Boczkowski, J. F. Savouret, M. T. Corvol, et al
2007
. Pharmacologic induction of heme oxygenase 1 reduces acute inflammatory arthritis in mice.
Arthritis Rheum.
56
:
2585
-2594.
22
Cavicchi, M., L. Gibbs, B. J. Whittle.
2000
. Inhibition of inducible nitric oxide synthase in the human intestinal epithelial cell line, DLD-1, by the inducers of heme oxygenase 1, bismuth salts, heme, and nitric oxide donors.
Gut
47
:
771
-778.
23
Chen, M., W. Bao, R. Aizman, P. Huang, O. Aspevall, L. E. Gustafsson, S. Ceccatelli, G. Celsi.
2004
. Activation of extracellular signal-regulated kinase mediates apoptosis induced by uropathogenic Escherichia coli toxins via nitric oxide synthase: protective role of heme oxygenase-1.
J. Infect. Dis.
190
:
127
-135.
24
Chen, J. C., F. M. Ho, P.-D. Lee Chao, C. P. Chen, K. C. Jeng, H. B. Hsu, S. T. Lee, W. W. Tung, W. W. Lin.
2005
. Inhibition of iNOS gene expression by quercetin is mediated by the inhibition of I-κB kinase, nuclear factor-κB and STAT1, and depends on heme oxygenase-1 induction in mouse BV-2 microglia.
Eur. J. Pharmacol.
521
:
9
-20.
25
Oh, G. S., H. O. Pae, B. S. Lee, B. N. Kim, J. M. Kim, H. R. Kim, S. B. Jeon, W. K. Jeon, H. J. Chae, H. T. Chung.
2006
. Hydrogen sulfide inhibits nitric oxide production and nuclear factor-κB via heme oxygenase-1 expression in RAW264.7 macrophages stimulated with lipopolysaccharide.
Free Radic. Biol. Med.
41
:
106
-119.
26
Lee, T. S., L. Y. Chau.
2002
. Heme oxygenase-1 mediates the anti-inflammatory effect of interleukin-10 in mice.
Nat. Med.
8
:
240
-246.
27
Nakahira, K., H. P. Kim, X. H. Geng, A. Nakao, X. Wang, N. Murase, P. F. Drain, X. Wang, M. Sasidhar, E. G. Nabel, et al
2006
. Carbon monoxide differentially inhibits TLR signaling pathways by regulating ROS-induced trafficking of TLRs to lipid rafts.
J. Exp. Med.
203
:
2377
-2389.
28
Kleinert, H., T. Wallerath, G. Fritz, I. Ihrig-Biedert, F. Rodriguez-Pascual, D. A. Geller, U. Forstermann.
1998
. Cytokine induction of NO synthase II in human DLD-1 cells: roles of the JAK-STAT, AP-1 and NF-κB-signaling pathways.
Br. J. Pharmacol.
125
:
193
-201.
29
Ceponis, P. J., D. M. McKay, J. C. Ching, P. Pereira, P. M. Sherman.
2003
. Enterohemorrhagic Escherichia coli O157:H7 disrupts Stat1-mediated γ interferon signal transduction in epithelial cells.
Infect. Immun.
71
:
1396
-1404.
30
Ryter, S. W., J. Alam, A. M. Choi.
2006
. Heme oxygenase-1/carbon monoxide: from basic science to therapeutic applications.
Physiol. Rev.
86
:
583
-650.
31
Otterbein, L. E., F. H. Bach, J. Alam, M. Soares, H. Tao Lu, M. Wysk, R. J. Davis, R. A. Flavell, A. M. Choi.
2000
. Carbon monoxide has anti-inflammatory effects involving the mitogen-activated protein kinase pathway.
Nat. Med.
6
:
422
-428.
32
Sarady-Andrews, J. K., F. Liu, D. Gallo, A. Nakao, M. Overhaus, R. Ollinger, A. M. Choi, L. E. Otterbein.
2005
. Biliverdin administration protects against endotoxin-induced acute lung injury in rats.
Am. J. Physiol.
289
:
L1131
-L1137.
33
Overhaus, M., B. A. Moore, J. E. Barbato, F. F. Behrendt, J. G. Doering, A. J. Bauer.
2006
. Biliverdin protects against polymicrobial sepsis by modulating inflammatory mediators.
Am. J. Physiol.
290
:
G695
-G703.
34
Wu, J., J. Ma, S. T. Fan, H. J. Schlitt, T. Y. Tsui.
2005
. Bilirubin derived from heme degradation suppresses MHC class II expression in endothelial cells.
Biochem. Biophys. Res. Commun.
338
:
890
-896.
35
Megias, J., J. Busserolles, M. J. Alcaraz.
2007
. The carbon monoxide-releasing molecule CORM-2 inhibits the inflammatory response induced by cytokines in Caco-2 cells.
Br. J. Pharmacol.
150
:
977
-986.
36
Poss, K. D., S. Tonegawa.
1997
. Heme oxygenase 1 is required for mammalian iron reutilization.
Proc. Natl. Acad. Sci. USA
94
:
10919
-10924.
37
Poss, K. D., S. Tonegawa.
1997
. Reduced stress defense in heme oxygenase 1-deficient cells.
Proc. Natl. Acad. Sci. USA
94
:
10925
-10930.
38
Wang, L. J., T. S. Lee, F. Y. Lee, R. C. Pai, L. Y. Chau.
1998
. Expression of heme oxygenase-1 in atherosclerotic lesions.
Am. J. Pathol.
152
:
711
-720.
39
Yet, S. F., A. Pellacani, C. Patterson, L. Tan, S. C. Folta, L. Foster, W. S. Lee, C. M. Hsieh, M. A. Perrella.
1997
. Induction of heme oxygenase-1 expression in vascular smooth muscle cells. A link to endotoxic shock.
J. Biol. Chem.
272
:
4295
-4301.
40
Barton, S. G., D. S. Rampton, V. R. Winrow, P. Domizio, R. M. Feakins.
2003
. Expression of heat shock protein 32 (hemoxygenase-1) in the normal and inflamed human stomach and colon: an immunohistochemical study.
Cell Stress Chaperones
8
:
329
-334.
41
Wang, W. P., X. Guo, M. W. Koo, B. C. Wong, S. K. Lam, Y. N. Ye, C. H. Cho.
2001
. Protective role of heme oxygenase-1 on trinitrobenzene sulfonic acid-induced colitis in rats.
Am. J. Physiol.
281
:
G586
-G594.
42
Paul, G., F. Bataille, F. Obermeier, J. Bock, F. Klebl, U. Strauch, D. Lochbaum, P. Rummele, S. Farkas, J. Scholmerich, et al
2005
. Analysis of intestinal haem-oxygenase-1 (HO-1) in clinical and experimental colitis.
Clin. Exp. Immunol.
140
:
547
-555.
43
Hegazi, R. A., K. N. Rao, A. Mayle, A. R. Sepulveda, L. E. Otterbein, S. E. Plevy.
2005
. Carbon monoxide ameliorates chronic murine colitis through a heme oxygenase 1-dependent pathway.
J. Exp. Med.
202
:
1703
-1713.
44
Suits, M. D., G. P. Pal, K. Nakatsu, A. Matte, M. Cygler, Z. Jia.
2005
. Identification of an Escherichia coli O157:H7 heme oxygenase with tandem functional repeats.
Proc. Natl. Acad. Sci. USA
102
:
16955
-16960.
45
Schmitt, M. P..
1997
. Utilization of host iron sources by Corynebacterium diphtheriae: identification of a gene whose product is homologous to eukaryotic heme oxygenases and is required for acquisition of iron from heme and hemoglobin.
J. Bacteriol.
179
:
838
-845.
46
Zhu, W., A. Wilks, I. Stojiljkovic.
2000
. Degradation of heme in gram-negative bacteria: the product of the hemO gene of Neisseriae is a heme oxygenase.
J. Bacteriol.
182
:
6783
-6790.
47
Witthoft, T., L. Eckmann, J. M. Kim, M. F. Kagnoff.
1998
. Enteroinvasive bacteria directly activate expression of iNOS and NO production in human colon epithelial cells.
Am. J. Physiol.
275
:
G564
-G571.
48
Resta-Lenert, S., K. E. Barrett.
2002
. Enteroinvasive bacteria alter barrier and transport properties of human intestinal epithelium: role of iNOS and COX-2.
Gastroenterology
122
:
1070
-1087.
49
Maresca, M., D. Miller, S. Quitard, P. Dean, B. Kenny.
2005
. Enteropathogenic Escherichia coli (EPEC) effector-mediated suppression of antimicrobial nitric oxide production in a small intestinal epithelial model system.
Cell. Microbiol.
7
:
1749
-1762.