TLR4-Dependent NF-κB Activation and Mitogen- and Stress-Activated Protein Kinase 1-Triggered Phosphorylation Events Are Central to Helicobacter pylori Peptidyl Prolyl cis-, trans-Isomerase (HP0175)-Mediated Induction of IL-6 Release from Macrophages¹

Helicobacter pylori is a Gram-negative microaerophilic bacterium that causes chronic gastritis and also peptic ulcer, gastric carcinoma, and gastric lymphoma (1). Infection is associated with the local production of chemokines and cytokines, such as IL-1β, IL-6, and IL-8 (2, 3). IL-6 is a pleiotropic cytokine with both pro- and anti-inflammatory properties (4) that is overexpressed at the margin of gastric ulcer in H. pylori-positive gastritis (5, 6). Its levels are high in H. pylori-infected early gastric cancer and fall significantly after the cure of H. pylori (7). Considering that activated macrophages are the main sources of IL-6, it is necessary to understand the effectors of H. pylori driving IL-6 gene induction in macrophages. Inflammation-associated factors, such as TNF-α, platelet-derived growth factor, and bacterial endotoxins, all enhance IL-6 gene expression.

TLRs play central roles in innate immunity by recognition and discrimination of specific conserved patterns of molecules derived from bacteria, fungi, or viruses (8, 9, 10). Activation of TLRs results in stimulation of signaling pathways widely involving recruitment of various adaptor molecules such as MyD88 (11, 12, 13), followed by the serine/threonine kinase IL-1R-associated kinase 1 (IRAK1)³ (14). IRAK1 becomes phosphorylated, dissociates from the complex, and associates with TNFR-associated factor 6 (15, 16), finally leading to the activation of MAPKs, transcription factors such as NF-κB, and concomitant production of cytokines (17, 18). The role of TLRs in the response of macrophages during H. pylori infection has not been studied extensively. TLR4 expression has been reported to be up-regulated in gastric biopsies obtained from H. pylori-positive patients compared with uninfected controls (19).

H. pylori has a large repertoire of secreted proteins, including the best studied virulence factors VacA and CagA. However, the clinical outcome of the disease does not necessarily correlate with the absence or presence of VacA or CagA, making it important to identify other factors that could modulate the clinical course of the disease. Till date, urease, and heat shock protein 60 of H. pylori have been reported to induce IL-6 production in macrophages (20, 21). By two-dimensional gel electrophoresis, followed by mass spectrometric analysis, Kim et al. (22) have demonstrated the secretion of HP0175, a peptidyl prolyl cis-, trans-isomerase (PPIase), in the supernatant when H. pylori is grown in vitro. HP0175 is one of the highly and consistently reactive Ags recognized by the sera of H. pylori-infected patients (23, 24). It would therefore not be unlikely for this Ag to have a role in the pathogenesis of H. pylori-associated disease. Our previous studies have identified HP0175 as a TLR4-interacting protein (25). Because TLR4 is one of the best studied receptors driving the innate immune response, we reasoned that it would be worthwhile to explore the effects of HP0175 on macrophage cytokine induction. The study described in this work provides evidence that H. pylori HP0175 induces the release of IL-6 from human macrophages in a TLR4-/MAPK-dependent manner by activating NF-κB-driven IL-6 gene transcription. The contribution of HP0175 in the release of IL-6 was confirmed by the observation that an isogenic mutant of H. pylori 26695 disrupted in the HP0175 gene was impaired in its ability to release IL-6. Immunodepletion of HP0175 from the aqueous extract of H. pylori also led to the inhibition of IL-6-releasing ability and further confirmed the novel role of this protein in the induction of cytokine production from macrophages.

Chromatin remodeling represents a determining factor controlling binding of transcription factors and the formation of preinitiation complexes. It therefore dictates selective induction of subsets of genes. Our studies present evidence of HP0175 triggering mitogen- and stress-activated kinase 1 (MSK1)-dependent phosphorylation of p65 (RelA) and histone H3. MSK1-mediated chromatin modifications most likely play a central role in the HP0175-dependent transcriptional activity of NF-κB and subsequent release of IL-6.

Wortmannin, SB203580, U0126, and PMA were products of EMD Biosciences. Anti-FLAG Ab, polymyxin B, and CREBTIDE were products of Sigma-Aldrich. Anti-p38 MAPK, anti-ERK1/2, and all phospho-specific Abs were from Cell Signaling Technology. Anti-p85, anti-MSK1, and anti-TLR4 Abs were from Santa Cruz Biotechnology. The human IL-6 ELISA kit, anti-Ras, and anti-Rac1 were from BD Biosciences. The human TLR4-neutralizing Ab and p65 ELISA kit were from Imgenex India. Antagonistic TLR2 Ab (T2.5) was from eBioscience. [γ-³²P]ATP was from Jonaki.

cDNAs for p65 and TLR2 were obtained by RT-PCR from a total RNA preparation from THP-1 cells and cloned in pCDNA3.1 and pFLAG-CMV-6a, respectively. For generating dominant-negative (dn) TLR2, a 15-aa deletion from the C terminus was conducted. p65(S276A) was generated by site-directed mutagenesis. TLR4 (1–643) (TLR4-dn) has been described previously (25). The extracellular domain (ECD) of human TLR4 (amino acids Met¹ to Lys⁶³¹) (26) was cloned in pFLAG-CMV 6a or in pET 28a. p1168 hu.IL-6-luc, a reporter construct that codes for the human IL-6 promoter-luciferase in pGL3-basic or mutants carrying mutations at the binding sites of NF-κB, AP1, CRE, NF-IL-6, ETS, and C/EBP individually were obtained from Laboratorium voor Moleculaire Biologie Plasmiden. MyD88 (152–196) (MyD88-dn) was obtained from M. Muzio and A. Mantovani (Mario Negri Institute for Pharmacological Research, Milan, Italy). FLAG-tagged JNK and its dn mutant JNK (T183A/Y185F) were obtained from R. Davis (University of Massachusetts Medical School, Worcester, MA). dn (K63W) TAK1 was a gift from K. Matsumoto (Nagoya University, Nagoya, Japan); wild-type and dn (D195A) MSK1 in pCMV-FLAG were gifts from D. Alessi (University of Dundee, Dundee, U.K.). dn IκBα (S32A, S36A) and dn Ras were purchased from BD Biosciences.

This was conducted by expressing His-tagged HP0175 in Escherichia coli BL21 (DE3), followed by purification of His-HP0175 from the lysates, as described earlier (25). Purified protein was used at a concentration of 1 μg/ml for treatments, unless otherwise stated. Constructs for expression of truncated versions of HP0175 encompassing amino acid residues 1–154 and 154–299 were generated by PCR amplification of required portions of the HP0175 gene and cloning in pET28a. Aqueous extract of H. pylori was prepared, as described earlier (27). HP0175 was depleted from the aqueous extract by incubation with anti-HP0175 Ab or with preimmune serum (as a control), followed by an additional incubation with protein A/G agarose to immunoprecipitate HP0175. The supernatant was used to study the release of IL-6 from THP-1 cells. For the expression of His-HP0175 in HEK293 cells, the gene encoding HP0175 was cloned between the XbaI and EcoRI sites of pcDNA myc-His, followed by transfection in HEK293 cells. HP0175 was purified from the cell lysate by chromatography on Ni²⁺-NTA agarose.

HP0175 was amplified by PCR using the primer pairs 5′-GGGGTACCATGAAAAAAAATATCTTAAA-3′ (sense) and 5′-GAAGATCTTTACTTGTTGATAACAATT-3′ (antisense). The resulting PCR product was cloned between the KpnI and BglII sites of the shuttle vector pHel2 (a gift from R. Haas, Max von Pettenkofer-Institut für Hygiene und Medizinische Mikrobiologie, Munich, Germany) (28). The shuttle plasmid was introduced into the knockout (KO) strain (described in Ref. 25) by natural transformation. Colonies were selected on plates containing 4 μg/ml chloramphenicol and 10 μg/ml kanamycin.

THP-1 and HEK293 cells were obtained from the National Centre for Cell Science (Pune, India). THP-1 was maintained in RPMI 1640 medium and treated with PMA to induce maturation of the monocytes to a macrophage-like adherent phenotype, as described (27). Blood was drawn from healthy adult volunteers, and PBMC were isolated, as described (27).

For treatments with bacteria, H. pylori strains were grown, as described (25), and incubated with PMA-treated THP-1 cells at a multiplicity of infection 50–100 on culture plates. HEK293 cells were grown in MEM supplemented with 10% FBS; 6 × 10⁵ cells/well were transfected with Polyfect complex, according to the manufacturer’s instructions. For THP-1 cells (4 × 10⁵/well), transfection was conducted with 2 μg of plasmid (empty vector or recombinant) using Fugene 6 (Roche), according to the manufacturer’s protocol.

Transfected or untransfected cells were either left untreated or treated with different inhibitors, followed by incubation with recombinant purified protein HP0175 (1 μg/ml) (or with bacteria). The conditioned medium was removed and assayed for IL-6 by ELISA using the human IL-6 assay kit.

Proteins were separated by SDS-PAGE and then transferred electrophoretically to polyvinylidene difluoride membranes. Western blotting was done, as described earlier (25).

Binding of HP0175 with TLR4-ECD was conducted by an ELISA-like binding assay. ELISA plates were coated with increasing concentrations of HP0175 purified from HEK293 cells. The plates were blocked with 1% BSA in PBS containing 1% Tween 20 for 1 h and incubated with FLAG-TLR4- or FLAG-TLR2-ECD at a concentration of 1 μg/ml for 2 h. After three washes, anti-FLAG mAb (1:5000) was added and incubated for 1 h. Wells were again washed, incubated with HRP-linked anti-mouse Ab, followed by the chromogenic substrate tetramethylbenzidine. Reactions were stopped with H₂SO₄, and bound TLR-ECD was determined spectrophotometrically at 450 nm. In a separate set of experiments, plates were first coated with TLR4- or TLR2-ECD (20 μg/ml), followed by blocking, as described above. Myc-tagged HP0175 was added at different concentrations, and bound HP0175 was determined using anti-myc mAb, as described above. Each experiment was performed in triplicate.

A total of 100 ng of RNA was reverse transcribed using the Titanium One-Step RT-PCR Kit (BD Biosciences). The primers 5′-GTACCCCCAGGAGAAAGATTCC-3′ (sense) and 5′-CAAACTGCATAGCCACTTTCC-3′ (antisense) were used to amplify 819 bp of IL-6 mRNA. The gapdh was amplified using the primers 5′-CCA TCA ATG ACC CCT TCA TTG ACC-3′ (sense) and 5′-GAA GGC CAT GCC AGT GAG CTT CC-3′ (antisense) to generate a 604-bp product. The PCR conditions for IL-6 mRNA were denaturation at 94°C for 30 s, annealing at 50°C for 1 min, and extension at 68°C for 1 min, for 35 cycles.

Ras and Rac1 activities were measured in cell extracts by assaying for activated Ras or Rac1 bound to immobilized Ras binding domain of Raf-1 (29) or immobilized p21 binding domain of PAK1, respectively (30), expressed as GST fusion proteins.

THP-1 cells were transfected with luciferase reporter plasmid along with β-galactosidase reporter construct, and promoter activation was analyzed by luciferase activity assays, as described (27).

MSK1 was immunoprecipitated from cell lysates by incubation with anti-MSK1 Ab, and immunoprecipitates were incubated with 30 μM substrate CREBTIDE and 0.1 mM [γ-³²P]ATP (20,000 cpm/pmol). The incorporation of phosphates in CREBTIDE was determined using p81 phosphocellulose paper (31).

Cells were transfected with 5–25 nM MSK1 Kinase ShortCut siRNA mix or Lit28i polylinker control ShortCut siRNA mix (New England Biolabs) using TransPass R2 SiRNA transfection reagent, following the manufacturer’s protocol. Silencing of MSK1 was confirmed by Western blotting of cell lysates with anti-MSK1 Ab.

After treatments, cells were washed with ice-cold PBS and lysed, as described above. Extracts were centrifuged at 1,500 × g and pellets were extracted with 0.4 N HCl for 1 h on ice. Extracts were centrifuged at 10,000 × g for 15 min at 4°C, and histones were precipitated with chilled acetone overnight at −20°C. Precipitated proteins were washed with acetone, dried, reconstituted, and separated by SDS-PAGE, followed by immunoblotting with Abs specific for phospho-Ser¹⁰ histone H3.

ChIP assay was conducted using the ChIP assay kit from Upstate Biotechnology. Briefly, cells after treatments were fixed with 1% formaldehyde; washed with PBS; resuspended in 1% SDS, 50 mM Tris, and 10 mM EDTA; and sonicated to generate DNA fragments with an average size of 1 kb. After centrifugation at 13,000 rpm for 10 min, the cell supernatant was diluted with ChIP dilution buffer, followed by preclearing of the samples with salmon sperm DNA/protein A/G agarose slurry for 30 min at 4°C. The supernatant was collected and incubated with appropriate primary Ab or no Ab overnight at 4°C. Salmon sperm/protein A/G agarose beads were added and incubated for 1 h at 4°C, and the beads were washed successively with low salt, high salt, and LiCl immune complex wash buffers (supplied by the manufacturer). Bound complexes were eluted with the above buffer containing 0.5% SDS at 60°C for 15 min. Eluates were decross-linked by incubation at 65°C for 12 h and digested with proteinase K. DNA was purified by phenol/chloroform extraction and ethanol precipitation in the presence of glycogen and analyzed for IL-6 by PCR using the primers 5′-GGCAAACCTCTGGCACAAGAG-3′ (sense) and 5′-AGGTCGTCATTGAGGCTAGCG-3′ (antisense) (32).

Macrophage-derived cytokine production is strongly up-regulated during H. pylori infection (33). Secreted Ags are candidates for stimulating cytokine induction from macrophages. We therefore evaluated the role of the secreted PPIase, HP0175, in modulating cytokine induction. ELISA showed that HP0175 elicited IL-6 release from PMA-differentiated THP-1 cells in a dose- (Fig. 1,A) and time-dependent (Fig. 1,B) manner. IL-6 release peaked at 24 h. HP0175-induced IL-6 release was not affected by polymyxin B treatment (Fig. 1,A), ruling out the possibility of the effect being due to LPS contamination. Heat treatment led to a loss of IL-6-inducing ability of the recombinant protein (Fig. 1,A), supporting the view that IL-6 induction was most likely due to the HP0175 protein. In addition, to rule out the contribution, if any, of bacterial products in the observed IL-6-inducing ability of HP0175, the protein was purified from HEK293 cells. HP0175 derived from this nonmicrobial source was also able to drive IL-6 induction in PMA-differentiated THP-1 (Fig. 1,A). To narrow down on the domain of HP0175 responsible for IL-6 induction, truncated versions of HP0175 were tested for IL-6-inducing ability. The data shown in Fig. 1 B suggest the importance of the C-terminal domain (V154-K299) in IL-6 induction. This domain also showed PPIase activity, whereas the N-terminal domain did not (data not shown).

An isogenic mutant (KO) of H. pylori 26695 inactivated in the HP0175 gene was attenuated in its ability to affect IL-6 release from THP-1 cells (Fig. 1,C). Complementation of the KO strain with the shuttle vector pHel2 harboring the HP0175 gene resulted in restoration of the IL-6-inducing ability (Fig. 1,C). An aqueous extract (HPE) of the wild-type strain (containing shed factors of the bacterium) was capable of driving IL-6 release from PMA-differentiated THP-1 (Fig. 1,D). IL-6-inducing ability of the HPE was attenuated in the KO strain (Fig. 1,D), suggesting that HP0175 was one of the major components responsible for inducing IL-6 release. In harmony with this, IL-6-inducing ability was compromised in HP0175-immunodepleted extracts of the wild type (Fig. 1 D). Taken together, these results suggested that HP0175 is a potent IL-6 inducer.

The TLR signaling pathway plays a role in regulation of cytokine production through a signaling cascade leading to the activation of NF-κB (12, 13). Pretreatment of cells with neutralizing Abs against TLR4 (but not against TLR2) before challenge with HP0175 blocked IL-6 release (Fig. 2,A). Transfection of cells with dn TLR4 (but not with dn TLR2) also inhibited HP0175-induced IL-6 release (Fig. 2,B). Furthermore, HP0175 could induce IL-6 release from HEK293 cells transfected with TLR4, but not from cells transfected with either vector or with TLR2 only (Fig. 2,C). We evaluated the role of HP0175-dependent TLR4 signaling in IL-6 induction in the context of the bacterium as a whole. H. pylori-mediated IL-6 induction was attenuated in cells treated with neutralizing Abs against TLR4 (but not TLR2) (Fig. 2 D). This suggested that TLR4 signaling plays a major role in H. pylori-mediated IL-6 induction in macrophages.

We have reported previously that immobilized HP0175 is capable of immunoprecipitating TLR4 from lysates of AGS cells (25). To establish that HP0175 and TLR4 interact directly without the need for any accessory molecule(s), the binding between the ECD of TLR4 and HP0175 was analyzed. ELISA plates were coated with increasing concentrations of purified HP0175. TLR4-ECD could bind to E. coli-derived HP0175-coated plates, and the binding increased with increasing concentrations of HP0175 (Fig. 3,A). In contrast, the TLR2-ECD did not bind to HP0175-coated wells, supporting the specificity of the HP0175-TLR4 interaction. The ECD of TLR4 did not bind to wells coated with an irrelevant His-tagged protein (data not shown). In a second ELISA-like assay, wells were coated with the ECD of TLR4 (or TLR2), and the binding of HP0175 to the coated wells was studied. HP0175 could bind to wells coated with TLR4 in a dose-dependent manner (Fig. 3 B), but not to wells coated with TLR2. These ELISA-like assays confirmed that HP0175 interacts with the ECD of TLR4 without the need for an accessory molecule. Similar results were obtained using HP0175 purified from HEK293 (data not shown), ruling out the possibility of microbial products serving as intermediary molecules in this interaction.

In the canonical pathway of NF-κB activation, TLR4 signals along the MyD88/IRAK/TNFR-associated factor 6 axis, leading to the phosphorylation of IκBα, its degradation by the proteasome, release of NF-κB, and movement of NF-κB into the nucleus. Transfection with dn constructs of MyD88, or the superrepressor IκBα (S32A, S36A) (IκBα (dn)), led to an inhibition of HP0175-induced IL-6 release, confirming the likely role of the canonical pathway of NF-κB activation in HP0175-mediated IL-6 release (Fig. 2 E).

MAPKs are known to regulate the upstream signaling events that control cytokine production. U0126 and SB203580, inhibitors of the ERK and p38 MAPK signaling pathways, respectively, could inhibit HP0175-induced IL-6 release from THP-1 cells as well as PBMCs (Fig. 4). A combination of U0126 and SB203580 led to a complete inhibition of IL-6 release (Fig. 4), suggesting a synergistic role of ERK and p38 MAPKs in HP0175-triggered IL-6 release. Neither dn JNK nor the JNK inhibitor SP600125 inhibited HP0175-induced IL-6 release (data not shown), suggesting that JNK signaling was not involved in IL-6 release. Besides their ability to phosphorylate and activate transcription factors, both ERK and p38 MAPKs transduce signals by activating downstream kinases such as MSK1 (34), which in turn phosphorylates transcription factors. PBMCs or THP-1 cells pretreated with H89 (a pharmacological inhibitor of MSK1 at the specific dose used in this study) led to a complete inhibition of IL-6 release, suggesting a role of MSK1 in IL-6 release (Fig. 4).

Semiquantitative RT-PCR analysis for IL-6 mRNA expression showed that the steady state IL-6 mRNA levels increased in cells treated with HP0175 (Fig. 5 A). HP0175-induced IL-6 gene expression could be attenuated partially either by U0126 or by SB203580 alone, and inhibited almost completely by a combination of U0126 and SB203580, suggesting that IL-6 gene expression is coordinately regulated by ERK and p38 MAPKs.

THP-1 cells were transfected with different constructs of the human IL-6 promoter coupled with the luciferase reporter gene. Treatment of cells with HP0175 resulted in a ∼20-fold increase in luciferase activity in cells transiently expressing the IL-6 promoter (p1168 hu.IL-6-luc) (35) compared with untreated cells (Fig. 5,B). THP-1 cells were also transfected with IL-6 promoter luciferase reporter constructs mutated individually at NF-κB, AP-1, C/EBP/β, CRE, and ETS sites (36), followed by stimulation with HP0175 for 24 h. Mutation at only the NF-κB site completely inhibited luciferase gene expression (Fig. 5,B), suggesting that the NF-κB element was necessary to drive HP0175-induced IL-6 gene expression. U0126 or SB203580 could partially inhibit luciferase gene expression, whereas a combination of both inhibitors completely inhibited luciferase gene expression (Fig. 5 C), suggesting that ERK and p38 MAPK activation are critical in sustaining HP0175-driven IL-6 promoter activation.

HP0175-triggered IL-6 release in PBMCs and THP-1 was inhibited by wortmannin, an inhibitor of PI3K (Fig. 4), and HP0175 was found to stimulate PI3K activity, as evidenced by the phosphorylation of the p85 subunit of PI3K (Fig. 6,A). This was abrogated when cells were pretreated with neutralizing anti-TLR4 Ab (Fig. 6,A), suggesting TLR4-dependent activation of PI3K. Wortmannin inhibited HP0175-triggered Ras (Fig. 6,B) and Rac1 activation (Fig. 6,C). Our results therefore suggested that HP0175 activates PI3K/Ras and PI3K/Rac1 signaling. HP0175-stimulated ERK activation was inhibited by dn Ras (Fig. 6,D), whereas p38 MAPK activation was partially inhibited by dn Rac1 (Fig. 6 E). Taken together, our results suggested that HP0175 signaling induced TLR4/PI3K/Ras-dependent ERK activation and TLR4/PI3K/Rac1-dependent p38 MAPK activation.

The MAPK kinase kinase TAK1 is known to phosphorylate nuclear factor-inducing kinase, MAPK kinase 3/6, and MAPK kinase 4 to regulate the NF-κB and the MAPK signaling pathways (37, 38). Kinase-dead TAK1 (K63W) inhibited HP0175-induced IL-6 release (Fig. 2,E). TAK1 K36W inhibited p38 MAPK activation partially (Fig. 6,E), suggesting that TAK1 signaling lies upstream of p38 MAPK. Cotransfection with TAK1 K36W and dn Rac1 lead to a greater inhibition of HP0175-induced p38 MAPK activation than observed using either construct alone (Fig. 6 E), suggesting that distinct PI3K/Rac1 and TAK1 signaling pathways converged upon p38 MAPK activation.

The MAPKs phosphorylate several downstream kinases, one of which is MSK1, which is constitutively localized in the nucleus, where it regulates transcription by phosphorylating among others, histone H3, p65/RelA, and CBP/p300 (39), leading to nucleosomal modification and increased accessibility of the transcription machinery to promoters. HP0175 activated MSK1 in a time-dependent manner (data not shown). MSK1 activity could be partially inhibited by pretreatment with either U0126 or SB203580 (Fig. 7,A). A combination of U0126 and SB203580 completely blocked MSK1 activation to the extent observed using H89, suggesting that ERK and p38 MAPK act synergistically to activate MSK1. Expression of MSK1 was effectively inhibited by siRNA directed against MSK1 (Fig. 7,B). Under these conditions, HP0175-induced IL-6 release was inhibited (Fig. 7 C), lending further support to the view that MSK1 plays a central role in HP0175-induced IL-6 release from THP-1.

Recruitment of NF-κB to its target depends on the integration of signals coming from stimulus-dependent activation of multiple signaling pathways, leading to covalent modifications of NF-κB that regulate its transcriptional activity. Protein kinase A and MSK1 phosphorylate p65 on Ser²⁷⁶ (40), enabling the p50-p65 complex to interact with the transcriptional coactivators p300 and CBP (41). HP0175-induced p65 phosphorylation was inhibited in cells transfected with dn MSK1 (Fig. 7,D). At the same time, transfection of cells with a p65 construct bearing the S276A mutation inhibited HP0175-stimulated IL-6 release (Fig. 2 E), suggesting a role of p65 phosphorylation on S276 in activating IL-6 production.

Covalent modification of histones by acetylation, phosphorylation, and methylation regulates transcription of a subset of genes in a context-specific manner (32). MSK1 phosphorylates histone on Ser¹⁰ (42). HP0175-induced H3 Ser¹⁰ phosphorylation was inhibitable by dn MSK1 (Fig. 7 E), affirming the role of MSK1 in H3 Ser¹⁰ phosphorylation.

To delineate the relationship of phosphorylation of H3 and p65 to IL-6 gene transcription, ChIP assays were performed. Immunoprecipitation with an Ab specific for p-Ser¹⁰ H3 after stimulation of THP-1 cells with HP0175 resulted in reproducible enrichment of IL-6-specific gene sequence in stimulated vs unstimulated samples by PCR (Fig. 7,F), affirming H3 phosphorylation at Ser¹⁰ on the nucleosomes distributed along the IL-6 promoter. Anti-phospho-Ser¹⁰ H3 immunoprecipitates from cells transfected with dn MSK1 contained greatly reduced amounts of promoter sequences (Fig. 7,F), confirming that MSK1 plays a critical role in nucleosomal modification at the IL-6 promoter. When Ab was omitted from the immunoprecipitation reaction, no promoter sequence was retrieved (data not shown), confirming the specificity of the reaction. p65-transfected cells when stimulated with HP0175 showed enrichment of IL-6-specific sequence in ChIP assays with p65-specific Ab (Fig. 7 G). Reduced amounts of promoter sequence were immunoprecipitated in cells transfected with p65(S276A), affirming that phosphorylation of p65 on Ser²⁷⁶ promotes its recruitment to the IL-6 promoter.

H. pylori infection leads to chronic inflammation and a Th1-skewed immune response with production of proinflammatory cytokines, one of which is IL-6. The role of TLRs in the pathophysiology of Helicobacter infection is a matter of debate. Most of the reports have focused on epithelial cells. Some groups have argued that the response of epithelial cells to H. pylori is TLR2, but not TLR4 dependent (43, 44). Others have reported that TLR4 expression is enhanced in the gastric epithelium following H. pylori infection (19) and that enhanced TLR4 expression stimulates NF-κB-driven IL-8 promoter activity (45). It is likely that TLRs 2, 4, and 5 all have a role in the context of H. pylori infection. It is pertinent to point out that our studies have focused not on gastric epithelial cells, but on cells of the monocytic lineage. Considering that mononuclear cell infiltration in the lamina propria and increased expression of inflammatory cytokines characterize H. pylori-induced chronic infection (33), it is of obvious importance to elucidate the effectors of H. pylori that drive the macrophage response to gain insight into how inflammation is induced by H. pylori. The present study is a step toward this end. TLR4-mediated signals play a major role in eliciting inflammatory cytokines from macrophages. TLR4-interacting molecules are likely to deliver signals of consequence in disease progression and outcome.

Based on our earlier observations that HP0175 interacts with TLR4 (25), we investigated the probable involvement of TLR signaling pathways in HP0175-induced IL-6 production in THP-1 cells. rHP0175 expressed in E. coli elicited IL-6 release from THP-1 and from PBMCs. In view of the fact that several reports in the past on the effects of recombinant proteins were later found to be due to contaminating bacterial products, care was taken to rule out this possibility. Contaminating LPS was removed by treating recombinant protein preparations with polymyxin-agarose. HP0175 was also purified from the HEK293 cell line. Both preparations were found to induce IL-6 release from THP-1, ruling out the possibility of contaminating bacterial products being responsible for the observed effect. This view was further strengthened by our observation that inactivation of the HP0175 gene led to >70% reduction in IL-6-inducing ability of H. pylori, and that significant reduction of IL-6-inducing ability occurred when HP0175 was immunodepleted from aqueous extracts of H. pylori. This suggested that HP0175 is one of the major IL-6-inducing factors. ELISA-like binding assays demonstrated that HP0175 (purified either from E. coli or from HEK293) could interact directly with TLR4, ruling out the requirement of accessory molecules or copurified bacterial products in this interaction. In vitro analysis has shown that HP0175 is secreted by H. pylori rather than being released by nonspecific lysis (22). Assuming that HP0175 release also occurs in vivo, it is likely to be a factor of potential importance in the pathophysiology of H. pylori infection, because HP0175 would be likely to reach mucosal macrophages after the disruption of the epithelial cell junction during infection.

HP0175-triggered TLR4-dependent signaling pathways activated the MAPKs ERK and p38 MAPK in a PI3K/Ras, Rac1-dependent manner. This in turn activated MSK1, a nuclear kinase central to HP0175-driven NF-κB activation (Fig. 8). NF-κB activation was necessary for driving HP0175-induced IL-6 gene expression. Considering that MSK1 has been reported to phosphorylate histone H3 as well as the p65 subunit of NF-κB, we conjectured that MSK1 was most likely influencing HP0175-induced chromatin modifications at the IL-6 promoter, thereby regulating IL-6 gene expression. HP1075-stimulated histone H3 phosphorylation on Ser¹⁰ and p65 phosphorylation were dependent on MSK1, and ChIP assays showed that association of p65 and phospho-Ser¹⁰ H3 with the IL-6 promoter was dependent on MSK1. MSK1 therefore plays a central role in TLR4-dependent IL-6 induction elicited by HP0175 in macrophages.

Based on drug specificity, PPIases have been divided into the cyclosporin A-binding cyclophilins, the FK506-binding proteins, and the parvulin-like PPIases that do not bind immunosuppressants (46). There is evidence that PPIases contribute to bacterial infection and virulence. The Mip protein of Legionella pneumophila, a member of the FK506-binding protein family, is involved in its entry into host cells and intracellular replication (47). The streptococcal cyclophilin SlrA is involved in pneumococcal colonization (48). However, there is no direct evidence of the requirement of PPIase activity in the interaction of these proteins with host cells. HP0175 represents a parvulin-like PPIase of relevance to interaction of H. pylori with host cells. The domain encompassing amino acid residues 154–299 is responsible for TLR4 binding as well as for PPIase activity. However, at present there is no evidence to link PPIase activity to TLR4-interacting ability. Systematic mutagenesis of HP0175 is underway to address this.

In summary, we describe a novel role of a secreted parvulin family PPIase of H. pylori in IL-6 induction in macrophages. HP0175-triggered TLR4 signaling leads to IκBα-mediated activation of NF-κB (Fig. 8). At the same time, TLR4-dependent MSK1 activation results in chromatin modification involving phosphorylation of histone H3 and NF-κB p65. This most likely generates a transcription factor code that turns on transcription of NF-κB-dependent transcription of IL-6 and possibly other, yet to be pinpointed genes. The findings are important in the context of H. pylori infection because such transcription factor codes generated by specific bacterial virulence factors are likely to contribute to the ultimate outcome of the infection.

We thank all of the scientists who provided the plasmids used in this study.

The authors have no financial conflict of interest.