Activation of Lipid Metabolism Contributes to Interleukin-8 Production during Chlamydia trachomatis Infection of Cervical Epithelial Cells

Abstract

Chlamydia trachomatis infection is the most common cause of bacterial sexually transmitted diseases. Infection of the urogenital tract by C. trachomatis causes chronic inflammation and related clinical complications. Unlike other invasive bacteria that induce a rapid cytokine/chemokine production, chlamydial infection induces delayed inflammatory response and proinflammatory chemokine production that is dependent on bacterial growth. We present data here to show that the lipid metabolism required for chlamydial growth contributes to Chlamydia-induced proinflammatory chemokine production. By gene microarray profiling, validated with biochemical studies, we found that C. trachomatis LGV2 selectively upregulated PTGS2 (COX2) and PTGER4 (EP4) in cervical epithelial HeLa 229 cells. COX2 is an enzyme that catalyzes the rate-limiting step of arachidonic acid conversion to prostaglandins, including prostaglandin E2 (PGE2) and other eicosanoids, whereas EP4 is a subtype of cell surface receptors for PGE2. We show that Chlamydia infection induced COX2 protein expression in both epithelial cells and peripheral blood mononuclear cells and promoted PGE2 release. Exogenous PGE2 was able to induce interleukin-8 release in HeLa 229 epithelial cells. Finally, we demonstrated that interleukin-8 induction by Chlamydia infection or PGE2 treatment was dependent on extracellular signal-regulated kinase/mitogen-activated protein activity. Together, these data demonstrate that the host lipid remodeling process required for chlamydial growth contributes to proinflammatory chemokine production. This study also highlights the importance of maintaining a balanced habitat for parasitic pathogens as obligate intracellular organisms.

Chlamydiae are gram-negative, obligate intracellular bacterial parasites that infect and multiply within a broad range of eukaryotic cells, including macrophages, smooth muscle, epithelial, and endothelial cells (12). As the common source for bacterial sexually transmitted diseases, Chlamydia trachomatis serovars D to K cause urogenital infection, which can evolve into chronic infections resulting in female pelvic inflammatory disease, tubal blockage, and infertility. Ocular infection by C. trachomatis serovars A to C yields the most preventable blindness worldwide. The lymphogranuloma venereum (LGV) serovars L1 to L3 cause disseminated infection of the regional lymph nodes, leading to lymphadenopathy. The pathogenesis of chlamydial diseases is unknown; however, a common feature of the infection is a damaging chronic inflammatory response to the parasite by the host (37, 43). Chlamydia infection of nonimmune cells produces proinflammatory factors such as interleukin-1α (IL-1α), IL-6, IL-8, IL-11, GROα, and granulocyte-macrophage colony-stimulating factor (6, 31, 37), which can lead to an acute inflammatory response characterized by neutrophil infiltration to the primary sites of infection, followed by a subepithelial accumulation of mononuclear leukocytes during the chronic phase of infection (28, 37, 43). These cellular responses promote cellular proliferation and tissue damage of affected organs (37).

The mechanisms that modulate host responses to Chlamydia infection have not been well defined. Most invasive bacterial pathogens often induce rapid, but transient, responses with microbial products such as lipopolysaccharide (LPS), peptidoglycan, or unmethylated DNA (22, 40). In contrast, chlamydial components such as LPS seem to play a less important role in the proinflammatory responses of epithelial cells (15, 18, 30, 31), the primary target of C. trachomatis infection. Furthermore, the production of proinflammatory factors in these cells is delayed and is dependent on bacterial replication (31). The obligate intracellular bacterial pathogen has evolved a unique biphasic life cycle of infection (20, 23, 36). After initial attachment and entry into the host cell, the infectious elementary body (EB) differentiates into the metabolically active reticulate body (RB) for replication approximately 8 to 12 h into the infection process. Mature RBs redifferentiate into the infectious EBs for eventual release (23, 36). Chlamydial replication involves trafficking of eukaryotic lipids (14, 39, 44). Purified chlamydiae contain glycerophospholipids, such as phosphatidylcholine and phosphatidylinositol, derived from eukaryotes (3, 14, 25, 44). However, these lipids differ from the eukaryotic glycerophospholipids at the sn-2 position. The straight-chain fatty acid of mammalian origin, commonly arachidonic acid, is replaced with a bacterially derived branched-chain fatty acid, a process modulated by the cytosolic phospholipase A2 (cPLA2) (14, 39, 44). As a member of the phospholipase A2 family proteins, cPLA2 catalyzes the release of arachidonic acid from membrane-bound phospholipids (5, 10, 13) and the production of lysophospholipids for bacterial uptake. The lipid mediators, including lysophopholipids and biochemical metabolites of arachidonic acid, are potent inflammatory stimuli (8, 10, 27, 34), although their role in Chlamydia-induced inflammatory cytokine production has not been investigated.

During the DNA microarray studies of host response to Chlamydia infection, we found that Chlamydia infection activated a lipid metabolism pathway for prostaglandin biosynthesis and prostaglandin E2 (PGE2) detection. Specifically, infection of HeLa 229 cervical epithelial cells with C. trachomatis upregulated prostaglandin-endoperoxide synthase 2 (PTGS2; also known as cyclooxygenase 2 [COX2]) and prostaglandin E receptor 4 (PTGER4 [EP4]), one of the four subtypes of PGE2 receptor (7, 21, 33). COX2 protein is a critical enzyme involved in the biosynthesis of prostanoids such as the proinflammatory PGE2 from arachidonic acid (42). We report here the identification and characterization of COX2 upregulation and PGE2 production by Chlamydia infection.

MATERIALS AND METHODS

Cell lines, Chlamydia organisms, and reagents.

The cervical epithelial adenocarcinoma HeLa 229 (CCL-2.1), vulva epidermoid carcinoma A431 (CRL-1555), laryngeal epidermoid carcinoma HEp-2 (CCL-23), and lung epithelial carcinoma A549 (CCL-185) cell lines were from the American Type Culture Collection (Manassas, VA), and the cells were maintained in Dulbecco modified Eagle medium (DMEM; Invitrogen, Carlsbad, CA) supplemented with 10% heat-inactivated defined fetal bovine serum (endotoxin < 10 IU/ml; HyClone), 2 mM nonessential amino acids, and 2 mM l-glutamate (Invitrogen). The C. trachomatis serovar L2 (434/Bu) was kindly provided by Guangming Zhong (University of Texas); serovar D and C. pneumoniae were provided by Harlan Caldwell (Rocky Mountain Laboratories, National Institutes of Health). The mitogen-activated protein (MAPK) antibody sampler kit (extracellular signal-regulated kinase [Erk], c-Jun N-terminal kinase [JNK], p38, and the corresponding phospho-proteins), anti-cPLA2, and phospho-cPLA2 (S505) were from Cell Signaling (Beverly, MA). Anti-COX2 antibody, PGE2, and LTB4 were from Cayman (Ann Arbor, MI), and the chemical inhibitors were from CalBiochem (La Jolla, CA) or Cell Signaling. The synthetic lipids were purchased from Avanti Polar Lipids (Alabaster, AL). The enzyme-linked immunosorbent assay (ELISA) kits for IL-6, IL-8, and PGE2 detection were from R&D Systems (Minneapolis, MN), and the Inflammation Antibody Array III kit was from RayBiotech (Norcross, GA). The fluorescent glycerophospholipid BODIPY FL C₅-HPC (D3803) was purchased from Invitrogen.

Chlamydia strains.

C. trachomatis serovars L2/434/Bu and D/UW3/Cx and C. pneumoniae were propagated in HeLa 229 cells as described previously (39). Briefly, cultures were grown in DMEM supplemented with 60 μg of vancomycin/ml and 20 μg of gentamicin/ml at 37°C. Cultures infected with C. trachomatis L2 were grown for 48 h, and those infected with C. trachomatis serovar D or C. pneumoniae were grown for 72 h in the presence of 1 μg of cycloheximide/ml. Infected monolayers were detached by scraping and sonicated to lyse the host cells. Cellular debris was removed by differential centrifugation. Chlamydial EBs were pelleted, resuspended in an isotonic sucrose-phosphate-glutamate buffer, and frozen at −80°C. Infectious titers were determined by titration on HeLa 229 cell monolayers and stained with a fluorescein isothiocyanate-labeled monoclonal antibody against chlamydial LPS (Meridian Diagnostics, Inc., Cincinnati, OH) and are expressed in inclusion-forming units (IFU).

Isolation of PBMC.

Human peripheral blood mononuclear cells (PBMC) were isolated from heparinized blood by separation on Ficoll density gradients. The blood was collected from healthy donors at the General Clinical Research Center at Scripps by using an institutionally approved human subject protocol. The whole blood was first diluted with equal volumes of HEPES-buffed saline (HBS, pH 7.4) and then overlaid onto 12 ml of Ficoll Histopaque-1077 (Sigma, St. Louis, MO) per 35 ml of diluted blood. The mononuclear cell layer was collected after spun at 1,600 rpm for 30 min in a Sorvall RT6000 centrifuge and washed with 4 volumes of HBS three times prior to lysing the remaining red blood cells in a buffer containing 150 mM NH₄Cl, 10 mM KHCO₃, and 0.1 mM EDTA. Cells were washed a final time with 4 volumes of HBS and resuspended in RPMI 1640 culture medium supplemented with 10% heat-inactivated fetal bovine serum at 10⁶ cells/ml. The cells were cultured at 37°C in a 5% CO₂ humidified incubator and used for infection assays.

Chlamydia infection and immunoblotting assays.

Cells were infected with LGV2 at a multiplicity of infection (MOI) of 1 IFU/cell or as specified in individual experiments. For infections with C. trachomatis serovar D or C. pneumoniae, the cell monolayers were treated with DEAE at 33 μg/ml in HBS prior to infection. The chlamydial inocula were removed 2 h postinfection, and cells were fed with complete DMEM. No cycloheximide was used for these infection assays. For immunoblotting analysis, the infected cells were lysed with a buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 1% Triton X-100, 0.1% sodium dodecyl sulfate, and a cocktail of protease inhibitors for 30 min on ice. The cell lysates were cleaned by centrifugation at 13,000 × g, and the soluble proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. After transfer to a polyvinylidene difluoride membrane, the proteins were detected by incubation with a primary antibody, followed by horseradish peroxidase-conjugated secondary antibody and the ECL reagent (Pierce, Rockford, IL).

Protein array analysis for inflammation cytokines.

HeLa 229 cells in a six-well tissue culture plate were infected with LGV2 at an MOI of 1. At 24 h after infection, the monolayer was rinsed twice with serum-free medium (SFM), and the cells were fed with 1 ml of SFM. The secretion of proteins from infected and uninfected cells was assayed 18 h later by using the RayBio Inflammation Antibody Array III kit (RayBiotech, Norcross, GA) according to the manufacturer's recommendation. This kit detects 40 soluble proteins associated with inflammation, including IL-1 to IL-17 (excluding IL-5), interferons, tumor necrosis factors (TNFs), TNF receptors (TNFRs), ligand 10, macrophage chemoattractant proteins (MCP) 1 and 2, tissue inhibitor of metalloproteinase (TIMP), and intracellular cell adhesion molecules (ICAMs). The sensitivity for these proteins varies from 1 pg/ml for IL-4, -5, -6, -8, -12, and -16 and TIMP-2 to 3 pg/ml for MCP-1 to as high as 1 ng/ml for IL-1α and TNF-β. A complete list of the proteins and the corresponding sensitivities for their detection is available online from the manufacturer (Raybiotech).

ELISAs (PGE2, IL-6, and IL-8).

HeLa 229 epithelial cells were seeded into six-well plates at a density of 10⁶ cells/well and cultured for 24 h. To assay cytokine production in the culture supernatants, cells, in duplicates, were infected with LGV2 at 5 IFU/cell unless otherwise stated. The cell monolayers were rinsed with SFM 2 h after inoculation to remove the free bacterium. The infected cells were fed with SFM. Cytokines or PGE2 in the culture supernatants were detected with ELISA kits. Under the serum-free condition, no IL-8 or PGE2 production was detected from the uninfected controls. For treatment with inhibitors, the compounds were present throughout the infection process.

To examine the effect of lysophospholipids or eicosanoids on IL-8 release, HeLa 229 cells, in duplicates, in a 24-well plate were treated with the individual compound for 24 h. All samples were measured in triplicates, and the concentrations were presented as means ± the standard deviations. The sensitivities of the ELISA were 3 pg for IL-6 and IL-8 and 7 pg for PGE2.

cPLA2 activity assay.

The anti-cPLA2 immunocomplexes were resuspended in 200 μl of Hanks balanced salt solution supplemented with 1 mM Ca²⁺, Mg²⁺, and 1 μM BODIPY FL C₅-HPC (9). After incubation at 37°C for 15 min with occasional vortex, the reaction was stopped by two extractions with 200 μl of methanol-chloroform. The organic layer was dried under vacuum. The total lipids were resuspended in 200 μl of methanol-chloroform (1:1) and applied onto thin-layer chromatography plates (60F₂₅₄; E Merck, Darmstadt, Germany). The plates were developed in xylene-diethyl ether-ethanol-acetic acid (50:40:2:0.2). The image was downloaded with a FluorImager 595 (Molecular Dynamics, Sunnyvale, CA), and the amount of released fluorescent fatty acid was quantified by using the proprietary software.

Transcriptional profile analysis with Affymetrix gene chip.

Total RNA from LGV2-infected or uninfected HeLa 229 cells was isolated with the TRIzol reagent (Invitrogen) and cleaned with Qiagen RNeasy columns. The samples were processed and analyzed at the DNA Microarray Core Facility at The Scripps Research Institute. The integrity of the RNA was verified by using the Agilent Bioanalyzer 2100 system. We used the HG-U133 Plus 2.0 Affymetrix gene chip arrays containing over 47,000 human genes and transcripts. The normalized average intensity value was used to determine the number of positive and negative probe pairs. The differential expression of genes of interests was verified by reverse transcription-PCR (RT-PCR) or fluorescence-based real-time PCR technology (Applied Biosystems, Foster City, CA), with GAPDH (glyceraldehyde-3-phosphate dehydrogenase) as a control.

RESULTS

Chlamydia infection upregulates PTGS2.

Previous studies using DNA microarray techniques have shown that Chlamydia infection upregulates genes of cell proliferation, transcriptional regulation, and inflammation (16, 45). To fully evaluate host responses of cervical epithelial cells to Chlamydia infection, we used the newly released Affymetrix Gene Chip Human Genome HG-U133 plus 2.0. The U133 plus 2.0 array completely covers the human genome and 6,500 additional genes that allows for thorough examination of cellular responses compared to arrays used in previous studies (16, 45). HeLa 229 cells were infected with Chlamydia trachomatis LGV2 at 1 IFU/cell for 16 and 34 h (representative of middle and late stages of infection, respectively). Total RNA was extracted from C. trachomatis-infected and uninfected control cells and subjected to microarray analysis. Consistent with previous reports that Chlamydia infection upregulated host gene expression, including genes associated with cell cycle, cell proliferation, transcription regulation, and inflammation (6, 16, 45), we also observed strong induction and activation of genes of JAK/STAT signal transduction at both the middle and late stages of Chlamydia infection. In addition, we observed selective upregulation of PTGS2 (COX2) and PTGER4 of prostanoid biosynthesis and PGE2 detection. PTGS2 was induced 612-fold, whereas the level of PTGER4 was 45 times higher than that of the uninfected control at 34 h postinfection. No other genes with known functions of arachidonic acid metabolism or detection were upregulated.

We next performed RT-PCR to validate the DNA microarray result. Total RNA from C. trachomatis-infected HeLa 229 cells or the uninfected control was analyzed with RT-PCR for PTGS2 and PTGER4 expression (Fig. 1A). Consistent with the result of microarray studies, PTGS2 was not detectable in the control HeLa 229 cells, and the levels of PTGS2 expression were induced by Chlamydia infection. We found that PTGER4 was constitutively expressed in HeLa 229 cells and that Chlamydia infection increased PTGER4 expression.

FIG. 1.

Chlamydia infection upregulates PTGS2/COX2. (A) Chlamydia infection upregulates PTGS2 and EP4 gene expression. HeLa 229 cells of human cervical carcinoma were infected with C. trachomatis LGV2 at 1 IFU/cell. The cells were harvested at 12 and 24 h postinfection and processed for isolation of total RNA. The expression of PTGS2 and EP4 of PGE2 receptor was determined by RT-PCR amplification. Consistent with the DNA microarray results, Chlamydia infection induced PTGS2 gene expression and EP4 upregulation. GAPDH was included as a control. (B and C) Chlamydia infection induces COX2 protein expression. COX2 protein, or Erk as a loading control, expression was determined by immunoblotting analysis. L2, LGV2; serovar D, C. trachomatis serovar D; Cpn, C. pneumoniae. (D) Chlamydia infection induces COX2 expression in epithelial cells and PBMC. Human cell lines of vulva epidermoid carcinoma A431 (CRL-1555), laryngeal epidermoid carcinoma HEp-2 (CCL-23), and lung epithelial carcinoma A549 were infected with C. trachomatis LGV2 at 1 and 3 IFU/cell for 24 h or PBMC for 48 h. COX2 protein expression was detected by immunoblotting analysis.

We next performed immunoblotting analysis to demonstrate COX2 upregulation by C. trachomatis infection. HeLa 229 cells were infected with LGV2 at various MOIs for 24 h. COX2 expression and induction was detected by using a monoclonal anti-COX2 antibody. As shown in Fig. 1B, the level of COX2 in the uninfected HeLa 229 control cells was very low or not detected. Chlamydia infection significantly induced COX2 protein expression in HeLa 229 cells in a dose-dependent manner. At MOIs of 0.3, 1, 3, and 10 IFU/cell, the levels of COX2 expression were induced by 1.6-, 5-, 9-, and 20-fold, respectively. The induction of COX2 protein was not restricted to the LGV2 serovar or to HeLa 229 cervical epithelial cells. As shown in Fig. 1C, both serovar D of urogenital infection and C. pneumoniae induced COX2 expression in HeLa 229 epithelial cells. Moreover, the expression of COX2 was also upregulated in A431, A549, and HEp-2, as well as PBMC infected with LGV2 (Fig. 1D). Together, these data demonstrated that Chlamydia infection induced COX2 expression.

PGE2 production during Chlamydia infection promotes IL-8 release.

Chlamydia infection was reported to induce host lipid remodeling (14, 44), resulting in the release of arachidonic acid. We hypothesized that COX2 upregulation promoted prostanoid production, including the proinflammatory PGE2, leading to host inflammation. Therefore, we investigated whether PGE2 was produced by Chlamydia-infected cells. Culture supernatants from LGV2-infected HeLa 229 or uninfected control cells were measured for PGE2 production by ELISA. As shown in Fig. 2A, no PGE2 was detected in the uninfected HeLa 229 cells. C. trachomatis infection induced PGE2 release at 900 pg/ml at 24 h postinfection by 10⁵ cells. The production of PGE2 was dependent on Chlamydia growth and was modulated by COX2 activity since treatment with the antibiotic chloramphenicol or with the specific COX2 inhibitors NS-398 at 20 μM or SC-791 at 1 μM, but not salicylic acid, completely blocked Chlamydia-induced PGE2 production.

FIG. 2.

Chlamydia infection induces IL-6 and IL-8 production, and PGE2 production induces IL-8 production. (A) Chlamydia infection promotes PGE2 production. Monolayers of HeLa 229 cells were infected with C. trachomatis LGV2 in duplicates. PGE2 production was determined by ELISA. We also included heat-inactivated C. trachomatis at an equivalent of 10 IFU/cell. In parallel experiments, the infected cells were treated with chloramphenicol to inhibit bacterial growth, selective COX2 inhibitors NS-398 and SC-791, or the nonselective COX1 and COX2 inhibitor salicylic acid at a subinhibitory concentration (20 μM). (B) IL-6 and IL-8 (as indicated with arrows) were the major cytokine/chemokines secreted by HeLa 229 cells during Chlamydia infection. HeLa 229 cells in six-well plate were infected with LGV2 for 18 h. Cytokine production was collected to serum-free medium for another 12 h and was detected by incubation with membranes from the RayBio Inflammation Antibody Array III kit for simultaneous detection of 40 inflammatory factors. HeLa 229 cellsconstitutively shed soluble TNF-RII and TIMP2. Chlamydia infection induced IL-6 and IL-8 production, which was determined as 2.1 and 1.3 ng/ml, respectively, by quantitative ELISA. In a parallel experiment, the infected cells were treated with chloramphenicol (LGV2+Chl) at 60 μg/ml to demonstrate the dependency of cytokine production on Chlamydia infection. (C) PGE2 promotes IL-8 production. Monolayers of HeLa 229 cells in a 24-well plate were treated with PGE2, LTB4, a panel of synthetic lipids, or infected with LGV2 at 1 IFU/cell for 24 h. IL-8 accumulation in the culture medium was determined by ELISA. PE 16:0 and PE 18:1, 10 μM lysophosphoethanoamine at 16:0 and 18:1, respectively; PC 16:0, 10 μM lysophosphocholine at 16:0; PA 18:1, 10 μM lysophosphatic acid at 18:1; PGE, 5 μM PGE2; LTB, 10 μM leukotriene B4. The data are representative of three independent experiments and are presented as means ± the standard deviations.

To investigate whether PGE2 production promoted Chlamydia-induced inflammation, we first examined cytokine production in Chlamydia-infected HeLa 229 cells with protein array analysis. The culture supernatants from Chlamydia-infected and uninfected cells were assayed against an array of 40 antibodies of inflammation factors. We found that HeLa 229 cells mainly secreted IL-6 and IL-8 in response to Chlamydia infection (Fig. 2B). The concentrations of IL-6 and IL-8 were determined to be 2.1 and 1.3 ng/ml, respectively, by quantitative ELISA. This result was consistent with our DNA microarray results, which showed strong upregulation of both IL-6 and IL-8 genes at the middle and late stages of Chlamydia infection (data not shown).

IL-8, also known as neutrophil-activating peptide 1, induces neutrophil infiltration and has been proposed as the underlying factor for Chlamydia-induced tissue damage (2, 37). Thus, we investigated whether PGE2 treatment promoted IL-8 secretion. As shown in Fig. 2C, PGE2 treatment induced IL-8 production at 130 pg/ml, a level ca. 27% of that induced by C. trachomatis. In contrast, treatment with LTB4, one of the major lipooxygenase metabolites of arachidonic acid, or lysolipids, potential products from lipase activation, did not significantly induce IL-8 production, indicating that lipid metabolism, specifically COX2-mediated PGE2 production, contributed to proinflammatory IL-8 production.

Chlamydia infection activates cPLA2.

To further establish that lipid metabolism contributed to proinflammatory chemokine production, we next examined cPLA2 activation during Chlamydia infection with immunoblotting analysis. Consistent with a previous report (39), Chlamydia infection induced cPLA2 phosphorylation (Fig. 3A). This result was verified by using an in vitro lipase activity assay. cPLA2 catalyzes the release of sn-2 position fatty acid, which can be measured by using 2-O-alkyl substituted phospholipids such as the fluorescent BODIPY FL C₅-HPC as a substrate (9). HeLa 229 cells were infected with LGV2 at an MOI of 5 IFU/cell in the presence or absence of chloramphenicol. Total cPLA2 protein was immunoprecipitated with a polyclonal anti-cPLA2 antibody and used for the cPLA2 activation assay. The fluorescent fatty acid was separated on a thin-layer chromatography plate and quantified with a FluorImager for fluorescent fatty acid production as a measurement of cPLA2 activity. The immunocomplex from Chlamydia-infected cells, but not from the uninfected control or chloramphenicol-treated cells, contained lipase activity, indicating cPLA2 activation during Chlamydia infection (Fig. 3B). Together, these data demonstrated that Chlamydia infection activated the eicosanoid metabolism pathway for PGE2 production which promoted IL-8 secretion during Chlamydia infection.

FIG. 3.

Chlamydia infection activates cPLA2. (A) Chlamydia infection induces cPLA2 phosphorylation. HeLa 229 cells were infected with LGV2 at 1 IFU/cell for times as indicated. cPLA2 activation was determined by immunoblotting analysis with a phospho-cPLA2 antibody (S505). (B) Chlamydia infection activates cPLA2. HeLa 229 cells were infected with C. trachomatis LGV2 at 1 IFU/cell in the presence or absence of chloramphenicol at 60 μg/ml for 24 h. The cell lysates were immunoprecipitated with a polyclonal anti-cPLA2 antibody. cPLA2 activity in the immunocomplexes was assayed by incubation with a fluorescent derivative of glycerophospholipid (BODIPY FL C₅-HPC). The lipid product was extracted with chloroform-methanol. After separation on thin-layer chromatography plate, the production of sn-2 fatty acid was visualized with a FluorImager. The upper portion of the panel shows the chemical structure of BODIPY FL C₅-HPC, and the cPLA2-mediated cleavage site as indicated by the arrow.

PGE2 promotes IL-8 production by activating Erk/MAPK.

The production of IL-8 is regulated at both transcriptional and translational levels (1, 17, 32). PGE2 was reported to promote cell proliferation through the activation of p42/44 Erk1/2 (26, 29), members of the MAPKs that activate transcriptional factors for cell proliferation, and stress responses, as well as the maximal activation of IL-8 promoter. To investigate the regulatory mechanisms of IL-8 production during Chlamydia infection, we first examined MAPK signaling pathway activation since IL-8 production involves MAPK. HeLa 229 cells were infected with Chlamydia at 1 IFU/cell for various times. MAPK activation was determined by detection of Erk, JNK, or p38 MAPK protein phosphorylation with specific antibodies. Chlamydia infection selectively activated Erk/MAPK (Fig. 4A). No activation of JNK or p38 was detected from Chlamydia-infected cells. In addition, the activation was dependent on bacterial replication since heat inactivation or treatment with chloramphenicol or rifampin blocked Chlamydia-induced Erk phosphorylation (Fig. 4B), a finding consistent with the previous observation that IL-8 production requires bacterial growth (31).

FIG. 4.

PGE2 activates Erk/MAPK, which is required for Chlamydia-induced IL-8 production. (A) Chlamydia infection selectively activates Erk/MAPK. HeLa 229 cells were infected with C. trachomatis LGV2 in the presence or absence of inhibitors against MAPK signal transduction. MAPK activation was determined by immunoblotting analysis with a phosphorylation-specific antibody. The inhibitors (specificity) were U0126 (Erk/MAPK) at 1 μM, SB203580 (p38) at 10 μM, or SP600125 (JNK) at 2 μM for 24 h. Anisomycin (Ani.) at 10 μM was included as a positive control for MAPK activation. (B) Erk activation is dependent on bacterial growth. HeLa 229 cells were infected with LGV2 in the presence or absence of chloramphenicol (Chl) or rifampin (Rif) for 24 h or treated with heat-inactivated LGV2 (HK) at a dose equivalent to 20 IFU/cell. Erk activation and bacterial replication, as determined by the expression of the major outer membrane protein (MOMP), were evaluated by immunoblotting analysis. (C) Inhibitors of the Erk signal pathway activation block Chlamydia-induced Erk activation and IL-8 production. HeLa 229 cells were infected with C. trachomatis LGV2 in serum-free DMEM. The infected cells were treated with Erk/MAPK inhibitor U0126 at 1 μM or PD98059 at 10 μM. IL-8 production and Erk/MAPK activation were determined by ELISA and immunoblotting analysis, respectively. (D) PGE2 treatment activates Erk/MAPK. Monolayers of HeLa 229 cells were treated with PGE2 at 5 μM for various times, and Erk activation was determined by immunoblotting analysis. For treatment with U0126, the compound was used at 10 μM 30 min prior to addition of PGE2 for 15 min. In parallel experiments, IL-8 production was determined by ELISA 24 h after PGE2 stimulation.

To demonstrate the requirement of Erk activity for Chlamydia-induced IL-8 production, we treated infected cells with either U0126 or PD98059, two structurally distinct inhibitors against Erk activation, and investigated whether inhibition of Erk activity affected IL-8 production. Treatment with U0126 or PD98059 at 1 or 10 μM, respectively, blocked IL-8 induction, an event coinciding with the inhibition of Erk/MAPK activation (Fig. 4C), indicating the essential role of Erk/MAPK activation in Chlamydia-induced IL-8 production.

We next addressed the question of whether PGE2 induced IL-8 production through the Erk/MAPK pathway. As shown in Fig. 4D, PGE2 treatment transiently activated Erk. Erk phosphorylation was detected as early as 5 min after PGE2 treatment and disappeared rapidly. Treatment with inhibitors against Erk/MAPK activation prior to addition of PGE2 abrogated Erk activation. In a parallel experiment, we found that both U0126 and PD98059 also inhibited PGE2-induced IL-8 production. Thus, Chlamydia infection promoted PGE2 release, which stimulated proinflammatory IL-8 production through the activation of Erk/MAPK.

DISCUSSION

By gene microarray profiling, validated by biochemical studies, we found that Chlamydia infection upregulated COX2 expression. COX2 is a key enzyme in the biosynthesis of prostaglandins. Membrane-bound phospholipids are cleaved by phospholipases for the production of arachidonic acid. Cyclooxygenases then convert arachidonic acid into prostaglandin H2, which can be transformed enzymatically and nonenzymatically into the primary prostanoids, including PGE2, PGF2α, PGD2, and TXBα2 (7). One of the major metabolites of cyclooxygenases is PGE2, which can induce inflammation, fever, and pain reaction through the interaction with its receptors (19, 38, 41, 42). We demonstrated that Chlamydia infection activated a lipid metabolism pathway for PGE2 production and proinflammatory chemokine IL-8 release. We observed that Chlamydia infection and/or replication induced cPLA2 activity and promoted COX2 expression and PGE2 secretion. PGE2 production by Chlamydia-infected epithelial cells stimulated IL-8 release through Erk/MAPK activation.

Chlamydia infection is the most common cause of sexually transmitted diseases, resulting in pelvic inflammatory disease, tubal blockage, and female infertility. Ocular infection leads to the most preventable blindness worldwide. A hallmark of the infection is chronic inflammation associated with proinflammatory cytokine production and IL-8-mediated neutrophil infiltration. The mechanisms of host inflammatory response and cytokine production by Chlamydia-infected epithelial cells remain undefined. There are conflicting reports on chlamydial LPS induction of proinflammatory cytokines. Chlamydial LPS does not induce cytokine production in cervical epithelial cells, the primary target of infection. Furthermore, proinflammatory cytokine production is dependent on bacterial growth (31). Consistent with this report, we found that treatment of infected cells with antibiotics that inhibit Chlamydia infection blocks IL-8 production. Our identification of a lipid metabolism pathway to Chlamydia-induced IL-8 production links cytokine production to bacterial growth.

The production of proinflammatory factors by nonimmune cells leads to the pathological conditions of chlamydial diseases. Chemokines such as IL-8 induces neutrophil infiltration to the primary sites of infection, causing subepithelial accumulation of mononuclear leukocytes and tissue damage. IL-6 and IL-8 were among the major secreted cytokine/chemokines by Chlamydia-infected HeLa 229 cells (Fig. 2A). IL-8 production is regulated both transcriptionally and posttranscriptionally. The promoter region of IL-8 contains binding sites for NF-κB, AP-1 and C/EBP, followed by a TATA box (4, 17, 32). Previous studies show that NF-κB activity is essential for TNF-α-induced IL-8 production, whereas AP-1 and C/EBP work in concert for maximal promoter activity (17, 24, 35). In contrast to TNF-α induction of IL-8, we found that Chlamydia infection did not significantly activate NF-κB (data not shown). Instead, it selectively activated Erk/MAPK. We showed that inhibitors against Erk/MAPK pathway activation blocked Chlamydia-induced IL-8 production, coinciding with inhibited Erk phosphorylation. Considering the importance of IL-8 in chlamydial pathogenesis, it will be interesting to elucidate the regulatory mechanisms leading to Erk activation and Erk-mediated IL-8 production during Chlamydia infection.

The mechanism of COX2 induction by Chlamydia infection is unknown. The stimuli known to induce COX2 are those associated with inflammation, such as bacterial LPS and cytokines such as IL-1, IL-2, and TNF-α. Rasmussen et al. observed that chlamydial LPS did not induce IL-8 and TNF-α release in cervical epithelial cells (31). Although we did not detect IL-1α, IL-1β, or TNF-α accumulation from Chlamydia-infected HeLa 229 cells with both protein array analysis and ELISAs, our DNA microarray studies showed strong induction of IL-1β and IL-1R by Chlamydia infection. It is possible that IL-1 produced at low concentration stimulates COX2 expression during Chlamydia infection.

Microbial pathogens can be classified into two broad categories from an evolutionary and ecological standpoint: those that infect the host accidentally and those that do so for growth advantages (11). The outcome of an accidental pathogen infection is often lethal, whereas pathogens such as Chlamydia spp. have developed efficient mechanisms to secure a favorable habitat for their replication while averting harm to the host. The findings presented here reflect the intimate relationship between Chlamydia as an obligate intracellular pathogen and its host. Chlamydia replication requires host lipid remodeling, which involves cPLA2 activity. The activation of cPLA2 promotes the release of arachidonic acid, a precursor for conversion to the proinflammatory PGE2 catalyzed by COX2 and related enzymes. A metabolically dormant Chlamydia may persist without triggering severe adverse effects to the host. However, the host may respond rapidly and effectively with such events as gene upregulation for antimicrobial as well as inflammatory responses if the balance is disrupted due to the onset of some vital events such as lipid uptake.