Differences in Innate Immune Responses (In Vitro) to HeLa Cells ...

INFECTION AND IMMUNITY, June 2002, p. 3234–3248 0019-9567/02/$04.00⫹0 DOI: 10.1128/IAI.70.6.3234–3248.2002 Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Vol. 70, No. 6

Differences in Innate Immune Responses (In Vitro) to HeLa Cells Infected with Nondisseminating Serovar E and Disseminating Serovar L2 of Chlamydia trachomatis Sophie Dessus-Babus,1 Toni L. Darville,2 Francis P. Cuozzo,1 Kaethe Ferguson,1 and Priscilla B. Wyrick1* Department of Microbiology, James H. Quillen College of Medicine, East Tennessee State University, Johnson City, Tennessee,1 and Department of Microbiology and Immunology, University of Arkansas for Medical Sciences, Little Rock, Arkansas2 Received 10 December 2001/Returned for modification 13 February 2002/Accepted 7 March 2002

The inflammatory response associated with Chlamydia trachomatis genital infections is thought to be initiated by the release of proinflammatory cytokines from infected epithelial cells. This study focuses on the interactions between C. trachomatis-infected HeLa cells and immune cells involved in the early stages of infection, i.e., neutrophils and macrophages. First, we showed that the expression of interleukin-11 (IL-11), an anti-inflammatory cytokine mainly active on macrophages, was upregulated at the mRNA level in the genital tracts of infected mice. Second, incubation of differentiated THP-1 (dTHP-1) cells or monocyte-derived macrophages (MdM) with basal culture supernatants from C. trachomatis serovar E- or serovar L2-infected HeLa cells resulted in macrophage activation with a differential release of tumor necrosis factor alpha (TNF-␣) and upregulation of indoleamine 2,3-deoxygenase (IDO) but not of Toll-like receptor 2 and 4 mRNA expression. Third, coculture of infected HeLa cells with dTHP-1 cells resulted in a reduction in chlamydial growth, which was more dramatic for serovar E than for L2 and which was partially reversed by the addition of anti-TNF-␣ antibodies for serovar E or exogenous tryptophan for both serovars but was not reversed by the addition of superoxide dismutase or anti-IL-8 or anti-IL-1␤ antibodies. A gamma interferon-independent IDO mRNA upregulation was also detected in dTHP-1 cells from infected cocultures. Lastly, with a two-stage coculture system, we found that (i) supernatants from neutrophils added to the apical side of infected HeLa cell cultures were chlamydicidal and induced MdM to express antichlamydial activity and (ii) although polymorphonuclear leukocytes released more proinflammatory cytokines in response to serovar E- than in response to L2-infected cells, MdM were strongly activated by serovar L2 infection, indicating that the early inflammatory response generated with a nondisseminating or a disseminating strain is different. Strains of the nondisseminating serovars D to K of Chlamydia trachomatis cause primarily asymptomatic lower genital tract infections but also symptomatic cervicitis and urethritis, as well as pelvic inflammatory disease that may lead to irreversible fallopian tube damage in women. Although genital infections caused by the nondisseminating serovars of C. trachomatis are limited to the mucosal domains, infections that involve the biovar lymphogranuloma venereum (serovars L1 to L3) are associated with the systemic spread of chlamydiae to the regional lymph nodes, leading to inguinal lymphadenopathy and complications such as genital ulcers, fistulas, and elephantiasis, and leading eventually to the destruction of the mucosal epithelium and scarring (32). The tissue damage associated with C. trachomatis infections is believed to be the result of an inflammatory process initiated at the infectious site and is exacerbated by reinfection (38). Histopathological studies showed that these infections are associated with a massive polymorphonuclear leukocyte (PMN) infiltration in the mucosal domains, followed by a migration of mononuclear leukocytes during the chronic phase (18, 24). In

vitro studies showed that epithelial cells—the prime targets of chlamydial infection—released several proinflammatory cytokines, including interleukin-8 (IL-8), IL-6, IL-1␣, IL-18, granulocyte-macrophage colony-stimulating factor, GRO-␣, GCP-2, and ENA-78 after infection (9, 20, 30, 41), suggesting that the acute host response to Chlamydia might primarily be initiated and sustained by the infected epithelial cells. Recently, the release of the anti-inflammatory cytokine IL-11 was also detected in polarized cultures of C. trachomatis-infected HeLa cells, interestingly in larger amounts in cultures infected with the disseminating serovar L2 than in cultures infected with the nondisseminating serovar E (9). IL-11 is a pleiotropic cytokine reported to inhibit the release of proinflammatory mediators such as tumor necrosis factor alpha (TNF-␣), IL-1␤, IL-6, IL-12, and nitrite oxide (NO) from lipopolysaccharide (LPS)activated macrophages (36, 37), as well as the production of IL-12 from dendritic cells (2), and recent studies revealed that recombinant IL-11 (rIL-11) can also interact with T cells and enhance the Th2 response (2, 33). Although a Th1 response appears to be crucial for the resolution of chlamydial infections (28, 39), the role of TNF-␣ is still unclear. In a recent study, Darville et al. (6) showed that blockade of the early TNF-␣ response in chlamydia-infected mice or guinea pigs had no effect on the number of organisms isolated from their genital tracts and did not alter the release of proinflammatory cytokines in infected mice. However, sev-

* Corresponding author. Mailing address: Department of Microbiology, Box 50759, VA#1–Rm. 1-41, James H. Quillen College of Medicine, East Tennessee State University, Johnson City, TN 37684. Phone: (423) 439-8079. Fax: (423) 439-8044. E-mail: pbwyrick@access .etsu.edu. 3234

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

VOL. 70, 2002

eral other studies suggested a protective role for TNF-␣ in the early response to infection (7, 8, 25, 40). For instance, Darville et al. (7) showed that C57BL/6 mice, which produced significantly higher levels of TNF-␣, were able to resolve their infection more rapidly than C3H/HeN or BALB/c mice. Considering that resident macrophages of the submucosal domains are likely to be a major source of TNF-␣, the higher IL-11 production from epithelial cells after C. trachomatis serovar L2 infection compared to that seen after serovar E infection may inhibit the release of proinflammatory cytokines, including TNF-␣ from activated macrophages which, if we assume that TNF-␣ plays a role in vivo in host defense against chlamydiae, could allow a better dissemination of L2 (9). The aims of the present study were (i) to determine whether or not IL-11 upregulation occurs during natural genital infection with C. trachomatis mouse pneumonitis (MoPn)-infected C3H mice as a model; (ii) to dissect the interactions between C. trachomatis-infected genital epithelial cells and macrophages whereby serovar E- or serovar L2-infected polarized HeLa cells were cocultivated in the presence of THP-1 cells or monocyte-derived macrophages (MdM) and the antichlamydial activity and the macrophage-derived cytokine response were compared; and lastly, (iii) to establish a two-stage coculture system by using, sequentially, PMN followed by MdM cocultivated with infected HeLa cells in order to mimic the chronology of events occurring during natural infection for a more valid analysis in vitro of the early immune responses. In the execution of these studies, it became clear that the choice of cell type used for the macrophage population makes a difference in the results obtained in vitro and also complicates the interpretation of results that are otherwise intended to aid in vivo findings. MATERIALS AND METHODS Cell lines. Endocervical epithelial cells HeLa 229 were grown at 37°C under 5% CO2 in minimal essential medium (MEM) with Earle’s salts and glutamine (Gibco-BRL) supplemented with 10% fetal bovine serum (FBS) and 10 ␮g of gentamicin/ml. McCoy cells, for growing chlamydia stocks, were cultivated at 37°C in MEM containing Hanks’ salts and glutamine and supplemented with 10% FBS. The human myelomonocytic cell line THP-1 was obtained from the American Type Culture Collection and cultivated in RPMI 1640 containing glutamine (Gibco-BRL) supplemented with FBS and gentamicin as detailed above. Before they were used in experiments, THP-1 cells were incubated at a concentration of 7 ⫻ 105 cells/well in 24-well tissue culture plates in the presence of 120 ng of phorbol 12-myristate 13-acetate (PMA; Sigma)/ml for 24 h to induce differentiation into adherent macrophages, designed as dTHP-1 cells, that were then washed twice with 1 ml of fresh RPMI. MdM and PMN isolation. Peripheral blood mononuclear cells and PMN were isolated from healthy donors by a standard Ficoll-Hypaque density gradient centrifugation. Peripheral blood mononuclear cells were washed three times with Hanks’ balanced salt solution devoid of Ca2⫹ and Mg2⫹, resuspended in RPMI, and seeded in wells containing a glass coverslip at a concentration of 3 ⫻ 106 to 4 ⫻ 106 cells/well. After 3 h of incubation at 37°C in 5% CO2, the nonadherent cells were removed by two washes with prewarmed RPMI, and the plates were reincubated. Then, half of the volume of medium was replaced with fresh RPMI every 3 days, and the MdM were used in experiments between 10 and 13 days postisolation. After Ficoll-Hypaque centrifugation, dextran separation, and red cell lysis, PMN were resuspended in fresh RPMI at a concentration of 107/ml and used in some of the coculture experiments. Chlamydia strains. C. trachomatis serovar E/UW-5/CX and C. trachomatis biovar lymphogranuloma venereum L2/434/Bu were grown in McCoy cells cultivated on microcarrier beads as described previously (42). The chlamydia harvests were aliquoted and stored at ⫺80°C.

3235

Antibodies. Anti-human IL-8 (anti-hIL-8; 5 ␮g/ml), anti-hTNF-␣ (0.5 ␮g/ml; Genzyme Diagnostics), and anti-hIL-1␤ (0.2 ␮g/ml; R&D Systems, Minneapolis, Minn.) mouse monoclonal antibodies, as well as goat anti-hIL-11 polyclonal antibodies (50 ␮g/ml; R&D Systems), were used in this study. Stimulation of dTHP-1 cells and MdM with basal supernatants from polarized HeLa cells infected with C. trachomatis. PMA-differentiated THP-1 or mature MdM, seeded in replicate wells of 24-well plates, were incubated for 24 h in the presence of 500 ␮l of basal supernatants from uninfected or C. trachomatis E- or L2-infected polarized HeLa cells. These basal supernatants were prepared as previously described (9). Briefly, polarized and confluent HeLa cells, seeded in 44-cm2 inserts coated with extracellular matrix (ECM; Matrigel; Becton Dickinson, Franklin Lakes, N.J.), were inoculated with a dilution of C. trachomatis serovar E or L2, giving a high level of infectivity. At 48 h postinfection (hpi), the basal supernatants of infected or uninfected cultures were collected and stored at ⫺80°C. These supernatants were determined to be positive by enzyme-linked immunosorbent assay (ELISA) for IL-8, IL-11, and IL-6 but negative for TNF-␣, and the amounts of IL-8 and IL-11 were higher in L2-infected cell supernatants (9). After 24 h of incubation, the supernatant from stimulated dTHP-1 cells or MdM was collected and stored at ⫺80°C for ELISAs. In some experiments, cells from replicate wells were lysed at 12 and 24 h poststimulation in the presence of Trizol reagent (Gibco-BRL) and pooled. Total RNA, isolated according to the manufacturer’s instructions, was then used for Toll receptor (TLR) and indoleamine deoxygenase (IDO) reverse transcription-PCR (RT-PCR) analysis. In all of these experiments, negative and positive controls, consisting of cells incubated in the presence of RPMI alone and Escherichia coli O111:B4 LPS (200 ng/ml; Sigma), respectively, were also included. In some experiments, as a positive control, dTHP-1 cells and MdM were incubated for 24 h in the presence of a mixture of rhIL-6 and rhIL-8 with or without rhIL-11 (R&D Systems), and the supernatants were then collected and stored at ⫺80°C for subsequent ELISA. The concentrations of cytokines used in this mixture were similar to those found in L2-infected HeLa cell basal supernatants used as described above, i.e., 45 pg of IL-6, 150 pg of IL-8, and 1 ng of IL-11/ml. Coculture experiments. HeLa cells were seeded, as previously described (9), in BioCoat commercial inserts (Becton Dickinson) coated with ECM and placed in the wells of 24-well cluster plates. Each experiment included duplicate wells and two sets of plates (one for the direct inclusion count and one for the subpassage). When confluent and polarized, the HeLa cells were inoculated with C. trachomatis E or L2 and, after 1 h of adsorption, the inoculum was removed and replaced with 500 ␮l of RPMI in the insert chamber and in the well. Subsequently, the infected cultures were incubated for 48 h in medium alone, in coculture with dTHP-1 cells or MdM seeded in the wells below the inserts, or in coculture with MdM prestimulated with supernatants from PMN incubated with C. trachomatis-infected HeLa cells. In some experiments, the cocultures were performed by using RPMI supplemented with 200 ␮g of tryptophan (Trp; Sigma)/ml, 300 U of superoxide dismutase (SOD; Sigma)/ml, or anti-hTNF-␣, anti-hIL-1␤, and anti-hIL-8 antibodies; the latter experiments included a 2-h preincubation of the dTHP-1 cells with SOD or antibodies. After 48 h of incubation, the basal supernatants were collected, centrifuged for 5 min at 500 ⫻ g to remove possible contaminant cells, and stored at ⫺80°C for subsequent ELISA. One set of infected cocultures was fixed with cold methanol and stained for chlamydial inclusions; HeLa cells from the second set of coculture plates were collected for progeny titration as described below. In order to determine whether the expression of IDO mRNA could be detected in HeLa and dTHP-1 cells grown in coculture, total RNA was isolated from both cell types at 24 or 48 hpi by the same protocol described below and analyzed by RT-PCR. In some preliminary coculture experiments, dTHP-1 cells and MdM were collected from culture plates by trypsinization, centrifuged, resuspended in 50 ␮l of ECM (diluted 1:3 in RPMI), and then reseeded directly onto insert filters (on the side opposite the infected cells) rather than down in the wells. After ca. 30 min of incubation at 37°C to allow for ECM polymerization, the inserts were reversed to their normal position and incubated in fresh RPMI medium. The protocol used for the cocultures of HeLa cells and MdM prestimulated with supernatants from PMN incubated with C. trachomatis-infected HeLa cells is outlined in Fig. 1. Titration of chlamydiae. (i) Direct counts. After methanol fixation, infected HeLa cells from control cultures or cocultures were stained for 30 min at 37°C with a pool of fluorescein isothiocyanate-conjugated monoclonal antibodies generated against the Chlamydia major outer membrane protein (Microtrak; Wampole Laboratories, Newark, N.J.) and then evaluated by fluorescence microscopy. For each experiment, the average number of inclusions per field (⫻400 magni-

3236

DESSUS-BABUS ET AL.

INFECT. IMMUN.

FIG. 1. Protocol used for the coculture of HeLa cells with MdM prestimulated with supernatants from PMN incubated with C. trachomatisinfected HeLa cells. Briefly, 106 PMN were added to the apical side of polarized HeLa cells infected with C. trachomatis for 36 h, and the cultures were reincubated for 12 h on a shaker (CULTURE 1). The apical medium was then collected and centrifuged to pellet the PMN. A portion of the supernatant was stored at ⫺80°C (SUPERNATANT I), and the rest was diluted 1:1 in fresh RPMI before being added to MdM. After 48 h of stimulation, the macrophage culture supernatant was collected and stored at ⫺80°C (SUPERNATANT II) and then replaced by 500 ␮l of fresh RPMI, and these stimulated MdM were next cocultivated with freshly inoculated cultures of polarized HeLa cells (CULTURE 2). At 48 hpi, basal supernatants from the cocultures were collected and stored at ⫺80°C (SUPERNATANT III), and inclusion counts and progeny titration were performed on HeLa cells.

fication) was determined. Since some differences in the absolute numbers of inclusions were found in independent experiments, inclusion counts were expressed, for each set of experiments, as percentages compared to the control cultures of infected HeLa cells alone. (ii) Subpassage. For subpassage, infected HeLa cells from control cultures or cocultures were collected at 48 hpi by using 100 ␮l of a dispase solution (Becton Dickinson), washed with fresh MEM, and centrifuged for 5 min at 500 ⫻ g. The infected cells were resuspended in 200 ␮l of MEM, disrupted by a freezingthawing cycle, and sonicated for 10 min. The suspension was centrifuged at 500 ⫻ g to remove large cell debris, and serial dilutions of the supernatant containing elementary bodies (EB) were adsorbed onto fresh, confluent HeLa cell monolayers grown on glass coverslips. After 48 h of incubation in MEM supplemented with cycloheximide (1 ␮g/ml), the infected cells were fixed and stained for detection of fluorescent chlamydial inclusions. The average number of inclusions on the duplicate coverslips was determined, and the progeny titer was expressed as the total number of inclusion-forming units (IFU) in the cell harvest. The final results were again expressed as percentages compared to the control infected HeLa cultures alone. Transmission electron microscopy. Sample processing for embedding in Epon-Araldite resin was performed as described previously by Wyrick et al. (44). Stained thin sections were examined on a Phillips 201 electron microscope operated at 60 kV. ELISA for cytokines. Culture supernatants were assayed for human IL-1␤, IL-6, IL-8, IL-12, TNF-␣, and gamma interferon (IFN-␥) by using Quantikine ELISA kits purchased from R&D Systems. Measurement of superoxide and NO production by dTHP-1 cells. Superoxide production was measured as the SOD-sensitive reduction of ferricytochrome c by

a previously described protocol (26). Spontaneous O2⫺ release was measured in supernatants from dTHP-1 cells seeded in 96-well microplates (105 cells/well) and incubated with only phenol red-free RPMI rather than from coculture supernatants for technical reasons. The cumulative O2⫺ production was expressed as nanomoles per 7 ⫻ 105 cells, i.e., the number of dTHP-1 cells generally used in coculture experiments. Nitrite concentration in supernatants from dTHP-1 cells seeded in 24-well plates (7 ⫻ 105 cells/well in phenol red-free RPMI for 48 h) was measured by the Griess method. In vivo study. Female C3H/HeN (H-2k) mice (6 weeks old) were purchased from Jackson Laboratories, Bar Harbor, Maine. Mice were given food and water ad libitum in an environmentally controlled room with a cycle of 12 h of light and 12 h of darkness. Female mice, 7 to 8 weeks of age, were used in this study. The agent of MoPn, a C. trachomatis biovar, was used for infection. This agent was originally obtained from the American Type Culture Collection and is maintained in McCoy cells (29). Mice were infected by placing 30 ␮l of 250 mM sucrose–10 mM sodium phosphate–5 mM L-glutamic acid (SPG) containing 107 IFU of MoPn into the vaginal vault. Infection was performed with the mice under sodium pentobarbital anesthesia. Control mice were inoculated with SPG. The mice received 2.5 mg of progesterone (Depo-Provera; Upjohn, Kalamazoo, Mich.) subcutaneously 7 days before vaginal inoculation. Infection was monitored by swabbing the vaginal vault and ectocervix with a Calgiswab (Spectrum Medical Industries, Los Angeles, Calif.) prior to sacrifice and by enumerating the IFU via isolation on McCoy cell monolayers (17). Four to five mice per group were sacrificed by cervical dislocation 48 and 72 h after vaginal inoculation of SPG or MoPn, and the genital tracts were removed and placed in RPMI 1640 containing 0.6 mg of collagenase Type I (Sigma)/ml. Genital tracts were exten-

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

VOL. 70, 2002

sively minced with a scalpel, and the suspension was incubated for 1 h at 37°C in 5% CO2. The cell suspensions were filtered through a 70-␮m (pore-size) nylon cell strainer (Becton Dickinson), and then washed twice and resuspended in RNA Later (Ambion, Inc., Austin, Tex.). Total RNA isolation and semiquantitative RT-PCR analysis. Mouse genital tissues stored in RNA Later were solubilized with 10 volumes of Trizol reagent, and the total RNA was isolated as recommended by the manufacturer and resuspended in RNase-free water. RNA isolation from cultured cells (HeLa, dTHP-1, or MdM) was performed by using the same protocol after removal of the culture medium. The contaminant genomic DNA was then eliminated from RNA preparations by using RNase-free DNase I (Clontech Laboratories, Palo Alto, Calif.). The total RNA concentration and purity were evaluated by absorbance readings at 260 and 280 nm, and RNA integrity was confirmed on a denaturing agarose gel. The RT reactions were performed with the RetroScript kit (Ambion) by using ca. 1 ␮g of total RNA in the presence of random decamers according to the conditions specified by the manufacturer. cDNAs were amplified by PCR by using the Expand High Fidelity PCR System (Roche Molecular Biochemicals, Indianapolis, Ind.) in a 50-␮l reaction mixture containing 1⫻ Expand HF buffer, 1.5 mM MgCl2, 200 ␮M (each) deoxynucleoside triphosphates, 1 ␮M forward and reverse primers, and 1.5 U of Taq polymerase. Subsequently, 2 ␮l of cDNA generated from mouse tissues or cultured cells was added to the PCR mixture. The PCR primers for human GAPDH (glyceraldehyde-3-phosphate dehydrogenase; 306 bp), for human IDO (324 bp), and for murine IL-11 (413 bp) have been described previously (4, 5, 12). The sequences of the primers used for murine GAPDH (536 bp) were 5⬘-CCACCAT GGAGAAGGCCGGG and 5⬘-AACCTGGTCCTCAGTGTAGCCCA. Except for GAPDH, the sets of PCR primers used in this study did not allow for amplification from genomic DNA, or the PCR product derived from DNA was different in size compared to that obtained from cDNA. For all PCRs, the conditions were 5 min at 95°C, followed by the appropriate number of cycles consisting of 2 min at 95°C, 2 min at 60°C, and 2 min at 72°C, except for the IDO PCR, which required an annealing temperature of 57°C. In preliminary experiments, a time course analysis of amplification was performed for each set of primers to determine the number of cycles necessary for semiquantitative analysis of PCR products, i.e., before the amplification curve reached the plateau. After amplification, 10 ␮l of each PCR product was electrophoresed on a 2% agarose gel and visualized by ethidium bromide staining. For the mouse study, PCR products were transferred to a nylon membrane and analyzed by Southern blotting using internal probes with the following sequences: 5⬘-CTTTGGCATT GTGGAAGGGCTCAT for GAPDH and 5⬘-AGACAAATTCCCAGCTGACG GAGA for IL-11. Then, 5 pmol of probe was labeled for 60 min at 37°C in the presence of 1⫻ NEB buffer (50 mM Tris [pH 7.5], 10 mM MgCl2, 0.1 mM EDTA, 0.1 mM spermidine), 5 mM dithiothreitol, 10 U of T4 polynucleotide kinase, and 5 ␮l of [␥-32P]dATP (Amersham Pharmacia, Piscataway, N.J.) in a 50-␮l reaction mixture. The probe was then purified on a MicroSpin G25 column (Amersham Pharmacia), and the incorporation yield was evaluated by using a scintillation counter (Beckman LS 6500). After a prehybridization step at 42°C for 60 min in hybridization buffer (0.5 M phosphate buffer, pH 7.5; 7% sodium dodecyl sulfate [SDS]), the membranes were incubated overnight at 42°C with 10 ml of fresh hybridization buffer containing 1 ⫻ 106 to 2 ⫻ 106 cpm of labeled probe. The membranes were subsequently washed for 30 min at room temperature in 2⫻ SSC (1⫻ SSC is 0.15 M NaCl plus 0.015 M sodium citrate)–0.1% SDS and for 30 min at 30°C in 0.1⫻ SSC–0.1% SDS and then exposed to a PhosphorImager screen for different times. The screen was scanned and analyzed with the Quantity One software (Bio-Rad), and the data were normalized by using the GAPDH housekeeping gene as a reference for constant level of expression. For TLR RT-PCR, scanning and densitometric analysis were directly performed from the agarose gels, and the PCR products were quantified by comparison to the GAPDH control.

RESULTS IL-11 mRNA expression is upregulated in mice infected with C. trachomatis MoPn. The first objective was to determine whether expression of IL-11 mRNA was upregulated in response to C. trachomatis infection in vivo. Female C3H mice were inoculated intravaginally with the mouse biovar C. trachomatis MoPn. Semiquantitative RT-PCR analysis of genital tracts, after normalization with GAPDH, showed 3.1- and 2.2fold increases in the levels of IL-11 mRNA expression in in-

3237

FIG. 2. Semiquantitative analysis of IL-11 mRNA expression in C3H mice infected with C. trachomatis MoPn. The PCR products from murine IL-11 (413 bp) primers after 28 cycles of amplification (lanes 1 to 5) and from GAPDH (536 bp) primers after 18 cycles (lanes 6 to 10) were analyzed on a 2% agarose gel (A) and after Southern blotting with an internal probe (B). Then, 10-␮l portions of sample were loaded onto the gels as follows: uninfected control mice (lanes 1 and 6), MoPn-infected mice at 48 hpi (lanes 2 and 7) and at 72 hpi (lanes 3 and 8), a negative control of amplification with no cDNA added to the reaction mix (lanes 4 and 9), and cDNA from uninfected McCoy cells used as positive control for the RT and PCRs (lanes 5 and 10).

fected mice at 48 hpi (Fig. 2, lane 2) and 72 hpi (Fig. 2, lane 3), respectively, compared to control uninfected mouse tissues. These data confirmed that the production of IL-11 by chlamydia-infected genital epithelial cells in vitro was not simply due to laboratory manipulation. Basal supernatants from C. trachomatis-infected HeLa cells induced TNF-␣ production in dTHP-1 cells and MdM. Since it has been proposed that one function of IL-11 may be to depress TNF-␣ release from macrophages and thereby prolong the establishment of infection, it seemed logical to compare macrophage responses to signals released by genital epithelial cells infected with the nondisseminating serovar E and the disseminating serovar L2 of C. trachomatis. In the literature, PMA-differentiated THP-1 cells are commonly used as a model for macrophages. However, we chose to employ and compare the responses of dTHP-1 cells and MdM. Both cell types were incubated separately for 24 h in the presence of basal supernatants obtained from HeLa cells infected with each serovar and the concentration of TNF-␣ released, used as a marker for macrophage activation, was determined by ELISA. The results are presented in Table 1. Baseline production of TNF-␣, i.e., in the presence of RPMI alone, which was higher for dTHP-1 cells (35.3 ⫾ 5 pg/ml) than for MdM (undetectable), was likely due to prior induction by PMA exposure used to stimulate the differentiation of THP-1 cells. The responsiveness to stimuli of the dTHP-1 cells, as well as of the MdM, was also controlled by using Escherichia coli LPS and resulted in the release of large amounts of TNF-␣ (11,167 ⫾ 3,617 pg/ml for dTHP-1 cells and 1,012 ⫾ 187 pg/ml for MdM). In both cell types, the TNF-␣ response to basal supernatants from infected HeLa cells was higher than for the uninfected control supernatants (Table 1). Interestingly, the amounts of

3238

DESSUS-BABUS ET AL.

INFECT. IMMUN.

TABLE 1. TNF-␣ secretion by dTHP-1 cells and MdM in response to different stimuli and measured by ELISA Inducera

U sup E sup L2 sup RPMI E. coli LPS (200 ng/ml) rhIL-6⫹rhIL-8 rhIL-6⫹rhIL-8⫹rhIL-11

Concn of TNF-␣ (pg/ml)b ⫾ SD dTHP-1

MdMc

27.3 ⫾ 6.7 89.7 ⫾ 13.6* 79.8 ⫾ 11.2* 35.3 ⫾ 5 11,167 ⫾ 3,617* 33.9 37.0

0.3 ⫾ 0.6 160.5 ⫾ 52.6* 52.3 ⫾ 20.6* – 1,012 ⫾ 187* 3.1 10.2

a U sup, E sup, and L2 sup represent basal supernatants from HeLa cells either uninfected or infected with C. trachomatis E and L2, respectively, and collected at 48 hpi. No TNF-␣ was detected in these basal supernatants. The rhIL-6, rhIL-8, and rhIL-11 concentrations in the mixture used to stimulate dTHP-1 and MdM were 45 pg/ml, 150 pg/ml, and 1 ng/ml, respectively. b The TNF-␣ concentration released from dTHP-1 cells or MdM was determined after 24 h of incubation in the presence of the inducing agent(s). The data shown in this table represent the average values from two or three independent experiments, and the standard deviation was calculated only for the experiments run three times. TNF-␣ concentrations significantly different (P ⬍ 0.05) from the negative control RPMI are indicated by an asterisk. –, Not detected. c Because the number of MdM per well after 10 to 13 days of culture was quite variable, each experiment included replicate wells to ensure that the differences in the TNF-␣ amounts detected after stimulation were a true reflection of a specific response rather than related to a difference only in the number of cells.

TNF-␣ induced by serovar E-infected cell supernatants were higher than those induced by serovar L2-infected cell supernatants, and this difference was more pronounced with MdM (160.5 ⫾ 52.6 pg/ml for serovar E versus 52.3 ⫾ 20.6 pg/ml for serovar L2) than with dTHP-1 cells (89.7 ⫾ 13.6 pg/ml for serovar E versus 79.8 ⫾ 11.2 pg/ml for serovar L2). To determine whether IL-11, present in higher amounts in serovar L2-infected than in serovar E-infected cell supernatants, might, at least partially, account for a differential suppressive effect on TNF-␣ response, HeLa cell basal supernatants were preincubated for 2 h with a neutralizing anti-hIL-11 antibody prior to being exposed to macrophages. An increase in TNF-␣ production was detected for all samples containing anti-IL-11 antibodies, including the RPMI control samples (data not shown). In the event that nonspecific activation of macrophages might result from the binding of antibody Fc portions to Fc receptors present on these cells, dTHP-1 cells and MdM were preincubated with human Fc fragment; there was no improvement in the specificity of the TNF-␣ response under these experimental conditions (data not shown). Surprisingly, incubation of MdM or dTHP-1 cells with a mixture of rhIL-6 and rhIL-8, at concentrations similar to that found in L2-infected cell supernatants (9), did not induce an increase in the TNF-␣ production compared to the control RPMI alone (Table 1), suggesting that these cytokines were not major TNF-␣ inducers in C. trachomatis-infected HeLa cell supernatants. Equally interesting, the supernatants from MdM or dTHP-1 cells, upon incubation for 24 h with infected HeLa cell supernatants or rhIL-6–rhIL-8 mixtures, did not contain the macrophage-derived cytokines IL-1␤ and IL-12 (data not shown). Coculture of infected HeLa cells with dTHP-1 cells induced a differential inhibition in C. trachomatis serovar E and L2 growth. In order to determine whether the presence of macrophages would affect chlamydial growth in HeLa cells, PMA-

differentiated THP-1 cells, seeded in the plastic wells below the inserts, were cocultivated with freshly inoculated polarized HeLa cells, and chlamydial growth was subsequently evaluated at 48 hpi by a direct count of inclusions or after a subpassage on fresh HeLa cell monolayers to determine the infectivity of the progeny. The results are summarized in Fig. 3. Interestingly, the presence of dTHP-1 cells consistently induced a strong reduction in the number of serovar E inclusions (53% ⫾ 2% compared to the control; Fig. 3A and Fig. 4A and B) but not in the number of serovar L2 inclusions (Fig. 3B). However, a decrease in the infectious progeny was detected for both strains and probably reflected the fact that the inclusions look much smaller in the cocultures with serovar E and L2 (serovar E, Fig. 4B; L2, data not shown) than in the control cultures (Fig. 4A). Again, this reduction in infectivity was more dramatic for serovar E (15.7% ⫾ 2.1% compared to the control for E versus 40.3% ⫾ 6.4% for L2, Fig. 3). The same pattern of chlamydia inhibition was found when the dTHP-1 cells were closely juxtaposed to infected HeLa cells, i.e., when they were suspended in the ECM on the underside of the filter (data not shown). Nevertheless, even though the filters in the inserts contained 8-␮m pore channels, we cannot exclude the possibility that some dTHP-1 cells migrated toward infected epithelial cells and therefore that chlamydial inhibition may be the result of cell-to-cell interactions, possibly accompanied by the release from activated macrophages of soluble factor(s) with an antichlamydial activity. Transmission electron microscopy analysis of serovar E inclusions revealed a delay in C. trachomatis growth in cocultures (Fig. 4D to F) compared to the controls (Fig. 4C). At 48 hpi, inclusions in control HeLa cells contained mostly mature EB and intermediate bodies and some reticulate bodies (RB) (Fig. 4C). In contrast, inclusions from HeLa cells cocultivated with dTHP-1 cells contained mostly normal and abnormal RB and only rarely EB. In addition, large and numerous outer membrane blebs were consistently observed in the cocultures and were often colocalized with a dark globular component, sometimes lining the inner surface of the inclusion membrane (Fig. 4D to F). This component was not characterized but might be chlamydia-derived glycogen released under stress conditions (41), such as coculture with activated macrophages. In addition, instead of being well individualized and separated within the inclusion, chlamydiae from the cocultures were commonly aggregated together (compare Fig. 4F with C). Similar ultrastructural observations were made for serovar L2 in coculture (data not shown). Taken together, these data showed that C. trachomatis growth and maturation in epithelial cells was affected by the presence of macrophage-like dTHP-1 cells and that the inhibition was more dramatic for serovar E than for serovar L2. Analysis of cytokine, NO, and superoxide release by dTHP-1 cells and coculture experiments in the presence of neutralizing antibodies or SOD. In an attempt to identify the factor(s) released by dTHP-1 cells that may be responsible for the strong and differential growth inhibition of C. trachomatis serovar E versus serovar L2 in HeLa cells, basal supernatants from coculture chambers were tested for several macrophage-derived cytokines by ELISA. Detectable amounts of TNF-␣, IL-1␤, IL-6, and IL-8 were found in the coculture supernatants with concentration ranges of 40 pg of TNF-␣, 110 pg of IL-1␤, 85 pg

VOL. 70, 2002

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

3239

FIG. 3. Titration of C. trachomatis serovar E and L2 after 48 h of coculture of polarized infected HeLa cells with dTHP-1 cells. HeLa cells, seeded in inserts and infected with C. trachomatis, were cultivated in the presence of dTHP-1 cells. At 48 hpi, chlamydial inclusions were counted after fluorescent immunostaining (direct count, shaded bars), and the titer of the infectious progeny was determined after passage on fresh HeLa cells (subpassage, open bars). Since some differences in the absolute numbers of inclusions were found between independent experiments, inclusion counts for serovar E (A) and serovar L2 (B) were expressed as percentages of chlamydial growth compared to the control cultures, i.e., infected HeLa cells cultivated alone, that represent 100% of growth. In some experiments, the cocultures were performed in the presence of anti-TNF-␣, anti-IL-1␤, or anti-IL-8 antibodies; Trp (200 ␮g/ml); or SOD (300 U/ml). The lack of or minimal effect of the antibodies, Trp, or SOD on chlamydia growth was tested on control cultures of HeLa cells alone. This figure shows average (⫾ the standard deviation) results from three independent experiments, and statistically significant recovery (P ⱕ 0.05) in chlamydial growth or infectivity compared to the control cocultures is indicated by an asterisk.

of IL-6, and ⬃1,000 pg of IL-8/ml but, surprisingly, no significant difference in cytokine concentration was found between the uninfected control and C. trachomatis-infected cocultures (Table 2). Importantly, no IL-12 nor IFN-␥ was detected in supernatants from infected HeLa cell cultures or cocultures (data not shown). In addition, the baseline production of NO and superoxide by dTHP-1 cells after PMA-induced differentiation was also monitored. NO amounts were either negligible or not detectable in supernatants from cocultures of infected HeLa and dTHP-1 cells and from dTHP-1 cells incubated in RPMI alone, respectively (Table 2). However, increasing amounts of O2⫺ were detected over time in supernatants from dTHP-1 cells incubated in RPMI, ranging from 0.4 nmol/7 ⫻ 105 cells released in the first hour to 18 nmol detected after 48 h of incubation. Even if these data may not reflect the actual

O2⫺ production from dTHP-1 cells in the cocultures, they showed that there was a substantial spontaneous superoxide accumulation in the culture medium over time. The low amounts of O2⫺ detected early were probably not responsible for the dramatic reduction in serovar E inclusion numbers; however, later in infection, O2⫺ accumulation may have an inhibitory effect on C. trachomatis growth. Taken together, these data indicated that dTHP-1 cells released an array of mediators with potential antichlamydial activity but that this production was independent of the presence of Chlamydia and probably resulted from a nonspecific preactivation of dTHP-1 cells during PMA-induced differentiation. However, the fact that coculture of infected HeLa cells with dTHP-1 cells resulted in a different pattern of inhibition of C. trachomatis serovar E and L2 growth was interesting and

3240

DESSUS-BABUS ET AL.

INFECT. IMMUN.

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

VOL. 70, 2002

3241

TABLE 2. Analysis of cytokine and NO release by dTHP-1 cells cultivated alone or in coculture with C. trachomatis-infected polarized HeLa cells Supernatanta

U sup E sup L2 sup U-cocult. sup E-cocult sup L2-cocult sup dTHP-1/RPMI sup

Concn (pg/ml) ⫾ SD TNF-␣

IL-1␤

IL-6

IL-8

NO

– – – 35.6 ⫾ 3.9 41.5 ⫾ 6.4 38.3 ⫾ 5.3 46.6 ⫾ 5.1

– – – 113.1 ⫾ 48 114.0 ⫾ 33 112.3 ⫾ 62.2 60.2 ⫾ 26.2

11 ⫾ 1.5 11.7 ⫾ 3.2 14.5 ⫾ 9.5 71 ⫾ 49 85.8 ⫾ 48.7 82.7 ⫾ 28.7 0.45 ⫾ 0.2

12.2 ⫾ 4.4 33.1 ⫾ 15.8* 47.4 ⫾ 18.2* 978 ⫾ 322 1,010 ⫾ 439 1,037 ⫾ 426 ND

ND ND ND – – – 14.9

a U, E, L2 sup, U-, E-, and L2-cocult sup represent basal supernatants from HeLa cells uninfected or infected with C. trachomatis E and L2 cultivated alone or in coculture with dTHP-1 cells, respectively, and collected at 48 hpi. Supernatant from dTHP-1 cells incubated for 48 h in the presence of RPMI only was used as a control for the spontaneous production of cytokines and NO as measured by ELISA and the Griess method, respectively. The cytokine concentrations presented here represent the average values from three independent experiments. Significant differences in cytokine concentrations present in infected HeLa cell supernatants (E sup, L2 sup) or in infected coculture supernatants (E-cocult sup, L2-cocult sup) compared to the respective negative controls (U sup and U-cocult sup, respectively) are indicated by an asterisk (P ⬍ 0.05). –, not detected; ND, not done.

prompted further investigation to identify the factor(s) responsible for this differential inhibition. Since the antichlamydial activity of TNF-␣ and IL-1␤ has been previously reported and since large amounts of IL-8 and some O2⫺ were also detected in coculture and/or dTHP-1 supernatants, coculture experiments were performed in the presence of antibodies specific for these cytokines or with SOD. Coculture in the presence of anti-TNF-␣ antibodies did not result in recovery in the number of serovar E inclusions (Fig. 3A, shaded bars); however, it did result in a partial recovery in infectivity (27% ⫾ 4.6% of the control HeLa alone compared to 15.7% ⫾ 2.1% infectivity in cocultures without antibodies; Fig. 3A, open bars). For serovar L2, the presence of anti-TNF-␣ antibodies did not significantly increase the infectivity yields. In addition, no consistent recovery was detected in the presence of anti-IL-1␤- and anti-IL-8-specific antibodies for either serovar, and the partial recovery in both serovar infectivity found after the addition of SOD to the coculture was not statistically significant (P ⬍ 0.05; Fig. 3). Importantly, the efficiency of neutralization of the TNF-␣ bioactivity present in coculture basal supernatants by the antiTNF-␣ antibodies used was confirmed by a cytotoxicity assay with L929 cells. For IL-1␤ and IL-8, efficiency in antibody binding to the cytokine was controlled by ELISA (data not shown). Interestingly, the addition of exogenous Trp (200 ␮g/ml) to the cocultures resulted in a good recovery in serovar E growth, with an average inclusion number corresponding to 78% ⫾ 3% that of the HeLa alone control compared to 53% ⫾ 2% in cocultures with no Trp (Fig. 3A, shaded bars) and in infectivity (59% ⫾ 4% versus 15.7% ⫾ 2.1%, Fig. 3A, open bars). The addition of Trp also resulted in the total recovery of L2 infectivity (100% ⫾ 14% compared to the control versus 40.3% ⫾

6.4% in the absence of Trp; Fig. 3B). One explanation for this recovery could be an induction of IDO activity in dTHP-1 and/or HeLa cells in the coculture, leading to Trp degradation and therefore to chlamydial inhibition. However, since IFN-␥, a cytokine known to induce IDO expression and abnormal chlamydial growth (19), was not detected in the coculture supernatants, the data obtained with the Trp were unexpected. Upregulation of IDO mRNA expression in dTHP-1 cells and MdM stimulated with basal supernatants of C. trachomatisinfected HeLa cells and in dTHP-1 cells from coculture. Since good recovery in chlamydial growth was found in coculture of infected HeLa and dTHP-1 cells in the presence of exogenous Trp despite the absence of detectable IFN-␥, IDO mRNA expression in both cell types was analyzed by RT-PCR. No IDO mRNA was found in uninfected or C. trachomatis-infected HeLa cells from control cultures or cocultures (data not shown). However, a low level of IDO mRNA expression was detected in dTHP-1 cells cultivated with infected HeLa cells for 24 h (Fig. 5, lanes 12 and 13) and 48 h (lanes 15 and 16). Furthermore, after 24 h of stimulation with serovar E- and serovar L2-infected HeLa cell supernatants, an upregulation in IDO mRNA expression was also detected in dTHP-1 cells (Fig. 5, lanes 2 and 3, respectively) and in MdM (Fig. 5, lanes 7 and 8, respectively) compared to the level of expression in macrophages incubated with supernatants from uninfected HeLa cells or RPMI alone (Fig. 5, lanes 1 and 4, respectively, for dTHP-1 cells and lanes 6 and 9, respectively, for MdM). A strong upregulation in IDO mRNA expression was also detected in macrophages exposed for 12 and 24 h to 200 ng of E. coli LPS/ml (Fig. 5, lanes 5 and 10). RT-PCR analysis of TLR2 and TLR4 mRNA expression in C. trachomatis-infected HeLa cells and in dTHP-1 cells and

FIG. 4. Morphology of chlamydial inclusions in infected HeLa cells cocultured with dTHP-1 cells. Fluorescence microscopy and transmission electron photomicrographs of polarized HeLa cells infected with C. trachomatis serovar E cultivated alone (A and C) or in the presence of dTHP-1 cells (B, D, E, and F). Serovar E inclusions, immunostained with an anti-major outer membrane protein fluorescein isothiocyanate-conjugated antibody, appeared much smaller and in reduced number in cocultures (B) compared to controls (A) (⫻400 magnification). (C) At 48 hpi, inclusions from control cultures contain mostly small, electron-dense EB and some RB. (D to F) A delay in chlamydial maturation was observed in inclusions from cocultures of HeLa and dTHP-1 cells, as well as the presence of damaged RB (arrows, E and F), many outer membrane blebs, and associated dark globular components (arrowheads, D and E). Magnifications: ⫻3,000 (C, E, and F); ⫻2,000 (D).

3242

DESSUS-BABUS ET AL.

FIG. 5. RT-PCR analysis of IDO mRNA expression in dTHP-1 cells and MdM stimulated for 24 h with C. trachomatis-infected HeLa cell supernatants and in dTHP-1 cells cocultivated with infected HeLa cells for 24 and 48 h. The cDNAs were amplified with IDO (324 bp, top panel) and GAPDH (306 bp, bottom panel) primers for 35 and 22 PCR cycles, respectively. A total of 10 ␮l of PCR products was loaded on a 2% agarose gel as follows. dTHP-1 cells and MdM were incubated with supernatants from HeLa cells left uninfected (lanes 1 and 6, respectively), infected with serovar E (lanes 2 and 7, respectively) or serovar L2 (lanes 3 and 8, respectively), treated with RPMI alone (lanes 4 and 9, respectively) or E. coli LPS alone (lanes 5 and 10, respectively), or dTHP-1 cells were incubated in coculture for 24 and 48 h with uninfected HeLa cells (lanes 11 and 14, respectively), with serovar E-infected HeLa cells (lanes 12 and 15, respectively), or with serovar L2-infected HeLa cells (lanes 13 and 16, respectively). A negative control for amplification (lane 17) wherein DNA was omitted and a positive control (lane 18) consisting of cDNA from HeLa cells exposed to rhIFN-␥ (10 ng/ml for 12 h) were also included in the analysis.

MdM stimulated with infected HeLa cell supernatants. According to a previous report, chlamydial LPS induced TNF-␣ release from whole blood via a CD14-dependent pathway (15). Since chlamydial LPS has been shown to be released from infected host cells (41) and since stimulation of dTHP-1 cells and MdM with supernatants from serovar E- and serovar L2infected HeLa cells resulted in TNF-␣ release, the possibility that this increased TNF-␣ production was correlated with an induction or upregulation in TLR expression in macrophages was investigated. The level of TLR2 and TLR4 mRNA expression in dTHP-1 cells and MdM was analyzed by RT-PCR after 12 and 24 h of stimulation with C. trachomatis-infected HeLa cell supernatants. A low level of expression for both TLRs was detected in dTHP-1 cells and in MdM, but semiquantitative analysis with GAPDH as a control for constant level of expression did not reveal a difference in the level of TLR expression in macrophages exposed to control or chlamydia-infected HeLa cell supernatants. Further, neither TLR2 nor TLR4 mRNA was detected in HeLa cells infected for 12 to 48 h with C. trachomatis serovar E or serovar L2 (data not shown). MdM preactivated with supernatants from C. trachomatisstimulated PMN induced an inhibition of chlamydial growth in HeLa cells. Since coculture experiments with PMA-differentiated THP-1 cells provided some information about the potential role of some of the factors released by these macrophages in their overall antichlamydial activity, the next step was to use a more relevant model for studying interactions between macrophages and C. trachomatis-infected epithelial cells. Therefore, cocultures of infected HeLa cells and MdM were undertaken, with MdM seeded either in wells below the culture inserts or suspended in ECM and polymerized to the filter surface opposite the polarized infected HeLa cells. To

INFECT. IMMUN.

our surprise, no inhibition in C. trachomatis serovar E or serovar L2 growth was found under any of the conditions tested (data not shown). This finding contradicted earlier findings by Manor and Sarov (21). Unlike dTHP-1 cells that were preactivated by PMA exposure before use in coculture experiments, MdM were used between 10 and 13 days after isolation without any additional activation. After infection of the genital mucosal epithelial layer during natural infection, there is first an influx of PMN in response to signals released by the infected cells; subsequently, mononuclear cells, including macrophages, migrate to the infectious site (18, 24). In order to mimic this chronology of events, a two-stage coculture experiment was established (see Fig. 1 for experimental details). Although it is well established that chemokine responses are stronger at mucosal basal domains, for technical reasons, PMN were incubated for 12 h with the apical surfaces of C. trachomatis-infected HeLa cells (36 hpi). The apical supernatants were then incubated with MdM for 48 h. Subsequently, the “stimulated” MdM in fresh medium were cocultured with HeLa cells newly infected with serovar E or serovar L2. The results of a representative experiment are presented in Fig. 6. The first apical supernatant obtained from infected HeLa cells incubated for 12 h with PMN inhibited C. trachomatis serovar E growth [71% compared to the control, (EC1/PMN)⫹EC2; Fig. 6A] and infectivity (27%), whereas no inhibition of chlamydial growth or infectivity was detected in the presence of supernatants from control-infected HeLa cells (EC1⫹EC2). More importantly, when the MdM used in the coculture experiments were prestimulated with EC1/PMN supernatants, a slight reduction in serovar E growth (inclusion numbers corresponding to 89% of the control EC2) but a dramatic reduction in infectivity (49% of the control) were found (Fig. 6A). When MdM were preincubated with supernatants from serovar E-infected HeLa cells (EC1) only or from PMN incubated with uninfected HeLa cells (UC1/PMN) or RPMI alone (RPMI/PMN), the level of infection in cocultures was similar to that of the control HeLa cells alone (Fig. 6A). These results clearly demonstrate that, in our two-stage cocultivation in vitro system, a preactivation of MdM by soluble factor(s) derived from PMN incubated with infected HeLa cells was required to initiate antichlamydial activity. For L2 infection and inhibition thereof, the pattern was comparable to that of serovar E (Fig. 6B). Indeed, the first apical supernatants from PMN incubated with L2-infected HeLa cells caused an inhibitory activity on L2 growth, even though the inclusion counts were higher than the inclusion counts in the control L2C1 (197% versus 100% for the control); this conclusion is supported by dramatic reductions in the inclusion size (data not shown) and in the titer of the infectious progeny (58% of the control cultures; Fig. 6B). Also, as was true for serovar E, a strong inhibition in L2 growth (62%) and infectivity (23%) was detected in cocultures by using stimulated MdM preincubated with supernatants from PMN exposed to L2-infected HeLa cells, with inclusion numbers and infectious yields corresponding to 62 and 23% of the controls, respectively [(L2C1/PMN⫹MdM)⫹L2C2, Fig. 6B]. In comparison, there was no effect or only a modest effect on L2 infection when MdM were preincubated with supernatants from serovar

VOL. 70, 2002

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

3243

FIG. 6. Titration of C. trachomatis serovar E and serovar L2 after a 48-h coculture of polarized infected HeLa cells with MdM prestimulated with different culture supernatants. HeLa cells, seeded in inserts and infected with C. trachomatis (culture 1, C1), were cultivated for 36 h, and then 106 PMN were added to the apical side of these cultures for 12 h. Control culture 1, corresponding to HeLa cells incubated for 48 h without the addition of PMN, was also included, as well as a control wherein 106 PMN were incubated in a tissue culture plate for 12 h in RPMI medium alone. The apical medium from HeLa cultures or PMN seeded in plates was then collected and centrifuged to pellet the cells. For convenience, the various supernatants are designated as follows: supernatants of PMN incubated in RPMI alone (RPMI/PMN), with uninfected HeLa cells (UC1/PMN), or with serovar E- or serovar L2-infected cells (EC1/PMN and L2C1/PMN, respectively, and EC1 and L2C1, respectively, corresponding to apical supernatants from E- or L2-infected HeLa cells). These supernatants were then diluted 1:1 in fresh RPMI before being added to MdM (EC1 or L2C1, RPMI/PMN, UC1/PMN, EC1/PMN, or L2C1/PMN ⫹ MdM). After 48 h of stimulation, the culture medium from macrophages was replaced by fresh RPMI, and these stimulated MdM were then cocultivated for 48 additional hours with freshly inoculated cultures of polarized HeLa cells (culture 2, EC2 and L2C2, corresponding to serovar E- and serovar L2-infected cultures). Controls consisting of infected HeLa cells incubated with EC1 and EC1/PMN supernatants diluted 1:1 with fresh medium were also included in these experiments. Serovar E (A) and serovar L2 (B) inclusion counts (shaded bars) and progeny titration after subpassage (open bars) were expressed as percentages of chlamydial growth compared to the control cultures, i.e., infected HeLa cells cultivated alone, representing 100% of growth. These coculture experiments were repeated twice but, since the actual number of MdM per well is a difficult parameter to control, some variations in the level of chlamydial inhibition were found between independent experiments. However, the pattern of inhibition was similar in both experiments. This figure shows the results from a representative experiment.

L2-infected HeLa cells (L2C1) or from PMN incubated with uninfected HeLa cells (UC1/PMN) or RPMI alone (RPMI/ PMN) (Fig. 6B). ELISA of cytokine production in supernatants from the

two-step coculture experiments. For preliminary identification of the cytokine production by transepithelial PMN and stimulated MdM, the supernatants from every step of the two-stage coculture experiments were collected and analyzed by ELISA

3244

DESSUS-BABUS ET AL.

INFECT. IMMUN.

TABLE 3. ELISA of the supernatants from the two-step coculture experiments Groupa

I

Supernatant source and designation

Apical sup from: EC1 EC1/PMN L2C1 L2C1/PMN UC1/PMN RPMI/PMN

Cytokine concn (pg/ml)b TNF-␣

73.2 52.8 0.7

– – –

IL-1␤

229

–

– 104.6 3.3 0.13

II

Sup from MdM stimulated for 48 h UC13MdM UC1/PMN3MdM EC13MdM EC1/PMN3MdM L2C13MdM L2C1/PMN3MdM

– 2.7 (2.7) 6.3 (6.3) 55.4 (18.8) 1.5 (1.5) 278.5 (252.1)

– 0.9 (–) 3.3 (3.3) 102.3 (–) 3.2 (3.2) 369.3 (317)

III

Basal sup from: EC2 EC1/PMN3EC2 UC13MdM3EC2 EC13MdM3EC2 RPMI/PMN3MdM3EC2 UC1/PMN3MdM3EC2 EC1/PMN3MdM3EC2 L2C2 L2C1/PMN3L2C2 UC13MdM3L2C2 L2C13MdM3L2C2 RPMI/PMN3MdM3L2C2 UC1/PMN3MdM3L2C2 L2C1/PMN3MdM3L2C2

– 13.6 (–) – 8.4 5.5 – 13.6 0.7 10.3 (–) – 326.6 3.8 – 588.6

– 101.1 (–) – 8 – – 11.5 0.3 47 (–) – 559 0.2 0.2 564.8

a Group I includes supernatants (Sup) from representative apical supernatants from culture 1 of HeLa cells uninfected or infected with C. trachomatis serovar E (EC1) or L2 (L2C1) and incubated in some experiments with PMN (UC1/PMN, EC1/PMN, and L2C1/PMN, respectively). Controls using PMN incubated with RPMI alone were also included. Group II represents supernatants from MdM incubated for 48 h with group I supernatants. Group III represents basal supernatants from cultures of serovar E-infected (EC2) or serovar L2-infected (L2C2) HeLa cells alone or in coculture with MdM preactivated with group I supernatants. b In some experiments, group I supernatants were used as cytokine inducers. Since these supernatants already contain known concentrations of TNF-␣ and IL-1␤ and since they were used after a 1:1 dilution in fresh medium, the actual cytokine release after induction, shown in parentheses, was calculated as follows: total amount detected ⫺ (concentration in group I supernatant/2). –, not detected; (⫺), no additional cytokine release after induction. The table shows average cytokine concentrations from duplicate wells of a representative experiment. These coculture experiments were repeated only twice because of their complexity, and the cytokine concentrations detected in these independent experiments were not averaged because some differences in the actual numbers were found; these were probably caused by a variation in the MdM responsiveness or in the number/well. However, the patterns of cytokine response were similar in both experiments.

for TNF-␣ and IL-1␤. The data are summarized in Table 3. First, TNF-␣ and IL-1␤ production, as well as IL-8 production, was detected in PMN incubated with C. trachomatis-infected HeLa cells but not in PMN incubated with uninfected cells or RPMI. Interestingly, higher concentrations of these cytokines were consistently released when PMN were stimulated with serovar E-infected cultures versus serovar L2-infected cultures (73.2 versus 52.8 pg of TNF-␣/ml and 229 versus 104.6 pg of IL-1␤/ml [Table 3] and 1,695 versus 1,089 pg of IL-8/ml). Second, when these supernatants were used to stimulate MdM for 48 h (after 1:1 dilution in fresh RPMI), a much greater TNF-␣ release was detected in response to supernatants from PMN stimulated with serovar L2-infected cells than in PMN stimulated with serovar E-infected cells (252.1 and 18.8 pg/ml, respectively). IL-1␤ production from MdM was also relatively high in response to serovar L2-stimulated PMN supernatants, since an additional release of 317 pg/ml was detected. In contrast, no additional IL-1␤ was released when MdM were stimulated with EC1/PMN supernatants. Supernatants from PMN incubated with uninfected HeLa cells (UC1/PMN) or from

uninfected and from serovar E- or serovar L2-infected HeLa cell cultures (UC1, EC1, and L2C1, respectively) induced no or very low TNF-␣ and IL-1␤ production from MdM (Table 3). Third, after the prestimulation step, MdM in fresh medium were used in coculture with serovar E- and serovar L2-infected HeLa cells and, after a 48-h coincubation, basal supernatants were tested for both cytokines. For serovar E infection, only low amounts of TNF-␣ and IL-1␤ were released from MdM even when these cells were preactivated with supernatants from serovar E-stimulated PMN (13.6 and 11.5 pg/ml, respectively; Table 3). On the other hand, large quantities of both cytokines were detected in basal supernatants from cocultures with MdM that were preincubated with apical supernatants from serovar L2-infected cultures (L2C1; 326.6 pg of TNF␣/ml and 559 pg of IL-1␤/ml) and detected even more dramatically with supernatants from serovar L2-stimulated PMN (588.6 and 564.8 pg of TNF-␣ and IL-1␤/ml, respectively). Most importantly, this MdM response was specific for serovar L2 since undetectable or negligible amounts of TNF-␣ and IL-1␤ were released in cocultures with macrophages preincu-

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

VOL. 70, 2002

bated with supernatants from PMN incubated with RPMI alone or with uninfected HeLa cells (Table 3). However, the relationship between TNF-␣ and/or IL-1␤ concentrations and the inhibition of chlamydial growth was not obvious. In summary, these in vitro data show that the cytokine responses of both PMN and MdM to C. trachomatis serovar E and serovar L2 infections are different. Surprisingly, the major inflammatory response, i.e., cytokines TNF-␣ and IL-1␤, released from PMN in response to serovar E-infected HeLa cells was stronger than the release in response to L2 infection, whereas MdM prestimulated with these PMN supernatants produced much higher amounts of proinflammatory cytokines in response to L2 infection. In essence, the role of TNF-␣ and IL-1␤ in chlamydial growth is still unclear, at least in our in vitro model, but one could argue that the more dramatic inhibition of serovar L2 compared to that of serovar E by stimulated MdM may correlate with the presence of higher concentrations of TNF-␣ and IL-1␤. DISCUSSION Tissue damage associated with C. trachomatis genital infections is believed to be immune mediated with a strong inflammatory response that appears to be initiated by the release of proinflammatory cytokines from infected epithelial cells. The release of the anti-inflammatory cytokine IL-11, first detected in vitro in C. trachomatis-infected polarized HeLa cells (9), was then confirmed in vivo in the present study, whereby an upregulation of IL-11 mRNA expression was found in genital tissues from C. trachomatis MoPn-infected mice. Even though IL-11 expression appears to be mainly regulated at the transcriptional level (11), it will be necessary to compare the amounts of IL-11 protein present from infected versus uninfected mice. IL-11 exerts an anti-inflammatory activity on activated macrophages (36, 37), as well as on dendritic cells (2). In a previous study, we detected IL-11 in C. trachomatis-infected cultures of polarized HeLa cells in significantly higher concentrations after infection with the disseminating serovar L2 strain compared to that seen in infection with the nondisseminating serovar E strain (9). One hypothesis to explain this is that the immunosuppressive effects of IL-11 may allow L2 to escape host innate defenses for better dissemination. Data from the present study are somewhat consistent with this hypothesis in that supernatants from serovar L2-infected HeLa cells resulted in a lower TNF-␣ response in macrophages than did supernatants from serovar E-infected cells. Unfortunately, due to unexpected technical difficulties, i.e., preincubation of HeLa cell supernatants with anti-IL-11 neutralizing antibodies resulted in a nonspecific activation of macrophages, a correlation could not be made between the amount of IL-11 released from infected epithelial cells and the intensity of TNF-␣ response from macrophages. Further, considering that the active IL-11 concentrations in MdM range from 10 to 100 ng/ml (37) and are ⬎100 ng/ml in dTHP-1 cells (the present study), C. trachomatis-infected HeLa cell supernatants may not contain a sufficient concentration of IL-11, i.e., 1 ng/ml at the most, to exhibit a detectable inhibitory activity on macrophages in vitro. Although basal supernatants from infected HeLa cells did induce a specific TNF-␣ response from MdM and dTHP-1

3245

cells, the presence of IL-8 and IL-6 in these supernatants did not seem to be the major macrophage activators. This finding suggests that macrophage activation may result from the release by infected HeLa cells of other cytokines, chemokines, or epithelial defensins, although mRNA expression of human beta-defensin 2 (HBD-2) and alpha-defensin 5 (HD5) could not be detected in infected HeLa cells (unpublished observations). Chlamydial LPS might actually be a prime candidate for inducing macrophage activation. Indeed, purified LPS from C. trachomatis serovar F, even if less potent than Salmonella enterica serovar Minnesota and Neisseria gonorrhoeae LPS, induced TNF-␣ release from whole blood ex vivo (15), and IL-1 production has been detected in medium from monocytes incubated with serovar L2 LPS (31). Since several studies reported that chlamydial LPS is present on the surface of infected cells late in the developmental cycle and may also be released into the medium (3, 43), it is likely that C. trachomatis-infected HeLa cell culture supernatants, collected at 48 hpi and used in this study, contained soluble LPS. Unfortunately, in the present study, experiments including a preincubation step of infected HeLa cell supernatants in the presence of antichlamydial LPS antibodies did not result in interpretable data. Recent studies have demonstrated that TLR4 acts as a transmembrane coreceptor to CD14 in the host cell response to bacterial LPS, resulting in NF-␬B activation (45). In the Ingalls et al. study (15), the TNF-␣ response from whole blood to C. trachomatis LPS—albeit structurally and biochemically different from E. coli LPS—seemed to involve the CD14 pathway, and the activation of CD14-transfected CHO cells with chlamydial LPS was associated with an NF-␬B translocation, suggesting that TLR4 may be involved in the CD14-positive cell response to C. trachomatis. In contrast, a recent study (27) showed that activation of bone marrow-derived dendritic cells by C. pneumoniae was mainly dependent on the presence of TLR2, a receptor usually involved in the recognition of grampositive bacteria. In our study, TLR2 and TLR4 mRNA expression was detected in MdM and dTHP-1 cells. However, there was no change in the level of TLR expression in activated macrophages compared to the controls, at least at the messenger level, during exposure to supernatants from C. trachomatisinfected HeLa cells. Perhaps baseline production of TLR protein, i.e., TLR2 and/or TLR4 or others, on macrophages was sufficient to mediate LPS-induced signal transduction and result in cytokine release. Recent data suggest that Nod proteins, which belong to the CARD family proteins, might be sensors of intracellular pathogen components—through the binding of microbial products such as LPS—just as TLR serves to detect conserved components of extracellular pathogens (22). Thus, it is possible that chlamydial products, present in infected epithelial cells, as well as possibly ingested by activated macrophages, could interact with Nod proteins, which are potent activators of NF-␬B (22), and by this pathway initiate a cytokine response in these cells. Interestingly, stimulation of dTHP-1 cells and MdM with C. trachomatis-infected HeLa cell supernatants resulted in the induction of IDO expression, which was also detected in dTHP-1 cells cocultivated with infected HeLa cells. In vitro studies with the monocytic cell line U937 showed that a low level of IDO expression was detected after PMA exposure and

3246

DESSUS-BABUS ET AL.

was strongly enhanced by coexposure to LPS (19). Although PMA and chlamydial LPS may have contributed to IDO induction in dTHP-1 cells, a different mechanism is obviously involved in the case of MdM. The most common IDO inducer in many cell types is IFN-␥; however, IFN-␣ and IFN-␤ can also induce IDO activity in mononuclear phagocytes but not in epithelial cells (19). No detectable IFN-␥ was released either by macrophages cocultivated with C. trachomatis-infected HeLa cells or activated with infected HeLa cell supernatants or by the infected epithelial cells. IFN-␣/␤ secretion from HeLa cells after chlamydial infection is controversial since a study from Devitt et al. (10) showed that an IFN-␣/␤ activity, induced after serovar E and serovar L2 infection in McCoy cells, could not be detected in infected HeLa cells, whereas in an earlier study Jenkins and Lu (16) reported an IFN induction in HeLa 229 cells after infection with the Bour strain. Since we did not test our HeLa cell subclone for IFN-␣/␤ production in response to C. trachomatis infection, we cannot exclude that IDO expression in macrophages was actually induced by IFN␣/␤ and enhanced by macrophage-derived cytokines, as well as chlamydial LPS, but it might also occur via an IFN-independent mechanism. The interactions between chlamydiae and monocytes/macrophages have focused primarily on the intracellular survival and growth of the different chlamydial species in these professional phagocytes (1, 13, 19, 23), and only a few studies have actually investigated the interactions between macrophages and infected epithelial cells. An early report from Manor and Sarov (21) demonstrated an inhibition of serovar L2 infectivity in HEp-2 cells in the presence of MdM that could be partially reversed by addition of anti-TNF-␣ antisera to the cocultures. These authors also reported that human rTNF-␣ inhibited L2 growth in HEp-2 cells (35); the inhibition was accompanied by an increase in epithelial production of prostaglandin E2 (14) and reversed by the addition of Trp or neutralizing antibodies to IFN-␤ (34). In an attempt to investigate more thoroughly the effects of IL-11 on macrophages relative to chlamydia infection consequences, we performed fairly extensive analyses with dTHP-1 cells. Two important conclusions resulted from these experiments. First, while supernatants from cocultures of dTHP-1 cells and C. trachomatis-infected HeLa cells contained an array of cytokines, the release of these effectors from dTHP-1 cells was apparently due more to PMA preactivation than to a response to chlamydia infection, leading us to question the value of this popular cell line in these types of studies. Admittedly, coculture of dTHP-1 cells with C. trachomatis-infected HeLa cells did reduce chlamydial development and the number of infectious progeny upon subpassage, and this effect was different for serovar E versus serovar L2. Further, there was a discriminating recovery in chlamydial inclusion number and progeny infectivity upon supplementation of the cocultures with anti-TNF-␣ antibodies and tryptophan (despite the absence of IFN-␥). However, our second important conclusion is that we believe that the coculture of macrophages, be they dTHP-1 cells or the more appropriate MdM, with chlamydiainfected epithelial cells—in the absence of PMN—may provide misleading information and thus such data should be analyzed with caution. In chlamydial infections of the genital epithelial mucosae, resident macrophages of the submucosa are likely

INFECT. IMMUN.

activated by signals released from both infected epithelial cells and chemotactic PMN, and this chronology of appearance of innate response cells is crucial. For these reasons, a two-stage coculture system was undertaken, which we believe might more closely mimic the natural early infection process. Serovar Eand serovar L2-infected HeLa cells were allowed to interact first with PMN. The supernatants from this coculture were added to MdM, and then the “stimulated” MdM were cocultured with serovar E- or serovar L2-infected HeLa cells. A summary of the progressive TNF-␣ response at each stage is illustrated in Fig. 7. The results were clear, reproducible, and dramatic. The TNF-␣ response in the serovar E infection scenario was modest and dropped during each subsequent stage. In marked contrast, the TNF-␣ response in the serovar L2 infection scenario was lower in the earlier stage but increased fivefold in stage II and doubled yet again in stage III, suggesting that presence of large amounts of epithelial IL-11, i.e., ca. 1 ng/ml, released in response to L2 infection (9) did not prevent a massive production of proinflammatory mediators from activated MdM, at least in this in vitro model. However, since several reports showed that IL-11, besides its activity on macrophages, appears also to affect dendritic cell functions and the Th1 response (2, 33), we cannot exclude that, if the expression of this cytokine is upregulated in vivo after chlamydial infection (as suggested by our in vivo study using infected mice), it may interfere somewhat with the immune response to chlamydia, particularly in the case of the disseminating serovar L2. Using this two-stage coculture system, we also found that significant numbers of viable chlamydial progeny were still present after coculture of infected HeLa cells with stimulated MdM for 48 h, i.e., 49 and 23% of the total infectious progeny of serovar E and L2, respectively (summarized in Fig. 7); these data suggested that the early proinflammatory response resulting from the successive activation of PMN and then MdM exposed to C. trachomatis-infected epithelial cells was not sufficient to efficiently kill all intracellular organisms. The significance of these in vitro results is uncertain, since during natural infection the inflammatory process is probably amplified over time and involves other immune cell types and effectors which may contribute to eradicate chlamydiae. In this regard, a recent study showed that TNF-␣ had no effect on the in vitro growth of the human strains serovar D and L2 or of the murine MoPn strain of C. trachomatis when murine intestinal epithelial cell lines were used (25). However, in the same study Perry et al. (25) found, in contrast, that genital inoculation of mice bearing targeted mutations in the TNF-␣ p55 receptor molecule with C. trachomatis serovar D or MoPn did result in a significant reduction in the rate of chlamydial clearance compared to that seen in wild-type mice. Since no direct inhibitory effect of TNF-␣ was determined in vitro in their study, the authors proposed that the delay in clearance was due to indirect contribution of this cytokine in vivo by activation of other cells and/or factors present in the environment of the genital mucosa. In summary, our two-stage coculture system highlighted new information at the cellular level in vitro about the complex interactions between infected epithelial cells and professional phagocytes, such as PMN and macrophages, and showed differences in the early inflammatory response generated after epithelial infection with a nondisseminating versus that occur-

INNATE IMMUNE RESPONSES TO C. TRACHOMATIS SEROVARS

VOL. 70, 2002

3247

FIG. 7. Summary of in vitro TNF-␣ production and chlamydial survival during chronological appearance of PMN and MdM to chlamydiainfected HeLa cells. The left part of the chart presents a summary of the protocol used (for more details, see Fig. 1). Briefly, polarized HeLa cells infected with C. trachomatis serovar E or serovar L2 for 36 h were incubated in the presence of PMN. After 12 h of incubation, the supernatants from infected HeLa/PMN cocultures (SUPERNATANT I) were collected and used, after 1:1 dilution in fresh medium, to activate MdM. After 48 h of incubation, MdM supernatants (SUPERNATANT II) were collected, and these “stimulated” MdM, in fresh medium, were then cocultured with fresh HeLa cells infected with the respective serovars of C. trachomatis (culture 2). At 48 hpi, basal supernatants from cocultures were collected (SUPERNATANT III), and chlamydial inclusions and infectious progeny were counted. The right portion of the chart presents the progression in TNF-␣ production at each step of the two-stage coculture experiments by using serovar E (E) or serovar L2 (L2), with TNF-␣ concentrations in supernatants I, II (TNF-␣ released by the “stimulated” MdM only), and III represented by the shaded, open, and solid bars, respectively. The numbers shown on the graph above the open and solid bars represent the infectivity yields obtained when serovar E- or serovar L2-infected HeLa cells (culture 2) were cultured in the presence of the respective supernatant I or cocultured in the presence of PMN preactivated MdM, respectively. The infectious progeny titers were expressed as percentages of the control infected HeLa cells cultivated alone. The graph summarizes the data presented in Table 3 (TNF-␣ ELISA) and Fig. 6 (chlamydia titration).

ring after epithelial infection with a disseminating strain of C. trachomatis. It should be emphasized that the use of HeLa cells as a model of cervical epithelial cells might not be ideal since fully differentiated primary cells are probably better armed to fight infection and may express additional mediators, particularly those involved in innate defenses such as defensins or TLR. Further, since it now appears that dendritic cells may be more efficient than macrophages in initiating cell-mediated immunity, use of a two-stage coculture system employing infected epithelial cells, dendritic cells, and PMN may be warranted. ACKNOWLEDGMENTS This study was supported by Public Health Service grants AI13446 (P.B.W.) and AI43337 (T.L.D.). We gratefully acknowledge the use of the Transmission Electron Microscopy Core Facility in the Department of Pathology, East Tennessee State University. REFERENCES 1. Airenne, S., H. M. Surcel, A. Bloigu, K. Laitinen, P. Saikku, and A. Laurila. 2000. The resistance of human monocyte-derived macrophages to Chlamydia pneumoniae infection is enhanced by interferon-gamma. APMIS 108:139– 144. 2. Bozza, M., J. L. Bliss, A. J. Dorner, and W. L. Trepicchio. 2001. Interleukin-11 modulates Th1/Th2 cytokine production from activated CD4⫹ T cells. J. Interferon Cytokine Res. 21:21–30.

3. Campbell, S., S. J. Richmond, P. S. Yates, and C. C. Storey. 1994. Lipopolysaccharide in cells infected by Chlamydia trachomatis. Microbiology 140: 1995–2002. 4. Cario, E., I. M. Rosenberg, S. L. Brandwein, P. L. Beck, H.-C. Reinecker, and D. K. Podolsky. 2000. Lipopolysaccharide activates distinct signaling pathways in intestinal epithelial cell lines expressing Toll-like receptors. J. Immunol. 164:966–972. 5. Currier, A. R., M. H. Ziegler, M. M. Riley, T. A. Babcock, V. P. Telbis, and J. M. Carlin. 2000. Tumor necrosis factor-␣ and lipopolysaccharide enhance interferon-induced antichlamydial indoleamine dioxygenase activity independently. J. Interferon Cytokine Res. 20:369–376. 6. Darville, T., C. W. Andrews, Jr., and R. G. Rank. 2000. Does inhibition of tumor necrosis factor alpha affect chlamydial genital tract infection in mice and guinea pigs? Infect. Immun. 68:5299–5305. 7. Darville, T., C. W. Andrews, K. K. Laffoon, W. Shymasani, L. R. Kishen, and R. G. Rank. 1997. Mouse strain-dependent variation in the course and outcome of chlamydial genital tract infection is associated with differences in host response. Infect. Immun. 65:3065–3073. 8. Darville, T., K. K. Laffoon, L. R. Kishen, and R. G. Rank. 1995. Tumor necrosis factor alpha activity in genital tract secretions of guinea pigs infected with chlamydiae. Infect. Immun. 63:4675–4681. 9. Dessus-Babus, S., S. T Knight, and P. B. Wyrick. 2000. Chlamydial infection of polarized HeLa cells induces PMN chemotaxis but the cytokine profile varies between disseminating and non-disseminating strains. Cell. Microbiol. 2:317–327. 10. Devitt, A., P. A. Lund, A. G. Morris, and J. H. Pearce. 1996. Induction of alpha/beta interferon and dependent nitric oxide synthesis during Chlamydia trachomatis infection of McCoy cells in the absence of exogenous cytokine. Infect. Immun. 64:3951–3956. 11. Du, X., and D. A. Williams. 1997. Interleukin-11: review of molecular, cell biology, and clinical use. Blood 89:3897–3908. 12. Du, X., E. T. Everett, G. Wang, W.-H. Lee, Z. Yang, and D. A. Williams. 1996.

3248

13. 14.

15. 16. 17.

18.

19. 20. 21. 22. 23.

24. 25. 26. 27.

28. 29.

DESSUS-BABUS ET AL.

Murine interleukin-11 is expressed at high levels in the hippocampus and expression is developmentally regulated in the testis. J. Cell. Physiol. 168: 362–372. Heinemann, M., M. Susa, U. Simnacher, R. Marre, and A. Essig. 1996. Growth of Chlamydia pneumoniae induces cytokine production and expression of CD14 in a human monocytic cell line. Infect. Immun. 64:4872–4875. Holtmann, H., Y. Shemer-Avni, K. Wessel, I. Sarov, and D. Wallach. 1990. Inhibition of growth of Chlamydia trachomatis by tumor necrosis factor is accompanied by increased prostaglandin synthesis. Infect. Immun. 58:3168– 3172. Ingalls, R. R., P. A. Rice, N. Qureshi, K. Takayama, J. S. Lin, and D. T. Gloenbock. 1995. The inflammatory cytokine response to Chlamydia trachomatis infection is endotoxin mediated. Infect. Immun. 63:3125–3130. Jenkins, H. M., and Y. K. Lu. 1967. Induction of interferon by the Bour strain of trachoma in HeLa 229 cells. Am. J. Ophthalmol. 63:1110–1115. Kelly, K. A., E. A. Robinson, and R. G. Rank. 1996. Initial route of antigen administration alters the T-cell cytokine profile produced in response to the mouse pneumonitis biovar of Chlamydia trachomatis following genital infection. Infect. Immun. 64:4976–4983. Kiviat, N. B., J. A. Paavonen, P. Wolner-Hanssen, C. W. Critchlow, W. E. Stamm, J. Douglas, D. A. Eschenbach, L. A. Corey, and K. K. Holmes. 1990. Histopathology of endocervical infection caused by Chlamydia trachomatis, herpes simplex virus, Trichomonas vaginalis, and Neisseria gonorrhoeae. Hum. Pathol. 21:831–837. La Verda, D., and G. I. Byrne. 1994. Interactions between macrophages and chlamydiae. Immunol. Ser. 60:381–399. Lu, H., C. Shen, and R. C. Brunham. 2000. Chlamydia trachomatis infection of epithelial cells induces the activation of caspase-1 and release of mature IL-18. J. Immunol. 165:1463–1469. Manor, E., and I. Sarov. 1988. Inhibition of Chlamydia trachomatis replication in HEp-2 cells by human monocyte-derived macrophages. Infect. Immun. 56:3280–3284. Martin, S. J. 2001. Dealing the CARDs between life and death. Trends Cell Biol. 11:188–189. Nettelnbreker, E., H. Zeidler, H. Bartels, U. Dreses-Werringloer, W. Daubener, H. Holtmann, and L. Kohler. 1998. Studies of persistent infection by Chlamydia trachomatis serovar K in TPA-differentiated U937 cells and the role of IFN-gamma. J. Med. Microbiol. 47:141–149. Patton, D. L., and C. C. Kuo. 1989. Histopathology of Chlamydia trachomatis salpingitis after primary and repeated reinfections in the monkey subcutaneous pocket model. J. Reprod. Fertil. 85:647–656. Perry, L. L., H. Su, K. Feilzer, R. Messer, S. Hughes, W. Whitmire, and H. D. Caldwell. 1999. Differential sensitivity of distinct Chlamydia trachomatis isolates to IFN-␥-mediated inhibition. J. Immunol. 162:3541–3548. Pick, E. 1986. Microassays for superoxide and hydrogen peroxide production and nitroblue tetrazolium reduction using an enzyme immunoassay microplate reader. Methods Enzymol. 132:407–421. Prebeck, S., C. Kirschning, S. Durr, C. da Costa, B. Donath, K. Brand, V. Redecke, H. Wagner, and T. Miethke. 2001. Predominant role of Toll-like receptor 2 versus 4 in Chlamydia pneumoniae-induced activation of dendritic cells. J. Immunol. 167:3316–3323. Rank, R. G., and P. M. Bavoil. 1996. Prospects for a vaccine against Chlamydia genital disease: immunity and vaccine development. Bull. Inst. Pasteur 94:55–82. Rank, R. G., K. H. Ramsey, E. A. Pack, and D. M. Williams. 1992. Effect of

Editor: R. N. Moore

INFECT. IMMUN.

30.

31. 32. 33. 34. 35. 36. 37.

38.

39. 40.

41. 42. 43. 44. 45.

gamma interferon on resolution of murine chlamydial genital infection. Infect. Immun. 60:4427–4429. Rasmussen, S. J., L. Eckmann, A. J. Quayle, L. Shen, Y. X. Zhang, D. J. Anderson, J. Fierer, R. S. Stephens, and M. F. Kagnoff. 1997. Secretion of proinflammatory cytokines by epithelial cells in response to Chlamydia infection suggests a central role for epithelial cells in chlamydial pathogenesis. J. Clin. Investig. 99:77–87. Rothermel, C. D., J. Schachter, P. Lavrich, E. V. Lipsitz, and T. Francus. 1989. Chlamydia trachomatis-induced production of interleukin-1 by human monocytes. Infect. Immun. 57:2705–2711. Schachter, J., and A. O. Osoba. 1983. Lymphogranuloma venereum. Br. Med. Bull. 39:151–154. Schwertschlag, U. S., W. L. Trepicchio, K. H. Dykstra, J. C. Keith, K. J. Turner, and A. J. Dorner. 1999. Hematopoietic, immunomodulatory and epithelial effects of interleukin-11. Leukemia 13:1307–1315. Shemer-Avni, Y., D. Wallach, and I. Sarov. 1989. Reversion of the antichlamydial effect of tumor necrosis factor by tryptophan and antibodies to beta interferon. Infect. Immun. 57:3484–3490. Shemer-Avni, Y., D. Wallach, and I. Sarov. 1988. Inhibition of Chlamydia trachomatis growth by recombinant tumor necrosis factor. Infect. Immun. 56:2503–2506. Trepicchio, W. L., L. Wang, M. Bozza, and A. J. Dorner. 1997. Interleukin-11 regulates macrophage effector function through the inhibition of nuclear factor-␬B. J. Immunol. 159:5661–5670. Trepicchio, W. L., M. Bozza, G. Pedneault, and A. J. Dorner. 1996. Recombinant human IL-11 attenuates the inflammatory response through downregulation of proinflammatory cytokine release and nitric oxide production. J. Immunol. 157:3627–3634. Van Voorhis, W. C., L. K. Barrett, Y. T. Cosgrove Sweeney, C.-C. Kuo, and D. L. Patton. 1997. Repeated Chlamydia trachomatis infection of Macaca nemestrina fallopian tubes produces a Th1-like cytokine response associated with fibrosis and scarring. Infect. Immun. 65:2175–2182. Ward, M. E. 1999. Mechanisms of Chlamydia-induced disease, p. 171–210. In R. S. Stephens (ed.), Chlamydia: intracellular biology, pathogenesis, and immunity. American Society for Microbiology, Washington, D.C. Williams, D. M., D. M. Magee, L. F. Bonewald, J. G. Smith, C. A. Bleicker, G. I. Byrne, and J. Schachter. 1990. A role in vivo for tumor necrosis factor alpha in host defense against Chlamydia trachomatis. Infect. Immun. 58: 1572–1576. Wyrick, P. B., S. T. Knight, T. R. Paul, R. G. Rank, and C. S. Barbier. 1999. Persistent chlamydial envelope antigens in antibiotic-exposed infected cells trigger neutrophil chemotaxis. J. Infect. Dis. 179:954–966. Wyrick, P. B., D. G. Gerbig, Jr., S. T. Knight, and J. E. Raulston. 1996. Accelerated development of genital Chlamydia trachomatis serovar E in McCoy cells grown on microcarrier beads. Microb. Pathog. 20:31–40. Wyrick, P. B., J. Choong, S. T. Knight, D. Goyeau, E. S. Stuart, and A. B. MacDonald. 1994. Chlamydia trachomatis antigens on the surface of infected human endometrial epithelial cells. Immunol. Infect. Dis. 4:131–141. Wyrick, P. B., J. Choong, C. H. Davis, S. T. Knight, M. O. Royal, A. S. Maslow, and C. R. Bagnell. 1989. Entry of genital Chlamydia trachomatis into polarized human epithelial cells. Infect. Immun. 57:2378–2389. Zhang, G., and S. Ghosh. 2001. Toll-like receptor-mediated NF-kB activation: a phylogenetically conserved paradigm in innate immunity. J. Clin. Investig. 107:13–19.