Expression of inflammatory host genes in Chlamydia trachomatis-infected human monocytes

Blood samples from three healthy donors were used. Peripheral blood mononuclear cells were separated according to the standard Ficoll-Histopaque procedure and incubated in a tissue-culture plate for 20 minutes at room temperature. The non-adherent cell fraction was carefully removed by washing the culture plate two times with AIM-V medium (Gibco-Invitrogen GmbH, Karlsruhe, Germany). The adherent cell fraction contained more than 80% of cells with macrophage-like appearance as determined by inverted microscope and described earlier [5, 6]. C. trachomatis was acquired from the Washington Research Foundation (Seattle, WA, USA) and multiplied in the human larynx carcinoma epithelial cell line (Hep2) in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum (Biochrom AG, Berlin, Germany), 1% (wt/vol) L-glutamine, and 0.1% (wt/vol) gentamycin. After inoculation for 48 hours, Chlamydia were harvested, purified on a discontinuous urographin gradient (Schering, Berlin, Germany) as described in Caldwell and colleagues [7], resuspended in sucrose-phosphate-glutamate (SPG) buffer, and finally stored at -80°C until use.

To determine chlamydial infectivity, sequentially diluted chlamydial probes were titrated on confluent monolayers of Hep2 cells. As controls, pure elementary bodies (EBs) at various multiplicities of infection (MOIs) were used. Multiple passages were performed to enhance recovery of released Chlamydia. After 48 hours, cultures were terminated by the addition of absolute methanol followed by an indirect immunoperoxydase assay (IPAzyme test; medac GmbH, Hamburg, Germany) for visualization of chlamydial inclusions [8]. For this purpose, serum of a patient positive for anti-chlamydial antibodies with a specific immunoglobulin G (IgG) titer of 1:1,024 (IPAzyme test) was used. After overnight incubation with the antibody, the second peroxidase-conjugated goat anti-human IgG antibody and subsequently 4-chloro-1-naphtol (Savyon Diagnostics Ltd., Beer Sheva, Israel) were added. Chlamydial inclusions were identified by light microscopy, and the number of inclusions was expressed as inclusion-forming units (IFU) per milliliter of the titrated lysate.

Using 3 × 10⁷ cells per well, monocytes were cultured for 4 hours in six-well plates in RPMI 1640 medium (Invitrogen Corporation) supplemented with 10% human serum, 1% L-glutamine, and 0.1% gentamycin at 37°C in an atmosphere of 5% CO₂ and subsequently inoculated with purified C. trachomatis EB serovar K (UW/31/Cx) for 4 hours at an MOI of 5:1. The chlamydial suspension, which was free of mycoplasma as determined by polymerase chain reaction (PCR), contained 1.4 × 10⁸ EB IFU per 50 μl.

Unabsorbed Chlamydia were removed 4 hours post-infection (pi) by washing the plates three times in RPMI growth medium containing 10% human serum. For recultivation, fresh medium was added on days 1 and 7. Morphology of monocytes in culture was monitored microscopically on a daily basis, and their viability was tested by trypan blue dye exclusion tests on days 1 and 7. At least 80% of infected monocytes were viable on day 7. The cells were harvested at the three time points mentioned below by gently scraping with a rubber policeman. Samples were stored at -80°C until use. The detailed procedure has been described earlier [9].

To investigate the expression of transcripts encoding inflammatory mediators over the time course, the three time points of 4 hours (4 h pi), 1 day (1 d pi), and 7 days (7 d pi) were chosen because C. trachomatis is known to start replication in monocytes within the first day (early phase of productive cycle, in this context called 'active' infection) and starts generating a persistent infection after 24 hours [10‐12].

Messenger RNA of C. trachomatis-infected monocytes of the three healthy donors was compared with the corresponding mock-infected samples. For mock infection, SPG buffer was used instead of the C. trachomatis suspension, and all other procedures were performed identically. Cells were resuspended in solution D, and total RNA was extracted by application of phenol/chloroform 5:1 (pH 4.5) (Ambion, Inc., Austin, TX, USA) followed by precipitation in isopropanol at -80°C for 1 hour and incubation with RQ1 DNase (Promega Corporation, Madison, WI, USA) at 37°C for 1 hour.

The signal intensity of the glyceraldehyde-3-phosphate dehydrogenase (G3PDH) housekeeping gene on the microarray membrane was used as an indirect quality marker for the RNA used because only experiments that revealed a G3PDH signal intensity within 1.5 times the standard deviation of all membranes evaluated in our study were included in the analysis. The entire microarray procedure and its analysis have been validated and reported in detail [6]. Total RNA of all donors was pooled, and for each microarray experiment, 150 μg was reverse-transcribed and amplified by SMART™-PCR (Clontech, Mountain View, CA, USA), a technology that allows reverse transcription (RT) of small amounts of total RNA and subsequent amplification of the entire cDNA. Probes were labeled with phosphorus 32 and subsequently hybridized overnight to a filter-based nylon membrane containing immobilized cDNA-specific sequences from a total of 1,184 genes (Human Atlas Array 1.2; BD Biosciences Clontech). For analysis, we focused solely on the 159 cytokines, chemokines, and their receptors as given by the manufacturer (see [13] for detailed information). Signal intensities were retrieved by a STORM 860 scanner (Molecular Dynamics, now part of GE Healthcare, Little Chalfont, Buckinghamshire, UK) in combination with the AtlasImage 2.0 software (Clontech, Mountain View, CA, USA). Data from all signal intensities were then subtracted by the local background intensity measured around each gene. The local background intensities of all individual genes were subsequently averaged, resulting in the mean background intensity of a particular membrane. Gene expression in our experiments was determined by a spot intensity of a single transcript that exceeded twice the mean background. Normalization to background and to the G3PDH housekeeping gene was achieved by the global (sum) normalization method. To test the reproducibility of array measurements between the mock-infected and the C. trachomatis-infected probes, correlation matrices using Pearson correlation revealed coefficients of 0.93 (time point 4 h pi), 0.85 (time point d 1 pi), and 0.91 (time point d 7 pi). The semiquatitative interpretation (-/+/++/+++) of the differential regulation can be found in the legends of Tables 1, 2, 3. Signal intensity data are given as the ratio of C. trachomatis-infected versus mock-infected probes. The GEO (Gene Expression Omnibus) accession number is GSE7601.

Expression levels of three of the four genes found to be expressed at all time points of infection (4 h, 1 d, and 7 d pi) were selected for RT-PCR measurements as described previously [9]. Total RNA from C. trachomatis-infected monocytes of five additional healthy donors was used for confirmation of the array data. Again, the corresponding mock-infected samples were used for comparison. Measurements of the pooled RNA were performed in triplicate using normalization to the G3PDH gene. Data are given as mean values. Primer sequences were retrieved by BLAST (basic local alignment search tool) search and obtained from MWG-Biotech AG (Ebersberg, Germany). They are given as follows:

Macrophage inflammatory protein-1-alpha (MIP-1-α): sense 5'-CCGACCGCCTGCTGCTTCA-3', antisense 5'-CTGCCGGCTTCGCTTGGTTAG-3'; MIP-1-beta (MIP-1-β): sense 5'-CACCGCCTGCTGCTTTTCT-3', antisense 5'-GACTTGCTTGCTTCTTTTGGTT-3'; interleukin-2 receptor-gamma (IL-2R-γ): sense 5'-CACTGGGGGAGCAATACTTCAAAA-3', antisense 5'-GGGGCATCGTCCGTTCC-3'.

All followed procedures have been approved by the local ethics committee of the Hannover Medical School.

Three time points pi were analyzed. Of the 159 genes encoding cytokines, chemokines, and their receptors, 15 (9%) were expressed at 4 h pi, 17 (11%) at 1 d pi, and 12 (8%) at 7 d pi. Among them were four genes found to be differentially expressed at all three time points: interferon-gamma (IFN-γ), MIP-1-α, MIP-1-β, and IL-2R-γ. Apart from genes confirming the monocytic origin of the samples, we identified several interferon-related transcripts. Data are given in Table 1. Ten mRNA transcripts were induced exclusively in early infection at 4 h pi. The highest factors of regulation compared to the corresponding mock-infected monocytes were observed at 4 h pi for IL-6, tumor necrosis factor-alpha (TNF-α), small inducible cytokine A1 (SCYA1, now termed CCL1), IL-4, IL-18, and IL-17. Eight mRNA transcripts were identified to be exclusively differentially regulated after 1 d. IL-5 was expressed at both time points (4 h and 1 d pi), showing a decreasing level of expression with ongoing infection. Four genes were induced exclusively in persistent infection (7 d pi): transforming growth factor-beta-1 (TGF-β-1), monocyte chemotactic protein-1 (MCP-1), MIP-2-α, and TNF receptor (TNFR). In general, most of the expression levels at 1 d and 7 d pi were clearly lower as compared with 4 h pi. Apart from IFN-γ, all transcripts found to be induced at all three time points of infection by the microarray analysis (MIP-1-α, MIP-1-β, and IL-2R-γ) were subjected to quantitative RT-PCR in order to confirm their level of expression (Figures 1, 2, 3).

To test whether chlamydial lipopolysaccharide (LPS) or other cell-wall components might have an effect on gene induction, we performed RT-PCR measurements on the selected genes (MIP-1-α, MIP-2-β, and IL-2R-γ) using both UV irradiation-inactivated and heat-inactivated C. trachomatis-infected monocytes. For this purpose, chlamydial EBs were treated with UV light for 1 hour (Stratalinker™; Stratagene, La Jolla, CA, USA) or attended to heat at 100°C for 20 minutes. For each sample, a control experiment was performed. Taken together, the overall level of MIP-1-α, MIP-1-β, and IL-2R-γ induction did not vary substantially for viable, UV-irradiated, and heat-inactivated C. trachomatis, respectively. However, there was an increase in expression for MIP-1-α in monocytes incubated with UV-inactivated Chlamydia (slight increase at 4 h and 1 d pi and a stronger increase on 7 d pi; data not shown), indicating that chlamydial LPS or other components of the cell wall might be relevant for the observed gene induction.