Eighty-six male wild-type C57BL/6 and 32 male homozygous TNF knockout (TNF^−/−) mice weighing 27–29 g were used in the study. The TNF^−/− mice (Alexander Fleming Institute, Vari, Athens, Greece) were generated on a C57BL/6 genetic background as described (8). The study received permission from the Veterinary Directorate of the Prefecture of Athens according to the Greek legislation in conformance to the 160/1991 Council Directive of the EU. Animals were housed in metal cages and maintained under specific pathogen-free conditions. Room temperature ranged between 18 and 22˚C, relative humidity between 55 and 65%, and the light/dark cycle was 6:00 a.m./6:00 p.m.

Pseudomonas aeruginosa 2 was isolated from the blood of a patient with acute pyelonephritis and severe sepsis. The isolate was multidrug-resistant (MDR) as defined by Clinical and Laboratory Standards Institute (CLSI) criteria. The isolate was stored in multiple aliquots in skim milk (Oxoid Ltd, London, UK) at −70˚C. Before each experiment, one aliquot was thawed and cultured on MacConkey agar plates (Becton Dickinson, Cockeysville, MD, USA). Single colonies were suspended in Mueller–Hinton broth (Oxoid) and incubated for 12 h at 37˚C in a shaking water bath. The resulting inoculum was washed three times with 0.9% NaCl to remove free endotoxin; it was diluted to 1 × 10⁷ cfu/ml and a final volume of 0.15 ml was administered using intraperitoneal (i.p.) injection under light ether anesthesia.

Animals were randomly assigned into four groups. Two animals per group were studied per day of experiment.

Six animals in each group were sacrificed 6 h after bacterial challenge. This time point was selected based on preliminary experiments with 20 mice sacrificed at serial time intervals post bacterial challenge showing that the peak expression of TREM-1 on neutrophils was observed 6 h after bacterial challenge (data not shown). Survival was recorded for the remaining 10 animals of each group at 6-h intervals for 7 days. Mice were sacrificed by intramuscular (i.m.) injection of 25 mg/kg ketamine. An abdominal midline incision was then performed under sterile conditions, and the intestines were displaced to the left to allow blood sampling by puncture of the lower vena cava with a sterile 26G needle. Blood was collected into EDTA-coated tubes (Vacutainer, Becton Dickinson) and centrifuged, and the serum was transferred to a new tube and stored at −70˚C until assayed. TNFα (R&D Inc, Minneapolis, MD, USA) was measured in plasma by an enzyme immunoabsorbent assay (lower limit of detection 15.6 pg/ml). The sedimented pellets that remained after removal of the serum were re-suspended in Phosphate Buffered Saline (PBS, pH 7.2, Biochrom, Berlin, Germany) and the erythrocytes were lysed after adding 1 mM NH₄Cl. Leukocytes were washed three times with PBS and incubated with mouse anti-TREM-1 antibody PE (R&D Inc) as described below. Leukocyte subsets were analyzed by the EPICS XL/MSL flow cytometer (Beckman Coulter, Miami, FL, USA) after selection for neutrophils by characteristic Forward/Side scattering. Neutrophils from two uninfected (sham) mice were used as negative controls. Expression of TREM-1 was given as % and mean fluorescence intensity (MFI).

In parallel, specimens of liver were aseptically excised and transferred to separate sterile containers. The livers were weighed and homogenized in 1 ml of Mueller–Hinton broth and diluted four consecutive times using serial 1:10 dilutions in 0.9% NaCl. A 0.1 ml aliquot of each dilution was plated onto MacConkey agar plates. Plates were incubated for 48 h at 35˚C, and the number of colonies were counted and multiplied by the corresponding dilution factor. The results were expressed as log10 cfu/g. The minimum detection limit was 10 cfu/g.

One million cells of the U937 human monocytic cell line were inoculated into each well of a 12-well plate (final volume 2.4 ml). The cell culture medium consisted of RPMI 1640 (Biochrom) supplemented with 2 mM of glutamine (Biochrom), 10% Fetal Bovine Serum (Biochrom), 100 U/ml of penicillin G and 0.1 mg/ml of streptomycin (Sigma Co). The cultures were incubated at 37˚C under 5% CO₂ for 24 h. Cultures were incubated for 1 h with medium alone (control), 10 μM dexamethasone, or 10 μM of hydrocortisone. Following pre-incubation, cells were stimulated by adding 10 ng/ml of Escherichia coli O55:B5 (LPS, Sigma) or 1 × 10⁶ cfu/ml of heat-killed P. aeruginosa 2. Heat killing was achieved after incubation for 6 h at 65˚C in a shaking water bath, and lack of growth was determined following inoculation in growth media. In separate experiments, U937 cells were pre-incubated with dexamethasone or hydrocortisone as described above, in the presence of either TNFα (100 pg/ml, Sigma) or anti-TNFα antibody (100 pg/ml, Serotec, Marseille, France). Cells were then stimulated with 10 ng/ml LPS. After 24 h of incubation, the cultured cells were transferred to a sterile tube, centrifuged, and the supernatants were transferred to separate tubes and stored until analysis. sTREM-1 levels were measured in the supernatants by ELISA (R&D Inc). The cell pellets were diluted with 1 ml of PBS and 100 μl of the cell suspension was transferred to a new tube. Ten microliters of human anti-TREM-1 monoclonal antibody labeled with phycoerythrin (PE emission 570 mm, R&D Inc) was added to the mixture and incubated for 40 min at 4˚C. The cells were then washed with PBS pH 7.2 and centrifuged, and the cells were analyzed by the EPICS XL/MSL flow cytometer. Unstained cells and antibody isotype controls [cells stained with IgG-PE (Beckman Coulter)] were used as negative controls. Expression of TREM-1 was expressed as the percent of TREM-1-positive cells compared to the entire population of leukocytes gated, and MFI was calculated based on the TREM-1-positive population after subtracting the background measured in the antibody isotype control cells.

To measure the effect of dexamethasone on expression of the TREM-1 gene, U937 monocytes were stimulated with 1 × 10⁶ cfu/ml of heat-killed P. aeruginosa 2 without or with 10 μM dexamethasone for 8 h. After stimulation, the cultured cells were collected and centrifuged, and the supernatant discarded. The cell pellets were treated with consecutive treatments of Trizol (AppliChem, GmbH, Germany) and chloroform, and contaminating DNA was removed by treatment for 30 min at 37˚C with 0.04 U/μl of DNAase (Ambion). To produce cDNA, 1.5 μg of RNA was incubated with 0.4 mM of dNTPs (New England BioLabs, Ipswitch, MA, USA), 1.0 U of RNA-sin (New England BioLabs), 10 mM DTT (New England BioLabs), and 5× reverse transcriptase buffer. In a Mastercycler 5330 (Eppendorf), the RNA was initially incubated for 10 min at 65˚C. One microunit of reverse transcriptase (New England BioLabs) was added and the reactions consecutively incubated at: 10 min at 25˚C; 50 min at 42˚C; and 15 min at 70˚C. Samples without reverse transcriptase were used as blanks, and cDNA was kept at −80˚C until assayed.

Quantitative PCR was performed in 96-well plates with a iQ5 Real Time PCR Detection System (BioRad Laboratories Inc, USA). Quantification of TREM-1 mRNA levels was calculated relative to the housekeeping gene encoding β2-microglobulin. Reactions were performed in a final volume of 20 μl containing 10 μl FluocycleTM II SYBR Master Mix (EuroClone S.p.a-Italy), 7.0 μl PCR-grade water (AppliChem), 1.0 μl cDNA, and 1.0 μl each of the 20 μM forward and reverse gene-specific primers. The reactions were incubated at 95˚C for 5 min, followed by 40 PCR cycles. Each cycle consisted of three steps: denaturation (95˚C for 10 s), annealing (60˚C for 30 s), and extension (75˚C for 30 s). Blanks were subjected to this protocol as well. Melting curve analyses was performed for all genes and confirmed the specificity of the PCR products by the presence of a single peak. PCR products were visualized on 3% agarose gels and stained with ethidium bromide (AppliChem, GmbH, Germany). PCR primers were: TREM-1, sense 5′-TGG TCT TCT CTG TCC TGT TTG-3′; antisense 5′-ACT CCC TGC CTT TTA CCT C-3; β2-microglobulin, sense 5′-ATG AGT ATG CCT GCC GTG TG-3′; antisense, 5′-CCA AAT GCG GCA TCT TCA AAC-3′ (10). The number of transcripts was measured using the PFAFFL equation (11).

Results are presented as means ± SE. Comparisons were done by the Mann–Whitney U test with Bonferroni corrections for multiple comparisons. Survival was calculated by Kaplan–Meier analysis; comparisons between groups were performed by the log-rank test. Percent changes after treatment with dexamethasone were calculated by the formula 100 × [Effect of stimulus − Effect of stimulus after dexamethasone treatment]/Effect of stimulus, where stimulus was either LPS or heat-killed P. aeruginosa. Any p value below 0.05 was considered statistically significant.

Pre-treatment of C57BL/6 mice with dexamethasone significantly prolonged animal survival (Figure 1A) compared to P. aeruginosa-infected mice treated with water (controls). However, pre-treatment with hydrocortisone did not affect survival (data not shown). Compared to control mice that were infected with P. aeruginosa but not treated with corticosteroids, the percent of circulating neutrophils expressing TREM-1 was decreased in mice pre-treated with dexamethasone [30% mean reduction, Figure 1B (p = 0.044)], and the MFI of TREM-1 on TREM-1-positive cells was also decreased (mean 14.2% reduction, p = 0.035). However, pre-treatment of mice with any concentration of hydrocortisone did not significantly affect TREM-1 expression (Figures 1B,C) compared to controls. Administration of dexamethasone decreased serum TNFα (94% mean reduction, p = 0.005) but did not affect liver bacterial load (Figures 1D,E).

The human monocytic U937 cell line was utilized to determine the effects of dexamethasone on TREM-1 expression. Cells were pre-incubated without (control) or with dexamethasone or hydrocortisone prior to stimulation with LPS or heat-killed P. aeruginosa. Compared to control cells, pre-treatment of U937 cells with dexamethasone prior to LPS stimulation decreased the percent of TREM-1-positive cells by 5.2% (p = 0.014, Figure 2A), and dexamethasone pre-treatment reduced the percent of TREM-1-positive cells stimulated with in heat-killed P. aeruginosa by 8.2% (not significant, Figure 2D). A similar reduction in TREM-1 MFI was observed in LPS-stimulated cells compared to control cells (Figure 2B, 41.3% mean reduction, p = 0.005). There were no significant changes in hydrocortisone-treated cells compared to controls (Figure 2E).

sTREM-1 levels decreased by 51.7% in dexamethasone-treated cells that were stimulated with LPS compared to control cells (Figure 2C, p = 0.011). Similarly, sTREM-1 levels decreased by 55.5% in dexamethasone-treated cells that were stimulated with P. aeruginosa (Figure 2F, p = 0.041). No effect of hydrocortisone was observed on sTREM-1 levels compared to control cells.

Since dexamethasone inhibits release of TNFα from whole blood of patients with sepsis (7), we investigated the role of TNFα on the observed effect of dexamethasone inhibition of TREM-1/sTREM-1 expression. Addition of TNFα to U937 cells increased the percent of TREM-1-positive cells stimulated with LPS compared to LPS alone (compare LPS bars in Figures 2A and 3A). Addition of dexamethasone completely suppressed LPS-stimulated production of TREM-1, similar to levels observed for untreated cell controls (Figure 3A, p < 0.0001). Similar dexamethasone suppressive effects were observed for TREM-1 MFI (Figure 3B, p = 0.001), and sTREM-1 levels (Figure 3C, p = 0.029) compared to the LPS-only controls.

To separately confirm the affects of TNFα on TREM-1 expression, U937 cells were incubated with an anti-TNFα antibody. Blockade of TNFα activity by the anti-TNFα antibody completely abolished the dexamethasone-dependent suppression of TREM-1 and sTREM-1 (Figures 3D–F).

In order to further understand the significance of TNFα for the suppressive effect of dexamethasone on TREM-1 expression, animal experiments were conducted in mice deficient for TNFα. TNF^−/− mice were pre-treated with either saline (control) or dexamethasone prior to infection with P. aeruginosa. As shown in Figure 4A, pre-treatment with dexamethasone did not offer any significant survival benefit. Expression of TREM-1 on circulating neutrophils was not significantly decreased in the dexamethasone-treated mice compared to the control mice (Figures 4B,C).

To examine if the dexamethasone mechanism of action is mediated through an affect on gene expression, U937 monocytes were stimulated with heat-killed P. aeruginosa. Treatment of cells with dexamethasone significantly decreased the number of TREM-1 copies (Figure 5) compared to stimulated cells (92.5% mean reduction, p = 0.021).