Toll-Like Receptor 2 Controls the Gamma Interferon Response to Francisella tularensis by Mouse Liver Lymphocytes^▿

Reagents.

Recombinant mouse IL-12 and granulocyte-Mφ colony-stimulating factor (GM-CSF) were purchased from R&D Systems (Minneapolis, MN). Recombinant mouse IL-18 was from Invitrogen (Carlsbad, CA). Neutralizing rat anti-mouse IL-12p40/p70 was obtained from BD Biosciences (San Jose, CA), and rat anti-mouse IL-18 was purchased from MBL International (Woburn, MA). Normal rat immunoglobulin G2a kappa (BD Biosciences) was used as a negative control for the antibody neutralization studies. Paired OptEIA antibody sets and cytokine standards for the detection of mouse IFN-γ, IL-12p40, and IL-12p70 were purchased from BD Biosciences. The paired antibodies for measuring IL-18 were from MBL International.

Mice.

Six-week-old female C57BL/6J (B6) mice purchased from Jackson Laboratories (Bar Harbor, ME) were used from 7 to 12 weeks of age throughout this study. Immunodeficient mutants on the B6 background (Jackson Laboratories) included the IL-12p35-deficient B6.129-IL12a^tm1Jm/J and the TLR2-deficient B6.129-Tlr2^tmiKir/J strains. All mice were maintained on a 12-h-light, 12-h-dark cycle with food and water ad libitum. Animal care and use protocols were approved by the University of Kansas Medical Center Animal Care and Use Committee.

Bacteria.

The LVS of F. tularensis subsp. holarctica (referred to here as F. tularensis LVS) was obtained from Jeannine Petersen (Centers for Disease Control and Prevention, Ft. Collins, CO) and grown in supplemented Mueller-Hinton broth as previously described (38). Frozen aliquots were stored at −70°C, thawed rapidly prior to use, washed once in Dulbecco's phosphate-buffered saline, and resuspended in antibiotic-free RPMI medium containing 2 mM l-glutamine and 10% heat-inactivated fetal bovine serum (FBS). Multiplicities of infection (MOI) were determined by serial dilution of bacterial suspensions and inoculation onto chocolate agar plates.

Bone marrow-derived Mφ and DC.

Mouse bone marrow-derived Mφ and DC were prepared by slight modifications of previously published techniques (30, 38). To prepare Mφ, bone marrow cells were flushed from the femurs of wild-type or mutant mice, counted, and grown in 75-cm² tissue culture flasks in Mφ growth medium (Dulbecco's modified Eagle's medium, 3.7 g/liter NaHC0₃, 10 mM HEPES, 7% L-cell-conditioned medium, 10% FBS, penicillin, and streptomycin). The cells were grown for 12 days in an atmosphere of 10% CO₂, with the Mφ growth medium being replaced after 7 days. Over 95% of the resulting Mφ expressed F4/80, as determined by flow cytometry. The adherent Mφ were dislodged from the flasks by treatment with 0.25% trypsin-EDTA for 5 min, resuspended in Mφ growth medium, and transferred to 96-well plates. After overnight incubation to allow attachment of the cells, Mφ growth medium was replaced with antibiotic-free complete tissue culture medium (RPMI 1640, 10 mM HEPES, 10% FBS), and the cultures were transferred to an atmosphere of 5% CO₂. Some cultures then also received HMC and F. tularensis LVS, while other Mφ cultures were challenged directly with bacteria.

For the preparation of DC, bone marrow cells were grown in DC medium (RPMI 1640, 10 mM HEPES, 10% FBS, 10 μM 2-mercaptoethanol, 20 ng/ml GM-CSF, penicillin, and streptomycin) in 60-mm-diameter bacteriological dishes in an atmosphere of 5% CO₂. After the initial 2 days of culture, an equal volume of fresh DC medium was added. On days 4, 6, and 8, most of the medium was removed from each culture and centrifuged at 300 × g for 8 min. The cell pellet was then resuspended in fresh DC medium, and the cells were returned to their original dishes. On day 10, the nonadherent DC were again removed, washed, resuspended in fresh DC medium containing 10 ng of GM-CSF per ml, and returned to their original dishes. Incubation was continued for an additional 2 days. The resulting nonadherent DC were 75 to 95% CD11c⁺ B220⁻, as determined by flow cytometry. Nonadherent DC were washed once to remove DC medium and resuspended in antibiotic-free complete tissue culture medium prior to their addition to 96-well plates on the morning of the experiment. The plates were then incubated at 37°C in an atmosphere of 5% CO₂.

Lymphocyte and accessory cell cultures.

HMC were prepared as previously described (38), by digesting mouse livers with a mixture of Liberase (Roche, Indianapolis, IN) and DNase and centrifuging the resulting cell suspension through 28% Percoll (GE Healthcare, Piscataway, NJ). The erythrocytes that passed through the Percoll were lysed with ACK lysis buffer (Cambrex, Walkersville, MD). The washed HMC preparation contained lymphocytes, Mφ, DC, and an occasional hepatocyte and was resuspended in antibiotic-free complete tissue culture medium. The cells were then plated at a concentration of 200,000 cells per well in flat-bottom 96-well tissue culture plates and incubated at 37°C in an atmosphere of 5% CO₂. To assess the ability of accessory cells to coactivate HMC for IFN-γ production, the HMC were added to wells that already contained bone marrow-derived Mφ or DC. To assess their ability to produce IL-12, cultures of Mφ or DC were challenged with bacteria in the absence of HMC.

Thawed aliquots of F. tularensis LVS were washed in phosphate-buffered saline and resuspended in antibiotic-free medium. They were then added to HMC, Mφ, or DC cultures in the absence of antibiotics. After 2 h of incubation, gentamicin (50 μg/ml) was added to the cultures and incubation was continued without removal of the bacteria for an additional 22 h. In control experiments, comparable IFN-γ responses were observed when HMC cultures were instead stimulated with gentamicin-treated F. tularensis LVS or when no antibiotic was added to cultures challenged with live bacteria.

In experiments in which neutralizing anti-cytokine antibodies were tested, these antibodies were added immediately prior to the addition of the bacteria. By contrast, coactivating cytokines were added at the end of the initial 2-h challenge period with F. tularensis LVS. Culture supernatant fluids were collected 24 h later and stored at −70°C until assayed.

Cytokine immunoassays.

Paired antibody sets and standards (BD Pharmingen) were used in enzyme-linked immunosorbent assays (ELISAs) to detect IFN-γ, IL-12p40, IL-12p70, and IL-18 in culture fluids. Absorbance was measured using a Biotek EL340 plate reader, and means and standard deviations for triplicate samples were calculated from concentrations extrapolated from standard curves by DeltaSOFT II software. The limit of detection of each ELISA was approximately 5 pg/ml. According to the manufacturer, the assay for IL-12p40 detected all cytokines containing the p40 subunit, whereas the IL-12p70 assay was specific only for the p35p40 IL-12 heterodimer. The ELISA for IL-12p70 did not show a reaction with the IL-12p40 standard from the p40 ELISA kit. Likewise, cells from p35-deficient mice produced cytokines detectable by the IL-12p40 but not the IL-12p70 ELISA.

Statistical analysis.

Each experiment was performed three times, and the most representative results are shown. Values shown are means and standard deviations. Analysis of variance (P < 0.05) was used to examine the differences between groups for all figures. For Fig. 8A, the comparison of responses in the presence of 30,000 cells was conducted by the two-sample t test.

HMC responses to F. tularensis LVS are augmented by Mφ and DC.

Mononuclear cells prepared from the livers of naïve B6 mice produced IFN-γ in vitro when challenged with F. tularensis LVS organisms, and these responses were augmented by the addition of low concentrations (<10 pg/ml) of recombinant mouse IL-12 (Fig. 1A). In a similar fashion, both bone marrow-derived Mφ and DC provided coactivating signals for the IFN-γ response to this pathogen. Significant augmentation was seen when as few as 10,000 Mφ or DC were added to cultures containing 200,000 HMC. Both accessory cell populations decreased the number of bacteria that were necessary to activate HMC for IFN-γ production and increased the magnitude of the response (Fig. 1B). These properties are typical of coactivating cytokines, such as IL-12 and IL-18 (38).

IL-12 mediates the coactivating properties of Mφ and DC.

To determine whether IL-12 or related cytokines played an essential role in the induction of IFN-γ by Mφ or DC, anti-IL-12 was added prior to challenging the cultures with F. tularensis LVS. As shown in Fig. 2, anti-IL-12 inhibited the production of IFN-γ by hepatic lymphocytes supplemented with bone marrow-derived Mφ or DC, suggesting that IL-12 mediated the coactivating functions of these two accessory cell populations. In support of this finding, DC derived from p35-deficient mice showed a marked decrease in their ability to coactivate HMC for high-level IFN-γ production (Fig. 3), whereas p35-deficient Mφ were indistinguishable from wild-type Mφ in this regard. Differences between DC and Mφ were also observed when their production of individual IL-12 subunits was measured (Fig. 4). Both unstimulated Mφ and DC constitutively produced relatively high levels of the IL-12p40 subunit, and this activity was increased following challenge with F. tularensis LVS (P < 0.05). By contrast, DC produced considerably higher levels of the IL-12p70 heterodimer than did Mφ when both cell types were challenged with bacteria. The quantities of IL-12p70 produced by Mφ varied from undetectable levels to several hundred picograms per ml but were always considerably less than the amounts produced by DC.

Because the wild-type HMC used in the experiments whose results are shown in Fig. 3 could themselves produce IL-12p70 (38), we next evaluated IL-12 as a coactivating accessory cell signal, using HMC prepared from p35-deficient mice. In this fashion, only the accessory cells (i.e., Mφ or DC) that were added to these HMC cultures could serve as a source of IL-12. When accessory cell function was evaluated in this manner, it was clear that both Mφ and DC from wild-type mice strongly coactivated p35-deficient HMC for IFN-γ production (Fig. 5). By contrast, neither Mφ nor DC from p35-deficient mice augmented the F. tularensis LVS-induced response. These observations established that the costimulatory activities of both Mφ and DC were IL-12 dependent.

TLR2 mediates the recognition of F. tularensis LVS by DC and Mφ but not the IFN-γ-producing cells.

A number of studies indicate that TLR2 is an important pattern recognition receptor mediating the recognition of F. tularensis LVS in mice (26, 28, 31). As an initial approach to determining whether TLR2 similarly controlled the liver IFN-γ response to this pathogen, HMC were prepared from wild-type, p35-deficient, and TLR2-deficient B6 mice and challenged with F. tularensis LVS in vitro. As can be seen in Fig. 6A, cells from mice deficient in either IL-12p35 or TLR2 failed to produce IFN-γ when activated with viable bacteria. The addition of exogenous recombinant IL-12 restored the IFN-γ responses of HMC from the two mutant strains and also enhanced the response by wild-type cells (Fig. 6B), as was seen in Fig. 1A. These findings suggested that TLR2-deficient HMC lack the ability to produce sufficient quantities of IL-12 or cytokines with similar functions when stimulated with F. tularensis LVS.

To test this hypothesis, both Mφ and DC were prepared from wild-type and TLR2-deficient mice and stimulated with bacteria. The production of IL-12p70, IL-12p40, and IL-18 was measured in culture fluids 24 h later. Similar to what was observed in Fig. 4, wild-type DC, but not Mφ, produced high levels of IL-12p70 (Fig. 7A). However, DC prepared from TLR2-deficient mice produced only small amounts of the cytokine, indicating that TLR2 regulates the IL-12 response of DC to F. tularensis LVS. When the same culture fluids were analyzed for IL-12p40 as a control, both cell types from each strain of mice produced the subunit and did so over a wide range of MOI. Both wild-type DC and Mφ produced IL-18 when activated with F. tularensis LVS (Fig. 7C), but TLR2 appeared to regulate this response differently in the two cell types. Whereas TLR2-deficient DC showed greatly diminished IL-18 production, the absence of TLR2 did not significantly decrease IL-18 production by Mφ.

Overall, these findings established a link between TLR2 on DC and the production of both IL-12 and IL-18 and suggested that TLR2 might regulate the IFN-γ response by this mechanism. To test this hypothesis, TLR2-deficient Mφ and DC were compared to wild-type accessory cells for their abilities to coactivate p35-deficient HMC that had been challenged with F. tularensis LVS. Whereas TLR2-deficient Mφ showed a significantly decreased ability to coactivate HMC for IFN-γ production (Fig. 8A), they were significantly more active than IL-12p35-deficient Mφ in this regard. By contrast, TLR2-deficient DC were nearly inactive and appeared functionally equivalent to DC from IL-12p35-deficient mice. The defects of both p35-deficient and TLR2-deficient DC in these cultures could be completely restored by adding exogenous recombinant IL-12 to the cultures (Fig. 8B). These findings support the conclusion that TLR2 regulates DC costimulatory activity by controlling IL-12p70 production. The data also indicate that TLR2-deficient Mφ produce IL-18 in response to F. tularensis LVS (Fig. 7C) and show costimulatory activity for the IFN-γ response (Fig. 8A).

Another potential mechanism for explaining the function of TLR2 in the IFN-γ response was considered, namely, that TLR2 mediates the recognition of F. tularensis LVS by the IFN-γ-producing cells themselves. However, TLR2-deficient HMC produced levels of IFN-γ that were comparable to those for wild-type HMC when either of these types of cells was cocultured with wild-type Mφ or DC and activated with F. tularensis LVS (Fig. 9). Thus, TLR2 appeared to control IFN-γ production by regulating IL-12 and IL-18 production rather than by controlling the ability of IFN-γ-producing cells to recognize the pathogen using TLR2.