Immunomodulatory Role of Endogenous Interleukin-18 in Gamma Interferon-Mediated Resolution of Replicative Legionella pneumophila Lung Infection

Abstract

The in vivo role of endogenous interleukin-18 (IL-18) in modulating gamma interferon (IFN-γ)-mediated resolution of replicative Legionella pneumophila lung infection was assessed using a murine model of Legionnaires' disease. Intratracheal inoculation of A/J mice with virulent bacteria (10⁶ L. pneumophila organisms per mouse) resulted in induction of IL-18 protein in bronchoalveolar lavage fluid (BALF) and intrapulmonary expression of IL-18 mRNA. Real-time quantitative RT-PCR analysis of infected lung tissue demonstrated that induction of IL-18 in BALF preceded induction of IL-12 and IFN-γ mRNAs in the lung. Blocking intrapulmonary IL-18 activity by administration of a monoclonal antibody (MAb) to the IL-18 receptor (anti-IL-18R MAb) prior to L. pneumophila infection inhibited induction of intrapulmonary IFN-γ production but did not significantly alter resolution of replicative L. pneumophila lung infection. In contrast, blocking endogenous IL-12 activity by administration of anti-IL-12 MAb) alone or in combination with anti-IL-18R MAb inhibited induction of intrapulmonary IFN-γ and resulted in enhanced intrapulmonary growth of the bacteria within 5 days postinfection. Taken together, these results demonstrate that IL-18 plays a key role in modulating induction of IFN-γ in the lung in response to L. pneumophila and that together with IL-12, IL-18 regulates intrapulmonary growth of the bacteria.

Legionella pneumophila, the causative agent of Legionnaires' disease, is an intracellular pathogen of host mononuclear phagocytic cells (MPCs), primarily alveolar macrophages (19, 24, 27). Resistance to primary replicative L. pneumophila lung infection is dependent on the induction of cellular immunity and is mediated in part by cytokines, including gamma interferon (IFN-γ) (8, 9). Growth of L. pneumophila within permissive MPCs requires iron. IFN-γ limits MPC iron, creating an intracellular environment that is nonpermissive for L. pneumophila replication (8, 9). IFN-γ in combination with other cytokines, including tumor necrosis factor alpha (TNF-α), facilitates elimination of L. pneumophila from infected MPCs, likely through the induction of effector molecules, including nitric oxide (7).

Interleukin-18 (IL-18) is a cytokine isolated from the livers of mice sequentially injected with heat-killed Propionibacterium acnes and lipopolysaccharide (28, 29). Originally termed IFN-γ-inducing factor because of its ability to induce IFN-γ in mice, IL-18 is now recognized to have pleotropic effects including (i) induction of proliferation of activated T cells; (ii) enhancement of the lytic activity of NK cells; (iii) induction of IFN-γ and granulocyte-macrophage colony-stimulating factor production by activated T cells, B cells, and/or NK cells; and (iv) promotion of T helper type 1 (Th1) responses (20, 22, 25, 29, 39, 40, 44). Responsiveness to IL-18 is conferred by IL-18 binding to its cognate receptor, which consists of the IL-1 receptor (IL-1R)-related protein 1 chain (IL-1Rrp1) (also known as IL-1R5) and the IL-1R accessory protein-like chain (IL-1RAcPL) (also known as IL-1R7) (4, 38, 42; R. Debets, J. C. Timans, T. Churakowa, S. Zurawski, R. de Waal-Malefyt, K. W. Moore, J. S. Abrams, A. O'Garra, J. F. Bazan, and R. A. Kastelein, unpublished data). Recent studies have demonstrated that IL-18-mediated cell activation can be prevented by inhibiting IL-18 ligand receptor interaction, by administration of anti-IL18 antibody (28) or by administration of monoclonal antibodies which recognize either the IL-1R5 chain (42) of the IL-1R7 chain (Debets et al., unpublished data) or the IL-18R.

Synergistic effects of IL-18 with other cytokines, including IL-12, have been described in vitro, including markedly increased IFN-γ production by T cells in comparison to that induced by either cytokine alone (1, 33, 37, 43). The molecular mechanism underlying the synergy between IL-18 and IL-12 may be explained in part by reciprocal modulation of cytokine receptor expression. Specifically, IL-18 has been demonstrated to upregulate IL-12R expression (42), while IL-12 has been shown to upregulate expression of the IL-18R (1, 43).

IL-18 has been shown to play a key role in innate immunity to intracellular pathogens, including Mycobacterium tuberculosis (35) and Cryptococcus neoformans (32). However, the potential role of endogenous IL-18 in the pathogenesis of replicative L. pneumophila lung infections has not been previously investigated. We have developed a model of Legionnaires' disease in A/J mice inoculated intratracheally with virulent bacteria (6). Resolution of replicative L. pneumophila lung infections in this animal model is mediated by cytokines, including IFN-γ (6, 7). In the present study, the biologic relevance and immunomodulatory role of endogenous IL-18 in IFN-γ-mediated resolution of replicative L. pneumophila lung infection were assessed using a monoclonal antibody (MAb) to the IL-1R7 chain of the IL-18R (Debets et al., unpublished data).

MATERIALS AND METHODS

Mice.

Female pathogen-free 6- to 8-week-old A/J mice (Jackson Laboratory, Bar Harbor, Maine) were used for all experiments. Animals were housed in microisolator cages and were cared for in accordance with standard guidelines.

Preparation of bacteria.

L. pneumophila serogroup 1, strain AA100, a redesignation of a primary clinical isolate from the Wadsworth Veterans Administration Hospital (Wadsworth, Calif.) was provided by Paul Edelstein. For preparation of the intratracheal inoculum, L. pneumophila was quantified on buffered charcoal-yeast extract (BCYE) agar plates that had been incubated for 48 h and resuspended in phosphate-buffered saline at 4 × 10⁷ organisms/ml (6, 10).

Infection of A/J mice with L. pneumophila.

A/J mice received intratracheal inoculations with L. pneumophila as previously described (6). Briefly, each mouse was anesthetized with ketamine (2.5 mg/mouse [intraperitoneally]) and tethered, and an incision was made through the skin of the ventral neck. The trachea was isolated, and 25 μl of the bacterial suspension (i.e., containing 10⁶ L. pneumophila organisms) followed by 10 μl of air was injected directly into the trachea with a 26-gauge needle. The skin incision was closed with a sterile wound clip.

Recovery of L. pneumophila from infected lung tissue.

At specific time points postinoculation, mice were humanely euthanatized, and the lungs were removed. Lung tissue was finely minced in sterile water (5 ml per lung) and homogenized (6). Lung homogenates were serially diluted in sterile water and cultured on BCYE agar containing polymyxin B, cefamandole, and anisomycin (BCYE-PAC; Baxter) for 72 h (6, 10). The lower limit of detection of L. pneumophila using this system was 10³ CFU per lung.

Collection of lung homogenate supernatant and BALF for cytokine analysis.

Lung homogenate supernatant was obtained by filtering lung homogenates prepared as described above through a 0.22-μm-pore-size filter (Gelman Sciences, Ann Arbor, Mich.) to remove the bacteria. Alternatively, for collection of bronchoalveolar lavage fluid (BALF), the mice were humanely euthanatized and their lungs were lavaged with 1.6 ml of phosphate-buffered saline (2). The resultant lavage fluid was subsequently filtered as described above. Filtered lung homogenates and BALF were stored at −80°C until use for cytokine analysis.

Cytokine analysis.

IL-18, IL-12, and IFN-γ protein levels in BALF and/or lung homogenates were measured by commercially available cytokine-specific murine enzyme linked immunosorbent assay (ELISA) kits (Quantikine mouse-IL-18, mouse IL-12p70, and mouse IFN-γ; R&D, Minneapolis, Minn.) according to the manufacturer's directions.

Quantitative RT-PCR.

IL-18 transcripts were quantified in L. pneumophila-infected lung tissues by competitive reverse transcription (RT)-PCR using a modification of previously described methodology (15, 21). Briefly, lungs were excised from L. pneumophila infected mice at specific time points (0 to 96 h postinfection [h.p.i.]) and flash frozen in liquid nitrogen until use. Total RNA was extracted using TriReagent (Molecular Research Center, Cincinnati, Ohio) according to the manufacturer's directions and was stored at −80°C in nuclease-free water containing 0.1 mM EDTA. Total RNA (1 μg) was reverse transcribed in the presence of oligo(dT)₁₅ using the Promega RT kit. cDNA samples were tested for integrity and amount of input RNA by RT-PCR for B-actin, which served as an endogenous control. Primer pairs specific for murine IL-18 (sense, 5′-ACT GTA CAA CCG CAG TAA TAC GC-3′; antisense, 5′-AGT GAA CAT TAC AGA TTT ATC CC-3′; PCR product, 434 bp [28]) and B-actin (sense, 5′-TGG AAT CCT GTG GCA TCC ATG AAA C-3′; antisense, 5′-TAA AAC GCA GCT CAG TAA CAG TCC G-3′; PCR product, 348 bp [3]) were designed and purchased from Research Genetics (Huntsville, Ala.).

For synthesis of the competitive IL-18 template, the 434-bp IL-18 PCR product was amplified using a forward primer which caused a 42-bp deletion at the 5′ end (5′ ACT GTA CAA CCG CAG TAA TAC GGG TGT TCG AGG ATA TGA CTG). The 392-bp PCR product (IL-18 competitive template [CT]) was analyzed by agarose gel electrophoresis and purified using the Qiagen gel extraction kit. During competitive PCR, the CT and sample cDNA compete for specific primers and are coamplified, resulting in PCR products that differ in size by 42 bp. Since we were able to demonstrate an equal PCR amplification efficiency of IL-18 cDNA and IL-18 CT by densitometric analysis of competitive PCR products (data not shown), the CT could be used for IL-18 cDNA quantification by competitive RT-PCR.

For competitive RT-PCR, the CT was diluted, resulting in the following concentrations (copies per microliter) 2 × 10⁷, 2 × 10⁶, 2 × 10⁵, 4 × 10⁴, 2 × 10⁴, and 4 × 10³. Equivalent amounts of individual cDNA reactions (prepared as described above) from similarly treated mice were combined to create pooled samples for competitive IL-18 RT-PCR. The competitive IL-18 PCR was performed in a final volume of 25 μl containing Qiagen 1× Taq PCR master mix, 50 ng of lung sample cDNA, CT (at the dilutions noted above), and 1 μM concentrations of 5′ and 3′ IL-18 primers. The PCR products derived from the CT and the sample cDNA were resolved on a 2% agarose gel. The mean density of the bands of the CT and sample IL-18 cDNA were quantified densitometrically, and the amount of IL-18 cDNA/50 ng of lung sample cDNA was determined using regression analysis. Due to their size differences, the copy number of the sample IL-18 cDNA was calculated for the cDNA/CT ratio: 434/392 = 1.1.

Quantitation of cytokine transcripts by real-time RT-PCR.

Real-time RT-PCR assays were performed to specifically quantitate mouse IL-12 and IFN-γ transcripts. Total cellular RNA was extracted as described above. Isolated RNA was incubated with 10 U of DNase I (Boehringer Mannheim) in the presence of RNasin (Promega) for 30 min at 37°C. The samples were then heat inactivated at 95°C for 10 min, chilled, and reverse transcribed with Superscript II reverse transcriptase (Gibco/BRL) with random hexamers according to the manufacturer's protocol. Equivalent amounts of individual cDNA reaction mixtures (prepared as described above) from similarly treated mice (6 to 8 mice/time point) were combined to create pooled samples for real-time RT-PCR. Primers for IL-12 and IFN-γ were obtained from Perkin-Elmer as predeveloped assay reagents (PDARs). Samples were then subjected to 40 cycles of amplification of 95°C for 15 s followed by 60°C for 1 min using an ABI Geneamp 5700 sequence detection system and SYBR green buffer according to the instructions of the manufacturer (Perkin-Elmer). PCR amplification of the housekeeping gene ubiquitin was performed for each sample to control for sample loading and to allow normalization between samples according to the instructions of the manufacturer (Perkin-Elmer). Both water and genomic DNA controls were included to ensure specificity. Each data point was examined for integrity by analysis of the amplification plot and disassociation curves. The ubiquitin-normalized data was expressed as the fold induction of gene expression in L. pneumophila-infected mice compared to that in uninfected mice.

Interventional studies.

Endogenous intrapulmonary IL-18 and IL-12 activities were blocked by pretreatment of the mice with a MAb to the IL-1R7 chain of the IL-18R (hereafter referred to as anti-IL-18R MAb) (1 mg/mouse [intraperitoneally]) and/or with anti-IL-12 MAb (1 mg/mouse [intraperitoneally]) respectively, 1 h prior to intratracheal inoculation with L. pneumophila (31; Debets et al., unpublished data). Similarly infected mice that had been administered an isotype-matched antibody immunoglobulin G2a (IgG2a) served as controls.

Statistical analysis.

The Student t test or analysis of variance was used to compare differences between treatment groups. A P value of <0.05 was considered significant.

RESULTS

Endogenous IL-18 is induced in the lung during primary replicative L. pneumophila lung infection.

Induction of intrapulmonary IL-18 mRNA during replicative L. pneumophila lung infection was assessed by competitive RT-PCR (Fig. 1). This methodology, which uses a CT as an internal standard, provides a sensitive, valid, and reliable tool for quantification of cytokine mRNA expression (15, 45). Approximately 20,000 copies of IL-18 mRNA were present in 50 ng of converted total RNA from uninfected lung tissue (Fig. 1a and b). While there was a minimal increase in IL-18 mRNA copy number in L. pneumophila-infected lung tissue at ≤8 h.p.i., IL-18 mRNA copy number was increased 2.1-fold in infected lung tissue at 24 h.p.i. (Fig. 1b). Analysis of BALF by ELISA demonstrated that IL-18 protein was significantly enhanced in BALF within 2 h.p.i. and was maximally induced at 4 h.p.i. (Fig. 1c). IL-18 protein was not induced in BALF from uninfected mice inoculated with an equivalent volume of saline, demonstrating that induction of IL-18 in response to L. pneumophila was specific.

FIG. 1.

IL-18 expression during replicative L. pneumophila lung infection. A/J mice were inoculated with L. pneumophila (10⁶ CFUs [intratracheally]). At specific time points postinoculation, mice were euthanatized and IL-18 expression was quantified in infected lung tissue and BALF by competitive RT-PCR and ELISA, respectively. (a) PCR products of the competitive IL-18 PCR were analyzed densitometrically using the inverted image of an ethidium bromide-stained 2% agarose gel. The competition was performed with the lung sample and various CT concentrations (in copies per microliter) as follows: 2 × 10⁷ [lanes A], 2 × 10⁶ [lanes B], 2 × 10⁵ [lanes C], 4 × 10⁴ [lanes D], 2 × 10⁴ [lanes E], and 4 × 10³ [lanes F]) at 0, 2, 8, and 24 h.p.i. Uninfected mouse lung contained 21,600 copies of IL-18 cDNA, infected mouse lung at 2 h.p.i. contained 22,200 copies of IL-18 cDNA, infected mouse lung at 8 h.p.i. contained 25,000 copies of IL-18 cDNA, and infected mouse lung at 24 h.p.i. contained 44,400 copies of IL-18 cDNA. STD, molecular size standard. (b) Time kinetics of IL-18 mRNA expression in L. pneumophila-infected lung tissue. (c) Demonstration of IL-18 protein in BALF by cytokine-specific ELISA. Symbols: ▾, BALF from L. pneumophila-infected mice; ×, BALF from saline-treated mice. Results represent the means ± standard errors of the means (error bars) of three to five animals per time point. ∗, P < 0.05 (considered significant).

Kinetics of endogenous IL-18, IL-12, and IFN-γ expression in lung of L. pneumophila-infected A/J mice.

We have previously demonstrated that cytokines, including IFN-γ, play a key role in the resolution of replicative L. pneumophila lung infection in A/J mice (6), thereby mimicking the immune response in human infection. Because IL-18 has been shown to act synergistically with other cytokines, including IL-12, to induce IFN-γ production by T cells and NK cells (1, 25, 30, 43), in initial studies the kinetics of induction of intrapulmonary IL-12 and IFN-γ mRNAs were evaluated by real-time RT-PCR. This methodology allows a rapid, accurate, and precise quantitation of gene transcripts (14, 17). As shown in Fig. 2a to c, IL-12p40, IL-12p35, and IFN-γ mRNAs, respectively, were enhanced in the lung within 12 h.p.i. In agreement with previously published results (5), immunoreactive IL-12 was also induced in BALF of L. pneumophila-infected mice within 12 h.p.i., with maximal induction at 48 h.p.i. (Fig. 2d). IFN-γ was similarly enhanced in BALF from infected mice within 48 h.p.i. (Fig. 2e). Neither IL-12 nor IFN-γ was induced in BALF of uninfected mice inoculated intratracheally with an equivalent volume of saline (data not shown). Taken together, these studies demonstrate that in response to intrapulmonary L. pneumophila, IL-18 protein is induced in BALF prior to induction of IL-12 and IFN-γ protein. Furthermore, induction of IL-18 in BALF (at ≤8 h [Fig. 1c]) preceded induction of IFN-γ mRNA expression in the lung of L. pneumophila-infected A/J mice (at ≥12 h [Fig. 2c]).

FIG. 2.

Kinetics of intrapulmonary IL-12 and IFN-γ induction during replicative L. pneumophila lung infection. A/J mice were inoculated with virulent L. pneumophila (10⁶ CFU/mouse). At specific time points p.i., the mice were euthanatized, and the lungs were excised. Total RNA was extracted, or the lungs were lavaged for collection of BALF. Transcript levels were quantified by real-time RT-PCR (IL-12p40 [a], IL-12p35 [b], and IFN-γ [c], while protein levels were quantified in BALF by cytokine-specific ELISA (IL-12p70 [d] and IFN-γ [e]). For mRNA quantification, PCR amplification of the housekeeping gene ubiquitin was performed for each sample to control for sample loading and to facilitate normalization between samples. The ubiquitin-normalized data was expressed as the fold induction of gene expression of L. pneumophila-infected mice compared to uninfected mice. Results represent fold induction of gene expression from six to eight mice per time point. For quantification of cytokines in BALF, results represent the means ± standard errors of the means of three to five animals per time point. ∗, P < 0.05 (considered significant).

Immunomodulatory activity of endogenous IL-18 and IL-12 on IFN-γ expression in the lung of L. pneumophila-infected A/J mice.

In subsequent experiments, the effect of inhibition of endogenous IL-18 and/or IL-12 activity on intrapulmonary levels of IFN-γ in L. pneumophila-infected mice was determined by cytokine-specific ELISA. As shown in Fig. 3, IFN-γ was significantly enhanced in lung homogenates from L. pneumophila-infected mice that were administered control antibody within 72 h.p.i. Pretreatment of mice with anti-IL-12MAb or anti-IL-18R MAb resulted in significant inhibition (93 and 62%, respectively) of pulmonary IFN-γ at 72 h.p.i., compared to similarly infected mice administered control antibody (IgG2a). Pretreatment with both anti-IL-18R MAb and anti-IL-12 MAb resulted in a 97% decrease in intrapulmonary levels of IFN-γ at 72 h.p.i. (Fig. 3). At 120 h.p.i. there was no significant difference in intrapulmonary levels of IFN-γ between the different treatment groups.

FIG. 3.

Immunomodulatory effect of endogenous IL-18 and IL-12 on intrapulmonary IFN-γ during replicative L. pneumophila lung infection. A/J mice were administered anti-IL-18R MAb, anti-IL-12 MAb, or both MAbs prior to intratracheal inoculation with L. pneumophila. At specific time points p.i., mice were euthanatized, and lungs were excised, homogenized, and filtered. Intrapulmonary levels of IFN-γ were quantified in filtered lung homogenates by ELISA. Mice were treated with control antibody (IgG2a) (solid bars), anti-IL-12 MAb (shaded bars), anti-IL-18RmAB (hatched bars), anti-IL-12 MAb and anti-IL-18R MAb (horizontally striped bars). Results represent means ± standard errors of the means for five animals per treatment group. ∗, P < 0.05 (considered significant).

Endogenous IL-18 and resolution of primary replicative L. pneumophila lung infection.

We have previously demonstrated that A/J mice receiving intratracheal inoculations with virulent L. pneumophila (10⁶ bacteria/mouse) develop replicative L. pneumophila lung infection, with logarithmic growth of the bacteria within the first 48 h.p.i. followed by IFN-γ-mediated clearance of bacteria from the lung at ≥72 h.p.i. (6). Because both IL-18 and IL-12 modulated IFN-γ production (Fig. 3), the potential role of these cytokines in resolution of primary replicative L. pneumophila lung infection was evaluated in mice that were administered anti-IL-18R MAb and/or anti-IL-12 MAb. As shown in Fig. 4, there was no significant difference in recovery of L. pneumophila from the lungs of mice administered anti-IL-18R MAb (1 mg/mouse [intraperitoneally]) when compared to similarly infected mice administered control antibody at any time point. In contrast, while there was no significant difference in recovery of L. pneumophila from the lungs of mice treated with anti-IL-12 MAb and similarly infected mice administered control antibody within the first 72 h.p.i., there was a 90-fold increase in recovery of intrapulmonary L. pneumophila in anti-IL-12 MAb-treated mice at 5 days p.i. Recovery of L. pneumophila from the lungs of mice administered both anti-IL-18R MAb and anti-IL-12 MAb was significantly greater than that of mice treated with either antibody alone at 72 h.p.i. and was similar to that of mice administered anti-IL-12 MAb alone at 120 h.p.i. Subsequent experiments were conducted to determine if coadministration of lower doses of anti-IL-12 MAb and anti-IL18R MAb would also cause prolonged infection. Mice were treated with anti-IL-12 MAb (50, 100, or 500 μg/mouse intraperitoneally) alone or in combination with anti-IL-18R MAb (1 mg/mouse [intraperitoneally]) 1 h prior to intratracheal inoculation with L. pneumophila (10⁶ CFU/mouse). L. pneumophila CFU in lung tissue were subsequently assessed at 5 days p.i. Coadministration of anti-IL-12 MAb (50, 100, or 500 μg/mouse [intraperitoneally]) and anti-IL-18R MAb resulted in a significant increase in L. pneumophila CFU in lung homogenates compared to values for similarly infected mice administered the same dose of anti-IL-12 MAb alone (data not shown). These results suggest that blocking endogenous IL-12 alone or in combination with inhibition of IL-18, but not blocking endogenous IL-18 alone, resulted in persistent replicative intrapulmonary L. pneumophila infection.

FIG. 4.

Role of endogenous IL-18 and IL-12 on resolution of replicative L. pneumophila lung infection. A/J mice were administered control antibody (IgG2a), anti-IL-18R MAb, anti-IL-12 MAb, or both MAbs prior to intratracheal inoculation with L. pneumophila (10⁶ bacteria/mouse). At specific time points thereafter, mice were euthanatized, and lungs were excised and homogenized. Growth of L. pneumophila in the lung tissue was quantified by culture of lung homogenates. Mice were treated with control antibody (IgG2a) (▾), anti-IL-18R MAb (○), anti-IL-12 MAb (×), anti-IL-18R MAb and anti-IL-12 MAb (▴). Results represent means ± standard errors of the means for five animals per treatment group. Symbols for statistical significance: ∗, significantly greater than mice treated with IgG2a (P < 0.05); t, significantly greater than mice treated with anti-IL-12 MAb alone (P < 0.05); ψ, significantly greater than mice treated with anti-IL-18R MAb alone (P < 0.05).

DISCUSSION

A role of IL-18 in P. acnes, M. tuberculosis, and C. neoformans infection has been previously reported (29, 32, 35). In this study, the role of endogenous IL-18 in innate immunity to replicative L. pneumophila lung infection was assessed in vivo, using a murine model of Legionnaires' disease in A/J mice inoculated intratracheally with virulent bacteria. IL-18 mRNA was constitutively expressed in uninfected mouse lungs and was increased in response to L. pneumophila infection at ≥24 h.p.i. In contrast, IL-18 protein was rapidly induced in BALF of infected mice (within 2 h.p.i.), reaching maximal levels by 4 h.p.i. These results suggest that regulation of intrapulmonary IL-18 protein during L. pneumophila infection may occur, at least in part, by a posttranscriptional mechanism. IL-18 is synthesized as a precursor molecule (pro-IL-18) devoid of a signal sequence and requires the IL-1β converting enzyme (ICE) (also known as caspase 1) for cleavage into a mature peptide (13, 16). Results of preliminary RT-PCR experiments on total RNA demonstrated that ICE mRNA was constitutively expressed in uninfected mouse lung and was induced (twofold) in response to L. pneumophila infection (data not shown). The potential relationship between intrapulmonary ICE activity and secretion of mature IL-18 during replicative L. pneumophila infection remains to be explored.

We have previously shown that resolution of L. pneumophila lung infection is mediated by IFN-γ (6). Because both IL-18 and IL-12 act synergistically to induce IFN-γ in vitro (30), the potential immunomodulatory role of endogenous IL-18 on intrapulmonary levels of IFN-γ during legionellosis was investigated. Inhibition of either endogenous IL-18 or IL-12 activity individually resulted in a significant reduction in intrapulmonary levels of IFN-γ compared to those in similarly infected control mice. Simultaneous inhibition of both endogenous IL-18 and IL-12 activity resulted in greater inhibition of intrapulmonary IFN-γ levels at 72 h.p.i. than that caused by inhibition of either cytokine alone. While the molecular basis for this additive inhibitory effect on IFN-γ has not been thoroughly investigated, it may be due, at least in part, to inhibition of multiple cell signaling pathways. IL-12-mediated cell activation is dependent on STAT-4 (26), while IL-18-mediated activation occurs by an IRAK-NFκB-dependent pathway (33). Therefore, inhibition of both signaling pathways, by simultaneous administration of anti-IL-18R MAb and IL-12 MAb, may facilitate greater inhibition of IFN-γ production than that caused by either MAb alone. In addition, it has recently been demonstrated that IL-12 and IL-18 reciprocally upregulate each other's receptors in vitro, leading directly to production of IFN-γ (30, 42, 43). Results of preliminary RT-PCR experiments in our laboratory demonstrated that intrapulmonary receptor expression for both IL-18 and IL-12 were enhanced during murine legionellosis (data not shown). Reciprocal regulation of IL-18R and IL-12R expression may also contribute to inhibition of IFN-γ production in mice administered anti-IL-12 MAb and/or anti-IL-18R MAb. These results demonstrate that both IL-18 and IL-12 are key immunomodulators of intrapulmonary IFN-γ during replicative L. pneumophila lung infection.

Results of our present study, demonstrating immunomodulation of intrapulmonary IFN-γ by endogenous IL-12, are in contrast to results of our previous study, which suggested that IL-12-mediated resolution of replicative L. pneumophila lung infection occurred by an IFN-γ-independent mechanism (5). While the reason for this disparity is not completely understood, it is likely due, in large part, to the use of different reagents to block IL-12 activity in the two studies. In the former study, mice were pretreated with anti-IL-12 antiserum (5), while in the present study, mice were administered anti-IL-12 MAb (31). An appropriate monoclonal antibody may be more efficient than antisera in inhibiting endogenous IL-12 activity. Furthermore, we cannot rule out the possibility that the anti-IL-12 antisera may have contained a contaminant which induced IFN-γ levels in the lung.

Results of subsequent studies demonstrated that while blocking endogenous IL-18 resulted in a >60% decrease in levels of intrapulmonary IFN-γ, the ability of the mice to resolve a replicative L. pneumophila lung infection was unimpaired. In contrast, blocking endogenous IL-12 activity or the simultaneous inhibition of both endogenous IL-18 and IL-12 activity decreased intrapulmonary levels of IFN-γ by >90% and resulted in a persistent replicative L. pneumophila infection. In agreement with previous studies demonstrating a key role of IL-12 in immunity to other intracellular pathogens (11, 12, 18, 23, 34, 36, 41), these results suggest that IL-12 is the dominant cytokine in IFN-γ-mediated resolution of replicative L. pneumophila lung infection. Whether persistent replicative L. pneumophila lung infection in mice treated with anti-IL-12 MAb is due at least in part to IL-12-induced modulation of IL-18R expression, resulting in modulation of IL-18-mediated cell activation, is currently being investigated.

While our studies have focused on elucidating immunomodulatory effects of IL-18 on IFN-γ-mediated resistance to replicative L. pneumophila lung infection, it is likely that endogenous IL-18 may also contribute to innate immunity to L. pneumophila via enhanced NK- and T-cell cytotoxicity (20). The potential role of cytotoxic T cells and/or NK cells in resistance to primary replicative L. pneumophila lung infections remains to be thoroughly explored.

In summary, using a murine model of Legionnaires' disease in A/J mice, we have demonstrated that endogenous IL-18 contributes to innate immunity to legionellosis through modulation of intrapulmonary IFN-γ. Future studies which identify the role of endogenous IL-18 in innate immunity to L. pneumophila are warranted, since an understanding of cytokine networking in the lung will facilitate development and evaluation of therapeutics for the treatment of Legionnaires' disease.