INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Legionella pneumophila causes Legionnaires' disease and Pontiac Fever (13). The initial phase of disease in humans (11) is characterized by symptoms that correspond to acute-phase cytokine mobilization (21). In BALB/c mice, infection results in an acute disease wherein the animals either survive or die during the first 60 h of infection (22, 28). Survival depends on the induction of innate immune mechanisms, including macrophage activation by gamma interferon (IFN-) (1, 17, 26, 36), protection by tumor necrosis factor alpha (TNF-) (2, 3, 25, 35, 37), and the production of interleukin-6 (IL-6) and IL-1 (21, 22, 44). Although the mobilization of these cytokines is generally protective (2, 35), they can also induce enhanced mortality if their levels in blood and tissue become excessive (22). The mortality is similar to septic shock (5, 16), and the mice can be rescued with anti-TNF- or anti-IL-6 antibodies (22). It appears, therefore, that the mobilization of acute-phase cytokines following L. pneumophila infection can be either protective or detrimental depending upon the extent of cytokine mobilization as well as other unknown factors.

L. pneumophila is a gram-negative, facultative intracellular bacterium, which primarily infects macrophages and monocytes (18). As with other intracellular pathogens, protective adaptive immunity depends on Th1 immunity and the associated cytokines, IFN- and IL-12 (19). These cytokines appear early during the course of infection and promote the development of Th1 cells (19, 31, 41). IL-4, on the other hand, is reported to be detrimental to the survival of animals, especially BALB/c mice, because of its role in induction of Th2 cells (15, 31). However, IL-4 was detected in mice within 3 h of infection with Listeria monocytogenes (6, 12, 15) and Mycobacterium tuberculosis (6, 7), and the transient IL-4 did not interfere with development of Th1 responses. More recently, IL-4 has been demonstrated to induce monocyte chemoattractant protein-1 (MCP-1) production during innate immunity to L. monocytogenes (6, 12, 20), and this induction of MCP-1 mediates the recruitment of monocytes, macrophages, and activated T cells (14). In the present study, we report that L. pneumophila infection also induces an IL-4 response along with MCP-1, IL-12, IFN-, TNF-, IL-1, and IL-6. Studies with IL-4-deficient mice suggest that IL-4 regulates the levels of TNF-, IL-1, and IL-6, independently of MCP-1.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mice. Female BALB/c and BALB/c-IL-4^tm2Nnt (29) mice, at 7 to 8 weeks of age (Jackson Laboratories, Bar Harbor, Maine), were used in these studies. They were housed and cared for in the University of South Florida Health Sciences Center animal facility, which is fully accredited by the American Association for Accreditation of Laboratory Animal Care.

Bacteria. L. pneumophila M124, a virulent serogroup 1 isolate from Tampa General Hospital (Tampa, Fla.), was grown on buffered charcoal-yeast extract agar (BCYE; Difco, Detroit, Mich.) for 48 h from a passage 3 stock maintained at 80°C. The bacteria were suspended in pyrogen-free saline, and the concentration was adjusted spectrophotometrically.

Mouse infections. For mortality studies, mice were infected intravenously in the tail vein with 1 × 10⁶ to 20 × 10⁶ L. pneumophila. Mice died within 60 h of infection, and surviving mice were monitored for an additional 2 to 3 weeks. In other experiments, infected mice were CO₂ asphyxiated at various times postinfection, and blood and spleens were harvested. Sera were processed and analyzed for cytokines. Splenocyte suspensions were prepared with a Stomacher 80 homogenizer (Tekmar, Cincinnati, Ohio) and processed for ex vivo cultures or RNA extraction.

Ex vivo cultures. Prepared splenocytes (5 × 10⁶ cells/ml) were cultured without additional stimulation in 24-well plates (Costar, Cambridge, Mass.) in RPMI 1640 medium (Sigma, St. Louis, Mo.) supplemented with 10% fetal calf serum (HyClone, Logan, Utah) and penicillin-streptomycin. Supernatants were collected either at 3 h of culture for IL-4 or at 24 h for IFN- and IL-12. Cytokine levels were measured by enzyme-linked immunosorbent assay (ELISA).

Neutralization of cytokines. BALB/c-IL-4^tm2Nnt mice were injected intravenously in the tail veins with 0.1 ml of neutralizing anti-cytokine antibodies at 15 and 24 h postinfection. The antibodies used were rat anti-IL-6 (10 µg/mouse; PharMingen, San Diego, Calif.), rabbit anti-TNF- polyclonal serum (0.1 ml/mouse; Genzyme, Cambridge, Mass.), hamster anti-IL-1 (10 µg/mouse; Genzyme), and the appropriate isotype control antibodies.

Macrophage infections. Mice were injected intraperitoneally with 3 ml of thioglycolate medium 4 days prior to macrophages being collected by peritoneal lavage. Adherence-purified macrophages (10⁶ cells/ml) were infected with L. pneumophila (10:1) for 30 min, washed, and cultured for 24 h. Alternatively, macrophages were exposed to killed bacteria (100:1) for 24 h. Recombinant IL-4 (PharMingen), at concentrations of between 100 and 5,000 pg/ml, was added to the cultures after infection or at the same time as the killed bacteria. In selected cultures, macrophages were lysed with 0.1% saponin (Sigma), and lysates were diluted and plated on BCYE agar for 72 h. CFU counts were determined by standard plate counting.

ELISAs. Cytokine levels of IFN-, IL-4, IL-6, and IL-12 p40-p70 were determined by sandwich ELISAs using antibody pairs from PharMingen according to protocols previously described (27). The antibody pairs for the IL-12 p40-p70 ELISA capture and detects p40 protein and thus detects both p70 and p40 proteins. Some serum samples were also analyzed using OptEIA Mouse IL-12 (p70) kit (PharMingen). MCP-1 was measured according to the above-described protocols using an anti-MCP-1 capture antibody (10 µg/ml), a biotinylated detection anti-MCP-1 antibody (2 µg/ml), and recombinant MCP-1 standard (starting at 25 ng/ml; PharMingen). TNF- and IL-1 assays were performed using DuoSet assay kits (Genzyme). The 3,3',5,5'-tetramethylbenzidine [TMB] Liquid Substrate System (Sigma) was added for 5 to 30 min, and the horseradish peroxidase reaction was stopped with 1 N sulfuric acid. The plates were read at 450 nm on an Emax Microplate Reader (Molecular Devices, Mento Park, Calif.). Units were calculated from a standard curve run with each plate. The low-end sensitivities for each ELISA were as follows: IFN- (200 pg/ml), IL-4 (20 pg/ml), IL-6 (200 pg/ml), IL-12p40/p70 (250 pg/ml), MCP-1 (500 pg/ml), TNF- (50 pg/ml), IL-1 (20 pg/ml), and IL-12 p70 (250 pg/ml).

RT-PCR. RNA was extracted by standard protocols with 1 ml of TriReagent (Sigma) per 2 × 10⁷ splenocytes collected from individual mice. Reverse transcription (RT) of total RNA was performed with avian myeloblastosis virus reverse transcriptase (Promega, San Diego, Calif.), and the RT product was PCR amplified as previously described (46). The IL-4 primer pairs (5'-CATCGGCATTTTGAACGAGGTCA and 5'-CTTATCGATGAATCAGGCATCG) were specifically designed for mRNA. The control gene product, -actin, was amplified using the following primer pair: 5'-ATGGATGACGATATCGCT and 5'-ATGAGGTAGTCTGTCAGGT. The PCR was performed in a Minicycler (MJ Research, Watertown, Miss.) at a 60°C annealing temperature for 30 cycles for IL-4 and for 25 cycles for -actin. The products were visualized in a 2% agarose gel with ethidium bromide.

Statistical analysis. Data were analyzed by one-way analysis of variance with Dunnett's test for comparing individuals using SigmaStat (Jandel Scientific, San Rafael, Calif.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

L. pneumophila infection induces early IL-12 and IFN- responses. To study cytokines produced early during innate immunity, mice were infected with 7 × 10⁶ L. pneumophila and at various times sera were analyzed for IL-12 p40-p70 and IFN-. As shown in Fig. 1A, IL-12 p40-p70 began to rise in the serum as early as 3 h postinfection and peaked at 10 h postinfection. Since IL-12 p40-p70 ELISA detected both p40 and p70 proteins, some serum samples were also analyzed with a p70-specific ELISA (OptEIA) for comparison. At 5 h postinfection, there was more p40 being produced because the levels detected by p40-p70 ELISA were higher than that of the p70 ELISA, e.g., 13.4 ± 2.8 versus 7.2 ± 1.4 ng/ml. However, at 8 h postinfection, the difference between the two ELISA assays was very slight (13.3 ± 1.06 versus 12.7 ± 1.8 ng/ml). Similar results were obtained with an IL-12 bioassay, which measured the induction of IFN- by IL-12 (45). At 5 h, more p40 protein was detected (8.7 ± 1 versus 3.7 ± 0.8 ng/ml), and at 12 h the levels were equivalent (8.9 ± 0.5 versus 8.0 ± 0.4 ng/ml). Thus, IL-12 detected in the serum by the IL-12 p40-p70 ELISA during the peak times appears to be primarily the bioactive p70 protein. IFN- was also measured in the serum, and the kinetics were similar to those of IL-12, with peak production occurring at 8 h postinfection (Fig. 1B). In addition to assaying sera for cytokines, splenocytes from these mice were cultured ex vivo without additional stimulation, and supernatants were collected at 24 h. The ex vivo cytokine patterns were similar to serum, although spleen production appeared to precede the cytokines in the serum, with IL-12 p40-p70 increasing within 1 h of infection (Fig. 1C) and IFN- increasing within 3 h (Fig. 1D). This suggested that the spleen was at least one source of the serum cytokine response. Cytokine levels, in all cases, returned to or near baseline levels by 24 h (Fig. 1).

View larger version (37K): [in this window] [in a new window]   FIG. 1.   L. pneumophila infection induces IL-12 and IFN- during the first 24 h after infection. Mice were infected with 7 × 106 bacteria and, at the indicated times after infection, blood and spleens were collected. (A and B) Serum cytokine levels. (C and D) Splenocyte ex vivo production. For panels C and D, the splenocytes were cultured without any additional stimulation for 24 h, and the supernatants were analyzed. Both sera and supernatants were measured by sandwich ELISAs. Each bar is the mean ± the standard error of the mean (SEM) of three to five experiments for a total of 6 to 16 mice.

Early IL-4 levels were detected in L. pneumophila-infected mice. Mice were infected as described above, and splenocytes were collected at 3, 5, and 24 h postinfection. The cells (5 × 10⁶/ml) were cultured without additional stimulation, and supernatants were harvested at 3 h for ELISA. As shown in Fig. 2A, IL-4 protein reached maximum production levels in the cells at 3 and 5 h postinfection and decreased to near baseline by 24 h. IL-12 p40-p70 and IFN- were not detected in these 3-h cultures but were detected at 24 h. We were unable to detect IL-4 in any of the serum samples. Therefore, in order to support the observation of splenic production of IL-4, we analyzed splenocytes for IL-4 mRNA by RT-PCR. Mice were infected and splenocytes were removed and analyzed at 1, 2, and 3 h postinfection. IL-4 mRNA expression was increased by 1 h after infection relative to -actin mRNA (Fig. 2B), supporting the conclusion that the spleen mobilizes IL-4 early after infection.

View larger version (29K): [in this window] [in a new window]   FIG. 2.   L. pneumophila infection induces IL-4 production. Mice were infected with 7 × 106 bacteria and, at the times indicated, spleens were collected. (A) Splenocytes were cultured without additional stimulation for 3 h, supernatants were collected, and IL-4 levels were determined by ELISA. Each bar is the mean ± the SEM of data from three experiments. (B) Splenocytes were lysed with TriReagent, RNA was extracted, and RT-PCR was performed. The bands were visualized with ethidium bromide in a 2% agarose gel. Each lane contains an upper IL-4 band and a lower -actin band. Each lane is represents an individual mouse, with N1 and N2 representing normal (uninfected) mice, and lanes 1 to 6 showing results for L. pneumophila infected mice at the indicated times postinfection. The gel is a representative experiment. *, P < 0.05.

Mice deficient in IL-4 are more susceptible to L. pneumophila infection. To examine possible functions for the induction of the early IL-4 protein, BALB/c and BALB/c-IL-4^tm2Nnt mice were infected with various concentrations of L. pneumophila, and the mortality was determined. As shown in Table 1, IL-4-deficient mice were more susceptible to L. pneumophila infection than were the wild-type mice. At a concentration at which all of the competent mice survived, 100% (six of six) of the IL-4-deficient mice died. This increased susceptibility suggested that IL-4 functioned to protect the mice. Because deaths occurred in both groups of mice at between 30 and 60 h postinfection, it was concluded that the effect of IL-4 occurred during innate immune responses.

Elevated acute-phase cytokines were detected in the IL-4-deficient mice and anti-TNF- antibodies were protective. We have previously observed that mortality in L. pneumophila-infected BALB/c mice is due to increased mobilization of acute-phase cytokines because the mice were protected by pretreatment with antibodies to TNF- or IL-6 (22). To determine if IL-4-deficient mice produced higher levels of TNF-, IL-1, and IL-6, mice were infected with 5 × 10⁶ L. pneumophila, a concentration that kills 75% of the deficient mice (see Table 1). Sera were collected at 24 h postinfection, and cytokine levels were determined by ELISAs. Compared competent mice, the BALB/c-IL-4^tm2Nnt mice had significantly higher levels of all three cytokines, especially TNF- and IL-6 (Fig. 3). Therefore, to examine the effect of antibody neutralization, BALB/c-IL-4^tm2Nnt mice were infected with 7 × 10⁶ bacteria and, at 15 and 24 h postinfection, were injected with either antibodies to IL-6, IL-1, or TNF- or to isotype controls. When the infected mice were treated with isotype controls, three of three mice succumbed to the infection. In contrast, the anti-TNF- antibodies protected three of three mice from death. Anti-IL-1 and anti-IL-6 antibodies had no effect in this experiment; however, these levels of antibodies had previously been shown to provide partial protection (22). The data suggested that IL-4 downregulated acute-phase cytokine mobilization and thus reduced the mortality associated with excessive levels of the cytokines, especially TNF-.

View larger version (25K): [in this window] [in a new window]   FIG. 3.   L. pneumophila infection induces higher serum levels of acute-phase cytokines in IL-4-deficient mice. Normal BALB/c and BALB/c-IL-4tm2Nnt mice were infected with 5 × 106 bacteria, and sera were collected at 24 h postinfection. Cytokine levels were determined by ELISA. Data are the means ± the SEMs for three experiments. *, P < 0.05.

IL-4 reduced acute-phase cytokine production in infected macrophage cultures. To examine the possible downregulation directly, thioglycolate-elicited macrophages were treated with either living or killed L. pneumophila in the presence or absence of IL-4. Supernatants from 24-h cultures were analyzed by ELISA for IL-1, TNF-, and IL-6. As shown in Fig. 4A, IL-4 at concentrations of as low as 100 pg/ml caused a significant decrease in IL-1 production. A suppressive effect on TNF- production was also observed, but at a 1,000-pg/ml concentration (Fig. 4B). Very little IL-6 was produced in response to live bacteria; however, treatment with killed bacteria increased the IL-6 production, and this response was attenuated by IL-4 treatment (Fig. 4C). The observed decreases in the cytokine levels were not due to fewer bacteria in the IL-4-treated cultures. In fact, the addition of IL-4 (5,000 pg/ml) significantly increased the number of bacteria detected in the cultures (P < 0.05; Fig. 4D). The concentration of IL-4 needed to suppress TNF- and IL-6 was fivefold greater than the concentration detected in ex vivo splenocyte cultures (see Fig. 2). However, the IL-4 concentration in these cultures was produced in a 1-ml volume by only 1/20th of the spleen. Extrapolation to an intact animal makes it more likely that the tissue levels were much higher than the 200 pg/ml seen in the ex vivo cultures.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   IL-4 treatment of L. pneumophila-stimulated macrophage cultures decreases the production of acute-phase cytokines. Thioglycolate-elicited peritoneal macrophages were infected with L. pneumophila (10:1) or stimulated with killed L. pneumophila (100:1). IL-4 (100 to 5,000 pg/ml) was added after bacterial infection or with the killed L. pneumophila. Supernatants were collected at 24 h and assayed by ELISAs for IL-1 (A), TNF- (B), and IL-6 (C). (D) In addition, the CFU of bacteria were determined on lysates of cells by the plate count method. Data are the means ± the SEMs for three experiments. *, P < 0.05.

IL-4-deficient and competent mice have elevated MCP-1 levels following infection. MCP-1 has been demonstrated to be induced by IL-4 in listeriosis (6, 12, 20). Additionally, MCP-1 has been shown to reduce lipopolysaccharide-induced mortality (47), suggesting that IL-4 could be controlling IL-1, TNF-, and IL-6 by inducing MCP-1. To examine this, BALB/c mice were infected with L. pneumophila, and sera were collected at several time points between 1 and 48 h. Elevated serum levels of MCP-1 were detected beginning 2 h postinfection and continued past 30 h, with the level peaking at ca. 8 h (Fig. 5A). To examine the effect of IL-4 on MCP-1, sera were collected from BALB/c-IL-4^tm2Nnt mice at 8 h postinfection, and the MCP-1 level was measured. The deficient mice had an equivalent level of MCP-1 compared to competent mice (Fig. 5B), indicating that IL-4 was not required for MCP-1 induction. Furthermore, since both groups of mice had equivalent levels of MCP-1 and since only the deficient mice had elevated levels of acute-phase cytokines, it appears unlikely that the chemokine was directly involved in the downregulation of the cytokines.

View larger version (19K): [in this window] [in a new window]   FIG. 5.   Serum MCP-1 levels are elevated following L. pneumophila infection in BALB/c and BALB/c-IL-4tm2Nnt mice. Mice were infected either with 7 × 106 (BALB/c) or with 5 × 106 (BALB/c-IL-4tm2Nnt) bacteria per mouse, and sera collected at the indicated times after infection. (A) Time kinetics for serum MCP-1 in BALB/c mice. (B) Serum level of MCP-1 at 8 h postinfection in BALB/c and BALB/c-IL-4tm2Nnt mice. Data are the means ± the SEMs for three experiments.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Much of what is known about the development of innate and cell-mediated (Th1) immunity to intracellular pathogens comes from studies with L. monocytogenes (34, 42). Protective immunity to L. monocytogenes involves both innate immunity, which limits the growth of the bacteria, and cell-mediated immunity, with its antigen-specific T cells, which clear the infection (8, 42). If an adequate cell-mediated immunity does not develop, the animals died within several days of infection. In contrast, L. pneumophila infection of BALB/c mice causes death within 60 h of infection, suggesting that innate responses are vital for survival of the animals. However, while Th1 responses do not appear to be important for resolution of the infection during the initial infection stage, they are crucial for survival of subsequent challenge (28). Therefore, our L. pneumophila infection model appears to be appropriate for studying the innate responses to intracellular bacteria as well as the corresponding development of Th1 responses.

In the present study, we examined cytokine production during the first 24 h after infection to define the early cytokine environment. Elevated levels of IL-12 p40-p70 and IFN- were observed in the serum within 3 to 5 h postinfection. These increases most likely promote the subsequent development of Th1 immunity that we have previously observed (28). This idea would agree with observations of others regarding intracellular pathogen infections (19, 33). An early IL-4 response was also detected. Moreover, IL-4-deficient mice were found to be more susceptible to the L. pneumophila infection, suggesting that the IL-4 had a protective role. Such a role was supported by additional findings that implied that IL-4 attenuated the mobilization of acute-phase cytokines during the early immune response. The IL-4-deficient mice had elevated levels of TNF-, IL-1, and IL-6, and IL-4 treatment of L. pneumophila-infected macrophage cultures suppressed their production. Furthermore, treating deficient mice with anti-TNF- antibodies protects the animals from L. pneumophila infection.

IL-4 has been shown to induce the production of MCP-1 in an L. monocytogenes infection model (12, 20). This did not occur following L. pneumophila infection, as evidenced by the finding that the mobilization of MCP-1 was equivalent in both IL-4-competent and IL-4-deficient mice. Moreover, the presence of MCP-1 in IL-4-deficient mice did not appear to effect the elevated levels of TNF-, IL-1, and IL-6, indicating that MCP-1 was not involved in the IL-4-mediated downregulation of these cytokines. The intracellular life cycles of L. pneumophila and L. monocytogenes are quite different. L. pneumophila, after entering the cell, avoids phagosome-lysosome fusion and the corresponding acidification and replicates within specialized endosomes, while L. monocytogenes lyses the vacuoles and survives in the cytoplasm (19, 33). Therefore, it is possible that these intracellular differences account for the observed variations in cytokine production and function. Additionally, L. pneumophila and L. monocytogenes are gram-negative and gram-positive bacteria, respectively. Different Toll-like receptors, which are involved in TNF- induction, have recently been shown to distinguish between the two types of bacteria (43). Therefore, while both are facultative intracellular bacteria, L. pneumophila and L. monocytogenes differ in their interactions with immune cells, and these differences could account for the disparities observed here.

An inverse correlation between IL-4 and TNF- production, however, has been reported in IFN--receptor-deficient mice after L. monocytogenes infection (39). As in our study, IL-4 production attenuated the TNF- response. However, unlike our study, the lowered TNF- response resulted in a more severe infection. As in our findings, several groups have observed increased susceptibility in IL-4-deficient mice to other pathogens, e.g., Toxoplasma gondii (32, 38) and Schistosoma mansoni (4). Moreover, Brunet et al. concluded that the increased susceptibility to S. mansoni was due to a failure to regulate TNF- production, leading to acute cachexia during the early stages of schistosomiasis (4). A regulatory role for IL-4 has also been proposed by Falcone et al. in an experimental allergic encephalomyelitis (EAE) model. In this study, IL-4-deficient BALB/c mice developed a more severe form of EAE due to an uncontrolled proinflammatory cytokine response, which included TNF- (10). Therefore, IL-4 has an ameliorating effect in several diseases, in addition to the one observed in our study.

It is also of interest to note that the IL-4 response did not appear to be required for Th1 development to L. pneumophila. IL-4-deficient mice developed a robust splenic, antigen-specific IFN- production at day 5 postinfection and survived a later challenge infection (data not shown). Others have also reported that IL-4-deficient mice develop Th1 responses to intracellular pathogens (23, 30, 32).

It is not clear at this time how IL-4 is regulating the production of TNF- and other cytokines. It is possible that IL-4 is functioning with the aid of other Th2 cytokines such as IL-10. However, we were unable to detect any IL-10 in our system. Additionally, IL-4 has been demonstrated by several groups to downregulate IL-6, IL-1, and TNF- protein and/or gene expression (9, 24, 40). Thus, IL-4 may be functioning to control these cytokines at a transcriptional level, and this is currently under investigation.

In summary, our data demonstrate that L. pneumophila infection of BALB/c mice rapidly induces IL-12 and IFN-, as well as an early, transient IL-4 response. Infections in IL-4-deficient mice suggest that IL-4 plays a regulatory function during the innate immune responses. In contrast to the proposed role for IL-4 in a Listeria infection, IL-4 produced during L. pneumophila infections appears to be involved in regulating acute-phase cytokines, especially TNF-.