Toll/IL-1R domain-containing adaptor inducing IFN-β
type III secretion system
Pseudomonas aeruginosa is an opportunistic Gram-negative human pathogen that is responsible for a broad range of infections in individuals with a variety of predisposing conditions. After infection, P. aeruginosa induces a marked inflammatory response in the host. However the mechanisms involved in bacterium recognition and induction of immune responses are poorly understood. Here we report that the Nod-like receptor family member Ipaf is required for optimal bacterial clearance in an in vivo model of P. aeruginosa lung infection. Further analysis showed that bacterial flagellin was essential for caspase-1 and IL-1β and this activity depended on Ipaf and the adaptor ASC but not TLR5. Notably, P. aeruginosa induced macrophage cell death and this event relied on flagellin and Ipaf but not on ASC. Analysis of Pseudomonas mutants revealed that different amino acid residues of flagellin were critical for sensing by Ipaf and TLR5. Finally, activation of caspase-1 and IL-1β secretion by P. aeruginosa required a functional type III secretion system, but not the effector molecules ExoS, ExoT and ExoY. These results provide new insight into the interaction of P. aeruginosa with host macrophages and suggest that distinct regions of flagellin are sensed by Ipaf and TLR5.
IL-1β plays an important role in the induction of immune responses and in the development of inflammatory disease, fever and septic shock 1. In response to proinflammatory stimuli including pathogenic bacteria, the IL-1β precursor is induced in monocytes and macrophages and processed into the biologically active IL-1β molecule by caspase-1 2–5. The protease caspase-1 is expressed in monocytes/macrophages as an inactive zymogen that is activated by self cleavage in large multi-protein complexes named ‘inflammasomes’ 6.
The mechanism responsible for activation of caspase-1 in response to microbial stimuli has remained poorly understood. Recent studies have revealed members of the Nod-like receptor (NLR) family as critical components of the inflammasomes by linking microbial sensing to caspase-1 activation 7, 8. For example, Ipaf, an NLR family member and the adaptor ASC have been implicated in activation of caspase-1 in response to Salmonella and Legionella through the cytosolic sensing of flagellin 7, 9, 10. Notably, flagellin is also recognized by TLR5 11. However, it is unclear whether Ipaf and TLR5 sense identical or distinct regions of flagellin. Similarly, Cryopyrin/Nalp3 is critical for caspase-1 activation and secretion of IL-1β and IL-18 in response to microbial RNA, synthetic purine-like compounds and endogenous urate crystals 12–14. In addition, Cryopyrin regulates caspase-1 activation triggered by exogenous ATP or pore-forming toxins in macrophages stimulated with several TLR agonists 15, 16.
Pseudomonas aeruginosa is a flagellated opportunistic Gram-negative human pathogen that is responsible for a broad range of infections in individuals with a variety of predisposing conditions including cystic fibrosis, immunodefiency, impaired pulmonary ventilation and loss of skin integrity 17. A component of P. aeruginosa that is critical for virulence is the type III secretion system (TTSS) that allows the bacterium to invade hosts and to overcome host defense mechanisms 18–20. P. aeruginosa uses the TTSS to directly inject effector proteins into the cytosol of the host cell 21. Four type III-secreted effectors, ExoS, ExoT, ExoY and ExoU, have been identified in P. aeruginosa. However, the expression of exoS and that of exoU appear to be mutually exclusive 22, 23. Once inside the host cell, these effector molecules promote cellular invasion by modulating host functions important in cytoskeletal organization and signal transduction 24, 25. In addition to the TTSS, other P. aeruginosa factors haven implicated in virulence including flagellin 26, 27, but the mechanism by which these factors contribute to host infection remain poorly understood.
P. aeruginosa infections are usually associated with marked inflammatory responses in host tissues 28. Immune recognition of bacterial pathogens is mediated by specific pattern recognition molecules, such as the TLR and NLR that sense microbial structures at the cell surface/endosomes and the cytosol, respectively 29, 30. Recent studies have implicated several TLR and their adaptors MyD88 and Toll/IL-1R domain-containing adaptor inducing IFN-β (TRIF) as well as Nod1 in the cytokine/chemokine response elicited upon recognition of P. aeruginosa by host cells 31–37. Caspase-1 and IL-1β are known to contribute to the inflammatory response induced by P. aeruginosa infection 38–40. However, the machinery whereby P. aeruginosa is sensed by innate immune cells to induce the activation of caspase-1 and secretion of IL-1β is unknown. In the present report, we have identified Ipaf as a critical NLR protein that is required for the activation of caspase-1 and secretion of IL-1β in response to P. aeruginosa.
P. aeruginosa induces IL-1β secretion in alveolar macrophages through Ipaf
Ipaf has been implicated in the regulation of IL-1β secretion 41, 42. Therefore, we first tested whether Ipaf is critical for IL-1β secretion induced by P. aeruginosa infection by comparing the response of WT and Ipaf-deficient macrophages isolated from mouse lungs. Exposure of alveolar macrophages to P. aeruginosa elicited production of IL-1β in WT macrophages, but this response was almost abolished in Ipaf-null macrophages (Fig. 1A).
To assess the role of Ipaf in vivo, WT and mutant mice were infected intratracheally with 5×105 colony-forming units (CFU) of P. aeruginosa and the production of IL-18 in serum was determined by ELISA. We assessed IL-18 as readout of caspase-1 activation because the levels of IL-1β induced by the bacteria in vivo were small and difficult to evaluate with reliability. At 5 h post-infection, there were reduced levels of IL-18 in Ipaf-deficient mice but not of TNF-α when compared to WT mice (Fig. 1B, C). At 6 h post-infection, there were similar numbers of P. aeruginosa CFU in the lung tissue of WT and mutant mice. By 18 h post-infection, however, there was a significant reduction of P. aeruginosa in the tissue of WT mice but not in Ipaf-deficient mice (Fig. 1D). By 48 h post-infection, P. aeruginosa were almost undetected in the lungs of both WT and Ipaf-deficient mice (Fig. 1D). Nonetheless, there was a modest but significant increase in the number of bacteria in the liver, but not spleen, of Ipaf-deficient mice when compared to WT mice 48 h after infection (Fig. 1E).
Consistently, mouse survival was not reduced in WT and Ipaf-deficient animals after intratracheal infection with 5×105 organisms (Fig. 1F). In addition, we did not observe a significant difference in mortality between mutant and WT mice when infected with a higher number of P. aeruginosa organisms (data not shown). These results indicate that Ipaf is important for IL-1β production by alveolar macrophages but it has a transient and modest role in the host response against P. aeruginosain vivo.
P. aeruginosa flagellin induces caspase-1 activation and IL-1β secretion
Given that flagellin has been implicated in IL-1β production and macrophage cell death induced by Salmonella and Legionella7, 9, we tested the ability of a P. aeruginosa mutant deficient in flagellin to induce IL-1β secretion and cell death. To ensure similar contact of WT and non-motile P. aeruginosa with macrophages, infections were followed by mild centrifugation as described 7, 9. Under these conditions, the secretion of IL-1β and cytotoxicity induced by P. aeruginosa lacking flagellin (ΔfliC) were greatly reduced when compared to that observed with WT bacteria (Fig. 2A, B). In contrast, secretion of IL-6 (Fig. 2C) and TNF-α (data not shown) was comparable after infection of macrophages with WT and ΔfliCP. aeruginosa.
To test whether expression of flagellin is important for caspase-1 activation, extracts were prepared from macrophages infected with WT and flagellin-deficient P. aeruginosa at different times post-infection and immunoblotted with an antibody that recognizes the p20 subunit of caspase-1. Infection with P. aeruginosa induced activation of caspase-1, but bacteria lacking flagellin did not (Fig. 2D). Thus, flagellin is important for caspase-1 activation, IL-1β secretion and cell death in response to P. aeruginosa.
TLR5 is not required for caspase-1 activation after P. aeruginosa infection
Flagellin is recognized by TLR5 11. We tested next whether TLR5 was required for IL-1β secretion, cell death and caspase-1 activation in response to P. aeruginosa. Both secretion of IL-1β and cell death were unimpaired in macrophages deficient in TLR5 when compared to WT macrophages (Fig. 3A, B). Similarly, the activation of caspase-1 was unaffected by the absence of TLR5 (Fig. 3C). These results indicate that the induction of caspase-1 activation and cell death is independent of TLR5 in P. aeruginosa-infected macrophages.
Ipaf and TLR5 sense different residues of flagellin
Flagellin is sensed by both Ipaf and TLR5. To determine whether the sensing of flagellin by these two host factors involves the same or different amino acid residues, we infected macrophages with P. aeruginosa mutants which express amino acid substitutions in a conserved region of flagellin predicted to interact with TLR5 43. Two flagellin P. aeruginosa mutants, L94A and Q83A, retained normal motility and the mutations in purified flagellin elicited slightly reduced and normal IL-8 production, respectively 43. Notably, both P. aeruginosa mutants, Q83A and L94A, were defective in the induction of IL-1β secretion and cell death when compared to WT bacteria (Fig. 4A, B). Consistently, the activation of caspase-1 triggered by infection with both mutants was reduced when compared to that observed with WT P. aeruginosa (Fig. 4C). These results indicate that Q83 and L94 of flagellin are critical for Ipaf-mediated caspase-1 activation, IL-1β production and cell death, but not for bacterial motility or TLR5 recognition.
P. aeruginosa activation of caspase-1 requires a functional TTSS
We determined next the role of the TTSS, a factor that is critical for virulence, in caspase-1 activation, IL-1β secretion and cell death induced by P. aeruginosa. To address this question, we infected macrophages with a P. aeruginosa mutant that lacks PscC, an essential component of the TTSS apparatus 25. These experiments showed that caspase-1 activation, IL-1β secretion and cell death were greatly reduced in macrophages infected with mutant P. aeruginosa (Fig. 5A–C). We infected next macrophages with P. aeruginosa mutants in which the genes encoding the TTSS effectors ExoS, ExoT and ExoY have been deleted individually or in combination 25. The analyses revealed that the TTSS effectors were dispensable for the induction of IL-1β secretion, cytoxicity and caspase-1 activation in P. aeruginosa-infected macrophages (Fig. 6A–F). These results indicate that a functional TTSS is critical for the induction of caspase-1 activity, IL-1β secretion and cell death, whereas the effectors ExoS, ExoT and ExoY are dispensable.
P. aeruginosa-induced caspase-1 activation requires ASC, but not Cryopyrin
We examined next the requirement for NLR proteins and the adaptor ASC in IL-1β production and caspase-1 activation triggered by P. aeruginosa. Analysis of WT and mutant macrophages lacking Ipaf, Cryopyrin or ASC revealed that P. aeruginosa-induced IL-1β secretion was greatly reduced in Ipaf- or ASC-null macrophages, but not in Cryopyrin-deficient macrophages (Fig. 7A). Consistent with the latter observations, both Ipaf and ASC, but not Cryopyrin, were required for caspase-1 activation induced by P. aeruginosa (Fig. 7B–E). In agreement with the results shown in Fig. 2, a P. aeruginosa mutant lacking flagellin (ΔfliC) was unable to induce caspase-1 activation (Fig. 7B–E). These results indicate that caspase-1 activation in response to P. aeruginosa relies on the Ipaf/ASC inflammasome.
Differential role for Ipaf and ASC in P. aeruginosa-induced macrophage cell death
We examined next the role of Ipaf, Cryopyrin and ASC in induction of cell death by P. aeruginosa. At 4 h post-infection, macrophage cell death was significantly reduced in Ipaf-deficient macrophages but not in macrophages lacking ASC or Cryopyrin when compared to WT macrophages (Fig. 8A–C). Further studies revealed similar kinetics of cell death in WT, ASC- and Cryopyrin-deficient macrophages after P. aeruginosa infection (Fig. 8D). In contrast, Ipaf-deficient macrophages were greatly protected against cell death at all times examined (Fig. 8D).
In this study, we demonstrate a critical role for Ipaf and its adaptor protein ASC in the activation of caspase-1 and IL-1β secretion in P. aeruginosa-infected macrophages. The activation of caspase-1 induced through Ipaf required expression of the bacterium flagellin but was independent of the TTSS effector molecules ExoS, ExoT and ExoY. The sensing of flagellin by Ipaf appears to involve different amino acid residues compared to those required for TLR5 recognition 43.
Remarkably, both caspase-1 activation and IL-1β secretion were abolished or greatly reduced in response to a P. aeruginosa mutant lacking an essential component of the TTSS machinery. A critical role for the TTSS apparatus in caspase-1 activation is also suggested by the observation that Salmonella requires SipB, a translocase of the TTSS, for the induction of caspase-1 and IL-1β secretion 10, 41. Similarly, the Legionella type IV secretion system has been shown to be essential for the induction of Ipaf-mediated caspase-1 in macrophages 9.
Although the precise mechanism by which the TTSS contributes to caspase-1 through Ipaf requires further investigation, a reasonable possibility is that small amounts of soluble flagellin might enter the cytosol during the assembly of the TTSS across the macrophage membrane or through the needle complex formed by the P. aeruginosa TTSS apparatus. Thus, Ipaf may sense flagellin directly or through another host factor in the cytosol to promote the activation of caspase-1. Such a mechanism has been proposed to explain the requirement of both flagellin and the TTSS for caspase-1 activation in response to Salmonella 10, 41. Similarly, peptidoglycan-derived molecules are delivered to the host cytosol by the Helicobacter pylori type IV secretion system for the activation of Nod1, another NLR family member 44.
Alternatively, the TTSS might induce an activity at the membrane of infected macrophages independent of ExoS, ExoT and ExoY that is critical cofactor for the activation of the Ipaf inflammasome. Because the TTSS forms a pore in the membrane of the contacted host cell, it may induce changes in cytosolic ion concentrations or another event across macrophage membranes that promote caspase-1 activation. However it should be noted that the Ipaf-inflammosome is not modulated by intracellular K+ concentration 42. Finally, there is the possibility that there may be undiscovered TTSS effectors that regulate caspase-1 activation. Further studies are needed to understand the contribution of the P. aeruginosa TTSS and flagellin to the activation of the Ipaf inflammasome.
We found a critical role for the TTSS, flagellin, Ipaf and caspase-1 in the induction of rapid cell death in macrophages infected with P. aeruginosa. This mode of bacteria-induced macrophage cell death that relies on caspase-1 is triggered by several bacterial pathogens including Salmonella and Shigella and refereed as pyroptosis 45. The induction of pyroptosis by P. aeruginosa proceeded normally in ASC-deficient macrophages despite the absence of caspase-1 activation. These results indicate that the function of Ipaf and ASC differ in a subtle manner and that the absence of caspase-1 activation is not sufficient to inhibit pyroptosis.
One possibility is that both caspase-1 activation and the failure to induce pro-survival signals are required for pyroptosis. In this model, ASC promotes survival signals and in the absence of ASC, but not caspase-1 or Ipaf, these ASC-mediated pro-survival signals will not be induced leading to pyroptosis. Consistent with this hypothesis, ASC mediates NF-κB activation 46–48 and thus NF-κB or another activity induced via ASC independently of caspase-1 might counter the induction of pyroptosis in P. aeruginosa-infected macrophages. An alternative possibility is that the inflammasome formed in the absence of Ipaf and ASC might vary in a subtle manner. For example, they may differ in the recruitment of host molecules that regulate the induction of pyroptosis in infected macrophages. Further studies are needed to understand the differential role of Ipaf and ASC in pyroptosis induced by bacterial infection.
Several studies have assessed the host mechanisms that mediate the immune response to P. aeruginosa. These experiments revealed a role for TLR2, TLR4 and TLR5 in the cytokine/chemokine response of epithelial cells and macrophages to the bacterium in vitro and in vivo32–36. However, while analyses of MyD88-null and TRIF-null mice have shown a critical role for this adaptor in bacterial resistance, there is no or little evidence that individual TLR are critical for susceptibility to P. aeruginosa and bacterial clearance in vivo32–36. The high susceptibility of mice deficient in the adaptor MyD88 to P. aeruginosa in contrast to individual TLR suggests redundancy of TLR in the host response to P. aeruginosa and/or involvement of IL-1R in that this pathway also uses MyD88 for signaling 49.
A role for IL-1 signaling in controlling P. aeruginosa lung infection in mice is controversial. Recent studies have suggested an important role for IL-1R using a chronic colonization model whereas no significant role was found in acute pulmonary infection with P. aeruginosa40. Similarly, some authors have reported that pre-treatment with IL-1β protected the mice against bacterial infection, but other studies showed that IL-1β-neutralizing antibody administered after P. aeruginosa protected mice from sepsis and acute pneumonia. These seemingly contradictory results are likely to reflect differences in the experimental models including the timing of administration and the dose of bacterial inocolum. In a model of corneal infection induced by P. aeruginosa, both caspase-1 and IL-1β were found to be important in eliciting acute inflammatory responses and tissue damage 39.
Our studies have revealed that the absence of Ipaf is associated with a transient defect in the clearance of P. aeruginosa in the lung tissue after intratracheal infection. This modest effect is comparable with results obtained in the Salmonella system in which no or minimal effects in bacterial clearance were observed after oral infection 50. These findings suggest redundancy between different NLR family members and other pattern recognition receptors in host defense against P. aeruginosa. Therefore, it will be important in future studies to assess the role of Ipaf in the presence and absence of other NLR and TLR in mouse models of P. aeruginosa infection.
Materials and methods
Mice and cells
Mice deficient in Ipaf, ASC, Cryopyrin, caspase-1, TLR5 have been previously described 12, 41, 51, 52. For in vivo experiments Ipaf-KO mice were backcrossed five times on a Balb/c background. Mice were housed in a pathogen-free facility.
Bone marrow-derived macrophages (BMDM) were isolated as previously described 53. Briefly, femurs and tibia were removed from the euthanized mouse and briefly sterilized in 70% ethanol. IMDM was used to wash out the marrow cavity plugs and bone marrow cells were resuspended in L cell-conditioned medium containing M-CSF to stimulate proliferation and differentiation of the marrow progenitors into macrophages. After 5–6 days, the resulting BMDM were replated and used within 2 days. Alveolar macrophages were prepared as previously described 54. The animal studies were conducted under approved protocols by the University of Michigan Committee on Use and Care of Animals.
Reagents and bacterial infection
Ultrapure Escherichia coli LPS was from Invivogen. All the bacterial strains used in this study were derived from the WT P. aeruginosa strain PAK. P. aeruginosa deletion mutants pscC, fliC, exoS, exoT, exoY, exoSTY used in this study were described elsewere. The bacteria were propagated in liquid Luria–Bertani broth or on Luria–Bertani agar plates. The bacteria were grown at 30°C overnight. The next day cultures were diluted 10–1 and grown for 4 h at 37°C to late exponential/early stationary phase before macrophage infection. Bacteria were diluted to the desired concentration in IMDM + 10% heat-inactivated FBS and used to infect macrophages at different bacterial/macrophage ratios. After 1 h, gentamycin (100 μg/mL) was added to limit the growth of extracellular bacteria. In all the experiments immediately after infection P. aeruginosa were spun onto the cells at 1500 rpm to synchronize the infection.
Cells were lysed together with the cell supernatant by the addition of 1% NP-40, complete protease inhibitor cocktail (Roche, Mannheim, Germany) and 2 mM dithiothreitol. Clarified lysates were resolved by SDS-PAGE and transferred to PVDF membranes by electro-blotting. The rabbit anti-mouse caspase-1 was a kind gift from Dr. Vandanabeele (Ghent University, Ghent, Belgium). Anti IL-1β was from R&D Systems, Minneapolis, MN.
Measurements of cytokines
Mouse cytokines were measured in culture supernatants, or serum, with ELISA kits (R&D Systems, Minneapolis, MN). Assays were performed in triplicate for each independent experiment.
Mouse infection and statistical analysis
WT and Ipa-KO mice were infected intratracheally with 5×105 Pseudomonas, and the number of bacteria in the lungs was determined at 6, 18 and 48 h post-infection by serial dilution plating. Comparisons between two experimental groups were performed with Student's t-test. Differences in data values were considered significant at a p value of less than 0.05.
We thank Anthony Coyle, Ethan Grant and John Bertin (Millennium Pharmaceuticals) for generous supply of mutant mice, S. Lory for bacterial strains, and P. Vandenabeele for anti caspase-1 antibody. This work was supported by NIH grants AI063331, AI064748 and AI064748 to G.N. L.F. is recipient of a postdoctoral fellowship from the Arthritis Foundation.Conflcit of interest: The authors declare no financial or commercial conflict of interest.