|Home | About | Journals | Submit | Contact Us | Français|
Macrophages play a crucial role in the innate immune response against the human pathogen Streptococcus pyogenes, yet the innate immune response against the bacterium is poorly characterized. In the present study, we show that caspase-1 activation and IL-1β secretion were induced by live, but not killed S. pyogenes, and required expression of the pore-forming toxin streptolysin O. Using macrophages deficient in inflammasome components, we found that both Nlrp3 and Asc were crucial for caspase-1 activation and IL-1β secretion, but dispensable for pro-IL-1β induction, in response to S. pyogenes infection. Conversely, macrophages deficient in the essential TLR adaptors Myd88 and Trif showed normal activation of caspase-1, but impaired induction of pro-IL-1β and secretion of IL-1β. Notably, activation of caspase-1 by TLR2 and TLR4 ligands in the presence of SLO required Myd88/Trif whereas that induced by S. pyogenes was blocked by inhibition of NF-κB. Unlike activation of the Nlrp3 inflammasome by TLR ligands, the induction of caspase-1 activation by S. pyogenes did not require exogenous ATP or the P2X7R. In vivo experiments revealed that Nlrp3 was critical for the production of IL-1β but was not important for survival in a mouse model of S. pyogenes peritoneal infection. These results indicate that caspase-1 activation in response to S. pyogenes infection requires NF-κB and the virulence factor streptolysin O, but proceeds independently of P2X7R and TLR signaling.
Recognition of invading microorganisms by multicellular organisms is pivotal for the induction of a rapid and effective immune response. Initial recognition of microorganisms is mediated by pattern recognition receptors (PRRs), ancient molecules of the innate immune system that can be found in plants, invertebrates and vertebrates (1). The most extensively studied PRRs are the TLRs, which comprise transmembrane proteins that recognize conserved bacterial components such as LPS, flagellin, lipoproteins, lipoteichoic acid and unmethylated CpG DNA (2). More recently, another class of PRRs called intracellular nucleotide-binding oligomerization domain (Nod)-like receptors (NLRs) have been identified. Whereas TLRs sense bacterial products present at the outer cell surface or in endosomes, NLRs mediate cytoplasmic recognition of bacterial products (3). The two best studied NLRs, nucleotide-binding oligomerization domain containing 1 and 2 (Nod1 and Nod2) sense bacterial molecules produced during the synthesis and/or degradation of peptidoglycan and mediate the activation of the transcription factor NF-κB and MAP kinases (4).
Another group of NLRs participates in the formation of a protein complex called the inflammasome which mediates the induction of caspase-1 activation in response to microbial and endogenous stimuli. The inflammasome includes a NLR protein and an adaptor protein called apoptosis associated speck-like protein (Asc) which can link the NLR to the pro-caspase-1. Several NLR proteins can form inflammasomes including NLR family pyrin domain- containing 3 (Nlrp3) (also known as Cryopyrin or Nalp3), NLR family CARD domain- containing 4 (Nlrc4) (also known as Ipaf) and Nlrp1 (also known as Nalp1) (5). Activation of the inflammasome results in self-cleavage and activation of pro-caspase-1 into the active protease. Caspase-1, in turn, mediates the processing of several targets including pro-IL-1β into its biological active form (6–8). IL-1β plays an important role in the induction of immune responses and the development of inflammatory conditions such as fever and septic shock (9). In addition, gain of function Nlrp3 mutations are associated with several autoinflammatory syndromes that are characterized by aberrant IL-1β processing and elevated IL-1β levels (10).
The inflammasomes are activated by different stimuli (8, 11). For example, mouse Nlrp1b is a sensor of lethal toxin produced by Bacillus anthracis (12); Nlrc4 detects cytosolic flagellin in cells infected with Salmonella (13, 14), Legionella (15) and Pseudomonas (16, 17). In addition, Nlc4 is activated by cytosolic Shigella in a flagellin-independent manner (18). The Nlrp3 inflammasome is activated by a variety of microbial ligands including muramyl dipeptide, bacterial and viral RNA (19), as well as endogenous signals such as urate crystals (20). In cells exposed to TLR-ligands, the activation of the Nlrp3 inflammasome requires a second signal that includes stimulation of the purinergic receptor P2X, ligand-gated ion channel 7 (P2X7R) by ATP (21), or addition of certain pore-forming toxins (22). Recent studies showed that TLR agonists promote activation of the Nlrp3 inflammasome via a priming effect that is mediated through TLR signaling and NF-κB activation (23, 24). In human monocytes, stimulation with TLR ligands induces the release of endogenous ATP that is thought to contribute to the production of IL-1β (25, 26). In addition, in macrophages exposed to TLR ligands both the P2X7R agonist ATP (27, 28) and certain bacterial pore-forming toxins (29) potentiate IL-1β production. However, the role of TLR, the P2X7R and pore-forming toxins in the regulation of IL-1β secretion in the context of more physiological conditions including bacterial infection remains largely unknown.
The Gram-positive bacterium Streptococcus pyogenes is an important human pathogen that causes various infections ranging from mild superficial skin and respiratory tract infections to life-threatening systemic diseases (30, 31). Recent studies showed an important role of macrophages in the host defense against S. pyogenes infection (32). However, the molecular interactions between S. pyogenes and macrophages are poorly understood. In the present study, we demonstrate that the production of IL-1β in macrophages infected with S. pyogenes is dependent on both TLR and Nlrp3 signaling. We further show that TLR signaling is required for induction of pro-IL-1β whereas activation of caspase-1 is mediated by the pore-forming toxin streptolysin O (SLO) and the host Nlrp3 inflammasome but proceeds independently of TLR signaling and the P2X7R.
Mice deficient in Caspase-1, Nlrc4, Asc, Nlrp3, Myd88/Trif and P2X7R in C57BL/6J background have been previously described (13, 19, 24, 33, 34). All mice were backcrossed onto a C57BL/6 background for at least 6 times. Wild-type (WT) C57BL/6J mice were originally purchased from Jackson Laboratories (Bar Harbor, ME) and maintained in our Animal Facility. The animal studies were conducted under approved protocols by the University of Michigan Committee on Use and Care of Animals. Mice were housed in a specific pathogen-free facility.
The Streptococcus pyogenes wild-type strain 950771 and its isogenic streptolysin O deficient mutant, a generous gift of Dr. Michael A. Wessels (Boston), have been described (35, 36). Bacteria were grown at 37°C in Todd-Hewitt broth (Difco) supplemented with 5% yeast extract (Difco). For stimulation, an overnight culture of the bacteria was diluted 1:10 in fresh broth and grown to late log-phase for 4 h. After centrifugation at 3000 × g for 5 min, bacteria were washed twice in PBS and resuspended in IMDM without FCS and antibiotics.
Mouse bone marrow-derived macrophages were obtained from femurs and tibia as described (37). For stimulation, macrophages were cultured in 48-well plates (2.5 × 105/well) or 6-well plates (2 × 106/well) and infected with S. pyogenes for 3.5 h and harvested for immunoblotting. For measurement of secreted cytokines, cells were washed twice with PBS and incubated for additional 16 h in IMDM containing 10% heat-inactivated fetal bovine serum and 300 μg/ml gentamycin to limit the growth of extracellular bacteria. Unless otherwise stated, experiments for cytokine secretion were performed at a bacterial/macrophage ratio of 4/1 and experiments for the evaluation of caspase-1 activation at a bacterial/macrophage ratio of 10/1. For experiments with recombinant streptolysin O, macrophages were stimulated with 10 μg/ml streptolysin O (Sigma) and 10 mM DTT for 30–60 min in PBS without Ca++ and Mg++ in the absence or presence of ligands, then rinsed, and incubated for 5 hrs in IMDM medium supplemented with heat inactivated serum and antibiotics. Ultrapure LPS from E. coli (10 μg/ml) and Pam3CSK (bacterial lipopeptide) at 10 μg/ml were from Invivogen.
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. After centrifugation at 20000 × g for 15 min, the supernatant was mixed with 5 × fold SDS buffer and boiled for 5 min, and samples were separated by SDS-PAGE and transferred to nitrocellulose membranes. Immunoblotting was performed with antibodies against caspase-1 (a gift of P. Vandenabeele, University of Ghent, Ghent, Belgium) or anti IL-1β antibody (R&D Systems) as described (37).
Mice were injected i. p. with 5 × 105 CFU of S. pyogenes in PBS. After 20 h, serum was harvested and levels of IL-1β were determined by ELISA. Results were analyzed with Student’s t test and differences in data values were considered significant at a p value of <0.05. For the animal survival experimnets, comparisons between animal groups were performed by the Kaplan-Meier method using Prism software. Differences in data values were considered significant at a p value of less than 0.05.
Mouse cytokines were measured in culture supernatants or serum by ELISA kits from R&D Systems (Minneapolis, MN). Assays were performed in triplicate for each independent experiment. Student’s t test was used to determine statistical significance. A p value of <0.05 was considered significant.
We first tested the secretion of IL-1β in bone-marrow derived macrophages infected with S. pyogenes. Live, but not heat-inactivated, S. pyogenes induced IL-1β production (Fig. 1A). Macrophage responses to heat-killed S. pyogenes were not generally impaired because heat-killed bacteria induced higher amounts of TNF-α than live bacteria (Fig. 1B). These results indicate that live bacteria specifically trigger IL-1β secretion.
We next investigated the role of caspase-1 in S. pyogenes-induced IL-1β secretion using YVAD-cmk, a specific caspase-1 inhibitory peptide. Incubation of S. pyogenes-infected macrophages with YVAD-cmk reduced the release of IL-1β by 70–80% (Fig. 2A). To determine more conclusively the role of caspase-1 in S. pyogenes-induced IL-1β production, we infected macrophages deficient in caspase-1. IL-1β induced by S. pyogenes was undetectable in macrophages deficient in caspase-1 (Fig. 2B). In contrast, the production of IL-1α was comparable in WT and caspase-1-null macrophages (Fig. 2C). These results indicate that caspase-1 is essential for the secretion of IL-1β, but not IL-1β, in macrophages infected with S. pyogenes.
Human monocytes stimulated with microbial ligands secrete IL-1β, which is in part mediated by the release of endogenous ATP and stimulation of the P2X7R (25, 26) In contrast, human and mouse macrophages stimulated with TLR ligands or certain extracellular bacteria do not secrete IL-1β unless ATP is added exogenously (33). To investigate the role of P2X7R in the secretion of IL-1β induced by S. pyogenes, we infected macrophages deficient in the P2X7R and assessed the production of IL-1β. WT and P2X7R-deficient macrophages secreted comparable amounts of IL-1β (Fig. 3A) and TNF-α (Fig. 3C) in response to S. pyogenes infection. In contrast, stimulation of WT macrophages with TLR ligands and ATP induced the production of IL-1β which was abrogated in macrophages deficient in the P2X7R (Fig. 3B). These results indicate that IL-1β production in macrophages infected with S. pyogenes is independent of the P2X7R.
The addition of recombinant SLO to cells exposed to bacterial molecules such as muramyl dipeptide or flagellin can induce the activation of caspase-1 (21). To test the role of SLO under the more physiological context of bacterial infection, macrophages were incubated with WT bacteria and an isogenic mutant S. pyogenes strain deficient in SLO. Notably, the S. pyogenes lacking SLO did not induce detectable IL-1β secretion but elicited even higher amounts of TNF-α than the WT bacterium (Fig. 4B) Consistently, S. pyogenes induced the activation of caspase-1, as determined by the induction of the p20 subunit which is generated by autoproteolytic processing of pro-caspase-1 into the active enzyme (Fig. 4C). In contrast, the mutant S. pyogenes strain deficient in SLO did not activate caspase-1 (Fig. 4C). Together, these results indicate that infection of macrophages with S. pyogenes induces caspase-1-dependent release of IL-1β, a process that requires the expression of SLO.
We assessed next the role of the different inflammasomes in S. pyogenes-induced caspase-1 activation. To this end, we infected WT or macrophages deficient in Nlrp3, Nlrc4 and Asc and analyzed the activation of caspase-1. Infection of WT and Nlrc4-deficient macrophages, but not macrophages deficient in Nlrp3 or Asc, with WT bacterium induced caspase-1 activation as revealed by the detection of the p20 subunit of caspase-1 (Fig. 5A). In agreement with these results, the secretion of IL-1β in S. pyogenes-infected macrophages was abrogated in macrophages deficient in Nlrp3 or Asc, but not Nlrc4 (Fig. 5B). In contrast, Nlrp3 and Asc as well as Nlrc4 were dispensable for the production of TNF-α in response to S. pyogenes (Fig. 5C). To further assess the role of the Nlrp3 inflammasome in S. pyogenes infection, we investigated the production of IL-1β in mice infected i. p. with S. pyogenes. Infection of WT mice with S. pyogenes induced the production of IL-1β in serum which was impaired in mice deficient in Nlrp3 (Fig. 5D). However, we found that WT and Nlrp3-deficient mice were equally susceptible to S. pyogenes infection as assessed by i. p. administration of 102 and 104 CFU (supplemental Fig. 1). These results indicate that the Nlrp3 inflammasome is critical for caspase-1 activation and IL-1β secretion in response to S. pyogenes but does not play an important role in the susceptibility to infection.
Induction of pro-IL-1β is required for IL-1β secretion in response to inflammatory stimuli. We examined next whether TLR signaling was involved in the production of pro-IL-1β in macrophages infected with S. pyogenes. We found that the induction of pro-IL-1β in response to S. pyogenes infection was abrogated in macrophages deficient in Myd88 and Trif, the adaptor molecules required for TLR signaling whereas Nlrp3 was dispensable (Fig. 6A). Consistently, secretion of IL-1β and TNF-α induced by S. pyogenes was impaired in Myd88/Trif-deficient macrophages (Fig. 6B and 6C). These results indicate that TLR signaling, but not Nlrp3, is essential for pro-IL-1β induction in response to S. pyogenes.
Stimulation of macrophages with ATP and TLR agonists including LPS (TLR4 agonist) and synthetic lipopeptide (TLR2 agonist) induces caspase-1 activation via the Nlrp3 inflammasome (23, 24). Furthermore, caspase-1 activation via Nlrp3 in response to TLR agonists and ATP requires a priming effect mediated through TLR signaling and NF-κB activation (23, 24). Consistently, activation of caspase-1 induced by LPS or synthetic lipopeptide in the presence of SLO was impaired in macrophages deficient in Myd88 and Trif (Fig. 7A). Notably, the activation of caspase-1 induced by infection with S. pyogenes was independent of Myd88 and Trif (Fig. 7B). However, caspase-1 activation triggered by infection with S. pyogenes was abrogated by pre-treatment of macrophages with BAY 11–7082 (Fig. 7C), a drug that inhibits NF-κB activation by targeting the Iκ-B kinase complex (38). These results suggest that TLR signaling is dispensable for S. pyogenes-mediated caspase-1 activation but this process relies on NF-κB activation.
S. pyogenes is a highly pathogenic bacterium but its interaction with the innate immune system remains poorly characterized. In this study, we show that infection with S. pyogenes induces the secretion of IL-1β and this response is mediated by the coordinated interaction between TLR signaling and the Nlrp3 inflammasome. In addition, the results indicate that active production of the pore-forming toxin SLO is required for caspase-1 and IL-1β secretion induced by S. pyogenes as heat-inactivated bacteria and bacteria lacking SLO were impaired in stimulating the Nlrp3inflammasome.
S. pyogenes is known to secrete several virulence factors among them the cytotoxin SLO, a member of a conserved family of cholesterol-dependent pore-forming cytolysins (39). It has been demonstrated that various cytolysins such as nigericin (40) and maitotoxin (29) can induce caspase-1-dependent release of IL-1β in macrophages pre-stimulated with TLR ligands (22). In the present study, we provide evidence that SLO is critical for activation of the Nlrp3 inflammasome in response to S. pyogenes infection. Unlike TLR ligands that require exogenous ATP stimulation for activation of caspase-1 via Nlrp3, S. pyogenes triggered caspase-1 activation via Nlrp3 independently of the P2X7R. These results indicate that the Nlrp3 inflammasome can be activated via P2X7R-dependent and P2X7R-independent mechanisms in response to microbial stimuli. One possibility is that SLO acts by mimicking the function induced by ATP activation thereby bypassing the requirement for P2X7R stimulation. Because the pore-forming SLO can mediate the delivery of microbial molecules to the host cytosol and cytosolic escape of S. pyogenes (41), it is possible that infection with S. pyogenes results in cytosolic internalization of bacterial molecules or the bacterium via SLO to trigger Nlrp3-dependent caspase-1 activation. Because activation of caspase-1 induced by LPS or lipopeptide and SLO requires TLR signaling, the results suggest that MyD88/Trif-independent activation of the Nlrp3 inflammasome in response to S. pyogenes infection cannot be explained by SLO-mediated internalization of TLR ligands. However, SLO could promote caspase-1 activation by mediating internalization of microbial molecules distinct of TLR ligands.
Recent studies showed that S. pyogenes induces macrophage cell death, a process that requires SLO but the cellular demise was largely independent of caspase-1 (42). Failure to induce cell death was associated with higher survival rates of the internalized bacteria inside macrophages (42). Consistent with these results, we found that the mutant strain SLO induced higher amounts of TNF-α than WT bacteria which may be explained by the presence of higher number of mutant S. pyogenes inside macrophages. In addition, SLO can induce TLR-independent production of type-I interferons (43). Collectively, these results indicate that SLO can induce different processes in host cells including cell death, caspase-1 activation, IL-1β and interferon production which are likely to contribute to bacterial virulence and host defense.
Activation of the Nlrc4 inflammasome is dependent on the presence of a functional type III/IV secretion system, a feature of pathogenic bacteria (11). Here we show that the activation of the Nlrp3 inflammasome by S. pyogenes is similarly dependent on the expression of a virulence factor, the pore-forming toxin SLO. A common feature of these virulence factors is the formation of pores in host membranes or alteration of membrane permeability which may allow the cytosolic localization of microbial molecules. Consistent with this hypothesis, co-stimulation of macrophages with recombinant SLO and certain TLR ligands, but not each stimulus alone, triggered caspase-1 activation through the Nlrp3 inflammasome (21). We found that activation of caspase-1 with TLR ligands and SLO required TLR signaling as it was abolished in Myd88/Trif-deficient macrophages. These results are consistent with recent findings showing that TLR ligands promote activation of the Nlrp3 inflammasome, at least in part, through a priming effect mediated via TLR signalingand NF-κB activation (23, 24). Using macrophages that are deficient in the adaptor proteins Myd88 and Trif, and that therefore cannot signal via TLRs, we showed that S. pyogenes-induced caspase-1 activation is independent of TLR signaling. One possible model to explain these results is that S. pyogenes induces TLR-independent priming of the Nlrp3 inflammasome. This priming effect may be mediated via NF-κB as caspase-1 activation induced by S. pyogenes infection was blocked by treatment with a NF-κB inhibitor. In a non-excluding model, SLO may mediate internalization of bacterial molecules that are important for priming and/or activation of the Nlrp3 inflammasome independently of TLR signaling. Activation of Nlrp3 through SLO is likely to be indirect as a physical interaction between microbial molecules and NLR proteins has not yet been identified. Such an indirect mechanism has been proposed for the Nlrp3 inflammasome by TLR agonists and ATP or particulate matter such as silica or urate crystals (44). Indeed, there is evidence that reactive oxygen species and cathepsin B may contribute to the activation of the Nlrp3 inflammasome by a mechanism that remains undefined (45–48). Irrespective of the mechanism involved, the activation of inflammasome(s) is triggered by pathogenic bacteria via virulence factors such as pore-forming toxins or bacterial secretion systems. Further work is needed to understand the role of the inflammasome(s) in host defense against bacterial pathogens.
We thank Michael Wessels for generous supply of S. pyogenes strains, Shizuo Akira (Osaka University), Richard Flavell (Yale University) and Millennium Pharmaceuticals for mouse mutant strains, Joel Whitfield and Peter Kuffa for technical support and Sherry Koonse for excellent animal husbandry.
1This work was supported by grants AI063331 and AI064748. J.H. was supported by the Heisenberg-program of the Deutsche Forschungsgemeinschaft, L. F by a Fellowship from the Arthritis Foundation and T.R. by a Fellowship from the Swiss National Science Foundation
The authors declare that no competing interest exist