Portal de Periódicos da CAPES

Inflammasome activation by Pseudomonas aeruginosa ' s ExlA pore‐forming toxin is detrimental for the host

2020; Wiley; Volume: 22; Issue: 11 Linguagem: Inglês

10.1111/cmi.13251

1462-5822

Stéphanie Bouillot, Stéphane Pont, Benoît Gallet, Christine Moriscot, Vincent Deruelle, Ina Attrée, Philippe Huber,

Clostridium difficile and Clostridium perfringens research

Cellular MicrobiologyVolume 22, Issue 11 e13251 RESEARCH ARTICLEFree Access Inflammasome activation by Pseudomonas aeruginosa's ExlA pore-forming toxin is detrimental for the host Stéphanie Bouillot, Stéphanie Bouillot Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorStéphane Pont, Stéphane Pont Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorBenoit Gallet, Benoit Gallet Institut de Biologie Structurale, CEA, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorChristine Moriscot, Christine Moriscot Institut de Biologie Structurale, CEA, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorVincent Deruelle, Vincent Deruelle Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorIna Attrée, Ina Attrée Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorPhilippe Huber, Corresponding Author Philippe Huber phuber@cea.fr orcid.org/0000-0002-4153-7694 Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, France Correspondence Philippe Huber, IRIG-BCI, CEA-Grenoble, 17 rue des Martyrs, 38054 Grenoble, France. Email: phuber@cea.frSearch for more papers by this author Stéphanie Bouillot, Stéphanie Bouillot Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorStéphane Pont, Stéphane Pont Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorBenoit Gallet, Benoit Gallet Institut de Biologie Structurale, CEA, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorChristine Moriscot, Christine Moriscot Institut de Biologie Structurale, CEA, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorVincent Deruelle, Vincent Deruelle Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorIna Attrée, Ina Attrée Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, FranceSearch for more papers by this authorPhilippe Huber, Corresponding Author Philippe Huber phuber@cea.fr orcid.org/0000-0002-4153-7694 Unité de Biologie Cellulaire et Infection, CEA, INSERM, CNRS, Université Grenoble-Alpes, Grenoble, France Correspondence Philippe Huber, IRIG-BCI, CEA-Grenoble, 17 rue des Martyrs, 38054 Grenoble, France. Email: phuber@cea.frSearch for more papers by this author First published: 11 August 2020 https://doi.org/10.1111/cmi.13251Citations: 1 Funding information: Agence Nationale de la Recherche, Grant/Award Numbers: ANR-10-INSB-05-02, ANR-15-CE11-0018-01, ANR-17-EURE-0003; Fondation pour la Recherche Médicale, Grant/Award Number: DEQ20170336705 AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract During acute Pseudomonas aeruginosa infection, the inflammatory response is essential for bacterial clearance. Neutrophil recruitment can be initiated following the assembly of an inflammasome within sentinel macrophages, leading to activation of caspase-1, which in turn triggers macrophage pyroptosis and IL-1β/IL-18 maturation. Inflammasome formation can be induced by a number of bacterial determinants, including Type III secretion systems (T3SSs) or pore-forming toxins, or, alternatively, by lipopolysaccharide (LPS) via caspase-11 activation. Surprisingly, previous studies indicated that a T3SS-induced inflammasome increased pathogenicity in mouse models of P. aeruginosa infection. Here, we investigated the immune reaction of mice infected with a T3SS-negative P. aeruginosa strain (IHMA879472). Virulence of this strain relies on ExlA, a secreted pore-forming toxin. IHMA879472 promoted massive neutrophil infiltration in infected lungs, owing to efficient priming of toll-like receptors, and thus enhanced the expression of inflammatory proteins including pro-IL-1β and TNF-α. However, mature-IL-1β and IL-18 were undetectable in wild-type mice, suggesting that ExlA failed to effectively activate caspase-1. Nevertheless, caspase-1/11 deficiency improved survival following infection with IHMA879472, as previously described for T3SS+ bacteria. We conclude that the detrimental effect associated with the ExlA-induced inflammasome is probably not due to hyperinflammation, rather it stems from another inflammasome-dependent process. 1 INTRODUCTION Pseudomonas aeruginosa is an opportunistic pathogen which exploits several virulence factors to achieve infection of various organs (Gellatly & Hancock, 2013; Hauser, 2009; Williams, Dehnbostel, & Blackwell, 2010). One of its most potent virulence factors is its type 3 secretion system (T3SS) through which it delivers several effectors (ExoS, ExoT, ExoY and ExoU) into the cytoplasm of target cells. Most strains also display a flagellum, type 4 pili, and LPS at their surface, collectively called pathogen-associated molecular patterns (PAMPs) These PAMPs can trigger the innate immune system (Hayward, Mathur, Ngo, & Man, 2018) through binding to toll-like receptors (TLRs) located at the surface of sentinel macrophages or dendritic cells. TLR triggering induces the synthesis of a large number of proteins, including inflammatory cytokines such as the IL-1β precursor (pro-IL-1β), and components of the inflammasome complex (see below). In addition, the T3SS machinery per se, by forming a pore in the macrophage plasma membrane, can activate an intracellular sensor of innate immunity, the NOD-like receptor (NLR) family CARD domain containing 4 (NLRC4) (Hayward et al., 2018; Man & Kanneganti, 2016). The NLRC4-associated inflammasome is a multiprotein complex which activates IL-1β and IL-18 by inducing their cleavage by caspase-1, and their release into the extracellular milieu following plasma membrane rupture, as a result of a cell death process called pyroptosis. Noncanonical inflammasome activation has also been described for Gram-negative bacteria. This process involves caspase-11, which can directly trigger pyroptosis and indirectly induce IL-1β/IL-18 cleavage through activation of the NLR family pyrin domain containing 3 (NLRP3)-associated inflammasome. The resulting IL-1β (and to a lesser extent IL-18) secretion contributes significantly to enhancing the inflammatory response during its initial phases. The immune reaction is subsequently amplified by several other pro-inflammatory cytokines and chemokines secreted by immune and non-immune cells, a process referred to as a ‘cytokine storm’, which leads to recruitment of circulating neutrophils to the infected organ (Garlanda, Dinarello, & Mantovani, 2013; Potey, Rossi, Lucas, & Dorward, 2019). In general, infection with P. aeruginosa strains leads to a massive recruitment of neutrophils. In P. aeruginosa-induced pneumonia, a high number of neutrophils penetrates the alveoli and accumulates in the bronchi, where they actively combat invading bacteria. However, P. aeruginosa is generally well adapted to hyperinflamed environments and can thrive in this bacteriocidic milieu by adapting its metabolism. The mechanisms deployed include producing a protective biofilm and injecting T3SS effectors into neutrophils. These effectors can block phagocytosis, inhibit reactive oxygen species (ROS) production, and induce apoptosis (ExoS and ExoT) (Rangel, Logan, & Hauser, 2014; Sun, Karmakar, Taylor, Rietsch, & Pearlman, 2012; Vareechon, Zmina, Karmakar, Pearlman, & Rietsch, 2017) or necrosis (ExoU) (Diaz et al., 2008). Although the inflammatory reaction remains the main weapon against P. aeruginosa in acute infections, several groups have demonstrated that it may also be deleterious for the host (Cohen & Prince, 2013; Faure et al., 2014; Schultz et al., 2002; Thakur, Barrett, Hobden, & Hazlett, 2004; Thakur, Barrett, McClellan, & Hazlett, 2004; Veliz Rodriguez et al., 2012). By using mice deficient in various inflammasome proteins, including caspase-1 and caspase-11, inhibitors of IL-1β, or genetically deficient mice for components of the IL-1β pathway, these groups showed that inactivating the inflammasome or preventing IL-1β action diminished the severity of disease in infected mice. However, contradictory data have been published in the literature regarding the role of the inflammasome in bacterial clearance, possibly owing to the genetic and phenotypic diversity of P. aeruginosa strains (Lin & Kazmierczak, 2017). In recent years, our group and others have identified phylogenic outliers of P. aeruginosa phylum, that were named PA7-like strains, in reference to the first fully sequenced isolate of the group, PA7 (Roy et al., 2010). PA7-like strains have now been isolated worldwide from patients with acute or chronic infections affecting a range of organs (Freschi et al., 2018; Reboud et al., 2016). Several features distinguish PA7-like strains from ‘classical’ P. aeruginosa strains, including their lack of the whole locus encoding the T3S machinery and genes for the T3SS effectors (Elsen et al., 2014). The pathogenicity of these strains is mainly based upon a two-partner secretion system, ExlB–ExlA, which is only found in PA7-like strains. ExlA (also called Exolysin) is a pore-forming toxin inducing necrosis in several mammalian cell types, and ExlB is the transporter located in the bacterial outer membrane that allows ExlA secretion into the extracellular medium (Basso, Ragno, et al., 2017; Reboud et al., 2016). Some of these strains are highly cytotoxic, a property that has been correlated with the amounts of ExlA secreted (Reboud et al., 2016). However, strain virulence in a mouse model of acute pneumonia was not directly linked to their toxicity towards cells nor to the amount of ExlA secreted (Reboud et al., 2016), suggesting that the immune reaction in response to the various PA7-like strains could differ. Further work from our group revealed that the pore formed in the plasma membrane by ExlA can activate the NLRP3 inflammasome in vitro—whereas classical strains activate the NLRC4 inflammasome—leading to macrophage pyroptosis and IL-1β secretion (Basso, Wallet, et al., 2017). Activation of the NLRP3 rather than the NLRC4 inflammasome may lead to different outcomes, as suggested by Iannitti et al. (2016). In this article, we studied the immune reaction induced in mice infected with an exlA+ strain that can trigger IL-1β secretion from macrophages. More specifically, we evaluated the effect of inflammasome inactivation on mouse survival and the response to infection. 2 RESULTS 2.1 Inflammasome activation in macrophages by exlA+ strains The capacity of bacteria to trigger inflammasome assembly and IL-1β release by resident macrophages has a major impact on the host response to infection (Garlanda et al., 2013; Wonnenberg et al., 2016). We therefore investigated inflammasome induction following exposure to several exlA+ strains by analysing levels of pro (p)- and mature (m)-IL-1β produced by infected macrophages. Mouse bone marrow-derived macrophages (BMDMs) were infected with 15 different exlA+ strains and the m-IL-1β content of cellular lysates and supernatants was analysed at 3.5 hr post-infection (hpi) by Western blotting (Figure 1a). Pro-IL-1β expression is induced by all exlA+ strains (except ATCC33359) at different levels, but they are all less efficient than the reference strain PAO1, an exlA− T3SS+ strain (ExoSTY+). In contrast, another exlA− T3SS+ (ExoUTY+) reference strain, PA14, was a poor inducer of p-IL-1β, indicating that T3SS+ strains also differentially stimulate IL-1β synthesis. m-IL-1β was only detectable when cells were infected with IHMA567230, IHMA879472, or BL043, however the signals were much less intense than observed following infection with PAO1. The profiles of IL-1β contents in supernatants were confirmed by ELISA (Figure 1b). It should be noted that the small discrepancies between the Western blot and the ELISA data are likely due to the fact that the ELISA technique does not distinguish between the different forms of IL-1β. FIGURE 1Open in figure viewerPowerPoint IL-1β secretion in culture supernatants from infected BMDM. BMDMs were incubated with 15 T3SS− exlA+ strains, or T3SS+ exlA− strains (PAO1 and PA14) at MOI 10, or were left NI. Culture media and cell lysates were collected at 3.5 hpi. (a) Both the lysates (above) and the supernatants (below) were assayed for IL-1β (either pro: p or mature: m) by Western blot. β-tubulin was used as loading control for the lysates. (b) IL-1β concentrations in the supernatants assayed by ELISA. This analysis was performed twice. Strains used in subsequent studies are indicated in red. BMDM, bone marrow-derived macrophage; NI, non-infected Altogether, there is no correlation between priming (induction of expression) and secretion of the mature form of IL-1β. In conclusion, secretion of the active form of IL-1β is more related to the capacity of exlA+ strains to activate the inflammasome than to induce its expression. IL-18 is another interleukin processed by the inflammasome and secreted after macrophage pyroptosis (Hayward et al., 2018; Man & Kanneganti, 2016). IL-18 was not detected by Western blot in any supernatants of BMDMs infected by exlA+ strains (data not shown), suggesting that IL-18 is not involved in the immune response to these types of strains; a result that was confirmed in vivo (see below). To combine virulence and m-IL-1β response, we selected strain IHMA879472 (hereafter IHMA) for further investigation. 2.2 Inflammasome activation is detrimental for mice infected with the IHMA strain As mentioned above, several reports indicated that in mouse pneumonia models involving T3SS+ strains (either ExoS+ or ExoU+), survival was enhanced by inhibiting inflammasome induction (Cohen & Prince, 2013; Faure et al., 2014; Schultz et al., 2002; Thakur, Barrett, Hobden, & Hazlett, 2004; Thakur, Barrett, McClellan, & Hazlett, 2004; Veliz Rodriguez et al., 2012). To examine the impact of inflammasome induction by IHMA (exlA+), we infected WT and caspase-1/caspase-11 KO (hereafter KO) mice by inhalation of a bacterial suspension. The KO mice used for this study induce neither conventional nor unconventional activation of IL-1β, and cannot trigger pyroptosis (Basso, Wallet, et al., 2017). As with T3SS+ strains, mice lacking caspase-1/caspase-11 genes survived significantly better than WT animals following infection with IHMA (Figure 2a), indicating that NLRP3-inflammasome activation is also detrimental in mice infected with an exlA+ strain. Mice (either WT or KO) infected with IHMAΔexlA (IHMAΔA) strain all survived, indicating that ExlA is the major virulence factor deployed by IHMA in vivo. FIGURE 2Open in figure viewerPowerPoint Infection of wild-type or caspase-1/caspase-11-deficient mice by IHMA or IHMAΔA. (a) Mice, either WT or caspase-1/caspase-11 deficient (KO), were infected intranasally with IHMA or IHMAΔA strains (107 bacteria), and their survival was monitored for 5 days. The number of mice per condition is indicated. Significance of differences between survival rates for WT and KO mice infected with IHMA, as well as differences in survival rates for WT mice infected with IHMA or IHMAΔexlA, were determined by the Log-Rank method, and the probability is shown on the graph. (b) Lung extracts were produced at 18 hpi and analysed by Western blotting using a caspase-1 antibody. β-actin was used as loading control. (c) Transmission electron micrographs of lung sections from IHMA- or IHMAΔA-infected mice (WT and KO) euthanized at 18 hpi. Two representative images are shown for IHMA/WT and IHMA/KO and one for IHMAΔA/WT and IHMAΔA/KO. Lack of pneumocytic coverage along alveoli is shown between arrowheads. A, alveolus; BM, basement membrane; C, capillary; E, endothelium; Eo, eosinophil; N, neutrophil; P, pneumocyte; RBC, red blood cell. (d) BALs were collected post-mortem from mice euthanized at 18 hpi (n = 4 per condition). Protein concentrations were measured and plotted; median values are shown (bars). Differences between WT and KO were not significant according to an unpaired t test. Differences between bacterial strains in WT mice were significant according to an ANOVA test; multiple comparisons were performed using a Tukey test, and the results are indicated by (*) or n.s. Differences were all significant compared to the NI condition (not shown). (e) β-actin contents in BALs was analysed by Western blot. (f) LDH activities were measured in BALs and plotted along with median values. Statistical analyses were performed as in (d). BAL, broncho-alveolar lavage; LDH, lactate dehydrogenase Caspase-1 expression in WT lungs was strongly induced by IHMA and moderately by IHMAΔA (Figure 2b). This difference suggests that more inflammatory cells are recruited in lungs infected by IHMA than by IHMAΔA. To directly assess the effects of IHMA on lung histology, infected mice were euthanized at 18 hpi and lung sections analysed by transmission electron microscopy (Figure 2c). Frequent interruptions of the pneumocytic lining of the alveoli (between arrowheads in Figure 2c) was observed in all mice (both WT and KO), confirming the cytolytic effect of IHMA (Basso, Wallet, et al., 2017). In contrast, no pneumocyte alteration was detected in IHMAΔA-infected lungs (WT or KO), confirming that pneumocyte disruption was caused by ExlA. As disruption of the pneumocyte layer can affect permeability of the alveolo-capillary barrier, and to determine whether inflammasome inactivation affects this parameter, we measured the protein content of broncho-alveolar lavage (BAL) fluids (Figure 2d). Consistent with the histological data, IHMA-infected lungs were more permeable than those infected with IHMAΔA. Inactivation of caspase-1/caspase-11 had no impact on permeability, indicating that the effect is a toxin-only effect. PAO1-infected mice showed a lower effect of this strain on barrier permeability, even though PAO1 is more virulent than IHMA in vivo (Reboud et al., 2016). Pseudomonas aeruginosa exlA+ strains can potentially induce necrosis of different cell types in the lung: epithelial cells, when ExlA is secreted, macrophages, when pyroptosis is activated, and neutrophils, when neutrophil extracellular traps are produced by a process called NETosis (Delgado-Rizo et al., 2017). To further examine cell necrosis induced in infected lungs, BAL fluids were assessed to determine the β-actin (Figure 2e) and lactate dehydrogenase (LDH) (Figure 2f) levels. These two intracellular proteins are released after plasma membrane rupture. IHMA, and to a lower extent IHMAΔA, caused a significant increase in LDH activity. The difference in LDH activity between mice infected with IHMA and IHMAΔA strains was consistent with the degree of disruption of the pneumocytic lining observed in Figure 2c. The fact that IHMAΔA also induced some necrosis (Figure 2f) indicates that bacterial factors other than ExlA can trigger cell death, possibly by promoting neutrophil NETosis (see the confirmation below on isolated neutrophils), rather than macrophage pyroptosis, that requires ExlA activation of the inflammasome. Thus, our data suggest that IHMA is capable to induce both pneumocyte and neutrophil necrosis in the lung. PAO1 also induced necrosis in the lungs, partially due to pyroptosis, as both markers were higher in BAL fluids from WT than from KO animals. In summary, taken together, our results indicate that IHMA has a major effect on pneumocyte and probably neutrophil cell death, whereas PAO1 mainly induces macrophage lysis. 2.3 Bacterial growth and dissemination Bacterial loads in lungs, liver and spleen were determined at 18 hpi to assess the role played by caspase-1/caspase-11 in the capacity of IHMA to proliferate in the lungs and to disseminate to distant organs (Figure 3). The bacterial amounts found in IHMA-infected lungs at 18 hpi were similar to the initial load (107), suggesting that IHMA did not thrive in lungs over this time period. In sharp contrast, the population of PAO1 (Figure 3) dramatically increased over the same time period. Conversely, the IHMAΔA bacterial load decreased, indicating that ExlA is required for sustained bacterial population of this organ, probably as it plays a role in IHMA's mechanisms of defence against the immune system. IHMA and PAO1 survived better in the presence of caspase-1/caspase-11 than in their absence, a feature previously observed for the T3SS+ strain PAK (Cohen & Prince, 2013). FIGURE 3Open in figure viewerPowerPoint Bacterial growth in lungs and dissemination in liver and spleen. Mice infected with PAO1, IHMA or IHMAΔA were euthanized at 18 hpi, and their lungs, liver and spleen were homogenised. Serial dilutions of the homogenates were plated on LB plates for CFU counting. Data are shown with medians (bars). Statistical differences between WT and KO mice were calculated using a Mann–Whitney test. The red arrows indicate the initial bacterial load In conclusion, IHMA appears to be less resistant than PAO1 to the immune system, probably because there is no equivalent of the T3SS effectors capable of preventing neutrophil bacteriocidic action. As for PAO1, IHMA disseminated to the liver and spleen, whereas IHMAΔA remained confined to the lungs, confirming that ExlA is key for bacterial spread in the body (Bouillot et al., 2017), like the T3SS for classical strains (Allewelt, Coleman, Grout, Priebe, & Pier, 2000; Kudoh, Wiener-Kronish, Hashimoto, Pittet, & Frank, 1994; Rangel, Diaz, Knoten, Zhang, & Hauser, 2015; Shaver & Hauser, 2004; Vance, Rietsch, & Mekalanos, 2005). 2.4 Mature-IL-1β is undetectable in IHMA-infected lungs, but IL-1α is highly produced To assess IHMA's capacity to promote IL-1β synthesis in vivo, we first used ELISA to measure the amounts of total IL-1β (pro and mature) in lung extracts (Figure 4a). For both IHMA- and IHMAΔA-infected lungs, there was a small non-significant increase in IL-1β at 8 hpi; no difference was observed between WT and KO lungs. In contrast, at this time point, PAO1 had already induced a significant IL-1β response in WT mice, but not in KO mice. This difference between mouse strains continued to 18 hpi. At this later time point, IL-1β levels were also significantly increased in lungs infected with IHMA, and to a lower extent in those infected with IHMAΔA, but again no difference between WT and KO mice was observed with these bacterial strains (Figure 4a). As only mature-IL-1β produced upon caspase-1-dependent cleavage is active (Afonina, Muller, Martin, & Beyaert, 2015), we analysed lung extracts (18 hpi) by Western blot to discriminate between the two forms of IL-1β (Figure 4b). Mature-IL-1β was only detected in PAO1-infected WT mice, whereas strong pro-IL-1β signals were detectable in all infected conditions. Taken together, we conclude that both IHMA and IHMAΔA can promote pro-IL-1β expression via TLR activation, although not as early or as strongly as PAO1. The slight difference observed between IHMA- and IHMAΔA-infected lungs in the ELISA results suggests that ExlA contributes to the induction of pro-IL-1β expression. However, no mature-IL-1β was detected in IHMA-infected lungs, suggesting that ExlA is a poor inflammasome activator, or that NLRP3 is less efficient than NLRC4 when it comes to activating caspase-1. It should be noted that electrophoretic gels had to be overloaded to detect m-IL-1β by Western blot in PAO1-infected lungs, indicating that only very small amounts of the mature form were present even for this strain, as previously reported (Galle et al., 2008). Thus, most of the pro-IL-1β synthetized remained unused, and only minute amounts of active IL-1β were detected following infection with PAO1, and any active IL-1β produced following IHMA infection was at undetectable levels. However, extremely low amounts of IL-1β is nevertheless sufficient to induce inflammatory reactions, as previously noted (Garlanda et al., 2013). FIGURE 4Open in figure viewerPowerPoint IL-1β, IL-18 and IL-1α production in infected lungs. (a) IL-1β concentration was measured by ELISA in lung extracts at 8 and 18 hpi. Results were plotted together with mean values (bars). (b) Lung extracts at 18 hpi were also analysed by Western blotting, using IL-1β antibody and β-actin as control. Two gels were used: one for pro-IL-1β/β-actin, and one for mature-IL-1β. To detect mature-IL-1β, gels were overloaded with protein, resulting in a distorted migration profile. (c) IL-18 concentration was measured by a bead-based immunoassay in lung extracts at 8 and 18 hpi. (d) IL-1α concentration was measured by ELISA in lung extracts at 8 and 18 hpi. (e) IL-1α concentration was measured by ELISA in BAL supernatants at 18 hpi. All statistical analyses (t test and ANOVA) were performed as described for Figure 2 As for infected BMDM supernatants, IL-18 was at background levels in lung extracts when mice were infected with IHMA or IHMAΔA (Figure 4c), confirming that IL-18 does not play a role in the immune response to IHMA. IL-1α, an interleukin targeting the same receptor as IL-1β (and thus triggering the same effects), but that does not require inflammasome activation, is rapidly and massively upregulated by IHMA and IHMAΔA strains, independent of the presence of caspase-1/caspase-11 as expected (Figure 4d). IL-1α was also detected in BAL supernatants from IHMA-infected mice (Figure 4e), showing that IL-1α is indeed secreted and thus can interact with the IL-1 receptor (IL-1R). Likewise, IL-1α production may compensate for the lack of substantial secretion of active IL-1β and may trigger the production of secondary inflammatory cytokines, like IL-6, IL-17a and KC. Interestingly, IL-1α contents of BAL supernatants from IHMAΔA-infected mice were significantly lower, suggesting that ExlA-dependent cell lysis was required for IL-1α release. 2.5 Induction of pro-inflammatory cytokines by IHMA Several pro-inflammatory cytokines (IL-6, IL-17a and TNF-α) and a chemokine (KC) were significantly induced in IHMA-infected lungs. Levels of all inflammatory mediators except TNF-α were increased between 8 and 18 hpi. The most striking increase was observed for IL-6, a cytokine involved in neutrophil maturation and recruitment (Figure 5). IL-6 levels were similar in PAO1- and IHMA-infected lungs, but were strikingly lower at 18 hpi in the absence of caspase-1/caspase-11. Furthermore, IHMAΔA failed to induce IL-6 expression in infected lungs at 18 hpi. Thus, ExlA is required for IL-6 expression, and both caspase-1/caspase-11-dependent and -independent pathways are involved. FIGURE 5Open in figure viewerPowerPoint Additional inflammatory cytokines and chemokine in infected lungs. Levels of IL-6, IL-17a, KC, TNF-α and IL-10 in lung extracts at 8 and 18 hpi were measured by ELISA and plotted alongside mean values. Significance of differences between WT and KO data were calculated using an unpaired t test, or Mann–Whitney test when distribution was not normal (i.e. for IL-17a at 18 hpi, TNF-α at 18 hpi, IL-10 at 18 hpi). For all cytokines, differences between bacterial strains in WT mice were significant according to a one-way ANOVA test; multiple comparisons were performed using a Tukey's test and the results are indicated with a (*) or n.s. Differences were all significant relative to NI, unless indicated with n.s Two other mediators of neutrophil recruitment—IL-17a and especially KC—were also induced by IHMA at 18 hpi, but at lower levels than with PAO1. Similar to our result for IL-6, caspase-1/caspase-11 deficiency was associated with lower IL-17a and KC levels, and their expression was negligible in IHMAΔA-infected lungs. Thus, in addition to ExlA, to enhance IL-6, IL-17a, and KC expression, the inflammasome must be activated. This conclusion is consistent with previous reports using other strains (Gasse et al., 2011; Rosales, 2018; Schultz et al., 2002; Slaats, Ten Oever, van de Veerdonk, & Netea, 2016). TNF-α is another chemo-attractant for neutrophils that also facilitates their extravasation by acting on endothelial cells. TNF-α expression was found to be induced to similar levels in all infected conditions, regardless of strain type or mouse genetic background. Taken together, our results show that IHMA-infected WT mice can mount an inflammatory cytokine response, which is mostly dependent on ExlA and partly on caspase-1/caspase-11 activation. As mentioned previously, massive IL-1α production in IHMA-infected lungs is likely to provide the alarming signal triggering pro-inflammatory cytokine synthesis in the infected lungs. Interestingly, the anti-inflammatory cytokine IL-10 was detected at 8 and 18 hpi in IHMAΔA-infected lungs, but not in IHMA-infected lungs. This result suggests that mice infected with the avirulent IHMAΔA strain were at a very early stage in the resolution phase of inflammation, a process characterised by IL-10 production. A similar increase in IL-10 levels was observed in PAO1-infected lungs at 18 hpi