Abstract
Neutrophils mediate essential immune and microbicidal processes. Consequently, to counteract neutrophil attack, pathogens have developed various virulence strategies. Here, we showed that Pseudomonas aeruginosa (P. aeruginosa) phospholipase ExoU drives pathological NETosis in neutrophils. Surprisingly, inhibition of ExoU activity uncovered a fully functional Caspase-1-driven pyroptosis pathway in neutrophils. Mechanistically, activated NLRC4 inflammasome promoted Caspase-1-dependent Gasdermin-D activation, IL-1β cytokine release and neutrophil pyroptosis. Whereas both pyroptotic and netotic neutrophils released alarmins, only NETosis liberated the destructive DAMPs Histones, which exacerbated Pseudomonas-induced mouse lethality. To the contrary, subcortical actin allowed pyroptotic neutrophils to physically limit poisonous inflammation by keeping Histones intracellularly. Finally, mouse models of infection highlighted that both NETosis and neutrophil Caspase-1 contributed to P. aeruginosa spreading. Overall, we established the host deleterious consequences of Pseudomonas-induced-NETosis but also uncovered an unsuspected ability of neutrophils to undergo Caspase-1-dependent pyroptosis, a process where neutrophils exhibit a self-regulatory function that limit Histone release.
Graphical abstract P.aeruginosaExoU (right) triggers phospholipid degradation and subsequent neutrophil lysis that associates to NETosis through F-Actin collapse/degradation and PAD4-dependent DNA decondensation. P.aeruginosaExoU- (left) triggers NLRC4-dependent pyroptosis in neutrophils, which leads to PAD4- depndent DNA decondensation but not expulsion due to a still function subcortical F- Actin network. Created with Biorender.com.
Introduction
Over the last 30 years, non-apoptotic forms of cell death have emerged as crucial processes driving inflammation, host defense against infections but also (auto) inflammatory pathologies (Galluzzi et al., 2018).
Unique among all regulated cell necrosis is the capacity of granulocyte neutrophils to undergo the process of NETosis (Brinkmann et al., 2004). NETosis is an antimicrobial and pro-inflammatory from of cell death that promotes the formation of extracellular web-like structures called Neutrophil Extracellular Traps (NETs) (Brinkmann et al., 2004).
NETosis consists in sequential steps that start with nuclear envelope disintegration, DNA decondensation, cytosolic expansion of DNA and its subsequent expulsion through plasma membrane (Thiam et al., 2020). Completion of DNA decondensation and expulsion requires various cellular effectors. Among them, neutrophil serine proteases (Neutrophil elastase, Cathepsin G, Proteinase 3) or caspase-11 can mediate histone cleavage, which relaxes DNA tension (Chen et al., 2018; Kenny et al., 2017; Knackstedt et al., 2019; Papayannopoulos et al., 2010; Sollberger et al., 2018). In addition, granulocyte-enriched Protein arginine deaminase 4 (PAD4), citrullinates histone-bound DNA, which neutralizes arginine positive charges, thus helping nuclear DNA relaxation and decondensation (Li et al., 2010; Thiama et al., 2020; Wang et al., 2009). Then, decondensed DNA is mixed with the neutrophil cytoplasmic granule content such as NE, CathG, PR3 and Myeloperoxidase (MPO) proteins (Chen et al., 2018; Papayannopoulos et al., 2010; Thiama et al., 2020). Finally, sub-cortical actin network disassembly is required to ensure efficient DNA extrusion through the permeabilized plasma membrane (Neubert et al., 2018; Thiama et al., 2020).
Depending on the initial trigger, various signaling pathways such as calcium fluxes (Kenny et al., 2017; Thiama et al., 2020), necroptosis-associated MLKL phosphorylation (D’Cruz et al., 2018), ROS-induced Neutrophil protease release (Papayannopoulos et al., 2010) or endotoxin-activated caspase-11 (Chen et al., 2018, 2021b; Kovacs et al., 2020) all bring neutrophil into NETosis. Common to both ROS- and caspase-11-dependent NETosis is the requirement of the pyroptosis executioner Gasdermin-D (GSDMD) cleavage by both neutrophil serine and caspase-11 proteases, which triggers neutrophil NETosis (Chen et al., 2018; Sollberger et al., 2018). Specifically, active GSDMD forms a pore on PIP2-enriched domains of the plasma and nuclear membrane of neutrophils, which ensures both IL-1-related cytokine release (Evavold et al., 2018; Heilig et al., 2018; Tsuchiya et al., 2021; Xia et al., 2021) and osmotic imbalance-induced DNA decondensation and expulsion (Chen et al., 2018; Sollberger et al., 2018). An intriguing feature of neutrophils is that, despite GSDMD activation, they resist canonical inflammasome-induced Caspase-1- dependent cell pyroptosis upon Salmonella Typhimurium and Burkholderia thaïlandensis-activated NLRC4 inflammasome or upon Nigericin/ATP-mediated NLRP3 inflammasome activation (Chen et al., 2014, 2018; Karmakar et al., 2020; Kovacs et al., 2020).
Although the importance of NETosis in host immunity to infections has been well established (Brinkmann et al., 2004; Chen et al., 2018; Kovacs et al., 2020; Li et al., 2010), NETosis dysregulation also associates to autoimmunity, host tissue damages, aberrant coagulation and thrombus that all contribute to pathology such as sepsis or autoimmune lupus (Apel et al., 2021; Biron et al., 2018; Fuchs et al., 2010; Kahlenberg et al., 2013; Knackstedt et al., 2019; Kumar et al., 2015; Lefrançais et al., 2018; Martinod et al., 2015). Specifically, P. aeruginosa bacterial strains that express the necrotizing ExoU phospholipase virulence factor of the patatin-like phospholipase A2 family, promote organ damage-dependent acute respiratory distress syndrome (Bagayoko et al., 2021; Howell et al., 2013; Phillips et al., 2003; Sato et al., 2003). Despite seminal studies underlined that neutrophils constitute one of the primary targets of ExoU (Diaz and Hauser, 2010), their importance in ExoU-driven pathology and the associated molecular mechanisms involved remain elusive. Therefore, in this study, we explored the role of regulated neutrophil necrosis upon P. aeruginosa infection and addressed the molecular pathways involved.
Using primary murine and human neutrophils, we showed that ExoU phospholipase activity triggers pathological PAD4-dependent NETosis of neutrophils. In addition, the lack of ExoU expression unveiled a compensatory cell necrosis of neutrophils driven by the canonical inflammasome NLRC4. Specifically, NLRC4-driven neutrophil death led to an incomplete NETosis where neutrophil DNA is decondensed and fills the host cell cytosol but is not expulsed out from the cells. Regarding this, whereas both netotic and pyroptotic neutrophils released HMGB1/2 DAMPs, only NETosis liberated the destructive DAMPs Histones, which contributed to Pseudomonas-dependent lethality. To the contrary, pyroptotic neutrophils limited poisonous inflammation by keeping DNA-bound Histones intracellularly thanks to a still assembled subcortical actin cytoskeleton. Finally, both ExoU-induced NETosis and neutrophil caspase-1 played a pro-microbial role in mice by contributing to P. aeruginosa spreading. Overall, we established the host deleterious consequences of Pseudomonas-induced-NETosis and -pyroptosis of neutrophils but also uncovered an unsuspected ability of neutrophils to undergo caspase-1-dependent pyroptosis, a process where neutrophils exhibit a self-regulatory function that limit specific DAMP release.
Results
NETs contribute to Pseudomonas aeruginosa ExoU-driven pathology
In order to determine if neutrophil might be involved in pathology induced by exoU- expressing P. aeruginosa (P. aeruginosaExoU), we first monitored for the presence of dying neutrophils in the lungs of infected mice. Using intravital microscopy in lungs from transgenic mice expressing GFP under the granulocyte promoter MRP8 (MRP8GFP), we observed that P. aeruginosaExoU triggered significant neutrophil death that exhibited extracellular DNA, a feature of NETosis (Figure 1A, Movie S1). Related to these observations, we detected increased amounts of netotic markers (MPO/DNA or Histone/DNA complexes) in bronchoalveolar fluids (BALFs) from mice infected with P. aeruginosaExoU, whereas those markers were decreased in mice infected with P. aeruginosaExoUS142A, carrying a catalytically inactivating mutation in exoU gene (Figure 1B). This suggested to us that neutrophils are important targets of ExoU, and that P. aeruginosaExoU-induced neutrophil death might contribute to pathology. A hallmark of neutrophil NETosis-mediated pathology is the dysregulated release and accumulation of DNA extracellular Traps that promote endothelial and epithelial cell damages. As addition of DNAse-1 allows Neutrophil Extracellular Trap (NET) degradation and elimination, we infected mice with P. aeruginosaExoU or P. aeruginosaExoUS142A in presence or not of DNAse-1 and monitored for mice survival, bacterial loads and NET markers (Figures 1C, D). The amount of NET markers in BALs of mice infected with P. aeruginosaExoU were strongly decreased, although we did not detect significant induction of NETs upon infection with P. aeruginosaExoUS142A (Figure 1C). As consequence, DNAse-1 treatment improved the survival of P. aeruginosaExoU-infected mice but not the overall bacterial load, suggesting that ExoU-induced NETosis mostly contributes to damage-driven lethal pathology (Figure 1D). Altogether, our results suggest that ExoU-induced neutrophil NETosis mostly contributes to Pseudomonas aeruginosa-dependent pathology.
ExoU phospholipase activity drives PAD4-dependent NETosis
ExoU exhibits a calcium-independent phospholipase A2 activity that triggers epithelial and macrophage cell necrosis. Therefore, we wondered if ExoU phospholipase activity could regulate neutrophil killing by ExoU. Hence, we infected WT murine bone marrow neutrophils (BMNs) or human blood neutrophils with P. aeruginosaExoU (MOI2) or its isogenic mutant P. aeruginosaExoUS142A (MOI2) in presence or absence of the phospholipase inhibitor MAFP. Pharmacological inhibition or genetic inactivation of ExoU phospholipase activity (ExoUS142A) inhibited ExoU-dependent plasma membrane permeabilization in murine (BMNs) and human blood neutrophils (Figure 2A), hence confirming the importance of the ExoU phospholipase activity at promoting neutrophil death.
As we observed NET markers in infected mice (Figure 1C), we hypothesized that ExoU-dependent neutrophil death might also promote the generation of NETs. Using scanning electron microscopy, we observed that P. aeruginosaExoU stimulated a strong extracellular DNA release in BMNs, a process inhibited by the use of MAFP (Figure 2B). Among others, Histone degradation and DNA citullination are two conserved mechanisms that promote DNA relaxation and delobulation (Thiam et al., 2020). Neutrophil elastase and caspase-11 promote Histone degradation (Chen et al., 2018; Sollberger et al., 2018) and PAD4 triggers Histone citrullination (Li et al., 2010; Thiama et al., 2020; Wang et al., 2009). Therefore, we first explored whether ExoU-induced neutrophil DNA delobulation and release required PAD4-dependent citrullination. We observed that P. aeruginosaExoU induced PAD4-dependent Histone Citrullination (Figure 2C). In addition, nuclear delobulation and DNA extrusion were strongly impaired in Pad4-/- BMNs, suggesting that PAD4 plays a central function at driving DNA release during P. aeruginosaExoU infection (Figures 2C, D).
As PAD4-mediated Histone citrullination requires calcium signaling, a process that can be mediated by membrane pore formation (Chen et al., 2018), we infected BMN neutrophils WT or deficient for various neutrophil NETosis regulators such as GP91 (GP91phox-/-), Neutrophil Elastase, Cathepsin G, Proteinase 3 (NE-/-CatG-/-Pr3-/-), PAD4 (Pad4-/-) or Gasdermin D (GsdmD-/-). GP91phox-/-, NE-/-CatG-/-Pr3-/-, Pad4-/- and GsdmD-/- neutrophils all lysed (LDH release) to the same extend (Figure 2E). In addition, Scanning Electron Microscopy (SEM) analysis of NET formation showed no differences in the ability of NE-/-CatG-/-Pr3-/- or GsdmD-/- to go into NETosis upon P aeruginosaExoU whereas Pad4-/- showed resistance to trigger NETosis (Figure 2F).
This suggested to us that not only ExoU-induced NETosis occurs independently from classical NETotic regulators but also that ExoU-dependent neutrophil lysis could be uncoupled from PAD4-driven DNA decondensation and expulsion (referred here as NETosis). Next, we wondered about the respective importance of neutrophil lysis and PAD4-driven NETosis on the cell-autonomous immune response of neutrophils to P. aeruginosaExoU. We infected WT or Pad4-/- BMNs in presence or absence of MAFP and we observed that MAFP-treated neutrophils, but not Pad4-/- cells had improved bacterial killing capabilities (Figure S2A), hence suggesting that ExoU-mediated neutrophil lysis is sufficient to escape the microbicidal action of neutrophils.
Finally, we noticed that the infection of neutrophils with P. aeruginosaExoU in presence of MAFP only decreased but not abrogated neutrophil plasma membrane permeabilization after 3 hours of infection (Figure 2G). In this context, MAFP-treated GsdmD-/- neutrophils showed an improved protection upon P. aeruginosaExoU-induced plasma membrane permeabilization (Figure 2). Consequently, phospholipase inhibition by MAFP or infection with P. aeruginosaExoUS142A specifically induced Caspase1- and Gasdermin-D-dependent IL1β release although P. aeruginosaExoU triggered few or no IL1β release (Figure S2B).
Altogether, our results show that Pseudomonas aeruginosa-induced neutrophil lysis requires ExoU phospholipase activity, which subsequently triggers PAD4-dependent DNA citrullination, decondensation and expulsion from neutrophils. In addition, inhibition of ExoU phospholipase activity uncovers a compensatory pathway that drives GasderminD-dependent neutrophil lysis and IL1β release upon P. aeruginosa infection.
ExoU-deficient P. aeruginosa exposes a fully functional NLRC4-Caspase-1- Gasdermin D-dependent pyroptosis axis in neutrophils
The observation that, in absence of ExoU phospholipase activity, GSDMD mediates neutrophil death, encouraged us to decipher the molecular mechanisms behind. The infection with P. aeruginosaExoUS142A (MOI2) of WT murine neutrophils or deficient for various inflammasome sensors or signaling components, namely Aim2-/-, Nlrp3-/-, Nlrc4-/-, Casp11-/-, Casp1-/- and GsdmD-/- showed that only Casp1-/-, GsdmD-/- and Nlrc4-/- BMNs were protected against P. aeruginosaExoUS142A-dependent cell necrosis and IL-1β release (Figure 3A). In parallel, we monitored for plasma membrane permeabilization upon infection with P. aeruginosaExoUS142A (MOI2) in WT and Casp1-/- BMNs or human blood neutrophils with (Figures 3B, C). Consistently, Sytox Green uptake was strongly reduced in Casp1-/- BMNs (Figure 3B), a process that the Caspase-1 inhibitor Z-YVAD also repressed in human blood neutrophils upon P. aeruginosaExoUS142A infection (Figure 3C).
Those observations were paralleled with the detection of Caspase-1 (p20) and Gasdermin-D (p30) processing fragments in WT but not in Nlrc4-/- BMNs (Figure 3D). P. aeruginosa flagellin and rod/needle components from its Type-3 Secretion System (T3SS) trigger NAIP-dependent NLRC4 inflammasome activation in macrophages. Similarly, the use of P. aeruginosaExoU- (deficient for ExoU), P. aeruginosaExoU-FliC- (lacking both ExoU and Flagellin expression) or P. aeruginosaExsA- (lacking T3SS expression) strains showed that neutrophil death occurred in a T3SS and Flagellin- dependent manner (Figure S3A). We extended our observations to other ExoU- expressing P. aeruginosa (PA14, PA103) strains invalidated or not for ExoU expression. Indeed, infections of WT or Nlrc4-/- BMNs with PA14 and PA103 P. aeruginosa strains showed that the lack of ExoU expression also triggered NLRC4- dependent pyroptosis in neutrophils (Figure S3B). This, suggests that ExoU expression controls the ability of neutrophils to perform NLRC4-dependent pyroptosis. Neutrophils resist NLRC4/Caspase-1-dependent pyroptosis upon Salmonella infection (Chen et al., 2014). Hence, to further analyze the specificity of our findings with P. aeruginosaExoU-, we next infected WT or Nlrc4-/- BMNs with various bacteria (Shigella flexnerii, Chromobacter violaceum, Burkholderia thailandensis) known to trigger a NLRC4 inflammasome response and monitored for cell death and IL-1β release (Figure 3E) (Chen et al., 2014; Kovacs et al., 2020; Kumari et al., 2021; Maltez et al., 2015; Zhao et al., 2011). None of the tested bacteria triggered a significant NLRC4- dependent neutrophil lysis although they promoted NLRC4-dependent IL-1β release and gasdermin D processing (Figures 3E, F). In contrast to other bacteria, P. aeruginosa can use its Type-3 Secretion System independently of its phagocytosis (Hauser and Engel, 1999; Man et al., 2014a). Hence, we hypothesized that in neutrophils, once phagocytosed or endocytosed, bacteria might be exposed to various unknown factors able to restrict the NLRC4 inflammasome response. Regarding this, the direct electroporation of flagellin into the neutrophil host cell cytosol triggered NLRC4-dependent pyroptosis, whereas flagellin transfection, which requires the endocytic pathway to access host cell cytosol, did not lead to detectable death (Figure 3G). This suggests that the route by which NLRC4-activating ligands neutrophil cell are delivered to the cell strongly influences the ability of neutrophils to undergo pyroptosis.
Finally, to determine if P. aeruginosaExoUS142A-induced neutrophil NLRC4 inflammasome activation also occurs in vivo, we infected ASC-Citrine mice with low doses (1.105 CFUs) of P. aeruginosaExoU- or P. aeruginosaExoU-/FliC-, deficient for the expression of Flagellin. ImageStreamX observation of neutrophils presenting an active ASC supramolecular speck (ASC speck+/LY6G+ neutrophils) showed that P. aeruginosaExoU- infection triggered inflammasome activation in neutrophils, which was reduced when mice where infected with P. aeruginosaExoU-/FliC- (Figure S3C).
Altogether, our results show that the NLRC4/CASP1/GSDMD axis is fully functional to promote neutrophil pyroptosis in response to P. aeruginosa or direct Flagellin electroporation but not to the other NLRC4-activating bacteria tested, suggesting that one or many bacterial virulence or neutrophil intracellular factors might regulate the lytic threshold level of the NLRC4 inflammasome in neutrophils.
Caspase-1-induced neutrophil pyroptosis generates PAD4-dependent intracellular but not extracellular DNA structures
Next, we sought to determine whether Caspase-1-induced neutrophil pyroptosis could also promote NETosis upon P. aeruginosaExoUS142A infection. Scanning electron microscopy experiments showed that Caspase-1- and GasderminD-induced neutrophil pyroptosis did not efficiently generate extracellular DNA structures upon P. aeruginosaExoUS142A infection, although P. aeruginosaExoU robustly induced NETs (Figure 4A). Rather, immunofluorescence experiments revealed that P. aeruginosaExoUS142A triggered efficient DNA decondensation as well as expulsion from the nuclear envelop (Lamin-B1 staining) but no or few DNA release from the neutrophil plasma membrane both in murine BMNs and human blood neutrophils (Figure 4B, S4A).
Next, we wondered about the mechanisms by which such process might occur. Immunoblotting and microscopy analysis of Histone citrullination showed that P. aeruginosaExoUS142A induced a robust Histone3 (H3)-Citrullination in a NLRC4- and GasderminD-dependent manner (Figure 4C, S4B), a process that was also seen in human blood neutrophils (Figure S4C, D). Conversely, ASC-Citrine BMNs revealed that NLRC4/CASP1/GSDMD-induced DNA decondensation required PAD4 as pharmacological inhibition of PAD4 (GSK484) abrogated both Histone citrullination as well as the DNA nuclear release (Figure 4D, E). In addition, measure of ASC specks (ASC+), cell lysis (LDH release) and IL-1β release in ASC-Citrine, WT, Pad4-/- and Nlrc4-/- BMNs highlighted that PAD4 was not involved in P. aeruginosaExoUS142A- induced NLRC4 inflammasome activation (Figure 4D-F).
Next, to determine if Caspase-1-induced neutrophil lysis and PAD4-dependent DNA decondensation plays a microbicidal function, we infected WT, Pad4-/- and Nlrc4-/- BMNs with P. aeruginosaExoUS142A and evaluated their cell-autonomous immune capacities. Nlrc4-/- BMNs had improved ability to restrict P. aeruginosaExoUS142A infection than WT and Pad4-/- neutrophils (Figure 4G). Such results were also observed in human blood neutrophils where only Caspase-1 inhibition (Z-YVAD) but not PAD4 inhibition (GSK484) improved neutrophil-mediated P. aeruginosaExoUS142A killing (Figure 4G), suggesting that neutrophil pyroptosis more than PAD4-driven DNA decondensation promotes neutrophil failure to restrict P. aeruginosaExoUS142A.
Finally, in order to determine if neutrophils can undergo Caspase-1-dependent death in mice, we infected MRP8-GFP+ mice and monitored for the granulocyte death features using intravital microscopy (Figure 4H, Movie S2). Although necrotic granulocytes exhibited NETotic features (e.g. extracellular DNA) upon exposure to P. aeruginosaExoU, P. aeruginosaExoUS142A infection led to the appearance of swelled- round necrotic granulocytes that exhibited intracellular decondensed DNA, similarly to what we observed in vitro (Figure 4H, Movie S2). This suggests that upon lung infection, Caspase-1-induced neutrophil pyroptosis is well occurring and displays morphological and immunological distinct characteristics to NETs. All in one, our results describe an original pathway by which NLRC4-induced neutrophil pyroptosis generates PAD4-dependent intracellular but not extracellular DNA structures.
Subcortical actin cytoskeleton limits the release of the damaging DAMPs Histones upon Caspase-1-induced pyroptosis
The observation that Caspase-1-induced neutrophil pyroptosis generates an “incomplete” NETosis brought the questions of what process/effectors determine the ability of DNA to breach or not the plasma membrane and what are the immunological consequences of keeping DNA intracellularly.
To address the first question, we took advantage of a recent study from the Waterman lab that found that stabilizing subcortical filamentous actin with the stabilizing agent jasplakinolide did not abrogate DNA decondensation but rather inhibited DNA expulsion from the neutrophil upon NETosis inuction (Thiama et al., 2020), an observation that our findings with P. aeruginosaExoUS142A mirror. Hence, we sought to determine the behavior of filamentous actin upon both P. aeruginosaExoU-induced NETosis and P. aeruginosaExoUS142A-induced “incomplete” NETosis. Fluorescence microscopy observations showed that upon infection with P. aeruginosaExoU NETotic neutrophils completely lost F-actin staining as well as the nuclear envelop cytoskeletal protein Lamin-B1 (Figure 5A). To the contrary, we observed that subcortical F-actin was condensed close to the plasma membrane upon P. aeruginosaExoUS142A-induced neutrophil pyroptosis (Figure 5A). Such observations were also validated when we immunoblotted actin in NETotic and pyroptotic neutrophils (Figure 5B). Indeed, we observed a disappearance of actin in NETotic samples, whereas actin was still present in samples from pyroptotic neutrophils (Figure 5B). Lamin B1, an essential nuclear envelope cytoskeletal component was found to be strongly cleaved upon P. aeruginosaExoU infection and to a lower extend by P. aeruginosaExoUS142A-induced NLRC4-dependent response, suggesting that alterations of various cytoskeletal components might also account for DNA release from the nucleus (Figure 5B) (Knight et al., 2019; Li et al., 2020). As Lamin B1 is mostly involved at regulating nuclear integrity but not extracellular DNA release, we mostly focused on the role of actin degradation and/or F-actin destabilization at regulating DNA-breached plasma membrane of neutrophils. To address this, we infected murine neutrophils with P. aeruginosaExoUS142A and after 2h of infection, a time where the neutrophil DNA was decondensed and could fill the intracellular compartment, we added the actin depolymerizing agent Latrunculin A and monitored for the formation of NETs (Figure 5C). We controlled that Latrunculin A use did not modify P. aeruginosaExoUS142A- induced neutrophil lysis (LDH release) or ASC supramolecular complexes formation (%ASC specks+ cells) (Figure 5C, S5A). We observed that Latrunculin A- depolymerized F-actin induced the appearance of NET-like structures in P. aeruginosaExoUS142A-infected neutrophils (Figure 5C). To challenge these findings, we also reasoned that if actin depolymerization could induce DNA expulsion upon pyroptosis, actin stabilization might inhibit DNA expulsion upon NETosis induction. In this context, we infected neutrophils with P. aeruginosaExoU for 30 minutes as ExoU triggers a very fast neutrophil NETosis and then added the stabilizing agent jasplakinolide (Figure S5B). Microscopy observation and quantification of DNA able to cross the plasma membrane showed that jasplakinolide-treated cells kept intracellularly their decondensed DNA upon P. aeruginosaExoU infection (Figure S5B). These results suggest that cortical F-actin acts as a physical barrier against DNA expulsion upon neutrophil pyroptosis, a capacity that disappears in NETosis contexts. Next, we addressed the immunological relevance of pyroptotic neutrophils. We hypothesized that, contrary to NETs, pyroptotic neutrophils might keep intracellularly several DAMPs in order to lower unnecessary tissue damages.
Subsequently, to determine to what extent the content of released protein might vary between neutrophil pyroptosis and NETosis, we performed a comparative mass spectrometry analysis of the secretomes of pyroptotic and NETotic neutrophils (Figure 5D). To ensure a clear comparison, we set up experimental conditions where both NETotic and pytopotic neutrophils died to the same extend (Figure 5D). Hence, BMNs were infected for 3 hours with P. aeruginosaExoU at a MOI of 0.5 and P. aeruginosaExoUS142A at an MOI of 2. Comparative analysis of supernatants detected approximatively 450 proteins that were mostly enriched extracellularly (more than two fold) upon NETosis, including Histones core components (Histone H3) or Histone- associated proteins (Histone H1) (Figure 5D, Table S1). To the contrary, both pyroptotic and NETotic neutrophils released the necrotic markers Ldh or the alarmins IL36γ, HMGB1 and HMGB2 (Henry et al., 2016; Phulphagar et al., 2021) to a similar extend, suggesting that pyroptosis of neutrophils might interfere with the release of specific DAMPs (Figure 5D). These features were then further validated by immunoblotting and ELISA against HMGB1 or Histones release upon both NETosis and pyroptosis induction (Figures 5E, S5C, D). In those experiments, only P. aeruginosaExoU promoted strong Histone H3 and Citrullinated Histone H3 release although HMGB1 DAMP was released in both situations (Figures 5E, S5C, D). In addition, we observed that latrunculin A-destabilized F-actin also induced Histone 3 release upon P. aeruginosaExoUS142A infection of neutrophils (Figure 5F). To the contrary, japlakinoloide strongly impaired ExoU-dependent release of extracellular Histones (Figure 5F), hence suggesting that F-actin integrity contributes to restrain DNA-associated Histone release upon neutrophil pyroptosis. When present extracellularly, Histones are extremely potent pro-inflammatory and damaging molecules (Silvestre-Roig et al., 2019; Xu et al., 2009, 2015). Indeed, direct injection of Histones into the blood stream of mice triggers a sepsis like response (Xu et al. 2009). In this context, we hypothesized that P. aeruginosaExoU-induced extracellular, release of Histones might be an important contributor of lethality in vivo. Therefore, we infected mice intranasally with P. aeruginosaExoU or P. aeruginosaExoUS142A in presence or absence of N-Acetyl Heparin (NAH), a non-anticoagulant heparin-derived molecule that inhibits Histone mediated cellular damages (Figures 5G, S5E) (Wen et al., 2016; Wildhagen et al., 2014; Zhang et al., 2014). P. aeruginosaExoUS142A-infected mice exhibited moderate lethality, which was not significantly modified by the use of NAH (Figure 5G). By contrast, NAH strongly improved survival of P. aeruginosaExoU-infected mice, hence suggesting that NETosis specifically contributes to Histone-associated lethality in response to P. aeruginosaExoU (Figure 5G).
Altogether, our results unveil an original process where F-actin act as physical barrier able to inhibit DNA extrusion out from neutrophils upon Caspase-1-induced pyroptosis, hence limiting Histone-mediated lethality.
Targeting both inflammasome and ExoU phospholipase activity synergistically improves P. aeruginosa elimination
Our results showed that both NETosis and neutrophil pyroptosis are inefficient at eliminating P. aeruginosa in vitro and that NETosis-induced Histone release contributes to P. aeruginosaExoU-driven mouse lethality. In this context, we aimed at determining the role of the Neutrophil Caspase-1 upon P. aeruginosa infection in mice. We infected MRP8Cre-Casp1flox and MRP8Cre+Casp1flox mice either intranasally or systemically with P. aeruginosaExoU or P. aeruginosaExoUS142A strains. We observed that, upon lung and systemic infections with P. aeruginosaExoU, MRP8Cre+Casp1flox mice did not show any differences in bacterial elimination, IL1β production or Histone/DNA complexes presence in BALs or plasma, confirming that P. aeruginosaExoU triggers successful infection independently of the inflammasome pathways (Figures 6A-D, S6A). To the contrary, MRP8Cre+Casp1flox mice infected with P. aeruginosaExoUS142A showed a slight but significant improved bacterial elimination in lungs, a phenotype that was further amplified in systemic infection as shown in spleen, liver and lung (Figure 6A, C). In parallel, BAL and plasma IL-1β levels, but not Histone/DNA complexes, were decreased in MRP8Cre+Casp1flox, hence suggesting that neutrophil Caspase-1 is also a contributor of IL-1β production upon P. aeruginosaExoUS142A infection (Figure 6B, D).
Finally, we aimed at determining if targeting both ExoU phospholipase activity and inflammasome response could synergistically protect mice against infection. Both Faure et al and Cohen et al previously showed that NLRC4 in macrophages played a deleterious immune response to P. aeruginosa lung infection (Cohen and Prince, 2013; Faure et al., 2014). Hence, as NLRC4 in macrophages and neutrophils is of importance in P. aeruginosa infection, we infected WT or Nlrc4-/- mice with P. aeruginosaExoU or P. aeruginosaExoUS142A and monitored for mice survival and bacterial loads (FigsS6B, C). We observed that NLRC4 did not protect against P. aeruginosaExoU-induced lethality or against bacterial growth (FigsS6B, C). However, P. aeruginosaExoUS142A infection was strongly attenuated in WT mice, a process that was even further amplified in Nlrc4- /- mice that almost entirely cleared bacteria from the lung, hence suggesting that targeting both ExoU-induced lethal pathology and inflammasome-promoted bacterial growth synergistically improve both mice survival and Pseudomonas elimination.
Altogether, our results highlight that neutrophil Caspase-1 activity contributes to P. aeruginosa spreading in mice in absence of ExoU.
Discussion
Induction of NETosis is a crucial process in the host defense against extra- and intra- cellular pathogens but its dysregulation also favors autoimmunity or infectious sepsis. Whereas infections involving P. aeruginosa are mostly studied through the chronic prism, which occurs with P. aeruginosa strains that initially express (then repress during the chronic infectious stage) the ExoS toxin, ExoU-expressing P. aeruginosa strains drive acute and lethal infections (Ozer et al., 2019). Here, we found that upon P. aeruginosaExoU infection, neutrophils undergo pathological NETosis, which induces both aberrant NETosis-induced lethality and bacterial escape from neutrophil uptake and killing. Specifically, ExoU induces phospholipid cleavage and degradation, which triggers neutrophil lysis. Subsequently, neutrophil membrane alterations trigger PAD4 activation, hence ensuring DNA relaxation and the full NETosis process. Although NE/PR3/CathG, ROS and Caspase-11 all play a strong role at promoting NETosis in various infectious and sterile contexts (Chen et al., 2018; Kenny et al., 2017; Papayannopoulos et al., 2010), none of them was important for ExoU-induced neutrophil lysis or NETosis, which suggests that additional factors might control ExoU- induced neutrophil lysis.
A key observation of our study was that, although neutrophils resist NLRC4- and NLRP3-induced Caspase-1-dependent pyroptosis upon various bacterial challenges (Chen et al., 2014; Karmakar et al., 2020; Kovacs et al., 2020), the lack of ExoU expression induced neutrophil pyroptosis through the engagement of the fully competent canonical NLRC4 inflammasome (Sutterwala et al., 2007). In agreement with previous studies, we did not find another bacterium among the well-known bacteria that trigger NLRC4 inflammasome response able to trigger neutrophil pyroptosis in neutrophils, at the exception of the various strains of P. aeruginosa tested. It is clear enough that neutrophils show more resistance than macrophages to NLRC4-dependent pyroptosis (Chen et al., 2014; Heilig et al., 2018; Nichols et al., 2017), yet P. aeruginosaEXOUS142A still trigger neutrophil pyroptosis. This suggests to us that beyond their intrinsic resistance to pyroptosis (e.g. ESCRT machinery, Caspase-1 expression levels, Ragulator pathway…) (Bjanes et al., 2021; Chen et al., 2018; Evavold et al., 2020; Rühl et al., 2018), neutrophils might have additional factors that restrict the ability of bacteria to promote NLRC4-dependent pyroptosis. Related to this, a seminal study from Zynchlynsky and colleagues found that neutrophil serine proteases could degrade the Type-3 Secretion System and flagellin virulence factors of S. flexneri (Weinrauch et al., 2002), hence limiting their ability to hijack the neutrophil autonomous immunity and restraining Shigella-induced neutrophil necrosis. Similarly, upon P. aeruginosa infection, neutrophils deficient for the Nadph oxidase enzyme Nox2 undergo some low degree of caspase-1-dependent pyroptosis, which drives a deleterious response of the host (Ryu et al., 2017). As P. aeruginosa can inject flagellin directly through the plasma membrane in absence of phagocytosis, one could speculate that bypassing the endocytic/phagocytic pathway of neutrophils, allows escaping T3SS and flagellin degradation or ROS-mediated alterations, hence allowing direct access of large amounts of flagellin into the neutrophil cytosol. Related to this, our results show that flagellin electroporation but not its transfection promotes NLRC4- driven neutrophil pyroptosis, hence suggesting that avoiding the endocytosis/phagocytosis pathway might be sufficient to trigger NLRC4-dependent pyroptosis in neutrophils. Interestingly, Chen and colleagues recently found that upon infection with Yersinia, murine neutrophils induce a pyroptotic program that involves virulence-inhibited innate immune sensing, hence promoting RIPK1-induced Caspase 3-dependent Gasdermin E cleavage and activation and neutrophil pyroptosis (Chen et al., 2021a), which also suggests that neutrophil pyroptosis can occur through different molecular pathways.
Similar to ExoU-mediated NETosis, Caspase-1 and Gasdermin-D also promoted PAD4-dependent Histone Citrullination, which stimulated DNA relaxation and release from the nucleus but not its extracellular expulsion. Why upon ExoU, Caspase-11 (Chen et al., 2018), MLKL (D’Cruz et al., 2018), NADPH (Kenny et al., 2017) or NE/CatG/Pr3 (Papayannopoulos et al., 2010) stimulation but not upon Caspase-1 activation neutrophils generate two different types of DNA structures remains yet to be investigated. Upon Caspase1-induced pyroptosis in neutrophils, a substantial proportion of subcortical actin was still present on its Filamentous form, which in our settings constrained neutrophil DNA intracellularly whereas ExoU-stimulated neutrophils completely lost cellular F-actin network. Similarly, Thiam et al., (Thiama et al., 2020) also observed that F-actin stabilization allowed the formation these peculiar structures upon Ionomycin-induced NETosis. Interestingly, neutrophil elastase has also been shown to degrade actin (Metzler et al., 2014), hence ensuring efficient NETosis induction. This, suggests that extracellular DNA release, the final step of NETosis, might actually be a cell-regulated process involving various, yet to be determined, regulators (Neubert et al., 2018; Thiam et al., 2020). An obvious question lies on both pyroptotic Caspase-11 and necroptotic MLKL proteins and their ability to directly or indirectly induce actin degradation or F-actin destabilization, and whether additional neutrophil components also determine the ability of neutrophil to specifically control extracellular DNA release in this context.
Regarding the immunological purpose of Caspase-1-induced neutrophil pyroptosis, we hypothesize that, mainly due to the broad diversity of inflammasomes, Caspase-1 has more possibilities to be activated than caspase-11. One guess would be that the decondensation of DNA but its conservation into the intracellular space might be a physical mean for neutrophils to trap some intracellular DAMPs, hence avoiding their passive release and a too strong exacerbation of the inflammatory response. Regarding this, our observations that DNA-bound Histones mostly remain trapped intracellularly, but not HMGB1 alarmin, both initially located in the nucleus show that additional regulatory pathways might be involved at promoting HMGB1 release but not Histones upon neutrophil pyroptosis. In the light of the recent discovery from Kayagaky and colleagues on the role for Ninjurin-1 at promoting active cell shrinkage and HMGB1/nucleosome DAMP release downstream of Gasdermin-D pores in macrophages, the use of Ninjurin-1 deficient mice are full of promises in order to determine if neutrophil extracellular vs intracellular DNA can be a mean to trap DAMPs, hence dampening inflammatory and pathological responses (Kayagaki et al., 2021). All in one, our results unveil an unsuspected ability of neutrophils to undergo caspase-1-dependent pyroptosis which drives an intriguing “incomplete NETosis”, hence expanding our understanding of neutrophil death mechanisms and opening novel research areas regarding the immunological importance of such process in various infectious and non-infectious contexts.
AUTHOR CONTRIBUTIONS
RP and EM designed the experiments. RP, KS and EM wrote the manuscript. RP and KS performed the experiments with the help of SB, DP, PJB, AH, MP, SALI, YR, RA, ED, FA, CC, AGDP, EL, RP. Specifically RP and RP performed SEM experiments, SM, EB and EL set up and performed intravital mouse experiments, RA, FA, YR and AGDP performed mass spectrometry experiments. JPG, OBS, CTNP, ML, NW and CP provided essential reagents, tools and inputs for the conduct of the project. EM, RP and KS supervised the entire study.
CONFLICT OF INTEREST
Authors have no conflict of interest to declare.
STAR Methods
All reagents, concentrations of use and their references are listed in the Reagent Table Mice Nlrc4−/− (Man et al., 2014b), Nlrp3−/− (Martinon et al., 2006), ASC−/−, Casp11−/− (Li et al., 1995; Wang et al., 1998), Casp1−/−Casp11−/− (Li et al., 1995; Wang et al., 1998), GsdmD−/−, Aim2−/−, Pad4−/−, Gp91phox−/−, NE−/−CatG−/−Pr3−/− (Yan et al., 2016), ASCCitrine, MRP8Cre+GFP+ were generated and described in previous studies. Mice were bred at the IPBS (Toulouse, France) and INRAE (Tours Nouzilly, France) animal facilities in agreement to the EU and French directives on animal welfare (Directive 2010/63/EU). Charles Rivers provided WT C57BL/6 mice. Mice experiments are under legal authorizations APAFIS#8521-2017041008135771 and APAFIS#12812- 2018031218075551, according to the local, French and European ethic laws.
MRP8CreCasp1flox mice genotyping
Casp1flox/flox mice were crossed to MRP8Cre mice to generate MRP8CreCasp1flox. Caspase-1 genotyping was performed using Primer Fw: CGAGGGTTGGAGCTCAAGTTGACC and Primer Rv: CACTTTGACTTCTCTAAGGACAG. Cre genotyping was performed using Primers Fw: CGCCGTAAATCAATCGATGAGTTGCTTC and Primers Rv: GATGCCGGTGAACGTGCAAAACAGGCTC.
Bacterial cultures
P. aeruginosa strains (PP34, PA14, PA103) and their isogenic mutants were grown overnight in Luria Broth (LB) medium at 37°C with constant agitation. Otherwise specified, all along the study the clinical isolate PP34, referred as P.aExoU, or its isogenic mutants referred as P.aExoUS142A, P.aExoU-, P.aExoU-FliC-, P.aExsaA- were used. Bacteria were sub-cultured the next day by diluting overnight culture 1:25 and grew until reaching an optical density (OD) O.D.600 of 0.6 – 0.8. Bacterial strains and their mutants are listed in reagent table.
Mice infections
Age and sex-matched animals (5–8 weeks old) per group were infected intravenously (venous-orbital, 50µL with 1.107CFUs) or intranasally with 5.105 (lethal doses) or 2.5.105 CFUs of various P. aeruginosa strains suspended in 25µL of PBS. When specified, intranasal addition of PBS, DNase 1 (4000U/mouse) or N-Acetyl Heparin (NAH) were performed 3 and 9 hours after infections. Specifically, NAH was injected both by intranasal aspiration (25 µL/mouse, 5mg/kg) and by orbital injection (50µL/mouse, 15mg/kg) 3 and 9 hours after infection to ensure both lung and BAL access of NAH. Animals were sacrificed at indicated times after infection and bronchoalveolar fluids (BALFs), blood and lungs were recovered. When specified, bacterial loads (CFU plating), cytokine levels (ELISA) and NET complexes (MPO/DNA, Histone/DNA, ELISA) were evaluated. No randomization or blinding were done.
Intravital microscopy experiments
We relied on the previously published lung intravital microscopy method using an intercoastal thoracic window (Headley et al., 2016; Looney et al., 2011), adapted at the IPBS CNRS-University of Toulouse TRI platform. MRP8-mTmG mice (8-12 weeks old) were infected intratracheally with 5.105 CFUs of P. aeruginosa ExoU or ExoUS142A strains resuspended in 50µL of PBS and imaged 6 to 8 hours after infection. 50µL of 50µM solution of Sytox blue (Life Technologies) was injected both intravenously (retroorbital) and intratracheally just before imaging to visualize extracellular DNA.
Next, mice were anesthetized with ketamine and xylazine and secured to a microscope stage. A small tracheal cannula was inserted, sutured and attached to a MiniVent mouse ventilator (Harvard Apparatus). Mice were ventilated with a tidal volume of 10 μl of compressed air (21% O2) per gram of mouse weight, a respiratory rate of 130-140 breaths per minute, and a positive-end expiratory pressure of 2-3 cm H2O. Isoflurane was continuously delivered to maintain anesthesia and 300 μl of 0.9% saline solution were i.p. administered in mice every hour for hydration. Mice were placed in the right lateral decubitus position and a small surgical incision was made to expose the rib cage. A second incision was then made into the intercostal space between ribs 4 and 5, through the parietal pleura, to expose the surface of the left lung lobe. A flanged thoracic window with an 8 mm coverslip was inserted between the ribs and secured to the stage using a set of optical posts and a 90° angle post clamp (Thor Labs). Suction was applied to gently immobilize the lung (Dexter Medical). Mice were placed in 30°C heated box during microscopy acquisition to maintain the body temperature and the 2-photon microscope objective was lowered over the thoracic window. Intravital imaging was performed using a Zeiss 7MP upright multi-photon microscope equipped with a 20×/1.0 objective and a Ti-Sapphire femtosecond laser, Chameleon-Ultra II (Coherent Inc.) tuned to 920 nm. Sytox Blue, GFP and Tomato emission signals were detected thanks to the respective bandpass filters: Blue (SP485), Green (500-550) and Red (565-610). Images were analyzed using Imaris software (Bitplane) and Zen (Zeiss).
Isolation of primary murine neutrophils
Murine Bone marrow cells were isolated from tibias and femurs, and neutrophils were purified by positive selection using Anti-Ly-6G MicroBead Kit (Miltenyi Biotech) according to manufacturer’s instructions. This process routinely yielded >95% of neutrophil population as assessed by flow cytometry of Ly6G+/CD11b+ cells.
Isolation of primary human neutrophils
Whole blood was collected from healthy donors by the “Ecole française du sang” (EFS, Tolouse Purpan, France) in accordance with relevant guidelines. Written, informed consent was obtained from each donor. Neutrophils were then isolated by negative selection using MACSxpress® Whole Blood Human Neutrophil Isolation Kit (Miltenyi Biotech) according to manufacturer’s instructions. Following isolations cells were centrifuged 10 min at 300 g and red blood cells were eliminated using Red blood cells (RBC) Lysis Buffer (BioLegend). This procedure gives >95% of neutrophil population as assessed by flow cytometry of CD15+/CD16+ cells. License to use human samples is under legal agreement with the EFS; contract n° 21PLER2017-0035AV02, according to Decret N° 2007-1220 (articles L1243-4, R1243-61).
Cell plating and treatment of Neutrophils
Following isolation, Neutrophils were centrifugated for 10 min at 300 g and pellet was resuspendent in serum free OPTI-MEM medium. Absolute cell number was determined with automated cell counter Olympus R1 with trypan blue cell death exclusion method (typically living cells represent >70% of cell solution) and cell density was adjusted at 106 / mL by adding OPTI-MEM culture medium. Neutrophils were then plated in either 96 well plates, 24 well plates or 6 well plates with 100 µL (105 cells), 500 µL (5.105 cells) or 2 mL (2.106 cells) respectively. When indicated cells were primed with Pam3CSK4 (100 ng/ml) or LPS (100 ng/ml) for 2 hrs and/or incubated with chemical inhibitors Z-VAD-fmk (20 µM), Y-VAD-fmk (40 µM), GSK484 (10 µM), Latrunculin A (2 µM), Jasplakinolide (200 nM) as indicated in each experimental setting. Neutrophils were infected with various bacterial strains and multiplicity of infections (M.O.I.) as indicated.
Electroporation / transfection of flagellin
Before electroporation/transfection neutrophils were first primed with Pam3CSK4 (1 µg.mL-1) for 2 hrs, and washed with PBS. Recombinant flagellin (FLA-PA, invivoGen) 1 to 5 µg was electroparated into ∼2.0 × 106 cells in 20 µL of Tampon R (R buffer) using the Neon transfection system (Life Technologies). Settings were the following: 1720 Voltage, 10 Width, 2 Pulse. After electroporation cells were plated in 6 or 24 well plates and incubated for 3hrs before further experiments. For transfection, 1 to 5 µg of flagellin was mixed with 0.1% v/v Lipofectamine 2000 (Life Technologies) in Optimem medium, incubated for 5 min at room temperature, and transfected into ∼2.0 × 106 cells in 6 or 24 well plates. Cells were incubated for 4hrs before further manipulations.
Kinetic analysis of Neutrophil’s permeability by SYTOX Green incorporation assay
Cells are plated at density of 1 x 105 per well in Black/Clear 96-well Plates (REF) in OPTI-MEM culture medium supplemented with SYTOX-Green dye (100ng/mL) and infected/treated as mentioned in figure legend. Green fluorescence are measured in real-time using Clariostar plate reader equipped with a 37°C cell incubator.
ELISA and plasma membrane lysis tests
Cell death was measured by quantification of the lactate dehydrogenase (LDH) release into the cell supernatant using LDH Cytotoxicity Detection Kit (Takara). Briefly, 100 μL cell supernatant were incubated with 100 μL LDH substrate and incubated for 15 min. The enzymatic reaction was stopped by adding 50 μL of stop solution. Maximal cell death was determined with whole cell lysates from unstimulated cells incubated with 1% Triton X-100. Human and mouse IL-1β secretion was quantified by ELISA kits (Thermo Fisher Scientific) according to the manufacturer’s instructions.
Quantification of NETs
Histone/DNA complexes in cells supernatant or serum sample were quantified using Cell Death Detection ELISAPLUS according to manufacturer’s instructions (roche). MPO/DNA complexes was assessed as describes previously (Lefrançais et al., 2018). Citrullinated Histone H3/DNA complexes quantification assay was build “in house” and performed similarly to standard ELISA procedure (Thermo Fisher Scientific). Specifically we used anti-Citrullinated H3 antibody (1/1000 in PBS) as capture antibody and anti H3 tot-biot (KIT Cell death roche 1/200) as detection antibody.
Preparation of neutrophil lysates and supernatant for immunoblot
At the end of the treatment 5 mM of diisopropylfluorophosphate (DFP) cell permeable serine protease inhibitor was added to cell culture medium. Cell’ Supernatant was collected and clarified from non-adherent cells by centrifugation for 10 min at 300 g. Cell pellet and adherent cells were lysed in 100 µL of RIPA buffer (150 mM NaCl, 50 mM Tris-HCl, 1% Triton X-100, 0.5% Na-deoxycholate) supplemented with 5 mM diisopropylfluorophosphate (DFP) in addition to the protease inhibitor cocktail (Roche). Cell scrapper was used to ensure optimal recovery of cell lysate. Collected cell lysate was homogenized by pipetting up and down ten times and supplemented with laemli buffer (1X final) before boiling sample for 10 min at 95°C. Soluble proteins from cell supernatant fraction were precipitated as described previously (Eren et al., 2020). Precipitated pellet was then resuspended in 100 µL of RIPA buffer plus laemli supplemented with 5 mM diisopropylfluorophosphate (DFP) and protease inhibitor cocktail (Roche) and heat denaturated for 10 min at 95°C. Cell lysate and cell supernatant fraction were then analysed by immunoblot either individually or in pooled sample of lysate plus supernatant (equal vol/vol).
Treatment of Neutrophils for Immunofluorescences
5.105 Cells were plated on 1.5 glass coverslips in 24 well plate and infected/treated as described above. At the end of the assay, cell supernatant was removed and cells were fixed with a 4% PFA solution (ref) for 10 min at 37°C. PFA was then removed and cells were washed 3 times with HBSS. When desired, plasma membrane was stained with Wheat Germ Agglutinin, Alexa Fluor™ 633 Conjugate (ThermoFisher Scientifique) at 1/100th dilution in HBSS, and incubated for 30 min under 100 rpm orbital shaking conditions. Then cells were washed with HBSS and processed for further staining. Permeabilization was performed by incubating cells for 10 min in PBS containing 0.1% Triton X-100. To block unspecific binding of the antibodies cells are Incubated in PBS- T (PBS+ 0.1% Tween 20), containing 2% BSA, 22.52 mg/mL glycine in for 30 min. 3 washes with PBS-T was performed between each steps. Primary antibodies staining was performed overnight at 4°C in BSA 2% - Tween 0.1% - PBS (PBS-T) solution. Coverslips were washed three times with PBS-T and incubated with the appropriate fluor-coupled secondary antibodies for 1 hour at room temperature. DNA was counterstained with Hoechst. Cells were then washed three times with PBS and mounted on glass slides using Vectashield (Vectalabs). Coverslips were imaged using confocal Zeiss LSM 710 (INFINITY, Toulouse) or Olympus Spinning disk (Image core Facility, IPBS, Toulouse) using a 63x oil objective. Unless specified, for each experiment, 5-10 fields (∼50-250 cells) were manually counted using Image J software.
Scanning Electron Microscopy experiments
For scanning electron microscopy observations, cells were fixed with 2.5% glutaraldehyde in 0.2M cacodylate buffer (pH 7.4). Preparations were then washed three times for 5min in 0.2M cacodylate buffer (pH 7.4) and washed with distilled water. Samples were dehydrated through a graded series (25 to 100%) of ethanol, transferred in acetone and subjected to critical point drying with CO2 in a Leica EM CPD300. Dried specimens were sputter-coated with 3 nm platinum with a Leica EM MED020 evaporator and were examined and photographed with a FEI Quanta FEG250.
ImageStreamX
Cells isolated from peritoneal washes were pelleted by centrifugation (10 min at 300 g). Neutrophils were stained prior to fixation with anti-Ly6G (APC-Vio770, Miltenyi- Biotec Clone: REA526 | Dilution: 1:50) in MACS buffer (PBS-BSA 0,5%-EDTA 2mM) in presence of FC block (1/100) and Hoechst (1 µM). Then, cells were fixed in 4% PFA. Data were acquired on ImageStreamXMKII (Amnis) device (CPTP Imaging and Cytometry core facility) and analyzed using IDEAS software v2.6 (Amnis). The gating strategy used to evaluate inflammasome activation in neutrophils was performed as follow: (i) a gate was set on cells in focus [Cells in Focus] and (ii) a sub-gate was [LY6G+] and second on (iv) ASC-citrine+ and Hoechst+ cells [Hoechst+/ASC-Citrine+] within LY6G+ population. (v) To distinguish cells with active (ASC-speck) versus inactive inflammasome (Diffuse ASC), we plotted the Intensity with the area of ASC- citrine. This strategy allow to distinguish cells with active inflammasome that were visualized and quantified. (Fig S3E).
Mass spectrometry
Tryptic peptides were resuspended in 2% acetonitrile and 0.05% trifluoroacetic acid and analyzed by nano-liquid chromatography (LC) coupled to tandem MS, using an UltiMate 3000 system (NCS-3500RS Nano/Cap System; Thermo Fisher Scientific) coupled to an Orbitrap Q Exactive Plus mass spectrometer (Thermo Fisher Scientific). Around 1µg of each sample was loaded on a C18 precolumn (300 µm inner diameter × 5 mm, Thermo Fisher Scientific) in a solvent made of 2% acetonitrile and 0.05% trifluoroacetic acid, at a flow rate of 20 µl/min. After 5 min of desalting, the precolumn was switched online with the analytical C18 column (75 µm inner diameter × 50 cm, in-house packed with Reprosil C18) equilibrated in 95% solvent A (5% acetonitrile, 0.2% formic acid) and 5% solvent B (80% acetonitrile, 0.2% formic acid). Peptides were eluted using a 5%-50% gradient of solvent B over 105 min at a flow rate of 300 nl/min. The mass spectrometer was operated in data-dependent acquisition mode with the Xcalibur software. MS survey scans were acquired with a resolution of 70,000 and an AGC target of 3e6. The 10 most intense ions were selected for fragmentation by high-energy collision induced dissociation, and the resulting fragments were analyzed at a resolution of 17500, using an AGC target of 1e5 and a maximum fill time of 50ms. Dynamic exclusion was used within 30 s to prevent repetitive selection of the same peptide.
Data processing
Raw MS files were processed with the Mascot software (version 2.7.0) for database search and Proline (Bouyssie et al, Bioinformatics 2020) for label- free quantitative analysis. Data were searched against Mus musculus entries of the UniProtKB protein database (release UniProtKB/Swiss-Prot+TrEMBL 2020_07, 87954 entries). Carbamidomethylation of cysteines was set as a fixed modification, whereas oxidation of methionine and protein N-terminal acetylation were set as variable modifications. Specificity of trypsin digestion was set for cleavage after K or R, and two missed trypsin cleavage sites were allowed. The mass tolerance was set to 10 ppm for the precursor and to 20 mmu in tandem MS mode. Minimum peptide length was set to 7 amino acids, and identification results were further validated in Proline by the target decoy approach using a reverse database at both a PSM and protein false-discovery rate of 1%. For label-free relative quantification of the proteins across biological replicates and conditions, cross-assignment of peptide ions peaks was enabled with a match time window of 1 min, after alignment of the runs with a time window of +/- 600s.
Statistical tests used
Statistical analysis was performed with Prism 8.0a (GraphPad Software, Inc.). Otherwise specified, data are reported as mean with SEM. T-test with Bonferroni correction was chosen for comparison of two groups. For in vivo mice experiments and comparisons we used Mann-Whitney tests and mouse survival analysis were performed using log-rank Cox-Mantel test. P values are shown in figures with the following meaning; NS non-significant and Significance is specified as *p ≤ 0.05; **p ≤ 0.01, ***p ≤ 0.001.
Table reagents
Reagents and Tools are available upon request to Etienne.meunier{at}ipbs.fr or Remi.planes{at}gmail.com
Supplemental information
Movie S1. Intravital microscopy visualization of granulocyte death in MRP8-GFP+ mice infected with 2.5.105 CFUs of P. aExoU in presence of Sytox Blue for 10 hours. Granulocyte death was observed in infected lungs by the appearance of Sytox blue fluorescence. Pseudo colors represent vessels (gray, mTG); Granulocytes (Blue, MRP8-GFP+); Dead cells (Yellow, Sytox blue). Scale bar: 20µm.
Movie S2. Intravital microscopy visualization of granulocyte death in MRP8-GFP+ mice infected with 2.5.105 CFUs of P. aExoU or P. aExoUS142A in presence of Sytox Blue for 10 hours. Granulocyte death was observed in infected lungs by the appearance of Sytox blue fluorescence. Pseudo colors represent vessels (gray, mTG); Granulocytes (Blue, MRP8-GFP+); Dead cells (Yellow, Sytox blue). Scale bar: 20µm.
ACKNOWLEDGEMENTS
Nlrc4−/− mice were provided by Clare E. Bryant (Man et al., 2014b) and generated by Millenium Pharmaceutical, GsdmD−/− mice (Demarco et al., 2020) came from P. Broz (Univ of Lausanne, Switzerland), Casp11−/− and Casp1−/−/ Casp11−/− came from B. Py (ENS Lyon, France) and Junying Yuan (Harvard Med School, Boston, USA) (Li et al., 1995; Wang et al., 1998). Virginie Petrilli (ENS Lyon, France) provided Nlrp3−/− mice that were generated by Fabio Martinon (Martinon et al., 2006). Christine T N Pham (Washington University, Seattle, USA) generated and provided the NE−/−/CatG−/−/Pr3−/− mice (Yan et al., 2016), Thomas Henry (CIRI, Lyon, France) provided ASC−/− and AIM2−/− mice upon agreement with Genentech (San Francisco, Roche, USA) and. ASC-Citrine (#030744) and Pad4−/− (#030315) mice came from Jaxson Laboratory (USA) and were generated by Douglas T Golenbock (University of Massachusetts Medical School, USA) and Kerri Mowen (The Scripps Research Institute, USA) respectively. MRP8Cre/Casp1flox mice are provided by Natalie Winter (INRAE Tours Nouzilly, France) and were generated by crossing MRP8Cre (Jackson # 021614) mice with Caspase1flox mice generated by Mohamed Lamkanfi (Univ. of Ghent, Belgium)(Van Gorp et al., 2016). MRP8CreGFP and mTmG mice were obtained from Jackson laboratories and generated respectively by Emmanuelle Passegue (UCSF, USA) and Liqun Luo, (Stanford University, USA). Pseudomonas aeruginosa strains were a kind gift of Ina Attrée (CNRS, Grenoble, France) and Julien Buyck (Univ. of Poitiers, France). Authors also acknowledge the animal facility and Cytometry/microscopy platforms of the INFINITY, CBI and IPBS institutes and particularly Valerie Duplan-Eche for Imagestream acquisition and analysis. This project was funded by grants from the Fonds de Recherche en Santé Respiratoire - Fondation du Souffle to EL, ATIP-Avenir program, FRM “Amorçage Jeunes Equipes” (AJE20151034460) and the ERC (StG INFLAME 804249) to EM, the European of Clinical Microbiology and Infectious Diseases (ESCMID, 2020) to RP, Invivogen-Society CIFRE collaborative PhD (to MP) and post-doctoral fellowships (to RP), from Mali and Campus France cooperative agencies to SB and of a NIH grant AR073752 to C.TN Pham. The project benefited of support from Labex/Investissements d’Avenir (French Ministry of research) and the Fondation Bettencourt to the IPBS Animal facility infrastructure development and use.