The mammalian inhibitor of apoptosis proteins (IAPs) are key regulators of cell death and inflammation. A major function of IAPs is to block the formation of a cell death–inducing complex, termed the ripoptosome, which can trigger caspase-8–dependent apoptosis or caspase-independent necroptosis. Recent studies report that upon TLR4 or TNF receptor 1 (TNFR1) signaling in macrophages, the ripoptosome can also induce NLRP3 inflammasome formation and IL-1β maturation. Whether neutrophils have the capacity to assemble a ripoptosome to induce cell death and inflammasome activation during TLR4 and TNFR1 signaling is unclear. In this study, we demonstrate that murine neutrophils can signal via TNFR1-driven ripoptosome assembly to induce both cell death and IL-1β maturation. However, unlike macrophages, neutrophils suppress TLR4-dependent cell death and NLRP3 inflammasome activation during IAP inhibition via deficiencies in the CD14/TRIF arm of TLR4 signaling.
The mammalian inhibitor of apoptosis proteins (IAPs), cellular IAP (cIAP)1, cIAP2, and X-linked IAP (XIAP), are E3 ubiquitin ligases that inhibit caspases, apoptosis, and necroptosis. However, it is increasingly clear that IAPs are also critical regulators of proinflammatory signaling pathways (1).
Two studies reported that a loss of IAPs induces the formation of the ripoptosome death-inducing complex, comprising the core components RIPK1, FADD, and caspase-8 (2, 3). IAPs are degraded and/or inhibited upon exposure to genotoxic stress, proapoptotic stimuli (4), TNF superfamily receptor ligation (5, 6), microbial infection (7), or the application of IAP antagonist compounds (8). Hereditary mutation can also reduce IAP expression, and in some cases, the clinical manifestation in humans resembles that of cryopyrin-associated periodic fever syndromes, which are driven by IL-1β (9). Consistent with this, LPS-TLR4 signaling in IAP-depleted macrophages induces ripoptosome-dependent caspase-8 activation, caspase-8–dependent IL-1β maturation, and NLRP3 inflammasome signaling (6, 10, 11). Similarly, TNF receptor 1 (TNFR1) ligation in XIAP-deficient dendritic cells activates caspase-8–dependent IL-1β maturation and NLRP3 activation (12). Notably, TLR4-induced NLRP3 activation upon IAP loss requires signaling by the TLR3/4 adaptor TRIF, which recruits RIPK3 and RIPK1 through RIP homotypic interaction motif (RHIM) interactions to activate the ripoptosome (6, 10). When caspase-8 activity is suppressed, RIPK3 in turn activates the mixed lineage kinase domain–like protein (MLKL) to trigger necroptosis and drive NLRP3 inflammasome assembly (13).
Neutrophils often exhibit specialized innate immune signaling pathways to prolong their otherwise short lifespan (14, 15). In this study, we investigated whether the ripoptosome induces neutrophils to undergo TNFR1- or TLR4-dependent apoptosis, necroptosis, or NLRP3 inflammasome assembly. We demonstrate that deficiencies in the CD14/TRIF arm of the TLR4 signaling pathway enable LPS-stimulated neutrophils to resist acute cell death and inflammasome activation upon IAP depletion. However, at later time points, IAP depletion sensitized neutrophils to LPS-induced autocrine TNF signaling, leading to caspase-8–dependent apoptosis and IL-1β activation. If caspase-8 activity is blocked, IFN-γ priming rendered neutrophils sensitive to TNFR1-dependent necroptosis and subsequent NLRP3 signaling. These findings highlight cell type specificity in the regulation of cell death and inflammasome activation.
Materials and Methods
Cell isolation, stimulation, and analysis
Mature bone marrow neutrophils and bone marrow–derived macrophages were prepared from murine bone marrow as previously described (10, 15, 20). The purity of neutrophil isolation was 98–100%. Cells were stimulated with 100 ng/ml ultrapure Escherichia coli K12 LPS (InvivoGen), 100 ng/ml human recombinant TNF (PeproTech), 5 μM nigericin (Sigma-Aldrich), 10 μM Q-VD-Oph (QVD; R&D Systems), 1 μM MCC950, 80 μM dynasore (Sigma-Aldrich), 100 ng/ml SuperFasLigand (Enzo Life Sciences), or compound A (Cp.A) (0.25–1 μM). Each batch of Cp.A was titrated and used at the most potent dose. Cell death was assessed by propidium iodide uptake (10) or lactate dehydrogenase (LDH) release into the culture supernatant (Promega). Quantitative RT-PCR relative to the reference gene Hprt was performed as described (15).
Statistical analyses were performed using the nonparametric Mann–Whitney t test. Data were considered significant when p values were ≤0.05.
Results and Discussion
IAP inhibition induces neutrophil cell death in the presence of exogenous TNF
To mimic IAP loss during physiological conditions, we treated mature bone marrow neutrophils with the bivalent SMAC mimetic Cp.A to degrade or inhibit cIAP1, cIAP2, and XIAP (8). Cp.A triggered rapid cIAP1 degradation in neutrophils at doses as low as 125 nM (Fig. 1A) and induced modest but significant neutrophil death (Fig. 1B), whereas Cp.A elicited robust macrophage death (Fig. 1C) by promoting autocrine TNF-TNFR1 signaling and the subsequent caspase-8 activation (6, 8). The modest induction of neutrophil death by Cp.A suggested that Cp.A induces insufficient autocrine TNF signaling for robust cell death induction at this time point. Indeed, Cp.A failed to significantly induce TNF secretion compared with LPS treatment (Fig. 1D), and the addition of exogenous TNF sensitized neutrophils to Cp.A-mediated killing (Fig. 1E, Supplemental Fig. 1). Thus, although Cp.A does not itself induce sufficient TNF production from neutrophils to induce robust cell death, IAP depletion licenses neutrophil apoptosis to exogenous TNF, demonstrating that neutrophils can signal cell death via the ripoptosome.
TLR4 signaling does not trigger caspase-8–dependent cell death and NLRP3 activation when IAPs are inhibited in neutrophils
TLR4 signaling in IAP-depleted macrophages triggers RIPK3-dependent ripoptosome-mediated cell death, IL-1β processing, and NLRP3 activation (6, 10, 11). Having established that neutrophils have the capacity to assemble a ripoptosome (Fig. 1B, 1E), we next examined whether IAPs modulate ripoptosome-dependent cell death and NLRP3 activation in LPS-stimulated neutrophils. Cells were primed with LPS for 3 h to induce pro–IL-1β and NLRP3 expression, and IAPs were then inhibited with Cp.A to induce ripoptosome assembly. As expected, LPS stimulation suppressed basal neutrophil apoptosis but, surprisingly, did not enhance Cp.A-mediated neutrophil death (Fig. 2A). In contrast, and consistent with previous reports (10, 11), LPS stimulation enhanced Cp.A-mediated cell death in macrophages (Fig. 2B). Because LPS/Cp.A treatment did not elicit neutrophil death (Fig. 2A), we hypothesized that IAP inhibition would be similarly unable to drive caspase-8 and NLRP3 activation in LPS-primed neutrophils. We investigated the cleavage status of caspase-8 and the caspase-8 substrate, Bid, in cells exposed to LPS/Cp.A. Full-length caspase-8 and Bid levels were unaffected by neutrophil treatment with Cp.A, regardless of LPS priming (Fig. 2C), whereas Fas ligand (FasL) triggered robust caspase-8 and Bid processing (Fig. 2C). In contrast, in macrophages, Cp.A readily triggered caspase-8 activation and the subsequent cleavage of Bid, similar to FasL stimulation (Fig. 2C).
Next, we assessed inflammasome activation by examining ASC polymerization and caspase-1 and IL-1β cleavage in cells exposed to LPS/Cp.A. The NLRP3 agonist nigericin stimulated ASC polymerization, caspase-1 cleavage, and IL-1β maturation in LPS-primed cells, whereas Cp.A did not trigger any hallmarks of inflammasome activation in LPS-primed or -unprimed neutrophils but did so in macrophages (Fig. 2D–F). As anticipated, given the dual pathways (caspase-8 versus caspase-8/NLRP3–dependent caspase-1) mediating IL-1β cleavage downstream of the ripoptosome (10, 11), caspase-1 deficiency (Ice−/−) only partially suppressed Cp.A-induced IL-1β secretion from LPS-primed macrophages (Fig. 2E, 2F). In all, the failure of LPS/Cp.A treatment to induce cell death, caspase-8 processing, ASC polymerization, and IL-1β cleavage in neutrophils indicates that TLR4 ligation cannot signal ripoptosome-induced cell death or NLRP3 activation in these cells.
Neutrophils appear unable to robustly signal via CD14-TRIF
In the absence of IAPs, LPS triggers RIPK3-dependent cell death and NLRP3 activation in macrophages by using the endosomal TLR4 signaling adaptor TRIF, which recruits and activates RIPK3 through RHIM interactions and so is required for ripoptosome-dependent cell death induction (10). We thus queried the capacity of neutrophils to induce TRIF signaling. We first examined levels of the LPS coreceptor CD14, which is required for the internalization of the TLR4-LPS complex into endosomes to initiate TRIF signaling (21), and resultant IFN regulatory factor (IRF)-3 phosphorylation and IFN-β production. CD14 was expressed on the surface of murine macrophages but was undetectable on neutrophils (Fig. 3A). Correspondingly, LPS induced IRF3 phosphorylation and IFN-β induction in macrophages but not in neutrophils (Fig. 3B, 3C). Despite the lack of TRIF-dependent signaling, neutrophils displayed robust TRIF-independent outputs, including LPS-induced NF-κB p65 phosphorylation (Fig. 3B) and TNF production (Fig. 1D).
We next sought to demonstrate that neutrophil CD14 deficiency blocks LPS/Cp.A–induced cell death and NLRP3 inflammasome activation. Because of the technical challenges associated with overexpressing CD14 in a short-lived cell type, such as primary neutrophils, we assessed the impact of blocking CD14 internalization on macrophage CD14/TLR4/TRIF/RIPK3 signaling, cell death, and IL-1β production using the dynamin inhibitor dynasore (21). In keeping with TRIF/RIPK3–mediated apoptosis and necroptosis (10), blocking CD14 internalization with dynasore significantly reduced macrophage apoptosis (LPS/Cp.A), necroptosis (LPS/Cp.A/pan-caspase inhibitor QVD that blocks caspase-8), and IL-1β secretion (Fig. 3D, 3E). The lack of CD14 expression in neutrophils, coupled with the requirement for CD14 internalization to induce LPS/Cp.A–induced RIPK3-dependent cell death and IL-1β activation, indicates that neutrophils suppress LPS-induced cell death induction and inflammasome assembly via deficiencies in the CD14/TRIF arm of TLR4 signaling.
Neutrophils undergo apoptosis and necroptosis via TNFR1
In macrophages, LPS/Cp.A induces rapid cell death and IL-1β activation in a TRIF-dependent manner (10), but at later time points, LPS-induced autocrine TNF engages TNFR1 to contribute to cell death and IL-1β production (10). Given that LPS induces TNF production from neutrophils (Fig. 1D), we investigated whether chronic LPS/Cp.A exposure elicits neutrophil death and IL-1β production via autocrine TNF-TNFR1 signaling. Although acute (5 h) LPS/Cp.A exposure was a poor inducer of neutrophil death and inflammasome activation (Fig. 2), LPS/Cp.A exposure for 16 h triggered substantial (∼50–60% above background) cell death (Fig. 4A, Supplemental Fig. 2A) and IL-1β secretion (Fig. 4B) from wild type neutrophils in a TNFR1-dependent manner (Fig. 4A, 4B). Because chronic LPS can upregulate CD14 expression in neutrophils (22), we confirmed that enhanced CD14 signaling was not responsible for increased death, as death was TRIF-independent (Supplemental Fig. 2B). We next examined whether inhibiting caspase-8 during this treatment regimen switches neutrophil apoptotic cell death to necroptosis by treating cells with the pan-caspase inhibitor QVD prior to Cp.A exposure. Surprisingly, QVD blocked chronic LPS/Cp.A–induced neutrophil cell death (Fig. 4A) and abrogated neutrophil IL-1β secretion (Fig. 4B). We considered the possibility that neutrophils resist LPS/Cp.A/QVD–mediated death because of insufficient TNFR1 signaling or expression of necroptotic death machinery. IFN-γ enhances TLR4-driven TNF production (23) and induces the expression of necroptotic drivers, RIPK3, and MLKL in macrophages (24), and so we investigated whether IFN-γ performs similar functions in neutrophils to sensitize these cells to necroptosis. IFN-γ did not induce expression of Ripk3 mRNA (Supplemental Fig. 2C) and moderately upregulated Mlkl mRNA expression in neutrophils (Supplemental Fig. 2D), but markedly enhanced LPS-stimulated TNF production from neutrophils (Fig. 4C). IFN-γ indeed sensitized wild type but not Tnfr1−/− neutrophils to LPS/Cp.A/QVD–mediated death and IL-1β secretion (Fig. 4A, 4B), and IFN-γ/LPS/Cp.A–treated wild type neutrophils cleaved caspase-8 and IL-1β (Supplemental Fig. 2E). IFN-γ did not sensitize neutrophils to TRIF-mediated cell death upon chronic LPS/Cp.A exposure (Supplemental Fig. 2B). In all, this indicates that IFN-γ sensitized neutrophils to TNF-mediated necroptosis, a conclusion also supported by the observation that exogenous TNF or FasL triggered Ripk3-dependent, necroptotic death in IFN-γ–untreated neutrophils (Supplemental Fig. 2F). Interestingly, IFN-γ/LPS/Cp.A–induced IL-1β secretion was largely insensitive to the NLRP3 inhibitor, MCC950 (Fig. 4C), suggesting that neutrophil IL-1β was predominantly processed by caspase-8 and not caspase-1 under these conditions. In macrophages, necroptotic cell death activates the NLRP3 inflammasome and drives IL-1β secretion in a cell-intrinsic manner (13). IL-1β production from neutrophils exposed to IFN-γ/LPS/Cp.A/QVD occurs via the same pathway because MCC950 suppressed IL-1β secretion from these necroptotic neutrophils (Fig. 4B). Taken together, our data indicate that when IAPs are blocked, TNFR1 signaling triggers apoptosis and IL-1β activation, whereas IFN-γ renders neutrophils sensitive to TNFR1-induced necroptosis and NLRP3 inflammasome signaling (Fig. 4D).
In this study, we found that when IAPs are blocked, primary murine neutrophils have the capacity to assemble the ripoptosome and signal cell death if exogenous TNF is supplied. Surprisingly, acute LPS exposure did not induce ripoptosome-dependent caspase-8–mediated cell death, IL-1β processing, or NLRP3 activation in neutrophils because of deficiencies in the TLR4/TRIF signaling arm. Although neutrophils were reported to express CD14 at low levels in some studies (reviewed in Ref. 25), we could not detect surface CD14 on purified mature bone marrow neutrophils. We and others detected strong MyD88- but not TRIF-dependent outputs in LPS-stimulated human and murine neutrophils (Fig. 3B) (26), suggesting that CD14 expression in these cells is insufficient for robust engagement of LPS-dependent TRIF signaling. Pyroptosis and necroptosis are important innate immune mechanisms for removing intracellular pathogens from their replicative niche. However, a key in vivo function of recruited neutrophils at a site of infection or injury is to clear infection, and rapid induction of cell death would compromise neutrophil antimicrobial mechanisms. Indeed, we previously reported that neutrophils resist caspase-1–dependent pyroptosis to prolong their lifespan and promote bacterial clearance (15). At later time points, LPS-induced TNF drives neutrophil apoptosis and concomitant IL-1β release under conditions of IAP inhibition, presumably to recruit and activate more neutrophils to the site of infection or injury. Although IAP inhibition sensitized neutrophils to TNF-mediated apoptosis, caspase inhibition by QVD did not trigger a switch to necroptosis unless cells were first stimulated with IFN-γ or exogenous TNF, indicating that circulating levels of proinflammatory cytokines dictate whether neutrophils undergo necroptosis in vivo. The finding that cell death pathways are distinctly regulated in neutrophils as compared with closely-related myeloid cells highlights the important influence of cell identity on immune signaling pathways. Such cell type–specific signaling adaptations have likely evolved to provide a coordinated inflammatory response in vivo.
We thank TetraLogic Pharmaceuticals, Prof. John Silke and Prof. Heinrich Korner for reagents, and Dr. Akshay D’Cruz for technical assistance. We apologize for omitting citations to relevant publications because of space constraints.
This work was supported by the National Health and Medical Research Council of Australia (Grant 1101405 and Fellowship 1141466 to J.E.V., Grants 1122240 and 1023297 and Fellowship 1141131 to K.S.) and the Australian Research Council (Fellowship FT130100361 to K.S.). K.W.C. was supported by an ANZ Trustee Medical Research Program, K.W.C. and D.B. were supported by The University of Queensland, B.A.C. was supported by the American Asthma Foundation and National Institutes of Health Grant 5R01HL124209-04, and M.G. was supported by the Israel Science Foundation (Grant 1416/15).
The online version of this article contains supplemental material.
K.S. holds a full-time Associate Professorial Research Fellow appointment at The University of Queensland and conducts research focused on inflammation and inflammasomes. K.S. is a coinventor on patent applications for NLRP3 inhibitors, which have been licensed to Inflazome Ltd. (Dublin, Ireland). Inflazome is developing drugs that target the NLRP3 inflammasome to address unmet clinical needs in inflammatory disease. K.S. served on the Scientific Advisory Board of Inflazome in 2016–2017. The other authors have no financial conflicts of interest.