Mitochondria and NLRP3: To die or inflame

Summary

Mitochondria play critical roles in intrinsic apoptosis and NLRP3 inflammasome activation, but how these processes are interconnected remains unclear. In this issue of Immunity, Saller et al. unveiled the complexity of NLRP3 activators, highlighting mitochondria’s roles in switching apoptosis to NLRP3 inflammasome activation.

The NLRP3 inflammasome is a cytosolic complex that recognizes pathogenic insults and cellular perturbations (1). Its activation typically requires two signals: The first priming signal, activated by pattern recognition receptors or cytokine receptors, transcriptionally upregulates inflammasome components and post-translationally modifies NLRP3 from an auto-inhibitory to a signal-competent state. The second signal involves detecting numerous damage-associated molecular patterns (DAMPs) and pathogen-associated molecular patterns (PAMPs), leading to NLRP3’s interaction with the adaptor molecule ASC and pro-caspase-1, nucleating the inflammasome. Subsequent signaling drives the cleavage of caspase-1, promoting the maturation of IL-1β, IL-18, and gasdermin D, which mediates the release of proinflammatory cytokines and pyroptotic cell death (1). In contrast to inflammasomes that respond to specific or structurally similar stimuli, NLRP3 recognizes a wide array of DAMPs or PAMPs, including ion flux, cellular stress, organelle dysfunction, and metabolic shifts (2), without evidence supporting the direct binding to these stimulators. How NLRP3 integrates such diverse and distinct signals indirectly for initiation remains elusive. Moreover, the involvement of multiple organelles, such as mitochondria, lysosomes, and trans-Golgi networks, adds complexity to NLRP3 signaling (2).

Mitochondria, as hubs of the respiratory chain and oxidative phosphorylation complexes, influence the NLRP3 inflammasome in different aspects. Mitochondrial dysfunction results in the accumulation of reactive oxygen species (ROS) and the release of oxidized mitochondrial DNA (mtDNA), both potent initiators of NLRP3 nucleation (3). Mitochondria also provide the platform for inflammasome assembly via molecules like cardiolipin, mitofusin 2, and mitochondrial antiviral-signaling protein (MAVS) (3). Additionally, as central regulators of intrinsic apoptotic cell death, mitochondria play critical roles in deciding between apoptosis and pyroptosis, which remains undefined.

In the current study (4), Saller et al. examined the interplay between pyroptosis and apoptosis during inflammasome activation. Using nigericin, a canonical NLRP3 activator, and raptinal, an inducer of intrinsic apoptosis, the authors demonstrated that simultaneous treatment of primed myeloid cells with both led to the execution of pyroptosis while suppressing apoptosis. This inhibition of apoptosis was not attributed to the NLRP3 inflammasome, as cells deficient in inflammasome components exhibited comparable apoptosis suppression. Furthermore, nigericin suppressed intrinsic apoptosis but not the caspase-8-mediated extrinsic apoptosis.

During the intrinsic apoptotic pathway, cytochrome c is released from the mitochondria to the cytoplasm, where it interacts with the adaptor protein Apaf-1 to assemble an apoptosome, subsequently activating caspase-9 (5). The authors found that nigericin dampened the cytochrome c release triggered by raptinal in a potassium (K⁺) efflux-independent manner. Isolated mitochondria from human and murine cells retained cytochrome c during nigericin and raptinal co-stimulation. Notably, even under conditions selectively permeabilizing the outer mitochondria membrane, nigericin still limited cytochrome c release. Using confocal and transmission electron microscopy, the authors observed significant rearrangement of mitochondrial cristae, the major structures for oxidative phosphorylation, in nigericin-treated cells. The crista membranes appeared to merge with the inner boundary membrane, forming an onion ring-like structure, and the frequency of the crista junction was markedly reduced. Such aberrant crista organization has been previously described in the mitochondrial contact site and cristae organizing system (MICOS) complex-deficient cells (6). In concert, raptinal no longer triggered apoptosis in IMMT-deficient cells (encoding MICOS component MIC60). This supports the model where nigericin traps cytochrome c within the crista lumen through crista junction closure, thus inhibiting intrinsic apoptosis.

Nigericin, a K⁺/H⁺ ionophore, dissipates the proton gradient across the mitochondrial membrane, uncoupling ATP synthesis from the electron transport chain (ETC) (7). The authors hypothesized a link between its uncoupling activity and apoptosis inhibition. Supporting this notion, a variety of chemically distinct K⁺/H⁺ ionophores with uncoupling properties, but not a Ca²⁺ ionophore, repressed raptinal-induced apoptosis. This aligns with the untargeted metabolomics analysis of nigericin-treated cells, which reveals disruptions in de novo pyrimidine biosynthesis and a decrease in phosphocreatine, a crucial molecule in ATP generation. Another uncoupler, FCCP mirrored the metabolic effect of nigericin, causing similar mitochondrial morphology changes, preventing cytochrome c release, and inhibiting apoptosis. Paradoxically, neither nigericin nor FCCP lowered cellular ATP levels in cancer cell lines, yet apoptosis inhibition persisted. In contrast, 2-deoxyglucose treatment had no impact on apoptosis and reduced cellular ATP as expected. This suggests that the apoptosis suppression by nigericin or FCCP is not attributed to cytoplasm ATP depletion.

Following the potential shared pathway between NLRP3 activation and apoptosis inhibition, the authors examined the capability of ionophores to activate NLRP3 inflammasome. Notably, while most ionophores activated the NLRP3 inflammasome, FCCP failed to do so, suggesting that additional ionophore effects are essential for NLRP3 activation beyond uncoupling or apoptosis inhibition. To further explore this connection, the authors employed non-ionophore ETC inhibitors. As reported previously by this group, imidazoquinoline family molecules (imiquimod and CL097), known as TLR7/8 agonists, activated the NLRP3 inflammasome by inhibiting ETC complex I and generating ROS (8). In the current study, various non-ionophore ETC inhibitors also suppressed raptinal-induced apoptosis. Oligomycin, a complex V inhibitor that preserves proton gradient and mitochondria membrane potential (instead of blocking the electron transport like other ETC inhibitors or disrupting the gradient like uncouplers), similarly inhibited apoptosis. This indicates that the ultimate ATP synthase activity rather than electron transport is the critical factor. Indeed, bypassing specific ETC complex inhibition using electron shunts restored the electron transport and apoptotic signaling but not for nigericin, since the latter altered ATP synthesis. Furthermore, both nigericin and oligomycin promoted F1F0 ATP synthase dimerization, a phenomenon associated with cristae morphology changes (9).

The authors then investigated if known NLRP3 activators could affect apoptosis inhibition. Extracellular ATP through the P2X7 channel triggers K⁺ efflux and activates the NLRP3 inflammasome (10). The authors discovered that extraneous ATP likewise suppressed raptinal-induced apoptosis in an inflammasome-independent but P2X7-dependent manner. Moreover, particulate activators, such as monosodium urate crystals (MSU), adjuvant alum, and virus infections (SARS-CoV2, MVA, and influenza virus) all interfered with raptinal-induced intrinsic apoptosis.

Of note, OXPHOS inhibitors alone did not activate the NLRP3 inflammasome, implying a requisite of additional signal(s). Building on their previous work (8), the authors combined imidazoquinolines family molecules, resiquimod, or gardiquimod, which did not trigger NLRP3 alone, with ETC inhibitors and detected a robust IL-1β response. This suggests that altered ATP production caused by OXPHOS inhibitors, while essential, is insufficient for NLRP3 activation, and resiquimond or gardiquimod provides the additional signal(s). Mechanistically, the authors noticed that the colocalization of NLRP3 aggregates with the trans-Golgi network protein TGN46 was caused by resiquimod and CL097 treatment, leading the authors to propose that the second signals might reside in the vesicles of NLRP3 aggregates. Additionally, the activity of mechanosensitive ion channel Piezo-1 may act as the second signal, as its activator Yoda-1 with ETC inhibitors activated NLRP3. However, the precise mechanisms remain unclear, and other potential second signals are yet to be identified.

Taken together, Saller et al. have elucidated that OXPHOS inhibition suppresses intrinsic apoptosis by disrupting mitochondrial cristae and restricting cytochrome c release in LPS-primed cells. While blocking mitochondrial ATP production is necessary for triggering NLRP3 inflammasome, it is not sufficient on its own. Classical OXPHOS inhibitors require additional signals for NLRP3 activation, and when combined with molecules, e.g., resiquimod, gardiquimod, or Yoda-1, these inhibitors elicit a robust IL-1β response. Based on these findings, the authors propose a “two activation signal” model, with one signal originating from mitochondrial dysfunction and an extra signal simultaneously needed for NLRP3 activation (Figure 1). This model differs from the wide-acknowledged “two signal” model for NLRP3 inflammasome activation. The authors provide the priming signal in their experiment design and emphasize the second activation signal, delineating a two-step activation process. One limitation of this model is its reliance on pharmaceutical reagents and cell lines precisely controlled under ideal conditions, which may not fully reflect natural settings. Future in vivo studies will be necessary to validate the fidelity and universality of this model. Despite these uncertainties, this model offers a unified perspective on the intricate signaling in the NLRP3 inflammasome activation step. A key unresolved issue is how mitochondrial signals are detected by cytosolic NLRP3. Also, the mechanisms underlying the second signals are undetermined and may involve other endosomal compartments, such as trans-Golgi networks or lysosomes. The switch from apoptosis to pyroptosis during inflammasome activation leads to inflammatory responses and cell death, both of which eliminate the pathogens. However, excessive pro-inflammatory cytokine release during pyroptosis is associated with severe tissue damage and worsened disease outcomes. This switch thus may cause collateral damage. Therefore, the intersection between NLRP3 and mitochondria in determining the choice of apoptosis vs, inflammation is important for establishing a balance between pathogen elimination vs. overzealous inflammation.

Figure 1.

“Two activation signals” model proposed in the current study. NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome requires two activation signals: one mitochondrial signal and an additional signal (e.g., resiquimod, gardiquimod or Yoda-1) that is not yet defined but involves. Reduced ATP production or oxidative phosphorylation (OXPHOS) inhibition accounts for the first/mitochondrial signal, triggering cristae closure and trapping cytochrome c (cyt c) within the lumen. Limited cyt c release further dampens apoptosome formation and subsequent intrinsic apoptosis while enabling pyroptosis. Apaf-1 = Apoptotic protease activating factor 1. IL-1β = interleukin-1β. ADP = Adenosine Diphosphate. ATP = Adenosine Triphosphate. I, II, III, IV, and V refer to electron transport chain complex I through V. Illustration created in Biorender.