The missing link: how the inflammasome senses oxidative stress

The inflammasome is a multiprotein complex that controls the activation of caspase-1 in the innate immune system. Much research interest has focused on the signaling pathways for inflammasome activation. In a recent study published in Nature Immunology, Zhou et al.1 found that thioredoxin (TRX)-interacting protein (TXNIP) can bind to NLRP3, the key inflammasome component in a reactive oxygen species (ROS)-responsive manner, thus elucidating the link between ROS production and inflammasome activation. The NOD-like receptors (NLRs) are a large family of cytosolic pattern recognition receptors (34 members in mice, 23 in human), which are able to recognize various pathogen-associated molecular patterns and danger-associated molecular patterns and thereby initiate innate immune response toward invading pathogens and cellular damage.2,3 Several members of this family can behave as central components of inflammasomes to regulate the activity of caspase-1, such as NLRP1, NLRP3 and NLRC4.3 The best characterized inflammasome is that consisting of NLRP3, ASC and caspase-1. NLRP3 contains a C-terminal leucine-rich repeat domain (LRR), a central nucleotide-binding and oligomerization domain (NACHT) and an N-terminal PYRIN domain. While LRR and NACHT domains are thought to be involved in ligand sensing and autoregulation respectively,2 the PYRIN domain of NLRP3 recruits the adaptor protein ASC, which subsequently associates with caspase-1. On assembly of the intact NLRP3 inflammasome, caspase-1 becomes activated and produces the mature and secreted forms of the pro-inflammatory cytokines interleukin IL-1β and IL-18 (Figure 1). Thus far, a broad range of stimuli have been found to activate the NLRP3 inflammasome. These include infectious microorganisms such as Sendai virus, Influenza virus, adenovirus, Candida albicans, Staphylococcus aureus, Listeria monocytogenes and Shigella flexneri; microbial components such as muramyl dipeptide and bacterial pore-forming toxins; host-derived stress signals such as extracellular ATP, monosodium urate (MSU) and amyloid-β; as well as crystalline and particulate substances from the environment such as silica, asbestos and aluminum hydroxide.2,3,4 However, the molecular mechanism by which these stimuli trigger the assembly and activation of the NLRP3 inflammasome is still poorly understood. Given the structural and chemical divergence of these stimuli, it is unlikely that there is a direct ligand–receptor interaction as proposed for most Toll-like receptors. Consistent with this, no study has shown that any of these stimuli directly binds to NLRP3. Therefore, it is possible that these diverse stimuli may elicit a common cellular signal that is recognized by the inflammasome. Currently, three signaling pathways have been proposed (Figure 1). First, multiple studies have shown that potassium (K+) efflux is necessary, although not sufficient, to drive the activation of the NLRP3 inflammasome.5 When K+ efflux is prevented experimentally by equilibrating intra- and extracellular-K+ in the cell culture media, activation of the NLRP3 inflammasome is abolished in response to almost all known activators, including extracellular ATP, bacterial pore-forming toxins and crystalline materials.2,3,5 Second, most crystals and particulates (alum, silica, amyloid-β and so on) are found to engage the phagocytic pathway to activate the NLRP3 inflammasome. In this model, uptake of these crystals causes the disruption of the phagolysosome acidic compartment and subsequent release of cathepsin B; inhibition of this process has been proven to block the inflammasome activation.4,6 In the third model, the generation of ROS is proposed to be critical for activation of the inflammasome in response to a number of stimuli. Supporting this, pharmacological inhibitors of NADPH oxidase or siRNA-mediated knockdown of the p22phox subunit of NADPH oxidase have been shown to diminish IL-1β production induced by ATP, MSU, asbestos and silica.7,8 Opposing this, however, silica, MSU or ATP-induced inflammasome activation remains unaffected in macrophages derived from mice deficient in the gp91phox subunit of NADPH oxidase.6,9 Furthermore, Meissner et al.9 recently found that increased ROS production in superoxide dismutase-1 (SOD-1)-deficient macrophages inhibited, rather than promoted, caspase-1 activation. On the basis of these observations, the exact role of ROS in inflammasome activation is still elusive. The recent study by Zhou et al.1 identified TXNIP as a link between ROS and inflammasome activation. It was known previously that TXNIP binds to TRX—an ROS detoxifying protein, and negatively regulates its reductase activity.10 Taking biochemical approaches, Zhou et al.1 first found that TXNIP was able to specifically interact with the LRR and NACHT domains of NLRP3 in vitro. Then, immunoprecipitation experiments revealed that TXNIP interacted with TRX in the steady state; on the addition of inflammasome activators (MSU, H2O2 and R-837) to THP-1 cells, ROS was produced, causing TXNIP to dissociate from TRX and bind to NLRP3 (Figure 1). Then, as a result of the NLRP3 inflammasome activation, mature IL-1β was secreted with corresponding kinetics. Zhou et al.1 determined that TXNIP was essential for the activation of the NLRP3 inflammasome in response to ATP, MSU and R-837 based on four pieces of evidence obtained from both in vitro and in vivo approaches: (1) reduced IL-1β production from THP-1 cells in which TXNIP was knocked down by siRNA, (2) reduced IL-1β production from macrophages derived from Txnip−/− mice, (3) increased IL-1β production from THP-1 cells in which TXNIP was overexpressed or TRX was knocked down and (4) reduced neutrophil influx in Txnip−/− mice in the MSU-induced peritonitis model.1 These observations suggest that TXNIP may behave as the direct ligand of NLRP3 to trigger inflammasome activation in the ROS-dependent pathway. On the basis of this scenario, Zhou et al.1 further investigated whether the known physiological functions of TXNIP can be ascribed to the involvement of NLRP3 inflammasome. Specifically, they showed that high glucose-induced IL-1β secretion from islets was dependent on the ROS–TXNIP–NLRP3 inflammasome activation pathway ex vivo.1 Because most previous studies have been focused on the role of the NLRP3 inflammasome in immune cells with only occasional studies looking at skin keratinocytes, it would be exciting to expand our knowledge of inflammasome function to another cell type. In this case of the islets, however, because the islet preparation contains β cells as well as resident immune cells, it is difficult to distinguish at this stage whether this glucose–TXNIP–NLPR3 inflammasome axis is intrinsic to β cells or to infiltrating immune cells. To actually prove the former situation, increased purity of the β-cell culture is required to show the expression of NLRP3 components and glucose-induced IL-1β production in this cell type. Alternatively, it has to be shown that the immune cells residing in islets (macrophages and dendritic cells) cannot respond to high glucose to produce IL-1β. Txnip−/− mice were previously found to exhibit hypoglycemia at baseline and better glucose tolerance than wild-type mice when kept on a regular chow diet.10 In comparison, Zhou et al.1 observed that chow-fed Nlrp3−/− mice had unaltered glucose baseline, whereas high-fat-fed Nlrp3−/− mice exhibited better glucose tolerance than wild-type mice. Because these parameters were measured on different experimental settings, it is difficult to make a direct comparison between the phenotype of Txnip−/− and Nlrp3−/− mice. Although the similar tendency observed in Txnip−/− and Nlrp3−/− mice supports the possibility that these molecules are components of the same signaling pathway, more definitive evidence has yet to be established to connect the ROS–TXNIP–NLRP3 inflammasome activation pathway to the pathogenesis of type II diabetes. It is worth mentioning that the three pathways proposed above may not be mutually exclusive. While phagocytosis of particulate matters in macrophages is well known to result in the generation of ROS, several studies have also linked K+ efflux to ROS production.3 Therefore, it is probable that all these signaling events triggered by an inflammasome activator contribute to the release, modification or recognition of one single ligand that directly binds to NLRP3.11 Future research is necessary to clarify the relative contribution of these pathways to inflammasome activation under different circumstances, and it will be interesting to determine whether TXNIP is generally required for inflammasome activation by stimuli other than MSU, ATP and R-837. Also, to further address whether the ligand–receptor interaction between TXNIP and NLRP3 per se is essential for ROS-induced inflammasome activation, the assembly and activation of the inflammasome must be investigated with TXNIP mutants that cannot bind to TRX and TXNIP mutants that cannot bind to NLRP3. Given the important roles of the NLRP3 inflammasome in the innate host defense against invading pathogens and in autoinflammatory disorders, the identification of TXNIP as a crucial signaling molecule in the NLRP3 inflammasome activation pathway is an important advance and, of note, may provide potential drug targets. The authors declare no conflict of interest. Proposed signaling pathways for the NLRP3 inflammasome activation, with the molecular mechanism described by Zhou et al.1 highlighted.