The two pore potassium channel THIK‐1 regulates NLRP3 inflammasome activation

Abstract The NLRP3 (NLR family, pyrin domain containing 3) inflammasome is a multi‐protein complex responsible for the activation of caspase‐1 and the subsequent cleavage and activation of the potent proinflammatory cytokines IL‐1β and IL‐18, and pyroptotic cell death. NLRP3 is implicated as a driver of inflammation in a range of disorders including neurodegenerative diseases, type 2 diabetes, and atherosclerosis. A commonly reported mechanism contributing to NLRP3 inflammasome activation is potassium ion (K+) efflux across the plasma membrane. Identification of K+ channels involved in NLRP3 activation remains incomplete. Here, we investigated the role of the K+ channel THIK‐1 in NLRP3 activation. Both pharmacological inhibitors and cells from THIK‐1 knockout (KO) mice were used to assess THIK‐1 contribution to macrophage NLRP3 activation in vitro. Pharmacological inhibition of THIK‐1 inhibited caspase‐1 activation and IL‐1β release from mouse bone‐marrow‐derived macrophages (BMDMs), mixed glia, and microglia in response to NLRP3 agonists. Similarly, BMDMs and microglia from THIK‐1 KO mice had reduced NLRP3‐dependent IL‐1β release in response to P2X7 receptor activation with ATP. Overall, these data suggest that THIK‐1 is a regulator of NLRP3 inflammasome activation in response to ATP and identify THIK‐1 as a potential therapeutic target for inflammatory disease.


| INTRODUCTION
Inflammation is a response of the immune system to harmful stimuli such as pathogens, damaged cells, and toxic compounds (Medzhitov, 2010), and functions to remove damaging stimuli and initiate healing (Ferrero-Miliani et al., 2007). Acute inflammatory responses initiate molecular and cellular pathways to minimize injury or infection to restore tissue homeostasis. However, uncontrolled inflammation can become chronic, contributing to a variety of diseases (Lamkanfi & Dixit, 2012). Understanding the mechanisms regulating the inflammatory response is therefore critical to identify new therapeutic targets for limiting damaging inflammation.
Activation of the protease caspase-1 by inflammasomes is a proinflammatory signaling pathway of the innate immune system (Broz & Dixit, 2016). Upon activation, caspase-1 drives the processing of proinflammatory cytokine precursors pro-interleukin (IL)-1β and pro-IL-18 into their mature biologically active forms which subsequently drive inflammatory responses (Dinarello et al., 2012). Of the inflammasomes identified, the inflammasome formed by the sensor NLRP3 (NLR family, pyrin domain containing 3) is the most researched due to its association with a number of inherited and acquired inflammatory diseases (Hoffman et al., 2001;Wen et al., 2012). NLRP3-dependent inflammation is suggested to drive neuroinflammation in neurodegenerative conditions including Alzheimer's and Parkinson's disease in addition to peripheral diseases such as atherosclerosis, type-2 diabetes, and others (Wang et al., 2020). Accordingly, there is a great interest in understanding the mechanisms regulating NLRP3 to identify potential therapeutic targets for limiting damaging inflammation.
NLRP3-inflammasome activation is induced by pathogen or damage associated molecular patterns (PAMPs and DAMPs respectively) (Broz & Dixit, 2016). Due to the structural diversity of NLRP3 activators, it is thought unlikely they directly bind to, and activate, NLRP3. Activators of NLRP3 are reported to indirectly induce NLRP3 activation via altering cellular homeostasis and inducing organelle dysfunction which is in turn sensed by NLRP3 (Seoane et al., 2020). Following activation, NLRP3 nucleates the oligomerization of the adaptor protein apoptosisassociated speck-like protein containing a caspase recruitment domain (ASC) (Lu et al., 2014). ASC itself then undergoes oligomerization into inflammasome specks which leads to caspase-1 recruitment and activation. Caspase-1 cleaves pro-IL-1β and pro-IL-18 to active released forms (Dick et al., 2016;Lu et al., 2014). In addition to activation of pro-inflammatory cytokines, active caspase-1 also triggers a pro-inflammatory form of programmed cell death termed pyroptosis via the cleavage of gasdermin D (GSDMD) (Shi et al., 2015). Following cleavage, GSDMD forms pores in the cell membrane which act as a conduit for IL-1β release (Heilig et al., 2018), and which also leads to cell swelling and nerve injury-induced protein 1 (NINJ1) dependent cell membrane rupture resulting in the release of pro-inflammatory intracellular contents (Ding et al., 2016;Kayagaki et al., 2021). Canonical activation of NLRP3 in vitro requires a two-step activation process. The first "priming" step can be induced by stimulation of Toll-like receptors which drives the expression of pro-IL-1β and NLRP3 (Hornung & Latz, 2010). The second step is NLRP3 inflammasome activation, which can be induced by a range of structurally unrelated stimuli such as the K + ionophore nigericin, extracellular ATP, and crystalline/particulate matter such as silica (Hornung et al., 2008;Mariathasan et al., 2006). Several studies have proposed mechanisms to explain how such a diverse range of stimuli converge on NLRP3 activation. One proposed mechanism is a decrease in intracellular K + (Pétrilli et al., 2007) which is suggested to be important for multiple NLRP3 activating stimuli (Muñoz-Planillo et al., 2013). However, the mechanism by which K + efflux regulates NLRP3 activation remains unclear. Recent studies have also shown small molecules such as imiquimod and the imidazoquinoline derivative CL097 trigger NLRP3 activation independently of K + efflux (Groß et al., 2016). K + efflux is therefore an important but not universal driver of canonical NLRP3 activation. Studies also suggest an involvement of Cl À channels in NLRP3 activation (Kelley et al., 2019). We have previously shown Cl À efflux can regulate NLRP3-dependent ASC oligomerization (Green et al., 2018). An alternative K + independent mechanism of NLRP3 activation is also described in human monocytes (Gaidt et al., 2016). K + channels regulate an array of cellular and immune responses including immune cell proliferation, cell volume regulation, cytokine production and surveillance (Bittner et al., 2010;Bobak et al., 2011;Madry et al., 2018;Meuth et al., 2008). Members of the two-pore domain K + (K2P) channel family in particular, have recently been implicated with NLRP3 inflammasome activation (Di et al., 2018;Madry et al., 2018). The Two-pore domain Weak Inwardly rectifying K + channel 2 (TWIK2) has been suggested to facilitate ATP-induced K + efflux and subsequent NLRP3 activation in macrophages (Di et al., 2018). Furthermore, a recent study identified the K + channel Tandem Pore Domain Halothane-Inhibited Potassium Channel 1 (THIK-1) regulates IL-1β release from hippocampal slices in response to ATP (Madry et al., 2018), suggesting that it may be important for NLRP3 activation. Using pharmacological and genetic approaches we identified the K + channel THIK-1 as a specific regulator of NLRP3 inflammasome activation in response to extracellular ATP, but not to other NLRP3 activating stimuli. These results suggest THIK-1 may represent a potential therapeutic target in limiting damaging NLRP3 inflammasome activation in inflammatory disease where ATP signaling is a component.

| Primary murine mixed glial culture preparation
Murine mixed glial cells were prepared from the brains of C57BL/6 male and female 2-4-day old mice that were sacrificed by cervical dislocation under S1. All experimental procedures were performed under Home Office UK project license in accordance with the Animals (Scientific Procedures) Act UK 1986 and approved by the University of Manchester AWERB (Animal Welfare and Ethical Review Body).
The brains were isolated, followed by dissection of hemispheres and removal of meninges as previously described

| Primary murine adult microglia preparation
WT and THIK-1 KO C57BL/6 male and female mice aged 6-10 weeks were sacrificed by cervical dislocation under S1. All experimental pro-

| Western blotting
IL-1β, caspase-1, and GSDMD processing in addition to NLRP3 and IL-1β protein expression were determined by western blotting. Both cell supernatant and cell lysates were collected together and precipitated in deoxycholate containing 20% trichloroacetic acid (Fisher) and washed with acetone followed by air drying at room temperature before concentration in 2x Laemmlii buffer. All samples were separated using Tris-glycine SDS/PAGE and then transferred using a semi- Amersham ECL detection reagent (GE Healthcare). Images of western blots were captured digitally using a G-Box Chemi XX6 (Syngene).

| Lactate dehydrogenase assay
Cell death was quantified in pBMDMs, iBMDMs and mixed glia following treatment by measuring the release of the enzyme lactate dehydrogenase (LDH). This was achieved using the Cytotox-96 assay (Promega) according to the manufacturer's instructions.

| ASC speck imaging
Real-time ASC speck assays were performed using iBMDMs stably expressing ASC-mCherry (ASC-mCherry iBMDMs) (Daniels et al., 2016). ASC-mCherry iBMDMs were seeded out overnight into 96 well plates at a density of 0.75 Â 10 6 cells ml À1 followed by priming for 3 h with 1 μg ml À1 LPS. To prevent pyroptosis and loss of ASC specks iBMDMs were pre-treated with the pan-caspase inhibitor 2.11 | YO-PRO-1 P2X7 assay P2X7 receptor-dependent membrane permeability was determined using the YO-PRO-1 fluorescent dye (Rat et al., 2017). iBMDMs were seeded out overnight into 96 well plates at a density of 0.75 Â 10 6 cells ml À1 followed by priming for 4 h with 1 μg ml À1 LPS. After priming, cells were

| Statistical analysis
Data are presented as the mean ± standard error of the mean (SEM).
Levels of significance accepted were *p < .05, **p < .01, ***p < .001, ****p < .0001. Statistical significance was calculated using GraphPad Prism version 9.2.0. Data with multiple groups were analyzed using a one-way ANOVA followed by Dunnet's post hoc comparison. Experiments with two independent variables were analyzed using two-way ANOVA followed by Bonferroni's post hoc correct analysis.

| Potassium channels shared by cultured BMDM and microglia
In order to identify potential K + channels involved in regulating NLRP3 activation in macrophages and microglia we analyzed K + channel RNA expression in both cultured microglia and iBMDM cells from existing datasets (Hoyle et al., 2018 Figure S1), in addition to the non-specific broad K + channel inhibitor TEA. The NLRP3 inhibitor MCC950 (Coll et al., 2015) was included as a positive control. Inhibitors and concentrations used were based on previously published work (Madry et al., 2018;Nguyen et al., 2017;Paul et al., 2001;Roy et al., 2010;Schaarschmidt et al., 2009;Schmalhofer et al., 2009;Schmitz, 2005).
BMDMs were then stimulated with ATP (5 mM, 1 h) to induce NLRP3 activation via the P2X7 receptor (Mariathasan et al., 2006), or with silica (300 μg ml À1 , 4 h) to stimulate NLRP3 activation via lysosome damage (Hornung et al., 2008). The Kir2.1 inhibitor ML133 did however significantly reduce cell death in response to ATP, but not silica, which is in contrast to all of the other K + channel inhibitors used, which failed to reduce cell death in response to ATP or silica (Supplementary Figure S2A).
Having established which K + channel inhibitors attenuate IL-1β release in response to ATP and silica, we next confirmed that the effects of TEA, TPA and quinine were due to their inhibitory action on K + channels, and not an alternative "off target" mechanism. In order to test this we examined the impact of the K + channel inhibitors on NLRP3-dependent IL-1β release using the K + efflux-independent NLRP3 activator imiquimod (Groß et al., 2016). Stimulation of LPSprimed pBMDMs for 1 h with 75 μM imiquimod induced IL-1β release, which was abolished by pre-treatment with quinine ( Figure 1aiii). In contrast, TEA, TPA and other inhibitors had no effect on imiquimod-induced IL-1β release. Imiquimod-induced cell death was not significantly affected by treatment with any inhibitor tested (Supplementary Figure S2A). In addition to imiquimod, we investigated the effect of TEA, TPA and quinine on nigericin-induced NLRP3 activation. Nigericin is a K + ionophore which activates NLRP3 by facilitating K + efflux independently of K + channels (Mariathasan et al., 2006).
Stimulation of LPS-primed pBMDMs for 1 h with 10 μM nigericin induced IL-1β release which was unaffected by pre-treatment with TEA or TPA (Figure 1aiv). Pretreatment with quinine also reduced nigericininduced IL-1β release (Figure 1aiv). Nigericin-induced cell death was not significantly affected by treatment with TPA, TEA or quinine (Supplementary Figure S2A). The effect of ATP, silica, imiquimod and nigericin on IL-1β release were all NLRP3-dependent as in all cases IL-1β release was inhibited by MCC950 (Figure 1a). Taken together these data suggest that the K + channels targeted by TEA and TPA may play a role in ATP-and silica-induced activation of NLRP3. The specific effect of TPA on K + channel dependent NLRP3-mediated IL-1β release indicated a role of K2P channels in this pathway. Although only weakly, TEA also inhibits K2P channels (Lotshaw, 2007). Therefore, TEA may also inhibit NLRP3 activation via inhibition of K2P channels. In contrast, since quinine could inhibit NLRP3 activation in response to imiquimod, a K + -efflux independent activator, and nigericin, a K + ionophore, suggests quinine may be inhibiting NLRP3 through alternative mechanisms in addition to blocking K + channels.
Western blot analysis showed that ATP-induced caspase-1, IL-1β, and GSDMD processing were all inhibited by TPA and TEA (Figure 1b) further indicating that TPA and TEA are inhibitors of NLRP3 inflammasome activation. TPA inhibited both ATP-and silica-induced Caspase-1 Glo, a quantitative measure of caspase-1 activity showed that TPA and TEA inhibited caspase-1 activity in pBMDMs in response to ATP treatment (Figure 1d) further suggesting TPA and TEA as inhibitors of NLRP3 activation. Interestingly, inhibition of NLRP3 activation, caspase-1 cleavage and subsequent GSDMD cleavage with TPA and TEA in response to ATP failed to inhibit cell death.
These results suggest the cell death observed in these studies is not pyroptosis. One potential explanation for these findings is the cell death is not pyroptosis but necrosis as has been described for NLRP3 activating stimuli previously (Cullen et al., 2015).
In addition to NLRP3, other well characterized inflammasomes AIM2 and NLRC4 also drive caspase-1 cleavage and subsequent IL-1β release in response to cytosolic DNA and intracellular bacteria respec- inflammasome activation, the impact of K + channel inhibition on AIM2 and NLRC4-dependent IL-1β release was tested. LPS-primed pBMDMs were pretreated with K + channel inhibitors as described above. AIM2 or NLRC4 inflammasome activation was then stimulated to induce IL-1β release by transfecting BMDMs with poly (dA:dT) (1 μg ml À1 , 4 h) or salmonella typhimurium flagellin (1 μg ml À1 , 4 h).
Pretreatment with K + channel inhibitors failed to reduce IL-1β release or cell death in response to AIM2 or NLRC4 inflammasome activation (Supplementary Figure S3). These data suggest K + channels and specifically K2P channels selectively regulate NLRP3 inflammasome activation without impacting other inflammasomes.
Following the findings that K2P channel inhibition attenuated canonical NLRP3 activation in mouse macrophages we next aimed to determine whether K2P channel inhibition also regulated canonical NLRP3 activation in human immune cells. As observed in mouse BMDMs, stimulation of LPS-primed THP-1 monocytes with silica (300 μg mL À1 , 4 h) induced IL-1β release which was inhibited by both TEA and TPA (Supplementary Figure S4). IL-1β release was inhibited in both MCC950 treated THP-1 cells and NLRP3 KO THP-1 cells, demonstrating the response was NLRP3 dependent. These data indicate that K2P channels may regulate canonical NLRP3 activation in both murine and human immune cells. In addition to two step canonical NLRP3 activation, LPS alone can stimulate caspase-1 activation and IL-1β release in human monocytes (Perregaux et al., 1996). This mechanism has been defined as alternative NLRP3 inflammasome activation and does not require K + efflux (Gaidt et al., 2016). The impact of the K2P channel inhibition on alternative NLRP3 activation in human monocyte THP-1 and also primary human CD14+ monocytes freshly isolated from healthy donors was therefore tested.
Treatment with LPS (1 μg ml À1 , 16 h) induced alternative NLRP3 activation and subsequent IL-1β release in both THP-1 and primary human monocytes which was inhibited by both TEA and TPA (Supplementary Figure S4). Together these data show K + channel inhibitors TEA and TPA inhibit both canonical and alternative NLRP3 activation in human monocytes. The findings that both TEA and TPA inhibit K + efflux independent alternative NLRP3 activation suggests both TEA and TPA can inhibit alternative NLRP3 activation through additional mechanisms independent from preventing K + efflux via K2P channels.
We also investigated the impact of K + channel inhibition on the priming step of canonical NLRP3 inflammasome activation. LPS stimulation results in the activation of the transcription factor NF-κB, which in addition to upregulating NLRP3 and pro-IL-1β expression, also upregulates other pro inflammatory cytokines such as IL-6 and TNF (Bauernfeind et al., 2009;Liu et al., 2017). iBMDMs were pretreated with K + channel inhibitors or the NF-κB inhibitor Bay11 for 15 min prior to priming with LPS for 4 h. TEA and TPA significantly inhibited both IL-6 and TNF release in response to LPS (Figure 2ai,ii), and while TEA showed some toxicity this was not the case for TPA (Figure 2b).
By western blot we confirmed TPA and TEA inhibited LPS-induced protein expression of NLRP3 and pro-IL-1β (Figure 2c). Together these results show that TPA and TEA inhibit both NLRP3 priming and activation steps, suggesting K2P channels may also play a role in NLRP3 priming as well as activation.
3.3 | Blocking two-pore domain potassium channels enhances ASC speck formation but does not trigger caspase-1 activation We previously reported Cl À flux to be required for ASC oligomerization while caspase-1 activation is K + efflux-dependent (Green et al., 2018). Having observed that inhibition of K2P channels with TPA increased ASC speck formation in response to ATP in this study, we next wanted to determine whether TPA was enhancing speck formation via blocking K + efflux through K2P channels. We therefore investigated whether blocking K + efflux also enhanced ASC speck formation in response to ATP. To understand the effect of blocking K + efflux on inflammasome activation we performed ion substitution experiments. LPS-primed iBMDMs were incubated in solutions with normal K + and normal Cl À , high K + and Cl À free, or high K + and normal Cl À . High K + will block K + efflux as previously reported (Green et al., 2018). NLRP3 activation was then stimulated with ATP. Incubation with both high K + or high K + and Cl À free completely abolished IL-1β release in response to ATP (Figure 3bi). These data show that blocking K + efflux is sufficient to block activation of the NLRP3 inflammasome supporting our findings that K + channel inhibitors block NLRP3-inflammasome activation (Figure 1). We next aimed to investigate the impact of incubating iBMDMs in the above-mentioned isotonic salt solutions on ASC speck formation. To test the impact of these isotonic salt solutions on ASC speck formation, ASC-mCherry iBMDMs were primed with LPS and incubated with (a) normal K + , normal Cl À , (b) high K + , normal Cl À , or (c) high K + , Cl À free solutions.
NLRP3 activation was then stimulated with ATP and ASC formation analyzed in real time. Blocking both K + and Cl À efflux with high K + and high Cl À solution completely inhibited the formation of ASC-specks in response to ATP (Figure 3bii,iii). In contrast, allowing Cl À efflux but blocking K + efflux with high K + and Cl À free solution enhanced ASC speck formation in response to ATP (Figure 3bii). These findings are consistent with our previous report that Cl À efflux serves as an ASC oligomerizing signal while K + efflux is required for NLRP3 activation (Green et al., 2018). These results suggested that blocking K + efflux directly or inhibiting K2P channels inhibited NLRP3-dependent caspase-1 activation but enhanced ASC speck formation.

| Two pore domain potassium channel inhibition blocks NLRP3 activation in mixed glia and adult microglia
Having observed that TPA inhibited NLRP3 activation in iBMDMs we next sought to determine whether TPA mediated K2P inhibition could block NLRP3 in microglia, a brain resident macrophage cell population.
Initially, we investigated the impact of TPA on ATP-, silica-, nigericin-,  Figure S6). These results show F I G U R E 3 Potassium efflux is required for NLRP3 inflammasome activation but not ASC speck formation in response to ATP. (a, i) ASC speck formation measured in real time and (a, ii) ASC speck formation after 165 min of ATP stimulation from ASC-mCherry iBMDMs primed with LPS (1 μg ml À1 , 4 h) followed by pretreatment vehicle control, TPA (50 μM) or MCC950 (10 μM) for 15 min before stimulation with ATP (5 mM) (n = 6). (a, iii) ASC speck formation after 165 min from ASC-mCherry iBMDMS primed with LPS (1 μg ml À1 , 4 h) followed by treatment with vehicle control, TPA (50 μM) or MCC950 (10 μM) in the absence of ATP (n = 6). (b, i) IL-1β ELISA of the supernatant of iBMDMs primed with LPS (1 μg ml À1 , 4 h) followed by incubation in a control (145 mM NaCl/ 5 mM KCl), high K + and normal Cl À (150 mM KCl), high K + and Cl À free (150 mM KGluconate) or control and MCC950 (10 μM) solution for 15 min before stimulation with ATP (5 mM, 1 h) (n = 6). (b, ii) ASC speck formation measured in real time and (b, iii) ASC speck formation after 165 min of ATP stimulation from iBMDMs stably expressing ASC-mCherry (ASC-mCherry iBMDMs) primed with LPS (1 μg ml À1 , 4 h) followed by incubation in a control (145 mM NaCl/ 5 mM KCl), high K + and normal Cl À (150 mM KCl), high K + and Cl À free (150 mM KGluconate) or control and MCC950 (10 μM) solution for 15 min before stimulation with ATP (5 mM) (n = 4). (b, iv) representative images of ASC-mCherry iBMDMs after 165 min ATP stimulation (scale bar, 50 μm, arrows denote ASC specks). ASC speck experiments were performed in the presence of ac-YVAD-CMK (50 μM) to prevent pyroptosis and loss of ASC specks. ****p < .0001, *p < .05 determined by one-way ANOVA with Dunnett's post hoc analysis. Values shown are the mean ± SEM TPA inhibited NLRP3 activation in mixed glia potentially through its ability to block K2P channels. To confirm TPA was inhibiting NLRP3 activation in microglial cells directly we evaluated the effect of TPA on ATP and silica induced NLRP3 activation in isolated adult microglia. We found TPA inhibited IL-1β release from isolated microglia in response to both ATP and silica (Figure 4b).
These results suggest TPA sensitive channels are important for regulating NLRP3 activation within CNS resident microglia as well as peripheral macrophages.
3.5 | THIK-1 specifically regulates ATP-induced NLRP3 inflammasome activation in macrophages Although our data using TPA suggested K2P channels played a role in NLRP3 activation, K + channel modulators are known to inhibit cellular signaling pathways independently of K + channels (Akopova, 2017;Humphries & Dart, 2015). We therefore utilized genetic approaches to further determine which specific K + channel regulates NLRP3 activation.
Previous research has already shown TWIK-2 to facilitate ATP-induced K + efflux and NLRP3 activation (Di et al., 2018). We therefore investi- release in response to ATP, and had no effect on the response to silica, imiquimod, or nigericin stimulation (Figure 5a). THIK-1 KO had no effect on cell death in response to any of the stimuli tested (Supplementary Figure S7). ATP triggers NLRP3 inflammasome activation via activation of the P2X7 receptor (Solle et al., 2001). We therefore wanted to determine whether THIK-1 regulation of ATP-induced activation was occurring upstream or downstream of P2X7 receptor activation. Activation of the P2X7 receptor by ATP leads to the formation of a pore, which permeabilizes the plasma membrane to molecules up to 900 Da including the dye YO-PRO-1 (Rassendren et al., 1997;Steinberg et al., 1987).
YO-PRO-1 can be used as a readout of P2X7 receptor activation (Rat et al., 2017). WT pBMDMs were primed with LPS and stimulated with ATP in the presence of YO-PRO-1. Pre-treatment with K2P inhibitor TPA, and general K + channel inhibitor TEA, had no effect on P2X7-dependent pore formation, but the P2X7 inhibitor, oxidized ATP (oATP) inhibited YO-PRO-1 uptake (Supplementary Figure S8). These data suggest that THIK-1 regulated ATP-induced NLRP3 activation downstream of P2X7 receptor activation.
Silica induced IL-1β release was still inhibited by TPA in THIK-1 KO BMDMs, indicating that TPA can also inhibit NLRP3 independently of THIK-1 inhibition (Supplementary Figure S9), potentially by targeting TWIK-2. We then used western blotting to further characterize the effect of THIK-1 KO on NLRP3 activation. Caspase-1, IL-1β, and GSDMD processing were reduced in THIK-1 KO BMDMs in comparison to WT in response to ATP (Figure 5b). Following our findings that TPA also inhibited NLRP3 priming we sought to clarify whether these effects of THIK-1 KO were due to priming or activation. THIK-1 KO had no effect of either IL-6 or TNF release in response to LPS ( Figure 5c). Using western blot, we confirmed knocking out THIK-1 did not inhibit NLRP3 or pro-IL-1β protein expression stimulated by LPS ( Figure 5d). These data suggest THIK-1 is specifically required for ATP-induced NLRP3 activation in pBMDMs but is dispensable for activation in response to other canonical stimuli and NLRP3 priming.
Furthermore, these data suggest TPA inhibited NLRP3 activation and priming independently from inhibiting THIK-1.
It is therefore possible ATP, and its metabolites stimulate purinergic receptors which indirectly induce K + efflux and NLRP3 activation through downstream opening of K + channels. In this study, we observed K2P channel inhibition to reduce ATP-induced NLRP3 activation without impacting P2X7 receptor activity suggesting THIK-1 regulates NLRP3 activation downstream of P2X7 receptor activation. P2X7 depletion blocks ATP-dependent NLRP3 activation and is thus fundamentally required for ATP-induced NLRP3 activation (Solle et al., 2001). However, deletion of P2Y12 also reduces NLRP3 activation in response to ATP (Suzuki et al., 2020). Together, previous findings and this study suggest P2X7 and P2Y12 may both be required for ATP-induced NLRP3 activation, potentially in part through regulation of K + currents through K2P channels such as THIK-1 and TWIK-2. The findings that genetic ablation of THIK-1 or TWIK-2 inhibits NLRP3 activation in response to ATP suggests the two channels are non-redundant in their regulation of the NLRP3 inflammasome. Therefore, indicating activation of both THIK-1 and TWIK-2 is required for the activation of the NLRP3 inflammasome in response to ATP signaling.
We recently reported a mechanism of NLRP3 activation in which a Cl À -dependent step is required to drive NLRP3-dependent ASC oligomerization (Green et al., 2018). Although Cl À efflux was required to form an ASC speck, K + efflux was required to permit activation of caspase-1 (Green et al., 2018). These previous findings are supported by recent research which demonstrated low intracellular K + levels trigger a conformational change in ASC oligomer structure resulting in enhanced caspase-1 recruitment and activation (Martín-Sánchez et al., 2020). In the present study, we show that inhibition of K2P channels, non-selective K + channel inhibition, and K + efflux blockage, all inhibited caspase-1 activation without blocking the formation of NLRP3-dependent ASC specks in response to ATP. Furthermore, we show both inhibition of K + and Cl À efflux together abolished ATP-induced speck formation. These data provide further evidence dissociating the impact of Cl À and K + efflux on NLRP3 formation and activation with Cl À driving ASC oligomerization and K + efflux dependent mechanism acting potentially via K2P channels driving caspase-1 activation. These data suggest that K2P channels may be required to enable full activation of the inflammasome and caspase-1 in response to ATP.
The present study identifies THIK-1 as a regulator of NLRP3 activation in mouse macrophages and microglia in response to the canonical stimuli ATP. Consistent with previous work (Green et al., 2018) we also report that the formation of ASC specks can occur without downstream activation of caspase-1 and IL-1β cleavage. We show THIK-1 is required for NLRP3 dependent caspase-1 activation and IL-1β release in response to ATP. These results demonstrate that multiple K + channels may be involved in P2X7 dependent NLRP3 activation and highlight the therapeutic potential of targeting K + channels to limit aberrant NLRP3-induced inflammation in disease. THIK-1 represents a viable therapeutic target for limiting NLRP3 inflammasome activation in peripheral and CNS diseases.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding authors upon reasonable request.