Healthy donor blood was purchased from the National Health Service Blood and Transplant (NHSBT) (Tooting, UK). Written informed consent was obtained by NHSBT. Ethical approval for the use of human blood purchased from the NHSBT was granted by Wales research ethics committee 6 (18/WA/0176). Peripheral blood mononuclear cells (PBMCs) were isolated using Ficoll-Paque gradients (Cedarlane, Burlington, Canada) as previously described (21). Peripheral blood monocytes were isolated from PBMCs by iso-osmotic Percoll gradient centrifugation as previously described (22). Monocytes were cultured at 37°C and 5% CO₂ for 4-24 hours. Cells were cultured in RPMI 1640 media containing 5% (v/v) FBS and 100U/ml penicillin/streptomycin or Opti-MEM containing 100U/ml penicillin/streptomycin.

Cells were stimulated with 100ng/ml PAM3CSK4 (Pam3) (Axxora, Nottingham, UK), 10ng/ml LPS (Axxora), 1ng/ml Pam2CGDPKHPKSF (FSL-1) (Invivogen, San Diego, CA, USA), or 2µg/ml Resiquimod (R-848) (Enzo Life Sciences, Lausen, Switzerland) in the presence or absence of 10µM MCC950 (gifted by Dr. Matthew Cooper from University of Queensland, Australia), 10µM nigericin (Invivogen), 20mM potassium chloride (KCl) (Carl Roth, Karlsruhe, Germany), 40μM Necrostatin-1 (Nec-1) (Thermo Fisher Scientific, Waltham, MA USA), 2.5μM GSK872 (Biotechne, Abingdon, UK) or 10µM necrosulfonamide (NSA) (Biovision, CA, USA). For caspase inhibition, cells were pre-incubated for 30 min with 10µM Z-YVAD-FMK (caspase-1 inhibitor) (Calbiochem, Millipore, Hertfordshire, UK) or 1µM Z-IETD-FMK (caspase-8 inhibitor) (Calbiochem) before stimulation. Staurosporine (1µM) (Calbiochem) was used to induce caspase-8 activation. Cell viability was assessed using the CellTiter-blue Cell Viability Assay (Promega, Madison, WI, USA) according to the manufacturer’s instructions. All experiments were performed as technical triplicates for each donor from which the mean is displayed in graphs where the data has been pooled from several donors.

The Pierce LDH Cytoxicity Assay Kit (Thermo Fisher Scientific, Waltham, MA USA) was used to quantify pyroptosis according to manufacturer’s instructions. To determine LDH activity, the 680nm absorbance value was subtracted from the 490 absorbance. Relative LDH release was calculated as LDH release [%] = (compound treated LDH activity – spontaneous LDH activity)/(maximum LDH activity – spontaneous LDH activity) x 100.

Caspase-1 and caspase-8 activity were measured in cell lysates and supernatants using Caspase-Glo 1 inflammasome assay and Caspase-Glo 8 assay (Promega) according to manufacturer’s instructions.

Sandwich ELISAs were applied to measure IL-1β and tumor necrosis factor-α (TNF) in cell supernatants. IL-1β capture antibody (MAB601, R&D systems, Abingdon, UK), and biotinylated IL-1β antibody (BAF201, R&D systems), TNF capture antibody (551220, BD Pharmingen, UK), TNF biotinylated antibody (554511, BD Pharmingen) were used according to the manufacturer’s instructions. Recombinant TNF and IL-1β were purchased from PeproTech (London, UK). Streptavidin horseradish peroxidase (DY998, R&D systems) and 3,3′,5,5′-Tetramethylbenzidine microwell peroxidase substrate kit (KPL Inc., USA) were used according to the manufacturer’s instructions. Colour formation was stopped using 0.16 M sulfuric acid. Absorbance was read at 450 nm on a Biotek synergy HT Microplate reader and analysed using Gen5 software (BioTek, Winooski, VT, USA). The IL-1β ELISA detects both mature and pro-IL-1β.

RNA was extracted using a RNeasy Mini Kit (QIAGEN, Stockach, Germany) and reverse transcribed using a High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Paisley, UK) according to the manufacturer’s instructions. The cDNA was analysed using Taqman qPCR assays (Life Technologies, Carlsbad, USA) using Taqman PCR master mix 2x (Thermo Fisher Scientific, Waltham, MA, USA). The expression of IL1B (assay: Hs00174097_m1) and NLRP3 (Hs00918082_m1) were determined relative to the geometric mean of the reference genes GAPDH (Hs02758991_g1) and HPRT1 (Hs02800695_m1). Reactions were performed using an Agilent AriaMX thermocycler (Agilent Technologies, Cheshire, UK) as per manufacturer’s instructions. Agilent Aria 1.6 software (Agilent Technologies) was used for data collection and analysis. The comparative threshold cycle method (2^-ΔCt) was used for quantification of gene expression.

Cells were lysed in Triton X buffer [1% Triton-X (Sigma-Aldrich], 2mM EDTA (Fisher Scientific, Loughborough, UK), 100mM NaCl (Fisher), 30mM HEPES pH7.5 (Melford, Ipswich, UK)] or NP-40 buffer [50mM TrisHCl pH8 (Sigma-Aldrich), 150mM NaCl (Fisher), 1% IGEPAL-CA-630 (Sigma-Aldrich)] containing freshly added protease inhibitor (Sigma-Aldrich). Protein concentrations were determined using a bicinchoninic acid protein assay kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Proteins in the supernatants were precipitated in acetone at -20°C overnight, clarified at 12000xg for 10min and then resuspended in 2X SDS-PAGE loading buffer. Lysate-proteins were denatured in 5X SDS-PAGE loading buffer, separated on tris-glycine denaturing SDS-PAGE and transferred onto nitrocellulose membranes (GE Healthcare Life Sciences, Little Chalfont, UK). After transfer, total protein staining was performed using Revert™ 700 Total Protein Stain (LI-COR, Lincoln, NE, USA) according to the manufacturer’s instructions. Membranes were blocked with 5% milk in Tris-buffered saline with 0.1% Tween 20 (TBS-T) for 1h at RT and incubated overnight at 4°C with anti-IL-1β (MAB601, R&D). For GSDMD detection, membranes were blocked and incubated with anti-GSDMDC1 Antibody (64-Y) (sc-81868, Santa Cruz Biotechnology) in intercept blocking buffer (LI-COR) in TBS-T. Membranes were then incubated with m-IgGκ BP-CFL 790 Antibody (sc-516181, Santa Cruz Biotechnology) in 5% milk in TBS-T. For protein visualisation, fluorescence was measured on a LI-COR Odyssey Fc (LI-COR).

Mean, standard deviation, standard error of the mean (SEM), and statistical significance were calculated using GraphPad version 8 (GraphPad Software Inc., USA). For statistical analysis, parametric data were compared using a two-tailed one sample t-test. Non-parametric data were compared with a two-tailed one sample Wilcoxon test. For the comparison of multiple data sets relative to a control, one-way ANOVA with posthoc Dunnett’s multiple comparisons test was applied for parametric distributed data, in case of a non-parametric distribution, Kruskal-Wallis test with posthoc Dunnett’s multiple comparisons test was used. In case of multiple comparisons of non-parametric distributed datasets, a Friedman test was used. Significance is shown as *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

To investigate if the alternative NLRP3 inflammasome pathway is activated by different TLRs in primary human monocytes, cells were stimulated for 24h with Pam3 (TLR1/2), FSL-1 (TLR2/6), LPS (TLR4) and R-848 that can stimulate both TLR7 and TLR8. All TLR ligands induced IL-1β and TNF secretion ( Figure 1A ). However, addition of the K⁺ ionophore nigericin during the last 2h of stimulation to activate the canonical NLRP3 pathway, as would occur in the canonical model of NLRP3 inflammasome activation, significantly increased the level of IL-1β secreted, apart from in the R-848 stimulated cells ( Figure 1B ). Conversely, TNF release which is independent of inflammasome activation, was unaffected ( Figure 1C ) (23).

Immunoblot analysis was then performed on cell lysates and supernatants to determine if the IL-1β released was pro-IL-1β or mature IL-1β. Detecting IL-1β in the supernatant was challenging, as these experiments were conducted in Optimem a reduced serum media, to reduce the background level of protein in the supernatant. Consequently, this led to a significant reduction in the already low level of IL-1β released from monocytes; particularly for the TLR2 ligands Pam3 and FSL-1 ( Supplementary Figure S1 ). Pro-IL-1β was detected in the cell lysate following activation with all TLR ligands and could be detected in the supernatant for some of the donors ( Figure 2 , Supplementary Figure S2 ). Mature IL-1β was also visible in the supernatant of cells activated with Pam3, FSL-1, LPS and R-848. However, for Pam3 and FSL-1, mature IL-1β was not visible in the supernatant of all donors, probably due to the low level of mature IL-1β being at the limit of detection by western blotting ( Figure 2 , Supplementary Figure S2 ). In comparison, mature IL-1β was detected in the supernatants of all cells activated by TLR-ligands followed by nigericin for all donors ( Figure 2B , Supplementary Figure S2 ). These results demonstrate that secretion of mature IL-1β from primary human monocytes following TLR activation alone, is not unique to TLR4.

In the canonical NLRP3 inflammasome pathway, NLRP3 inflammasome assembly leads to activation of caspase-1 and consequent maturation of IL-1β, in addition to cleavage of GSDMD which promotes release of the mature IL-1β from the cell (4, 11, 13). To determine if NLRP3 activation was required for IL-1β release upon TLR stimulation alone, cells were treated with the selective NLRP3 inhibitor MCC950 (24). IL-1β secretion was significantly inhibited upon TLR activation in the presence of MCC950 compared to untreated cells ( Figure 3A ). TNF release was unaffected, except following stimulation with R-848 where TNF was significantly enhanced in the presence of MCC950 ( Figure 3B ). MCC950 alone did not affect the cell viability but was able to limit the reduction in viability of cells incubated with nigericin ( Supplementary Figures S3A, B ).

As NLRP3 was activated by TLR stimulation in the absence of a second signal, further experiments were performed to explore if TLR induced IL-1β release required K⁺ or Cl^- efflux, which can activate the NLRP3 inflammasome (9, 25). Incubation in media containing KCl to impede K⁺ and Cl^- efflux, had no significant effect on the release of IL-1β ( Figures 4A, B ). However, in the presence of nigericin (a K⁺ ionophore), excess KCl significantly reduced the amplification of IL-1β secretion following activation of TLR1/2, 2/6 and 4 ( Figures 4A, C ). Following TLR stimulation alone, TNF secretion was moderately enhanced in the presence of excess KCl ( Figures 4D, E ). This was also evident in the presence of nigericin where R-848 induced TNF was significantly elevated ( Figure 4F ). Although incubation with nigericin reduced cell viability as would be expected due to induction of pyroptosis, KCl did not affect cell viability compared to the relevant controls ( Supplementary Figure S3C ). Together, these data demonstrate that IL-1β is released in a NLRP3 dependent manner that did not require K⁺ efflux.

NLRP3 inflammasome activation results in cleavage and activation of caspase-1, which in turn cleaves pro-IL-1β to its mature form and GSDMD leading to pore formation in the cell membrane (4, 11, 13). Accordingly, when monocytes were stimulated with TLR ligands in the presence of the caspase-1 inhibitor Z-YVAD-FMK, IL-1β secretion was significantly suppressed ( Figure 5A ). TNF was mostly unaltered in all apart from FSL-1 activated cells where a modest but significant reduction was observed ( Figure 5B ). Incubation with Z-YVAD-FMK did not affect the cell viability ( Supplementary Figure S3D ).

These results indicated that caspase-1 was active even in the absence of a NLRP3-activating second signal, thus caspase-1 activity was measured to determine if TLR activation alone led to increased activity. Although a trend towards enhanced caspase-1 activity was observed following stimulation with all TLRs tested compared to unstimulated cells, these levels did not reach significance ( Figure 5C ). Caspase-1 has previously been reported to be constitutively active in primary human monocytes, thus unstimulated cells were next treated with the caspase-1 inhibitor Z-YVAD-FMK to confirm if there was a constitutive level of activation (26). Inhibition of caspase-1 in resting unstimulated cells led to a significant decrease in caspase-1 activity ( Figure 5D ).

In the alternative NLRP3 inflammasome pathway, caspase-8 was required for TLR4 induced NLRP3 activation (19). To investigate if caspase-8 is also required by other TLRs for the secretion of IL-1β from primary human monocytes, cells were treated with 1µM of the caspase-8 inhibitor Z-IETD-FMK 30 min before stimulation with TLR ligands. Catalytic inhibition of caspase-8 significantly reduced IL-1β release for all TLRs ( Figure 6A ). However, TNF secretion was unaffected in all except R848-activated cells where the caspase-8 inhibitor led on average to a 1.4 fold higher secretion of TNF ( Figure 6B ). Cell viability was not compromised upon caspase-8 inhibitor treatment ( Supplementary Figure S3E ). A caspase-8 activity assay was then performed to assess whether TLR stimulation triggers the activation of caspase-8. Staurosporine, a known activator of caspase-8, induced a significant increase in caspase-8 activity but, interestingly, TLR stimulation did not enhance caspase-8 activity ( Figure 6C ) (27). To evaluate whether there was a constitutive level of caspase-8 activity in monocytes, similar to that observed for caspase-1, unstimulated cells were treated with the caspase-8 inhibitor Z-IETD-FMK, which led to a significant decrease in caspase-8 activity ( Figure 6D ). These findings suggest that constitutive caspase-1 and caspase-8 activity are sufficient to facilitate IL-1β release following TLR activation of human monocytes.

In addition to caspase-8, RIPK1 and FADD are also suggested to be required upstream of NLRP3 activation within the alternative pathway in BLaER1 cells. However, in BLaER1 cells the dependence on RIPK1 could only be observed in necroptosis deficient cells, as disruption of RIPK1 led to a necroptotic release of IL-1β (19). To investigate the requirement for RIPK1 in primary human monocytes, cells were therefore stimulated with TLR ligands in the presence of the RIPK1 inhibitor Nec-1 with or without GSK872, an inhibitor of RIPK3 to prevent necroptosis ( Figure 7 ). Inhibition of RIPK1 kinase activity with Nec-1 alone did not affect IL-1β release and did not induce LDH release ( Figures 7A–D , Supplementary Figure S4 ). In addition, IL-1β secretion was not significantly affected in the presence of dual inhibition with Nec-1 and GSK872 ( Figures 7A–D ). Furthermore, no effect was observed on LPS induced caspase-1 activation by these inhibitors when used alone or in combination ( Figure 7E ).

In the canonical NLRP3 inflammasome pathway, the release of IL-1β is facilitated by the formation of GSDMD pores, which also induce pyroptosis (11, 13, 14). However, in the alternative NLRP3 pathway, TLR4-induced IL-1β is independent of pyroptosis (19). Thus, the involvement of pyroptotic cell death was assessed by measuring LDH release upon TLR stimulation. Pam3, FSL-1, LPS or R-848 alone, did not significantly increase LDH release compared to unstimulated monocytes. Although stimulation of cells with R-848 moderately increased LDH levels in some donors, this did not reach statistical significance. When incubated with nigericin, LDH release increased for all conditions ( Figure 8A ). This corresponded with cleavage of 53kDa full length GSDMD (GSDMD-FL) to the 31kDa pore-forming N-terminus (GSDMD-NT) for all cells stimulated with TLR ligands in the presence of nigericin. In cells stimulated in the absence of nigericin, only R-848 led to detectable GSDMD cleavage ( Figure 8B , Supplementary Figure S5 ). This suggests that TLR induced IL-1β secretion is independent of pyroptosis.

As R-848, the most potent TLR ligand used in this study, induced GSDMD cleavage, it was possible that the other ligands may also cleave GSDMD but to a lower level than detected by immunoblotting. Thus, an alternative approach was explored using necrosulfonamide (NSA), which directly binds to GSDMD and inhibits the p30-GSDMD oligomerization in the membrane (28). NSA significantly suppressed IL-1β release for all TLRs tested in both the presence or absence of nigericin ( Figures 9A–C ). Conversely, secretion of TNF from TLR stimulated monocytes was not significantly affected but did show a small increase in cells incubated with nigericin ( Figures 9D–F ). To ensure that these results were not due to an off-target effect of NSA on the induction of pro-IL-1β or NLRP3 upregulation, the gene expression of IL1B and NLRP3 were measured following TLR1/2 and TLR4 activation. Both genes were upregulated following TLR stimulation, but this was not affected by NSA ( Supplementary Figure S6 ). Cell viability of TLR activated monocytes was also unaffected by NSA and the reduction in cell viability in cells incubated with nigericin was partially reversed in the presence of NSA ( Supplementary Figure S3F ). These data suggest that the release of TLR-induced IL-1β may in part be facilitated by GSDMD pores.