Survival and gene expression information were downloaded from the TCGA data portal (https://portal.gdc.cancer.gov/) for SKCM (skin cutaneous melanoma) and LUSC (lung squamous cell carcinoma) datasets. We specifically downloaded preprocessed expression data using the quantile-normalized FPKM values. Next, patients were either stratified by expression of NLRP3 or the NLRP3-inflammasome-related gene set as defined by Ju et al. (24). For the NLRP3 inflammasome gene set, we defined a module score inspired by Seurat single-cell analysis (25), using each patient’s average expression of all NLRP3 inflammasome gene-set genes subtracted by the average expression of all genes. For survival analysis, we used the bottom and top 33% of patients with either NLRP3 expression or NLRP3 inflammasome module score. Statistical evaluation and Kaplan–Meier plot representation were performed with the R package survminer version 0.4.9, using the log-rank test for p-value estimation and 95% confidence interval of patient survival.

C57BL/6J mice were purchased from the Jackson Laboratory; Nlrp3-deficient (Nlrp3^-/- ) mice (on a C57BL/6 background) were kindly provided by Jürg Tschopp (Department of Biochemistry, Center of Immunity and Infection, University of Lausanne, Switzerland) (26). Disruption of the mouse Nlrp3 gene was based on the insertion of an EGFP cassette, which was accompanied by SV40 poly(A) tail, fused in frame with the ATG of exon 2. A PGK-neo selection cassette was also inserted in intron 2, which was flanked by two loxP sites and was deleted by the backcrossing of the mice, with the targeting vector, with a Cre-expressing strain (C57BL/6) resulting in a Nlrp3^-/- mouse on a C57BL/6 background. Foxp3^EGFP.KI mice (on a C57BL/6 background) were kindly provided by Alexander Rudensky (Department of Immunology, Memorial Sloan-Kettering Cancer Center, New York, USA).

All mice were maintained in the animal facility of the Biomedical Research Foundation of the Academy of Athens [BRFAA] and Institute of Molecular Biology and Biotechnology Institute [IMBB]. All procedures were in accordance with institutional guidelines and were approved by the Institutional Committee of Protocol Evaluation of the BRFAA and the Institutional Committee of Protocol Evaluation of the IMBB together with the Directorates of Agricultural Economy and Veterinary, Region of Crete, Greece (14/10/2020 Heraklion, Greece, protocol 234446). Unless indicated otherwise, all experiments used sex- and age-matched mice aged between 6 and 12 weeks.

Nlrp3-deficient mice were screened by PCR genotyping on tail genomic DNA using the following primers: 5’GCTCAGGACATACGTCTGGA3’ (forward in intron 1) and 5’TGAGGTCCACATCTTCAAGG3’ (reverse in exon 2). Nlrp3^+/+ (wild type) mice gave a product of 327 base pairs (bp), whereas Nlrp3^-/- mice did not give any PCR product. The program used for the PCR genotyping is as follows: 94°C (3 min), 30 × [94°C (30 s), 58°C (30 s), 72°C (1 min)], and 72°C (5 min).

The murine melanoma cancer cell line B16.F10 and the murine Lewis Lung carcinoma (LLC) cell line that were used for the solid tumor induction models were kindly provided by A. Eliopoulos (Medical School, National and Kapodistrian University of Athens, Athens, Greece) and were negative for Mycoplasma spp., tested by PCR.

B16.F10 and LLC cancer cells were cultured at 37°C under 5% CO₂ in RPMI-1640 (GlutaMAX™, Gibco, #61870) medium supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco, #10270), 100 U/ml penicillin–streptomycin (10,000 U/ml, Gibco, #15140), and 50 μM 2-mercaptoethanol (50 mM, Gibco, #31350). Cells were split when they were 90%–100% confluent. All experiments were performed with early passage (p2–3) cells.

Splenocytes and sorted MDSCs were obtained as described below. Mouse splenocytes and MDSCs were grown in RPMI-1640 culture medium containing 10% heat-inactivated FBS, 100 U/ml penicillin–streptomycin, and 50 μM 2-mercaptoethanol. The stimuli at the cultures were added where indicated, as mentioned below.

The transplantation of solid tumors in the tumor models was performed as described previously (27). Briefly, mice were implanted subcutaneously, at the back, with 3 × 10⁵ B16.F10 melanoma or LLC cells (viability assessed by Trypan blue exclusion). Tumor volume was monitored during the days indicated in the legends of corresponding curves. The tumor growth was monitored by measurement of two perpendicular diameters of palpable tumors every day by a caliper and was calculated using the equation ( l e n g t h   ×   width 2 ) 2 . Mice were sacrificed and analysis was performed 15 days after tumor induction or as indicated each time. Mice with tumors larger than 1,100 mm³ were euthanized. Mice that manifested tumor ulceration were excluded for the experimental processes. At the endpoint of each experiment, the tumor weight was also determined.

For the application of the combinational therapy protocol, each mouse was treated with anti-CTLA-4 Ab (clone 4F10, Bioceros LB) at 100 μg per 100-μl dose and anti-PD-1 Ab (clone RMP1–14, Bioceros LB) at 200 μg per 100-μl dose intraperitoneally (i.p.) every 3 days after tumor implantation, whereas NLRP3 inhibitor (MCC950, Sigma-Aldrich, #5.38120.0001) was administered at 10 mg/kg dose to each mouse by i.p. injection every other day. Control mouse cohort was administered PBS on the same days.

For the analysis of tumor-infiltrating lymphocytes (TILs), single-cell suspensions were generated by dissecting and dissociating tumor tissue in the presence of collagenase D (1 mg ml⁻¹, Roche) and DNase I (0.25 mg ml⁻¹, Sigma), diluted in RPMI-1640 medium (Gibco), for 45 min at 37°C and then were homogenized and strained passing through a 40-μm pore size cell strainer (BD Falcon). For the analysis and isolation of different immune populations, single-cell suspensions from spleen and lymph nodes (LNs) were prepared by homogenization of the tissue and passing through a 40-μm pore size cell strainer. Isolated femoral and tibial bones from the hindlimbs were flushed with ice-cold 5% FBS in PBS for BM single-cell suspensions. Single-cell suspensions from spleen and BM were prepared after erythrocyte lysis with red blood cell lysis buffer.

For extracellular marker staining, single-cell suspensions from TILs, spleen, LNs, or BM were incubated for 20 min at 4°C with the following anti-mouse conjugated antibodies: CD45-PerCP/Cy5.5 (BioLegend, clone 30-F11, #103132, diluted 1:200), CD11c-PE/Cy7 (BioLegend, clone N418, #117318, diluted 1:200), CD11b-Brialliant Violet 510 (BioLegend, clone M1/70, #101263, diluted 1:200), CD11b-PE (BD Pharmingen, clone M1/70, #553311, diluted 1:200), Gr1-Pacific Blue (BioLegend, clone RB6-8C5, #108430, diluted 1:200), Gr1-Brilliant Violet 421 (BioLegend, clone RB6-8C5, #108434, diluted 1:200), Gr1-PE (eBioscience, clone RB6-8C5, #12-5931-82, diluted 1:200), Ly6G-PE (BioLegend, clone 1A8, #127608, diluted 1:200), Ly6C-Brilliant Violet 421 (BioLegend, clone RB6-8C5, #108430, diluted 1:200), CD8a-PE/Cy7 (BioLegend, clone 53-6.7, #100722, diluted 1:200), CD8a-PE (BioLegend, clone 53-6.7, #100708, diluted 1:200), CD4-PE (BioLegend, clone RM4-4, #116006, diluted 1:200), NK-1.1-APC (BioLegend, clone PK136, #108710, diluted 1:200), CD16/32-PE (BioLegend, clone 93, #101308, diluted 1:200), CD16/32-PerCP/Cy5.5 (BioLegend, clone 93, #101323, diluted 1:200), TER-119/Erythroid Cells-PE (BioLegend, clone TER-119, #116208, diluted 1:200), CD45R/B220-PE (BioLegend, clone RA3-6B2, #103208, diluted 1:200), Ly-6A/E (Sca-1)-APC (BioLegend, clone E13-161.7, #122512, diluted 1:200), CD117 (c-kit)-PE/Cy7 (BioLegend, clone 2B8, #105813, diluted 1:200), CD34-Brilliant Violet 421 (BioLegend, clone MEC14.7, #119321, diluted 1:200), CD44-PerCP/Cy5.5 (BioLegend, clone IM7, #103032, diluted 1:200), and CD25-PE (BioLegend, clone 3C7, #101904, diluted 1:200). Dead cells in cultured splenocytes were stained by the addition of 7-AAD Viability Staining Solution (BioLegend, #420404). For NLRP3-APC (R&D Systems, clone 768319, #IC7578A, diluted 1:25) and IL-1β-FITC (R&D Systems, clone 166931, #IC4013F, diluted 1:50) intracellular staining, cells were stained for the extracellular markers, and then permeabilized and stained using the intracellular Fixation & Permeabilization buffer set (eBioscience) according to the vendor’s instructions. Rat IgG1 kappa Isotype Control-APC (eBioscience, clone eBRG1, #17-4301-81, diluted 1:100) and Rat IgG2b kappa Isotype Control-FITC (eBioscience, clone eB149/10H5, # 11-4031-82, diluted 1:100) were used as controls for NLRP3 and IL-1β, respectively. For IFN-γ intracellular staining, tumor cells were incubated with 50 ng ml⁻¹ of phorbol 12-myristate 13-acetate (PMA, Sigma-Aldrich), 2 μg ml⁻¹ of ionomycin (Sigma-Aldrich), and brefeldin (1/1,000; Becton Dickinson Biosciences) for 4 h at 37°C and 5% CO₂, stained for extracellular markers, and fixed and stained for IFN-γ (BioLegend, clone XMG1.2, #505808, diluted 1:50) or Rat IgG2a kappa Isotype Control-PE (eBioscience, clone eBR2a, #12-4321-81, diluted 1:50) using Foxp3 Transcription Factor Staining Buffer Set (eBioscience) according to the manufacturer’s instructions. All samples were analyzed using FACS ARIA III (BD Biosciences) and FACS Canto II (BD Biosciences). Flow cytometry data were analyzed with FlowJo v.8.7 and 10.8.1 software. MDSCs were sorted on a FACS ARIA III (BD Biosciences) and the BD FACSDIVA v8.0.1 software (BD Biosciences). Cell purity was above 95%.

Calculation of TIL numbers per gram of tumor tissue was performed, by flow cytometry, upon tumor tissue isolation, weighing, digestion, and suspension in a 0.1 g/100 μl volume of 5% FBS in PBS prior to staining.

Splenocytes as well as sorted splenic MDSCs from naïve and tumor-bearing mice were isolated as previously described. We seeded 5 × 10⁵ splenocytes or 1 × 10⁵ MDSCs per well in 96-well flat-bottom and 96-well round-bottom plates, respectively, and stimulated them for 24 h with 1 μg/ml LPS from Escherichia coli O55:B5 (Sigma, L2880). The next day, 1 h before the ending of the stimulation time, the culture medium was supplemented with 5 mM adenosine 5’-triphosphate disodium salt hydrate (ATP) inflammasome activator (Jena Bioscience, NU-1010-1G). Supernatants were removed and analyzed using Mouse IL-1 beta/IL-1F2 Quantikine ELISA Kit (R&D Systems, MLB00C), according to the manufacturer’s instructions, and cultured cells were prepared for staining of extracellular and intracellular markers, as described above.

Total splenocytes from naïve mice were seeded 5 × 10⁵ cells per well in 96-well flat-bottom plates and stimulated for 24 h with 1 μg/ml LPS (Sigma, L2880). The following day, in the medium from the overnight culture was added water for injection (control) or MCC950 NLRP3 inhibitor (2 μM) (Sigma-Aldrich, 5381200001) for the indicated time points. One hour before the end of priming, cells were stimulated with ATP inflammasome activator (Jena Bioscience, NU-1010-1G). Supernatants were removed and used for mouse IL-1β ELISA and cultured splenocytes were prepared for 7-AAD viability staining, as described above.

For the suppression assay of MDSC subsets, CD4⁺Foxp3^- effector T cells (Teff) were sorted from the LNs of naïve Foxp3^EGFP.KI mice, as previously described, and stained with the division-tracking dye CellTrace Violet (CTV, Invitrogen, #C34557) according to the manufacturer’s protocol. A total of 75 × 10³ labeled Teff cells were then seeded in 96-well round-bottom plate in each well. M-MDSC (CD11b^highLy6C⁺Ly6G^–) and G-MDSC (CD11b^highLy6C^–Ly6G⁺) subsets sorted from the spleens of WT or Nlrp3^-/- B16.F10 inoculated mice were added at the culture, at a ratio Teff/M-MDSCs 1:1 and Teff/G-MDSCs 3:1. Then, Dynabeads mouse T-activator conjugated with monoclonal antibody (mAb) to the invariant signaling protein CD3 plus mAb to CD28 (Gibco, #11456D) were supplemented into culture, at a ratio of one bead per one Teff cell. Cells were cultured in DMEM (Gibco, #11965) supplemented with 10% heat-inactivated FBS, 100 U/ml penicillin–streptomycin, and 50 μM 2-mercaptoethanol. As positive and negative controls, we used Teff cells cultured with or without anti-CD3/anti-CD28 activation beads, respectively. The plate was incubated at 37°C under 5% CO₂ for 64 h and then cultured cells were prepared for staining of extracellular markers, as described above, for the determination of proliferation and activation of Teff cells with flow cytometry analysis.

Tumor homogenates were generated in PBS that was supplemented with a cocktail of protease inhibitors (Roche) using a pestle inside an Eppendorf. The conditioned media of cultured splenocytes and MDSCs from in vitro cultures were also collected. The homogenates and cell culture supernatants were centrifuged and were assayed for mouse IL-1β using the Mouse IL-1 beta/IL-1F2 Quantikine ELISA Kit (R&D Systems, MLB00C), according to the manufacturer’s instructions.

M-MDSCs and G-MDSCs were sorted from spleens of B16.F10 melanoma-bearing WT and Nrlp3^-/- mice, and total RNA was extracted using the Macherey-Nagel NucleoSpin RNA kit as described by the manufacturer’s protocol (NucleoSpin^® RNA). Each RNA sample was representative of one mouse.

NGS libraries were generated using 300 ng of total RNA as input on average with the QuantSeq 3′ mRNA-Seq Library Prep Kit FWD for Illumina kit from Lexogen according to manufacturer’s protocol, using 15 or 17 cycles of amplification. Libraries were sequenced on Illumina Nextseq 500 on 1 × 75 High flowcell.

Fastq files were downloaded from Illumina-BaseSpace and mapped to mm10 genes (iGenomes UCSC/mm10) using hisat2 version 2.1.0 (–score-min L 0,-0.5) (28). Gene counts were computed with htseq-count (-s yes, version 0.11.2) (29).

Further processing was performed with the R Bioconductor (Bioconductor) package edgeR v.3.14.0 (edgeR). Reads were normalized for intra- and inter-sample variances using the functions “calcNormFactors” and “estimateTagwiseDisp”, and further by the gene length as per Ensembl V103 Genes annotations, resulting in fragments per kilobase per million (FPKM) for each gene. Differential gene expression analysis was performed as previously described (30). Genes with FDR< 0.05 and fold change |FC| > 1.5 were considered statistically significant. Heatmaps and boxplots were created in R with an in-house developed script that is based on the ggplot package.

Significant differentially expressed genes (DEGs) were used for gene ontology (GO) analysis using the g:Profiler web-server. Gene set enrichment analysis (GSEA) was also performed in order to reveal enriched signatures in our gene sets based on the Molecular Signatures Database (MSigDB) v7.4. Gene sets were ranked by taking the –log10 transform of the p-value multiplied by the FC. Significantly upregulated genes were at the top and significantly downregulated genes were at the bottom of the ranked list. GSEA pre-ranked analysis was then performed using the remapped Mouse Gene Symbol dataset and collapsing probe sets while keeping only the max probe value. The rest of the parameters were left to default. Enrichment was considered significant FDR (q-value)<5%.

Data are presented as mean ± S.D., as bar graphs represent the mean and standard deviation (SD) between biologically independent mouse samples or technical repeats, as indicated each time. For statistical analysis, all data were analyzed using Prism 8 (GraphPad Software, Inc., La Jolla, USA). Data were analyzed using the two-tailed, parametric, unpaired Student’s t test or the two-tailed, nonparametric Mann–Whitney test, as appropriate after testing for normality of the values with the F test, with 95% confidence intervals. For multiple-group comparisons, the one-way ANOVA test and Tukey’s multiple comparison test were performed. Kaplan–Meier statistics were done with the log(rank) (Mantel–Cox) test. The p-value of<0.05 was considered to be statistically significant for each dataset.

Despite the established role of inflammasome pathway in the orchestration of an inflammatory response, through release of pro-inflammatory cytokines IL-1β and IL-18, its role in cancer development, progression, and immunotherapy response remains contradictive. To investigate the impact of NLRP3 inflammasome in the progression of LUSC and SKCM, Kaplan–Meier survival analysis was performed based on high and low NLRP3 gene expression or NLRP3 pathway gene-set expression [median expression of the 30 pathway genes as defined by Ju et al. (24), Supplementary Figure 1A], using the top/bottom 33% of samples with the highest/lowest expression, respectively. Patients with the lowest expression of both NLRP3 gene and its pathway gene set were found to display prolonged survival of LUSC patients (p = 0.068 and p = 0.047, respectively, Figures 1A, B, left panels). In accordance, low expression of the NLRP3 pathway gene set was also associated with prolonged survival of SKCM patients (p = 0.059, Figure 1A, right upper panel), whereas in these patients, low expression of NLRP3 gene was associated with increased morbidity rate (p = 0.004, Figure 1B, right lower panel). In summary, these data indicate that differential expression of NLRP3 pathway genes is associated with survival probability of cancer patients and extend the findings that NLRP3 inflammasome may have pro- and anti-tumorigenic roles.

To examine the functional importance of NLRP3 pathway during tumor development, we utilized the Nlrp3-deficient mice (Nlrp3^-/- ), in which the Nlrp3 gene has been disrupted by the insertion of an EGFP cassette (Supplementary Figure 2A). Effective inactivation of NLRP3 inflammasome was confirmed by the absence of IL-1β in culture supernatants of LPS/ATP-stimulated splenocytes from Nlrp3^-/- animals compared to WT mice (Supplementary Figure 2B). Furthermore, Nlrp3^-/- animals did not exhibit significant alterations in the composition of lymphoid (Supplementary Figures 2C, D) and myeloid compartments (Supplementary Figures 2E, F) compared to WT mice at the steady state, suggesting that inactivation of Nlrp3 gene does not disturb the immune homeostasis.

Therefore, this model allows us to study the role of the NLRP3 pathway in host cells through implantation of NLRP3-sufficient tumor cell line. To this end, upon inoculation with B16.F10 melanoma cells, Nlrp3^-/- mice exhibited significantly reduced tumor growth compared to WT mice, as assessed by the measurement of tumor volume and weight (Figure 2A). The tumor-suppressive effect of NLRP3 deficiency was not restricted only to the melanoma model, since inoculation with the LLC cell line demonstrated significantly decreased tumor growth in Nlrp3^-/- animals (Figure 2B). Tumor regression in melanoma-bearing Nlrp3^-/- mice was accompanied by significantly increased frequencies of tumor-infiltrating CD45⁺ leukocytes (Figure 2C), CD8⁺ lymphocytes, which exhibited elevated IFN-γ expression, and NK1.1⁺ cells, as well as increased frequencies of CD4⁺ T cells, compared to WT mice (Figures 2D, E). These findings were further confirmed upon extrapolation of cell subset frequencies to respective numbers per gram of tumor tissue (Supplementary Figure 3A). Interestingly, CD11c^–CD11b⁺Gr1⁺ MDSCs frequencies were comparable between WT and Nlrp3^-/- mice (Figure 3A), whereas assessment of MDSC subsets revealed an increased accumulation of CD11b^highLy6C⁺Ly6G^– M-MDSCs and markedly decreased levels of CD11b^highLy6C^–Ly6G⁺ G-MDSCs into the tumor site of Nlrp3^-/- mice (Figure 3B). The MDSC subset re-arrangement was also evident in the spleen of tumor-inoculated WT and Nlrp3^-/- mice (Figures 3C, D). Overall, these findings indicate an important role of NLRP3 in the suppression of anti-tumor immune responses and tumor development, associated with a major rewiring of the MDSC compartment.

Based on the established expression of NLRP3 in the myeloid compartment (31) combined with the enhanced accumulation of MDSCs compared to other myeloid cell subsets in the TME (Figure 4A and Supplementary Figure 4A) and the extended re-arrangement of MDSC subsets in the tumor-inoculated Nlrp3^-/- mice, we sought to determine the NLRP3 and pro-IL-1β expression in the MDSC population during tumor development. To this end, MDSCs demonstrated enhanced expression of NLRP3, which was accompanied by increased expression of pro-IL-1β, in spleen of melanoma-bearing animals, compared to isotype controls (Figure 4B). This was also confirmed to M-MDSC and G-MDSC subpopulations (Figures 4C, D), with the latter to express higher levels of both NLRP3 and pro-IL-1β (Supplementary Figure 4B).

Next, we asked whether NLRP3 inflammasome is functionally active in MDSCs, and to address this, we exposed them to cellular insults that include “priming” signal with a TLR agonist, such as LPS, and an activation stimulus, such as ATP. Flow cytometry analysis demonstrated robust expression of NLRP3 and pro-IL-1β in MDSCs from spleens of naïve mice (Figure 4E) as well as in highly pure MDSCs isolated from tumor-bearing mice (Figure 4F), and this is accompanied by increased levels of secreted IL-1β as determined in culture supernatants (Figure 4G). Overall, these data indicate that the NLRP3 inflammasome pathway is activated and functional in MDSCs during tumor development.

MDSCs are BM-derived cells, and recent evidence highlights the importance of BM hematopoiesis and in particular of myeloid progenitors in shaping of anti-tumor immunity (32). Thus, we asked whether Nlrp3^-/- animals present alterations of BM progenitors during tumor growth, which may affect the generation of MDSC subsets. To this end, flow cytometry analysis revealed that tumor-bearing Nlrp3^-/- mice did not exhibit significant differences in frequencies of hematopoietic progenitors (LSKs; Lin^–cKit⁺Sca1⁺) in their BM as compared to the tumor-bearing WT group (Supplementary Figures 5A, B). In line with this, assessment of total myeloid progenitor frequencies (MyPs; Lin^–cKit⁺Sca1^–) did not demonstrate any significant difference between Nlrp3^-/- and WT mice (Supplementary Figures 5C, D) with granulocyte macrophage progenitors (GMPs; Lin^–cKit⁺Sca1^–CD16/32⁺CD34⁺) to be decreased in Nlrp3^-/- mice while common myeloid progenitors (CMPs; Lin^–cKit⁺Sca1^–CD16/32^–CD34⁺) were not affected (Supplementary Figures 5C, E). To conclude, these data suggest that Nlrp3^-/- mice did not exhibit differences in the hematopoietic progenitors but granulocyte progenitors are decreased in tumor-bearing animals.

To provide evidence for the functional re-programming of MDSC subsets in Nlrp3-deficient mice as suggested by the transcriptomic analysis, we set up an in vitro suppression assay. To this end, highly pure M-MDSCs and G-MDSCs were isolated from the spleens of melanoma-bearing Nlrp3^-/- and WT animals and co-cultured with CellTrace Violet (CTV)-labeled T effector (CD4⁺Foxp3^-) cells sorted from naïve Foxp3^GFP mice in the presence of anti-CD3/anti-CD28 activation beads (Figure 7A). Both MDSC subsets from Nlrp3^-/- mice displayed reduced suppressive ability compared with their WT counterparts as evidenced by decreased CTV dilution (Figure 7B). Furthermore, absence of T-cell suppression by Nlrp3^-/- MDSC subsets was accompanied by enhanced activation of T cells as demonstrated by the CD25 and CD44 expression (Figure 7B). Overall, these data reveal that NLRP3 deficiency attenuates the suppressive activity of MDSC subsets.

Considering that immunotherapy has demonstrated impressive results, yet in a small proportion of cancer patients, we sought to investigate whether pharmacologic inhibition of NLRP3 inflammasome in combination with ICI immunotherapy may demonstrate a synergistic therapeutic effect in tumor-bearing animals. For this reason, we first examined whether the MCC950 inhibitor, which prevents ATP hydrolysis, thus preserving an inactive conformation of NLRP3 inflammasome (39, 40), could efficiently inhibit NLRP3 inflammasome in primary splenocytes from naïve mice. Indeed, treatment with MCC950 of LPS-exposed splenocytes demonstrated a significant reduction of IL-1β secretion in culture supernatants confirming the potency to inhibit NLRP3 inflammasome (Supplementary Figure 7B). In the meantime, MCC950 did not demonstrate cytotoxic effects since assessment of 7-AAD expression by flow cytometry showed no differences between treated and non-treated cells (Supplementary Figure 7A).

In order to examine the therapeutic efficacy of MCC950 in vivo and whether it may possess a synergistic effect if combined with immune checkpoint immunotherapy, WT animals were inoculated with B16.F10 melanoma cells and treated with MCC950 alone or in combination with anti-PD-1 and anti-CTLA-4 monoclonal antibodies (Figure 8A). Strikingly, systemic pharmacological inhibition of NLRP3 inflammasome significantly reduced tumor growth in tumor-bearing compared to control-treated mice (Figure 8B). However, no significant differences in tumor growth were observed between MCC950-treated mice and those that received the combination treatment protocol (MCC950 + anti-PD-1/anti-CTLA-4 ICIs) (Figure 8B). The therapeutic effect of MCC950 was accompanied by modulation of the myeloid compartment and specifically a significant reduction of both MDSCs and CD11c⁺ DCs was observed (Figure 8C). Of interest, DC frequencies were further reduced in mice treated with the combination of MCC950 inhibitor and anti-PD-1/anti-CTLA-4 antibodies compared to MCC950 alone (Figure 8C). Importantly, pharmacologic MCC950-mediated inflammasome inhibition resulted in a significant increase in the spleen M-MDSCs and subsequent reduction of G-MDSCs as compared to the control-treated group (Figure 8D), recapitulating the results obtained in Nlrp3^-/- tumor-inoculated animals. Collectively, therapeutic targeting of NLPR3 in tumor-bearing mice results in tumor regression and re-programming of the MDSC compartment, while no synergistic effect in combination with ICIs was observed.