Deficiency of circadian clock gene Bmal1 exacerbates noncanonical inflammasome-mediated pyroptosis and lethality via Rev-erbα-C/EBPβ-SAA1 axis

C57BL/6 mice were purchased from Orient Bio and employed as wild-type (WT) mice. Arntl^flox/flox (JAX 007668), Nlrp3^–/– (JAX 017971), and Gsdmd^–/– (JAX 032410) mice were obtained from The Jackson Laboratory. Gsdmd^flox/flox (RBRC10762) mice were provided by RIKEN BioResource Research Center. Ifnar1^–/– mice were provided by Dr. Sung Jae Shin (Yonsei University). LysMcre mice were provided by Dr. Myung-Shik Lee (Soonchunhyang University). Casp1^–/–Casp11^–/– mice were provided by Dr. Nam-Hyuk Cho (Seoul National University). All genetically modified mice had a C57BL/6 background. For generation of myeloid-specific Bmal1-deficient mice, Arntl^flox/flox mice harboring a loxP-flanked neomycin-resistant cassette in exon 8 of Arntl were crossed with LysMcre mice. For generation of myeloid-specific Gsdmd^–/– mice, Gsdmd^flox/flox mice were crossed with LysMcre mice. Eight- to 12-week-old mice were used in the experiments. All mice were bred and maintained under specific pathogen-free conditions at Yonsei University College of Medicine. The animal experimental protocols were approved by the Institutional Ethical Committee of Yonsei University College of Medicine (2021-0122). All experiments were performed in accordance with the approved guidelines of the Institutional Ethical Committee.

For induction of noncanonical inflammasome activation, mice were intraperitoneally injected with poly(I:C) (10 mg/kg) at ZT0, followed by intraperitoneal injection of lipopolysaccharide (LPS, L4391, 1 or 2 mg/kg) 7 h post-poly(I:C) injection at ZT7, according to previous studies with some modifications^18,19. Alternatively, mice were administered a high dose of LPS (L3012, 20 mg/kg, intraperitoneal injection) at ZT4 to induce in vivo noncanonical inflammasome activation²⁰. Mouse survival was monitored three times per day for 3 days.

LPS (L3012, L4391), ATP, nigericin, poly(I:C), poly(dA:dT), PI, SR9009, and SR8278 were purchased from Sigma‒Aldrich. Recombinant mouse serum amyloid A1 and recombinant mouse IFN-β were obtained from R&D Systems. Flagellin was purchased from InvivoGen. LLME was purchased from Cayman. SAA1 ELISA kits were purchased from R&D Systems. The following antibodies were used in this study: anti-caspase-1 (AG-20B-0042, Adipogen), anti-NLRP3 (AG-20B-00140, Adipogen), anti-mouse IL-1β (AF-401-NA, R&D Systems), anti-caspase-11 (NB120-10454, Novus Biologicals), anti-Bmal1 (93806, Abcam), anti-GSDMD (209845, Abcam), anti-cathepsin B (AF965, R&D Systems), anti-cleaved caspase-3 (9579, Cell Signaling Technology), anti-fibrin (MVI-1007, MERU-VasImmune), anti-phospho-STAT1 (#7649, Cell Signaling Technology), anti-STAT1 (#14994, Cell Signaling Technology), anti-β-actin (sc-47778, Santa Cruz Biotechnology), anti-LAMP-1 (sc-19992, Santa Cruz Biotechnology) and anti-C/EBPβ (sc-7962, Santa Cruz Biotechnology).

Mouse bone marrow cells were isolated from femurs and tibias of mice and cultured in L929-conditioned Dulbecco’s modified Eagle’s medium for 5–7 days for differentiation into BMDMs. BMDMs were maintained in L929-conditioned Dulbecco’s modified Eagle’s medium with 10% fetal bovine serum (FBS) and antibiotics. For isolation of alveolar macrophages, bronchoalveolar lavage fluid (BALF) was obtained by washing with 2% FBS-containing PBS, and alveolar macrophages were harvested by BALF through centrifugation for 10 min at 1500 RPM and 4 °C. For isolation of resident peritoneal macrophages, we collected mouse peritoneal lavage by rinsing with PBS containing 2% FBS. The peritoneal cells were then harvested by centrifugation, resuspended in RPMI 1640 media, and allowed to adhere to a cell culture dish. Following overnight incubation, the adherent cells were washed three times with warm PBS and employed as peritoneal macrophages.

Cells were lysed in buffer containing 20 mM HEPES (pH 7.5), 0.5% Nonidet P-40, 50 mM KCl, 150 mM NaCl, 1.5 mM MgCl₂, 1 mM EGTA, and protease inhibitors. Soluble lysates were fractionated by sodium dodecyl sulfate‒polyacrylamide gel electrophoresis and then transferred to polyvinylidene difluoride membranes. The membranes were stained with appropriate antibodies, developed for visualization using an enhanced chemiluminescence kit, and assessed using an image analyzer (Amersham^TM ImageQuant^TM 800). In some experiments, cell culture supernatants were precipitated with methanol/chloroform as described previously²¹ and then immunoblotted. All blots shown are representative of at least three independent experiments.

Total RNA was isolated using TRIzol reagent (Invitrogen) or the AccuPrep^® Universal RNA Extraction Kit (Bioneer) and was reverse transcribed using the PrimeScript RT Master Mix Kit (TaKaRa). qPCRs were run using SYBR Premix Ex Taq (TaKaRa) and the following primers: 5′-GCC CAT CCT CTG TGA CTC AT-3′ and 5′-AGG CCA CAG GTA TTT TGT CG-3′ (mouse Il1β); 5′-AGT TGC CTT CTT GGG ACT GA-3′ and 5′-TCC ACG ATT TCC CAG AGA AC-3′ (mouse Il6); 5′-CAA TGA GCC AGA CAA CGA GG-3′ and 5′-TAC GCC AAA ATA GCT GTC GC-3′ (mouse Arntl); 5′-TTC CTG CTG TGC TTC TTC AC-3′ and 5′-CTT TCC ATT CAG CTG CTC CA-3′ (mouse Ifnb); 5′-AGC TGG TGA AGA CAT GAC GA-3′ and 5′-GGT GGG AAG TAT GTG GGA CA-3′ (mouse Nr1d1); 5′-ATG ACC TGC ACT CAA TGG GA-3′ and 5′-TGA GTG TTT CCT GCA AAG CC-3′ (mouse Nr1d2); 5′-GGA CAT GAG GAC ACC ATT GC-3′ and 5′-GTA GGA AGA AGC CCA GAC CC-3′ (mouse Saa1); 5′-TTC AAC CAC AGC AGC AAG AC-3′ and 5′-CTG GAT AAT GTT GCT GGG CC-3′ (mouse Lpl); 5′-AGC AGA ACC ACG ATA ACC CA-3′ and 5′-ATC TTG TCT GAG CTG CTG GT -3′ (mouse Nfil3); 5′-AGC CCT GTG AAT ACC GTG AA-3′ and 5′-TGA CTC GTT CAA TGT CCC CA-3′ (mouse Hdc); 5′-TCC AGG AGA GCT GTG TTT GT-3′ and 5′-CTG TGC AGT TCG AGC CTA AC-3′ (mouse Prok2); 5′-ACT TGG CAT CCC TGT AGC AT-3′ and 5′-GTG AAG GCC AGG AAG CAA AA-3′ (mouse Pydc3); 5′-CTG GAA AAC TTC ACC TGC CC-3′ and 5′-AGA CGT TCT GGG GCA TGT AA-3′ (mouse Gcnt2) and 5′-CGC GGT TCT ATT TTG TTG GT-3′ and 5′-AGT CGG CAT CGT TTA TGG TC-3′ (mouse Rn18s). The PCR conditions were 10 min at 95 °C, followed by 40 cycles of 15 sec at 95 °C and 1 min at 60-64 °C. Rn18s was used to normalize target gene expression to calculate changes in mRNA levels with respect to the control.

For induction of conventional NLRP3 inflammasome activation, BMDMs were primed with LPS (0.25 μg/ml, 3 h), followed by ATP treatment (2–2.5 mM, 30–45 min). For stimulation of AIM2 and NLRC4 inflammasome activation, cells were transfected with poly(dA:dT) (1 μg/ml) or flagellin (250 ng/ml) using Lipofectamine 2000 for 2 h. Inflammasome activation was determined by the presence of active caspase-1 p20 and active IL-1β in culture supernatants as evaluated by immunoblotting and by extracellular IL-1β quantification using enzyme-linked immunosorbent assays. For induction of noncanonical inflammasome activation^18,22, cells were treated with IFN-β (500 pg/ml), followed by infection with Escherichia coli BL21 (MOI 20) or LPS transfection (2 μg/ml) using Lipofectamine 2000. Noncanonical inflammasome activation was assayed by the secretion of caspase-11 and the processing of GSDMD.

For quantification of pyroptosis, extracellular LDH release was measured using the CytoTox96 nonradioactive cytotoxicity assay kit (Promega) and was calculated as follows: Extracellular LDH release = [extracellular LDH/(intracellular LDH + extracellular LDH) × 100]. Cell death was also determined by quantifying PI uptake. BMDMs were seeded in black 96-well plates and incubated with 10 μg/ml PI in Opti-MEM. Subsequently, PI fluorescence was measured using a fluorescent plate reader.

After the treatments, mouse lungs were isolated and fixed in 10% formalin for 48 h. Furthermore, the lungs were embedded in paraffin, sectioned, and stained with hematoxylin and eosin. For quantification of cell death in the spleen, spleen tissues were lysed in TEN buffer (50 mM Tris-HCl, 2 mM EDTA, 150 mM NaCl, and protease inhibitors), and the soluble lysates were immunoblotted using anti-caspase-3 antibody. Changes in renal function were analyzed by measuring serum urea levels using a commercial kit. Urea was measured using a blood urea nitrogen (BUN) colorimetric detection kit (Arbor Assays).

For isolation of peritoneal macrophages, mouse peritoneal lavage was obtained by washing with phosphate-buffered saline (PBS) and incubation with an anti-CD16/32 antibody (1 μg/10⁶ cells, 20 min) to block the Fc receptor. Subsequently, the peritoneal cells were incubated with biotinylated anti-CD45R/B220 (0.25 μg/10⁶ cells) and anti-CD3 (0.125 μg/10⁶ cells) antibodies for 15 min and then with streptavidin-coated microbeads (10 μl/10⁷ cells, 15 min). The mixtures were applied to a magnetic-activated cell sorting system to collect the flow-through fraction using an autoMACS Pro Separator (Miltenyi Biotec).

Total RNA was isolated using TRIzol reagent (Invitrogen). RNA quality was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies), and the RNA was quantified using an ND-2000 Spectrophotometer (Thermo). Libraries were prepared from the total RNA using the NEBNext Ultra II Directional RNA-Seq Kit (New England BioLabs). mRNA was isolated using the Poly(A) RNA Selection Kit (Lexogen) and was used for cDNA synthesis and shearing, following the manufacturer’s instructions. Indexing was performed using Illumina indices 1–12. The libraries were enriched using PCR. Fragment size was evaluated using the Agilent 2100 Bioanalyzer (DNA High Sensitivity Kit). Quantification was performed using a library quantification kit and a StepOne RT‒PCR System (Life Technologies). High-throughput sequencing was performed as paired-end 100 sequencing using a NovaSeq 6000 instrument (Illumina).

Raw sequencing data were quality-controlled using FastQC. Adapter and low-quality reads ( < Q20) were removed using FASTX_Trimmer²³ and BBMap²⁴. The trimmed reads were mapped to the reference genome using TopHat²⁵. Gene expression levels were estimated based on fragments per kilobase per million reads values using Cufflinks²⁶. The fragments per kilobase per million reads values were normalized based on quantile normalization using the EdgeR package within R²⁷. Data mining and graphic visualization were performed using ExDEGA (Ebiogen).

Cells were washed twice with PBS and fixed in 1% formaldehyde, and fixation was quenched by the addition of 125 mM glycine. After the cells were harvested, they were lysed with 0.5% NP-40-containing buffer A (5 mM PIPES at pH 8.0 and 85 mM KCl) and sonicated using a Bioruptor (low power, Diagenode). After the removal of cell debris by centrifugation at 14,000 rpm for 15 min, the cell lysates were sonicated using a Bioruptor (high power) to shear the DNA to fragments of 200–400 base pairs. After preclearing with protein G Dynabeads, the appropriate antibody was added, and the cells were incubated at 4 °C overnight. The next day, antibody-chromatin complexes were captured with protein G Dynabeads and washed with low- and high-salt wash buffers and LiCl wash buffer. The chromatin antibody immobilized on the magnetic beads was then subjected to tagmentation. Eluted DNA was purified using the MinElute kit (Qiagen) and subjected to RT‒qPCR analysis.

Cells were grown on coverslips in 24-well plates. Following treatments, the cells were fixed using 4% formaldehyde and permeabilized using 0.2% Triton X-100. After blocking with 4% bovine serum albumin, the cells were incubated with anti-LAMP1 (1:250), anti-C/EBPβ (1:250), and anti-cathepsin B antibodies (1:200) followed by incubation with Cy3-conjugated anti-rat and mouse IgG and Alexa Fluor 488-conjugated anti-goat IgG antibodies. F-actin was stained using rhodamine-phalloidin (Thermo), and nuclei were visualized by counterstaining with DAPI. Images were acquired using a confocal microscope (LSM 700, Zeiss) and processed using ZEN 3.5 software.

All values are expressed as the mean ± standard error of the mean (SEM). All data were pooled from at least three independent experiments. The exact n number of independent experiments is described in each figure legend. Data were analyzed using one-way analysis of variance followed by Dunnett’s post hoc tests for comparison of all groups with the control group, two-way analysis of variance with Bonferroni post hoc correction for comparison between control and Bmal1-deficient mouse groups, or Student’s t tests. The level of statistical significance was set at p ≤ 0.05. Analyses were performed using GraphPad Prism 5.

To investigate whether circadian clock deficiency in macrophages affects the sensitivity of the inflammasome response, we generated myeloid-specific Bmal1-knockout (KO) mice. BMAL1 expression was effectively abolished in bone marrow-derived macrophages (BMDMs) as well as peritoneal or alveolar macrophages but not in peripheral organs, such as the spleen, kidneys, and liver, from myeloid Bmal1-deficient mice (LysM^cre/+Bmal1^flox/flox) (Fig. 1a). We next examined the inflammasome-activating potential of BMDMs from control (Bmal1^flox/flox) and Bmal1-deficient mice (Bmal1^{mye -/-}) by evaluating the presence of active interleukin (IL)-1β or caspase-1 in culture supernatants. Bmal1-deficient macrophages exhibited similar levels of IL-1β secretion (Supplementary Fig. 1a) and caspase-1 cleavage (Fig. 1b and Supplementary Fig. 1b) as control macrophages in response to NLRP3 inflammasome-activating stimulation (lipopolysaccharide (LPS) + ATP or nigericin). Similarly, there was no significant difference in caspase-1 processing between control and Bmal1-KO BMDMs in response to poly(dA:dT) (AIM2 inflammasome agonist) or flagellin (NLRC4 inflammasome agonist) transfection (Fig. 1c and Supplementary Fig. 1c). These data demonstrate that Bmal1 deficiency does not significantly alter the responsiveness of macrophages to canonical NLRP3-, NLRC4-, and AIM2-mediated inflammasome signaling at the cellular level.

We then examined the noncanonical inflammasome response in macrophages stimulated with interferon (IFN)-β priming and subsequent E. coli infection. Intriguingly, Bmal1 deficiency caused a marked increase in caspase-11 secretion and gasdermin D (GSDMD) cleavage in macrophages in response to IFN-β-E. coli stimulation (Fig. 1d). Consistent with this finding, Bmal1-deficient BMDMs showed significantly increased release of lactic acid dehydrogenase (LDH), a typical pyroptosis indicator²⁸, upon IFN-β-E. coli stimulation compared to the control cells (Fig. 1e). IFN-β priming and E. coli infection caused caspase-1/-11-dependent pyroptosis (Fig. 1f), indicating that these stimulants trigger caspase-11-mediated noncanonical inflammasome signaling. Supporting these observations, Bmal1-deficient macrophages exhibited enhanced LDH release by E. coli infection, which also induced caspase-1/-11-dependent cell death (Supplementary Fig. 2a, b). However, there was no difference in canonical inflammasome-mediated pyroptosis between control and Bmal1-deficient BMDMs (Supplementary Fig. 2c, d).

To validate the relevance of Bmal1 deficiency in noncanonical inflammasome-mediated pyroptosis, we examined intracellular LPS-induced pyroptosis in both types of macrophages using LPS transfection. Intracellular LPS-driven LDH release was significantly increased in the Bmal1-deficient BMDMs compared to the control cells (Fig. 1g). LPS transfection-mediated cell death was dependent on GSDMD, a crucial molecule of pyroptosis²⁹ (Fig. 1h). Then, we further confirmed the increased pyroptosis in Bmal1-deficient macrophages by measuring the uptake of propidium iodide (PI). Consistently, myeloid Bmal1-KO BMDMs exhibited significantly increased PI uptake after IFN-β-E. coli stimulation (Fig. 1i) or LPS transfection (Fig. 1j). These results demonstrate that Bmal1 deficiency leads to an enhanced noncanonical inflammasome response and pyroptosis in macrophages.

Next, we assessed the in vivo susceptibility of control and myeloid Bmal1-deficient mice to noncanonical inflammasome activation. To induce noncanonical inflammasome activation in vivo, mice were given an intraperitoneal injection of poly(I:C), a Toll-like receptor 3 (TLR3) agonist, and a low dose of LPS (1 or 2 mg/kg). Poly(I:C) priming and subsequent LPS administration caused Gsdmd-dependent pyroptosis-mediated lethality in mice (Fig. 2a). Notably, poly(I:C)-LPS-induced mouse lethality was significantly increased by myeloid Bmal1 deficiency (Fig. 2b). Consistent herewith, Bmal1-deficient mice were more susceptible to a high dose of LPS challenge (20 mg/kg) (Fig. 2c), which also induced Gsdmd-dependent lethality (Supplementary Fig. 3a). To examine whether poly(I:C)-LPS-induced mouse lethality is mediated by myeloid pyroptosis, we assessed our in vivo noncanonical inflammasome-mediated sepsis model in myeloid-specific Gsdmd-deficient mice. Consequently, mouse lethality induced by poly(I:C)-LPS was completely abolished in the myeloid Gsdmd-deficient mice (Supplementary Fig. 3b). Furthermore, poly(I:C) priming, followed by LPS stimulation, caused elevated myeloid cell death in the peritoneum of Bmal1-deficient mice compared to control mice (Fig. 2d). These results indicate that myeloid circadian clock BMAL1 is closely implicated in the susceptibility of mice to noncanonical inflammasome activation.

We next evaluated the potential role of BMAL1 in the innate immune response in vitro by stimulating BMDMs with LPS or poly(I:C). Bmal1 deficiency did not significantly alter LPS-induced IL-6 production in BMDMs (Fig. 2e and Supplementary Fig. 4a). Likewise, treatment with poly(I:C) induced similar levels of type I IFN production in the Bmal1-deficient and control BMDMs (Supplementary Fig. 4b). Furthermore, there was no significant difference in retinoic acid-inducible gene I-mediated IFN-β production induced by poly(I:C) transfection between control and Bmal1-KO macrophages (Supplementary Fig. 4c).

Next, we examined the phenotypes of control and myeloid Bmal1-deficient mice after poly(I:C) + LPS administration. Bmal1 ablation caused an increase in the serum level of IL-1β, but not IL-6, upon poly(I:C) + LPS challenge (Fig. 2f and Supplementary Fig. 4d). Poly(I:C) priming with LPS administration enhanced caspase-3 cleavage in the spleen more strongly in the Bmal1-deficient mice than in the control mice (Fig. 2g), indicating that Bmal1 deficiency caused more severe spleen damage upon noncanonical inflammasome activation. Increased blood clotting as determined by fibrin or prothrombin was proposed as a Gsdmd-dependent septic lethality³⁰. Consistently, we found an elevation of fibrin levels in the spleen of Bmal1-deficient mice. (Fig. 2h). Consistent with these findings, Bmal1-deficient mice exhibited more severe kidney and lung damage upon poly(I:C) + LPS administration than control mice (Fig. 2i, j). These results demonstrate that myeloid Bmal1 deficiency increases the susceptibility of mice to noncanonical inflammasome-mediated organ damage and pyroptosis-dependent lethality.

To elucidate the molecular mechanism by which myeloid cell-specific deletion of Bmal1 increases mouse lethality upon noncanonical inflammasome stimulation, we performed RNA-sequencing (RNA-seq) analysis to comprehensively compare gene expression levels in peritoneal myeloid cells isolated from control and Bmal1-deficient mice. Prior to the analysis, we enriched myeloid cells from isolated peritoneal cells by removing T (CD3⁺) and B (B220⁺) cells through cell-specific antibody-mediated magnetic-activated cell sorting (Fig. 3a and Supplementary Fig. 5).

Gene Ontology (GO) analysis of genes with a more than 2-fold change in expression after Bmal1 KO (P < 0.05) revealed that myeloid Bmal1 ablation caused robust changes in the expression of circadian rhythm-related genes (Fig. 3b). To further characterize the effect of Bmal1 deletion on the transcriptome, we performed gene set enrichment analysis (GSEA) using Kyoto Encyclopedia of Genes and Genomes (KEGG). The analysis revealed that the rhythmic behavior pathway was negatively regulated by Bmal1 KO; in particular, Nr1d1, Nr1d2, and Per3 were significantly decreased in the Bmal1-KO peritoneal myeloid cells (Fig. 3c). Based on the RNA-seq data, we generated a heatmap of the top 10 up- and downregulated genes in Bmal1-deficient versus control peritoneal myeloid cells (Fig. 3d) and a volcano plot of the differentially expressed genes (Fig. 3e). We then validated the expression of selected highly up- and downregulated candidate genes in peritoneal cells using RT‒qPCR analysis. The expression levels of Lpl and Nfil3 were significantly elevated in peritoneal myeloid cells of the myeloid Bmal1-deficient mice (Supplementary Fig. 6a, b). However, their expression levels in the Bmal1-deficient BMDMs were not significantly different from those in the control cells (Supplementary Fig. 6c, d). In contrast, the expression levels of Nr1d1 and Nr1d2, encoding Rev-erbα and Rev-erbβ, were decreased in mouse peritoneal myeloid cells (Fig. 3f, g) and BMDMs (Fig. 3h, i) of the myeloid Bmal1-deficient mice. These findings indicate that Bmal1 deficiency leads to a significant reduction in the expression of Nr1d1 and Nr1d2 in myeloid cells at steady state.

To provide molecular insight into the increased lethality of Bmal1-deficient mice upon poly(I:C) + LPS challenge, we profiled gene expression in peritoneal myeloid cells from control and Bmal1-deficient mice after poly(I:C) stimulation. Poly(I:C) injection caused a significant upregulation ( > 5-fold, P < 0.05) of 397 genes in peritoneal myeloid cells from mice of both groups. GO analysis of these 397 genes revealed that genes related to antiviral and antibacterial responses were mainly increased in response to poly(I:C) administration (Fig. 4a). Furthermore, we sorted 397 genes by their expression in the poly(I:C)-treated Bmal1-deficient versus control myeloid cells (Fig. 4b). Among the top five upregulated genes in Bmal1-deficient cells, we examined whether their increased expression is correlated with Nr1d1 expression using the Rev-erbα agonist SR9009 because Rev-erbα was remarkably downregulated in the Bmal1-deficient macrophages in our RNA-seq analysis. Notably, poly(I:C)-induced expression of Saa1 mRNA was significantly reduced by SR9009 (Fig. 4c), while the expression of other top upregulated genes, including Hbc, Prok2, Pydc3 and Gcn2, was not impaired by SR9009 (Fig. 4d and Supplementary Fig. 7). These data indicate that Rev-erbα activity is potentially implicated in the poly(I:C)-induced expression of Saa1, which encodes serum amyloid A1 (SAA1), a major acute phase protein in the regulation of the inflammatory process³¹. We then validated the mRNA expression of Saa1 in wild-type and Bmal1-deficient macrophages. Consistent with the transcriptome analysis, Saa1 mRNA expression was significantly increased in the peritoneal myeloid cells of myeloid Bmal1 KO mice compared to control mice upon poly(I:C) stimulation (Fig. 4e). Supporting these findings, Bmal1-deficient macrophages exhibited a significant increase in Saa1 mRNA levels in BMDMs upon poly(I:C) (Fig. 4f) or IFN-β stimulation (Fig. 4g) compared to control cells. In addition, myeloid Bmal1-KO mice exhibited a significant increase in SAA1 levels in peritoneal lavage fluid (Fig. 4h) compared to control mice in response to poly(I:C) administration. These data demonstrate that Bmal1 deficiency leads to increased SAA1 production in myeloid cells of mice upon poly(I:C) or IFN-β treatment.

To examine the potential role of SAA1 in the noncanonical inflammasome-mediated response, we administered recombinant SAA1 to mice during poly(I:C) and LPS injection. Notably, SAA1 administration enhanced mouse lethality induced by poly(I:C) + LPS challenge (Fig. 5a). Since poly(I:C) stimulation caused a notable elevation in SAA1 in the peritoneal lavage of myeloid Bmal1-deficient mice (Fig. 4h), we assessed systemic SAA1 levels in mice under noncanonical inflammasome-mediated septic conditions. In contrast to the increased SAA1 levels in the peritoneum, there was no significant difference in SAA1 production levels in the serum and liver between control and myeloid Bmal1-deficient mice following poly(I:C) + LPS administration (Supplementary Fig. 8a, b). These findings suggest that the elevated production of SAA1 in the peritoneum of Bmal1-deficient mice may contribute to the increased lethality. In accordance with mouse lethality, E. coli infection-induced pyroptosis of macrophages was significantly elevated by the addition of SAA1 (Fig. 5b, c), indicating that SAA1 augments the intensity of noncanonical inflammasome response. In support of these results, we found that exogenous SAA1 potentiated the secretion of caspase-11 and GSDMD cleavage from BMDMs upon E. coli infection (Fig. 5d). However, exogenous addition of SAA1 did not enhance LPS/ATP-induced canonical inflammasome activation, as shown by the processing and secretion of caspase-1 and IL-1β in macrophages (Fig. 5e). Instead, SAA1 treatment caused robust lysosome destabilization in macrophages, as did the lysosomotropic agent l-leucyl-l-leucine methyl ester (LLME) (Fig. 5f). At steady state, the luminal lysosomal protease cathepsin B colocalized with LAMP1, a lysosomal membrane protein. However, treatment with SAA1 or LLME caused lysosomal membrane rupture, as shown by the dispersion of cathepsin B in the cytosol and the disruption of colocalization with LAMP1. These results collectively demonstrate that SAA1 treatment increases pyroptosis in macrophages and mouse lethality upon noncanonical inflammasome-activating stimulation.

To elucidate the mechanism by which Bmal1 deletion facilitates the upregulation of SAA1, we evaluated the potential relevance of type 1 IFN receptor signaling for SAA1 expression because IFNβ-induced SAA1 expression was stronger in the Bmal1-deficient macrophages than in the control cells (Fig. 4g). Of note, the ablation of the IFN receptor significantly impaired poly(I:C)- and IFNβ-stimulated Saa1 mRNA induction (Fig. 6a). Supporting this finding, the JAK1 inhibitor ruxolitinib markedly suppressed Saa1 expression (Fig. 6b) as well as STAT1 phosphorylation by IFNβ (Supplementary Fig. 11a). These observations indicate that poly(I:C)-mediated SAA1 expression is a downstream event of type 1 IFN receptor signaling. Then, we examined a potential difference in type 1 IFN receptor-mediated signaling between control and Bmal1-deficient macrophages. STAT1 phosphorylation following IFNβ treatment exhibited a similar level regardless of Bmal1 expression (Fig. 6c). Consistent with this finding, there was no significant difference in Gbp2 mRNA expression, an interferon-stimulated gene (ISG), between control and Bmal1-deficient macrophages (Fig. 6d), indicating that Bmal1 deficiency did not affect conventional type 1 IFN receptor signaling events.

Instead, we observed that Rev-erbα plays a specific role in regulating Saa1 mRNA expression in response to poly(I:C) stimulation, as evidenced by the use of Rev-erbα agonist SR9009 (Fig. 4c). Consistently, SR9009 significantly inhibited IFN-β-stimulated Saa1 expression (Fig. 6e). However, SR9009 did not reduce poly(I:C)- or IFN-β-promoted mRNA expression of IL-6 and IFN-β in BMDMs (Supplementary Fig. 9a–c). These results suggest that Rev-erbα is a direct regulator of poly(I:C)- or IFN-β-induced SAA1 production and that Rev-erbα-mediated SAA1 expression is different from conventional TLR and type 1 IFN receptor signaling.

In this context, we searched for a potential transcription factor for SAA1 expression and found a C/EBPβ-binding site in the promotor region of Saa1 (Fig. 6f). Our RNA-seq analysis revealed that mouse peritoneal myeloid cells express high levels of Cebpb (Supplementary Fig. 10a). In BMDMs, poly(I:C) stimulation led to a slight but significant increase in C/EBPβ expression (Supplementary Fig. 10b). C/EBPβ was clearly detected in the nucleus of BMDMs, regardless of poly(I:C) stimulation (Fig. 6g). Intriguingly, chromatin immunoprecipitation (ChIP) analysis revealed that poly(I:C) stimulation (4 h) induced robust binding of C/EBPβ to the promoter region of Saa1 in BMDMs (Fig. 6h). This binding was significantly impaired by the Rev-erbα agonist SR9009 (Fig. 6i, j), indicating that Rev-erbα potentially inhibits poly(I:C)-induced association of C/EBPβ with the Saa1 promoter region. Furthermore, type 1 IFN receptor deficiency significantly attenuated C/EBPβ-Saa1 promoter binding induced by poly(I:C) or IFNβ stimulation (Fig. 6k). These results demonstrate that type 1 IFN receptor signaling is essential for C/EBPβ-mediated SAA1 induction in a Rev-erbα-dependent manner. Supporting these findings, the inhibition of JAK1, a downstream molecule of type 1 IFN receptor, markedly reduced Saa1 expression (Fig. 6b) and noncanonical inflammasome-mediated pyroptosis (Supplementary Fig. 11b).

Consistent with the above data, Bmal1-deficient macrophages, which expressed substantially lower levels of Rev-erbα than control cells, showed greatly enhanced binding of C/EBPβ with the Saa1 promoter in response to 1 h of poly(I:C) treatment (Fig. 7a, b). We then examined the physiological relevance of Rev-erbα activity in poly(I:C)-induced SAA1 production using SR8278, a Rev-erbα antagonist. Similar to Bmal1-deficient mice, SR8278-treated mice exhibited significantly increased Saa1 mRNA levels in peritoneal myeloid cells upon poly(I:C) stimulation (Fig. 7c). However, SR8278 administration did not alter poly(I:C)-induced Il6 and Ifnb mRNA levels in peritoneal myeloid cells (Fig. 7d, e). To further confirm whether the blockade of Rev-erbα by SR8278 could regulate noncanonical inflammasome-mediated SAA1 production and pyroptosis in macrophages, we examined the potential effect of SR8278 treatment in BMDMs upon E. coli infection. Consequently, SR8278 treatment caused a significant increase in E. coli-induced Saa1 mRNA production (Fig. 7f) and pyroptosis (Fig. 7g, h).

Finally, we assessed whether inhibition of Rev-erbα would affect the in vivo susceptibility of mice to a noncanonical inflammasome agonist. Notably, SR8278 administration led to a significant reduction in mouse survival after poly(I:C) + LPS challenge (Fig. 7i). In support of this finding, SR8278-mediated blockade of Rev-erbα increased mouse lethality induced by a high dose of LPS, which also triggered noncanonical inflammasome signaling (Fig. 7j). In contrast, SR8278 treatment did not enhance poly(I:C)-LPS-induced lethality in myeloid Gsdmd-deficient mice (Fig. 7k), indicating that Rev-erbα can affect myeloid pyroptosis-mediated mouse lethality. These results demonstrate that Rev-erbα is a key regulator of the noncanonical inflammasome response and consequent mouse lethality.

All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials. RNA-Seq data are available through the National Center for Biotechnology Information Gene Expression Omnibus under accession number GSE222759.