Scutellarin (≥98% purity) (110842) was obtained from Guangdong Institute for Drug Control (Guangzhou, China). Dulbecco’s Modified Eagle’s Medium (DMEM) medium with high glucose, Opti-MEM, fetal bovine serum (FBS), streptomycin and penicillin were products of ThermoFisher/Gibco (Carlsbad, CA, United States). MDL12330A (M182), adenosine triphosphate (ATP) (A6419), lipopolysaccharide (LPS) (Escherichia coli O111:B4) (L4391), dimethyl sulfoxide (DMSO) (D8418), propidium iodide (P4170), Hoechst 33342 (B2261), Tween 80 (P8074) and disuccinimidyl suberate (S1885) were purchased from Sigma–Aldrich (St. Louis, MO, United States). H89 (S1643), cell lysis buffer for Western and IP (P0013) and phenylmethanesulfonyl fluoride (PMSF) (ST505) were bought from Beyotime Biotechnology (Haimen, China). Nigericin (#tlrl-nig) was purchased from InvivoGen (San Diego, CA, United States). The antibody against caspase-1p10 (sc-514) was purchased from Santa Cruz (Dallas, TX, United States). The antibody against NLRP3 (Cryo-2) (AG-20B-0014) was purchased from Adipogen AG (Liestal, Switzerland). Specific antibodies against IL-1β (#12242), HMGB1 (#3935), ASC (#67824), p-(Ser/Thr) PKA substrate (#9621), and β-tubulin (#2128) were bought from Cell Signaling Technology (Danvers, MA, United States). The horse-radish peroxidase (HRP)-conjugated horse anti-mouse IgG (#7076), HRP-conjugated goat-anti-rabbit IgG (#7074) and Protein G Agarose beads (#37478) were also from Cell Signaling Technology. CF568 goat-anti-rabbit IgG (H + L), highly cross-adsorbed (#20103) and CF488A-conjugated goat-anti-mouse IgG (#20018), highly cross-adsorbed were obtained from Biotium (Hayward, CA, United States). Scutellarin was dissolved in DMSO at 100 mM, and stored at -20°C.

Female C57BL/6 mice (6–8 weeks of age) were purchased from the Experimental Animal Center of Southern Medical University (Guangzhou, China). Mice were acclimatized for 1 week before experiments. All animal experiments were performed in accordance with the guidelines for the care and use of animals approved by the Committee on the Ethics of Animal Experiments of Jinan University (approved No. JNU20160315).

Bone marrow-derived macrophages (BMDMs) were differentiated as described previously (Kayagaki et al., 2011; Li et al., 2017). In brief, C57BL/6 mice were sacrificed and bone marrow cells were differentiated in culture medium (DMEM supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin and 20% M-CSF-conditioned medium) for 6 days at 37°C in a humidified incubator of 5% CO₂. BMDMs were cultured in 24-well plates at 1.5 × 10⁵ cells/well (0.5 ml) or in 6-well plates at 1.2 × 10⁶ cells/well (2 ml) or in 10-cm petri dish at 7.5 × 10⁶ cells/dish (10 ml) with complete DMEM medium (DMEM supplemented with 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin) at 37°C overnight.

Cell death was measured by propidium iodide (PI) incorporation as described previously (Py et al., 2014; Li et al., 2017). Briefly, BMDMs were primed in Opti-MEM with 500 ng/ml LPS for 4 h. Then the cells were treated with different doses of scutellarin for 1 h followed by ATP (3 mM) or nigericin (10 μM) for indicated time periods. The nuclei were revealed by Hoechst 33342 (5 μg/ml) staining and dead cells revealed by PI (2 μg/ml) staining at room temperature for 10 min. The cells were observed immediately by live imaging using Zeiss Axio Observer D1 microscope equipped with a Zeiss LD PlanNeofluar 20×/0.4 Korr M27 objective lens (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). Fluorescence images were captured with a Zeiss AxioCam MR R3 cooled CCD camera controlled with ZEN software (Carl Zeiss).

Soluble protein in culture supernatants was precipitated as previously described (Kayagaki et al., 2011; Li et al., 2017). Briefly, soluble proteins in culture supernatants (equal volume for each sample) were precipitated with 7.2% trichloroacetic acid plus 0.15% sodium deoxycholate. The precipitated proteins were re-dissolved in 2× sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample loading buffer and subjected to western blot analysis for released mature IL-1β, caspase-1p10 and HMGB1.

Immunoblotting was performed essentially as previously described (Lin et al., 2016). In brief, equal amounts of proteins were separated by SDS-PAGE followed by electrotransfer to polyvinylidene difluoride (PVDF) membranes (#03010040001; Roche Diagnostics GmbH, Mannheim, Germany). Membranes were blocked and incubated with indicated primary antibodies at 4°C overnight, followed by incubation with appropriate HRP-conjugated secondary antibody. Bands were revealed by a BeyoECL Plus kit (P0018; Beyotime, Haimen China) and recorded on X-ray films (Carestream, Xiamen, China). The densitometry of each band was quantified by FluorChem 8000 (Alpha Innotech; San Leandro, CA, United States).

Soluble cytokine IL-1β in culture supernatants of BMDMs was detected by Cytometric Bead Array (CBA) Mouse IL-1β Flex Set (BD Biosciences, San Jose, CA, United States) according to the manufacturer’s instructions. Data were acquired on a flow cytometer (FACSCalibur; Becton Dickinson) equipped with CELLQuest Pro software (Becton Dickinson).

Immunofluorescence analysis was performed as previously described (Zhao et al., 2015; Zha et al., 2016). Briefly, BMDMs were seed in glass-bottomed dishes (4 × 10⁵ cells/dish) and cultured in complete DMEM medium at 37°C overnight. Cells were primed in Opti-MEM with 500 ng/ml LPS for 4 h, and then treated with different doses of scutellarin for 1 h followed by treatment with 3 mM ATP for 30 min or 10 μM nigericin for 1 h in Opti-MEM. After fixation, permeabilization and blocking, the cells were incubated with anti-ASC antibody (1:300) and anti-NLRP3 antibody (1:300) overnight at 4°C, followed by staining with CF568-conjugated goat-anti-rabbit IgG and CF488A-conjugated goat-anti-mouse IgG. Nuclei were revealed by Hoechst 33342 and ASC speck formation was observed under a Zeiss Axio Observer D1 microscope with a Zeiss LD Plan-Neofluar 40×/0.6 Korr M27 objective (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). Fluorescence images were captured by a Zeiss AxioCam MR R3 cooled CCD camera controlled with ZEN software (Carl Zeiss).

The cross-linking of ASC oligomers was performed as described previously (He et al., 2014; Coll et al., 2015). Briefly, BMDMs were seeded in 6-well plates at 1.0 × 10⁶ cells/well. After appropriate treatments, cells were lysed with cold PBS containing 0.5% Triton X-100, and the cell lysates were centrifuged at 6000 × g for 15 min at 4°C. The pellets were washed twice with PBS and then re-suspended in 200 μl PBS. Freshly prepared disuccinimidyl suberate (2 mM) was added to the re-suspended pellets and the suspension was incubated at room temperature for 30 min with rotation. The cross-linked pellets were collected by centrifugation at 6000 × g for 15 min at 4°C and re-dissolved in 25 μl of 1× SDS-PAGE sample loading buffer. Samples were boiled for 5 min and subjected to western blot analysis.

After appropriate treatments, cells were rinsed once with ice-cold PBS and lyzed with 0.5 ml ice-cold cell lysis buffer for Western blot and IP (containing 1 mM PMSF). Cell lysates were centrifuged at 13,000 × g for 10 min at 4°C and pre-cleared with a 10% volume of Protein G agarose beads for 30 min at 4°C with gentle agitation. The pre-cleared cell lysates were then incubated with anti-NLRP3 antibody (1:200) overnight at 4°C with gentle shaking. Antibody-NLRP3 complexes were collected with a 10% volume of Protein G agarose beads for 2 h at 4°C with gentle shaking. The beads were washed five times with cell lysis buffer, boiled for 5 min in 3× SDS-PAGE sample loading buffer, and resolved by Western blot analysis.

The murine model of bacterial sepsis was established according to the protocol described previously (Wegiel et al., 2014; Zha et al., 2016). In brief, thirty C57BL/6 mice were acclimated for 1 week, randomly divided into three groups (10 mice/group), and intragastrically administered once with scutellarin solution (100 mg/kg or 200 mg/kg body weight) or vehicle (2% Tween-80 in PBS), which has referred to previous studies (Li et al., 2011; Yuan et al., 2014; Wang W. et al., 2016). Three hour later, viable E. coli (DH5α) cells (2.5 × 10⁹ colony-forming units (CFU)/mouse, in 0.5 ml of PBS) were injected into the peritoneal cavity of each mouse. One hour after bacterial infection, mice were intragastrically administered once again with scutellarin solution or vehicle, respectively. Mouse survival was monitored every 6 h for 5 consecutive days.

In another experiment, 54 mice (6 mice/group) were treated with scutellarin (100 or 200 mg/kg body weight) or vehicle once by oral gavage, respectively. Three hour later, viable E. coli cells (2.0 × 10⁹ CFU/mouse, in 0.5 ml of PBS) were injected into the peritoneal cavity of all mice. One hour later, mice were gavaged once again with scutellarin or vehicle, respectively. The mice were sacrificed 4 and 8 h after bacterial infection, respectively. Their sera were collected, and serum IL-1β levels were measured by CBA mouse IL-1β Flex Set. The bacterial loads were measured as described previously (Pan et al., 2015). The liver and intestine were isolated and fixed in 4% neutral formaldehyde. Paraffin slices of the tissues were stained with hematoxylin and eosin. Images were captured under the Zeiss Axio Observer D1 microscope armed with a color CCD (ZEISS).

Experiments were performed three times independently, with one representative experiment shown. Data were expressed as mean ± standard deviation (SD). Statistical analysis was performed using GraphPad Prism5.0 (GraphPad Software Inc, San Diego, CA, United States). One-way analysis of variance (ANOVA) followed by Tukey post hoc test and unpaired Student’s t-test were used to analyze the statistical significance among multiple groups and between two groups, respectively. Kaplan–Meier survival curves were used for analysis of mouse survival, and the statistical difference between 2 groups was determined using the log-rank (Mantel-Cox) test. P-value < 0.05 was considered statistically significant.

By using bone marrow-derived macrophages (BMDMs) primed with LPS, we initially sought to explore the influence of scutellarin on NLRP3 inflammasome activation upon canonical NLRP3 activator stimulation. Western blot analysis showed that ATP treatment of LPS-primed macrophages induced a rapid release of active caspase-1p10 (10 kDa) and mature IL-1β (17 kDa) into culture supernatants, indicative of NLRP3 inflammasome activation. Notably, scutellarin pre-treatment dose-dependently reduced ATP-induced release of active caspase-1p10 and mature IL-1β from macrophages, whereas scutellarin per se did not induce the release of these proteins (Figures 1A–C). Consistent with previous studies (Kayagaki et al., 2013), the pro-IL-1β and NLRP3 proteins were highly expressed after LPS priming, whereas pro-caspase-1 and ASC were constitutively expressed regardless of LPS priming in BMDMs (Figure 1A). We also detected the mature IL-1β levels in the culture supernatants by using a bead-based immunoassay (CBA), and found that scutellarin markedly reduced IL-1β release upon ATP stimulation (Figure 1D), thus confirming the result of western blot analysis of IL-1β release (Figure 1A). Besides, scutellarin had minimal effects on the expression of pro-caspase-1, pro-IL-1β, NLRP3, and ASC in the cell lysates (Figures 1A,E–H). These results indicated that scutellarin inhibited NLRP3 inflammasome activation in macrophages upon ATP treatment.

Apart from extracellular ATP, nigericin is another commonly used activator for the NLRP3 inflammasome (He et al., 2016). Different from ATP that acts on the plasma membrane P2X7 purinergic receptor to induce K⁺ efflux thus triggering NLRP3 activation, nigericin is a potassium ionophore thus bypassing the P2X7 receptor to trigger NLRP3 inflammasome activation (Mariathasan et al., 2006). Similar to those results obtained from ATP-treated macrophages (Figure 1), scutellarin dose-dependently reduced caspase-1p10 and mature IL-1β secretion from LPS-primed macrophages upon nigericin stimulation (Figures 2A–D), suggesting that it suppressed nigericin-induced NLRP3 activation. Different from ATP stimulation, under nigericin stimulation, scutellarin treatment drastically reduced pro-IL-1β expression while also moderately decreasing the expression of pro-caspase-1 but not NLRP3 and ASC in the cell lysates (Figures 2A,E–H); the reason underlying this difference is unclear and needs further investigation. Together, these results indicated that scutellarin inhibited both ATP- and nigericin-induced NLRP3 inflammasome activation in murine macrophages.

As NLRP3 inflammasome activation results in caspase-1 activation leading to pyroptosis and release of danger signals including HMGB1, we next explored whether scutellarin had any effect on ATP- or nigericin-induced pyroptosis in LPS-primed BMDMs. The pyroptotic cell death was revealed by propidium iodide (PI) staining. As shown in Figures 3A,B, ATP treatment induced pyroptosis in ∼35% of the cells. Notably, when compared with ATP alone group, scutellarin treatment before ATP stimulation markedly reduced pyroptosis rates in a dose-dependent manner. To further corroborate this, scutellarin also dose-dependently reduced the release of HMGB1 from the macrophages into the culture supernatants (Figures 3C,D). Without ATP triggering, however, scutellarin alone induced neither pyroptosis nor HMGB1 release in these cells. Similarly, scutellarin dose-dependently reduced nigericin-induced pyroptosis in LPS-primed macrophages (Figures 4A,B), and HMGB1 release into the culture supernatants (Figures 4C,D). Together these results showed that scutellarin treatment inhibited NLRP3 inflammasome activation and pyroptosis in LPS-primed murine macrophages upon ATP or nigericin stimulation.

Upon treatment with NLRP3 activators, the adaptor ASC is recruited by NLRP3 to form one large speck in each macrophage, which can be revealed by immunofluorescence microscopy. The results showed that ASC was diffusely distributed in LPS or LPS with scutellarin-treated cells. Upon ATP or nigericin treatment, ASC specks were formed in ∼45% or ∼40% of the cells, whereas scutellarin pre-treatment before ATP or nigericin stimulation markedly reduced the percentages of cells containing ASC specks to 10% or 3%, respectively (Figures 5A,B, 6A,B). Consistent with previous reports (Wang et al., 2013; Granata et al., 2015), ASC specks were highly co-localized with NLRP3 in the cytoplasm, indicative of the recruitment of ASC by NLRP3 (Figures 5A, 6A). These results indicated that scutellarin inhibited NLRP3 activation by blocking ASC speck formation.

Apart from assaying ASC speck formation, we used chemical cross-linking to detect ASC oligomerization, which also reflects the activation of NLRP3 inflammasome (He et al., 2014; Coll et al., 2015). Different ASC multimers (dimers, trimers and higher oligomers) together with monomers were detected in ATP- or nigericin-treated macrophages, but were not detectable in both vehicle- and scutellarin-treated cells without ATP or nigericin. Similar to ASC speck assay, ATP- or nigericin-induced formation of ASC multimers was markedly attenuated by scutellarin pre-treatment (Figures 5C, 6C). Therefore, these data further confirmed that scutellarin inhibited NLRP3 inflammasome activation by blocking ASC recruitment to form ASC specks in macrophages.

The above-mentioned data indicated that scutellarin suppressed NLRP3 inflammasome activation by blocking ASC speck formation. As PKA signaling has been recently demonstrated to phosphorylate and suppress NLRP3 activation (Yan et al., 2015; Guo et al., 2016; Mortimer et al., 2016), we next explored whether PKA signaling was involved in scutellarin-mediated suppression of ASC speck formation upon NLRP3 inflammasome activation. We thus sought to determine whether scutellarin could enhance phosphorylation of NLRP3 on PKA-specific sites. A specific antibody against p-Ser/Thr PKA substrate was used to analyze NLRP3 phosphorylation. The results showed that scutellarin treatment significantly enhanced NLRP3 phosphorylation on PKA-specific sites in LPS-primed macrophages. Further, scutellarin-mediated augmentation of NLRP3 phosphorylation was completely abrogated by PKA inhibitor H89 (Figures 7A,B).

Next, we detected ASC speck formation in the presence or absence of PKA inhibitor H89 by immunofluorescence microscopy. Notably, blocking PKA signaling by H89 not only significantly enhanced ATP-induced ASC speck formation, but also completely abrogated scutellarin-mediated suppression of ATP-induced ASC speck formation in macrophages (Figures 7C,D). H89 alone did not induce ASC speck formation. These data indicated that scutellarin inhibited ASC speck formation upon NLRP3 inflammasome activation by regulating PKA signaling. Together, these results indicated that scutellarin-induced inhibition of ASC speck formation upon NLRP3 inflammasome activation was mediated by PKA signaling.

We next investigated the role of PKA signaling induced by scutellarin on NLRP3 activation and pyroptosis. BMDMs were primed with LPS, pre-treated with the selective PKA inhibitor H89, and then co-incubated with scutellarin followed by ATP treatment to induce NLRP3 inflammasome activation and pyroptosis. The results showed that ATP-induced pyroptosis was markedly enhanced by H89 treatment. Of note, scutellarin-mediated suppression of ATP-induced pyroptosis was completely abrogated by H89 treatment (Figures 8A,B). These results suggested that scutellarin-induced inhibition of NLRP3 activation was dependent on PKA signaling.

Further, we sought to explore whether adenyl cyclase (the upstream component of PKA signaling) also blocked scutellarin’s inhibitory action on NLRP3 inflammasome activation. The specific adenyl cyclase inhibitor MDL12330A was used to block cAMP production. Similar to H89 treatment, MDL12330A not only increased ATP-induced pyroptosis but also significantly reversed scutellarin-mediated suppression of ATP-induced pyroptosis (Figures 9A,B). Consistent with this, scutellarin-mediated suppression of caspase-1p10 and mature IL-1β (17 kDa) release into the culture supernatants of BMDMs were also markedly reversed by MDL12330A treatment (Figures 9C–E). Together, these results indicated that scutellarin inhibited NLRP3 inflammasome activation and pyroptosis by augmenting PKA signaling in murine macrophages.

As ATP-induced NLRP3 activation and pyroptosis plays a critical role in sepsis of bacterial infection (Wegiel et al., 2014), we finally explored whether scutellarin could attenuate bacterial sepsis in a mouse model of intraperitoneal infection. As shown in Figure 10A, mice infected with viable E. coli (2.5 × 10⁹ CFU/mouse) died within 24 h, whereas oral administration of scutellarin significantly prolonged mouse survival of bacterial sepsis. When administered with low-dose (100 mg/kg) and high-dose (200 mg/kg) scutellarin, 20 and 40% of mice had survived the period of observation (120 h), respectively (Figure 10A). In addition, scutellarin treatment significantly reduced serum IL-1β levels and attenuated infiltration of inflammatory cells in the liver of mice infected with E. coli in comparison to vehicle (Figures 10B,C), indicating that scutellarin inhibited systemic inflammation in the mice. But there was no obvious lesion in the jejunum and colon tissues with or without scutellarin administration (Supplementary Figure S1). Scutellarin had no influences on IL-1β levels in uninfected (control) mice (Figure 10B). The bacterial loads were also unchanged by scutellarin administration (Supplementary Figure S2). As IL-1β secretion in vivo upon Gram-negative bacterial infection has been shown to be NLRP3 dependent (He et al., 2013), these results suggested that scutellarin protected mice against bacterial sepsis probably by suppressing NLRP3 inflammasome activation upon microbial infection.