YVAD-CHO, ammonium chloride (NH₄Cl), chloroquine diphosphate (CQ), N-acetyl cysteine (NAC), potassium chloride (KCl), glibenclamide, probenecid, carbenoxolone, LPS (Escherichia coli O111:B4), cyclosporine A, CA-074-Me, nordihydroguaiaretic acid (NDGA) were purchased from Sigma-Aldrich (St. Louis, MO). TLR2 shRNA lentiviral particles (sc-40257-V), control shRNA lentiviral particles (sc-108080), P₂X₇ shRNA plasmids (sc-42576-SH), control shRNA plasmids (sc-108060), Cryopyrin CRISPR/Cas9 KO plasmids (sc-432122), Manganese (III) tetrakis (4-benzoic acid) porphyrin chloride (MnTBAP), and antibodies against ASC (SC-22514-R, polyclonal antibody), IL-18 (SC-6177, polyclonal antibody), P₂X₇ (SC-514962, monoclonal antibody), and actin (SC-47778, monoclonal antibody) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies against NLRP3 (AG-20B-0014, monoclonal antibody) and mouse caspase-1 (AG-20B-0044, monoclonal antibody) were purchased from Adipogen International (San Diego, CA). Antibodies against IL-1β (AB-401-NA, polyclonal antibody) were purchased from R&D Systems (Minneapolis, MN). MCC950 was purchased from TargetMol (Wellesley Hills, MA). DiOC₂(3) and ELISA kits for IL-1β and tumor necrosis factor-α (TNF-α) were purchased from Thermo Fisher Scientific (Waltham, MA). Phorbol 12-myristate 13-acetate was purchased from Merck Millipore (Bedford, MA). Macrophage Colony Stimulating Factor (M-CSF) was purchased from Peprotech (London, UK).

Mouse J774A.1 macrophages and human THP-1 monocytes were purchased from the American Type Culture Collection (Rockville, MD). THP-1 macrophages were differentiated from THP-1 monocytes by treatment with 50 nM PMA for 48 h. Human peripheral blood mononuclear cells (PBMCs) were separated from whole blood from healthy volunteers by density gradient centrifugation using Histopaque-1077 (23), and all experimental protocols were performed in accordance with the guidelines and regulations provided and accepted by the Institutional Review Board of the Tri-Service General Hospital, National Defense Medical Center and the volunteers' informed consent (TSGH-IRB-2-106-05-190 and TSGH-IRB-2-106-05-009). Mouse primary bone marrow derived macrophages (BMDM) were prepared from bone marrow collected from C57BL/6 mouse femur and tibia by differentiating in the M-CSF containing medium for 7 days. Animal experiments were performed with the approval of the Institutional Animal Care and Use Committee of the National Ilan University (approval number: No. 106-13). TLR2-knockdown and scramble control J774A.1 macrophages were generated by stably infection of TLR2 shRNA lentiviral particles and control shRNA lentiviral particles, respectively. P₂X₇-knockdown and scramble control J774A.1 macrophages were generated by stably transfection of P₂X₇ shRNA plasmids and control shRNA plasmids, respectively. NLRP3-knockout J774A.1 macrophages were generated by transfection of Cryopyrin CRISPR/Cas9 KO plasmids, and the clone with significantly reduced NLRP3 protein expression was selected for further studies. All cells were cultured in RPMI-1640 medium containing 10% fetal bovine serum in a 37°C CO₂ incubator.

S. sonnei (strain 25,931) was purchased from American Type Culture Collection. Generally, the bacteria were grown and subcultured twice a week on chocolate agar purchased from Creative Lifesciences (Taipei, Taiwan) at 37°C in 5% CO₂. S. sonnei was subcultured again 1 day before infection. J774A.1 macrophages, THP-1 macrophages, or BMDM were primed for 4 h with 1 μg/ml LPS and then infected with S. sonnei at different multiplicities of infections (MOIs) for 1 h at 37°C. The extracellular bacteria were washed with sterile PBS and cultured in fresh medium for an additional 20 h with 3 μg/ml gentamicin (to kill the residual extracellular bacteria and avoid excessive bacterial growth in the medium). The intracellular penetration of gentamicin is low and did not have significant effect on intracellular bacteria (24). If the inhibitor (YVAD-CHO, MCC950, NH4Cl, CQ, NAC, KCl, glibenclamide, probenecid, carbenoxolone, cyclosporine A, MnTBAP, CA-074-Me, and NDGA) was used, it was added to the medium after LPS priming and 30 min before S. sonnei infection. The expression levels of IL-1β and TNF-α in the culture medium were measured by ELISA as described previously (25). In addition, to detect the expression levels of proIL-1β/IL-1β, proIL-18/IL-18, p45/p10, NLRP3, and ASC in the culture medium, the medium was concentrated with methanol/chloroform as described previously (26) and then analyzed by Western blotting. To detect the expression levels of proIL-1β, NLRP3, and actin in the cells, the cell lysates were analyzed by Western blotting. Heat-killed S. sonnei were prepared by incubating the bacteria in 80°C hot-plate for 30 min. Freeze/thaw-killed S. sonnei were prepared by repeatedly freezing the bacteria at −80°C for 1 h and thawing at room temperature five times. The loss of bacterial viability was confirmed by plating on a chocolate agar plate.

Intracellular ROS levels were measured by staining the cells with the general oxidative stress indicator CM-H₂DCFDA (Thermo Fisher Scientific). J774A.1 macrophages were primed with 1 μg/ml LPS for 4 h and then infected with 50 MOI S. sonnei for an additional 20 h. Then, the cells were stained with 2 μM CM-H₂DCFDA for 15 min, and the intracellular fluorescence intensity was detected by flow cytometry (Cytomics FC500 Flow Cytometry CXP, Beckman Coulter Life Sciences).

To detect mitochondrial ROS production, J774A.1 macrophages were primed with 1 μg/ml LPS for 4 h and then incubated with or without 20 μM MnTBAP for 30 min. The cells were infected with S. sonnei at 50 MOI for an additional 20 h. Then, the cells were stained with 5 nM MitoSOX for 15 min, and the fluorescence signal was acquired by flow cytometry. To detect the mitochondrial membrane potential, J774A.1 macrophages were primed with 1 μg/ml LPS for 4 h and then infected with 50 MOI S. sonnei for an additional 20 h. Then, the cells were stained with 50 nM DiOC₂(3) for 15 min. The fluorescence signals were detected by flow cytometry.

For short-term treatment, wild-type and NLRP3 knockout J774A.1 macrophages were infected with S. sonnei at 50 MOI for 15 min at 37°C in a CO₂ incubator. The extracellular bacteria were washed out with sterile PBS and incubated in PBS containing 300 μg/ml gentamicin for an additional 1 h at 37°C in a CO₂ incubator to completely kill the extracellular bacteria. The cells were washed with PBS and lysed with 300 μl distilled water for 40 min at 37°C in a CO₂ incubator. The lysate was diluted 300 times with PBS, and 200 μl diluted lysate was then inoculated on chocolate agar plates and incubated at 37°C in a CO₂ incubator overnight. The number of colony-forming units (CFUs) was counted and calculated. For long-term treatment, wild-type and NLRP3 knockout J774A.1 macrophages were infected with S. sonnei at 50 MOI for 15 min at 37°C in a CO₂ incubator. The extracellular bacteria were washed out with sterile PBS and cultured in medium containing 100 μg/ml gentamicin for an additional 20 h at 37°C in a CO₂ incubator. The cells were washed with PBS and lysed in 300 μl distilled water for 40 min at 37°C in a CO₂ incubator. The lysate was diluted 50 times with PBS, and then 200 μl diluted lysate was inoculated on chocolate agar plates and incubated at 37°C in a CO₂ incubator overnight. The number of CFUs was counted and calculated. Bactericidal activity was presented as the number of killed bacteria, which was calculated by subtracting the long-term CFU assay result from the short-term CFU assay result.

Statistical significance was determined by GraphPad Prism 7.0 software. Two-tailed t-tests were used for two groups, and ANOVA with Dunnett's multiple comparisons test was used for three or more groups. ^*, ^**, and ^*** indicate a significant difference at the levels of p < 0.05, p < 0.01 and p < 0.001, respectively.

To investigate whether S. sonnei infection induces IL-1β secretion, untreated or LPS-primed mouse J774A.1 macrophages were infected with S. sonnei at 25, 50, or 100 MOI for 24 h. We found that S. sonnei induced IL-1β secretion in LPS-primed J774A.1 macrophages but not in untreated J774A.1 macrophages (Figure 1A). These results indicated that S. sonnei provided the activation signal but not the priming signal of inflammasome in macrophages. S. sonnei infection also induced IL-1β secretion in LPS-primed human THP-1 macrophages, human PBMCs, and mouse BMDM (Figure 1A). The S. sonnei-mediated induction of IL-1β was confirmed by detecting IL-1β expression in the culture medium of J774A.1 and THP-1 macrophages by Western blotting (Figure 1B). In addition, S. sonnei infection increased the expression of another inflammasome product, IL-18, in the culture medium of J774A.1 and THP-1 macrophages, as analyzed by Western blotting (Figure 1C). LPS- and ATP-activated J774A.1 macrophages were a positive control for IL-1β and IL-18 secretion, because ATP strongly activates the NLRP3 inflammasome in mouse macrophages, but only slightly activates the NLRP3 inflammasome in THP-1 macrophages. As the maturation and secretion of IL-1β and IL-18 are regulated by caspase-1, we asked whether caspase-1 is activated by S. sonnei infection. We found that the level of active caspase-1 (p10) in the culture medium of J774A.1 macrophages was increased by S. sonnei infection, as analyzed by Western blotting (Figure 1D). It has been demonstrated that upon the activation of inflammasome, inflammasome components are released from macrophages, act as extracellular danger signals and amplify the inflammatory response (27). We found that S. sonnei infection induced NLRP3 and ASC release into the culture medium of J774A.1 macrophages (Figure 1E).

To investigate whether S. sonnei-induced IL-1β secretion requires the activation of the NLRP3 inflammasome, LPS-primed J774A.1 macrophages were incubated for 0.5 h with the caspase-1 inhibitor YVAD-CHO before S. sonnei infection. We found that YVAD-CHO reduced IL-1β secretion in a dose-dependent manner (Figure 2A). In addition, the NLRP3-specific inhibitor MCC950 reduced S. sonnei-induced IL-1β secretion in LPS-primed J774A.1 macrophages and BMDM, indicating the important role of NLRP3 in IL-1β secretion (Figure 2B). The role of NLRP3 in S. sonnei-mediated IL-1β secretion was confirmed in J774A.1 macrophages by CRISPR/Cas9-mediated NLRP3 knockout, as S. sonnei and ATP failed to induce IL-1β secretion (Figure 2C) and caspase-1 activation in LPS-primed NLRP3 knockout cells (Figure 2D). These results demonstrate that S. sonnei induces IL-1β secretion through the NLRP3 inflammasome.

Potassium efflux plays important roles in NLRP3 inflammasome activation in response to various NLRP3 stimulators (28). To investigate whether S. sonnei mediates NLRP3 inflammasome activation through potassium efflux, high extracellular potassium concentrations (12.5 and 25 mM KCl in the culture medium) were used to block potassium efflux. We found that S. sonnei-mediated IL-1β secretion and caspase-1 activation in J774A.1 macrophages were significantly inhibited by extracellular KCl (Figure 3A). To confirm the role of potassium efflux in S. sonnei-mediated NLRP3 inflammasome activation, potassium efflux was blocked by the adenosine triphosphate-sensitive potassium channel blocker glibenclamide. We found that S. sonnei-mediated IL-1β secretion and caspase-1 activation in J774A.1 macrophages were significantly inhibited by glibenclamide in a dose-dependent manner (Figure 3B). To further determine the receptor involved in S. sonnei-mediated potassium efflux, the effects of probenecid (a P₂X₇ receptor inhibitor) and carbenoxolone (a pannexin-1 inhibitor) on S. sonnei-induced NLRP3 inflammasome activation were investigated. We found that the P₂X₇ receptor inhibitor probenecid, but not the pannexin-1 inhibitor carbenoxolone, reduced IL-1β secretion in S. sonnei-infected J774A.1 macrophages (Figure 3C). To provide direct evidence of the important role of the P₂X₇ receptor in S. sonnei-mediated NLRP3 inflammasome activation, we knocked down the P₂X₇ receptor in J774A.1 macrophages by shRNA technology (26). Although the cell surface expression of P₂X₇ receptor was partially reduced by shRNA, S. sonnei-, and ATP-mediated IL-1β secretion was significantly reduced in P₂X₇ receptor knockdown cells compared to scramble shRNA-treated control cells (Figure 3D). Notably, although ATP-mediated caspase-1 activation was significantly reduced in P₂X₇ receptor knockdown cells, S. sonnei-mediated caspase-1 activation was not reduced in P₂X₇ receptor knockdown cells compared to control cells (Figure 3E). To understand why P₂X₇ receptor knockdown reduced IL-1β secretion without affecting caspase-1 activation, we investigated the expression levels of NLRP3 and proIL-1β in LPS-activated control and P₂X₇ receptor knockdown J774A.1 macrophages. We found that LPS-induced NLRP3 and proIL-1β expression was reduced in P₂X₇ receptor knockdown cells (Figure 3F). These results indicate that the P₂X₇ receptor participates in the priming of the NLRP3 inflammasome.

It has been proposed that ROS are crucial elements for NLRP3 inflammasome activation (29). We investigated whether S. sonnei infection activates the NLRP3 inflammasome through the generation of ROS. We found that S. sonnei infection increased H₂O₂ production in LPS-primed J774A.1 macrophages, as analyzed by H₂DCFDA staining (Figure 4A). The inhibition of H₂O₂ by the antioxidant NAC reduced IL-1β secretion and caspase-1 activation in S. sonnei-infected J774A.1 macrophages (Figure 4B). NAC also reduced IL-1β secretion in S. sonnei-infected BMDM (Figure 4B). In addition, the inhibition of ROS generation enzyme lipoxygenase by NDGA reduced IL-1β secretion in S. sonnei-infected J774A.1 macrophages and BMDM (Figure 4C), confirming the importance of ROS in S. sonnei-mediated IL-1β secretion. Furthermore, the inhibition of the lysosomal cysteine protease cathepsin B by CA-074-me attenuates NLRP3 inflammasome activation in Mycobacterium tuberculosis- and Neisseria gonorrhoeae-infected cells (30, 31), suggesting an important role for lysosomes in bacterial infection-mediated NLRP3 inflammasome activation. As cathepsin B can be released from damaged lysosomes to drive NLRP3 inflammasome activation through binding to NLRP3 (32), NH₄Cl, and CQ, which both inhibit endosomal/lysosomal acidification, were used to block lysosomal damage. We found that both NH₄Cl and CQ significantly reduced IL-1β secretion in S. sonnei-infected J774A.1 macrophages (Figure 4D), confirming the role of lysosomes in NLRP3 inflammasome activation in response to S. sonnei infection. Furthermore, the cathepsin B inhibitor CA-074-me reduced IL-1β secretion in S. sonnei-infected J774A.1 macrophages (Figure 4E). These results indicate that S. sonnei activates the NLRP3 inflammasome through H₂O₂ production and lysosomal damage.

It has been demonstrated that mitochondrial damage drives downstream signaling that leads to NLRP3 inflammasome activation (33). Mitochondrial ROS are one downstream signals of damaged mitochondria that activate the NLRP3 inflammasome. Using the mitochondrial ROS indicator MitoSOX, we demonstrated that S. sonnei infection significantly induced mitochondrial ROS production in J774A.1 macrophages and that this effect was inhibited by MnTBAP, a mimic of superoxide dismutase (Figure 5A). We also found that the inhibition of mitochondrial ROS production by MnTBAP inhibited IL-1β secretion in S. sonnei-infected J774A.1 macrophages (Figure 5B). In addition, we found that S. sonnei infection caused mitochondrial integrity loss in J774A.1 macrophages, as analyzed by the mitochondrial membrane-potential-sensitive probe DiOC₂(3) (Figure 5C). Furthermore, preserving mitochondrial integrity with cyclosporine A, an inhibitor of mitochondrial membrane permeability transition (33), reduced IL-1β secretion in S. sonnei-infected J774A.1 macrophages (Figure 5D). These results suggest that S. sonnei activates the NLRP3 inflammasome through mitochondrial damage.

To investigate whether the activation of the NLRP3 inflammasome requires phagocytosis of S. sonnei, J774A.1 macrophages were incubated with cytochalasin D, a cell-permeable actin polymerization inhibitor, which blocks the phagocytosis of macrophages. We found that cytochalasin D significantly reduced IL-1β secretion but only slightly reduced TNF-α secretion in S. sonnei-infected J774A.1 macrophages (Figure 6A). These results indicate that phagocytosis of S. sonnei is required for the full activation of the NLRP3 inflammasome but plays a smaller role in NLRP3-independent TNF-α secretion. In addition, our previous study indicated that TLR2 and phagocytosis of live Neisseria gonorrhoeae are important for the NLRP3 inflammasome activation (26). To investigate the role of TLR2 and live S. sonnei in NLRP3 inflammasome activation, we generated TLR2 knockdown J774A.1 macrophages by shRNA technology (26). We found that TLR2 mRNA expression was significantly reduced in sh-TLR2 cells compared to the scramble control (sh-SC) cells, and TLR2 ligand Pam3CSK4-mediated TNF-α secretion was significantly reduced in sh-TLR2 cells compared to sh-SC cells, indicating the functional knockdown of TLR2 in the cells (Figure 6B). We investigated the effect of live, heat-killed, or freeze/thaw-killed S. sonnei on IL-1β and TNF-α in sh-TLR2 and sh-SC cells. We found that heat-killed and freeze/thaw-killed S. sonnei induced less IL-1β and TNF-α secretion than that induced by live S. sonnei (Figure 6C). These results indicate that the NLRP3 inflammasome activation and TNF-α expression requires phagocytosis of live S. sonnei. In addition, we found that live, heat-killed, and freeze/thaw-killed S. sonnei induced less IL-1β and TNF-α secretion in sh-TLR2 cells than that in sh-SC cells (Figure 6C). These results indicate that TLR2 plays an important role in S. sonnei-mediated IL-1β and TNF-α secretion in macrophages.

Our previous study demonstrated that the knockout of NLRP3 by CRISPR/Cas9 technology enhances phagocytosis of pHrodo Green E. coli BioParticles Conjugate and Neisseria gonorrhoeae by macrophages (26). In this study, we assessed the functional consequence of NLRP3 knockout on phagocytosis of S. sonnei by macrophages. The number of engulfed bacteria in wild-type and NLRP3 knockout J774A.1 macrophages was determined by the CFU assay after 15 min of infection. The extracellular bacteria were washed out and completely killed by gentamicin. We found that the number of engulfed bacteria in NLRP3 knockout cells (17,400 ± 1,665 CFUs) was higher than that in wild-type cells (3,240 ± 332 CFUs; Figures 7A,B). These results indicated that NLRP3-knockout increased phagocytosis of S. sonnei by macrophages. To further investigate the functional consequence of NLRP3 knockout on the bactericidal activity of macrophages against S. sonnei, the number of intracellular live S. sonnei cells after 20 h of infection was determined by the CFU assay. We found that the number of intracellular live S. sonnei cells in NLRP3 knockout cells and wild-type cells was 7,370 ± 451 and 760 ± 57, respectively (Figures 7A,B). The number of killed S. sonnei cells was calculated by subtracting the 20 h CFU assay result from the 15 min CFU assay result. We found that NLRP3 knockout significantly increased the bactericidal activity of macrophages against S. sonnei (10,030 S. sonnei cells were killed) compared to wild-type cells (2,480 S. sonnei cells were killed; Figure 7B). These results indicate that NLRP3 knockout increases the phagocytic and bactericidal activity of macrophages against S. sonnei.