- Email: [email protected]
Gossypol induces pyroptosis in mouse macrophages via a non-canonical inflammasome pathway
Toxicology and Applied Pharmacology 292 (2016) 56–64
Contents lists available at ScienceDirect
Toxicology and Applied Pharmacology journal homepage: www.elsevier.com/locate/ytaap
Gossypol induces pyroptosis in mouse macrophages via a non-canonical inflammasome pathway Qiu-Ru Lin a,1, Chen-Guang Li a,1, Qing-Bing Zha b,1, Li-Hui Xu c, Hao Pan a, Gao-Xiang Zhao a, Dong-Yun Ouyang a,⁎, Xian-Hui He a,⁎ a b c
Department of Immunobiology, College of Life Science and Technology, Jinan University, Guangzhou 510632, China Department of Fetal Medicine, The First Affiliated Hospital of Jinan University, Guangzhou 510632, China Department of Cell Biology, College of Life Science and Technology, Jinan University, Guangzhou 510632, China
a r t i c l e
i n f o
Article history: Received 31 July 2015 Revised 10 December 2015 Accepted 31 December 2015 Available online 4 January 2016 Keywords: Gossypol Pyroptosis Inflammasome HMGB1 Interleukin-1β
a b s t r a c t Gossypol, a polyphenolic compound isolated from cottonseeds, has been reported to possess many pharmacological activities, but whether it can influence inflammasome activation remains unclear. In this study, we found that in mouse macrophages, gossypol induced cell death characterized by rapid membrane rupture and robust release of HMGB1 and pro-caspase-11 comparable to ATP treatment, suggesting an induction of pyroptotic cell death. Unlike ATP, gossypol induced much low levels of mature interleukin-1β (IL-1β) secretion from mouse peritoneal macrophages primed with LPS, although it caused pro-IL-1β release similar to that of ATP. Consistent with this, activated caspase-1 responsible for pro-IL-1β maturation was undetectable in gossypol-treated peritoneal macrophages. Besides, RAW 264.7 cells lacking ASC expression and caspase-1 activation also underwent pyroptotic cell death upon gossypol treatment. In further support of pyroptosis induction, both pan-caspase inhibitor and caspase-1 subfamily inhibitor, but not caspase-3 inhibitor, could sharply suppress gossypol-induced cell death. Other canonical pyroptotic inhibitors, including potassium chloride and N-acetyl-L-cysteine, could suppress ATP-induced pyroptosis but failed to inhibit or even enhanced gossypol-induced cell death, whereas nonspecific pore-formation inhibitor glycine could attenuate this process, suggesting involvement of a non-canonical pathway. Of note, gossypol treatment eliminated thioglycollate-induced macrophages in the peritoneal cavity with recruitment of other leukocytes. Moreover, gossypol administration markedly decreased the survival of mice in a bacterial sepsis model. Collectively, these results suggested that gossypol induced pyroptosis in mouse macrophages via a non-canonical inflammasome pathway, which raises a concern for its in vivo cytotoxicity to macrophages. © 2016 Elsevier Inc. All rights reserved.
1. Introduction The inflammasome is a multimeric protein platform that is formed in innate immune cells, epithelial cells, and many other cell types to mount a coordinated molecular defense against pathogenic microbes and tissue injury (Lamkanfi and Dixit, 2014). The assembly of inflammasome complex is initiated upon the recognition of pathogenAbbreviations: ASC, apoptosis-associated speck-like protein containing a caspaserecruitment domain; GOS, gossypol; DAMPs, danger-associated molecular patterns; DMEM, Dulbecco's modified Eagle's medium; DMSO, dimethyl sulfoxide; EDTA, ethylenediamine tetra-acetic acid; FBS, fetal bovine serum; HMGB1, high-mobility group box 1; HRP, horse-radish peroxidase; IL-1β, interleukin-1β; LPS, lipopolysaccharide; NAC, N-acetyl-L-cysteine; PI, propidium iodide; ROS, reactive oxygen species; PAMPs, pathogen-associated molecular patterns; SD, standard deviation; SDS, sodium dodecyl sulfate; SDS-PAGE, SDS-polyacrylamide-gel electrophoresis; siRNA, small interfering RNA; TG, thioglycolate. ⁎ Corresponding authors. E-mail addresses: [email protected] (D.-Y. Ouyang), [email protected] (X.-H. He). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.taap.2015.12.027 0041-008X/© 2016 Elsevier Inc. All rights reserved.
associated molecular patterns (PAMPs), or danger-associated molecular patterns (DAMPs) by nucleotide-binding domain and leucine-rich repeat receptors (NLRs) or absent in melanoma 2 (AIM2)-like receptors (ALRs) (de Zoete et al., 2014). Such sophisticated mechanisms by which inflammasomes respond to danger signals lead to secretion of proinflammatory interleukin (IL)-1β and IL-18, which is crucial for clearance of infectious agents, as well as pyroptosis, an inflammatory form of cell death (Man and Kanneganti, 2015). Pyroptosis is a programmed cell death that is characterized by cell swelling, rapid plasma-membrane rupture, and release of proinflammatory contents including IL-1β, IL-18, ATP and high-mobility group box 1 (HMGB1) (Fink and Cookson, 2006), thus facilitating recruitment of neutrophils and monocytes, and leading to clearance of pathogenic microbes (Scaffidi et al., 2002). It can be induced by a wide range of PAMPs (i.e., flagellin, lipopolysaccharide (LPS), and bacterial toxins) and DAMPs (i.e., ATP, uric acid crystals, and amyloid-β fibrils) and generally mediates host defenses through the release of proinflammatory components thus being beneficial to the host during an infection. On the other hand, pyroptosis also has critical roles in the development and
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
progression of various chronic diseases, including gout, atherosclerosis, and metabolic syndrome, demonstrating the inflammatory response as a “double-edged sword” (Man and Kanneganti, 2015). It has recently been recognized that two distinct inflammasome activation pathways exist: the canonical and non-canonical pathways (Lamkanfi and Dixit, 2014; Man and Kanneganti, 2015). Under most circumstances, activated NLRs and ALRs recruit a bipartite protein known as apoptosis-associated speck-like protein containing a caspaserecruitment domain (ASC) to induce the activation of caspase-1. This canonical inflammasome pathway is dependent on ASC for the selfproteolysis and activation of caspase-1, culminating in the cleavage of pro-IL-1β and pro-IL-18 into their secreted forms and pyroptotic cell death. In contrast to the canonical inflammasome pathway, the recently identified non-canonical pathway is mediated by caspase-11 but independent of ASC (Kayagaki et al., 2011, 2013; Hagar et al., 2013). Activation of caspase-11 by intracellular LPS induces caspase-1-independent pyroptosis as well as IL-1α/HMGB1 production and caspase-1-dependent IL-1β/IL-18 secretion. Thus, caspase-11-mediated pyroptosis may have a critical role in clearing intracellular bacteria and constitutes a potential target for modulating innate inflammatory responses (Stowe et al., 2015). Gossypol (GOS) is a polyphenolic compound existed in cottonseed oil (Gadelha et al., 2014). Early studies of GOS mainly focused on how to reduce its toxicity as it exists at high abundance in cotton seeds used for feeding animals. Subsequent studies revealed that GOS possesses many pharmacological properties, including anti-fungal, anti-inflammatory, anti-tumor, and anti-fertility activities (Turco et al., 2007; Moon et al., 2011). Mechanical studies showed that GOS causes mitochondrial dysfunction by inhibiting cell respiration and stimulation of reactive oxygen species (ROS) generation (Keshmiri-Neghab and Goliaei, 2014). It has been reported that GOS exhibits immunosuppressive effects on mouse lymphocytes in vitro and suppresses delayed-type hypersensitivity in vivo in a mouse model, which is likely mediated by inhibiting lymphocyte proliferation and inducing apoptotic cell death (Xu et al., 2009). One recent study reported that GOS inhibits the expression of proinflammatory cytokines in mouse RAW 264.7 cells through attenuating multiple signaling pathways (Huo et al., 2013). In addition, GOS has been found to inhibit tumor necrosis factor-αinduced intercellular adhesion molecule-1 expression in breast cancer cells by suppressing the NF-κB pathway (Moon et al., 2011). In view of its potential immunomodulatory activity, it is of interest to unravel the potential effect of GOS on inflammasome activation and pyroptosis in macrophages. In this study, we aimed to explore whether GOS could induce pyroptosis and release of proinflammatory cytokines in mouse peritoneal macrophages and RAW 264.7 cells. Our data indicated that GOS could induce a robust pyroptotic cell death but a weak release of mature IL-1β in peritoneal macrophages in a manner different from that of ATP. In vivo administration of gossypol had likely eliminated the peritoneal macrophages in thioglycolate-treated mice and markedly reduced the survival rate of mice in a bacterial sepsis model. Our findings highlight a concern regarding its potential cytotoxicity to macrophages by induction of pyroptosis when used in vivo.
57
MD, USA). Mouse anti-caspase-1 monoclonal antibody (clone 14F468) was obtained from Novus Biologicals (Littleton, CO, USA). Specific antibodies against IL-1β, caspase-11, HMGB1, β-tubulin, and horseradish peroxidase (HRP)-conjugated goat anti-rabbit/mouse/rat IgG were obtained from Cell Signaling Technology (Danvers, MA, USA). 2.2. Animals C57BL/6 mice (6–8 weeks of age) were purchased from the Experimental Animal Center of Southern Medical University (Guangzhou, China). All animal experiments were performed according to the guidelines for the care and use of animals approved by the Committee on the Ethics of Animal Experiments of Jinan University. 2.3. Cell line and cell culture The RAW 264.7 cells was obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and maintained in DMEM supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine (DMEM complete medium) at 37 °C in a humidified incubator of 5% CO2; and sub-cultured every 2–3 days. 2.4. Isolation of peritoneal macrophages Mice were injected with 1 ml of 3% TG medium and 4 days later killed by cervical dislocation and sterilized by 75% ethanol. Peritoneal macrophages were immediately extracted by washing the peritoneal cavity with washing buffer (sterile PBS containing 5% newborn calf serum and 0.5 mM EDTA). The extracted solution was centrifuged at 300 × g for 10 min and isolated cells were cultured at 37 °C in DMEM complete medium. After 2-h incubation, unattached cells were discarded and attached macrophages were further cultured in fresh complete medium. 2.5. Cytotoxicity assay Peritoneal macrophages, which were seeded in 24-well plates (3 × 105 cells/well), stimulated with 1 μg/ml LPS for 5 h, and subsequently incubated with GOS for indicated time periods. Cell death was measured by PI incorporation (Py et al., 2014). PI (2 μg/ml) was added to cell culture media at room temperature for 10 min, and cells were observed immediately by live imaging using a Zeiss Axio Observer D1 microscope equipped with a Zeiss LD Plan-Neofluar 20 ×/0.4 Korr M27 objective lens. Fluorescence images were captured with a Zeiss AxioCam MR R3 cooled CCD camera controlled with ZEN software (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). 2.6. Determination of soluble IL-1β
2. Materials and methods
Soluble IL-1β in cell culture supernatants was determined by Cytometric Bead Array (CBA) Mouse IL-1β Flex Set (BD Biosciences, San Jose, CA, USA) according to the manufacturer's instruction. Data were acquired using CELLQuest software on a flow cytometer (FACSCalibur; Becton Dickinson, Mountain View, CA, USA).
2.1. Chemicals and antibodies
2.7. Western blot analysis
Gossypol (GOS), propidium iodide (PI), dimethyl sulfoxide (DMSO), Hoechst 33342, and LPS (Escherichia coli O111:B4) were purchased from Sigma-Aldrich (St. Louis, MO, USA). GOS was dissolved in DMSO at 60 mM, stored at −20 °C, and working solution was freshly prepared. Z-VAD-FMK (VAD), VX-765 (VX), and Z-DEVD-FMK (Z.D.) were obtained from MedChem Express (Princeton, NJ, USA). DMEM, Opti-MEM, L-glutamine, fetal bovine serum (FBS), penicillin and streptomycin were products of Invitrogen (Carlsbad, CA, USA). Thioglycollate (TG) medium (Brewer modified) was bought from Becton Dickinson (Sparks,
Western blotting was performed as described previously (Zhang et al., 2014). Total proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by electrotransfer to polyvinylidene difluoride (PVDF) membranes (Hybond-P; GE Healthcare Life Sciences, Piscataway, NJ, USA). The membranes were blocked with blocking buffer (50 mM Tris-buffered saline (pH 7.4) containing 5% nonfat milk and 0.1% Tween-20) and then incubated overnight with indicated antibodies, followed by HRPconjugated goat anti-rabbit/mouse/rat IgG. Bands were revealed with
58
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
an enhanced chemiluminescence kit (BeyoECL Plus; Beyotime, Haimen, China) and recorded on X-ray films (Kodak, Xiamen, China). The blot images were captured by FluorChem 8000 imaging system (AlphaInnotech, San Leandro, CA, USA).
3. Results
2.8. Precipitation of soluble proteins
Although the inhibitory effect of gossypol (GOS) on proliferation of cancer cells has been extensively explored (Vela and Marzo, 2015), it is still unknown whether GOS is acutely cytotoxic to macrophages. To investigate the acute cytotoxicity of GOS on macrophages, we used TG-elicited peritoneal macrophages as a cellular model to evaluate GOS-induced cell death. Cells were attached onto 24-well plates and unprimed or primed with LPS for 5 h, followed by incubation with graded doses of GOS for different time periods in the absence of LPS. Cell death was evaluated by propidium iodide (PI) staining and live imaging was carried out with an inverted microscope. Both untreated (control) and LPS-primed cells had PI-positive staining only in a minor fraction of cells, whereas a large fraction of PI-positive cells could be observed in GOS- or ATP-treated ones (Supplementary Fig. 1), indicating a robust induction of cell death. Quantification of PI-positive cells showed that GOS induced cell death in both LPS-primed or unprimed macrophages in a dose- (Fig. 1(A)) and time-dependent (Fig. 1(B)) manner, but LPS-priming enhanced the cell death by GOS treatment (Fig. 1(A)). Morphologically, GOS-treated cells displayed rounding, swelling, and disruption of the plasma membranes, much like those treated with ATP (Supplementary Fig. 1). These results suggested that GOS induced pyroptosis-like cell death in TG-elicited peritoneal macrophages. It has been reported that caspase-1 and caspase-11 are critical proteases for the maturation of proinflammatory cytokines, such as IL-1β (de Zoete et al., 2014; Stowe et al., 2015), and induction of pyroptosis in macrophages. Caspase-1 activation and IL-1β maturation have become the typical indicators of canonical inflammasome activation, while caspase-11 activation is indicative of non-canonical inflammasome activation. To characterize GOS-induced cell death, we determined the activation of these caspases and the release of IL-1β in TG-elicited peritoneal macrophages. Western blotting was used to detect caspase1, caspase-11, and IL-1β in cells and culture supernatants. As shown in Fig. 1(C), pro-caspase-11 was expressed in low levels in unprimed cells but significantly up-regulated by LPS stimulation, whereas proIL-1β was not expressed in unprimed cells but was induced by LPS stimulation. In contrast, caspase-1 was constitutively expressed and was unaffected by LPS, consistent with previous study in mouse bone marrowderived macrophages (Kayagaki et al., 2013). In contrast to the robust release of mature IL-1β (17 kDa) in ATP-treated cells, GOS only induced pro-IL-1β but no detectable mature IL-1β release into culture supernatants when assayed by Western blotting. To corroborate this result, cytometric bead array (CBA) showed that GOS induced a much lower level of soluble IL-1β in supernatants than ATP did (Fig. 1(D)). Consistent with this, cleaved caspase-1 p10 was undetectable in GOS-treated culture supernatants but was detected in ATP-treated supernatants (Fig. 1(C)), although the band was relatively weak probably due to the low reactivity of anti-caspase-1 antibody to p10 fragment. Different from mature IL-1β release, similar levels of HMGB1 (a critical DAMP released from cells) were observed in GOS- and ATP-treated supernatants (Fig. 1(D)). Concomitant with increased cell death (Fig. 1(B)), GOS treatment time-dependently increased caspase-11 release in LPS-primed cells but not in unprimed cells (Fig. 1(C)). No overt cleavage of caspase-11 was observed either in cell lysates or in culture supernatants, which is in line with previous studies (Hagar et al., 2013). Altogether, these results suggested that GOS could induce robust pyroptotic cell death but weak activation of caspase-1 and IL-1β maturation in TG-elicited mouse peritoneal macrophages.
Cells were primed with 1 μg/ml LPS for 5 h and then stimulated in Opti-MEM with GOS or ATP for indicated time periods. Proteins in culture supernatants were precipitated with 7.2% trichloroacetic acid plus 0.15% sodium deoxy cholate as previously described (Kayagaki et al., 2011). The precipitated proteins were re-dissolved in SDS-PAGE loading buffer and subjected to Western blot analysis. 2.9. Small interfering RNA (siRNA) The siRNA (5′-GGGCAACCUUGACGAGAUAdTdT-3′) duplexes targeting mouse caspase-11 and negative control (NC) siRNA were designed and synthesized by RiboBio (Guangzhou, China). Transfection was performed using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. In brief, one day before transfection, RAW 264.7 cells were plated in 6-well plates at 30–50% confluency, and transfected with 100 nM caspase-11 siRNA or NC siRNA for 72 h, followed by stimulation with LPS for 5 h. PI staining was used to measure cell death after GOS treatment for 2 h. Western blotting was used to determine caspase-11 expression levels. 2.10. Immunofluorescence microscopy Immunofluorescence analysis was performed as previously described (Pan et al., 2015). In brief, peritoneal macrophages were cultured in glass-bottom dishes, fixed, permeabilized and immunostained with AlexaFluor488-CD11b (eBioscience, San Diego, CA, USA). Nuclei were revealed by Hoechst 33342 staining. Cells were observed using a Zeiss Axio Observer D1 microscope with a Zeiss EC Plan-Neofluar 100 ×/1.30 Oil M27 objective (Carl Zeiss). Fluorescence images were captured with a Zeiss AxioCam MR R3 cooled CCD camera controlled with ZEN software (ZEISS). 2.11. Bacterial sepsis Bacterial infection was performed as described previously (Pan et al., 2015). Briefly, E. coli strain DH5α was grown in Luria Broth (LB) media and shaken overnight at 37 °C, and then re-inoculated into fresh LB at 10% for 3 h at 37 °C. Bacteria cell density was determined using an ultraviolet–visible spectrophotometer (NanoDrop2000, Thermo Scientific) and the corresponding colony-forming units (CFU) were determined on LB agar plates. Mice were acclimated for one week and injected intraperitoneally (i.p.) with one sub-lethal dose (1 × 109 CFU) of live bacteria suspended in 0.5 ml of PBS. One hour after bacterial injection, gossypol (dissolved in 2% Tween-80 in PBS as a suspension) or vehicle (2% Tween-80 in PBS) was given intragastrically (i.g.) once. Mouse survival was monitored every 6 h for 5 days. 2.12. Statistical analysis All 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 Prism 4.0 (GraphPad Software Inc., San Diego, CA, USA). One-way ANOVA, followed by the Tukey post-hoc test was used to analyze the statistical significance among multiple groups. Kaplan– Meier survival curves were adopted for analysis of mouse survival, and the statistical difference between 2 groups was determined using the nonparametric Mann–Whitney U test. P-values b0.05 were considered statistically significant.
3.1. Gossypol induces pyroptotic cell death in TG-elicited peritoneal macrophages
3.2. GOS induces pyroptosis in RAW 264.7 macrophages Next, we explored whether GOS could induce pyroptosis in RAW 264.7 cells. This cell line is known as lacking the expression of ASC
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
59
Fig. 1. Induction of pyroptosis by gossypol (GOS) in TG-elicited peritoneal macrophages. (A and B) Cells were stimulated with or without LPS (1 μg/ml) for 5 h, followed by incubation with various concentrations of GOS for 5 h (A) or indicated time periods (B) in the absence of LPS. Cell death was assayed by propidium iodide (PI) staining and quantified by counting 6 randomly chosen fields containing around 100 cells each. Data are shown as mean ± SD (n = 6) (**P b 0.01, ***P b 0.001 versus respective control). (C) Cells were stimulated with or without LPS (1 μg/ml) for indicated time periods, followed by incubated with GOS (5 μM) for 5 h without LPS. Western blotting was used to assess the expression and secretion of caspase-11, caspase-1, and IL-1β in cell lysates and TCA-precipitated culture supernatants, respectively. β-Tubulin was used as a loading control for cell lysates. ATP (5 mM) stimulation for 5 h was used as a positive control. #nonspecific band. (D) Cells were stimulated with LPS (1 μg/ml) for 5 h followed by incubation with GOS (5 μM) or ATP (5 mM) for 5 h in the absence of LPS. IL-1β in supernatants was determined by cytometric bead array (CBA) analysis (lower panel). HMGB1 in supernatants was determined by Western blotting (upper panel). Data are shown as mean ± SD (n = 3) (**P b 0.01; ***P b 0.001).
(Verhoef et al., 2003). As ASC is an indispensable component for the activation of caspase-1 through the canonical NLRP3 inflammasome pathway (Mariathasan et al., 2004, 2006), it is a useful cellular model for studying non-canonical ASC/caspase-1-independent pyroptosis. Probably due to the lack of ASC, the canonical inflammasome in LPSprimed cells was unable to be activated by 2 mM ATP for 5 h or by 5 mM ATP for less than 2 h. However, longer treatment (N2 h) with 5 mM ATP could induce strong cell death in these cells (Supplementary Fig. 2(A) and (B) and Supplementary Fig. 3). Interestingly, GOS also induced pyroptosis in RAW 264.7 cells: it robustly induced cell death in LPS-primed cells (by 10 μM and 20 μM GOS) or unprimed cells (by 20 μM GOS) (Fig. 2(A) and Supplementary Fig. 3). Western blot analysis of cell lysates showed that pro-caspase-11 was expressed in low levels in unprimed cells but significantly up-regulated after LPS stimulation, whereas pro-IL-1β was not expressed in unprimed cells but was highly induced after LPS stimulation (Fig. 2(B)). GOS treatment caused a decrease in pro-caspase-11 and pro-IL-1β levels in the cell lysates, concomitant with a significant increase in their levels in the supernatants, similar to ATP treatment (Fig. 2(B)). However, activated caspase-1 (p10) was not detectable both in the cell lysates and supernatants, although pro-caspase-1 existed in these samples, suggesting that caspase-1 activation might not be required for GOS-induced pyroptosis.
Supporting this, only pro-IL-1β, but not mature IL-1β, could be detected by Western blotting in the GOS-treated culture supernatants. No caspase-3 activation was detected in either GOS- or ATP-treated cells (data not shown), indicating that such cell death is not caspase-3dependent apoptosis. Consistent with a previous report (Antonopoulos et al., 2013), release of pro-IL-1β and pro-caspase-1 in LPS-primed RAW 264.7 cells was earlier than pro-caspase-11 and HMGB1. Together, these results indicated that GOS-induced pyroptosis was independent of ASC-mediated canonical inflammasome activation. 3.3. Inhibitors of pan-caspase and caspase-1 subfamily suppress GOSinduced pyroptosis in RAW 264.7 cells As GOS could induce pyroptosis in ASC-deficient RAW 264.7 cells without caspase-1 activation and IL-1β maturation, we sought to further verify whether other inflammatory caspases including caspase-11 (a member of caspase-1 subfamily) were responsible for such an effect. As shown in Fig. 3(A), GOS-induced pyroptosis in unprimed cells was robustly suppressed by either the pan-caspase inhibitor Z-VAD-FMK (VAD) or the caspase-1 subfamily inhibitor VX-765 (VX) (Wannamaker et al., 2007). In cells primed with LPS, such an effect was also markedly inhibited by VAD or VX (Fig. 3(B)). However,
60
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
Fig. 2. Induction of pyroptosis by GOS in RAW 264.7 cells. (A) Cells were stimulated with or without LPS (1 μg/ml) for 5 h, and then incubated with indicated doses of GOS or ATP (5 mM) for 5 h in the absence of LPS. Cell death was assayed by propidium iodide (PI) staining. Data are shown as mean ± SD (n = 6) (*P b 0.05, ***P b 0.001 versus control). (B) Cells were stimulated with or without LPS (1 μg/ml) for 5 h, and then incubated with or without gossypol (20 μM) for indicated time periods or ATP (5 mM) for 5 h in the absence of LPS. Western blot analysis was used to evaluate the expression and secretion of caspase-11, caspase-1, IL-1β, and HMGB1 in cell lysates and TCA-precipitated culture supernatants, respectively. β-Tubulin was used as a loading control for cell lysates. #TCA-precipitated supernatants of TG-elicited peritoneal macrophages primed with LPS followed by stimulation with ATP (5 mM) for 5 h in the absence of LPS were used as a positive control.
caspase-3 inhibitor Z-DEVD-FMK had no significant effect on GOSinduced cell death either in unprimed or in LPS-primed RAW 264.7 cells (Fig. 3(C) and (D)), corroborating pyroptotic cell death rather than caspase-3-dependent apoptosis. Given that VX inhibits caspase-1 subfamily members including caspase-1 and caspase-4 (human) and that VX is more effective in inhibiting caspase-4 than caspase-1 (Wannamaker et al., 2007), it is expected to be effective in inhibiting caspase-11 (the mouse homolog of caspase-4). Thus, there was a possibility that caspase-11 might be involved in GOS-induced pyroptosis. To explore whether caspase-11 is responsible for GOS-induced pyroptosis, we knocked down its expression by siRNA. Western blotting showed that the knockdown yield for caspase-11 was approximately 50% (Supplementary Fig. 4(A)). Unexpectedly, caspase-11 knockdown had no significant influence on GOS-induced cell death when compared with negative control (Supplementary Fig. 4(B)). Nevertheless, we currently could not exclude the role of caspase-11 in GOS-induced pyroptosis as the knockdown yield was relatively low and the residual caspase-11 might be sufficient to mediate GOS-caused cell death. Further investigation using other strategies is warranted to clarify this issue.
3.4. Glycine attenuates GOS-induced pyroptosis in TG-elicited peritoneal macrophages To further explore the features of GOS-induced cell death, we next sought to explore whether glycine (Gly) and high extracellular K+ ion had any effect on this process. Gly is a nonspecific cytoprotective agent that can suppress pore-formation in plasma membrane (Frank et al., 2000). K+ efflux has been shown to be essential for canonical caspase-1 activation, which can be blocked by high extracellular KCl (Kahlenberg and Dubyak, 2004; Franchi et al., 2007). PI staining revealed that high extracellular KCl, but not Gly, could reduce ATPinduced cell death in LPS-primed peritoneal macrophages (Fig. 4), consistent with the notion that ATP induces pyroptosis through activation of P2X7 receptor ion channel and K+ efflux (Kahlenberg and Dubyak, 2004; Franchi et al., 2007). In contrast, LPS-primed macrophages pretreated with Gly, but not KCl, displayed reduced cell death when expose to GOS treatment (Fig. 4). These results suggested that GOS-induced pyroptosis was associated with pore-formation in the plasma membrane but not with K+ efflux, which seems consistent with non-
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
61
Fig. 3. Effects of caspase inhibitors on GOS-induced pyroptosis in RAW 264.7 cells. (A) Cells were pretreated with pan-caspase inhibitor Z-VAD-FMK (VAD) (50 μM) or caspase-1 subfamily inhibitor VX-765 (VX) (50 μM) for 1 h, followed by incubated with GOS (20 μM) for additional 2 h. (B) Cells were stimulated with LPS (1 μg/ml) for 5 h and then pretreated with VAD or VX for 1 h before incubation with GOS (20 μM) for another 2 h. (C) Cells were pre-treated with caspase-3 inhibitor Z-DEVD-FMK (Z.D.) (50 μM) for 1 h, followed by incubated with GOS (20 μM) for additional 2 h. (D) Cells were stimulated with LPS (1 μg/ml) for 5 h and then pre-treated with Z.D. for 1 h before incubation with GOS (20 μM) for another 2 h. Cell death was assayed by PI staining. Data are shown as mean ± SD (n = 6) (***P b 0.001).
canonical inflammasome activation (Ruhl and Broz, 2015) but different from the action of ATP. 3.5. GOS-induced pyroptosis in RAW 264.7 cells is aggravated by ROS scavenger As reactive oxygen species (ROS) has been shown to drive the activation of the canonical inflammasome, we explored whether ROS was required for GOS-induced pyroptosis in RAW 264.7 cells. PI staining was used to detect cell death in cells treated with GOS or ATP in the presence or absence of ROS inhibitor N-acetyl-L-cysteine (NAC). As expected, inhibition of ROS by NAC robustly suppressed ATP-induced cell death in both unprimed and LPS-primed cells (Fig. 5(A) and (B)). In stark contrast, NAC did not inhibit, but instead enhanced the pyroptotic cell death induced by GOS in LPS-primed and unprimed RAW 264.7 cells, indicating that ROS did not have a critical role in GOS-induced pyroptosis. These results also suggested that GOS might induce pyroptosis via a non-canonical inflammasome pathway.
Therefore we analyzed the features of peritoneal macrophages by fluorescence microscopy. Most of the peritoneal macrophages displayed a cell size around 10 μm, with a bit higher expression of CD11b on the plasma membranes as compared to its distribution in the intracellular compartments (Fig. 6(A) and (B)). However, after GOS treatment, such peritoneal macrophages were replaced by a large number of relatively smaller cells (~ 7 μm). Moreover, their CD11b was mainly
3.6. GOS eliminates peritoneal macrophages in vivo and decreases the survival of mice with bacterial sepsis As GOS could induce pyroptotic cell death in macrophages in vitro, we sought to explore whether it had any effects on TG-elicited peritoneal macrophages in vivo. Mice injected with TG were subjected to LPS stimulation followed by GOS administration intraperitoneally. The peritoneal cells were isolated 5 h after GOS treatment. As the in vivo system of animals may rapidly remove any dead cells, we were unable to observe such dead cells in freshly isolated cells (data not shown). Probably due to similar situation in the in vivo system, a previous study had established an ex vivo human lymphoid aggregate culture system to explore CD4+ T cell pyroptosis during HIV-1 infection (Doitsh et al., 2014).
Fig. 4. Effects of glycine (Gly) and high extracellular potassium ion on GOS-induced pyroptosis in TG-elicited peritoneal macrophages. Cells were stimulated with LPS (1 μg/ml) for 5 h, and pretreated with KCl (100 mM) or Gly (5 mM) for 30 min before co-treatment with GOS (5 μM) or ATP (5 mM) for 5 h in the absence of LPS. Cell death was assayed by PI staining. Data are shown as mean ± SD (n = 6) (*P b 0.05; ***P b 0.001; ns, not significant).
62
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
Fig. 5. Effect of N-acetyl-L-cysteine (NAC) on GOS-induced pyroptosis in RAW 264.7 cells. (A) Cells were pretreated with or without NAC (10 mM) for 30 min before incubation with indicated doses of GOS or ATP (5 mM) for 5 h. (B) RAW 264.7 cells were primed with LPS (1 μg/ml) for 5 h, and then treated similarly as in (A) without LPS. Cell death was assayed by PI staining and quantified by counting 6 randomly chosen fields containing around 100 cells each. Data are shown as mean ± SD (n = 6) (*P b 0.05; ***P b 0.001; ns, not significant).
distributed in the intracellular compartments and many of their nuclei were multi-lobed as revealed by Hoechst staining (Fig. 6(A)), suggesting that they were newly recruited inflammatory leukocytes including neutrophils and monocytes. This result indirectly suggested that the TG-elicited peritoneal macrophages might have undergone pyroptotic cell death after GOS-treatment and that the dead cells might have been removed by other phagocytes in the peritoneal cavity. Finally, we sought to explore the in vivo relevance of GOS-induced macrophage pyroptosis in a bacterial sepsis model. A sub-lethal dose of live bacteria caused the death of 20% mice during 5-day observation (Fig. 6(C)). Similar to the previously report (Zhao et al., 2015), GOS gavage at 20 mg/kg had no significant cytotoxicity on normal mice
(data not shown), but significantly decreased the survival of bacterialinfected ones (Fig. 6(C)), indicating that GOS treatment exacerbated bacterial sepsis in this mouse model. 4. Discussion In the present study, we found for the first time to our knowledge that GOS induced a rapid pyroptotic cell death in both unprimed and LPS-primed mouse peritoneal macrophages and RAW 264.7 cells, culminating in a robust release of pro-caspase-11, pro-IL-1β, and HMGB1 but only low levels of mature IL-1β. GOS-induced pyroptosis was independent of canonical caspase-1 activation in that it could cause cell
Fig. 6. In vivo effects of GOS on peritoneal macrophages and mouse survival in bacterial sepsis. (A) Thioglycollate-treated mice (n = 3) were injected (i.p.) with LPS (1 μg/mouse) for 3 h followed by administration (i.p.) with GOS (20 mg/kg) or vehicle (2% Tween-80 in PBS) for additional 5 h. Peritoneal cells were isolated and seeded in glass-bottom culture dishes for 2 h. After washing, the attached cells were fixed and stained with CD11b-specific antibody and Hoechst 33342. Representative images captured by fluorescence microscopy are shown. Scale bars, 10 μm. (B) Quantification of average cell diameter in each group using ZEN software. Data are shown as mean ± SD (n = 50 cells) (***P b 0.001). (C) Mice were infected (i.p.) with live E. coli. (1 × 109 CFU/mouse) for 1 h, followed by administration (i.g.) with GOS (20 mg/kg) or vehicle (2% Tween-80 in PBS) once, and were monitored for survival. Kaplan–Meier survival curves were used to analyze the data (n = 10) (nonparametric Mann–Whitney U test, *P b 0.05).
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
death in ASC-deficient RAW 264.7 cells, suggesting activation of a noncanonical inflammasome pathway in GOS-treated macrophages. Several lines of evidence in this study suggested that GOS induced pyroptosis was different from the canonical pathway activated by ATP treatment. Firstly, GOS-induced cell death in peritoneal macrophages was accompanied by the release of pro-caspase-11 and danger signal HMGB1 but not the activation of caspase-1. Secondly, being different from ATP-treated peritoneal macrophages primed with LPS, GOS treatment could only induce low levels of pro-IL-1β processing into its 17 kDa mature form, consistent with the observation that activation of caspase-1 was undetectable in these cells. Thirdly, in RAW 264.7 cells lacking ASC expression and thus having no caspase-1 activation (Verhoef et al., 2003), GOS could still induce robust pyroptosis as well as release of pro-caspase-11 and HMGB1. The release of such intracellular components was consistently correlated with cell death, whereas the release of pro-IL-1β and pro-caspase-1 was not associated with cell death induction by GOS (Fig. 2B). Fourthly, the non-specific cytoprotective glycine could attenuate GOS-induced cell death, whereas high extracellular K+ concentration did not, in line with previous studies showing that K+ efflux is dispensable for the induction of cell death via the non-canonical pathway but is required for canonical NLRP3 activation (Case et al., 2013; Ruhl and Broz, 2015). Finally, ROS scavenger NAC markedly suppressed ATP-induced cell death in RAW 264.7 cells; in contrast, GOS-induced cell death was significantly enhanced by NAC co-treatment, although the underlying mechanism still awaits further research. Thus, all the evidence supports a notion that GOS may induce pyroptosis by activating a non-canonical inflammasome pathway, which is independent of ASC-mediated caspase-1 processing. Although multiple canonical inflammasome pathways have been identified, there is only one non-canonical pathway having been well characterized currently: it is the caspase-11-dependent inflammasome pathway that senses the cytoplasmic LPS (Hagar et al., 2013; Kayagaki et al., 2013; Shi et al., 2014). Caspase-11 can induce pyroptotic cell death similarly to caspase-1 but cannot by itself induce IL-1β maturation (Shi et al., 2014). Recently, it has been reported that caspase-11 activates the canonical NLRP3 inflammasome by promoting K+ efflux leading to caspase-1 activation and IL-1β maturation (Ruhl and Broz, 2015). Based on these reports and our observations in this study, it seems possible that caspase-11 activation may be involved in GOSinduced pyroptosis. Although the result from siRNA knockdown of caspase-11 was somewhat disappointed, we could not exclude the role of caspase-11 in the induction of a non-canonical pyroptosis since the knockdown efficiency was low and the remaining caspase-11 protein after siRNA knockdown might still be sufficient to mediate GOSinduced cell death. Nevertheless, other inflammatory caspases such as caspase-12 (Stowe et al., 2015) or even caspase-8 (Antonopoulos et al., 2013) might also have a role in mediating such a non-canonical pyroptosis. Further investigation is thus needed to clarify this issue. It has been reported that long-time incubation (such as 18 h) of GOS together with LPS could inhibit the release of IL-1β in RAW 264.7 macrophages (Huo et al., 2013). However, it has been reported that RAW 264.7 cells are defective of ASC expression and caspase-1 activation (Verhoef et al., 2003). This suggests that the released IL-1β detected in the previous study (Huo et al., 2013) is possibly the pro-IL-1β form. While in our study, GOS treatment immediately (within 1 h) and robustly stimulated pro-IL-1β release from the cells primed with LPS. Possibly due to the cytotoxicity of GOS, a prolonged incubation with GOS might have suppressed the expression of pro-IL-1β and thus its release, which might account for the discrepancy between our results and the previous study. In view of the important role of pyroptosis in lethal septic response (Stowe et al., 2015), it was expected that GOS administration might exacerbate bacterial sepsis. Indeed, we observed that Gram-negative bacterial sepsis was markedly aggravated by GOS treatment. It is presumed that GOS might act as a pyroptosis-inducer to promote septic death of bacterium-infected mice, although we were unable to detect the
63
in vivo pyroptosis of peritoneal macrophages due to the dynamic status of in vivo system. Alternatively, as revealed by phenotype analysis of TG-elicited peritoneal macrophages, GOS administration might have eliminated the tissue-resident macrophages by inducing pyroptosis, though such a possibility remains to be explored. 5. Conclusion Our findings indicate that in addition to its anticancer effect via inducing apoptosis, GOS may also act as a pyroptosis-inducer in macrophages through an as-yet-unidentified non-canonical inflammasome pathway, thus having potential cytotoxic effects on host innate immune cells including macrophages. Considering that the non-canonical inflammasome pathway plays a critical role in both host defense against infection and bacterial sepsis (Jorgensen and Miao, 2015; Stowe et al., 2015), GOS-triggered pyroptosis in macrophages should be considered in its potential clinical application. Conflicts of interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found, in online version. Acknowledgments This work was supported by the grants from the National Natural Science Foundation of China (No. 81373423 and No. 81173604). The funder had no role in study design, data analysis, manuscript preparation, and decision to submit the article for publication. We thank Miss Mei-Yun Huang for her excellent help in gavaging. Appendix A. Supplementary data Supplementary data are provided in a separate file. Supplementary data associated with this article can be found in the online version, at http://dx.doi.org/10.1016/j.taap.2015.12.027. References Antonopoulos, C., El Sanadi, C., Kaiser, W.J., Mocarski, E.S., Dubyak, G.R., 2013. Proapoptotic chemotherapeutic drugs induce noncanonical processing and release of IL-1beta via caspase-8 in dendritic cells. J. Immunol. 191, 4789–4803. Case, C.L., Kohler, L.J., Lima, J.B., Strowig, T., de Zoete, M.R., Flavell, R.A., Zamboni, D.S., Roy, C.R., 2013. Caspase-11 stimulates rapid flagellin-independent pyroptosis in response to Legionella pneumophila. Proc. Natl. Acad. Sci. U. S. A. 110, 1851–1856. Doitsh, G., Galloway, N.L., Geng, X., Yang, Z., Monroe, K.M., Zepeda, O., Hunt, P.W., Hatano, H., Sowinski, S., Munoz-Arias, I., Greene, W.C., 2014. Cell death by pyroptosis drives CD4 T-cell depletion in HIV-1 infection. Nature 505, 509–514. Fink, S.L., Cookson, B.T., 2006. Caspase-1-dependent pore formation during pyroptosis leads to osmotic lysis of infected host macrophages. Cell. Microbiol. 8, 1812–1825. Franchi, L., Kanneganti, T.D., Dubyak, G.R., Nunez, G., 2007. Differential requirement of P2X7 receptor and intracellular K+ for caspase-1 activation induced by intracellular and extracellular bacteria. J. Biol. Chem. 282, 18810–18818. Frank, A., Rauen, U., de Groot, H., 2000. Protection by glycine against hypoxic injury of rat hepatocytes inhibition of ion fluxes through nonspecific leaks. J. Hepatol. 32, 58–66. Gadelha, I.C., Fonseca, N.B., Oloris, S.C., Melo, M.M., Soto-Blanco, B., 2014. Gossypol toxicity from cottonseed products. ScientificWorldJournal 2014, 231635. Hagar, J.A., Powell, D.A., Aachoui, Y., Ernst, R.K., Miao, E.A., 2013. Cytoplasmic LPS activates caspase-11: implications in TLR4-independent endotoxic shock. Science 341, 1250–1253. Huo, M., Gao, R., Jiang, L., Cui, X., Duan, L., Deng, X., Guan, S., Wei, J., Soromou, L.W., Feng, H., Chi, G., 2013. Suppression of LPS-induced inflammatory responses by gossypol in RAW 264.7 cells and mouse models. Int. Immunopharmacol. 15, 442–449. Jorgensen, I., Miao, E.A., 2015. Pyroptotic cell death defends against intracellular pathogens. Immunol. Rev. 265, 130–142. Kahlenberg, J.M., Dubyak, G.R., 2004. Mechanisms of caspase-1 activation by P2X7 receptor-mediated K+ release. Am. J. Physiol. Cell Physiol. 286, C1100–C1108.
64
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
Kayagaki, N., Warming, S., Lamkanfi, M., Vande Walle, L., Louie, S., Dong, J., Newton, K., Qu, Y., Liu, J., Heldens, S., Zhang, J., Lee, W.P., Roose-Girma, M., Dixit, V.M., 2011. Noncanonical inflammasome activation targets caspase-11. Nature 479, 117–121. Kayagaki, N., Wong, M.T., Stowe, I.B., Ramani, S.R., Gonzalez, L.C., Akashi-Takamura, S., Miyake, K., Zhang, J., Lee, W.P., Muszynski, A., Forsberg, L.S., Carlson, R.W., Dixit, V.M., 2013. Noncanonical inflammasome activation by intracellular LPS independent of TLR4. Science 341, 1246–1249. Keshmiri-Neghab, H., Goliaei, B., 2014. Therapeutic potential of gossypol: an overview. Pharm. Biol. 52, 124–128. Lamkanfi, M., Dixit, V.M., 2014. Mechanisms and functions of inflammasomes. Cell 157, 1013–1022. Man, S.M., Kanneganti, T.D., 2015. Regulation of inflammasome activation. Immunol. Rev. 265, 6–21. Mariathasan, S., Newton, K., Monack, D.M., Vucic, D., French, D.M., Lee, W.P., Roose-Girma, M., Erickson, S., Dixit, V.M., 2004. Differential activation of the inflammasome by caspase-1 adaptors ASC and Ipaf. Nature 430, 213–218. Mariathasan, S., Weiss, D.S., Newton, K., McBride, J., O'Rourke, K., Roose-Girma, M., Lee, W.P., Weinrauch, Y., Monack, D.M., Dixit, V.M., 2006. Cryopyrin activates the inflammasome in response to toxins and ATP. Nature 440, 228–232. Moon, D.O., Choi, Y.H., Moon, S.K., Kim, W.J., Kim, G.Y., 2011. Gossypol decreases tumor necrosis factor-alpha-induced intercellular adhesion molecule-1 expression via suppression of NF-kappaB activity. Food Chem. Toxicol. 49, 999–1005. Pan, H., Xu, L.H., Huang, M.Y., Zha, Q.B., Zhao, G.X., Hou, X.F., Shi, Z.J., Lin, Q.R., Ouyang, D.Y., He, X.H., 2015. Piperine metabolically regulates peritoneal resident macrophages to potentiate their functions against bacterial infection. Oncotarget 6, 32468–32483. Py, B.F., Jin, M., Desai, B.N., Penumaka, A., Zhu, H., Kober, M., Dietrich, A., Lipinski, M.M., Henry, T., Clapham, D.E., Yuan, J., 2014. Caspase-11 controls interleukin-1beta release through degradation of TRPC1. Cell Rep. 6, 1122–1128. Ruhl, S., Broz, P., 2015. Caspase-11 activates a canonical NLRP3 inflammasome by promoting K(+) efflux. Eur. J. Immunol. 45, 2927–2936. Scaffidi, P., Misteli, T., Bianchi, M.E., 2002. Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 418, 191–195.
Shi, J., Zhao, Y., Wang, Y., Gao, W., Ding, J., Li, P., Hu, L., Shao, F., 2014. Inflammatory caspases are innate immune receptors for intracellular LPS. Nature 514, 187–192. Stowe, I., Lee, B., Kayagaki, N., 2015. Caspase-11: arming the guards against bacterial infection. Immunol. Rev. 265, 75–84. Turco, E., Vizzuso, C., Franceschini, S., Ragazzi, A., Stefanini, F.M., 2007. The in vitro effect of gossypol and its interaction with salts on conidial germination and viability of Fusarium oxysporum sp. vasinfectum isolates. J. Appl. Microbiol. 103, 2370–2381. Vela, L., Marzo, I., 2015. Bcl-2 family of proteins as drug targets for cancer chemotherapy: the long way of BH3 mimetics from bench to bedside. Curr. Opin. Pharmacol. 23, 74–81. Verhoef, P.A., Estacion, M., Schilling, W., Dubyak, G.R., 2003. P2X7 receptor-dependent blebbing and the activation of Rho-effector kinases, caspases, and IL-1 beta release. J. Immunol. 170, 5728–5738. Wannamaker, W., Davies, R., Namchuk, M., Pollard, J., Ford, P., Ku, G., Decker, C., Charifson, P., Weber, P., Germann, U.A., Kuida, K., Randle, J.C., 2007. (S)-1-((S)-2-{[1-(4-amino3-chloro-phenyl)-methanoyl]-amino}-3,3-dimethyl-butanoy l)-pyrrolidine-2carboxylic acid ((2R,3S)-2-ethoxy-5-oxo-tetrahydro-furan-3-yl)-amide (VX-765), an orally available selective interleukin (IL)-converting enzyme/caspase-1 inhibitor, exhibits potent anti-inflammatory activities by inhibiting the release of IL-1beta and IL-18. J. Pharmacol. Exp. Ther. 321, 509–516. Xu, W.B., Xu, L.H., Lu, H.S., Ou-Yang, D.Y., Shi, H.J., Di, J.F., He, X.H., 2009. The immunosuppressive effect of gossypol in mice is mediated by inhibition of lymphocyte proliferation and by induction of cell apoptosis. Acta Pharmacol. Sin. 30, 597–604. Zhang, Y.T., Xu, L.H., Lu, Q., Liu, K.P., Liu, P.Y., Ji, F., Liu, X.M., Ouyang, D.Y., He, X.H., 2014. VASP activation via the Galpha13/RhoA/PKA pathway mediates cucurbitacin-Binduced actin aggregation and cofilin–actin rod formation. PLoS One 9, e93547. Zhao, G.X., Xu, L.H., Pan, H., Lin, Q.R., Huang, M.Y., Cai, J.Y., Ouyang, D.Y., He, X.H., 2015. The BH3-mimetic gossypol and noncytotoxic doses of valproic acid induce apoptosis by suppressing cyclin-A2/Akt/FOXO3a signaling. Oncotarget 6, 38952–38966. de Zoete, M.R., Palm, N.W., Zhu, S., Flavell, R.A., 2014. Inflammasomes. Cold Spring Harb. Perspect. Biol. 6, a016287.
Contents lists available at ScienceDirect
Toxicology and Applied Pharmacology journal homepage: www.elsevier.com/locate/ytaap
Gossypol induces pyroptosis in mouse macrophages via a non-canonical inflammasome pathway Qiu-Ru Lin a,1, Chen-Guang Li a,1, Qing-Bing Zha b,1, Li-Hui Xu c, Hao Pan a, Gao-Xiang Zhao a, Dong-Yun Ouyang a,⁎, Xian-Hui He a,⁎ a b c
Department of Immunobiology, College of Life Science and Technology, Jinan University, Guangzhou 510632, China Department of Fetal Medicine, The First Affiliated Hospital of Jinan University, Guangzhou 510632, China Department of Cell Biology, College of Life Science and Technology, Jinan University, Guangzhou 510632, China
a r t i c l e
i n f o
Article history: Received 31 July 2015 Revised 10 December 2015 Accepted 31 December 2015 Available online 4 January 2016 Keywords: Gossypol Pyroptosis Inflammasome HMGB1 Interleukin-1β
a b s t r a c t Gossypol, a polyphenolic compound isolated from cottonseeds, has been reported to possess many pharmacological activities, but whether it can influence inflammasome activation remains unclear. In this study, we found that in mouse macrophages, gossypol induced cell death characterized by rapid membrane rupture and robust release of HMGB1 and pro-caspase-11 comparable to ATP treatment, suggesting an induction of pyroptotic cell death. Unlike ATP, gossypol induced much low levels of mature interleukin-1β (IL-1β) secretion from mouse peritoneal macrophages primed with LPS, although it caused pro-IL-1β release similar to that of ATP. Consistent with this, activated caspase-1 responsible for pro-IL-1β maturation was undetectable in gossypol-treated peritoneal macrophages. Besides, RAW 264.7 cells lacking ASC expression and caspase-1 activation also underwent pyroptotic cell death upon gossypol treatment. In further support of pyroptosis induction, both pan-caspase inhibitor and caspase-1 subfamily inhibitor, but not caspase-3 inhibitor, could sharply suppress gossypol-induced cell death. Other canonical pyroptotic inhibitors, including potassium chloride and N-acetyl-L-cysteine, could suppress ATP-induced pyroptosis but failed to inhibit or even enhanced gossypol-induced cell death, whereas nonspecific pore-formation inhibitor glycine could attenuate this process, suggesting involvement of a non-canonical pathway. Of note, gossypol treatment eliminated thioglycollate-induced macrophages in the peritoneal cavity with recruitment of other leukocytes. Moreover, gossypol administration markedly decreased the survival of mice in a bacterial sepsis model. Collectively, these results suggested that gossypol induced pyroptosis in mouse macrophages via a non-canonical inflammasome pathway, which raises a concern for its in vivo cytotoxicity to macrophages. © 2016 Elsevier Inc. All rights reserved.
1. Introduction The inflammasome is a multimeric protein platform that is formed in innate immune cells, epithelial cells, and many other cell types to mount a coordinated molecular defense against pathogenic microbes and tissue injury (Lamkanfi and Dixit, 2014). The assembly of inflammasome complex is initiated upon the recognition of pathogenAbbreviations: ASC, apoptosis-associated speck-like protein containing a caspaserecruitment domain; GOS, gossypol; DAMPs, danger-associated molecular patterns; DMEM, Dulbecco's modified Eagle's medium; DMSO, dimethyl sulfoxide; EDTA, ethylenediamine tetra-acetic acid; FBS, fetal bovine serum; HMGB1, high-mobility group box 1; HRP, horse-radish peroxidase; IL-1β, interleukin-1β; LPS, lipopolysaccharide; NAC, N-acetyl-L-cysteine; PI, propidium iodide; ROS, reactive oxygen species; PAMPs, pathogen-associated molecular patterns; SD, standard deviation; SDS, sodium dodecyl sulfate; SDS-PAGE, SDS-polyacrylamide-gel electrophoresis; siRNA, small interfering RNA; TG, thioglycolate. ⁎ Corresponding authors. E-mail addresses: [email protected] (D.-Y. Ouyang), [email protected] (X.-H. He). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.taap.2015.12.027 0041-008X/© 2016 Elsevier Inc. All rights reserved.
associated molecular patterns (PAMPs), or danger-associated molecular patterns (DAMPs) by nucleotide-binding domain and leucine-rich repeat receptors (NLRs) or absent in melanoma 2 (AIM2)-like receptors (ALRs) (de Zoete et al., 2014). Such sophisticated mechanisms by which inflammasomes respond to danger signals lead to secretion of proinflammatory interleukin (IL)-1β and IL-18, which is crucial for clearance of infectious agents, as well as pyroptosis, an inflammatory form of cell death (Man and Kanneganti, 2015). Pyroptosis is a programmed cell death that is characterized by cell swelling, rapid plasma-membrane rupture, and release of proinflammatory contents including IL-1β, IL-18, ATP and high-mobility group box 1 (HMGB1) (Fink and Cookson, 2006), thus facilitating recruitment of neutrophils and monocytes, and leading to clearance of pathogenic microbes (Scaffidi et al., 2002). It can be induced by a wide range of PAMPs (i.e., flagellin, lipopolysaccharide (LPS), and bacterial toxins) and DAMPs (i.e., ATP, uric acid crystals, and amyloid-β fibrils) and generally mediates host defenses through the release of proinflammatory components thus being beneficial to the host during an infection. On the other hand, pyroptosis also has critical roles in the development and
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
progression of various chronic diseases, including gout, atherosclerosis, and metabolic syndrome, demonstrating the inflammatory response as a “double-edged sword” (Man and Kanneganti, 2015). It has recently been recognized that two distinct inflammasome activation pathways exist: the canonical and non-canonical pathways (Lamkanfi and Dixit, 2014; Man and Kanneganti, 2015). Under most circumstances, activated NLRs and ALRs recruit a bipartite protein known as apoptosis-associated speck-like protein containing a caspaserecruitment domain (ASC) to induce the activation of caspase-1. This canonical inflammasome pathway is dependent on ASC for the selfproteolysis and activation of caspase-1, culminating in the cleavage of pro-IL-1β and pro-IL-18 into their secreted forms and pyroptotic cell death. In contrast to the canonical inflammasome pathway, the recently identified non-canonical pathway is mediated by caspase-11 but independent of ASC (Kayagaki et al., 2011, 2013; Hagar et al., 2013). Activation of caspase-11 by intracellular LPS induces caspase-1-independent pyroptosis as well as IL-1α/HMGB1 production and caspase-1-dependent IL-1β/IL-18 secretion. Thus, caspase-11-mediated pyroptosis may have a critical role in clearing intracellular bacteria and constitutes a potential target for modulating innate inflammatory responses (Stowe et al., 2015). Gossypol (GOS) is a polyphenolic compound existed in cottonseed oil (Gadelha et al., 2014). Early studies of GOS mainly focused on how to reduce its toxicity as it exists at high abundance in cotton seeds used for feeding animals. Subsequent studies revealed that GOS possesses many pharmacological properties, including anti-fungal, anti-inflammatory, anti-tumor, and anti-fertility activities (Turco et al., 2007; Moon et al., 2011). Mechanical studies showed that GOS causes mitochondrial dysfunction by inhibiting cell respiration and stimulation of reactive oxygen species (ROS) generation (Keshmiri-Neghab and Goliaei, 2014). It has been reported that GOS exhibits immunosuppressive effects on mouse lymphocytes in vitro and suppresses delayed-type hypersensitivity in vivo in a mouse model, which is likely mediated by inhibiting lymphocyte proliferation and inducing apoptotic cell death (Xu et al., 2009). One recent study reported that GOS inhibits the expression of proinflammatory cytokines in mouse RAW 264.7 cells through attenuating multiple signaling pathways (Huo et al., 2013). In addition, GOS has been found to inhibit tumor necrosis factor-αinduced intercellular adhesion molecule-1 expression in breast cancer cells by suppressing the NF-κB pathway (Moon et al., 2011). In view of its potential immunomodulatory activity, it is of interest to unravel the potential effect of GOS on inflammasome activation and pyroptosis in macrophages. In this study, we aimed to explore whether GOS could induce pyroptosis and release of proinflammatory cytokines in mouse peritoneal macrophages and RAW 264.7 cells. Our data indicated that GOS could induce a robust pyroptotic cell death but a weak release of mature IL-1β in peritoneal macrophages in a manner different from that of ATP. In vivo administration of gossypol had likely eliminated the peritoneal macrophages in thioglycolate-treated mice and markedly reduced the survival rate of mice in a bacterial sepsis model. Our findings highlight a concern regarding its potential cytotoxicity to macrophages by induction of pyroptosis when used in vivo.
57
MD, USA). Mouse anti-caspase-1 monoclonal antibody (clone 14F468) was obtained from Novus Biologicals (Littleton, CO, USA). Specific antibodies against IL-1β, caspase-11, HMGB1, β-tubulin, and horseradish peroxidase (HRP)-conjugated goat anti-rabbit/mouse/rat IgG were obtained from Cell Signaling Technology (Danvers, MA, USA). 2.2. Animals C57BL/6 mice (6–8 weeks of age) were purchased from the Experimental Animal Center of Southern Medical University (Guangzhou, China). All animal experiments were performed according to the guidelines for the care and use of animals approved by the Committee on the Ethics of Animal Experiments of Jinan University. 2.3. Cell line and cell culture The RAW 264.7 cells was obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and maintained in DMEM supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine (DMEM complete medium) at 37 °C in a humidified incubator of 5% CO2; and sub-cultured every 2–3 days. 2.4. Isolation of peritoneal macrophages Mice were injected with 1 ml of 3% TG medium and 4 days later killed by cervical dislocation and sterilized by 75% ethanol. Peritoneal macrophages were immediately extracted by washing the peritoneal cavity with washing buffer (sterile PBS containing 5% newborn calf serum and 0.5 mM EDTA). The extracted solution was centrifuged at 300 × g for 10 min and isolated cells were cultured at 37 °C in DMEM complete medium. After 2-h incubation, unattached cells were discarded and attached macrophages were further cultured in fresh complete medium. 2.5. Cytotoxicity assay Peritoneal macrophages, which were seeded in 24-well plates (3 × 105 cells/well), stimulated with 1 μg/ml LPS for 5 h, and subsequently incubated with GOS for indicated time periods. Cell death was measured by PI incorporation (Py et al., 2014). PI (2 μg/ml) was added to cell culture media at room temperature for 10 min, and cells were observed immediately by live imaging using a Zeiss Axio Observer D1 microscope equipped with a Zeiss LD Plan-Neofluar 20 ×/0.4 Korr M27 objective lens. Fluorescence images were captured with a Zeiss AxioCam MR R3 cooled CCD camera controlled with ZEN software (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). 2.6. Determination of soluble IL-1β
2. Materials and methods
Soluble IL-1β in cell culture supernatants was determined by Cytometric Bead Array (CBA) Mouse IL-1β Flex Set (BD Biosciences, San Jose, CA, USA) according to the manufacturer's instruction. Data were acquired using CELLQuest software on a flow cytometer (FACSCalibur; Becton Dickinson, Mountain View, CA, USA).
2.1. Chemicals and antibodies
2.7. Western blot analysis
Gossypol (GOS), propidium iodide (PI), dimethyl sulfoxide (DMSO), Hoechst 33342, and LPS (Escherichia coli O111:B4) were purchased from Sigma-Aldrich (St. Louis, MO, USA). GOS was dissolved in DMSO at 60 mM, stored at −20 °C, and working solution was freshly prepared. Z-VAD-FMK (VAD), VX-765 (VX), and Z-DEVD-FMK (Z.D.) were obtained from MedChem Express (Princeton, NJ, USA). DMEM, Opti-MEM, L-glutamine, fetal bovine serum (FBS), penicillin and streptomycin were products of Invitrogen (Carlsbad, CA, USA). Thioglycollate (TG) medium (Brewer modified) was bought from Becton Dickinson (Sparks,
Western blotting was performed as described previously (Zhang et al., 2014). Total proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by electrotransfer to polyvinylidene difluoride (PVDF) membranes (Hybond-P; GE Healthcare Life Sciences, Piscataway, NJ, USA). The membranes were blocked with blocking buffer (50 mM Tris-buffered saline (pH 7.4) containing 5% nonfat milk and 0.1% Tween-20) and then incubated overnight with indicated antibodies, followed by HRPconjugated goat anti-rabbit/mouse/rat IgG. Bands were revealed with
58
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
an enhanced chemiluminescence kit (BeyoECL Plus; Beyotime, Haimen, China) and recorded on X-ray films (Kodak, Xiamen, China). The blot images were captured by FluorChem 8000 imaging system (AlphaInnotech, San Leandro, CA, USA).
3. Results
2.8. Precipitation of soluble proteins
Although the inhibitory effect of gossypol (GOS) on proliferation of cancer cells has been extensively explored (Vela and Marzo, 2015), it is still unknown whether GOS is acutely cytotoxic to macrophages. To investigate the acute cytotoxicity of GOS on macrophages, we used TG-elicited peritoneal macrophages as a cellular model to evaluate GOS-induced cell death. Cells were attached onto 24-well plates and unprimed or primed with LPS for 5 h, followed by incubation with graded doses of GOS for different time periods in the absence of LPS. Cell death was evaluated by propidium iodide (PI) staining and live imaging was carried out with an inverted microscope. Both untreated (control) and LPS-primed cells had PI-positive staining only in a minor fraction of cells, whereas a large fraction of PI-positive cells could be observed in GOS- or ATP-treated ones (Supplementary Fig. 1), indicating a robust induction of cell death. Quantification of PI-positive cells showed that GOS induced cell death in both LPS-primed or unprimed macrophages in a dose- (Fig. 1(A)) and time-dependent (Fig. 1(B)) manner, but LPS-priming enhanced the cell death by GOS treatment (Fig. 1(A)). Morphologically, GOS-treated cells displayed rounding, swelling, and disruption of the plasma membranes, much like those treated with ATP (Supplementary Fig. 1). These results suggested that GOS induced pyroptosis-like cell death in TG-elicited peritoneal macrophages. It has been reported that caspase-1 and caspase-11 are critical proteases for the maturation of proinflammatory cytokines, such as IL-1β (de Zoete et al., 2014; Stowe et al., 2015), and induction of pyroptosis in macrophages. Caspase-1 activation and IL-1β maturation have become the typical indicators of canonical inflammasome activation, while caspase-11 activation is indicative of non-canonical inflammasome activation. To characterize GOS-induced cell death, we determined the activation of these caspases and the release of IL-1β in TG-elicited peritoneal macrophages. Western blotting was used to detect caspase1, caspase-11, and IL-1β in cells and culture supernatants. As shown in Fig. 1(C), pro-caspase-11 was expressed in low levels in unprimed cells but significantly up-regulated by LPS stimulation, whereas proIL-1β was not expressed in unprimed cells but was induced by LPS stimulation. In contrast, caspase-1 was constitutively expressed and was unaffected by LPS, consistent with previous study in mouse bone marrowderived macrophages (Kayagaki et al., 2013). In contrast to the robust release of mature IL-1β (17 kDa) in ATP-treated cells, GOS only induced pro-IL-1β but no detectable mature IL-1β release into culture supernatants when assayed by Western blotting. To corroborate this result, cytometric bead array (CBA) showed that GOS induced a much lower level of soluble IL-1β in supernatants than ATP did (Fig. 1(D)). Consistent with this, cleaved caspase-1 p10 was undetectable in GOS-treated culture supernatants but was detected in ATP-treated supernatants (Fig. 1(C)), although the band was relatively weak probably due to the low reactivity of anti-caspase-1 antibody to p10 fragment. Different from mature IL-1β release, similar levels of HMGB1 (a critical DAMP released from cells) were observed in GOS- and ATP-treated supernatants (Fig. 1(D)). Concomitant with increased cell death (Fig. 1(B)), GOS treatment time-dependently increased caspase-11 release in LPS-primed cells but not in unprimed cells (Fig. 1(C)). No overt cleavage of caspase-11 was observed either in cell lysates or in culture supernatants, which is in line with previous studies (Hagar et al., 2013). Altogether, these results suggested that GOS could induce robust pyroptotic cell death but weak activation of caspase-1 and IL-1β maturation in TG-elicited mouse peritoneal macrophages.
Cells were primed with 1 μg/ml LPS for 5 h and then stimulated in Opti-MEM with GOS or ATP for indicated time periods. Proteins in culture supernatants were precipitated with 7.2% trichloroacetic acid plus 0.15% sodium deoxy cholate as previously described (Kayagaki et al., 2011). The precipitated proteins were re-dissolved in SDS-PAGE loading buffer and subjected to Western blot analysis. 2.9. Small interfering RNA (siRNA) The siRNA (5′-GGGCAACCUUGACGAGAUAdTdT-3′) duplexes targeting mouse caspase-11 and negative control (NC) siRNA were designed and synthesized by RiboBio (Guangzhou, China). Transfection was performed using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. In brief, one day before transfection, RAW 264.7 cells were plated in 6-well plates at 30–50% confluency, and transfected with 100 nM caspase-11 siRNA or NC siRNA for 72 h, followed by stimulation with LPS for 5 h. PI staining was used to measure cell death after GOS treatment for 2 h. Western blotting was used to determine caspase-11 expression levels. 2.10. Immunofluorescence microscopy Immunofluorescence analysis was performed as previously described (Pan et al., 2015). In brief, peritoneal macrophages were cultured in glass-bottom dishes, fixed, permeabilized and immunostained with AlexaFluor488-CD11b (eBioscience, San Diego, CA, USA). Nuclei were revealed by Hoechst 33342 staining. Cells were observed using a Zeiss Axio Observer D1 microscope with a Zeiss EC Plan-Neofluar 100 ×/1.30 Oil M27 objective (Carl Zeiss). Fluorescence images were captured with a Zeiss AxioCam MR R3 cooled CCD camera controlled with ZEN software (ZEISS). 2.11. Bacterial sepsis Bacterial infection was performed as described previously (Pan et al., 2015). Briefly, E. coli strain DH5α was grown in Luria Broth (LB) media and shaken overnight at 37 °C, and then re-inoculated into fresh LB at 10% for 3 h at 37 °C. Bacteria cell density was determined using an ultraviolet–visible spectrophotometer (NanoDrop2000, Thermo Scientific) and the corresponding colony-forming units (CFU) were determined on LB agar plates. Mice were acclimated for one week and injected intraperitoneally (i.p.) with one sub-lethal dose (1 × 109 CFU) of live bacteria suspended in 0.5 ml of PBS. One hour after bacterial injection, gossypol (dissolved in 2% Tween-80 in PBS as a suspension) or vehicle (2% Tween-80 in PBS) was given intragastrically (i.g.) once. Mouse survival was monitored every 6 h for 5 days. 2.12. Statistical analysis All 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 Prism 4.0 (GraphPad Software Inc., San Diego, CA, USA). One-way ANOVA, followed by the Tukey post-hoc test was used to analyze the statistical significance among multiple groups. Kaplan– Meier survival curves were adopted for analysis of mouse survival, and the statistical difference between 2 groups was determined using the nonparametric Mann–Whitney U test. P-values b0.05 were considered statistically significant.
3.1. Gossypol induces pyroptotic cell death in TG-elicited peritoneal macrophages
3.2. GOS induces pyroptosis in RAW 264.7 macrophages Next, we explored whether GOS could induce pyroptosis in RAW 264.7 cells. This cell line is known as lacking the expression of ASC
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
59
Fig. 1. Induction of pyroptosis by gossypol (GOS) in TG-elicited peritoneal macrophages. (A and B) Cells were stimulated with or without LPS (1 μg/ml) for 5 h, followed by incubation with various concentrations of GOS for 5 h (A) or indicated time periods (B) in the absence of LPS. Cell death was assayed by propidium iodide (PI) staining and quantified by counting 6 randomly chosen fields containing around 100 cells each. Data are shown as mean ± SD (n = 6) (**P b 0.01, ***P b 0.001 versus respective control). (C) Cells were stimulated with or without LPS (1 μg/ml) for indicated time periods, followed by incubated with GOS (5 μM) for 5 h without LPS. Western blotting was used to assess the expression and secretion of caspase-11, caspase-1, and IL-1β in cell lysates and TCA-precipitated culture supernatants, respectively. β-Tubulin was used as a loading control for cell lysates. ATP (5 mM) stimulation for 5 h was used as a positive control. #nonspecific band. (D) Cells were stimulated with LPS (1 μg/ml) for 5 h followed by incubation with GOS (5 μM) or ATP (5 mM) for 5 h in the absence of LPS. IL-1β in supernatants was determined by cytometric bead array (CBA) analysis (lower panel). HMGB1 in supernatants was determined by Western blotting (upper panel). Data are shown as mean ± SD (n = 3) (**P b 0.01; ***P b 0.001).
(Verhoef et al., 2003). As ASC is an indispensable component for the activation of caspase-1 through the canonical NLRP3 inflammasome pathway (Mariathasan et al., 2004, 2006), it is a useful cellular model for studying non-canonical ASC/caspase-1-independent pyroptosis. Probably due to the lack of ASC, the canonical inflammasome in LPSprimed cells was unable to be activated by 2 mM ATP for 5 h or by 5 mM ATP for less than 2 h. However, longer treatment (N2 h) with 5 mM ATP could induce strong cell death in these cells (Supplementary Fig. 2(A) and (B) and Supplementary Fig. 3). Interestingly, GOS also induced pyroptosis in RAW 264.7 cells: it robustly induced cell death in LPS-primed cells (by 10 μM and 20 μM GOS) or unprimed cells (by 20 μM GOS) (Fig. 2(A) and Supplementary Fig. 3). Western blot analysis of cell lysates showed that pro-caspase-11 was expressed in low levels in unprimed cells but significantly up-regulated after LPS stimulation, whereas pro-IL-1β was not expressed in unprimed cells but was highly induced after LPS stimulation (Fig. 2(B)). GOS treatment caused a decrease in pro-caspase-11 and pro-IL-1β levels in the cell lysates, concomitant with a significant increase in their levels in the supernatants, similar to ATP treatment (Fig. 2(B)). However, activated caspase-1 (p10) was not detectable both in the cell lysates and supernatants, although pro-caspase-1 existed in these samples, suggesting that caspase-1 activation might not be required for GOS-induced pyroptosis.
Supporting this, only pro-IL-1β, but not mature IL-1β, could be detected by Western blotting in the GOS-treated culture supernatants. No caspase-3 activation was detected in either GOS- or ATP-treated cells (data not shown), indicating that such cell death is not caspase-3dependent apoptosis. Consistent with a previous report (Antonopoulos et al., 2013), release of pro-IL-1β and pro-caspase-1 in LPS-primed RAW 264.7 cells was earlier than pro-caspase-11 and HMGB1. Together, these results indicated that GOS-induced pyroptosis was independent of ASC-mediated canonical inflammasome activation. 3.3. Inhibitors of pan-caspase and caspase-1 subfamily suppress GOSinduced pyroptosis in RAW 264.7 cells As GOS could induce pyroptosis in ASC-deficient RAW 264.7 cells without caspase-1 activation and IL-1β maturation, we sought to further verify whether other inflammatory caspases including caspase-11 (a member of caspase-1 subfamily) were responsible for such an effect. As shown in Fig. 3(A), GOS-induced pyroptosis in unprimed cells was robustly suppressed by either the pan-caspase inhibitor Z-VAD-FMK (VAD) or the caspase-1 subfamily inhibitor VX-765 (VX) (Wannamaker et al., 2007). In cells primed with LPS, such an effect was also markedly inhibited by VAD or VX (Fig. 3(B)). However,
60
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
Fig. 2. Induction of pyroptosis by GOS in RAW 264.7 cells. (A) Cells were stimulated with or without LPS (1 μg/ml) for 5 h, and then incubated with indicated doses of GOS or ATP (5 mM) for 5 h in the absence of LPS. Cell death was assayed by propidium iodide (PI) staining. Data are shown as mean ± SD (n = 6) (*P b 0.05, ***P b 0.001 versus control). (B) Cells were stimulated with or without LPS (1 μg/ml) for 5 h, and then incubated with or without gossypol (20 μM) for indicated time periods or ATP (5 mM) for 5 h in the absence of LPS. Western blot analysis was used to evaluate the expression and secretion of caspase-11, caspase-1, IL-1β, and HMGB1 in cell lysates and TCA-precipitated culture supernatants, respectively. β-Tubulin was used as a loading control for cell lysates. #TCA-precipitated supernatants of TG-elicited peritoneal macrophages primed with LPS followed by stimulation with ATP (5 mM) for 5 h in the absence of LPS were used as a positive control.
caspase-3 inhibitor Z-DEVD-FMK had no significant effect on GOSinduced cell death either in unprimed or in LPS-primed RAW 264.7 cells (Fig. 3(C) and (D)), corroborating pyroptotic cell death rather than caspase-3-dependent apoptosis. Given that VX inhibits caspase-1 subfamily members including caspase-1 and caspase-4 (human) and that VX is more effective in inhibiting caspase-4 than caspase-1 (Wannamaker et al., 2007), it is expected to be effective in inhibiting caspase-11 (the mouse homolog of caspase-4). Thus, there was a possibility that caspase-11 might be involved in GOS-induced pyroptosis. To explore whether caspase-11 is responsible for GOS-induced pyroptosis, we knocked down its expression by siRNA. Western blotting showed that the knockdown yield for caspase-11 was approximately 50% (Supplementary Fig. 4(A)). Unexpectedly, caspase-11 knockdown had no significant influence on GOS-induced cell death when compared with negative control (Supplementary Fig. 4(B)). Nevertheless, we currently could not exclude the role of caspase-11 in GOS-induced pyroptosis as the knockdown yield was relatively low and the residual caspase-11 might be sufficient to mediate GOS-caused cell death. Further investigation using other strategies is warranted to clarify this issue.
3.4. Glycine attenuates GOS-induced pyroptosis in TG-elicited peritoneal macrophages To further explore the features of GOS-induced cell death, we next sought to explore whether glycine (Gly) and high extracellular K+ ion had any effect on this process. Gly is a nonspecific cytoprotective agent that can suppress pore-formation in plasma membrane (Frank et al., 2000). K+ efflux has been shown to be essential for canonical caspase-1 activation, which can be blocked by high extracellular KCl (Kahlenberg and Dubyak, 2004; Franchi et al., 2007). PI staining revealed that high extracellular KCl, but not Gly, could reduce ATPinduced cell death in LPS-primed peritoneal macrophages (Fig. 4), consistent with the notion that ATP induces pyroptosis through activation of P2X7 receptor ion channel and K+ efflux (Kahlenberg and Dubyak, 2004; Franchi et al., 2007). In contrast, LPS-primed macrophages pretreated with Gly, but not KCl, displayed reduced cell death when expose to GOS treatment (Fig. 4). These results suggested that GOS-induced pyroptosis was associated with pore-formation in the plasma membrane but not with K+ efflux, which seems consistent with non-
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
61
Fig. 3. Effects of caspase inhibitors on GOS-induced pyroptosis in RAW 264.7 cells. (A) Cells were pretreated with pan-caspase inhibitor Z-VAD-FMK (VAD) (50 μM) or caspase-1 subfamily inhibitor VX-765 (VX) (50 μM) for 1 h, followed by incubated with GOS (20 μM) for additional 2 h. (B) Cells were stimulated with LPS (1 μg/ml) for 5 h and then pretreated with VAD or VX for 1 h before incubation with GOS (20 μM) for another 2 h. (C) Cells were pre-treated with caspase-3 inhibitor Z-DEVD-FMK (Z.D.) (50 μM) for 1 h, followed by incubated with GOS (20 μM) for additional 2 h. (D) Cells were stimulated with LPS (1 μg/ml) for 5 h and then pre-treated with Z.D. for 1 h before incubation with GOS (20 μM) for another 2 h. Cell death was assayed by PI staining. Data are shown as mean ± SD (n = 6) (***P b 0.001).
canonical inflammasome activation (Ruhl and Broz, 2015) but different from the action of ATP. 3.5. GOS-induced pyroptosis in RAW 264.7 cells is aggravated by ROS scavenger As reactive oxygen species (ROS) has been shown to drive the activation of the canonical inflammasome, we explored whether ROS was required for GOS-induced pyroptosis in RAW 264.7 cells. PI staining was used to detect cell death in cells treated with GOS or ATP in the presence or absence of ROS inhibitor N-acetyl-L-cysteine (NAC). As expected, inhibition of ROS by NAC robustly suppressed ATP-induced cell death in both unprimed and LPS-primed cells (Fig. 5(A) and (B)). In stark contrast, NAC did not inhibit, but instead enhanced the pyroptotic cell death induced by GOS in LPS-primed and unprimed RAW 264.7 cells, indicating that ROS did not have a critical role in GOS-induced pyroptosis. These results also suggested that GOS might induce pyroptosis via a non-canonical inflammasome pathway.
Therefore we analyzed the features of peritoneal macrophages by fluorescence microscopy. Most of the peritoneal macrophages displayed a cell size around 10 μm, with a bit higher expression of CD11b on the plasma membranes as compared to its distribution in the intracellular compartments (Fig. 6(A) and (B)). However, after GOS treatment, such peritoneal macrophages were replaced by a large number of relatively smaller cells (~ 7 μm). Moreover, their CD11b was mainly
3.6. GOS eliminates peritoneal macrophages in vivo and decreases the survival of mice with bacterial sepsis As GOS could induce pyroptotic cell death in macrophages in vitro, we sought to explore whether it had any effects on TG-elicited peritoneal macrophages in vivo. Mice injected with TG were subjected to LPS stimulation followed by GOS administration intraperitoneally. The peritoneal cells were isolated 5 h after GOS treatment. As the in vivo system of animals may rapidly remove any dead cells, we were unable to observe such dead cells in freshly isolated cells (data not shown). Probably due to similar situation in the in vivo system, a previous study had established an ex vivo human lymphoid aggregate culture system to explore CD4+ T cell pyroptosis during HIV-1 infection (Doitsh et al., 2014).
Fig. 4. Effects of glycine (Gly) and high extracellular potassium ion on GOS-induced pyroptosis in TG-elicited peritoneal macrophages. Cells were stimulated with LPS (1 μg/ml) for 5 h, and pretreated with KCl (100 mM) or Gly (5 mM) for 30 min before co-treatment with GOS (5 μM) or ATP (5 mM) for 5 h in the absence of LPS. Cell death was assayed by PI staining. Data are shown as mean ± SD (n = 6) (*P b 0.05; ***P b 0.001; ns, not significant).
62
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
Fig. 5. Effect of N-acetyl-L-cysteine (NAC) on GOS-induced pyroptosis in RAW 264.7 cells. (A) Cells were pretreated with or without NAC (10 mM) for 30 min before incubation with indicated doses of GOS or ATP (5 mM) for 5 h. (B) RAW 264.7 cells were primed with LPS (1 μg/ml) for 5 h, and then treated similarly as in (A) without LPS. Cell death was assayed by PI staining and quantified by counting 6 randomly chosen fields containing around 100 cells each. Data are shown as mean ± SD (n = 6) (*P b 0.05; ***P b 0.001; ns, not significant).
distributed in the intracellular compartments and many of their nuclei were multi-lobed as revealed by Hoechst staining (Fig. 6(A)), suggesting that they were newly recruited inflammatory leukocytes including neutrophils and monocytes. This result indirectly suggested that the TG-elicited peritoneal macrophages might have undergone pyroptotic cell death after GOS-treatment and that the dead cells might have been removed by other phagocytes in the peritoneal cavity. Finally, we sought to explore the in vivo relevance of GOS-induced macrophage pyroptosis in a bacterial sepsis model. A sub-lethal dose of live bacteria caused the death of 20% mice during 5-day observation (Fig. 6(C)). Similar to the previously report (Zhao et al., 2015), GOS gavage at 20 mg/kg had no significant cytotoxicity on normal mice
(data not shown), but significantly decreased the survival of bacterialinfected ones (Fig. 6(C)), indicating that GOS treatment exacerbated bacterial sepsis in this mouse model. 4. Discussion In the present study, we found for the first time to our knowledge that GOS induced a rapid pyroptotic cell death in both unprimed and LPS-primed mouse peritoneal macrophages and RAW 264.7 cells, culminating in a robust release of pro-caspase-11, pro-IL-1β, and HMGB1 but only low levels of mature IL-1β. GOS-induced pyroptosis was independent of canonical caspase-1 activation in that it could cause cell
Fig. 6. In vivo effects of GOS on peritoneal macrophages and mouse survival in bacterial sepsis. (A) Thioglycollate-treated mice (n = 3) were injected (i.p.) with LPS (1 μg/mouse) for 3 h followed by administration (i.p.) with GOS (20 mg/kg) or vehicle (2% Tween-80 in PBS) for additional 5 h. Peritoneal cells were isolated and seeded in glass-bottom culture dishes for 2 h. After washing, the attached cells were fixed and stained with CD11b-specific antibody and Hoechst 33342. Representative images captured by fluorescence microscopy are shown. Scale bars, 10 μm. (B) Quantification of average cell diameter in each group using ZEN software. Data are shown as mean ± SD (n = 50 cells) (***P b 0.001). (C) Mice were infected (i.p.) with live E. coli. (1 × 109 CFU/mouse) for 1 h, followed by administration (i.g.) with GOS (20 mg/kg) or vehicle (2% Tween-80 in PBS) once, and were monitored for survival. Kaplan–Meier survival curves were used to analyze the data (n = 10) (nonparametric Mann–Whitney U test, *P b 0.05).
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
death in ASC-deficient RAW 264.7 cells, suggesting activation of a noncanonical inflammasome pathway in GOS-treated macrophages. Several lines of evidence in this study suggested that GOS induced pyroptosis was different from the canonical pathway activated by ATP treatment. Firstly, GOS-induced cell death in peritoneal macrophages was accompanied by the release of pro-caspase-11 and danger signal HMGB1 but not the activation of caspase-1. Secondly, being different from ATP-treated peritoneal macrophages primed with LPS, GOS treatment could only induce low levels of pro-IL-1β processing into its 17 kDa mature form, consistent with the observation that activation of caspase-1 was undetectable in these cells. Thirdly, in RAW 264.7 cells lacking ASC expression and thus having no caspase-1 activation (Verhoef et al., 2003), GOS could still induce robust pyroptosis as well as release of pro-caspase-11 and HMGB1. The release of such intracellular components was consistently correlated with cell death, whereas the release of pro-IL-1β and pro-caspase-1 was not associated with cell death induction by GOS (Fig. 2B). Fourthly, the non-specific cytoprotective glycine could attenuate GOS-induced cell death, whereas high extracellular K+ concentration did not, in line with previous studies showing that K+ efflux is dispensable for the induction of cell death via the non-canonical pathway but is required for canonical NLRP3 activation (Case et al., 2013; Ruhl and Broz, 2015). Finally, ROS scavenger NAC markedly suppressed ATP-induced cell death in RAW 264.7 cells; in contrast, GOS-induced cell death was significantly enhanced by NAC co-treatment, although the underlying mechanism still awaits further research. Thus, all the evidence supports a notion that GOS may induce pyroptosis by activating a non-canonical inflammasome pathway, which is independent of ASC-mediated caspase-1 processing. Although multiple canonical inflammasome pathways have been identified, there is only one non-canonical pathway having been well characterized currently: it is the caspase-11-dependent inflammasome pathway that senses the cytoplasmic LPS (Hagar et al., 2013; Kayagaki et al., 2013; Shi et al., 2014). Caspase-11 can induce pyroptotic cell death similarly to caspase-1 but cannot by itself induce IL-1β maturation (Shi et al., 2014). Recently, it has been reported that caspase-11 activates the canonical NLRP3 inflammasome by promoting K+ efflux leading to caspase-1 activation and IL-1β maturation (Ruhl and Broz, 2015). Based on these reports and our observations in this study, it seems possible that caspase-11 activation may be involved in GOSinduced pyroptosis. Although the result from siRNA knockdown of caspase-11 was somewhat disappointed, we could not exclude the role of caspase-11 in the induction of a non-canonical pyroptosis since the knockdown efficiency was low and the remaining caspase-11 protein after siRNA knockdown might still be sufficient to mediate GOSinduced cell death. Nevertheless, other inflammatory caspases such as caspase-12 (Stowe et al., 2015) or even caspase-8 (Antonopoulos et al., 2013) might also have a role in mediating such a non-canonical pyroptosis. Further investigation is thus needed to clarify this issue. It has been reported that long-time incubation (such as 18 h) of GOS together with LPS could inhibit the release of IL-1β in RAW 264.7 macrophages (Huo et al., 2013). However, it has been reported that RAW 264.7 cells are defective of ASC expression and caspase-1 activation (Verhoef et al., 2003). This suggests that the released IL-1β detected in the previous study (Huo et al., 2013) is possibly the pro-IL-1β form. While in our study, GOS treatment immediately (within 1 h) and robustly stimulated pro-IL-1β release from the cells primed with LPS. Possibly due to the cytotoxicity of GOS, a prolonged incubation with GOS might have suppressed the expression of pro-IL-1β and thus its release, which might account for the discrepancy between our results and the previous study. In view of the important role of pyroptosis in lethal septic response (Stowe et al., 2015), it was expected that GOS administration might exacerbate bacterial sepsis. Indeed, we observed that Gram-negative bacterial sepsis was markedly aggravated by GOS treatment. It is presumed that GOS might act as a pyroptosis-inducer to promote septic death of bacterium-infected mice, although we were unable to detect the
63
in vivo pyroptosis of peritoneal macrophages due to the dynamic status of in vivo system. Alternatively, as revealed by phenotype analysis of TG-elicited peritoneal macrophages, GOS administration might have eliminated the tissue-resident macrophages by inducing pyroptosis, though such a possibility remains to be explored. 5. Conclusion Our findings indicate that in addition to its anticancer effect via inducing apoptosis, GOS may also act as a pyroptosis-inducer in macrophages through an as-yet-unidentified non-canonical inflammasome pathway, thus having potential cytotoxic effects on host innate immune cells including macrophages. Considering that the non-canonical inflammasome pathway plays a critical role in both host defense against infection and bacterial sepsis (Jorgensen and Miao, 2015; Stowe et al., 2015), GOS-triggered pyroptosis in macrophages should be considered in its potential clinical application. Conflicts of interest The authors declare that there are no conflicts of interest. Transparency Document The Transparency document associated with this article can be found, in online version. Acknowledgments This work was supported by the grants from the National Natural Science Foundation of China (No. 81373423 and No. 81173604). The funder had no role in study design, data analysis, manuscript preparation, and decision to submit the article for publication. We thank Miss Mei-Yun Huang for her excellent help in gavaging. Appendix A. Supplementary data Supplementary data are provided in a separate file. Supplementary data associated with this article can be found in the online version, at http://dx.doi.org/10.1016/j.taap.2015.12.027. References Antonopoulos, C., El Sanadi, C., Kaiser, W.J., Mocarski, E.S., Dubyak, G.R., 2013. Proapoptotic chemotherapeutic drugs induce noncanonical processing and release of IL-1beta via caspase-8 in dendritic cells. J. Immunol. 191, 4789–4803. Case, C.L., Kohler, L.J., Lima, J.B., Strowig, T., de Zoete, M.R., Flavell, R.A., Zamboni, D.S., Roy, C.R., 2013. Caspase-11 stimulates rapid flagellin-independent pyroptosis in response to Legionella pneumophila. Proc. Natl. Acad. Sci. U. S. A. 110, 1851–1856. Doitsh, G., Galloway, N.L., Geng, X., Yang, Z., Monroe, K.M., Zepeda, O., Hunt, P.W., Hatano, H., Sowinski, S., Munoz-Arias, I., Greene, W.C., 2014. Cell death by pyroptosis drives CD4 T-cell depletion in HIV-1 infection. Nature 505, 509–514. Fink, S.L., Cookson, B.T., 2006. Caspase-1-dependent pore formation during pyroptosis leads to osmotic lysis of infected host macrophages. Cell. Microbiol. 8, 1812–1825. Franchi, L., Kanneganti, T.D., Dubyak, G.R., Nunez, G., 2007. Differential requirement of P2X7 receptor and intracellular K+ for caspase-1 activation induced by intracellular and extracellular bacteria. J. Biol. Chem. 282, 18810–18818. Frank, A., Rauen, U., de Groot, H., 2000. Protection by glycine against hypoxic injury of rat hepatocytes inhibition of ion fluxes through nonspecific leaks. J. Hepatol. 32, 58–66. Gadelha, I.C., Fonseca, N.B., Oloris, S.C., Melo, M.M., Soto-Blanco, B., 2014. Gossypol toxicity from cottonseed products. ScientificWorldJournal 2014, 231635. Hagar, J.A., Powell, D.A., Aachoui, Y., Ernst, R.K., Miao, E.A., 2013. Cytoplasmic LPS activates caspase-11: implications in TLR4-independent endotoxic shock. Science 341, 1250–1253. Huo, M., Gao, R., Jiang, L., Cui, X., Duan, L., Deng, X., Guan, S., Wei, J., Soromou, L.W., Feng, H., Chi, G., 2013. Suppression of LPS-induced inflammatory responses by gossypol in RAW 264.7 cells and mouse models. Int. Immunopharmacol. 15, 442–449. Jorgensen, I., Miao, E.A., 2015. Pyroptotic cell death defends against intracellular pathogens. Immunol. Rev. 265, 130–142. Kahlenberg, J.M., Dubyak, G.R., 2004. Mechanisms of caspase-1 activation by P2X7 receptor-mediated K+ release. Am. J. Physiol. Cell Physiol. 286, C1100–C1108.
64
Q.-R. Lin et al. / Toxicology and Applied Pharmacology 292 (2016) 56–64
Kayagaki, N., Warming, S., Lamkanfi, M., Vande Walle, L., Louie, S., Dong, J., Newton, K., Qu, Y., Liu, J., Heldens, S., Zhang, J., Lee, W.P., Roose-Girma, M., Dixit, V.M., 2011. Noncanonical inflammasome activation targets caspase-11. Nature 479, 117–121. Kayagaki, N., Wong, M.T., Stowe, I.B., Ramani, S.R., Gonzalez, L.C., Akashi-Takamura, S., Miyake, K., Zhang, J., Lee, W.P., Muszynski, A., Forsberg, L.S., Carlson, R.W., Dixit, V.M., 2013. Noncanonical inflammasome activation by intracellular LPS independent of TLR4. Science 341, 1246–1249. Keshmiri-Neghab, H., Goliaei, B., 2014. Therapeutic potential of gossypol: an overview. Pharm. Biol. 52, 124–128. Lamkanfi, M., Dixit, V.M., 2014. Mechanisms and functions of inflammasomes. Cell 157, 1013–1022. Man, S.M., Kanneganti, T.D., 2015. Regulation of inflammasome activation. Immunol. Rev. 265, 6–21. Mariathasan, S., Newton, K., Monack, D.M., Vucic, D., French, D.M., Lee, W.P., Roose-Girma, M., Erickson, S., Dixit, V.M., 2004. Differential activation of the inflammasome by caspase-1 adaptors ASC and Ipaf. Nature 430, 213–218. Mariathasan, S., Weiss, D.S., Newton, K., McBride, J., O'Rourke, K., Roose-Girma, M., Lee, W.P., Weinrauch, Y., Monack, D.M., Dixit, V.M., 2006. Cryopyrin activates the inflammasome in response to toxins and ATP. Nature 440, 228–232. Moon, D.O., Choi, Y.H., Moon, S.K., Kim, W.J., Kim, G.Y., 2011. Gossypol decreases tumor necrosis factor-alpha-induced intercellular adhesion molecule-1 expression via suppression of NF-kappaB activity. Food Chem. Toxicol. 49, 999–1005. Pan, H., Xu, L.H., Huang, M.Y., Zha, Q.B., Zhao, G.X., Hou, X.F., Shi, Z.J., Lin, Q.R., Ouyang, D.Y., He, X.H., 2015. Piperine metabolically regulates peritoneal resident macrophages to potentiate their functions against bacterial infection. Oncotarget 6, 32468–32483. Py, B.F., Jin, M., Desai, B.N., Penumaka, A., Zhu, H., Kober, M., Dietrich, A., Lipinski, M.M., Henry, T., Clapham, D.E., Yuan, J., 2014. Caspase-11 controls interleukin-1beta release through degradation of TRPC1. Cell Rep. 6, 1122–1128. Ruhl, S., Broz, P., 2015. Caspase-11 activates a canonical NLRP3 inflammasome by promoting K(+) efflux. Eur. J. Immunol. 45, 2927–2936. Scaffidi, P., Misteli, T., Bianchi, M.E., 2002. Release of chromatin protein HMGB1 by necrotic cells triggers inflammation. Nature 418, 191–195.
Shi, J., Zhao, Y., Wang, Y., Gao, W., Ding, J., Li, P., Hu, L., Shao, F., 2014. Inflammatory caspases are innate immune receptors for intracellular LPS. Nature 514, 187–192. Stowe, I., Lee, B., Kayagaki, N., 2015. Caspase-11: arming the guards against bacterial infection. Immunol. Rev. 265, 75–84. Turco, E., Vizzuso, C., Franceschini, S., Ragazzi, A., Stefanini, F.M., 2007. The in vitro effect of gossypol and its interaction with salts on conidial germination and viability of Fusarium oxysporum sp. vasinfectum isolates. J. Appl. Microbiol. 103, 2370–2381. Vela, L., Marzo, I., 2015. Bcl-2 family of proteins as drug targets for cancer chemotherapy: the long way of BH3 mimetics from bench to bedside. Curr. Opin. Pharmacol. 23, 74–81. Verhoef, P.A., Estacion, M., Schilling, W., Dubyak, G.R., 2003. P2X7 receptor-dependent blebbing and the activation of Rho-effector kinases, caspases, and IL-1 beta release. J. Immunol. 170, 5728–5738. Wannamaker, W., Davies, R., Namchuk, M., Pollard, J., Ford, P., Ku, G., Decker, C., Charifson, P., Weber, P., Germann, U.A., Kuida, K., Randle, J.C., 2007. (S)-1-((S)-2-{[1-(4-amino3-chloro-phenyl)-methanoyl]-amino}-3,3-dimethyl-butanoy l)-pyrrolidine-2carboxylic acid ((2R,3S)-2-ethoxy-5-oxo-tetrahydro-furan-3-yl)-amide (VX-765), an orally available selective interleukin (IL)-converting enzyme/caspase-1 inhibitor, exhibits potent anti-inflammatory activities by inhibiting the release of IL-1beta and IL-18. J. Pharmacol. Exp. Ther. 321, 509–516. Xu, W.B., Xu, L.H., Lu, H.S., Ou-Yang, D.Y., Shi, H.J., Di, J.F., He, X.H., 2009. The immunosuppressive effect of gossypol in mice is mediated by inhibition of lymphocyte proliferation and by induction of cell apoptosis. Acta Pharmacol. Sin. 30, 597–604. Zhang, Y.T., Xu, L.H., Lu, Q., Liu, K.P., Liu, P.Y., Ji, F., Liu, X.M., Ouyang, D.Y., He, X.H., 2014. VASP activation via the Galpha13/RhoA/PKA pathway mediates cucurbitacin-Binduced actin aggregation and cofilin–actin rod formation. PLoS One 9, e93547. Zhao, G.X., Xu, L.H., Pan, H., Lin, Q.R., Huang, M.Y., Cai, J.Y., Ouyang, D.Y., He, X.H., 2015. The BH3-mimetic gossypol and noncytotoxic doses of valproic acid induce apoptosis by suppressing cyclin-A2/Akt/FOXO3a signaling. Oncotarget 6, 38952–38966. de Zoete, M.R., Palm, N.W., Zhu, S., Flavell, R.A., 2014. Inflammasomes. Cold Spring Harb. Perspect. Biol. 6, a016287.