|Home | About | Journals | Submit | Contact Us | Français|
Shigella infection, the cause of bacillary dysentery, induces caspase-1 activation and cell death in macrophages, but the precise mechanisms of this activation remain poorly understood. We demonstrate here that caspase-1 activation and IL-1β processing induced by Shigella are mediated through Ipaf, a cytosolic pattern-recognition receptor of the nucleotide-binding oligomerization domain (NOD)-like receptor (NLR) family, and the adaptor protein apoptosis-associated speck-like protein containing a C-terminal caspase recruitment domain (ASC). We also show that Ipaf was critical for pyroptosis, a specialized form of caspase-1-dependent cell death induced in macrophages by bacterial infection, whereas ASC was dispensable. Unlike that observed in Salmonella and Legionella, caspase-1 activation induced by Shigella infection was independent of flagellin. Notably, infection of macrophages with Shigella induced autophagy, which was dramatically increased by the absence of caspase-1 or Ipaf, but not ASC. Autophagy induced by Shigella required an intact bacterial type III secretion system but not VirG protein, a bacterial factor required for autophagy in epithelial-infected cells. Treatment of macrophages with 3-methyladenine, an inhibitor of autophagy, enhanced pyroptosis induced by Shigella infection, suggesting that autophagy protects infected macrophages from pyroptosis. Thus, Ipaf plays a critical role in caspase-1 activation induced by Shigella independently of flagellin. Furthermore, the absence of Ipaf or caspase-1, but not ASC, regulates pyroptosis and the induction of autophagy in Shigella-infected macrophages, providing a novel function for NLR proteins in bacterial–host interactions.
Shigella are bacterial pathogens that are the cause of bacillary dysentery known as shigellosis. A crucial aspect of the propensity of Shigella to cause diseases lies in its ability to invade the cytoplasm of epithelial cells as well as macrophages. The bacterial invasion of macrophages induces pyroptosis, the proinflammatory cell death associated with caspase-1 activation. Activated caspase-1 then cleaves and activates prointerleukin (proIL)-1β and proIL-18, which are proinflammatory cytokines involved in host inflammatory responses. However, the precise mechanisms of caspase-1 activation induced by Shigella infection remain poorly understood. Ipaf, a cytosolic pattern-recognition receptor of the nucleotide-binding oligomerization domain (NOD)-like receptor (NLR) family, is a crucial host factor that activates caspase-1 through the sensing of flagellin produced by some bacteria, such as Salmonella or Legionella. We discovered that Ipaf and the adaptor protein ASC are required for caspase-1 activation induced by non-flagellated Shigella infection. Thus, Ipaf and ASC mediate caspase-1 activation by sensing an unknown bacterial factor, but not flagellin. Autophagy, a cellular system for eliminating intracellular pathogens, was dramatically enhanced in Shigella-infected macrophages by the absence of caspase-1 or Ipaf, but not ASC. The inhibition of autophagy promoted Shigella-induced cell death, suggesting that autophagy protects infected macrophages from pyroptosis. This study provides evidence that in Shigella-infected macrophages, autophagy is inhibited by Ipaf and caspase-1, but positively regulated by ASC, providing a novel function for NLR proteins in bacterial–host interactions.
An effective immune response against microbial pathogens relies on the ability of the host to sense the presence of the infectious agent as well as the ability to destroy the invading pathogen. The presence of infection is detected through pathogen recognition molecules that sense unique microbial components called pathogen-associated molecular patterns (PAMPs) [1,2]. The recognition of bacterial PAMPs is mediated by several host molecules, including Toll-like receptors (TLRs) that are present on the cell surface and endosomal compartments, as well as nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs) that sense the presence of PAMPs in the cytosol [1,2]. The NLR protein family contains more than 20 members, including Nod1, Nod2, cryopyrin (also called as Nalp3), Nalp1, and Ipaf. NLR proteins contain C-terminal leucine-rich repeats that are linked to microbial recognition, a centrally located NOD domain that mediates oligomerization, and an N-terminal effector domain that includes caspase-activation and recruitment domain or pyrin domain [3–5]. Cryopyrin and Ipaf have been implicated in caspase-1 activation and interleukin (IL)-1β processing induced by TLR agonists, gout-associated uric acid crystals, and specific bacterial infection [6–9]. Ipaf has been shown to mediate caspase-1 activation, IL-1β processing, and caspase-1-dependent cytotoxicity induced by intracellular Salmonella or Legionella [10–14]. Caspase-1 activation and IL-1β processing induced through Nalp3 or Ipaf also required the adaptor protein ASC (apoptosis-associated speck-like protein containing a C-terminal caspase recruitment domain), which is thought to be important for the formation of the inflammasome, a multiprotein complex that mediates caspase-1 activation [6,10,15]. NLR proteins such as cryopyrin and Ipaf play a crucial role in processing mature IL-1β (also IL-18), which are important inflammatory cytokines in host defense against infection and pathogenesis of inflammatory disorders [16–18].
Shigella are highly adapted human pathogens that cause bacillary dysentery (also referred to as shigellosis). The prominent pathogenic feature of Shigella is their ability to invade a variety of host cells, including epithelial cells, macrophages, and dendritic cells, which leads to severe inflammatory responses in intestinal tissue. Internalized Shigella multiply in the cytoplasm of epithelial cells and induce actin polymerization at one pole of the bacterium, allowing intracellular bacteria to move within the cytoplasm and to spread into adjacent epithelial cells [19,20]. Shigella also invade resident macrophages in the intestinal tissue, and the bacteria escape from the phagosome into the cytosol. Infected macrophages undergo caspase-1-mediated cell death, termed pyroptosis, which is a newly identified pathway of programmed cell death associated with an inflammatory response that is accompanied by plasma membrane permeability and nuclear condensation [21–23]. In the macrophage cytosol, Shigella induce pyroptosis through activation of caspase-1 that is dependent on IpaB, a protein secreted via the type III secretion system (TTSS) or by lipopolysaccharide (LPS) moiety released from the bacteria, leading to the processing and secretion of IL-1β [23,24]. As a result, pro-inflammatory chemokines and cytokines produced by the macrophages and epithelial cells infected with Shigella elicit strong inflammation in the intestinal tissue.
Flagellin, a major subunit of the flagellum, is a prerequisite for pyroptosis and caspase-1 activation of infected macrophages with Salmonella and Legionella [11,12,14,25,26]. In the case of Salmonella, Ipaf has been shown to be essential for caspase-1 activation and pyroptosis by sensing flagellin in the cytosol [11,12]. In this study, we examined the mechanisms that regulate caspase-1 activation or cell death induced by Shigella in macrophages. Surprisingly, we find that Ipaf mediates caspase-1 activation and cell death independently of flagellin in Shigella-infected macrophages. In addition, autophagy was induced by Shigella infection, and this process was negatively regulated by Ipaf and caspase-1, but not ASC.
Flagellin is required for pyroptosis and caspase-1 activation of macrophages infected with Salmonella and Legionella [11,12,14,25,26]. On the other hand, Shigella flexneri strains are not motile and are non-flagellated bacteria. To examine whether S. flexneri express flagellin, genome sequence information from the strains 2457T  and 301  was compared in silico with that of S. enterica serovar typhimurium LT2 . Forty-two genes are associated with flagella formation in Salmonella, including fliC and fljB, two genes that encode flagellin proteins . Several genes that are known to be essential for flagellar assembly in Salmonella were absent in Shigella. Eight genes (flgC, F, K, L, flhD, fliF, J, and P) were absent in the 2457T strain, and seven genes (flgC, D, F, K, L, flhD, and fliF) were deleted in the 301 Shigella strain. In particular, we found that flhD, the master regulatory gene that controls the transcription of flagellar genes , was not present in S. flexneri. These results suggest that both flagellar assembly and flagellin expression is deficient in Shigella. Consistent with this notion, Salmonella, but not Shigella, expressed FliC by western blot analysis with anti-FliC antibody (Figure 1A). In addition, the expression of the Shigella fliC gene was not detected by RT-PCR analysis (Figure 1B). However, the open reading frame of the Shigella flagellin gene (fliC) was intact in that expression of flagellin was induced under the inducible promoter after fliC plasmid complementation (Figure 1A).
We next examined the involvement of FliC in Shigella-induced pyroptosis. Mouse bone marrow–derived macrophages (BMMs) were infected with wild-type Shigella, ΔfliC mutant, or ΔfliC mutant strain complemented with fliC (ΔfliC/fliC), and cell death was examined by lactose dehydrogenase (LDH) release in infected cells. All the strains examined had similar kinetics of LDH release (Figure 1C). Thus, FliC expression is not essential for pyroptosis induction in Shigella-infected BMMs. Moreover, infection of BMMs with wild-type Shigella and ΔfliC mutant, but not TTSS-deficient mutant (S325, mxiA::Tn5), induced caspase-1 activation and caspase-1-mediated IL-1β processing (Figure 1D and and1E).1E). These results indicate that an intact TTSS, but not flagellin, is required for caspase-1 activation and IL-1β processing in Shigella-infected BMMs.
Ipaf and ASC are required for caspase-1 activation in Salmonella-infected macrophages [10–12]. To gain insight into the molecular mechanism responsible for caspase-1 activation induced by Shigella infection, we analyzed caspase-1 activation in infected BMMs isolated from wild-type, caspase-1-deficient, Ipaf-deficient, or ASC-deficient mice (Figure 2). After infection, the processed p10 fragment from procaspase-1 was detected in wild-type BMMs, but this proteolytic cleavage was impaired in Ipaf-deficient and ASC-deficient BMMs (Figure 2C and and2E).2E). Similarly, IL-1β processing was impaired in Ipaf-deficient and ASC-deficient BMMs, but not in wild-type BMMs (Figure 2D and and2F).2F). At 2 or 3 h post infection, low levels of mature IL-1β were detected after infection of Ipaf-deficient BMMs, and at an earlier time in ASC-deficient cells, suggesting that in addition to caspase-1, other proteases can contribute to proIL-1β cleavage. This is consistent with detection of low levels of mature IL-1β in caspase-1-deficient BMMs at later time points (Figure 2B). The presence of residual proIL-1β cleavage at earlier time points in ASC-deficient macrophages may reflect increased cell death in response to Shigella when compared to Ipaf- and caspase-1-deficient macrophages (see below). It is also possible that other NLRs may contribute to caspase-1 activation in Ipaf- or ASC-deficient BMMs during Shigella infection. These events are reminiscent of the Salmonella system in which at high multiplicity of infection (MOI), there is residual IL-1β secretion that is induced independently of cytosolic flagellin and, presumably, of Ipaf . The uptake of bacteria was similar in wild-type, caspase-1-deficient, Ipaf-deficient, and ASC-deficient BMMs (unpublished data), suggesting that deficient internalization of Shigella was not responsible for the phenotype. These results indicate that Ipaf and ASC are important for Shigella-inducing caspase-1 activation and subsequent IL-1β processing. Furthermore, unlike in Salmonella, the activation of the Ipaf-ASC inflammasome is independent of flagellin.
BMMs derived from Ipaf-deficient mice, but not ASC-deficient mice, are resistant to caspase-1-dependent Salmonella-induced pyroptosis [10–12]. To investigate the role of Ipaf and ASC in Shigella-inducing pyroptosis, we analyzed LDH release from infected wild-type, Ipaf-deficient, and ASC-deficient BMMs. We found that LDH release, a marker of pyroptosis, was abrogated in Ipaf-deficient BMMs within 2 h of Shigella infection when compared to wild-type BMMs, but the release was induced after 3 h of infection and by 5 h was comparable to that of wild-type BMMs (Figure 3B). The kinetics of LDH release induced by Shigella in Ipaf-deficient BMMs was similar to that in caspase-1-deficient BMMs (Figure 3A) . However, the kinetics of LDH release induced by Shigella in ASC-deficient BMMs was indistinguishable from that observed in wild-type BMMs (Figure 3C), suggesting that ASC is not important in cell death induced by Shigella, despite ASC being required for caspase-1 activation (Figure 2E). Notably, Ipaf- and ASC-deficient BMMs did not undergo rapid LDH release when infected with the Shigella TTSS mutant (Figure 3E and and3F),3F), indicating that an intact TTSS is required for Ipaf-dependent induction of pyroptosis. Thus, Ipaf, but not ASC, is required for the rapid pyroptotic response induced by Shigella infection in BMMs. Furthermore, caspase-1 activation may be dispensable for pyroptotic cell death induced by Shigella infection in the absence of ASC.
Autophagy is induced by diverse death stimuli, including that associated with caspase-independent death, but the regulation of autophagy triggered by bacterial infection is poorly understood . In pathogen-infected cells, autophagy appears to function as a host defense mechanism that can be subverted by certain virulent bacteria to enhance their intracellular replication [33–37]. To study the role of Ipaf and ASC in autophagy, we first examined whether autophagy is induced in Shigella-infected BMMs. To assess autophagy, wild-type and caspase-1-deficient BMMs were transfected with GFP-LC3 (Atg8, a marker protein of autophagy) using a retroviral vector, and the GFP-LC3 labeling pattern was visualized by fluorescence microscopy. Autophagy induced by amino acid starvation was not affected by caspase-1, because a similar number of GFP-LC3 aggregates that are typically associated with the formation of autophagosomal vesicles were observed in wild-type and caspase-1-, Ipaf-, and ASC-deficient BMMs (Figure 4A and and4B).4B). As another approach, endogenous LC3-I to LC3-II conversion, which is an indicator of autophagosome maturation, was examined by western blotting after rapamycin treatment to induce autophagy [38–40]. Although LC3-II was more abundant than LC3-I in steady state in all BMMs, LC3 conversions (an increase of the amounts of LC3-II) were actually observed in wild-type and caspase-1-, Ipaf-, and ASC-deficient BMMs (Figure 4C), suggesting that rapamycin-induced autophagy is not affected by these genetic deficiencies. The cellular localization of GFP-LC3 was examined 30 min after infection with Shigella, since at this early time the membrane integrity of the majority of wild-type BMMs was retained (unpublished data). As shown in Figure 5A and and5B,5B, nearly 20% of intracellular Shigella was associated with accumulated GFP-LC3 in wild-type BMMs, whereas the percentage increased to about 90% in caspase-1-deficient BMMs. No accumulation of GFP alone was observed in infected cells. These results indicate that the absence of caspase-1 promotes autophagosome maturation induced by Shigella infection. Interestingly, a large number of GFP-LC3-containing vesicles, which were not associated with bacteria, were observed in caspase-1-deficient BMMs infected with Shigella (Figure 5A), suggesting that endogenous autophagy was also activated during infection. The endogenous LC-I to LC3-II conversion was also detected by Shigella infection in caspase-1-deficient BMMs (Figure 6). In epithelial cells, internalized Shigella can escape from autophagy by secreting the IcsB effector, which interferes with the interaction of host Atg5 with bacterial surface protein VirG . VirG is not only a bacterial factor essential for actin polymerization, but also a molecular target of host autophagy. Indeed, ΔvirG, a Shigella mutant lacking VirG, did not induce GFP-LC3 accumulation in epithelial cells . Notably, we found that unlike that observed in epithelial cells, ΔvirG induced a similar level of GFP-LC3 aggregates as wild-type Shigella in BMMs (Figures 5A, A,5B,5B, and and6).6). These results indicate that factors other than VirG are involved in autophagy formation in caspase-1-deficient BMMs. In both wild-type and caspase-1-deficient BMMs, GFP-LC3 accumulation around the phagocytosed Shigella TTSS mutant (S325) and endogenous LC3 conversion were not induced (Figures 5A, A,5B,5B, and and6),6), suggesting that bacterial escape from the phagosome is required for autophagosome maturation in Shigella-infected BMMs.
We next examined the role of Ipaf and ASC in autophagosome maturation in Shigella-infected BMMs. Similar to that observed in caspase-1-deficient BMMs, GFP-LC3 accumulation and endogenous LC3-I to LC3-II conversion were enhanced after Shigella infection in Ipaf-deficient BMMs when compared to wild-type cells (Figures 6, ,7A,7A, and and7B).7B). Because caspase-1 activation induced by Shigella is deficient in caspase-1- and Ipaf-deficient BMMs, these results suggested that caspase-1 activation inhibits the induction of Shigella-induced autophagy. However, when ASC-deficient BMMs were infected with Shigella, the levels of autophagy associated with intracellular bacteria were similar to those observed in wild-type BMMs (Figures 6, ,7A,7A, and and6B),6B), indicating that autophagosome maturation is not enhanced by ASC deficiency, even though caspase-1 activation is abrogated upon Shigella infection (Figure 2E). GFP-LC3-associated autophagic vesicles triggered by amino acid starvation and endogenous LC3-I to LC3-II conversion by rapamycin treatment were still induced in ASC-deficient BMMs, suggesting that the autophagic machinery is intact in the absence of ASC. Together with the results presented in Figure 3, these results indicate that Ipaf and ASC differentially regulate the induction of autophagy and suggest that autophagy in caspase-1- and Ipaf-deficient BMMs is associated with resistance to pyroptotic cell death.
To begin to examine the functional role of autophagy in Shigella-induced cell death of infected macrophages, we incubated caspase-1- or Ipaf-deficient BMMs with 3-methyladenine (MA), a well known inhibitor of autophagy, after Shigella infection. Because 3-MA inhibits uptake of bacteria by macrophages , the compound was added after phagocytosis of bacteria by BMMs (10 min after infection). As shown in Figure 8A, the addition of 3-MA did not affect pyroptosis induced by Shigella infection in wild-type BMMs. Also, viability and multiplication of intracellular Shigella were not significantly affected by addition of 3-MA under microscopic observation (unpublished data). The treatment with 3-MA enhanced the LDH release from caspase-1- and Ipaf-deficient BMMs infected with Shigella (Figure 8B and and8C),8C), suggesting that the inhibition of autophagy promotes Shigella-induced cell death in caspase-1- and Ipaf-deficient BMMs. In contrast, the addition of 3-MA did not affect LDH release from macrophages infected with the TTSS mutant (Figure 8D–8F), indicating that the cytosolic invasion is required for 3-MA to enhance membrane permeability associated with pyroptosis. These results suggest that autophagy induced by Shigella infection protect infected macrophages from pyroptosis.
Intracellular pathogenic bacteria trigger immune responses distinct from extracellular bacteria, which are mainly recognized by TLRs. Recent reports have indicated that cytosolic recognition of flagellin by Salmonella and Legionella mediates caspase-1 activation and IL-1β maturation [11,12,14,25,26]. The host protein Ipaf is required for activation of caspase-1 and IL-1β processing as well as for the inducement of rapid cell death through the sensing of intracellular flagellin during Salmonella and Legionella infection [11,12,14]. In this study, we demonstrate that Ipaf and its adaptor protein ASC are required for caspase-1 activation and IL-1β processing in Shigella-infected macrophages, but these processes, unlike in Salmonella, are independent of flagellin. The results suggest that unknown bacterial factor(s) are released from intracytosolic Shigella or secreted via the TTSS and are sensed directly or indirectly by Ipaf to promote caspase-1 activation. Shigella-induced activation of caspase-1 was previously attributed to IpaB . Since IpaB is an integral component of the TTSS transmembrane pore complex that is inserted into the host cell membrane, the ipaB mutant is unable to translocate many effector proteins via TTSS. Thus, it is difficult or impossible to definitively attribute caspase-1 activation to IpaB alone by the use of the Shigella ipaB mutant. In Salmonella, the IpaB homolog SipB has been suggested to directly interact with caspase-1 and mediate its activation . However, caspase-1 activation induced by Salmonella depends on the sensing of intracellular flagellin by Ipaf, but not SipB, in that flagellin mutants do not induce caspase-1 activation even though their SipB function is intact [11,12]. Because the TTSS in both Shigella and Salmonella forms a pore in the membrane of infected macrophages, the TTSS apparatus may induce a potassium ion efflux or another activity across the membrane, a signal that has been suggested as activating Nalp3 through the activation of the purigenic P2X7 receptor . However, Nalp3 plays no function in caspase-1 activation induced by Salmonella infection  or Shigella infection (unpublished data). It was suggested that the recognition of intracellular flagellin by Ipaf is indirect , but the molecular mechanism by which flagellin is sensed by Ipaf is unclear. Thus, it is possible that both flagellin and the Ipaf-activating factor of Shigella interact with a common signaling machinery, and this host factor(s) is sensed by Ipaf to mediate caspase-1 activation. Further studies are needed to fully understand the molecular mechanism by which intracellular Shigella induces caspase-1 activation through the Ipaf-ASC-caspase-1 inflammasome.
The induction of caspase-1-independent pyroptotic cell death by Shigella infection was induced in Ipaf-deficient BMMs as well as in caspase-1-deficient BMMs . We initially assumed that this phenotype was due to the lack of caspase-1 activity. However, ASC-deficient BMMs were not resistant to pyroptosis induced by Shigella despite the absence of caspase-1 activation. These results indicate that the function of Ipaf and ASC differ in a subtle manner and suggest that pyroptosis can proceed in the absence of caspase-1 activation. One possibility is that in caspase-1-deficient or Ipaf-deficient macrophages, anti-pyroptotic signals might be induced, leading to transient protection of macrophages from pyroptosis caused by bacterial infection. In this model, ASC might promote such a survival signal in the absence of caspase-1 or Ipaf. In certain experimental systems, ASC is known to mediate NF-κB activation [4,44], and thus NF-κB or another activity induced via ASC independently of caspase-1 might provide survival signals to counter the induction of pyroptosis in Shigella-infected macrophages.
We found that the induction of autophagy was facilitated by Shigella infection in the absence of caspase-1- or Ipaf-deficient BMMs but not in ASC-deficient BMMs. Because autophagy induced by amino acid starvation or by rapamycin treatment was normally induced in the absence of caspase-1 or Ipaf, the results indicate that the autophagic machinery is intact in the mutant cells and that caspase-1 activation inhibits autophagy formation in wild-type BMMs infected with Shigella. We hypothesized that, in wild-type macrophages, activated caspase-1 may degrade some factors that are essential for the induction of autophagy pathway, and that inhibition of autophagy and consequent rapid pyroptosis may serve to promote efficient induction of host inflammatory responses. Previous studies suggested that Naip5, another NLR family member, regulates autophagy in mouse macrophages infected with Legionella pneumophila . The mechanism by which Naip5 and Ipaf/ASC/caspase-1 regulate autophagy in response to bacterial infection remains poorly understood. However, it is likely that these NLR proteins act through different mechanisms, as recent studies suggest that Ipaf, but not Naip5, controls caspase-1 activation . The connection between caspase activation and autophagy are complex in that both events shared regulatory and mechanistic components . Our results indicate that 3-MA, an inhibitor of autophagy, enhances cell death, thus raising the possibility that autophagy induction protects macrophages from cell death caused by Shigella infection. The mechanism by which Shigella or other intracellular bacteria trigger autophagy remains poorly understood. We have found no role for Shigella VirG in the induction of autophagy, in contrast to that reported in epithelial cells infected with Shigella . It is likely that components released from intracellular bacteria into the host cytosol activate autophagy in that the Shigella TTSS mutant did not activate autophagy. Our results raise the possibility that caspase-1 activation and necrotic cell death, the two important activities for induction of pyroptosis after Shigella infection, represent independent phenomena. Our results also suggest that, at least in part, the delayed cell death observed in caspase-1-deficient BMMs may be a consequence of induction of autophagy. Further studies are needed to understand the molecular link between caspase-1 activation, pyroptosis, and autophagy, as well as their role in regulating host innate immune responses against intracellular bacteria.
The wild-type S. flexneri 2a YSH6000 strain has been described previously . Shigella mutants, S325 (mxiA::Tn5) , and a cell-to-cell spreading deficient virG null mutant (ΔvirG) , were used in this study. The fliC mutant (ΔfliC) was constructed by allele replacement strategies according to the procedures described previously . The wild-type S. enterica serovar Typhimurium SR-11 χ3181 and the isogenic fliA::Tn10 were provided by H. Matsui (Kitasato Institute for Life Science, Tokyo, Japan) . The Shigella and Salmonella strains were grown routinely in brain-heart infusion broth (Becton Dickinson, http://www.bd.com/) or Luria-Bertani broth, respectively. A fliC gene of Shigella was cloned downstream of the ptac promotor of expression vector pTB101-Tp . FliC expression was driven by adding of 10 μM IPTG in the bacterial culture for 1 h.
3-MA was purchased from Sigma (http://www.sigmaaldrich.com/). Anti-FliC antibody was provided from H. Matsui (Kitasato Institute for Life Science, Tokyo, Japan). The following antibodies were obtained commercially: rabbit anti-mouse caspase-1 (Santa Cruz Biotechnology, http://www.scbt.com/), goat anti-mouse IL-1β (R&D Systems, http://www.rndsystems.com/), and rabbit anti-LC3 antibody (MBL International, http://www.mblintl.com/).
The bacteria grown to the exponential phase were stabilized by addition of RNAprotect bacteria reagent (Qiagen, http://www.qiagen.com/). Total RNA from the cells was prepared by using RNeasy mini kit and RNAse-free DNase (Qiagen) according to the protocols of the manufacturer, and converted to cDNA with ReverTra Ace (Toyobo Life Science, http://www.toyobo.co.jp/) as a template for PCR reactions. The primers for amplification of cDNA fragments were as follows: fliC, forward primer, 5′-CGTATTAACAGCGCGAAGGA-3′, reverse primer, 5′-AGACAGAACCTGCTGCGGTA-3′; phoA, forward primer, 5′-ATGTCACGGCCGAGACTTATAG-3′, reverse primer, 5′-GTGAATATCGACGCCCAGCG-3′; ipaB, forward primer, 5′-GCAGCAGTCGTTCTCGTAGC-3′, reverse primer, 5′-TCAAGCAGTAGTTTGTTGCAAAATTG-3′.
BMMs were prepared from the femurs and tibias of caspase-1-deficient mice , Ipaf-deficient mice , and ASC-deficient mice  by culture for 5 d in 10% FCS-RPMI 1640 supplemented with 30% L-cell supernatant. All mice are C57BL/6 background or backcrossed with C57BL/6 mice. Mice were housed in a pathogen-free facility. Animal studies used protocols approved by the University of Michigan Committee on Use and Care of Animals (Ann Arbor, Michigan, United States), and the Animal Care and Use Committee of the Institute of Medical Science, University of Tokyo (Tokyo, Japan).
BMMs were seeded at 5 × 105 cells in 24-well plates containing 10% FCS-RPMI 1640. The cells were infected with Shigella at an MOI of ~10 per cell. The plates were centrifuged at 600g for 10 min to synchronize the stage of infection, and gentamicin (100 μg/ml) and kanamycin (60 μg/ml) were added 30 min later. At the times indicated after infection, the LDH activity of the culture supernatants of infected cells was measured by using a CytoTox 96 assay kit (Promega, http://www.promega.com/) according to the manufacturer's protocol. For immunofluorescence study, the infected cells were fixed and immunostained as described previously , and they were analyzed with a confocal laser-scanning microscope (LSM510; Carl Zeiss, http://www.zeiss.com/).
BMMs seeded at 1 × 106 cells in 6-well plates were infected with Shigella at an MOI of ~10 per cell. Cells were lysed and combined with supernatants precipitated with 10% trichloroacetic acid. The samples were loaded onto 15% SDS-PAGE, and the cleaved form of caspase-1 and IL-1β were detected with anti-caspase-1 or anti- IL-1β antibody, respectively.
Plat-E cells were transfected with pMX-puro-GFP or pMX-puro-GFP- rat LC3 using FuGENE 6 (Roche, http://www.roche.com/) [41,52]. Two-day cultures of BM cells were transfected with resulting retrovirus and cultured for an additional 3 d. In our experiments, GFP-transfected cells were 40%–50% and GFP-LC3-transfected cells were 30%–40% after recombinant virus infection, respectively. GFP and GFP-LC3 expression in BMMs were confirmed by the observation using a confocal laser-scanning microscope with the same threshold level. For induction of endogenous autophagy in amino acid–starved conditions, the BMMs were incubated with Earle's Balanced Salt Solution buffer (Sigma) for 2 h. To score autophagosome formation, a macrophages was defined as positive if it contained >10 donut-like shaped GFP-LC3-labeled structures. For induction of autophagy by rapamycin treatment, macrophages without retrovirus infection were incubated with rapamycin (25 μg/ml; LC Laboratories, http://www.lclabs.com/). At the indicated time, total cell lysates were prepared and analyzed by western blotting for detecting LC3-I to LC3-II conversion.
Statistical analyses were performed by the Mann–Whitney U test. Differences were considered significant at p < 0.05.
We thank Anthony Coyle, Ethan Grant, and John Bertin (Millennium Pharmaceuticals) for generous supply of mutant mice, Ichiro Nakagawa and Noboru Mizushima for helpful discussions and suggestions, and Hiroko Tsutsui for materials.
Author contributions. TS designed the research. TS, LF, CT, HA, MO, YY, and HM performed the research. TS, NI, CS, and GN analyzed the data. TS and GN wrote the paper.
Funding. This work was supported by Grant-in-Aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, and Technology in Japan, Japan Science and Technology Agency, the Naito Foundation, Takeda Science Foundation, Waksman Foundation of Japan, and by grants from the US National Institutes of Health (R01 AI064748 and R01 AI063331). LF is recipient of a postdoctoral fellowship from the Arthritis Foundation of the United States of America.
Competing interests. The authors have declared that no competing interests exist.