The multiple mechanisms of Ca2+ signalling by ... - Wiley Online Library

Cellular Microbiology (2007) 9(8), 2008–2021

doi:10.1111/j.1462-5822.2007.00932.x First published online 5 April 2007

The multiple mechanisms of Ca2+ signalling by listeriolysin O, the cholesterol-dependent cytolysin of Listeria monocytogenes

Nelson O. Gekara,1* Kathrin Westphal,1 Bin Ma,2 Manfred Rohde,3 Lothar Groebe4 and Siegfried Weiss1 Departments of 1Molecular Immunology, 2Molecular Biotechnology, 3Microbiology and 4Mucosal Immunity, Helmholtz Centre for Infection Research Inhoffenstrasse 7, 38124 Braunschweig, Germany. Summary Cholesterol-dependent cytolysins (CDCs) represent a large family of conserved pore-forming toxins produced by several Gram-positive bacteria such as Listeria monocytogenes, Streptococcus pyrogenes and Bacillus anthracis. These toxins trigger a broad range of cellular responses that greatly influence pathogenesis. Using mast cells, we demonstrate that listeriolysin O (LLO), a prototype of CDCs produced by L. monocytogenes, triggers cellular responses such as degranulation and cytokine synthesis in a Ca2+-dependent manner. Ca2+ signalling by LLO is due to Ca2+ influx from extracellular milieu and release of from intracellular stores. We show that LLO-induced release of Ca2+ from intracellular stores occurs via at least two mechanisms: (i) activation of intracellular Ca2+ channels and (ii) a Ca2+ channels independent mechanism. The former involves PLC-IP3R operated Ca2+ channels activated via G-proteins and protein tyrosine kinases. For the latter, we propose a novel mechanism of intracellular Ca2+ release involving injury of intracellular Ca2+ stores such as the endoplasmic reticulum. In addition to Ca2+ signalling, the discovery that LLO causes damage to an intracellular organelle provides a new perspective in our understanding of how CDCs affect target cells during infection by the respective bacterial pathogens.

Received 15 December, 2006; revised 31 January, 2007; accepted 1 February, 2007. *For correspondence. E-mail nelson.gekara@ helmholtz-hzi.de; Tel. (+49) 531 6181 5108; Fax (+49) 531 6181 51002. © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd

Introduction Transient increase in cytosolic-free Ca2+ is a general process used by eukaryotic cells to regulate many cellular functions. Essentially two mechanisms are responsible for this: Ca2+ influx across the plasma membrane and/or release from intracellular Ca2+ stores. In recent years, calcium flux induction has emerged as a widespread mechanism by which pathogenic bacteria influence host cells (Tran Van-Nhieu et al., 2004). Alterations of metabolism, activation of apoptosis, induction of pro-inflammatory mediators as well as cytoskeletal reorganization have been reported in this context (Tran Van-Nhieu et al., 2004). A prominent example of bacterial factors known to participate in bacteria-induced Ca2+ signalling are poreforming toxins (Tran Van-Nhieu et al., 2004). This has largely been attributed to influx of extracellular Ca2+ via the pores (Repp et al., 2002; Tran Van-Nhieu et al., 2004). Cholesterol-dependent cytolysins (CDCs) represent an expansive family of highly conserved pore-forming toxins produced by several Gram-positive bacterial species. They include listeriolysin O (LLO), perfringolysin O, pneumolysin and streptolysin O (Billington et al., 2000; Palmer, 2001). Although evidence exist that these toxins play a role in pathogenesis (Berry et al., 1989; Boulnois et al., 1991; Van Epps and Andersen, 1974), the most compelling evidence for the direct involvement of a CDC is established for LLO, in listeriosis (Gaillard et al., 1986; Kathariou et al., 1987; Portnoy et al., 1988). One of the roles of LLO is allowing the bacteria to breach the phagosomal membrane. Of equal importance are the signalling cascades that LLO triggers (Kayal et al., 1999; Dramsi and Cossart, 2003; Carrero et al., 2004a,b). Lipid rafts aggregation and subsequent tyrosine phosphorylation has been described recently by us (Gekara and Weiss, 2004; Gekara et al., 2005). In addition, Ca2+ signal induction by LLO has been reported (Wadsworth and Goldfine, 1999; Billington et al., 2000; Dramsi and Cossart, 2003). Ca2+ signalling by CDCs and for that matter LLO has mainly been ascribed to Ca2+ influx from the extracellular milieu (Repp et al., 2002). Whether these toxins also induce Ca2+ release from intracellular stores without influx

Multiple mechanisms of Ca2+ signalling by LLO 2009 Fig. 1. Ca2+ signalling by L. monocytogenes and LLO involves influx as well as release of Ca2+ from intracellular stores. Oscillatory Ca2+ response in RBL-2H3 cells exposed to L. monocytogenes (blue) as monitored by confocal microscopy. i–viii show single frames from an image series of the intracellular Ca2+ concentration at different times after exposure to L. monocytogenes. Ca2+ elevations are represented by a colour shift from green to red. The traces 1, 2 and 3 in B represent the time-course profiles of intracellular Ca2+ levels of the corresponding cells 1, 2, 3 in A. The arrow heads in B indicate the time points corresponding to the images i–viii in A. (See Video S1) Ca2+ signals are represented in arbitrary fluorescence units (a.f.u.), where a.f.u. indicate F405/F460 ratios calibrated into arbitrary fluorescence. Scale bar; 4 mm.

of extracellular Ca2+ is not clear. The aim of this study was to characterize mechanisms of Ca2+ induction by LLO in details. In particular, the role of the intracellular Ca2+ stores was addressed. The data herein demonstrate that in addition to influx, LLO triggers Ca2+ release from intracellular stores, hence triggering cellular responses such as degranulation and cytokine synthesis. We show that Ca2+ release is mediated by LLO via Ca2+ channeldependent and -independent mechanisms. The former involves activation of the phospholipase C-inositol triphosphate receptor (PLC-IP3R) pathways via tyrosine kinases and G-proteins. For the latter, our data suggest a hitherto undescribed mechanism of intracellular Ca2+ release involving injury of Ca2+-rich intracellular organelles such as the endoplasmic reticulum (ER). Results Listeria monocytogenes releases Ca2+ from intracellular stores in an LLO-dependent manner To characterize mechanisms of Ca2+ signalling in host cells by L. monocytogenes (L.m) and LLO, we used mast cells. Mast cells were considered an ideal cell type for functional studies on Ca2+ induction because. (i) They are

known to act as sentinels at host–environment interfaces hence play important roles in host defence against various pathogens (Marshall, 2004). In particular, mast cells have been found to have a protective role in a peritonitis model of L.m infection (Gekara et al. manuscript submitted; Edelson, 2004; Edelson, 2006). (ii) They rapidly secrete several preformed pro-inflammatory mediators stored in their granules via degranulation and can freshly synthesize such mediator in response to stimuli such as Ca2+ mobilization. First, we employed confocal microscopy to monitor cytosolic Ca2+ elevation at single cell level using the adherent mast cell line RBL-2H3. Addition of L.m to RBL2H3 in Ca2+ containing medium resulted in gradual elevations of intracellular Ca2+ with a transitory pattern (Fig. 1 and Video S1). Intracellular Ca2+ mobilization began soon after cells came in contact with bacteria and was sustained for at least 60 min. To characterize the role of extracellular and intracellular Ca2+ in such response, Ca2+ signals were analysed in Ca2+-free medium. Under these conditions, exposure of cells to L.m-elicited cytosolic Ca2+ elevation, although at lower levels and without the transients observed before (Fig. 2A). This suggested that the magnitude of Ca2+

© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2008–2021

2010 N. O. Gekara et al. unlike the listerial phospholipases, LLO is an indispensable player in Ca2+ signal induction by L.m. We therefore decided to characterize the mechanisms of intracellular Ca2+ release by LLO. For that, we resorted to flow cytometry to analyse Ca2+ induction in primary bone marrow-derived mast cells (BMMCs) by purified LLO. To exclude contamination by LPS or listerial phospholipases, LLO was purified from LLO-overexpressing Listeria innocua, a listerial strain not synthesizing any of such components. Like L.m (Fig. 1A), LLO was found to trigger cytosolic Ca2+ elevation in BMMCs in the presence or absence of extracellular Ca2+. As expected, the Ca2+ levels induced in the absence of extracellular Ca2+ were lower (Fig. 3). Preincubation of LLO with cholesterol abrogates pore formation but not the toxin’s ability to bind to host cells and to activate pore-independent signals (Jacobs et al., 1998; Kayal et al., 1999; Gekara et al., 2005). Cholesterolinactivated LLO (CL-LLO) did not induce any Ca2+ flux in BMMCs (Fig. 3). This suggested that influx and release of Ca2+ from intracellular stores is exclusively poredependent. As a control, the calcium ionophore ionomycin also induced Ca2+ elevation in normal and Ca2+-free medium but with faster kinetics (Fig. 3). The role of intracellular stores in overall calcium response was also confirmed by stimulating cells with LLO after chelating or depleting intracellular Ca2+ with Fig. 2. LLO is indispensable for Ca2+ release from intracellular stores by L. monocytogenes. A. Induction of cytosolic Ca2+ elevation in individual RBL-2H3 cells by L. monocytogenes in normal (+Ca) or Ca2+-free (–Ca) medium monitored by confocal microscopy. The traces show calcium signals from individual cells. B. L. m induces Ca2+ release from intracellular stores independently of PI-PLC and PC-PLC. Each of the traces in B represent the average response of 8 cells. Note that the oscillatory pattern observed with L. m + Ca in A is masked in B due to averaging. As emphasized and for reasons stated in Experimental procedure, note that EGTA was used in the washing steps after indo-1AM labelling but not during calcium measurements.

elevation elicited by L.m is a product of Ca2+ influx from extracellular medium as well as release from intracellular stores. Listeriolysin O in synergy with listerial phospholipases PI-PLC and PC-PLC encoded by plcA and plcB, respectively, has been proposed to play a role in intracellular Ca2+ release by L.m (Sibelius et al., 1996; Wadsworth and Goldfine, 1999). When tested, the mutant strain deficient in both plcA and plcB (L.mDplcA/B) was found to mobilize Ca2+ in normal and Ca2+-free medium (Fig. 2B and data not shown). However, consistent with a role of PI-PLC and PC-PLC, Ca2+ elevation by L.mDplcA/B was lower than that by wild-type L.m. This was in contrast to the LLOdeficient mutant (L.mDhly) which did not induce Ca2+ elevation at all (Fig. 2B). These results suggested that

Fig. 3. Purified LLO induces influx and release of Ca2+ from intracellular stores. Indo 1-AM loaded BMMCs cells suspended in normal or Ca2+-free medium were analysed by flow cytometry before and after stimulation with LLO (0.25 mg ml-1) or ionomycin (1 mM). Each trace represents the average of intracellular Ca2+ levels in the cells during the time of acquisition. Ca2+ signals are represented as fold increase in a.f.u. over the 30 s prestimulation baseline. The arrow indicates the time point of stimulation. The data are representative of least four independent experiments.


Multiple mechanisms of Ca2+ signalling by LLO 2011

Fig. 4. Listeria monocytogenes and LLO induce de novo synthesis and secretion of pro-inflammatory factors in mast cells in a calcium-dependent manner. A. BMMCs exposed to either L monocytogenes, or a sublytic dose (0.25 mg ml-1) of LLO or cholesterol-inactivated LLO (CL-LLO) for 1 h were analysed for degranulation by measuring the activity of b-hexosaminidase in culture supernatants. Note that the slight elevation of b-hexosaminidase by CL-LLO over the –Ctrl could have been due to incomplete toxin inactivation. *P < 0.05, **P < 0.005 and compared –Ctrl. B. Culture supernatants of BMMCs exposed to L. monocytogenes or LLO (0.25 mg ml-1) for 3 h were tested for secreted TNF-a in a bioassay. *P < 0.001 compared with –Ctrl. ***P < 0.001. C–E. (C) TNF-a transcriptional activation by LLO is pore-dependent, involves NFAT activation, and can be induced independently of extracellular Ca2+. BMMCs were stimulated with 0.25 mg ml-1 of LLO or CL-LLO for 3 h in normal (+Ca) or Ca2+-free (–Ca) medium then analysed for TNF-a mRNA upregulation by RT-PCR. The mRNA of the house-keeping gene RPS9 was used as an internal control for the amount of cDNA used (C and E). (D) BMMCs were stimulated or not with 0.25 mg ml-1 LLO or 1 mM ionomycin for 45 min at 37°C in normal or Ca2+-free medium, then stained with an anti-NFATc1 antibody. The arrows indicate NFATc1 accumulated in the nuclei. (E) Transcriptional activation of TNF-a by LLO can be blocked by cyclosporin A (CSA). BMMCs in Ca2+-free medium were either unstimulated or stimulated with 0.25 mg ml-1 LLO or 1 mM ionomycin for 3 h in the presence or absence of CSA (1 mM) after which samples were analysed for TNF-a mRNA upregulation by RT-PCR. Scale bar; 2 mm.

BAPTA-AM or thapsigargin respectively. As expected, a diminished Ca2+ response was obtained in BAPTA-AM and thapsigargin pretreated cells (Fig. S1 and data not shown). Listeriolysin O and L.m induce secretion and de novo synthesis of pro-inflammatory factors in a Ca2+-dependent manner To investigate the functional consequence of Ca2+ flux induction by L.m or LLO, we exposed BMMCs to L.m or

LLO and then determined degranulation by measuring release of b-hexosaminidase, an enzyme stored in mast cell granules. Upon exposure of BMMCs to L.m or LLO, b-hexosaminidase became readily detectable in supernatants (Fig. 4A). Release was due to the pore-forming activity of LLO because neither CL-LLO nor the LLOdeficient mutant L.mDhly triggered this response (Fig. 4A and data not shown). Consistent with the Ca2+ responses displayed in Figs 2 and 3, degranulation induced by LLO or L.m was diminished in Ca2+-free medium (Fig. 4A), indicating that


2012 N. O. Gekara et al. although intracellular Ca2+ release by LLO can induce degranulation, a combination of influx and release of Ca2+ from intracellular stores is required for an optimal response. Tumour necrosis factor alpha (TNF-a) is one of the presynthesized mediators stored by mast cells (Gordon and Galli, 1991; Marshall, 2004). This cytokine is very important in the control of L.m infection. Mice that lack TNF-a or its receptor succumb quickly to L.m (Pfeffer et al., 1993; Rothe et al., 1993; Pasparakis et al., 1996). TNF-a has also has been reported to play a role in the recruitment of various inflammatory cells during infection by various pathogens (Malaviya et al., 1996; McLachlan et al., 2003). Upon activation, mast cells undergo degranulation to rapidly release the preformed mediators such as TNF-a. Independent of degranulation, mast cells also synthesize such mediators when triggered with the appropriate stimuli. Therefore, as a read out of calcium-dependent LLOinduced activity, we decided to analyse mast cells for secretion and transcription of TNF-a in response to LLO or L.m. After 3 h stimulation, LLO and L.m had induced secretion of TNF-a, which increased up to 24 h (Fig. 4B and data not shown). We then addressed whether Ca2+ mobilization stimulated by LLO plays a role in TNF-a transcription. BMMCs stimulated with LLO in normal or Ca2+-free medium upregulated TNF-a mRNA while CL-LLO, which does not induce Ca2+ signals, elicited minimal or no response (Fig. 4C). This suggested that the pore-forming activity of LLO and by extension Ca2+ signals induced thereof, might play a significant role in TNF-a gene activation. The calcium-dependent family of transcription factors, nuclear factor of activated T cells (NFAT) is known to regulate the transcription of a broad range of cytokines and chemokines. To determine whether TNF-a gene activation by LLO involves NFAT, we tested whether LLO activates nuclear translocation of this factor. LLO-induced NFATc1 nuclear translocation when stimulated in normal as well as Ca2+-free medium (Fig. 4D). Thus, Ca2+ released from intracellular stores by LLO is sufficient to activate NFAT. To explicitly demonstrate the role of Ca2+ and NFAT in the LLO-induced transcription of TNF-a, BMMCs in Ca2+free medium were stimulated with LLO in the presence of cyclosporine A (CSA). CSA is an inhibitor of calcineurin – a cytosolic phosphatase that activates NFAT in the presence of Ca2+. As shown in Fig. 4E, stimulation of BMMCs with LLO in presence of CSA completely inhibited upregulation of TNF-a mRNA. Together, the data implied that Ca2+ release from intracellular stores by LLO was sufficient to drive TNF-a expression, which involves NFAT activation.

Listeriolysin O induces Ca2+ release from intracellular stores via Ca2+ channel-dependent and -independent mechanisms In view of the significance of Ca2+ release from intracellular stores by LLO in cytokine synthesis and secretion, next we sought to elucidate the underlying mechanisms. One possible mechanism via which LLO could release Ca2+ from intracellular stores is Ca2+-induced Ca2+ release following the initial influx of Ca2+ via the toxin pores. However, the fact that LLO could induce cytosolic Ca2+ elevation in the absence of extracellular Ca2+ was indicative of mechanism(s) not-dependent on the initial influx of Ca2+. To find out whether this mechanism(s) of Ca2+ release involve Ca2+ channels, first, the effect of Lanthanum chloride (LaCl3), a non-specific Ca2+ channel blocker on cytosolic Ca2+ elevation was analysed. In presence of extracellular Ca2+, LaCl3 partially inhibited cytosolic Ca2+ elevation by LLO (data not shown) hence suggesting a role of Ca2+ channels. When cells were analysed in Ca2+free medium, a strong inhibitory effect of LaCl3 was also observed (Fig. 5A). Surprisingly, this inhibitory effect was only partial despite using saturating LaCl3 concentrations. Thus, although indicating an important role of Ca2+ channels in Ca2+ release from intracellular stores, these results also pointed towards an additional mechanism notdependent on Ca2+ channels. Ca2+ release from intracellular stores by many agonists is induced via PLC. PLC generates inositol triphosphate (IP3), which then activates the IP3 receptor (IP3R) Ca2+ channels. To assess whether LLO-induced Ca2+ release from intracellular stores involves this pathway, BMMCs in Ca2+-free medium were stimulated with LLO in the presence of PLC or IP3R inhibitors U73122 and 2-APB (2-aminoethoxydiphenyl borate). Both agents blocked Ca2+ release (Fig. 5B). However, like LaCl3, inhibition by these compounds was incomplete. In contrast, Ca2+ release by Na3VO4 that triggers release via PLC-IP3Rgated stores, was completely inhibited by 2-APB and U73122 (Fig. 5C and data not depicted). Hence, the residual Ca2+ released by LLO in the presence of the above inhibitors was not due to incomplete inhibition of the PLC-IP3R pathway.

Ca2+ release from intracellular stores by LLO partly involves protein tyrosine phosphorylation and G-protein activation Next, we sought to investigate upstream signalling pathways by which LLO activates the PLC-IP3R pathway. Generation of IP3 is mediated by several PLC isoforms activated via different mechanisms such as protein phosphorylation (PLC-g) and G-protein activation (PLC-b) (Berridge et al., 2003). Our recent study showed that LLO


Multiple mechanisms of Ca2+ signalling by LLO 2013 antiphosphotyrosine followed by immunoblotting with antiPLC-g1 antibodies revealed PLC-g1 as one of the proteins tyrosine phosphorylated following stimulation of BMMCs with LLO (Fig. 6B). For confirmation, cell lysates from a separate experiment were immunoprecipitated with antiPLC-g1 before immunoblotting with antiphosphotyrosine (Fig. S2). To directly demonstrate the role of protein tyrosine phosphorylation in intracellular Ca2+ release by LLO, the effect of the tyrosine kinase inhibitor genistein was tested. Genistein inhibited Ca2+ mobilization but only partially as seen before (Fig. 6C). Release of Ca2+ from intracellular stores via G-proteindependent activation of PLC-b was reported for other pore-forming toxins (Krause et al., 1998). To test the role of this pathway, we stimulated BMMCs with LLO in the presence of pertussis toxin, a G-protein inhibitor. This toxin impaired Ca2+ release from intracellular stores but again only partially (Fig. 6D). Apparently, LaCl3 and U73122, the Ca2+ channel and PLC inhibitors, respectively, were more effective inhibitors than pertussis toxin (Fig. 6D), suggesting that the latter mechanism might only account for a fraction of Ca2+ released via PLC-IP3R operated Ca2+ channels. However, even with combined inhibition of G-proteins and tyrosine kinases using pertussis toxin together with genistein, LLO was still able to trigger residual release of Ca2+ from intracellular stores. Apparently both compounds were more effective when used in combination than separately. (Fig. 6D). Collectively, these results show that PLC-IP3Rdependent Ca2+ release from intracellular stores by LLO is triggered via at least two independent mechanisms; (i) protein tyrosine phosphorylation and (ii) G protein activation. They also demonstrate that in addition to PLCIP3R operated Ca2+ channels, LLO mediates Ca2+ release via Ca2+ channel-independent mechanism. Fig. 5. Ca2+ release from intracellular stores by LLO partly involves the PLC-IP3R operated Ca2+ channels. A and B. BMMCs loaded with Indo 1-AM were incubated with or without 1 mM LaCl3(A) or 10 mM U73122 or 0.5 mM 2-APB (B) for 30 min at 37°C. Cells resupended in Ca2+-free medium were then analysed for Ca2+ mobilization by flow cytometry before and after stimulation with 0.25 mg ml-1 LLO. C. Indo 1-AM loaded BMMCs in Ca2+-free medium were preincubated with or without 0.5 mM 2-APB and stimulated with LLO or 1 mM Na3VO4. Note that Ca2+ mobilization by Na3VO4 – a general activator of PLC-dependent Ca2+ channels was completely blocked by 2-APB whereas Ca2+ release by LLO was only partially inhibited.

activates protein tyrosine phosphorylation in macrophages (Gekara and Weiss, 2004; Gekara et al., 2005). Similarly, LLO was found to induce a robust but transient protein tyrosine phosphorylation in BMMCs (Fig. 6A). Immunoprecipitations of lysates displayed in Fig. 6A with

Damage to intracellular Ca2+ stores – as a potential Ca2+ channel-independent mechanism of Ca2+ release from intracellular stores by LLO The ER is the principal intracellular Ca2+ store. Because CDCs form large pores that can allow passage of ions and macromolecules into and out of the cytosol (Darji et al., 1995a; Walev et al., 2001), injury to the ER, hence release of its contents was considered as a potential Ca2+ channel-independent mechanism of Ca2+ release by LLO. To assess this possibility, first we tried immunofluorescence to analyse the subcellular distribution of LLO but the results were inconclusive. Although LLO could readily be detected on the cell surface, as shown in our recent study (Gekara and Weiss, 2004; Gekara et al., 2005), at the sublytic toxin concentrations used throughout the present study, only a very weak cytosolic staining was


2014 N. O. Gekara et al. Fig. 6. Ca2+ release by LLO via the IP3R-gated Ca2+ channels is due to protein phosphorylation and G-protein activation. BMMCs at 37°C were stimulated with or without 0.25 mg ml-1 LLO. At the indicated time points, cells were lysed, then either blotted and developed directly with an antiphosphotyrosine antibody (A) or first immunoprecipitated with the antiphosphotyrosine antibody before immunoblotting the immunoprecipitates with an anti-PLC-g1 antibody (B). BMMCs resuspended in Ca2+-free medium were preincubated with or without 100 mM genistein at 37°C for 30 min then analysed for Ca2+ induction by LLO (C). In a separate experiment, BMMCs in Ca2+-free medium were preincubated with or without either 1 mM LaCl3 or 1 mg ml-1 pertussis toxin or 10 mM U73122 and Ca2+ signals before and after addition of LLO were analysed (D). The arrows indicate the time points of stimulation with LLO (0.25 mg ml-1).

obtained. Because such staining was hard to distinguish from background it was not possible to do any meaningful colocalization with intracellular organelle marker (data not shown). Bearing in mind that only a very minor fraction of LLO added extracellularly possibly reaches the intracellular Ca2+ stores, this was probably not surprising. We thus concluded that such tiny amount of LLO was beyond detection limits of immunofluorescence. We thus resorted for an indirect approach. The ER-Tracker BODIPYFL glibenclamide (Invitrogen Molecular probes) is a fluorescent probe which accumulates in the ER lumen. Because this probe leaks out in the event of damage to the ER (Rivers et al., 2005), it is a convenient tool not only for visualizing but for monitoring the integrity of this organelle in live cells. To investigate effects of LLO on the ER, RBL-2H3 cells were loaded with ER-Tracker then monitored by confocal microscopy for several minutes following exposure to L.m, L.mDhly or LLO.

Whereas ER labelling of untreated cells remained steady throughout 2 h of image acquisition (Fig. 7A–C and Video S2), a rapid swelling of ER accompanied by release of ER-Tracker was observed in cells exposed to LLO (Fig. 7D–F and Video S3) and L.m (Fig. 7G–I and Video S4) but not L.mDhly (data not shown). This indicated that LLO causes injury to the ER hence release of molecules. Figure 7G–I show profiles of ER-Tracker efflux from two individual cells (1 and 2). Three-dimensional (3D) analysis showed bacteria attachment to 2 but not 1 (data not shown). Because cells without bacterial contact showed no change in the ER-Tracker levels (e.g. cell 1 Fig. 7G–I), direct contact of L.m with the target cells seemed to be necessary for release of the probe from the ER. The requirement for bacterial contact with cells is most likely due to the localized concentration of the toxin at bacteria–cell interfaces where LLO secreted by bacteria should accumulate. In contrast, LLO secreted by bac-

Fig. 7. Perforation of the ER membrane – the Ca2+ channel-independent mechanism of intracellular Ca2+ release by LLO. A–I. RBL-2H3 mast cells loaded with the fluorescent ER specific probe ER-Tracker, were exposed to either medium control (A–C) or LLO (0.25 mg ml-1) (D–F) or fluorescently labelled L. monocytogenes (G–I) then monitored for several minutes by confocal microscopy. The images in A and B show the ER-Tracker in untreated cells at the beginning (0 s) and 30 min after image acquisition while D and E show the ER-Tracker 30 s and 30 min after exposure to LLO. G and H show the ER-Tracker at the beginning and 30 min after exposure to L. monocytogenes. (Time series of these images can be viewed as Videos S2–S4 in Supplementary material.) C, F and I show the kinetics of ER-Tracker efflux from the cells shown in A–B, D–E and G–H respectively. 3D confocal construction of image stacks revealed bacterial attachment to the cell marked 2 but not 1 (not shown). Note that the ER-Tracker efflux is detectable in cell 2 but not 1. J–K. Cytosolic invasion is not a prerequisite for listerial-induced efflux of ER-Tracker from mast cells. BMMCs were exposed to L. monocytogenes for 3 h then analysed by electron microscopy. The scanning (J) and transmission (K) micrographs show bacterial adherence to BMMCs but no internalization. Scale bars; 1 mm. © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2008–2021

Multiple mechanisms of Ca2+ signalling by LLO 2015


2016 N. O. Gekara et al. teria not in contact with cells might rapidly diffuse into the medium and is therefore diluted out. Interestingly, despite the intracellular effects, careful analysis of cells associated with bacteria by 3D microscopy showed that none of the bacteria associated was intracellular (data not shown). That L.m gains no entry into mast cells was confirmed in BMMCs by electron microscopy. As displayed in Fig. 7J and K many bacteria were found adherent to BMMCs while none was found inside such cells. Thus, even in the absence of bacterial internalization, LLO secreted by bacterial L.m could still cause injury to the ER. Listeriolysin O-induced ER injury is reversible Disintegration of intracellular organelles due to loss of cell viability could be an interpretion for release of molecules from the ER. We thus employed flow cytometry to analyse cell viability at various time points after exposure to LLO. Propidium iodide (PI) uptake was used as indicator for membrane perforation and/or loss of cell viability by LLO. While untreated cells remained almost completely negative for PI (Fig. 8A), more than 90% of the cells showed PI staining after 15 min of incubation with LLO (Fig. 8B). However, when such cells were washed after 1 h and allowed to recover for another 1 h before testing for PI upatke, only a minor fraction (25%) stained positive for PI (Fig. 8C), an indication of recovery from toxin attack. To correlate viability and efflux of ER-Tracker, labelled BMMCs were exposed to LLO or L.m. After 1 h, cells were allowed to recover in penicillin/streptomycin containing medium for 1 h before analysing ER-Tracker levels in PI negative cells. Despite substantial release of ER-Tracker, a high percentage of cells preincubated with L.m or a sublethal LLO dose (0.25 mg ml-1) still retained full viability (Fig. 8D and data not shown). Interestingly, although higher toxin concentration (> 1 mg ml-1), resulted in higher loss of viability, ER-Tracker levels in the remaining viable cells were comparable to those exposed to L.m or low toxin concentrations (Fig. 8D). This indicates that efflux of ER-Tracker is neither due to loss of cell viability nor does it linearly correlate with toxin concentration. Indeed, when monitored for several days, cells exposed to sublethal doses of LLO were still found to proliferate normally (not shown). Together, the data suggest that perforation of plasma and leakage of ER membranes caused by LLO is reversible. This is consistent with the known ability of cells to repair membrane lesions (Walev et al., 2001; McNeil and Steinhardt, 2003; McNeil and Kirchhausen, 2005). Cytosolic Ca2+ elevation can cause stress to the ER (Cywes-Bentley et al., 2005). Conceivably, the observed ER swelling and release of molecules could be the consequence of cytosolic Ca2+ elevation rather than the

Fig. 8. Perforation of cell membranes by LLO is reversible. A. Flow cytometric analysis of propidium iodide (PI) staining in BMMCs before treatment with LLO. B. PI staining of BMMCs 15 min after exposure to LLO (1 mg ml-1). C. PI staining in BMMCs treated with LLO (1 mg ml-1) for 1 h, washed and then cultured for another 1 h. The numbers in A–C show the percentage of cells in the PI negative and positive gates. Note that PI positive (B) staining does not necessarily mean loss of cell viability because when cultured for longer, most of the cells became impermeant to PI (C). D. Efflux of molecules from the ER is not associated with loss of cell viability. BMMCs loaded with ER-Tracker were exposed to either L. monocytogenes (moi 100) or sublethal to near-lethal doses of LLO (0.25–1 mg ml-1). After 1 h, cells were washed then allowed to recover in culture for another 1 h in the presence of penicillin/streptomycin, before simultaneously analysing their ER-Tracker levels and PI staining. The histograms show the ER-Tracker levels in PI negative cells (i.e. viable cells). Control cells not loaded with the ER-Tracker are shown in black.

cause. To verify, we monitored the release of ER-Tracker in Ca2+-free medium in presence of the calcium chelator EGTA and the PLC and IP3R inhibitors U73122 and 2-APB respectively. Even under these conditions, LLO still induced efflux of ER-Tracker (Video S5). In contrast, ionomycin, the Ca2+ ionophore, which induces Ca2+ influx as well as release from the intracellular stores, was not found to release the probe (data not shown). Therefore, LLOinduced release of ER contents is not due to the LLOinduced cytosolic Ca2+ elevation. To gain more insight, we also investigated whether LLO affects other Ca2+-rich intracellular organelles such as lysosomes. To that end, cells were loaded with the lysosome specific dye the Lysotracker before treatment with LLO. Like in the case of ER-Tracker, LLO was found to cause efflux of the dye from the lysosome as well (data not shown). The recent findings in which L.m was shown to cause perforations hence efflux of molecules including Ca2+ from the phagosome in an LLO-dependent manner


Multiple mechanisms of Ca2+ signalling by LLO 2017 (Shaughnessy et al., 2006) are consistent with those in the present study. Taken together, these results show that at sublethal concentrations, LLO causes reversible permeabilization of the plasma and the Ca2+-rich intracellular organelles, hence releasing their contents. This phenomenon may account for the Ca2+ channel-independent mechanism of Ca2+ release by LLO. Discussion Induction of Ca2+ signals by L.m in target cells is extremely important in context of bacterial survival in the host. As demonstrated in this work for mast cells, Ca2+ induced by LLO triggers de novo synthesis and secretion of proinflammatory mediators which could recruit more host cells to the site of infection (N.O. Gekara and S. Weiss, submitted). Similarly, effects like apoptosis induced during Listeria infection might also be attributed to Ca2+ signalling by LLO (Carrero et al., 2004a). Thus far, Ca2+ signalling by CDCs was mainly thought to be due to Ca2+ influx from extracellular milieu (Repp et al., 2002; Dramsi and Cossart, 2003). However, we present data, which clearly show that LLO, the prototype CDC, by itself induces Ca2+ release from intracellular stores, independent of initial Ca2+ influx. Ca2+ release from intracellular stores is mediated via Ca2+ channel-dependent and -independent mechanisms. Our data show that the former mechanism involves IP3R-gated Ca2+ channels whereas the latter is most likely due to injury of intracellular Ca2+ stores induced by LLO. The IP3R operated Ca2+ release can be mediated by several PLC isoforms activated by different mechanisms (Berridge et al., 2003). We discovered that LLO induces a robust tyrosine phosphorylation of several proteins including PLC-g1. This, plus the fact that Ca2+ release could partially be blocked by the tyrosine kinase inhibitor genistein and the G-protein inhibitor pertussis toxin, suggested that LLO activates the IP3R Ca2+ channels via two independent pathways, namely: protein tyrosine phosphorylation and G-protein activation of PLC-g and PLC-b isoforms respectively. We have recently shown that binding and oligomerization of LLO on host cell membranes spontaneously aggregates lipid rafts and hence triggers signals in target cells (Gekara and Weiss, 2004; Gekara et al., 2005). Because cholesterol-inactivated LLO which binds and aggregates rafts like the active form of LLO (Gekara and Weiss, 2004; Gekara et al., 2005), could not induce Ca2+ release, we think that membrane binding or lipid rafts aggregation per se is not sufficient to activate these IP3R-dependent pathways. Thus, oligomerization and transmembrane insertion, the two essential steps of pore formation are crucial for the activation of these PLC-IP3R pathways by LLO.

Whereas, G-proteins activation might be a common consequence of pore formation by a variety of proteins and peptides (Cybulsky et al., 1989; Grimminger et al., 1991; Krause et al., 1998), to our knowledge, no report has documented the role of protein tyrosine kinaseactivated PLC-g in Ca2+ release by a pore forming toxin. Whether this is unique to CDCs like LLO or also extends to other pore forming toxins remains to be tested. Despite evidence for the role of IP3R operated Ca2+ channels in LLO-induced Ca2+ release, the inability of various Ca2+ channel and PLC-IP3R inhibitors to completely block Ca2+ release by LLO suggests the existence of a channel-independent mechanism. For this, we have provided indirect evidence that argues for a mechanism of Ca2+ release involving injury of intracellular Ca2+ stores such as the ER and lysosomes. This is based on the fact that LLO and L.m do cause the release of fluorescent probes from such Ca2+-rich intracelular organelles. A number of possibilities could account for the ability of LLO to cause release of molecules from the ER. It is conceivable that the LLO-mediated damage to the plasma membrane (albeit reversible) could create osmotic or ionic disruptions of the cytoplasm leading the characteristic vacuolations and loss of organelle integrity. Such indirect effects have been observed when cells were treated with high concentrations of ATP (Steinberg et al., 1987). Alternatively, LLO could directly reach the ER and cause release of Ca2+ from intracellular stores. This could be due to retrograde transport of LLO to the ER although the endocytosis inhibitor cytochalasin D did not interfere with intracellular Ca2+ release (data not shown). On the other hand, the large membrane pores that LLO forms (Darji et al., 1995a) allows delivery of macromolecules including LLO into the cytosol (Darji et al., 1995a; Sibelius et al., 1996; Wadsworth and Goldfine, 1999). Thus, although the cholesterol content of ER is rather low (Lange et al., 1999), the possibility of a direct interaction and hence perforation of ER by LLO cannot be ruled out entirely. Whether permeabilization of such intracellular organelles is exclusively applicable to CDCs or extends to other pore forming toxins such as aerolysin or stapylococcus-a toxin which trigger intracellular calcium release (Krause et al., 1998) remains to be determined. Whereas the limited pore size of such toxins militates against direct damage by the toxin due to self-injection, retrograde transportation or injury of intracellular organelles as a result of osmotic or ionic disruptions is feasible. That LLO permeabilizes the plasma and internal membranes provides a new perspective in our understanding of how CDCs affect target cells. In fact, implications of these findings stretch far beyond intracellular Ca2+ release and should open new avenues of research aimed at understanding effects of such toxins on intracellular


2018 N. O. Gekara et al. organelles during pathogenesis. Thus, how CDCs interacts with and modifies the functions of various intracellular organelles and the significance thereof in pathogenesis is now worth looking into.

Experimental procedures Antibodies and reagents Anti-NFATc1, anti-phosphotyrosine (PY99) and anti-PLCg1 were obtained from Santa Cruz Biotechnology. p-nitrophenyl-N-acetylb-D-glucosaminide, Indo 1-AM, U73122 and Genistein from Sigma. BAPTA-AM and ER-Tracker, from Invitrogen, Pertussis toxin from Fluka Biochemika, 2-APB from Alexis Biochemicals (Lausen, Switzerland), Iscove’s modified Dulbecco’s medium (IMDM), Dulbecco’s modified Eagle’s medium (DMEM) and Ca2+free DMEM were from Gibco. LLO was purified from overexpressing L. innocua as previously described (Darji et al., 1995b) while CL-LLO was prepared by inactivating LLO with cholesterol as described (Gekara et al., 2005).

Generation of BMMCs Bone marrow-derived mast cells were matured by culturing bone marrow cells in the presence of IL-3 for 4–8 weeks as previously described (Razin et al., 1984) Rat basophilic leukemia cell line (RBL-2H3) was kindly provided by Pecht I (The Weizmann Institute of Science, Israel).

Ca2+ flux measurements by flow cytometry A total of 5 ¥ 106 BMMCs in 500 ml of DMEM were incubated with 50 mM Indo 1-AM in complete medium at 37°C. After 45 min, cells were washed once in Ca2+-free medium supplemented with 10 mM EGTA to ensure removal of any residual extracellular calcium from cells, then twice in unsupplemented Ca2+-free medium to remove residual EGTA. This second step was necessary because influx of residual EGTA into the cell via the LLO pore could chelate intracellular Ca2+ hence marring the Ca2+ signals due to release from intracellular store. Therefore, to evaluate the relative contribution of extracellular and intracellular Ca2+ pools in the overall Ca2+ signals triggered by LLO, after this washing procedure cells were resuspended in either normal or Ca2+-free medium then kept on ice until ready for measurement. Cells were warmed up to 37°C before start of Ca2+ measurements. Measurements were carried out in a MoFlo highspeed cell sorter (DakoCytomation) equipped with an UV argon ion laser (351–363 nm). Indo 1-AM emissions were detected with 405/30 (Ca2+-bound Indo 1-AM) and 515/30 (Ca2+-free Indo 1-AM) fluorescence filters. First the ratio of the fluorescence emitted at the two Indo 1-AM excitation wavelengths (F405/30/ F515/30) was calculated then calibrated into arbitrary fluorescence intensity units (a.f.u.) by the equation F405/30/F515/ 30 ¥ 128, where the constant 128 is the median of the instrument’s fluorescence spectrum. The data were subsequently normalized for fluctuations in the initial baseline measurements in Excel software, by dividing all the a.f.u. with the average a.f.u. of the initial 30 s baseline period. Therefore, the arbitrary ratiometric units represent the fold increase in F405/30/F515/30 over that of the initial baseline.

Ca2+ flux and ER-Tracker efflux measurements by confocal laser microscopy Bacteria were labelled with 100 nM BacLight Red stain (Molecular Probes) at room temperature for 15 min before addition to cells. RBL-2H3 cells were labelled with Indo 1-AM and thoroughly washed as described above before exposure to either LLO (0.25 mg ml-1), or BacLight Red stained L.m or Dhly or DPlcA/B [multiplicity of infection (moi) 100]. BacLight Red fluorescence was detected at 630 nm while Indo 1-AM emissions were detected with 405 (Ca2+-bound Indo 1-AM) and 460 (Ca2+-free Indo 1-AM) fluorescence filters by scanning laser confocal microscopy (Zeiss LSM510 Meta) using a Plan Apo lens (40¥, NA 1.4, oil immersion). The calcium measurements were calculated as ratio of fluorescence intensities of bound Ca2+/free Ca2+(F405/ F460) then calibrated into a.f.u. as already described above (a.f.u. = F405/F460 ¥ 128). For ER-Tracker efflux measurements, RBL-2H3 mast cells were loaded with 2 mM BODIPY ER-Tracker in phenol red-free medium for 30 min at 37°C, washed, then exposed to either LLO or BacLight Red stained bacteria as described above. The ER-Tracker and bacteria emissions were detected at 520 nm and 630 nm respectively. Images were captured at an interval of 30 s. Time series of the images were converted into an avi movie using the Zeiss LSM 510 software (Carl Zeiss, Jena, Germany). and finally into QuickTime format using QuickTimepro software.

Determination of NFATc1 nuclear translocation by Immunofluorescence microscopy Bone marrow-derived mast cells untreated or treated with LLO (0.25 mg ml-1) or ionomycin (1 mM) for 45 min, were fixed, permeabilized in a 50:50 Acetone/methanol fixative -20°C, then blocked in 3% BSA for 30 min before staining with a mouse anti-NFATc1 (Santa Cruz Biotechnol, CA), followed by a Cy3Goat anti-mouse antibody. Samples were mounted using Fluoprep (bioMerieux, Marcy l’Etoile, France) and examined using an Axiovert 135 TV microscope (Zeiss Oberkochen, Germany) equipped with Cy3, FITC filters and a Plan-Apochromat 100¥/1.40 NA oil immersion objective. Images were recorded with a cooled (-25°C), back-illuminated CCD camera.

Measurement of cell viability and ER-Tracker efflux by flow cytometry Bone marrow-derived mast cells were treated with LLO doses 0.25–1 mg ml-1. At different time points they were stained with PI and analysed by flow cytometry. Alternatively, BMMCs loaded with 2 mM ER-Tracker (30 min at 37°C) were exposed to L.m (moi 100) or sublethal to near-lethal doses of LLO (0.25–1 mg ml-1). After 1 h, cells were washed, cultured for another 1 h in the presence of penicillin/streptomycin before simultaneously analysing their ER-Tracker levels and PI staining.

Scanning and transmission electron microscopy After exposure to L.m (moi 100) for 3 h, BMMCs were prepared as already described (Chakravortty et al., 2005) then examined in a Zeiss field emission scanning electron microscope DSM 982 Gemini at an acceleration voltage of 5 kV applying the EverhartThornley SE-detector and the inlens detector at a 50:50 ratio. For


Multiple mechanisms of Ca2+ signalling by LLO 2019 ultrastructural analysis samples were examined in a Zeiss transmission electron microscope TEM910 at an acceleration voltage of 80 kV and calibrated magnifications. Images were stored on MO-disks and contrast and brightness was adjusted with Adobe Photoshop 7.0.

b-Hexosaminidase degranulation A total of 2 ¥ 105 BMMCs in 200 ml IMDM were seeded into each well of a 96 well plate, stimulated with either LLO/CL-LLO (0.25 mg ml-1) or exposed to L.m or Dhly (moi 100). After incubation at 37°C for 1 h, supernatants were transferred to a 96-well plate and cells were lysed in 0.5% Triton X-100 solution. Release of b-hexosaminidase was measured as an index of mast cell degranulation using a standard method (de Bernard et al., 2005).

TNF-a bioassay Following exposure of BMMC to LLO or L.m, the biological activity of TNF in the supernatants was assayed using the TNF-sensitive murine fibroblasts L929. Into each well of a 96-well plate containing 4 ¥ 104 cells/well, 100 ml of medium containing actinomycin D (6.25 mg ml-1) and 50 ml of the sample supernatant was added. After 24 h incubation supernatants were aspirated and cell viability determined using the EASY FOR YOU KIT (BIOMEDICA). Serial dilutions of recombinant TNF-a were used as reference.

Reverse transcription polymerase chain reaction (RT-PCR) analysis mRNA was isolated from BMMCs using the RNEasy Mini Kit from Qiagen and reverse transcribed into cDNA using the SuperScript II RNAseH Reverse Transcriptase kit from Invitrogen. PCRs were performed with primers; 5′-TCT CAT CAG TTC TAT GGC CC-3′, 5′-GGG AGT AGA CAA GGT ACA-3′ for TNF-a and 5′-CTG GAC GAG GGC AAG ATG AAG C-3′, 5′-TGA CGT TGG CGG ATG AGC ACA-3′ for ribosomal protein S9 (RPS9). Amplification conditions were: denaturation at 94°C for 1 min followed by 27 and 32 cycles (for RPS9 and TNF-a), respectively, of repeated denaturation (20 s at 94°C), annealing (20 s at 58°C) and extension (20 s at 72°C). PCR products were run on a 2% agarose gel.

Immunoprecipitation and immunoblotting A total of 107 BMMCs resuspended in 1 ml of IMDM were stimulated with 0.25 mg ml-1 LLO at 37°C. After 2, 4, 8, 20 and 30 min, cell pellets were lysed in 600 ml of RIPA buffer (10 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% NP40, 2 mM Na3VO4 and 10 mM NaF containing a cocktail of protease inhibitors). For immunoprecipitation, 500 ml of each of the fractions was filled up to 1 ml, then incubated at 4°C with either 1 mg of the mouse antiphsophotyrosine antibody (PY99) or rabbit anti-PLC-g1. After 2 h 20 ml of ProteinA/G-Agarose was added and lysates incubated overnight at 4°C. Agarose pellet were washed three times in RIPA buffer, boiled SDS sample buffer then subjected to SDS– PAGE and immunoblotting with rabbit anti-PLC-g1 or mouse antiphsophotyrosine antibody (PY99). Sample aliquots of preimmunoprecipitation lysates were also analysed for total phosphotyrosines, and anti-PLC-g1.

Acknowledgements Thanks to Klemens Rottner and Theresa Stradal for the help in processing the Videos and Riki Graham for critical reading of the manuscript, Kurt Dittmar for help with confocal micoscopy. Regina Lesch, Susanne zur Lage and Ina Schleicher for technical help. This work was supported by a grant from the European Community (EC), Deutsche Krebshilfe and Deutsche Forschungsgemeinschaft (DFG) to S.W. The authors have no conflicting financial interests.

References de Bernard, M., Cappon, A., Pancotto, L., Ruggiero, P., Rivera, J., Del Giudice, G., and Montecucco, C. (2005) The Helicobacter pylori VacA cytotoxin activates RBL-2H3 cells by inducing cytosolic calcium oscillations. Cell Microbiol 7: 191–198. Berridge, M.J., Bootman, M.D., and Roderick, H.L. (2003) Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Cell Biol 4: 517–529. Berry, A.M., Yother, J., Briles, D.E., Hansman, D., and Paton, J.C. (1989) Reduced virulence of a defined pneumolysinnegative mutant of Streptococcus pneumoniae. Infect Immun 57: 2037–2042. Billington, S.J., Jost, B.H., and Songer, J.G., (2000) Thiolactivated cytolysins: structure, function and role in pathogenesis. FEMS Microbiol Lett 182: 197–205. Boulnois, G.J., Paton, J.C., Mitchell, T.J., and Andrew, P.W. (1991) Structure and function of pneumolysin, the multifunctional, thiol-activated toxin of Streptococcus pneumoniae. Mol Microbiol 5: 2611–2616. Carrero, J.A., Calderon, B., and Unanue, E.R., (2004a) Listeriolysin O from Listeria monocytogenes is a lymphocyte apoptogenic molecule. J Immunol 172: 4866–4874. Carrero, J.A., Calderon, B., and Unanue, E.R., (2004b) Type I interferon sensitizes lymphocytes to apoptosis and reduces resistance to Listeria infection. J Exp Med 200: 535–540. Chakravortty, D., Rohde, M., Jager, L., Deiwick, J., and Hensel, M., (2005) Formation of a novel surface structure encoded by Salmonella Pathogenicity Island 2. EMBO J 24: 2043–2052. Cybulsky, A.V., Salant, D.J., Quigg, R.J., Badalamenti, J., and Bonventre, J.V. (1989) Complement C5b-9 complex activates phospholipases in glomerular epithelial cells. Am J Physiol 257: F826–F836. Cywes-Bentley, C., Hakansson, A., Christianson, J., and Wessels, M.R. (2005) Extracellular group A Streptococcus induces keratinocyte apoptosis by dysregulating calcium signalling. Cell Microbiol 7: 945–955. Darji, A., Chakraborty, T., Wehland, J., and Weiss, S. (1995a) Listeriolysin generates a route for the presentation of exogenous antigens by major histocompatibility complex class I. Eur J Immunol 25: 2967–2971. Darji, A., Chakraborty, T., Niebuhr, K., Tsonis, N., Wehland, J., and Weiss, S., (1995b) Hyperexpression of listeriolysin in the nonpathogenic species Listeria innocua and high yield purification. J Biotechnol 43: 205–212. Dramsi, S., and Cossart, P. (2003) Listeriolysin O-mediated calcium influx potentiates entry of Listeria monocytogenes


2020 N. O. Gekara et al. into the human Hep-2 epithelial cell line. Infect Immun 71: 3614–3618. Edelson, B.T., Li, Z., Pappan, L.K., and Zutter, M.M. (2004) Mast cell-mediated inflammatory responses require the alpha 2 beta 1 integrin. Blood 103: 2214–2220. Edelson, B.T., Stricker, T.P., Li, Z., Dickeson, S.K., Shepherd, V.L., Santoro, S.A., and Zutter, M.M. (2006) Novel collectin/ C1q receptor mediates mast cell activation and innate immunity. Blood 107: 143–150. Gaillard, J.L., Berche, P., and Sansonetti, P. (1986) Transposon mutagenesis as a tool to study the role of hemolysin in the virulence of Listeria monocytogenes. Infect Immun 52: 50–55. Gekara, N.O., and Weiss, S. (2004) Lipid rafts clustering and signalling by listeriolysin O. Biochem Soc Trans 32: 712– 714. Gekara, N.O., Jacobs, T., Chakraborty, T., and Weiss, S. (2005) The cholesterol-dependent cytolysin listeriolysin O aggregates rafts via oligomerization. Cell Microbiol 7: 1345–1356. Gordon, J.R., and Galli, S.J., (1991) Release of both preformed and newly synthesized tumor necrosis factor alpha (TNF-alpha)/cachectin by mouse mast cells stimulated via the Fc epsilon RI. A mechanism for the sustained action of mast cell-derived TNF-alpha during IgE-dependent biological responses. J Exp Med 174: 103–107. Grimminger, F., Sibelius, U., Bhakdi, S., Suttorp, N., and Seeger, W. (1991) Escherichia coli hemolysin is a potent inductor of phosphoinositide hydrolysis and related metabolic responses in human neutrophils. J Clin Invest 88: 1531–1539. Jacobs, T., Darji, A., Frahm, N., Rohde, M., Wehland, J., Chakraborty, T., and Weiss, S. (1998) Listeriolysin O: cholesterol inhibits cytolysis but not binding to cellular membranes. Mol Microbiol 28: 1081–1089. Kathariou, S., Metz, P., Hof, H., and Goebel, W. (1987) Tn916-induced mutations in the hemolysin determinant affecting virulence of Listeria monocytogenes. J Bacteriol 169: 1291–1297. Kayal, S., Lilienbaum, A., Poyart, C., Memet, S., Israel, A., and Berche, P. (1999) Listeriolysin O-dependent activation of endothelial cells during infection with Listeria monocytogenes: activation of NF-kappa B and upregulation of adhesion molecules and chemokines. Mol Microbiol 31: 1709– 1722. Krause, K.H., Fivaz, M., Monod, A., and van der Goot, F.G., (1998) Aerolysin induces G-protein activation and Ca2+ release from intracellular stores in human granulocytes. J Biol Chem 273: 18122–18129. Lange, Y., Ye, J., Rigney, M., and Steck, T.L. (1999) Regulation of endoplasmic reticulum cholesterol by plasma membrane cholesterol. J Lipid Res 40: 2264–2270. McLachlan, J.B., Hart, J.P., Pizzo, S.V., Shelburne, C.P., Staats, H.F., Gunn, M.D., and Abraham, S.N. (2003) Mast cell-derived tumor necrosis factor induces hypertrophy of draining lymph nodes during infection. Nat Immunol 4: 1199–1205. McNeil, P.L., and Kirchhausen, T. (2005) An emergency response team for membrane repair. Nat Rev Mol Cell Biol 6: 499–505. McNeil, P.L., and Steinhardt, R.A. (2003) Plasma membrane

disruption: repair, prevention, adaptation. Annu Rev Cell Dev Biol 19: 697–731. Malaviya, R., Ikeda, T., Ross, E., and Abraham, S.N., (1996) Mast cell modulation of neutrophil influx and bacterial clearance at sites of infection through TNF-alpha. Nature 381: 77–80. Marshall, J.S. (2004) Mast-cell responses to pathogens. Nat Rev Immunol 4: 787–799. Palmer, M. (2001) The family of thiol-activated, cholesterolbinding cytolysins. Toxicon 39: 1681–1689. Pasparakis, M., Alexopoulou, L., Episkopou, V., and Kollias, G. (1996) Immune and inflammatory responses in TNF alpha-deficient mice: a critical requirement for TNF alpha in the formation of primary B cell follicles, follicular dendritic cell networks and germinal centers, and in the maturation of the humoral immune response. J Exp Med 184: 1397– 1411. Pfeffer, K., Matsuyama, T., Kundig, T.M., Wakeham, A., Kishihara, K., Shahinian, A., et al. (1993) Mice deficient for the 55 kd tumor necrosis factor receptor are resistant to endotoxic shock, yet succumb to L. monocytogenes infection. Cell 73: 457–467. Portnoy, D.A., Jacks, P.S., and Hinrichs, D.J., (1988) Role of hemolysin for the intracellular growth of Listeria monocytogenes. J Exp Med 167: 1459–1471. Razin, E., Ihle, J.N., Seldin, D., Mencia-Huerta, J.M., Katz, H.R., LeBlanc, P.A., et al. (1984) Interleukin 3: a differentiation and growth factor for the mouse mast cell that contains chondroitin sulfate E proteoglycan. J Immunol 132: 1479–1486. Repp, H., Pamukci, Z., Koschinski, A., Domann, E., Darji, A., Birringer, J., et al. (2002) Listeriolysin of Listeria monocytogenes forms Ca2+-permeable pores leading to intracellular Ca2+ oscillations. Cell Microbiol 4: 483– 491. Rivers, D.B., Crawley, T., and Bauser, H. (2005) Localization of intracellular calcium release in cells injured by venom from the ectoparasitoid Nasonia vitripennis (Walker) (Hymenoptera: Pteromalidae) and dependence of calcium mobilization on G-protein activation. J Insect Physiol 51: 149–160. Rothe, J., Lesslauer, W., Lotscher, H., Lang, Y., Koebel, P., Kontgen, F., et al. (1993) Mice lacking the tumour necrosis factor receptor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature 364: 798–802. Shaughnessy, L.M., Hoppe, A.D., Christensen, K.A., and Swanson, J.A. (2006) Membrane perforations inhibit lysosome fusion by altering pH and calcium in Listeria monocytogenes vacuoles. Cell Microbiol 8: 781–792. Sibelius, U., Chakraborty, T., Krogel, B., Wolf, J., Rose, F., Schmidt, R., et al. (1996) The listerial exotoxins listeriolysin and phosphatidylinositol-specific phospholipase C synergize to elicit endothelial cell phosphoinositide metabolism. J Immunol 157: 4055–4060. Steinberg, T.H., Newman, A.S., Swanson, J.A., and Silverstein, S.C. (1987) Macrophages possess probenecidinhibitable organic anion transporters that remove fluorescent dyes from the cytoplasmic matrix. J Cell Biol 105: 2695–2702. Tran Van-Nhieu, G., Clair, C., Grompone, G., and Sansonetti,


Multiple mechanisms of Ca2+ signalling by LLO 2021 P. (2004) Calcium signalling during cell interactions with bacterial pathogens. Biol Cell 96: 93–101. Van Epps, D.E., and Andersen, B.R. (1974) Streptolysin O inhibition of neutrophil chemotaxis and mobility: nonimmune phenomenon with species specificity. Infect Immun 9: 27–33. Wadsworth, S.J., and Goldfine, H. (1999) Listeria monocytogenes phospholipase C-dependent calcium signaling modulates bacterial entry into J774 macrophage-like cells. Infect Immun 67: 1770–1778. Walev, I., Bhakdi, S.C., Hofmann, F., Djonder, N., Valeva, A., Aktories, K., and Bhakdi, S. (2001) Delivery of proteins into living cells by reversible membrane permeabilization with streptolysin-O. Proc Natl Acad Sci USA 98: 3185– 3190.

Supplementary material The following supplementary material is available for this article online: Fig. S1. Chelation of intracellular Ca2+ diminishes cytosolic Ca2+ elevation by LLO. BMMCs were incubated with 100 mM BAPT-AM for 37°C for 30 min to chelate intracellular Ca2+. After washing, cells were loaded with Indo 1-AM before stimulation with LLO or CL-LLO in normal or Ca2+-free medium. A 30 s baseline was recorded each time before stimulation. The arrow indicates the time point of stimulation. Fig. S2. Activation of PLC-g1 by LLO. BMMCs at 37°C were stimulated with or without 0.25 mg ml-1 LLO. At the indicated time points, cells were lysed then immunoprecipitated with an antiPLC-g1 antibody before immunoblotting the immunoprecipitates with the antiphosphotyrosine antibody. The lower panel shows the total PLC-g1 in the lysates before immunoprecipitation. Video S1. Ca2+ flux induction in RBL-2H3 cells by L. monocytogenes. The movie shows Ca2+ mobilization in RBL2H3 mast cells upon exposure to L. monocytogenes (blue). It represents a period of 1 h after addition of bacteria to cells.

Cytosolic Ca2+ elevation is represented by a shift in fluorescence emission of Indo-1-AM from 460 nm (green) to 405 (red). Video S2. Control RBL-2H3 mast cells loaded with the fluorescent ER-Tracker. The movie shows the ER-Tracker labelling in RBL-2H3 mast cells in untreated cells. Note that the levels remain steady throughout. It represents a time period of 30 min. Video S3. LLO causes efflux of molecules from the ER. RBL2H3 mast cells were loaded with the ER-Tracker then treated with 0.25 mg ml-1 LLO. Note that even by 30 s post LLO addition (i.e. the starting time of image recording), the ER has already undergone changes. This is then followed by rapid efflux of the ER-Tracker. The movie represent a time period of 30 min. Video S4. Listeria monocytogenes causes efflux of molecules from the ER. L. monocytogenes (red) was added to ER-Tracker loaded RBL-2H3 mast cells. Recording started 1 min after addition of bacteria. Note that the ER-Tracker efflux is detectable only in the cell marked 2 but not in 1. The movie represents a time period of 30 min. Video S5. LLO-induced efflux of ER-Tracker is not a consequence of cytosolic Ca2+ elevation. RBL-2H3 mast cells loaded with the fluorescent ER specific probe ER-Tracker and resuspended in Ca2+-free medium containing the calcium antagonists EGTA, 2-APB and U73122, were exposed (or not) to LLO (0.1 mg ml-1) then monitored for several minutes by confocal microscopy. The video on the left show the ER-Tracker levels in untreated cells while the video in the right shows LLO-induced ER-Tracker efflux in the combined presence of the calcium antagonists EGTA, 2-APB and U73122. Image acquisition was started 12 s after exposure to LLO. Videos represent time frame of 60 min. This material is available as part of the online article from: http://www.blackwell-synergy.com/doi/abs/10.1111/ j.1462-5822.2007.00932.x Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
