Jul 25, 2007 - to convert the bulk of cell surface SM to ceramide in polarized MDCK, human PMN ... enously added SMs, cholesterol or ceramides to reverse ...... analysed using ImageJ software (National Institutes of Health). Fluid-phase ...
Cellular Microbiology (2008) 10(1), 67–80
doi:10.1111/j.1462-5822.2007.01015.x First published online 25 July 2007
Conversion of apical plasma membrane sphingomyelin to ceramide attenuates the intoxication of host cells by cholera toxin David E. Saslowsky* and Wayne I. Lencer GI Cell Biology, Children’s Hospital, Boston, the Harvard Digestive Diseases Center, and the Department of Pediatrics, Harvard Medical School, Boston, MA 02115, USA. Summary Cholera toxin (CT) enters host cells by binding to ganglioside GM1 in the apical plasma membrane (PM). GM1 carries CT retrograde from the PM to the endoplasmic reticulum (ER), where a portion of the toxin, the A1-chain, retro-translocates to the cytosol, causing disease. Trafficking in this pathway appears to depend on the association of CT–GM1 complexes with sphingomyelin (SM)- and cholesterol-rich membrane microdomains termed lipid rafts. Here, we find that in polarized intestinal epithelia, the conversion of apical membrane SM to ceramide by bacterial sphingomyelinase attenuates CT toxicity, consistent with the lipid raft hypothesis. The effect is reversible, specific to toxin entry via the apical membrane, and recapitulated by the addition of exogenous long-chain ceramides. Conversion of apical membrane SM to ceramide inhibits the efficiency of toxin endocytosis, but retrograde trafficking from the apical PM to the Golgi and ER is not affected. This result suggests that the cause for toxin resistance occurs at steps required for retro-translocation of the CT A1-chain to the cytosol. Introduction The bacterial AB5-subunit toxin cholera toxin (CT) enters host intestinal epithelial cells by binding to the ganglioside GM1 in the apical plasma membrane (PM). GM1 is the vehicle that carries the toxin retrograde through the apical endosome, trans-Golgi network (TGN), and into the endoplasmic reticulum (ER). Upon entering the ER, a portion of the enzymatic A-subunit (A1-chain) unfolds and retrotranslocates to the cytosol where it induces cAMP accuReceived 4 April, 2007; revised 22 June, 2007; accepted 26 June, 2007. *For correspondence. E-mail david.saslowsky@childrens. harvard.edu; Tel. (+1) 617 919 2550; Fax (+1) 617 730 0498. © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd
mulation by activating adenylate cyclase (Lencer and Saslowsky, 2005). In the intestine, this causes the massive secretory diarrhoea seen in Asiatic cholera. The retrograde pathway from the PM to the ER typifies the mechanism by which other AB5-subunit toxins (including Escherichia coli and Shiga toxins) and some viruses enter host cells (Johannes and Goud, 1998; Pelkmans et al., 2004; Lencer and Saslowsky, 2005). So far, in all cells tested, only the gangliosides that associate with detergent-resistant membrane fragments (DRMs) can transport the AB5 toxins retrograde into the ER (Wolf et al., 1998; Falguieres et al., 2001; Fujinaga et al., 2003). The correlation between the fractionation of lipid-anchored toxins with DRMs and toxicity of the AB5-subunit toxins suggests that lipid rafts might act as platforms for toxin coupling to the intracellular sorting machinery required for membrane trafficking and toxin function. In this paper, we test this hypothesis by altering the content of SM and ceramide in host cells. Lipid rafts are thought to be small, highly dynamic membrane microdomains that depend on cholesterol and sphingolipids for structure and function (Simons and Ikonen, 1997; Brown and London, 2000). Certain sphingolipids, such as sphingomyelin (SM), hydrogen bond with cholesterol, promoting the formation of lipid microdomains (Bittman et al., 1994). The precursor of SM, ceramide, however, has a lower affinity for cholesterol (Megha and London, 2004), and yet still forms distinct ordered microdomains in vitro (Kolesnick et al., 2000; Wang and Silvius, 2003) and large, lipid raft-like structures in the outer leaflet of the PM in vivo (Grassme et al., 2003). The content of SM and ceramide in eukaryotic cells is dynamic and can be rapidly altered by the enzymatic activities of sphingomyelinases (SMases) and SM synthases. Sphingomyelinases are enzymes produced by both prokaryotes and eukaryotes that cleave the phosphorylcholine head group of SM, leaving the ceramide component intact in the lipid bilayer (Gulbins et al., 2004). Various SMase isoforms have been described and implicated in physiologic processes, including cell differentiation, proliferation, apoptosis and degradation of dietary SM (Nilsson, 1968; Nyberg et al., 1997); also in diseases such as Niemann–Pick syndrome, pulmonary oedema (Goggel et al., 2004), Alzheimer’s disease (Lee et al.,
68 D. E. Saslowsky and W. I. Lencer 2004) and atherosclerosis (Tabas, 1999). Membrane SM may also affect infection by HIV and prion protein, and trafficking of endogenous glycolipids (Mahfoud et al., 2002; Cheng et al., 2006; Rawat et al., 2006). Because the environment of the human intestine contains a variety of SMases contributed by both host cells (Cremesti et al., 2002) as well as the intestinal microbial flora (Chen et al., 1992; Nilsson and Duan, 1999), these enzymes may affect membrane structure and function so as to regulate intestinal physiology, as well as the response to microbial factors such as CT.
Results Enzymatic conversion of SM to ceramide in the apical PM inhibits CT-induced Cl – secretion To test whether the membrane content of SM affects CT toxicity, we pretreated apical membranes of polarized intestinal T84 cells with SMase and then intoxicated the cells apically with CT. Because T84 cells form polarized monolayers with high trans-epithelial resistance (TER), the only way apically applied CT can enter the cell is through the apical membrane (Lencer and Saslowsky, 2005). After entering the cytosol, the toxin rapidly induces a cAMP-dependent Cl– secretory response, which can be measured in real time using standard methods of electrophysiology. Apical treatment of T84 cells with SMase caused a ~50% reduction in CT-induced Cl– secretion (Isc; Fig. 1A). This was partially rescued by the addition of exogenous long-acyl-chain SM (primarily 18:0 and 24:0, Fig. 1A). Apical SMase treatment had no effect on Isc induced by CT entering the cell via the basolateral membrane (Fig. 1B and D). Thus, apical SMase must act locally at the site of application; it cannot be acting non-specifically
or on general mechanisms of membrane trafficking or retro-translocation in the ER. SMase, by itself, had no effect on resting trans-epithelial potential, on the Isc induced by the cAMP-agonist forskolin (Fig. 1A and B) or vasointestinal peptide (data not shown), or on TER. Thus, the inhibition of CT-induced Isc caused by conversion of SM to ceramide in the apical PM cannot be explained by confounding effects on the cAMP-dependent Cl– secretory pathway or on the structure and function of intercellular tight junctions. These results also indicate that the ceramide generated by hydrolysis of SM does not induce transient pores in the apical PM of intestinal cells as reported for liposomes and erythrocyte ghosts (Ruiz-Arguello et al., 1996), and also that cell function is generally intact. To demonstrate dose and time dependency for SMase action, T84 monolayers were incubated apically with increasing concentrations of SMase or for increasing times and then intoxicated apically with CT. The higher concentration of SMase (1 U ml-1) induced a stronger inhibitory response (Fig. 1C). Short incubations with SMase (5 min) had no observable effect on CT function, whereas increasingly longer incubations at this concentration (up to 50 min) attenuated CT toxicity (Fig. 1D). The prolonged incubation time required for SMase to inhibit CT function is consistent with the time required for SMase to convert the bulk of cell surface SM to ceramide in polarized MDCK, human PMN and BHK cells (Slotte et al., 1990; van Helvoort et al., 1994; Feldhaus et al., 2002). To show that the exogenous SMase effectively removed the bulk of phosphoryl choline groups on endogenous SM, we used the SM domain binding protein lysenin fused to red fluorescent protein (lysenin–RFP) (Kiyokawa et al., 2005). Lysenin–RFP was applied to apical surfaces of untreated and SMase-treated T84 monolayers at 4°C. In
Fig. 1. Conversion of SM to ceramide in the apical PM of polarized T84 cells attenuates CT function. A. Time-course of Cl– secretion (Isc, mean ⫾ SD, n = 2) induced by apically applied CT (3 nM, t = 0) in T84 monolayers pretreated apically with 1 U ml-1 SMase for 50 min at 37°C (open circles) or not treated (filled squares). For some monolayers, SM (100 mM) was added after SMase treatment (filled triangles). For each condition, unintoxicated monolayers (filled diamonds, open squares, filled circles) were treated with the cAMP-agonist forskolin at 115 min. B. Isc induced by CT applied basolaterally to monolayers treated apically with SMase (open squares) or not treated (filled diamonds). Unintoxicated controls are similar to panel A. C. Maximal Isc (mean ⫾ SD at 120 min) induced by CT in cells incubated apically with the indicated concentrations of SMase for 50 min prior to apical intoxication with 3 nM CT or with buffer alone (–CT). Sample size (n) for each condition indicated under bars. D. As in panel C. Apical surfaces were incubated in 1 U ml-1 SMase for the indicated times or with buffer only (–). E. Apical surfaces of polarized T84 monolayers were pretreated for 50 min with or without SMase, incubated apically with lysenin–RFP for 1 h at 4°C, fixed and imaged at the apical pole by confocal microscopy as described in Experimental procedures. Scale bars = 10 mm. F. Monolayers were treated with SMase (+) or not treated (–) as indicated. Prior to CT application, some monolayers were treated for 1 h with 30 mM mixed long-chain SM or ceramides, U18666A (10 mg ml-1), cholesterol (M/C), buffer, or allowed to recover for 3 h. Expressed as maximal Isc. G. Apical surfaces of T84 monolayers were treated with buffer or SMase in media for 1 h. Some inserts were washed with sterile media and allowed to recover for 16 h prior to apical intoxication with 3 nM CT (+), while others were intoxicated directly after the SMase/buffer treatment (–). Data are expressed as the percentage of maximal Isc induced by CT in buffer control samples. n = 3 for each sample. H. Maximal Isc induced by CT in cells incubated apically with 30 mM of each indicated lipid (or buffer) for 1 h prior to apical or basolateral intoxication with CT (3 nM). Addition of the lipid species had little effect on baseline cellular secretion (bottom panel). Statistical significance was tested by t-test. *P < 0.001 or 0.0001. © 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
Regulation of CT function by SM and ceramide 69
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
70 D. E. Saslowsky and W. I. Lencer monolayers incubated with buffer alone, fluorescence emanating from the lysenin fusion protein was distributed heterogeneously across the apical pole of cells composing the monolayer (Fig. 1E, left panel). However, in SMase-treated monolayers (apical treatment 50 min, 1 U ml-1), no fluorescence was detected (Fig. 1E, right panel). Thus, the bulk of SM in the apical PM had been hydrolysed to ceramide under these conditions. We next examined the specificity for SM in the action of SMase on T84 cells by testing for the ability of exogenously added SMs, cholesterol or ceramides to reverse the SMase-induced inhibition of CT function. In these studies, T84 cells were treated with SMase, washed and incubated with the indicated reagents before the addition of CT (Fig. 1F). The addition of exogenous SM partially rescued (~40%) the inhibition of CT toxicity caused by SMase (Fig. 1F, compare columns 4 and 5, and Fig. 1A). The addition of exogenous ceramides, however, had no detectable effect, suggesting that the phosphorylcholine head group is required. Rescue was also observed when cells were incubated in serum-free media for an extended period after SMase treatment (3 or 16 h recovery, Fig. 1F and G), presumably due to de novo synthesis and transport of SM to the cell surface (Slotte et al., 1990), or compensatory degradation of membrane ceramide by endogenous ceramides (Huwiler et al., 2000), or both. Thus, the effect of apical SMase treatment on CT-induced Isc is completely reversible. Because cells can respond to SM degradation by rapidly internalizing cholesterol (Slotte et al., 1989; Simons and Ikonen, 2000), and cholesterol depletion from the PM also inhibits CT function (Orlandi and Fishman, 1998; Wolf et al., 2002), we next tested whether SMase may act by affecting membrane cholesterol content. To do this, we used the compound U18666A, which inhibits cholesterol transport from the PM to intracellular compartments (Harmala et al., 1994), and cholesterol-saturated methyl-b-cyclodextrin (MbCD; M/C), which loads T84 cell apical membranes with cholesterol. Neither treatment rescued the SMase-induced inhibition of CT toxicity (Fig. 1F). Thus, gross changes in apical PM cholesterol concentration cannot explain the attenuation of CT function observed in SMase-treated T84 cells. We also considered the possibility that SMase may act by stimulating or inhibiting PKC, either by hydrolysing phosphatidylcholine to produce DAG (PLC-like activity) or through downstream products of ceramide metabolism (Huwiler et al., 2000; Stonehouse et al., 2002). First, we found that pretreatment of cells with the small molecule U73122 to inhibit PLA2 and PLC (Yule and Williams, 1992) did not rescue the inhibition of CT action caused by SMase (data not shown). We also found that the pan-PKC agonist PMA strongly inhibited forskolin-induced Isc in the absence of toxin. This does not phenocopy the effects of
SMase treatment on CT-induced Isc [data not shown and consistent with previous studies (Matthews et al., 1993)]. In addition, pretreatment of monolayers with the PKC inhibitor, calphostin C, did not rescue the SMase-induced attenuation of CT toxicity (data not shown). Finally, as discussed above, we find no effect on toxin function when CT enters the cell through the basolateral membrane. Thus, it is unlikely that activation or inhibition of PLA2, PLC or PKC explains how SMase functions to attenuate CT action. To confirm that the conversion of SM to ceramide caused the inhibition in CT function and to identify the SM/ceramide species involved, we applied the endproducts of the SMase reaction directly to apical membranes of intestinal cells and tested for effects on CT toxicity. Mixed long-chain porcine brain ceramides recapitulated the effect of SMase treatment, reducing the CT-induced Isc by 36% (Fig. 1H, compare columns 1 and 2). No effect on toxicity, however, was observed when mixed long-chain porcine brain SM, synthetic short-chain ceramides (C6, C8), or a non-raft lipid (dioleoylphosphatidylcholine; DOPC) were similarly added (Fig. 1H). A galactose-derivatized short-chain ceramide (C8-GalCer, column 7) actually enhanced CT toxicity, consistent with a previous report (Sharma et al., 2004a). CT toxicity was not affected in cells pre-incubated with 30 mM phosphorylcholine, the isolated SM head group liberated by SMase action (data not shown). As predicted, the apical addition of long-chain ceramides did not alter the toxicity of basolaterally applied CT (Fig. 1H, right two bars), suggesting again that only the apical membrane is affected. Addition of the various lipids had no effect on baseline Isc in the absence of CT (Fig. 1H, bottom bar set) or on forskolin-induced Isc (data not shown). These studies show that the conversion of SM to ceramide in apical membranes of T84 cells attenuates the toxicity of CT entering the cell through the apical PM. The effect is reversible, dose- and time-dependent, and apparently specific to the PM content of SM/ceramide species possessing long acyl chains. Inhibition of CT action in SMase-treated polarized Caco2 monolayers To determine whether these findings can be applied as a general rule for SM and ceramide function in polarized epithelial cells, we tested whether SMase might inhibit CT action on another polarized intestinal cell line, Caco2. CT-induced toxicity in this cell line, which does not have a Cl– secretory phenotype, was measured as an increase in intracellular cAMP. Caco2 cells responded to CT with a robust increase in intracellular cAMP, nearly equivalent to the level of cAMP response caused by addition of the adenylate cyclase agonist forskolin (Fig. 2). As found in
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
fmol cAMP
Regulation of CT function by SM and ceramide 71 apical PM at 4°C as described above, and the content of CT in DRM fractions was assessed by SDS-PAGE and immunoblot using antibodies against CTB. SMase treatment did not prevent the CT–GM1 complex from associating with DRMs (Fig. 3B, fractions 9–12 upper and lower panels), suggesting that SM, like cholesterol, may be dispensable for this interaction. However, we reproducibly found that a larger mass of cell-associated CT was recovered in both the DRM and soluble fractions obtained from T84 monolayers treated with SMase. The proportion of DRM-associated CTB from buffer- and SMase-treated samples was 19% and 20% of total CTB respectively. Effects on toxin trafficking Fig. 2. Apical treatment of the Caco2 intestinal cell line with SMase attenuates cholera toxin function. Apical surfaces of polarized cells were treated with 1 U ml-1 SMase (+) or buffer (–) for 80 min prior to apical intoxication with 10 nM CT or buffer only (–CT) for 60 min. cAMP levels were determined by ELISA as per the manufacturer’s protocol. Data from three independent experiments were normalized using measurements from forskolin-treated monolayers and reported as fmol cAMP per 0.33 cm2 insert. *P < 0.0001.
Trafficking of CT from the PM to the ER is required for toxicity. We first tested for an effect of SMase treatment on
T84 cells, SMase treatment of the apical membrane inhibited CT action (by ~8-fold, Fig. 2). These data show that the attenuation of CT toxicity induced by the conversion of SM to ceramide in the apical PM is not a phenomenon restricted to T84 cells. Conversion of SM to ceramide has no effect on toxin binding to GM1 or association with DRMs To explain why SMase treatment inhibits CT action, we first tested whether SMase somehow reduced the content of GM1 in the apical membrane as assessed by toxin binding to apical cell surfaces at 4°C. Here, the mass of CT bound to GM1 was measured in total cell lysates by SDS-PAGE and immunoblot using antibodies against the CT A-subunit (CTA, Fig. 3A). SMase had no detectable effect on CT binding to T84 cells (compare lanes 1 and 2 with lanes 3 and 4). Binding was specific for GM1 because competition with a molar excess of the CT B-subunit (CTB) completely inhibited CT binding (compare lanes 5 and 6 with lanes 3 and 4). Immunoblotting for b-actin on the same membrane (see stripping protocol) shows that each lane was loaded equally (Fig. 3A, lower panel). Thus, the conversion of SM to ceramide in T84 intestinal cells does not alter the GM1 concentration in the outer leaflet of the apical PM or the ability of CT to bind it. Because association of the CT–GM1 complex at the cell surface with DRMs is closely correlated with toxin function in live cells, we also tested whether the toxin still fractionated with DRMs isolated from T84 monolayers after treatment with SMase. For these studies, CT was bound to the
Fig. 3. Conversion of SM to ceramide has no apparent effect on toxin binding to GM1 or association with DRMs. A. Polarized T84 cells were incubated apically either with buffer or 1 U ml-1 SMase for 50 min prior to apical exposure to 10 nM CT holotoxin at 4°C. As a control for binding specificity, some cells were pre-incubated with a 100-fold molar excess of CTB (CTB-block + SMase). Crude lysates were separated by SDS-PAGE and immunoblotted using an A-subunit pAb (this antibody cross-reacts with the B-subunit). The blot was stripped and reprobed with a b-actin mAb (bottom panel). Each condition is shown in duplicate. B. Polarized T84 cells were treated as in A. Cells were washed and lysed in ice-cold DRM isolation buffer. Lysates (40% sucrose) were separated in a linear 5–30% sucrose density gradient, and fractions (loaded by vol) were analysed by SDS-PAGE and immunoblot (for CTB). Fractions 9–12 correspond to 21.6–26.2% sucrose, consistent with the buoyant density of DRMs in our system (Wolf et al., 2002).
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
72 D. E. Saslowsky and W. I. Lencer Fig. 4. Conversion of SM to ceramide inhibits endocytosis of CT–GM1 complexes but not toxin trafficking to the TGN or ER. A. Polarized T84 cells were incubated apically either with buffer (–) or 1 U ml-1 SMase (+) for 50 min at 37°C prior to apical exposure to 3 nM CT holotoxin at 4°C for 1 h. The samples were shifted to 37°C for the indicated times to promote endocytosis and subsequently returned to 4°C. Cell surface-bound CT was removed by low pH, cells were lysed, and CT was immunoprecipitated from all samples using a CTB pAb. Protein was separated by SDS-PAGE and immunoblotted using CTB pAb (upper panels). Soluble fractions of crude extracts were separated by SDS-PAGE and probed with a b-actin mAb antibody as a loading control (middle panels). B. Densitometry was performed on the immunoblot depicted in A to determine intensities for CTB bands (⫾SMase), where each time point was normalized to 0 min, acid-stripped bands (buffer and SMase, lanes 1 and 2 upper panel respectively). AU, arbitrary units. C. Monolayers were treated as in A, but were exposed to 1 mg ml-1 fluorescent dextran at 4°C, either maintained at 4°C (bars 1 and 2) or warmed to 37°C for 10 min (bars 3 and 4) and then rapidly cooled to 4°C. After washing, cell lysates were analysed by fluorimetry and reported in arbitrary fluorescence units (AFU; mean ⫾ SD, n = 3; left panel). Averaged total crude protein concentration for each sample is also shown (right panel). D. CT-GS, the CT mutant toxin appended with sulfation and N-glycosylation motifs at the C-terminus of the B-subunit, was applied to apical surfaces of T84 cells (20 nM, 37°C) pretreated with buffer (N) or 1 U ml-1 SMase (S) for the indicated times (min). 35S-sulfate was detected by phosphorimaging (upper panels); the B-subunit of CT-GS and b-actin by immunoblot (lower panels). Asterisk denotes band of increased mass due to N-glycosylation. Lower bands in CTB-GS immunoblot are degradation products.
the endocytosis of apical CT–GM1 complexes. T84 monolayers (⫾SMase) were incubated with CT applied to the apical PM at 4°C and either kept at 4°C (Fig. 4A, lanes 1 and 2), or warmed to 37°C for varying times to allow for endocytosis (Fig. 4A, lanes 3–10). After returning the monolayers to 4°C, the CT remaining bound to the cell
surface was removed by incubation at low pH. The mass of CT remaining cell-associated was then assessed by immunoprecipitation of CTB from cell lysates, and analysis by SDS-PAGE and immunoblot. Approximately 99% of toxin is removed from the cell surface by acid stripping (buffer, 98.6 ⫾ 0.6% and SMase, 98.8 ⫾ 0.3%;
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
Regulation of CT function by SM and ceramide 73 Figure S1, compare lanes 2 with 4, and 3 with 5; and data not shown). When monolayers were incubated at 37°C, the amount of toxin associated with the cells after acid stripping was much higher than background (~3- to 20-fold enhanced), and this represents the mass of CT–GM1 internalized by endocytosis (Fig. 4A, compare lanes 1 and 2 with lanes 3–10). Monolayers pretreated with SMase internalized less CT than untreated cells at all time points studied (Fig. 4A, upper panel and quantified by densitometry in Fig. 4B). After 10 min of endocytosis, the mass of CT internalized by SMase treated cells was ~2-fold less than untreated controls (Fig. 4A, compare lanes 5 and 6, and Figure S1). It was determined that in buffer-treated monolayers, ~6% of total surface-bound CT (range 3.4–8%, 730 ⫾ 64 pg, n = 2) was internalized after 10 min at 37°C, whereas ~2% (range 1.7–2%, 309 ⫾ 42 pg, n = 2) was internalized in monolayers treated apically with SMase (determined from Figure S1 and data not shown). The efficiency of CT endocytosis in T84 monolayers is lower than that reported for Caco2 and HeLa cells (Torgersen et al., 2001), possibly due to variations in cell phenotype. As a control for specificity, we also assessed the effect of SMase on fluid-phase pinocytosis of fluorophore-labelled dextran. Here, we found that the conversion of SM to ceramide in the apical PM had no effect on endocytosis of dextran after 10 min of uptake at 37°C (Fig. 4C, left panel, compare columns 3 with 4). Total protein content of the cell lysates was measured for each condition in this assay to control for the equivalency of cell density (Fig. 4C, right panel). We were not able, however, to further verify the specificity of SMase action on the toxin-specific endocytic pathway by measuring internalization of other proteins via the apical membrane. Nonetheless, these results indicate that conversion of SM to ceramide inhibits the efficiency of toxin internalization via the apical PM without grossly affecting fluid-phase uptake or membrane integrity to passive diffusion of large solutes. Cholera toxin–GM1 complexes can enter cells by multiple mechanisms of endocytosis (Torgersen et al., 2001; Massol et al., 2004; Kirkham et al., 2005), and only a fraction of the internalized CT–GM1 is trafficked retrograde to the TGN and ER [~1–2% of cell-associated CT (Fujinaga et al., 2003)]. Thus, to test whether SMase treatment affected toxin transport into the ER, we utilized a recombinant CT mutant (CT-GS) that can biochemically report on trafficking through the Golgi and ER by virtue of a sulfation and N-glycosylation motif appended to the C-terminus of the B-subunit (Fujinaga et al., 2003). When applied to monolayers of polarized T84 cells loaded with 35 S-sulfate, CT-GS becomes radio-labelled by sulfotransferases located in the TGN, but only if the toxin reaches this organelle. Likewise, if the toxin reaches the ER, the B-subunit is N-glycosylated, increasing its mass suffi-
ciently to be detected by SDS-PAGE (Fujinaga et al., 2003). These studies show that, for all time points tested, pretreatment of T84 monolayers with SMase had no effect on the mass of CT-GS reaching the TGN or ER (Fig. 4D). The time required for CT-GS to reach the TGN (detectable after only 20 min of endocytosis; lanes 1 and 2) and ER (slower migrating band denoted with asterisk; lanes 5–10) was similar to previous reports (Fujinaga et al., 2003). Equivalency of toxin input and gel loading was demonstrated by immunoblotting for CTB-GS and b-actin respectively (bottom panels, lanes 1–10). Thus, even though the efficiency of toxin endocytosis is decreased in SMase-treated cells, the mass and efficiency of toxin transport from the apical PM to the ER remains unaltered. Expression and distribution of SMase in intestinal epithelial cells It is possible that SMases endogenous to host cells, in addition to the enzymes produced by intestinal microbes, might physiologically regulate SM-rich membrane microdomains. As such, we examined the expression, location and membrane topology of SMase in our T84 intestinal epithelial cell model. In the human gut, alkaline SMase (alk-SMase) is the predominant isoform expressed (Nilsson and Duan, 1999). It is a transmembrane ectoenzyme that localizes both to mucosal surfaces and intracellular vesicles in situ and in HT-29 cells (Duan et al., 2003), although the latter possess a deleterious mutation in the alk-SMase gene (Wu et al., 2004). Immunoblot analysis of T84 extracts using antibodies raised against alk-SMase (Cheng et al., 2002) detects a protein band with a molecular mass of ~60–65 kDa (Fig. 5A), consistent with the mass of human alk-SMase (Duan et al., 2003). Conversely, acid-SMase (~57 kDa) protein is undetectable (Fig. 5A). Immunofluorescence microscopy of T84 monolayers shows that alk-SMase localizes close to and diffusely across the apical PM (Fig. 5B and C). No signal was detected at the basolateral membrane (data not shown), and immunofluorescence staining for acidSMase was negative (Fig. 5D). Selective cell surface biotinylation was performed to determine whether alk-SMase is oriented ectopically so as to be in a position to hydrolyse membrane SM as hypothesized. Total cell extracts from monolayers that were selectively biotinylated on the apical or basolateral surface, or not biotinylated, were incubated with avidinagarose and the resulting affinity-purified biotinylated proteins subjected to SDS-PAGE and immunoblot analysis using antibodies raised against alk-SMase. An immunoreactive band was detected at the expected mass in extracts from only those monolayers that were apically biotinylated (Fig. 5E, upper panel). Supernatants probed for b-actin controls for gel loading and that resin samples
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
74 D. E. Saslowsky and W. I. Lencer Fig. 5. Alk-SMase is expressed ectopically in the apical PM of polarized T84 cells. A. Crude T84 lysates were separated by SDS-PAGE and analysed by immunoblot using antibodies to either alk- or acid-SMase. B. Optical section acquired at the apical pole of a T84 monolayer fixed and immunostained with antibodies against alk-SMase and the tight junctional protein, ZO-1. The top and side panels are X–Z and Y–Z reconstructions respectively. C. Zoom of the boxed area in B. D. Monolayer treated as in B except immunostained for acid-SMase and ZO-1. Scale bars = 5 mm in B and D, and 1 mm in C. E. Apical or basolateral surfaces of T84 cells were biotinylated as described in Experimental procedures. After avidin-agarose precipitation, samples were analysed by SDS-PAGE and immunoblot. A portion of the avidin-agarose resin along with supernatant was also analysed for b-actin content as control (lower panels). F. SMase activity was measured at apical or basolateral surfaces of polarized T84 monolayers grown on Transwell inserts and reported in arbitrary fluorescence units (AFU). As controls, components of the reaction mixture were omitted, including SM, alkaline phosphatase (AP) and choline oxidase (CO).
contain no detectable b-actin shows that the biotin crosslinking reagent did not gain access to cytosolic proteins (Fig. 5E, bottom two panels). This result shows that at steady state in T84 intestinal cells, alk-SMase is exposed ectopically only at the apical PM. To confirm that the extracellular portion of alk-SMase contained a functional active site, we measured SMase activity on apical and basolateral surfaces of T84 monolayers as assessed by the liberation of phosphorylcholine into the apical and basolateral reservoirs. We detected SMase activity in the apical chamber regardless of the inclusion of exogenous SM as substrate (Fig. 5F, compare bars 3 and 4). Thus, sufficient SM is present in the apical PM and available to SMase for hydrolysis. Negligible activity was apparent in reactions where critical enzymatic components of the assay were omitted (bars 1
and 2, alkaline phosphatase and choline oxidase respectively). Also, much less SMase activity was detected on the basolateral PM (~5-fold less; compare bars 4 and 5), consistent with the immunofluorescence microscopy data. Throughout these studies, monolayer resistance was only nominally affected by the reaction components, indicating that the cells remained intact (Figure S2). These data show that T84 monolayers possess extracellular SMase activity primarily at the apical pole, which we attribute to alk-SMase that is anchored in the apical PM. To confirm in vivo expression of the enzyme, we examined mouse small intestine and found that alk-SMase is abundant in the apical brush border of surface epithelia (Figure S3) as reported before for the rat and human intestine (Cheng et al., 2002; Duan et al., 2003). Thus, it is
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
Regulation of CT function by SM and ceramide 75 possible that SMases endogenous to both host intestinal cells and the intestinal microflora may affect the dynamics of brush border membrane microdomains, which bind the AB5 toxins; however, this remains to be directly tested. Discussion The results of this study show that the enzymatic conversion of SM to ceramide in the apical membrane of intestinal cells attenuates CT toxicity if the toxin enters the cell through the apical PM. This reaction is physiologically relevant because SMases are endogenous to both prokaryotic and eukaryotic cells in the intestinal microenvironment, and they have direct access to SM located in apical membranes of epithelial cells lining the mucosal surface. Although the conversion of SM to ceramide somehow inhibits endocytosis of CT–GM1 complexes, trafficking to the TGN and ER is not detectably altered. Thus, effects on the efficiency of endocytosis, while almost certainly physiologically relevant, cannot fully explain the mechanism of SM/ceramide action on CT function. Rather, the conversion SM to ceramide appears to also affect a reaction(s) located remotely in the ER that is required for retro-translocation of the CT A1-chain to the cytosol. Endocytosis In many systems, endocytosis of the CT–GM1 complex occurs by multiple clathrin-mediated and clathrinindependent mechanisms of endocytosis (Orlandi and Fishman, 1998; Shogomori and Futerman, 2001; Massol et al., 2004; Kirkham et al., 2005). Some cell types internalize CT only by clathrin-independent mechanisms (Deinhardt et al., 2006), and some predominantly by clathrin-mediated mechanisms (Shogomori and Futerman, 2001; Torgersen et al., 2001). We still do not know for certain whether the specific mechanism of toxin uptake explains how the CT–GM1 complex sorts from the PM to ER so as to induce toxicity. In all of our studies, we find that only a small fraction of internalized CT enters the ER (1–2%) (Fujinaga et al., 2003). As endocytosis of CT can be strongly inhibited (as measured morphologically) with no effect on CT-induced toxicity, we have taken the view that in most cell types, including intestinal cells, the bulk of CT enters the cell on GM1 receptors unable to move retrograde from PM to ER. The results of our current study are consistent with this idea. The phenomenon that a subset of GM1 receptors appears incompetent to traffic retrograde might occur because not all GM1 molecules (all of which can bind CT) are structurally identical due to variations in the ceramide domain (our unpublished data and Degroote et al., 2004). Such differences in ceramide structure, we believe, likely affect how GM1 interacts with neighbouring lipids or pro-
teins that impart differential behaviour to CT–GM1 complexes in the plane of the membrane, and thus the dynamics of CT–GM1 vesicular trafficking. It is also possible that structurally identical GM1 molecules are not uniformly able to transport CT to the ER based simply on competition for association with other, as yet unknown, membrane components that are rate-limiting for toxin transport or retro-translocation. Mechanism of action Because apically applied SMase acts only on the fraction of CT entering the cell via the apical cell surface, we believe that SMase must act locally to alter the structure and function of the apical membrane. This appears to occur in a way that somehow affects one or more steps in the mechanism of retro-translocation in the ER, at least with respect to CT as substrate. Our attempts to confirm this idea in T84 cells by direct measurement of CT A1-chain retro-translocation to the cytosol were not successful due to insufficient sensitivity of the assay (described in Forster et al., 2006), but our interpretation of these data is consistent with the results of a recent study on shiga-like toxin (SLTx) entry into HeLa and Vero cells (Smith et al., 2006). In that study, Smith and colleagues also found that altering PM structure by inhibiting glucosylceramide synthesis blocked SLTx function without altering the mass of toxin reaching the ER. The authors postulated, as we do here, that SLTx retro-translocation from the ER to the cytosol must be affected. Presumably, the action of SMase on the apical membrane of T84 intestinal cells is conveyed to the ER by vesicular transport. CT moves from the PM to ER in a fully folded conformation, bound to GM1 (Fujinaga et al., 2003), and it is possible that the apical CT–GM1 complex assembles with SM (and likely other components) to sort from the PM into a specialized region of the ER containing the machinery required for retro-translocation of the A1-chain. How could this work? Unlike SM, ceramides are not amphipathic, but they still efficiently self-assemble into microdomains. In this way, the conversion of SM to ceramide in lipid rafts might cause the ceramide, and perhaps other membrane components, to coalesce into different and possibly much larger structures (Grassme et al., 2003; Wang and Silvius, 2003; Gulbins et al., 2004; Megha and London, 2004). Ceramides can also flip between the outer and inner leaflets of membrane bilayers (Bai and Pagano, 1997; Contreras et al., 2003; LopezMontero et al., 2005), potentially affecting how certain membrane proteins or lipids may phase-partition into the CT–GM1-containing lipid rafts on both sides of the PM. Cholesterol is enriched in lipid rafts and stabilizes such microdomains, in part by forming hydrogen bonds with SM
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
76 D. E. Saslowsky and W. I. Lencer in the plane of the bilayer (Bittman et al., 1994). As ceramide microdomains have a lower affinity for cholesterol than those composed of SM (Megha and London, 2004), the acute enrichment of ceramide in apical membrane lipid rafts after SMase treatment might cause the displacement of cholesterol. Previous studies have shown that acute depletion of membrane cholesterol affects lipid raft function in most cells (Brown and London, 2000; Sharma et al., 2004b), and causes a similar inhibition of CT function in intestinal cells (Orlandi and Fishman, 1998; Wolf et al., 2002). Because the biosynthesis, intracellular transport and homeostasis of SM and cholesterol are tightly coupled, it is not surprising that a perturbation of either species has similar ramifications on GM1 and, presumably, lipid raft function. In fact, cholesterol depletion induces SM degradation by endogenous SMases and, conversely, SM degradation results in the translocation of cholesterol from the PM to intracellular pools in a variety of cell types (Slotte et al., 1989; Simons and Ikonen, 2000). Our results, however, show that the structural change in the apical PM induced by the conversion of SM to ceramide cannot be explained by cholesterol depletion alone. We cannot rescue the SMase-induced inhibition of CT by addition of cholesterol. Rescue only occurs with the addition of SM composed of ceramides containing long-chain fatty acids. We also note that the depletion of membrane cholesterol greatly reduces the TER of T84 monolayers (Orlandi and Fishman, 1998; Wolf et al., 2002), whereas conversion of SM to ceramide in the apical PM has no impact. Thus, the rapid conversion of SM to ceramide likely affects PM microdomains containing CT–GM1 complexes differently compared with cholesterol depletion. Because we were able to partially recapitulate the effect of SMase on CT function by adding exogenous long-chain (but not short-chain) ceramides to the apical PM, we propose that an acute change in the relative content of SM and ceramide in the apical membrane of T84 intestinal cells might be the decisive factor that explains the observed attenuation of CT toxicity after SMase treatment. Such a dependence on the ratio of SM/ceramide would be consistent with previous studies on SMase-induced membrane blebbing during apoptosis (Tepper et al., 2000). It is unlikely that the effect of SMase action on CT function and endocytosis is due to the rapid activation of signal transduction cascades initiated by SM degradation products. This is evidenced most clearly by the failure of chemical inhibitors of ceramide-associated signalling pathways to reverse the phenotype, by the strict cell polarity of the SMase effect (so that freely diffusible factors cannot explain the result), and by the prolonged time required for SMase to act on CT function (consistent with the time required for SMase to hydrolyse the bulk of SM on the apical PM). When applied to basolateral mem-
branes of the same cells, however, SMase acts rapidly within minutes by initiating signal transduction cascades to severely inhibit cAMP-dependant Isc (our unpublished data). This is not seen when ceramide is produced in the apical PM. Although we favour the idea that SMase acts on the apical membrane to affect the structure and function of lipid rafts carrying the CT–GM1 complex into the cell (also suggested by our unpublished data), we cannot exclude the possibility that CT function might be affected by ceramide-induced downstream signalling events that require longer periods of time for actuation, but are nonetheless initiated by fast-acting signalling pathways localized to the apical membrane.
Summary Although the conversion of SM to ceramide in T84 cells affects both toxicity and endocytosis in parallel, our data show that CT transport from PM to the TGN and ER remains intact. Thus, the conversion of SM to ceramide in the apical membrane of host cells must somehow affect a step located remotely in the ER that is required for retrotranslocation of the A1-chain to the cytosol. As SMases are produced by bacteria in the human intestine as well as by intestinal enterocytes (Huwiler et al., 2000; Goni and Alonso, 2002), the modulation of SM and ceramide levels by this class of enzyme may serve as a physiologic regulatory point for lipid raft-based membrane dynamics involved in various aspects of intestinal function and host defence.
Experimental procedures Reagents Long-chain porcine brain ceramides and SMs (acyl chain carbon length primarily C18 and C24 species), synthetic C6, C8 and C8-galactosyl ceramides, and DOPC are from Avanti Polar Lipids (Alabaster, AL); bacterial SMase (Staphylococcus aureus) from Sigma-Aldrich (St Louis, MO); and CT holotoxin and CTB from CalBioChem (La Jolla, CA). HBSS, pH 7.4, was used for all manipulations with live cells. Recombinant lysenin fused to RFP was a kind gift from Dr Toshihide Kobayashi. This fusion protein contains a truncated form of lysenin that cannot oligomerize, but still binds SM specifically (Kiyokawa et al., 2005). Rat anti-human alkaline-SMase was a kind gift from Dr Rui-Dong Duan. Unless indicated, all other reagents were from Sigma-Aldrich (St Louis, MO).
Cell culture Polarized monolayers of T84 cells were cultured on 0.33 or 5 cm2 polyester Transwell® inserts (Corning, Acton, MA) as previously described (Lencer et al., 1995a). Polarized monolayers of Caco2 cells were cultured on 0.33 cm2 polyester Transwell® inserts as previously described (Turner et al., 1996).
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
Regulation of CT function by SM and ceramide 77 Lipid and cholesterol incorporation into cell membranes
Endocytosis assay
Lipids or solvent (100% ethanol) were titrated into buffer containing 0.034% defatted BSA to a final concentration of 30 mM as previously described (Pagano et al., 1991). For incorporation of C6 and mixed long-chain ceramides into apical membranes, these species were titrated to a final concentration of 10 mM into 8 mM a-cyclodextrin in HBSS at 50°C to form a water-soluble complex. Cholesterol was added using MbCD loaded with cholesterol by titration to form a water-soluble complex (M/C, 100 mM final cholesterol) as previously described (Pike and Miller, 1998). The ceramide and cholesterol complexes were applied apically for 2 h or 45 min respectively. For cholesterol depletion, monolayers were incubated in 3 mM MbCD apically for 50 min.
Cells grown on 5 cm2 inserts were treated as above (binding assay). Ice-cold buffer containing 3 nM CT holotoxin was applied apically for 1 h, and the experimental samples were shifted to 37°C for the indicated times (pre-warmed buffer + toxin, apically). Controls for acid stripping efficiency and total bound toxin were kept at 4°C (Figure S1). After extensive washing, monolayers were further washed in either acid solution (20 mM glycine, pH 2.5) or HBSS as follows: 2 ¥ 2 ml of cold acid solution was added to the apical chamber of inserts for 5 min and then 2 ¥ 2 ml of cold HBSS (pH 7.4) was added for 5 min This was repeated twice. Unstripped controls were treated similarly using only cold HBSS. All samples were dunk-washed for 10 min in 500 ml of cold HBSS, excised into lysis buffer, and CTB was immunoprecipitated from cell extracts and analysed by SDS-PAGE and immunoblot as previously described (Lencer et al., 1995b). Results are representative of three independent experiments. For densitometric measurement of band intensities, films were scanned using an Epson Perfection 4990 backlit scanner and analysed using ImageJ software (National Institutes of Health).
Electrophysiology Short-circuit current (Isc) and resistance (R) measurements in electrophysiological studies on polarized T84 monolayers (0.33 cm2 inserts) were performed as previously described (Lencer et al., 1992).
cAMP ELISA Polarized Caco2 cells grown on 0.33 cm2 inserts were treated apically with 1 U ml-1 SMase or HBSS (buffer alone) for 80 min prior to apical intoxication with 10 nM CT holotoxin for 60 min. Parallel samples stimulated basolaterally for 15 min with 10 mM forskolin served as positive controls for data normalization. Filters were excised into 400 ml of lysis buffer (50 mM acetate buffer, pH 5.8, 0.02% BSA, 0.25% dodecyltrimethylammonium bromide), shaken vigorously for 15 min, and centrifuged for 3 min at 15 000 r.p.m. Lysates were analysed for cAMP by ELISA (protocol 3, GE Healthcare Bio-Sciences, Piscataway, NJ).
Immunoblot analysis SDS-PAGE (10–20% Tris-HCl gradient gels) and standard immunoblotting methodologies, including the use of rabbit polyclonal antisera (pAb) against CTA (1:1200) and CTB (1:1500), were as previously described (Lencer et al., 1995a). Where indicated, nitrocellulose membranes were stripped of antibodies for 1 h in stripping buffer [62.5 mM Tris, pH 6.7; 100 mM b-mercaptoethanol (b-me); 2% SDS] and reprobed with b-actin mAb (Sigma) (1:4000).
Toxin binding assay Cells grown on 5 cm2 inserts were incubated for 50 min in buffer ⫾ 1 U ml-1 SMase apically at 37°C, dunk-washed in icecold buffer and allowed to equilibrate in buffer containing 0.1% BSA (included in all subsequent toxin incubations) for 10 min on ice. To show binding specificity for GM1, cells were pre-incubated with a 100-fold molar excess of only CTB. Ten nanomolar CT holotoxin was applied apically for 1 h, and cells were washed five times. The cells were placed in 300 ml of lysis buffer [1% SDS; 1X Protease Complete (PC, Roche Applied Science, Indianapolis, IN); 1 mM b-me; TBS], and boiled 5 min Crude lysates were separated by SDS-PAGE and analysed by immunoblot using antibodies against CTA.
Fluid-phase uptake Cells grown on 0.33 cm2 inserts were assayed for fluid-phase uptake similarly to the endocytosis assay with the following exceptions: AlexaFluor 488-conjugated to 10 000 mol weight dextran (1 mg ml-1; Molecular Probes, Eugene, OR) was substituted for CT, and unbound excess dextran was washed from monolayers using cold buffer (HBSS, pH 7.4). Parallel treatments without fluorescent dextran served as autofluorescence controls. Cells were lysed by excising filters into 150 ml of TBS (pH 7.2), 1% Triton X-100, 1X PC, for 10 min at 37°C and then 3 min at 60°C. Crushed filters were centrifuged at 5000 r.p.m. for 3 min at room temperature, and 125 ml of lysate was transferred to a round-bottom Black-Fluor microtiter plate (Dynex, St Paul, MN) and analysed by fluorimetry. Baseline controls (no fluorescent dextran) were subtracted from all experimental readings (performed in triplicate and averaged). After fluorimetry, protein concentrations were determined for each extract using a standard BCA Protein Assay Kit (Pierce, Rockford, IL).
Detergent-resistant membrane isolation Cells grown on 5 cm2 inserts were treated as above. All subsequent steps were performed at 4°C using ice-cold reagents. DRMs were isolated as previously described (Wolf et al., 1998) with the following exceptions: cells were incubated apically with 10 nM CT holotoxin or buffer (both with 0.1% BSA), washed five times in HBSS, and were scraped into 500 ml of DRM isolation buffer (1% Triton X-100; 1X PC; TBS, pH 7.4). Eighteen or 20 fractions were collected from the top of the gradient, and a portion of each mixed with an equal volume of 2X loading dye (Bio-Rad, Hercules, CA), boiled 5 min, and analysed by SDSPAGE and immunoblot. Sucrose refractometry was performed on fractions to determine sucrose concentration.
Trans-Golgi network/ER trafficking assay A mutant CT (CT-GS) harbouring sulfation and glycosylation motifs appended to the C-terminus of the B-subunit was used
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
78 D. E. Saslowsky and W. I. Lencer to gauge entry to the TGN and ER as previously described (Fujinaga et al., 2003).
Sphingomyelin and SMase localization by fluorescence microscopy Lysenin fused to RFP (2 mg ml-1) was applied apically at 4°C for 60 min to untreated and SMase-treated (1 U ml-1, 50 min) T84 monolayers. Monolayers were then washed, fixed with 4% PFA for 20 min, mounted on slides using ProLong Gold Antifade reagent (Molecular Probes), and imaged by confocal microscopy on a Nikon TE2000 inverted microscope (Nikon Instruments, Melville, NY) coupled to a Perkin-Elmer spinning disk confocal unit (Boston, MA), using a Nikon PlanFluor 40 ¥ (1.3 NA) oil immersion objective lens and an Orca AG scientific-grade cooled CCD camera (Hamamatsu Photonics K.K., Japan). For SMase immuno-localization, polarized T84 monolayers were treated and fixed as above and permeabilized using 0.2% saponin for 30 min at room temperature. Cells were washed, blocked in 10% nonimmune goat serum and immunostained using mouse or rabbit anti-ZO-1 antibodies and either affinity-purified goat anti-human acid-SMase (Santa Cruz Biotechnology, Santa Cruz, CA) or rabbit anti-alkaline-SMase. Samples were incubated with fluorescently labelled secondary antibodies (Molecular Probes) before mounting on slides using ProLong Gold Antifade reagent (Molecular Probes). Confocal images were collected en face to the apical PM using the microscopy equipment described above. For rendering X–Z and Y–Z planes, stacks of X–Y images were acquired at 1 mm intervals.
Cell surface biotinylation Apical or basolateral surfaces of T84 monolayers (⫾SMase) were selectively biotinylated as previously described (Lencer et al., 1995a). After avidin affinity purification, extracts were analysed for alkaline-SMase by SDS-PAGE and immunoblot. Both supernatants and a fraction of the avidin-agarose resin were also analysed for b-actin content by immunoblot.
Assay for SMase activity Sphingomyelinase activity was assayed at either apical or basolateral surfaces of polarized T84 cells grown on 0.33 cm2 inserts using the Amplex Red SMase Assay Kit (Molecular Probes) as indicated by the manufacturer.
Acknowledgements We thank Jessica Wagner for microscopy assistance, Sophie Chabot for immunostaining of mouse intestinal tissue, Eli Kern and Wendy Hamman for invaluable help with tissue culture, and members of the Lencer laboratory for general support and discussions. This work was supported by the National Institutes of Health Grants DK48106 (W.I.L), DK073480 (D.E.S) and DK34854 (Harvard Digestive Disease Center).
References Bai, J., and Pagano, R.E. (1997) Measurement of spontaneous transfer and transbilayer movement of BODIPY-labeled lipids in lipid vesicles. Biochemistry 36: 8840–8848.
Bittman, R., Kasireddy, C.R., Mattjus, P., and Slotte, J.P. (1994) Interaction of cholesterol with sphingomyelin in monolayers and vesicles. Biochemistry 33: 11776–11781. Brown, D., and London, E. (2000) Structure and function of sphingolipid- and cholesterol-rich membrane rafts. J Biol Chem 275: 17221–17224. Chen, H., Born, E., Mathur, S.N., Johlin, F.C. Jr and Field, F.J. (1992) Sphingomyelin content of intestinal cell membranes regulates cholesterol absorption. Evidence for pancreatic and intestinal cell sphingomyelinase activity. Biochem J 286: 771–777. Cheng, Y., Nilsson, A., Tomquist, E., and Duan, R.D. (2002) Purification, characterization, and expression of rat intestinal alkaline sphingomyelinase. J Lipid Res 43: 316– 324. Cheng, Z.J., Singh, R.D., Sharma, D.K., Holicky, E.L., Hanada, K., Marks, D.L., and Pagano, R.E. (2006) Distinct mechanisms of clathrin-independent endocytosis have unique sphingolipid requirements. Mol Biol Cell 17: 3197– 3210. Contreras, F.X., Villar, A.V., Alonso, A., Kolesnick, R.N., and Goni, F.M. (2003) Sphingomyelinase activity causes transbilayer lipid translocation in model and cell membranes. J Biol Chem 278: 37169–37174. Cremesti, A.E., Goni, F.M., and Kolesnick, R. (2002) Role of sphingomyelinase and ceramide in modulating rafts: do biophysical properties determine biologic outcome? FEBS Lett 531: 47–53. Degroote, S., Wolthoorn, J., and van Meer, G. (2004) The cell biology of glycosphingolipids. Semin Cell Dev Biol 15: 375–387. Deinhardt, K., Berninghausen, O., Willison, H.J., Hopkins, C.R., and Schiavo, G. (2006) Tetanus toxin is internalized by a sequential clathrin-dependent mechanism initiated within lipid microdomains and independent of epsin1. J Cell Biol 174: 459–471. Duan, R.D., Cheng, Y., Hansen, G., Hertervig, E., Liu, J.J., Syk, I., et al. (2003) Purification, localization, and expression of human intestinal alkaline sphingomyelinase. J Lipid Res 44: 1241–1250. Falguieres, T., Mallard, F., Baron, C., Hanau, D., Lingwood, C., Goud, B., et al. (2001) Targeting of shiga toxin B-subunit to retrograde transport route in association with detergentresistant membranes. Mol Biol Cell 12: 2453–2468. Feldhaus, M.J., Weyrich, A.S., Zimmerman, G.A., and McIntyre, T.M. (2002) Ceramide generation in situ alters leukocyte cytoskeletal organization and beta 2-integrin function and causes complete degranulation. J Biol Chem 277: 4285–4293. Forster, M.L., Sivick, K., Park, Y.N., Arvan, P., Lencer, W.I., and Tsai, B. (2006) Protein disulfide isomerase-like proteins play opposing roles during retrotranslocation. J Cell Biol 173: 853–859. Fujinaga, Y., Wolf, A.A., Rodigherio, C., Wheeler, H., Tsai, B., Allen, L., et al. (2003) Gangliosides that associate with lipid rafts mediate transport of cholera toxin from the plasma membrane to the ER. Mol Biol Cell 14: 4783–4793. Goggel, R., Winoto-Morbach, S., Vielhaber, G., Imai, Y., Lindner, K., Brade, L., et al. (2004) PAF-mediated pulmonary edema: a new role for acid sphingomyelinase and ceramide. Nat Med 10: 155–160.
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
Regulation of CT function by SM and ceramide 79 Goni, F.M., and Alonso, A. (2002) Sphingomyelinases: enzymology and membrane activity. FEBS Lett 531: 38–46. Grassme, H., Jendrossek, V., Riehle, A., von Kurthy, G., Berger, J., Schwarz, H., et al. (2003) Host defense against Pseudomonas aeruginosa requires ceramide-rich membrane rafts. Nat Med 9: 322–330. Gulbins, E., Dreschers, S., Wilker, B., and Grassme, H. (2004) Ceramide, membrane rafts and infections. J Mol Med 82: 357–363. Harmala, A.S., Porn, M.I., Mattjus, P., and Slotte, J.P. (1994) Cholesterol transport from plasma membranes to intracellular membranes is inhibited by 3 beta-[2-(diethylamino) ethoxy]androst-5-en-17-one. Biochim Biophys Acta 1211: 317–325. van Helvoort, A., van’t Hof, W., Ritsema, T., Sandra, A., and van Meer, G. (1994) Conversion of diacylglycerol to phosphatidylcholine on the basolateral surface of epithelial (Madin-Darby canine kidney) cells. Evidence for the reverse action of a sphingomyelin synthase. J Biol Chem 269: 1763–1769. Huwiler, A., Kolter, T., Pfeilschifter, J., and Sandhoff, K. (2000) Physiology and pathophysiology of sphingolipid metabolism and signaling. Biochim Biophys Acta 1485: 63–99. Johannes, L., and Goud, B. (1998) Surfing on a retrograde wave: how does shiga toxin reach the endoplasmic reticulum? Trends Cell Biol 8: 158–162. Kirkham, M., Fujita, A., Chadda, R., Nixon, S.J., Kurzchalia, T.V., Sharma, D.K., et al. (2005) Ultrastructural identification of uncoated caveolin-independent early endocytic vehicles. J Cell Biol 168: 465–476. Kiyokawa, E., Baba, T., Otsuka, N., Makino, A., Ohno, S., and Kobayashi, T. (2005) Spatial and functional heterogeneity of sphingolipid-rich membrane domains. J Biol Chem 280: 24072–24084. Kolesnick, R.N., Goni, F.M., and Alonso, A. (2000) Compartmentalization of ceramide signaling: physical foundations and biological effects. J Cell Physiol 184: 285–300. Lee, J.T., Xu, J., Lee, J.M., Ku, G., Han, X., Yang, D.I., et al. (2004) Amyloid-beta peptide induces oligodendrocyte death by activating the neutral sphingomyelinase-ceramide pathway. J Cell Biol 164: 123–131. Lencer, W.I., and Saslowsky, D. (2005) Raft trafficking of AB (5) subunit bacterial toxins. Biochim Biophys Acta 1746: 314–321. Lencer, W.I., Delp, C., Neutra, M.R., and Madara, J.L. (1992) Mechanism of cholera toxin action on a polarized human epithelial cell line: role of vesicular traffic. J Cell Biol 117: 1197–1209. Lencer, W.I., Moe, S., Rufo, P.A., and Madara, J.L. (1995a) Transcytosis of cholera toxin subunits across model human intestinal epithelia. Proc Natl Acad Sci USA 92: 10094– 10098. Lencer, W.I., Constable, C., Moe, S., Jobling, M., Webb, H.M., Ruston, S., et al. (1995b) Targeting of cholera toxin and E. coli heat labile toxin in polarized epithelia: role of C-terminal KDEL. J Cell Biol 131: 951–962. Lopez-Montero, I., Rodriguez, N., Cribier, S., Pohl, A., Velez, M., and Devaux, P.F. (2005) Rapid transbilayer movement of ceramides in phospholipid vesicles and in human erythrocytes. J Biol Chem 280: 25811–25819.
Mahfoud, R., Garmy, N., Maresca, M., Yahi, N., Puigserver, A., and Fantini, J. (2002) Identification of a common sphingolipid-binding domain in Alzheimer, prion, and HIV-1 proteins. J Biol Chem 277: 11292–11296. Massol, R.H., Larsen, J.E., Fujinaga, Y., Lencer, W.I., and Kirchhausen, T. (2004) Cholera toxin toxicity does not require functional Arf6- and dynamin-dependent endocytic pathways. Mol Biol Cell 15: 3631–3641. Matthews, J.B., Awtrey, C.S., Hecht, G., Tally, K.J., Thompson, R.S., and Madara, J.L. (1993) Phorbol ester sequentially downregulates cAMP-regulated basolateral and apical Cl- transport pathways in T84 cells. Am J Physiol 265: C1109–C1117. Megha and London, E. (2004) Ceramide selectively displaces cholesterol from ordered lipid domains (rafts): implications for lipid raft structure and function. J Biol Chem 279: 9997–10004. Nilsson, A. (1968) Metabolism of sphingomyelin in the intestinal tract of the rat. Biochim Biophys Acta 164: 575–584. Nilsson, A., and Duan, R.D. (1999) Alkaline sphingomyelinases and ceramidases of the gastrointestinal tract. Chem Phys Lipids 102: 97–105. Nyberg, L., Nilsson, A., Lundgren, P., and Duan, R.-D. (1997) Localization and capacity of sphingomyelin digestion in the rat intestinal tract. J Nutr Biochem 8: 112–118. Orlandi, P.A., and Fishman, P.H. (1998) Filipin-dependent inhibition of cholera toxin: evidence for toxin internalization and activation through caveolae-like domains. J Cell Biol 141: 905–915. Pagano, R.E., Martin, O.C., Kang, H.C., and Haugland, R.P. (1991) A novel fluorescent ceramide analogue for studying membrane traffic in animal cells: accumulation at the Golgi apparatus results in altered spectral properties of the sphingolipid precursor. J Cell Biol 113: 1267–1279. Pelkmans, L., Burli, T., Zerial, M., and Helenius, A. (2004) Caveolin-stabilized membrane domains as multifunctional transport and sorting devices in endocytic membrane traffic. Cell 118: 767–780. Pike, L.J., and Miller, J.M. (1998) Cholesterol depletion delocalizes phosphatidylinositol bisphosphate and inhibits hormone-stimulated phosphatidylinositol turnover. J Biol Chem 273: 22298–22304. Rawat, S.S., Viard, M., Gallo, S.A., Blumenthal, R., and Puri, A. (2006) Sphingolipids, cholesterol, and HIV-1: A paradigm in viral fusion. Glycoconj J 23: 189–197. Ruiz-Arguello, M.B., Basanez, G., Goni, F.M., and Alonso, A. (1996) Different effects of enzyme-generated ceramides and diacylglycerols in phospholipid membrane fusion and leakage. J Biol Chem 271: 26616–26621. Sharma, D.K., Brown, J.C., Choudhury, A., Peterson, T.E., Holicky, E., Marks, D.L., et al. (2004a) Selective stimulation of caveolar endocytosis by glycosphingolipids and cholesterol. Mol Biol Cell 15: 3114–3122. Sharma, P., Varma, R., Sarasij, R.C., Ira, Gousset, K., Krishnamoorthy, G., et al. (2004b) Nanoscale organization of multiple GPI-anchored proteins in living cell membranes. Cell 116: 577–589. Shogomori, H., and Futerman, A.H. (2001) Cholera toxin is found in detergent-insoluble rafts/domains at the cell surface of hippocampal neurons but is internalized via a raftindependent mechanism. J Biol Chem 276: 9182–9188.
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80
80 D. E. Saslowsky and W. I. Lencer Simons, K., and Ikonen, E. (1997) Functional rafts in cell membranes. Nature 387: 569–572. Simons, K., and Ikonen, E. (2000) How cells handle cholesterol. Science 290: 1721–1726. Slotte, J.P., Hedstrom, G., Rannstrom, S., and Ekman, S. (1989) Effects of sphingomyelin degradation on cell cholesterol oxidizability and steady-state distribution between the cell surface and the cell interior. Biochim Biophys Acta 985: 90–96. Slotte, J.P., Harmala, A.S., Jansson, C., and Porn, M.I. (1990) Rapid turn-over of plasma membrane sphingomyelin and cholesterol in baby hamster kidney cells after exposure to sphingomyelinase. Biochim Biophys Acta 1030: 251–257. Smith, D.C., Sillence, D.J., Falguieres, T., Jarvis, R.M., Johannes, L., Lord, J.M., et al. (2006) The association of Shiga-like toxin with detergent-resistant membranes is modulated by glucosylceramide and is an essential requirement in the endoplasmic reticulum for a cytotoxic effect. Mol Biol Cell 17: 1375–1387. Stonehouse, M.J., Cota-Gomez, A., Parker, S.K., Martin, W.E., Hankin, J.A., Murphy, R.C., et al. (2002) A novel class of microbial phosphocholine-specific phospholipases C. Mol Microbiol 46: 661–676. Tabas, I. (1999) Secretory sphingomyelinase. Chem Phys Lipids 102: 123–130. Tepper, A.D., Ruurs, P., Wiedmer, T., Sims, P.J., Borst, J., and van Blitterswijk, W.J. (2000) Sphingomyelin hydrolysis to ceramide during the execution phase of apoptosis results from phospholipid scrambling and alters cellsurface morphology. J Cell Biol 150: 155–164. Torgersen, M.L., Skretting, G., van Deurs, B., and Sandvig, K. (2001) Internalization of cholera toxin by different endocytic mechanisms. J Cell Sci 114: 3737–3747. Turner, J.R., Lencer, W.I., Carlson, S., and Madara, J.L. (1996) Carboxyl-terminal vesicular stomatitis virus G protein-tagged intestinal Na+-dependent glucose cotransporter (SGLT1): maintenance of surface expression and global transport function with selective perturbation of transport kinetics and polarized expression. J Biol Chem 271: 7738–7744. Wang, T.Y., and Silvius, J.R. (2003) Sphingolipid partitioning into ordered domains in cholesterol-free and cholesterolcontaining lipid bilayers. Biophys J 84: 367–378. Wolf, A.A., Jobling, M.G., Wimer-Mackin, S., Madara, J.L., Holmes, R.K., and Lencer, W.I. (1998) Ganglioside structure dictates signal transduction by cholera toxin in polarized epithelia and association with caveolae-like membrane domains. J Cell Biol 141: 917–927. Wolf, A.A., Fujinaga, Y., and Lencer, W.I. (2002) Uncoupling of the cholera toxin-G (M1) ganglioside receptor complex from endocytosis, retrograde Golgi trafficking, and downstream signal transduction by depletion of membrane cholesterol. J Biol Chem 277: 16249–16256.
Wu, J., Cheng, Y., Nilsson, A., and Duan, R.D. (2004) Identification of one exon deletion of intestinal alkaline sphingomyelinase in colon cancer HT-29 cells and a differentiation-related expression of the wild-type enzyme in Caco-2 cells. Carcinogenesis 25: 1327–1333. Yule, D.I., and Williams, J.A. (1992) U73122 inhibits Ca2+ oscillations in response to cholecystokinin and carbachol but not to JMV-180 in rat pancreatic acinar cells. J Biol Chem 267: 13830–13835.
Supplementary material Fig. S1. Endocytosis assay with quantification of internalized CT. Pretreatment of apical surfaces of T84 monolayers with SMase or buffer and incubation with CT is as in Fig. 4A. A. The indicated samples (lanes 6 and 7) were shifted to 37°C for 10 min to promote endocytosis and subsequently returned to 4°C. Cell surface-bound CT was removed by low pH (acid stripped), the efficiency of which is illustrated in samples that were kept at 4°C throughout the experiment (lanes 4 and 5 of the immunoblot; probed with pAb against CTB; upper panel). Total bound CT (not acid stripped) is shown in lanes 2 and 3 (diluted 1:50 from original sample) and was used (along with duplicate experiments, not shown) to estimate the proportion of CT internalized after 10 min at 37°C. Lane 1 shows a defined mass of purified CT standard. Total soluble protein from each sample was analysed by SDS-PAGE and immunoblot using a mAb against b-actin (lower panel). B. Raw densitometric data (arbitrary densitometric units) from the bands in the immunoblot shown in A. Fig. S2. Effects of SMase Amplex Red Assay Kit reaction buffer on T84 monolayer resistance. SMase assay reaction buffer (see Experimental procedures) was applied to the apical surface of high-resistance T84 monolayers (> 1500 W cm-2; addition denoted by asterisk) and resistance was measured. A modest decrease in resistance was observed after the apical addition of SMase assay reagent as compared with buffer-treated monolayers (A); however, these monolayers were still very resistant to electrical current. Inclusion of Triton X-100 (TX-100), which disrupts monolayer integrity, abolished monolayer resistance when included in the reaction buffer (B). These results show that the integrity of both the monolayer and the apical PM were not compromised by the SMase reaction buffer. Therefore, the assay results shown in Fig. 5F likely represent extracellular, not intracellular, SMase activity. Fig. S3. Alk-SMase is expressed at the apical surface of follicle-associated epithelia (FAE) and crypt cells from mouse small intestine. Fresh-frozen sections of mouse small intestine were immunostained with either alk-SMase antibody (A, B, D, E) or rabbit sera from an isotype control animal (C and F) and subsequently with fluorophore-conjugated secondary antibody. Nuclei were counter-stained with DAPI (blue channel). Bar = 10 mm.
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 10, 67–80