Chemistry and Physics of Lipids 159 (2009) 51–57
Contents lists available at ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
Phospholipase C and sphingomyelinase activities of the Clostridium perfringens ␣-toxin Patricia Urbina a , Marietta Flores-Díaz b , Alberto Alape-Girón b,c,d , Alicia Alonso a , Félix M. Goni a,∗ a
Unidad de Biofísica (Centro Mixto CSIC-UPV/EHU), and Departamento de Bioquímica, Universidad del Paˇııs Vasco, Aptdo. 644, 48080 Bilbao, Spain Instituto Clodomiro Picado, Facultad de Microbiología, Universidad de Costa Rica, Costa Rica Departamento de Bioquímica, Facultad de Medicina, Universidad de Costa Rica, Costa Rica d Centro de Investigaciones en Estructuras Microscópicas (CIEMic), Universidad de Costa Rica, San José 2060, Costa Rica b c
a r t i c l e
i n f o
Article history: Received 5 December 2008 Received in revised form 22 January 2009 Accepted 3 February 2009 Available online 28 February 2009 Keywords: Phospholipase C Sphingomyelinase ␣-Toxin Liposome aggregation Clostridium
a b s t r a c t ␣-Toxin is a major pathogenic determinant of Clostridium perfringens, the causative agent of gas gangrene. ␣-Toxin has been known for long to be a phospholipase C, but up to now its hydrolytic properties have been studied only through indirect methods, e.g. release of cell contents, or under non-physiological conditions, e.g., in micelles, or with soluble substrates. In this report we characterize the phospholipase C and sphingomyelinase activities of ␣-toxin using a direct assay method (water-soluble phosphorous assay) with phospholipids in bilayer form (large unilamellar vesicles) in the absence of detergents. The simplest bilayer compositions allowing measurable activities under these conditions were DOPC:Chol (2:1 mol ratio) and SM:PE:Chol (2:1:1 mol ratio) for the PLC and SMase activities respectively. PLC activity was five times higher than SMase activity. Both activities gave rise to vesicle aggregation, after a lag time during which ca. 10% of the substrate was hydrolyzed. Vesicle aggregation, measured as an increase in light scattering, was a convenient semi-quantitative method for estimating the enzyme activities. The optimum pH for the combined PLC and SMase activities was in the 5–7 range, in agreement with the proposed role of ␣-toxin in aiding the bacterium to escape the fagosome and survive within the cytosol. © 2009 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Clostridium perfringens is a Gram-positive anaerobic bacterium that is able to form spores. It is widespread in the enviroment (e.g. in soil and sewage) and is commonly found in the intestines of animals, including humans, where it can be pathogenic under certain circumstances. C. perfringens type A is the most important causative agent of gas gangrene in humans. The organism produces a multiplicity of toxins of which ␣-toxin (370 residues) is the most widely studied (McDonel, 1986). It was the first bacterial toxin to be identified as an enzyme, namely as a phospholipase C (PLC) (Evans, 1943). The preferred substrates are phosphatidylcholine and sphingomyelin (Krug and Kent, 1984). The hemolytic activity of the toxin is unable to be separated from its phospholipase C activity (Nagahama et al., 1995; Guillouard et al., 1996). It is therefore possible that the initial step of hemolysis induced by the toxin is the hydrolysis of phospholipids by the phospholipase C activity. However, little is known about the interaction between the toxin and phospholipids in biological membranes, although the toxin is known to hydrolyse phosphatidylcholine and
sphingomyelin in a dispersed solution or emulsion (Krug and Kent, 1984; Jepson et al., 1999, 2001; Nagahama et al., 2006). Eukaryotic cells have been used as assay models to study the toxin phospholipase and sphingomyelinase activities since ␣-toxin discovery. The hemolytic and cytotoxic effects have been assayed in erythrocytes from different species (Ochi et al., 2003, 2004; Oda et al., 2008) and other types of mammallian cells, e.g. chinese hamster lung fibroblasts and cells from a murine melanoma (Flores-Díaz et al., 2005). Because of the complexity of biological membranes, liposomes composed of phospholipids and cholesterol are often used as simple model systems. In liposomes the ␣-toxin activity was assayed indirectly as liposome aqueous content leakage (Nagahama et al., 1996, 2007; Jepson et al., 1999). The ␣-toxin activity was also studied in monolayers (Moreau et al., 1988; Bianco et al., 1990). In the present paper we have used direct enzyme assay methods, with liposomes as a model membrane system, in order to characterize the ␣-toxin PLC and sphingomyelinase activities. 2. Materials and methods 2.1. Materials
∗ Corresponding author. Tel.: +34 94 601 26 25; fax: +34 94 601 33 60. E-mail address: felix.goni@ehu.es (F.M. Goni). 0009-3084/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.chemphyslip.2009.02.007
Wild type recombinant C. perfringens ␣-toxin from the strain 8–6 expressed in Escherichia coli was purified as described
52
P. Urbina et al. / Chemistry and Physics of Lipids 159 (2009) 51–57
in Alape-Girón et al. (2000). N-2-hydroxyethylpiperazine-N2-ethanesulfonic acid (HEPES) and Tris ultrapure (Tris–HCl) were purchased from Apollo Scientific, NaCl from Fluka and CaCl2 and ZnSO4 from Prolabo. BIS–TRIS and fatty acid-free bovine serum albumin (BSA) were from Sigma. Egg-yolk phosphatidylethanolamine (PE (egg)) was Grade 1 from Lipid Products, egg sphingomyelin (SM (egg)), cholesterol (Chol) and dioleoylphosphatidylcholine (DOPC) were supplied by Avanti Polar Lipids (Alabaster, AL). Chloroform, methanol were purchased from Merck and hydrochloric acid from Carlo Erba. 2.2. Methods The appropriate lipids were mixed in chloroform:methanol (2:1, v/v) and the solvent evaporated thoroughly. Multilamellar vesicles (MLV) were prepared by hydration of the lipid film in buffer with intensive vortexing. The suspension was frozen in liquid nitrogen and defrozen at 37 ◦ C 10 times. Symmetric LUV were prepared by the extrusion method as described in Nieva et al. (1989). Liposomes to be used in phospholipase and sphingomyelinase activity assays were routinely prepared in buffer A (10 mM Tris–HCl, 0.9% (w/v) NaCl, 3 mM CaCl2 , 0.005 mM ZnSO4 , pH 7.5). When studying the buffer effect on the sphingomyelinase and phospholipase acivity, Tris–HCl was substituted by HEPES or BIS–TRIS buffer, maintaining the buffer concentration at 10 mM. Concentrations of other components were kept constant in all cases. For protein dilution buffer A was used (buffer A plus 0.1%, w/v fatty acid-free BSA). SMase and PLC activities were assayed as an increase in water-soluble phosphorous (Ruiz-Argüello et al., 2002). Final LUV concentration was 300 M. The reaction was developed in eppendorf vials at 37 ◦ C with shaking. 50 l aliquots were taken at different times and lipid extraction was performed as follows. Aliquots were added to glass vials containing 250 l of chloroform/methanol/HCl (66:33:1, v/v/v) and vortexed. Samples were centrifuged at 8000 rpm for 1 min in a Minispin Plus eppendorf centrifuge and the aqueous and organic phases were separated for phosphate measurements as described in chapter 2.1.2. The percentage of phosphate liberated at time 0 was subtracted from the values obtained at each time point. SMase and PLC activities on LUV are known to induce extensive vesicle aggregation (Nieva et al., 1989; Basanez et al., 1996b, 1997; Goni and Alonso, 2000). Changes in the light scattering of LUV suspensions caused by the toxin was measured using a SLMAminco Model 8100 spectrofluorimeter (SLM-Aminco, Spectronic Instrument, Barcelona, Spain), setting both excitation and emission wavelengths at 500 nm. Total sample volume was 1000 l, liposome concentration was 300 M, except for the experiments at different LUV concentrations where it varied from 0.05 to 1.3 mM. The reaction was allowed to proceed at 37 ◦ C with continous stirring for approximately 600 s. Maximum slopes were measured on the resulting “light scattering versus time” plots. 3. Results Preliminary experiments had shown that at low concentrations, of the order of ng/ml, the protein lost its activity against liposomes (Fig. S1). To stabilize the enzyme in such a diluted state, we used buffer A , which contains 0.1% (w/v) of BSA, used previously to this aim by other groups, e.g. for the phosphatidylinositol-specific phospholipase C from Bacillus thuringiensis (Qian et al., 1998). After dilution, the protein was incubated for 10 min at 37 ◦ C before starting the reaction. This dilution method enabled us to reach lower toxin concentrations. The best conditions for the ␣-toxin SMase and PLC activities assays include diluting the protein in buffer A and preparing the liposomes in buffer A.
Clostridium perfringens ␣-toxin has been reported to hydrolyze PC and SM in ethanolic dispersions (Krug and Kent, 1984), and disrupt model membranes and release aqueous content from PCor SM-containing liposomes (Nagahama et al., 1996; Jepson et al., 1999). In order to assay the phospholipase C activity of this toxin using a direct method, we performed our experiments on liposomes composed of DOPC:Chol at 2:1 molar ratio. This liposome composition was used in previous experiments, with the Listeria monocytogenes sphingomyelinase/phospholipase toxin, with good results (Montes et al., 2004). DOPC was chosen as enzyme substrate since its gel-fluid transition temperature Tm is −22 ◦ C, ensuring a bilayer fluid state during the experiments, carried out at 37 ◦ C. Membrane fluidity has been claimed to affect toxin binding to phospholipids in liposome bilayers (Nagahama et al., 1996). After toxin addition, aliquots were taken at different times, and phosphorylcholine was quantitated to determine the PLC activity of the toxin (Fig. 1). Experiments at 20 ng/ml toxin showed a linear hydrolysis of DOPC against time with relative low standard desviations. In fact, at 20 ng/ml of toxin within a minute, less than the 10% of the substrate was hydrolysed, a condition usually considered as ideal to measure maximum enzyme rates without the risk of endproduct inhibition (Fig. 1A). The initial reaction rate calculated for the phospholipase activity under these conditions is 1.9 × 10−2 mol phospholipid/g protein s. It is well known that phospholipase activity induces aggregation of liposomes containing PC (Nieva et al., 1989; Goni and Alonso, 2000). Agregation was assayed measuring changes in light scattering. After ␣-toxin addition, DOPC hydrolysis started, accompanied by a slow increase in the light scattering signal (Fig. 1B). This reaction period corresponds to the latency period, or lag phase (Basanez et al., 1996a). After the lag phase, when more than ≈10% of the substrate was hydrolysed (after 1 min), phospholipid hydrolysis displayed a burst in activity, shown in Fig. 1B as a large increase in the light scattering reaction rate. Higher enzyme concentrations lead to higher activities, that exhibited saturation phenomena, and virtual lack of latency periods (Fig. 1C and D). The toxin sphingomyelinase activity was studied on LUV with the composition SM(egg):PE(egg):Chol (2:1:1), that have been used in previous experiments in our group with other toxins with sphingomyelinase activity (Montes et al., 2002) or with a similar dual PLC and SMase activity (Listeria monocytogenes sphingomyelinase/phospholipase) with good results (Montes et al., 2004). Preliminary experiments had shown that SM:Chol (2:1) bilayers allowed only an extremely low SMase activity of ␣-toxin, thus the SM:PE:Chol mixture was used. At 80 ng/ml ␣-toxin, less than the 10% of the substrate was hydrolysed in 1 min, so that the inital reaction rate could be reliably measured (Fig. 2A). The initial reaction rate under the above conditions was 3.8 × 10−3 mol phospholipid/g protein s. Sphingomyelinases, as a result of their activity on SM-containing liposomes, release ceramide (Cer), and in turn Cer induces liposome aggregation, so that the light scattered by the sample increases with time (Nieva et al., 1989; Basanez et al., 1996b, 1997; Goni and Alonso, 2000). We confirmed the correlation between the ␣toxin sphingomyelinase activity and the increase in light scattering (Fig. 2B). After toxin addition, as ␣-toxin hydrolysed the sphingomyelin, the vesicles started aggregating and the light scattering signal increased. When low protein concentration was used (Fig. 2A and B), after ␣-toxin addition, toxin hydrolysed sphingomyelin without an observable increase in the light scattering signal. This reaction period corresponds to the lag phase (Basanez et al., 1996a). After the lag phase, when more than 10% of the substrate was hydrolysed (after 2 min), the phospholipid hydrolysis displayed a burst in activity, shown in Fig. 2B as an increase in light scattering. As with the PLC activity, higher ␣-toxin concentrations lead to a faster and non-linear increase in hydrolysed SM and in light scattering,
P. Urbina et al. / Chemistry and Physics of Lipids 159 (2009) 51–57
53
Fig. 1. Correlation between changes in light scattering and ␣-toxin phospholipase C activity in LUV. Protein concentration is 20 ng/ml (A and B) and 0.5 g/ml (C and D). The upper plot represents the lipid hydrolysed by the toxin and the plot below represents the increase in light scattering after ␣-toxin addition. LUV prepared in buffer A. Toxin previously diluted in buffer A and incubated at 37 ◦ C for 10 min. LUV are composed of DOPC:Chol at 2:1 molar ratio. Lipid concentration is 0.3 mM. Mean values ± S.D. of three independent experiments.
and to a decrease in lag times (Fig. 2C and D). In view of the constant observation of an increase in light scattering accompanying the hydrolytic activity, either PLC or SMase, changes in light scattering were routinely used in the experiments below as a convenient method of enzyme assay. The increase in light scattering induced by the ␣-toxin enzyme activities was recorded at different protein concentrations (Figs. 3 and 4). The experiments were carried out at pH 7.5 in buffer
A. The plot corresponding to PLC activity (Fig. 3) shows that both the maximum slope and the extent of light scattering increase with protein concentration, although they do so in a non-linear way. A change in regime is observed at ca. 100 ng/ml protein. Increasing toxin concentrations cause a dose-dependent decrease in lag time (Fig. 3B). When similar experiments are carried out measuring SMase activity on SM:PE:Chol vesicles, the plot shows a linear relation between the maximum rate of increase in light scattering, and
Fig. 2. Correlation between changes in light scattering and ␣-toxin sphingomyelinase activity in LUV. Protein concentration is 80 ng/ml (A and B) and 1 g/ml (C and D). The upper plot represents the lipid hydrolysed by the toxin and the plot below represents the increase in light scattering after ␣-toxin addition. LUV prepared in buffer A. Toxin previously diluted in buffer Aˇı and incubated at 37 ◦ C for 10 min. LUV are composed of SM (egg):PE (egg):Chol at 2:1:1 molar ratio. Lipid concentration is 0.3 mM. Lower panels: mean values ± S.D. of three independent experiments.
54
P. Urbina et al. / Chemistry and Physics of Lipids 159 (2009) 51–57
optimal activity at pH 6 while the optimum pH for SMase activity is rather observed at ca. pH 7. Tris–HCl buffer has been used in the study of the ␣-toxin activity, but some data had suggested that this buffer was not suitable for the enzyme. Tris–HCl has been demonstrated to be a divalent cation chelator (Aakre and Little, 1982). The ␣-toxin activity depends on the presence of Ca2+ and Zn2+ , and in the absence of these ions the ␣-toxin activity cannot be detected (Klein et al., 1975; Takahashi et al., 1981). We assayed the PLC and sphingomyelinase activities using the same composition as buffer A, but changing the Tris–HCl buffer component for BIS–TRIS or HEPES. There is no report of chelant ability of BIS–TRIS or HEPES. In fact, HEPES has been used in experimentation with proteins whose activities depend on divalent cations (Montes et al., 2002, 2004). Sample pH was 5–6 for PLC, 6–7 for SMase, according to the optimum range of the toxin activities. In all cases, Tris–HCl buffer allowed greater hydrolytic activity than BIS–TRIS or HEPES (Fig. 7). Nagahama et al. (1996) using indirect methods, had also indicated a higher PLC than SMase activity. 4. Discussion
Fig. 3. Phospholipase C activity at different ␣-toxin concentrations. LUV are composed of DOPC:Chol at 2:1 molar ratio and the concentration is 0.3 mM. Protein concentration varies from 0.02 to 1 g/ml. (A) Maximum slope (䊉) and maximum extent of the increase in light scattering ( ) caused by the reaction are plotted against the toxin concentration. (B) Lag time plotted of the above experiments against toxin concentration. Mean values ± S.D. of three independent experiments.
the protein concentration (Fig. 4). A linear correlation is also found between the amplitude (extent) of the change in light scattering, and toxin concentration (Fig. 4A). Fixing the substrate concentration, the maximum aggregation rate is not reached by the toxin even at 1 g/ml protein. Lag times decrease in a dose-dependent way with ␣-toxin concentration (Fig. 4B). Knowing the order of reaction the mechanistic relationship between substrate concentration and rate can be established. Reaction order determines how the amount of a reagent speeds up or retards a reaction. To determine the apparent reaction order we chose a protein concentration in the middle zone of the plot in Figs. 3 and 4, 0.5 g/ml, and assayed the PLC or sphingomyelinase activity as an increase in the sample light scattering at different liposome concentrations, from 0.05 to 1.3 mM. The observed rates of aggregation (light scattering) were corrected for the lipid concentration relative to 50 M (the lowest liposome concentration used) and plotted in a double-logarithmic manner against the total lipid concentration, as described by Wilschut et al. (1980). The resulting plot can be seen in Fig. 5. The corresponding slope indicates the reaction order with respect to the lipid (or vesicle) concentration. Vesicle aggregation is expected to be second-order with respect to lipid concentration, since it depends on vesicle–vesicle contacts. A slope of approximately 1.8 is found experimentally for the PLC activity, which is compatible with a second-order reaction. For the SMase activity, a slope of 1.6 is found experimentally, suggesting that the rate-limiting phenomenon is not only vesicle–vesicle contacts. The optimum pH for each of the ␣-toxin activities was studied. LUV concentration used was 0.3 mM and protein concentration was 20 ng/ml for PLC and 80 ng/ml for the SMase activity. The reaction was developed at 37 ◦ C in buffer A adjusted to different pHs and the change in light scattering was recorded. Maximum slopes at different pH are represented in Fig. 6. Phospholipase C activity shows an
The present study provides direct evidence of phospholipase C and sphingomyelinase activities of C. perfringens ␣-toxin. Under comparable conditions, the initial reaction rate is five-fold higher for phospholipase activity than for sphingomyelinase activity (PLC inital rate = 1.9 × 10−2 mol phospholipid/g protein s, vs. SMase initial rate = 3.8 × 10−3 mol phospholipid/g protein s). In the present manuscript a direct enzyme assay method, using liposomes as a model membrane system, has been established, and shown to correlate with vesicle aggregation. This kind of direct measurements
Fig. 4. Sphingomyelinase activity at different ␣-toxin concentrations. LUV are composed of SM (egg):PE (egg):Chol at 2:1:1 molar ratio, and the concentration is 0.3 mM. Protein concentration varies from 0.08 to 1 g/ml. (A) Maximum slope (䊉) and maximum extent of the increase in light scattering ( ) caused by the reaction are plotted against the toxin concentration. (B) Lag time plotted of the above experiments against toxin concentration. Mean values ± S.D. of three independent experiments.
P. Urbina et al. / Chemistry and Physics of Lipids 159 (2009) 51–57
Fig. 5. Influence of lipid concentration on the rate of vesicle aggregation induced by ␣-toxin. Rates are multiplied by the “corrected lipid concentration”, relative to the lower lipid concentration tested (50 M). Effect of varying the concentration of total lipid at a constant enzyme concentration 0.5 g/ml. In (A), PLC activity where LUV are composed of DOPC:Chol at 2:1 molar ratio, and in (B) SMase activity where LUV are composed of SM (egg):PE (egg):Chol at 2:1:1 molar ratio. Vesicle suspensions are prepared in buffer A at pH 7.5.
of phospholipase activities are key in understanding the behaviour and properties of enzymes, and in principle they do not require the use of purified enzyme sources. In recent years, similar protocols have been developed for the assay of related enzymes, such as Bacillus cereus sphingomyelinase (Ruiz-Argüello et al., 2002), or Listeria monocytogenes phospholipase C/sphingomyelinase (Montes et al., 2004). ␣-Toxin is sensitive to the dilution effect which inactivates its enzymatic activity. The dilution buffer must contain 0.1% (w/v) BSA in order to avoid toxin inactivation, and enable the protein to interact effectively with the liposome bilayer. A similar effect of toxin inactivation due to high dilution, and the protective effect of 0.1% (w/v) BSA had been shown previously (Takahashi et al., 1981). Phospholipase and sphingomyelinase activities of the toxin, in addition to being different in magnitude, have different optimum pH values. This might imply the possibility of two active sites in the protein structure. However this would be in disagreement with the structural data published by Naylor et al. (1998). The latter authors determined the three-dimensional structure of the ␣-toxin by Xray crystallography. In the structure they accommodated a modeled phospholipid in a single active site cleft without major stereochemical clashes with the protein residues. In the ␣-toxin model for accommodating a PC molecule in the active site only the N-domain is needed, while for sphingomyelin docking the ceramide moiety interacts also with the C-domain residues 295–305 in the model proposed by Naylor et al. (1998) (Fig. 8). This model is consistent with both the observation that the presence of the C-domain enhances sphingomyelinase activity (Titball, 1993) and the fact that a truncated ␣-toxin, which has no
55
C-domain, cannot hydrolyse sphingomyelin. The difference in optimum pH for both activities can be attributed to the implication of more protein residues in the toxin sphingomyelinase activity. Our results showed that ␣-toxin has phospholipase C and sphingomyelinase optimum activities at pH 6 and 7 respectively. pH did not change significantly during the enzyme assay (data not shown). It was expected that, as the bacteria is phagocytosed by the macrophages (O’Brien and Melville, 2004), an acidic optimum pH (around pH 5, the lysosomal pH) could explain its physiological role to escape from the phagosome. Both activities, and particularly phospholipase C, appear to operate very well at pH 5 (Fig. 6). Thus the bacteria can escape the early phagosome and stay in the cytoplasm for a long period, until suitable conditions for infection appear. Cytoplasmic pH is more neutral, but again from the data in Fig. 6, it appears that the two optimum pH are well suited for the toxin to be involved in escaping and killing the cell. In agreement with these findings, O’Brien and Melville (2004) have reported that C. perfringes requires ␣-toxin to survive intracelluarly within the pahgocytes, in order to escape from the endosomes to the cytosol, where it stays for long periods, until suitable conditions for growth appear. Buffer election is an important step in order to assay any protein activity. In our case, divalent cations (Zn2+ and Ca2+ ) are essential for the toxin activity. The optimal concentrations for these ions are used in our experiments (Zn2+ ) = 5 M and (Ca2+ ) ≈ 5 mM (Guillouard et al., 1997). Any chelator in the buffer composition could affect our results. As described by Aakre and Little (1982), Tris–HCl buffer is a chelator but at concentrations ten times above the one we use in our experiments. Despite this, we tested different buffers used previously in other experiments with enzymes whose activity is
Fig. 6. Phospholipase C (A) and SMase (B) activities of ␣-toxin at different pHs. LUV are composed of DOPC:Chol (2:1) for PLC activity and SM (egg):PE (egg):Chol at 2:1:1 molar ratio for SMase activity. Protein concentration is 20 ng/ml (A) and 80 ng/ml (B). Lipid concentration is 0.3 mM. The reaction maximum slope is plotted against pH of the reaction medium. Mean values ± S.D. of three independent experiments.
56
P. Urbina et al. / Chemistry and Physics of Lipids 159 (2009) 51–57
divalent cation dependent. Surprisingly, our results showed that Tris–HCl was the most suitable buffer assayed as compared with BIS–TRIS or HEPES. Looking at the interacting groups of the EDTA molecule, hydroxyl groups and the nitrogen atom can interact with a divalent cation and chelate it. The three buffer molecules have these interacting groups in its structure which can imply that all three buffers are good candidates to be chelators. Despite this, apart from the data published in relation to Tris–HCl chelant ability, no published data have been found about any chelant properties of the other two buffers assayed. For our purpose, Tris–HCl is a good buffer for experimentation with ␣-toxin and it has been used with this toxin in other experiments in model membranes (Moreau et al., 1988; Nagahama et al., 1996). In many kinetic problems, the first order of business is to determine the reaction order. The reaction order is simply the sum of the exponents on the concentration terms for a rate law: Rate = k(A)x (B)y
(1)
Reaction order = x + y
(2)
The exponents x, y, . . . (Eqs. (1) and (2)) can be positive or negative, integral or rational nonintegral numbers. They are the reaction orders with respect to A, B, . . . and are sometimes called ‘partial reaction orders’. Our results shows a second reaction order for the phospholipase activity so that the reaction rate increases with the square of the substrate concentration (Eq. (3)). Rate = k(DOPC)2
Fig. 7. Buffer effect on the ␣-toxin activity. (A) For phospholipase activity LUV are composed of DOPC:Chol (2:1) and protein concentration is 20 ng/ml toxin. (B) For sphingomyelinase activity LUV are composed of SM (egg):PE (egg):Chol at 2:1:1 molar ratio and protein concentration is 80 ng/ml. Lipid concentration is 0.3 mM. Maximum slope of the reaction is plotted for each buffers used at different pH. Mean values ± S.D. of three independent experiments.
(3)
For the sphingomyelinase activity the apparent reaction order is not an integer, 1.6. The non-integral order of reaction for sphingomyelinase activity might be due to the simultaneous participation of alternative reaction mechanisms that can in turn be first- and second-order. This non-integral number is an average that shows the proportion in which each mechanism takes part in the total reaction. It is possible to find non-integral numbers in many
Fig. 8. GRASP39 electrostatic potential surface of ␣-toxin (shown with the active site cleft and membrane-binding surface uppermost). The protein has only one active site where a modeled phospholipid can be accommodated without major stereochemical clashes with the protein residues. The positive oxyanion hole for the phosphate head group is visible at the base of the cleft, as indicated by a modeled sphingomyelin. Some residues thought to be important for membrane binding have been indicated. The electrostatic potential is scaled from red for −20.0 to blue for +20.0 (Naylor et al., 1998).
P. Urbina et al. / Chemistry and Physics of Lipids 159 (2009) 51–57
reactions that imply intermediates. In our case, through changes in light scattering, we are in fact observing a two-step phenomenon, namely lipid hydrolysis and subsequent vesicle aggregation. In some cases enzymes change partially their specifity and start hydrolysing other substrates. Among other explanations, it is possible that ␣-toxin once is hydrolysing SM, starts hydrolysing PE (egg) as well. Some indirect data in the literature suggest that ␣-toxin may hydrolyze phospholipids other than PC or SM under certain conditions (Stahl, 1973; Krug and Kent, 1984). This could explain the complex apparent reaction order. Acknowledgements This work was supported in part by grants from the Spanish Ministerio de Ciencia e Innovación No. BFU-2005-06095 (A.A) and No. BFU-2007-062062 (F.M.G.) and from the Basque Goverment, No. IT-461-07 (F.M.G.), CR-USA foundation (CT13-02), Vicerrectoria de Investigación, Universidad de Costa Rica (741-A3-503). P.U. was a Ph. D. student supported by the Basque Government. The authors also acknowledge financial help from the CSIC/Universidad de Costa Rica Joint Project 2008CR0030. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.chemphyslip.2009.02.007. References Aakre, S.E., Little, C., 1982. Inhibition of Bacillus cereus phospholipase C by univalent anions. Biochem. J. 203, 799–801. Alape-Girón, A., Flores-Díaz, M., Guillouard, I., Naylor, C.E., Titball, R.W., Rucavado, A., Lomonte, B., Basak, A.K., Gutiérrez, J.M., Cole, S.T., Thelestam, M., 2000. Identification of residues critical for toxicity in Clostridium perfringens phospholipase C, the key toxin in gas gangrene. Eur. J. Biochem. 267, 5191–5197. Basanez, G., Nieva, J.L., Goni, F.M., Alonso, A., 1996a. Origin of the lag period in the phospholipase C cleavage of phospholipids in membranes. Concomitant vesicle aggregation and enzyme activation. Biochemistry 35, 15183–15187. Basanez, G., Nieva, J.L., Rivas, E., Alonso, A., Goni, F.M., 1996b. Diacylglycerol and the promotion of lamellar-hexagonal and lamellar-isotropic phase transitions in lipids: implications for membrane fusion. Biophys. J. 70, 2299–2306. Basanez, G., Ruiz-Argüello, M.B., Alonso, A., Goni, F.M., Karlsson, G., Edwards, K., 1997. Morphological changes induced by phospholipase C and by sphingomyelinase on large unilamellar vesicles: a cryo-transmission electron microscopy study of liposome fusion. Biophys. J. 72, 2630–2637. Bianco, I.D., Fidelio, G.D., Maggio, B., 1990. Effect of sulfatide and gangliosides on phospholipase C and phospholipase A2 activity. A monolayer study. Biochim. Biophys. Acta 1026, 179–185. Evans, D.G., 1943. The protective properties of the alpha toxin antitoxin and the theta haemolysin in Cl. welchii type A antiserum. Br. J. Exp. Pathol. 24, 81–88. Flores-Díaz, M., Alape-Girón, A., Clark, G., Catimel, B., Hirabayashi, Y., Nice, E., Gutierrez, J.M., Titball, R., Thelestam, M., 2005. A cellular deficiency of gangliosides causes hypersensitivity to Clostridium perfringens phospholipase C. J. Biol. Chem. 280, 26680–26689. Goni, F.M., Alonso, A., 2000. Membrane fusion induced by phospholipase C and sphingomyelinases. Biosci. Rep. 20, 443–463. Guillouard, I., Garnier, T., Cole, S.T., 1996. Use of site-directed mutagenesis to probe structure-function relationships of alpha-toxin from Clostridium perfringens. Infect. Immun. 64, 2440–2444. Guillouard, I., Alzari, P.M., Saliou, B., Cole, S.T., 1997. The carboxy-terminal C2-like domain of the alpha-toxin from Clostridium perfringens mediates calciumdependent membrane recognition. Mol. Microbiol. 26, 867–876. Jepson, M., Howells, A., Bullifent, H.L., Bolgiano, B., Crane, D., Miller, J., Holley, J., Jayasekera, P., Titball, R.W., 1999. Differences in the carboxy-terminal (Putative
57
phospholipid binding) domains of Clostridium perfringens and Clostridium bifermentans phospholipases C influence the hemolytic and lethal properties of these enzymes. Infect. Immun. 67, 3297–3301. Jepson, M., Bullifent, H.L., Crane, D., Flores-Díaz, M., Alape-Girón, A., Jayasekeera, P., Lingard, B., Moss, D., Titball, R.W., 2001. Tyrosine 331 and phenylalanine 334 in Clostridium perfringens alpha-toxin are essential for cytotoxic activity. FEBS Lett. 495, 172–177. Klein, R., Miller, N., Kemp, P., Laser, H., 1975. The activation of phospholipase C from Clostridium welchii by quinine: an absolute requirement for calcium ions. Chem. Phys. Lipids 15, 15–26. Krug, E.L., Kent, C., 1984. Phospholipase C from Clostridium perfringens: preparation and characterization of homogeneous enzyme. Arch. Biochem. Biophys. 231, 400–410. McDonel, J.L., 1986. Toxins of Clostridium perfringens type A, B, C, D and E. In: Dorner, F., Drews, J. (Eds.), Pharmacology of Bacterial Toxins. Montes, L.R., Ruiz-Argüello, M.B., Goni, F.M., Alonso, A., 2002. Membrane restructuring via ceramide results in enhanced solute efflux. J. Biol. Chem. 277, 11788–11794. Montes, L.R., Goni, F.M., Johnston, N.C., Goldfine, H., Alonso, A., 2004. Membrane fusion induced by the catalytic activity of a phospholipase C/sphingomyelinase from Listeria monocytogenes. Biochemistry 43, 3688–3695. Moreau, H., Pieroni, G., Jolivet-Reynaud, C., Alouf, J.E., Verger, R., 1988. A new kinetic approach for studying phospholipase C (Clostridium perfringens alpha toxin) activity on phospholipid monolayers. Biochemistry 27, 2319–2323. Nagahama, M., Okagawa, Y., Nakayama, T., Nishioka, E., Sakurai, J., 1995. Site-directed mutagenesis of histidine residues in Clostridium perfringens alpha-toxin. J. Bacteriol. 177, 1179–1185. Nagahama, M., Michiue, K., Sakurai, J., 1996. Membrane-damaging action of Clostridium perfringens alpha-toxin on phospholipid liposomes. Biochim. Biophys. Acta 1280, 120–126. Nagahama, M., Otsuka, A., Sakurai, J., 2006. Role of tyrosine-57 and -65 in membrane-damaging and sphingomyelinase activities of Clostridium perfringens alpha-toxin. Biochim. Biophys. Acta 1762, 110–114. Nagahama, M., Otsuka, A., Oda, M., Singh, R.K., Ziora, Z.M., Imagawa, H., Nishizawa, M., Sakurai, J., 2007. Effect of unsaturated bonds in the sn-2 acyl chain of phosphatidylcoline on the membrane-damaging action of Clostridium perfringens alpha-toxin towards liposomes. Biochim. Biophys. Acta 1768, 2940–2945. Naylor, C.E., Eaton, J.T., Howells, A., Justin, N., Moss, D.S., Titball, R.W., Basak, A.K., 1998. Structure of the key toxin in gas gangrene. Nat. Struct. Biol. 5, 738– 746. Nieva, J.L., Goni, F.M., Alonso, A., 1989. Liposome fusion catalytically induced by phospholipase C. Biochemistry 28, 7364–7367. O’Brien, D.K., Melville, S.B., 2004. Effects of Clostridium perfringens alpha-toxin (PLC) and perfringolysin O (PFO) on cytotoxicity to macrophages, on escape from the phagosomes of macrophages, and on persistence of C. perfringens in host tissues. Infect. Immun. 72, 5204–5215. Ochi, S., Oda, M., Nagahama, M., Sakurai, J., 2003. Clostridium perfringens alpha-toxininduced hemolysis of horse erythrocytes is dependent on Ca2+ uptake. Biochim. Biophys. Acta 1613, 79–86. Ochi, S., Oda, M., Matsuda, H., Ikari, S., Sakurai, J., 2004. Clostridium perfringens alphatoxin activates the sphingomyelin metabolism system in sheep erythrocytes. J. Biol. Chem. 279, 12181–12189. Oda, M., Matsuno, T., Shiihara, R., Ochi, S., Yamauchi, R., Saito, Y., Imagawa, H., Nagahama, M., Nishizawa, M., Sakurai, J., 2008. The relationship between the metabolism of sphingomyelin species and the hemolysis of sheep erythrocytes induced by Clostridium perfringens alpha-toxin. J. Lipid Res. 49, 1039–1047. Qian, X., Zhou, C., Roberts, M.F., 1998. Phosphatidylcholine activation of bacterial phosphatidylinositol-specific phospholipase C toward PI vesicles. Biochemistry 37, 6513–6522. Ruiz-Argüello, M.B., Veiga, M.P., Arrondo, J.L., Goni, F.M., Alonso, A., 2002. Sphingomyelinase cleavage of sphingomyelin in pure and mixed lipid membranes. Influence of the physical state of the sphingolipid. Chem. Phys. Lipids 114, 11–20. Stahl, W., 1973. Phospholipase C and specificity with respect to individual phospholipids and brain microsomal membrane phospholipids. Arc. Biochem. Biophys. 154, 47–55. Takahashi, T., Sugahara, T., Ohsaka, A., 1981. Phospholipase C from Clostridium perfringens. Methods Enzymol. (71 Pt C), 710–725. Titball, R.W., 1993. Bacterial phospholipases C. Microbiol. Rev. 57, 347–366. Wilschut, J., Duzgunes, N., Fraley, R., Papahadjopoulos, D., 1980. Studies on the mechanism of membrane fusion: kinetics of calcium ion induced fusion of phosphatidylserine vesicles followed by a new assay for mixing of aqueous vesicle contents. Biochemistry 19, 6011–6021.