/ . Biochem. 82, 1225-1230 (1977)

Hydrolytic Action of Phospholipases on Bacterial Membranes

Ryo TAGUCHI and Hiroh IKEZAWA Faculty of Pharmaceutical Sciences, Nagoya City University, Mizuho-ku, Nagoya, Aichi 467 Received for publication, May 9, 1977

Substrate specificities of phospholipases C [EC 3.1.4.3] from Clostridium novyi, Clostridium perfringens. Bacillus cereus, and Pseudomonas aureofaciens were studied under the same conditions. Phospholipases C from Clostridium novyi and Bacillus cereus show wide substrate specificities while those of Clostridium perfringens and Pseudomonas aureofaciens show relatively narrow specificities. On the basis of these results, the hydrolytic actions of these phospholipases on membrane lipids of Escherichia coli, Bacillus cereus, and Clostridium novyi were examined under the same conditions. The enzymes of Clostridium novyi and Bacillus cereus attacked all the membranes and their lipid extracts, hydrolyzing phosphatidylethanolamine, phosphatidylglycerol, lysophosphatidylethanolamine, and o-aminoacylphosphatidylglycerol. Phospholipase C from Pseudomonas aureofaciens attacked these three membranes and their lipid extracts, hydrolyzing phosphatidylethanolamine. Phospholipase C from Clostridium perfringens hardly attacked the phospholipids of these bacterial membranes. However, phospholipase C from Clostridium perfringens hydrolyzed phosphatidylethanolamine in a mixture containing lipid extract from Escherichia coli membrane and purified phosphatidylcholine from egg yolk.

Phospholipases C [EC 3.1.4.3] obtained from Bacillus cereus and Clostridium perfringens have been used to investigate the structure and function of biological membranes (1-4). We have investigated the actions of phospholipases C from Clostridium novyi, Clostridium perfringens, and Pseudomonas aureofaciens on the membranes of mammalian erythrocytes (5). In the present paper, we describe the substrate specificities and hydrolytic actions on bacterial membranes of phospholipases under the same conditions for the purpose of discrimination of their properties and actions on the membranes.

Abbreviation: SDC, sodium deoxycholate. Vol. 82, No. 5, 1977

1225

MATERIALS AND METHODS All the chemicals used were of analytical reagent grade unless otherwise stated. Preparation of Substrates—Phosphatidylcholine and phosphatidylethanolamine were prepared from egg yolk, and sphingomyelin was obtained from bovine brain by silicic acid column chromatography according to the procedure described by Spanner (6). The purity of these lipid substrates was determined by silica gel H thinlayer chromatography with chloroform/methanol/ water (65:25:4, by vol.) and chloroform/methanol/ acetic acid/water (50:30:8:4, by vol.). Phospholipases—Phospholipases C obtained

1226 from Clostridium novyi, Clostridium perfringens, Bacillus cereus, and Pseudomonas aureofaciens were used in our experiments. The specific activity of each enzyme preparation was expressed in terms of units (micromoles per min) per mg protein, using 2 mM phosphatidylcholine as a substrate in the presence of 0.16% sodium deoxycholate (SDQ and 4 mM MgCl,. Phospholipase C (f-toxin) from Clostridium novyi type A, strain IID 140, was purified as described in the preceding paper (7). This enzyme preparation had a specific activity of 100 and was free from protease, lipase and oxygen-labile 3-hemolysin. Highly purified phospholipase C from Clostridium perfringens type A, strain PB6K (ATCC 10543), was kindly supplied by Dr. Yamakawa and Dr. Ohsaka of the National Institute of Health (8). This enzyme, which had a specific activity of 700, showed a single band in immunodiffusion analysis. Phospholipase C from Bacillus cereus strain I AM 1208 was purified in this laboratory, according to the method of Zwaal et al. (9) with some modifications, by (NH4),SO4 precipitation, and CMSephadex, DEAE-cellulose and Sephadex G-75 column chromatographies. This enzyme, which had a specific activity of 1550, was found to be 90% pure by polyacrylamide disc electrophoresis. Phospholipase C from Pseudomonas aureofaciens strain IFO 3521 was also purified in this laboratory by the method of Sonoki and Ikezawa (70). The enzyme preparation used showed a specific activity of 30, and was found to be 80% pure by disc electrophoresis (the eluate of the first CM-Sephadex chromatography). Assay of Phospholipase C—Phospholipase C activity was determined as described in the preceding paper (7). The standard reaction mixture contained 0.1 ml of 10 mM phosphatidylcholine, 0.1 ml of 0.8% sodium deoxycholate (SDQ, 0.1 ml of 0.2 M borax-HCl buffer (pH 7.0), 0.1ml of 20 ITIM MgCl,, and 0.1 ml of enzyme (about 0.01 unit) in 0.1 % bovine serum albumin. Incubations were carried out for 10 min at 37°C. The reaction was terminated by the addition of 2.5 ml of chloroform/methanol/HCl (66:33:1, by vol.), then the mixture was well shaken and centrifuged. From the upper methanol/water layer, a 0.2 ml aliquot was withdrawn and subjected to phosphate analysis according to the method of Eibl and Lands (11), after decomposition of organic phosphate by

R. TAGUCHI and H. IKEZAWA the method of Fiske and Subbarow (12). Preparation of Bacterial Membranes—Cells of E. coli B, B. cereus strain IAM 1208 and C. novyi strain IID 140 were harvested at the early stationary growth phase by centrifugation at 10,000 X g for 20 min. These cells were suspended in an equal volume of isotonic borate buffer (pH 7.6), consisting of 0.57 g of Na,B 4 O,-10H,O, 2.1 g of H.BO,, 1.8 g of CaCl,-2H,O, and 7.8 g of NaCl per liter, then sonicated for 10 min with cooling in ice bath, and centrifuged at 105,000 xg for 90 min. Precipitates were suspended in the same buffer and centrifuged at 3,000 x g for 10 min. The resulting supernatants were used as the membrane preparations, after determination of their phospholipid contents. Extraction of lipids from each bacterial membrane was carried out by the method of Bligh and Dyer (73). Conditions of the Enzyme Reaction and Phospholipid Analysis—Bacterial membrane or extracted lipid mixture containing about 300 nmol of phospholipids and 0.6 unit of enzyme in 1 ml of isotonic borate buffer (pH 7.6) was incubated for 90 min at 37°C. The reaction mixture was extracted according to the method of Bligh and Dyer (13). This extract was subjected to silica gel H thin-layer chromatography with chloroform/ methanol/water (65:25:4, by vol.) as the solvent. The spot corresponding to each phospholipid was scraped off and the lipids were extracted with formic acid/chloroform/methanol (2:1:1, by vol.) (14). The phosphorus determination of individual phospholipids was performed as described in "Assay of phospholipase C." RESULTS Substrate Specificity of Phospholipase C— Table I shows the relative activity of phospholipase C derived from the four microorganisms towards phospholipids. Sodium deoxycholate (SDQ stimulated these four phospholipase C activities, as reported by several authors (7, 8,10). The sodium deoxycholate concentration optimal for stimulation of these four enzyme activities was 0.1-0.2%, critical for micelle formation at 2 mM of each pure substrate. Phosphatidylcholine-hydrolytic activity of phospholipase C was also stimulated by 1 mM-10 mM MgCl, or CaCl,. However,

J. Biochem.

ACTION OF PHOSPHOLIPASES ON BACTERIAL MEMBRANES

1227

TABLE I. Substrate spccificies of phospholipascs C. The incubation system and assay method described in the text were used, except that the substrates were varied. Activity toward pure phospholipids was examined in the absence (A) or presence (B) of MgClt. Values of activity are expressed in terms of relative activity, taking phosphatidylcholine-hydrolyzing activity in the presence of MgCl, as 1,000. The figures in parentheses represent the specific activity (units/mg protein) of each enzyme, using 2 nw phosphatidylcholine as a substrate in the presence of 0.16% SDC and 4mM MgCl,. Origin of phospholipase C C. novyi C. perfringens B. cereus P. aureofaciens

Phosphatidylcholine A 340 40 790 815

B

1,000(100) 1,000(700) 1,000(1,550) 1,000(30)

Phosphatidylethanolamine Phosphatidylserine A 350 25 415

B 206 13 216

1,560

1,170

the degree of stimulation by these divalent ions varied with the origin of the enzyme and the substrate, as will be described later. In general, the highest activity was observed either in the presence of 1 mM-10 mM or in the absence of MgCl,. Thus the results shown in Table I were obtained in the presence of 0.16% SDC and 2 mM of each substrate, with or without 4 mM MgCl,. Phospholipase C obtained from C. novyi hydrolyzed all of the substrates listed, and the enzyme from B. cereus hydrolyzed most of the substrates except for sphingomyelin (a very low activity toward sphingomyelin seemed to be due to contamination with sphingomyelinase), while the enzyme from C. perfringens mainly hydrolyzed phosphatidylcholine and sphingomyelin, and the enzyme from P. aureofaciens specifically hydrolyzed phosphatidylcholine and phosphatidylethanolamine, as described by Sonoki and Ikezawa {10). Hydrolysis of phosphatidylethanolamine by all of these enzymes was slightly depressed in the presence of 4 mM MgCl,. Furthermore, phospholipase C of C. perfringens did not show any significant activity toward pure phosphatidylethanolamine in the presence of lO-'-lO"1 M MgCl,. Action of Phospholipases C on Lipid Mixtures Extracted from Bacterial Membranes—Table II shows the activity of phospholipases C on lipid mixtures obtained as extracts from the membranes of E. coli, B. cereus, and C. novyi. Phospholipases C from C. novyi, B. cereus, and P. aureofaciens hydrolyzed phospholipids in these three mixtures and released water-soluble organic phosphorus. However, the enzyme from C. perfringens did not Vol. 82, No. 5, 1977

Sphingomyelin

B 71 7 189 2

B 127 206 16 0

TABLE II. Action of phospholipases on lipid mixtures extracted from bacterial membranes. The reaction mixtures, containing about 300 nmol of phospholipids extracted from bacterial membranes and 0.6 unit of enzyme in 1 ml of isotonic borate buffer (pH 7.6), were incubated at 37°C for 90 min. Origin of phospholipase C C. novyi C. perfringens B. cereus P. aureofaciens

Lipid mixture extracted from E. coli

B. cereus

271 16

220 10 136 175

239 204

C. novyi 126* 13 114 72

a

Values in the table are expressed in terms of total phosphorus (nmol) liberated in the methanol/water layer under the conditions described in the text. appreciably hydrolyze any of the phospholipids in these mixtures.

Action of Phospholipases C on Bacterial Membranes—Figure 1 shows the activity of phospholipases C on E. coli membrane. Each column shows the phospholipid content of E. coli membrane after incubation at 37°C for 90 min with (B-E) or without (A) phospholipase C. As a result of autolytic degradation, the phospholipid contents in column A of Figs. 1, 2, and 3 became slightly different from those of the native bacterial membranes. E. coli membrane was attacked by phospholipases C from C. novyi, B. cereus, and P. aureofaciens, but not by phospholipase C from C. perfringens. Similarly, the membranes of both

1228

R. TAGUCHI and H. IKEZAWA

B. cereus and C. novyi were attacked by phospholipases C from C. novyi, B. cereus, and P. aureofaciens, but not by the enzyme from C. perfringens, as shown in Figs. 2 and 3. In these cases, phospholipases C from C. novyi and B. cereus hydrolyzcd phosphatidylethanolamine, phosphatidylglycerol, lysophosphatidylethanolamine, and perhaps also cardiolipin and o-aminoacylphos•150

phatidylglycerol, while phospholipase C from P. aureofaciens mainly hydrolyzed phosphatidylethanolamine. Furthermore, phospholipase C from C. perfringens did not hydrolyze any phospholipid under the conditions examined. Action of Phospholipase C from C. perfringens on a Mixture Containing Lipid Extract from E. coli Membrane and Phosphatidylcholine Purified from Egg Yolk—As described above, phospholipase C from C. perfringens hardly hydrolyzed phosphatidylethanolamine in a pure state (Table I), in bacterial membranes (Figs. 1-3) or in the lipid

•100 rtX)

£ SO •50

ABCDE CL

ABCDE PE

ABCDE PG

ABCDE

L

jI

ABCDE

ABCDE

CL

PE

PG

lysoPE

Fig. 1. Action of phospholipases on E. co//membrane. The reaction mixture, which consisted of E. coli membrane containing about 300 nmol phospholipids and 0.6 unit of enzyme in 1 ml of isotonic borate buffer (pH 7.6), was incubated at 37°C for 90 min. A, Without enzyme; B, phospholipase C from C. novyi; C, phospholipase C from C. perfringens; D, phospholipase C from B. cereus; E, phospholipase C from P. aureofaciens. Abbreviations: CL, cardiolipin; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; lyso PE, lysophosphatidylethanolamine; 0-amino PG, Oaminoacylphosphatidylglycerol.

I

ABCDE

0

ABCDE 0-amino PG • ty»o PE

Fig. 3. Action of phospholipases on C. novyi membrane. The incubation system was as described in Fig. 1, except that C. novyi membrane was used. The symbols A-E and abbreviations are the same as in Fig. 1.

H50

•too

A •oo

50

AC

AC

CL

PE

-0 PG

PC

lysoPE

Fig. 4. Action of phospholipase C from C. perfringens on a mixture containing lipid extract from E. coli memABCDE ABCDE ABCDE ABCDE brane and phosphatidylcholine purified from egg yolk. 0-amino PG The incubation system was as described in Table II, PG CL • lyso PE except that a mixture of about 150 nmol of phospholipids Fig. 2. Action of phospholipases on B. cereus mem- extracted from E. coli membrane and about 150 nmol of brane. The incubation system was as described in Fig. phosphatidylcholine purified from egg yolk was used as 1, except that B. cereus membrane was used. The a substrate. The symbols A, C, and abbreviations are symbols A-E and abbreviation are the same as in Fig. \. the same as in Fig. 1. /. Biochem.

ACTION OF PHOSPHOLIPASES ON BACTERIAL MEMBRANES mixture extracted from these membranes (Table II). However, when choline-containing phospholipids were present in the lipid mixture, phosphatidylethanolamine was hydrolyzed by this enzyme to an extent comparable to that of sphingomyelin, as described in our preceding paper on erythrocyte membranes (5). Thus, we investigated the action of phospholipase C from C. perfringens on a mixture containing lipids extracted from E. coli membrane and phosphatidylcholine purified from egg yolk. In this case, phospholipase C from C. perfringens hydrolyzed about 50% of phosphatidylethanolamine, as shown in Fig. 4. DISCUSSION It was found that phospholipases C from C. novyi and B. cereus exhibit substrate specificities broader than those of the enzymes from C. perfringens and P. aureofaciens (7, 8, 15, 16). Recently, using the culture broth of C. novyi, phosphatidylinositol phospholipase C was separated from the major enzyme activity which hydrolyzed phosphatidylcholine, sphingomyelin and phosphatidylethanolamine (Taguchi and Ikezawa, in preparation). Furthermore, in the case of B. cereus, phosphatidylinositol phospholipase C (17) and sphingomyelinase (to be reported later) were purified separately from the major enzyme activity which hydrolyzes phosphatidylcholine and phosphatidylethanolamine. Yamakawa and Ohsaka (8) showed that with increasing concentration of phosphatidylcholine, higher SDC concentrations were required for maximum stimulation of phospholipase C activity of C. perfringens. According to their report, the optimal molar ratio of SDC to phosphatidylcholine for maximal hydrolytic activity was about 0.5 for dipalmitoyl phosphatidylcholine and about 1.0 for egg phosphatidylcholine. As will be reported elsewhere, a similar molar ratio of SDC to egg phosphatidylcholine was found for phospholipase C of C. novyi, ranging from 1 to 3 at 4 mM MgClj. As reported by Yamakawa and Ohsaka (8) phosphatidylcholine and SDC seem to form conjugated "mixed micelle (s)" at the molar ratio which gives the maximal degree of hydrolysis. In our experiments, however, phosphatidylcholine-SDC mixture was more unstable in the presence of Ca 1+ or Mg t + than in the absence of these divalent Vol. 82, No. 5, 1977

1229

cations, so that the situation is complicated. Phospholipase C from C. perfringens hardly hydrolyzed phosphatidylethanolamine in single micelles or in a lipid mixture extracted from E. coli membrane, but the enzyme hydrolyzed phosphatidylethanolamine when phosphatidylcholine from egg yolk was added to the lipid mixture obtained from E. coli membrane. Bangham and Dawson (18), and De Gier et al. (19) reported that phosphatidylethanolamine was readily attacked in the presence of phosphatidylcholine. However, as we reported earlier, when C. perfringens phospholipase C was incubated with lipid extract of sheep erythrocyte membrane, which contains about 50% sphingomyelin and less than 1 % phosphatidylcholine among total phospholipids, phosphatidylethanolamine was rapidly hydrolyzed at the same rate as sphingomyelin. Thus phosphatidylethanolamine coexisting with choline-containing phospholipids, such as phosphatidylcholine and sphingomyelin, becomes more susceptible to the attack of C. perfringens phospholipase C than it is in the pure state. Furthermore, the accessibility of each phospholipid in a pure state may be different from that in mixed micelles formed with other phospholipids. Bangham and Dawson (18) have reported that a cationic detergent such as cetyltrimethyl ammonium bromide activated the phosphatidylcholine-hydrolyzing activity of phospholipase C of C. perfringens. However, cetyltrimethyl ammonium bromide did not activate the hydrolysis of phosphatides such as phosphatidylcholine and phosphatidylethanolamine by phospholipases C of P. aureofaciens (10) and C. novyi (data not shown). As demonstrated by Takahashi et al. (20) phospholipase C of C. perfringens (a-toxin) exhibits high lethal toxicity towards animals. Phospholipases C from other sources have not been reported to show such toxicity. The lethal toxicity of phospholipase C of C. perfringens may be partly ascribed to the selectivity of its action on the membranes of animal cells, which contain large quantities of choline-containing phospholipids. Op den Kamp et al. (3), Nanninga et al. (2), and Duckworth et al. (1) described the actions of phospholipase C from B. cereus on the membranes of E. coli and B. subtilis. Our results show that the action of phospholipase C from C. novyi on bac-

1230

R. TAGUCHI and H. IKEZAWA

5. Taguchi, R. & Ikezawa, H. (1976) Arch. Biochem. Biophys. 173, 538-545 6. Spanner, S. (1973) in Form and Function of Phospholipids (Ansell, G.B., Hawthorne, J.N., & Dawson, R.M.C., eds.) pp. 43-65, Elsevier, Amsterdam 7. Taguchi, R. & Ikezawa, H. (1975) Biochim. Biophys. Ada 409, 75-85 8. Yamakawa, Y. & Ohsaka, A. (1977) J. Biochem. 81, 115-126 9. Zwaal, R.F.A., Roelofsen, B., Comfurious, P., & Van Deenen, L.L.M. (1971) Biochim. Biophys. Ada 233, 474-479 10. Sonoki, S. & Ikezawa, H. (1975) Biochim. Biophys. Ada 403, 412-424 11. Eibl, H. & Lands, W.E.M. (1969) Anal. Biochem. 36, 51-57 12. Fiske, H. & Subbarow, Y. (1925) /. Bioi. Chem. 66, 375-400 13. Bligh, E.G. & Dyer, W.J. (1959) Canad. J. Biochem. Physiol. 37,911-917 14. Burger, S.P., Fujii, T., & Hanahan, D.J. (1968) The authors thank Dr. Yamakawa and Dr. Ohsaka of Biochemistry 7, 3682-3700 the National Institute of Health for the gift of purified 15. Zwaal, R.F.A. & Roelofsen, B. (1974) in Methods phospholipase C of Clostridium perfringens. in Eniymology (Colowick, S.P. & Kaplan, N.O., This study was supported in part by a Scientific eds.) Vol. 32, pp. 154-161, Academic Press, New Research Grant from the Ministry of Education, Science York and Culture of Japan. 16. Macfarlane, M.G. & Knight, B.C.J.G. (1941) Biochem. J. 35, 884-902 REFERENCES 17. Ikezawa, H., Yamanegi, M., Taguchi, R., Miyasita, T., & Ohyabu, T. (1976) Biochim. Biophys. Ada 1. Duckworth, D.H., Bevers, E.M., Verkleij, AJ., Op 450, 154-164 den Kamp, J.A.F., & Van Deenen, L.L.M. (1974) 18. Bangham, A.D. & Dawson, R.M.C. (1962) Biochim. Arch. Biochem. Biophys. 165, 379-387 Biophys. Ada 59, 103-115 2. Nanninga, N., Tijssen, F.C., & Op den Kamp, J.A.F. 19. De Gier, J., De Haas, G.H., & Van Deenen, L.L.M. (1973) Biochim. Biophys. Ada 298, 184-194 (1961) Biochem. J. 81, 33p-34p 3. Op den Kamp, J.A.F., Kauerz, M.T., & Van Deenen, L.L.M. (1972) /. Bacteriol. 112, 1090-1098 20. Takahashi, T., Sugahara, T., & Ohsaka, A. (1974) Biochim. Biophys. Ada 351, 155-171 4. Verkleij, A.J., Zwaal, R.F.A., Roelofsen, B., Comfurius, P., Kastelijn, D., & Van Deenen, L.L.M. (1973) Biochim. Biophys. Ada 323, 178-193

terial membranes is similar to that of the enzyme from B. cereus. However, as shown in Table II and Fig. 2, phospholipids of the B. cereus membrane were relatively insensitive to attack by the enzyme from B. cereus. In the absence of enzyme, the phospholipids were significantly hydrolyzed by membrane-bound phospholipase through an autolytic process, but further hydrolysis occurred more readily with the enzyme from C. novyi than that from B. cereus. These results suggest that, in the case of hydrolysis of phospholipids in the membrane of B. cereus, the enzyme from C. novyi is more effective. In addition, the behavior of the enzyme from P. aureofaciens was different from those of both the above enzymes. These differences may be useful for investigations of the function of each phospholipid in bacterial membranes by application of these phospholipases.

/. Biochem.

Hydrolytic action of phospholipases on bacterial membranes.

/ . Biochem. 82, 1225-1230 (1977) Hydrolytic Action of Phospholipases on Bacterial Membranes Ryo TAGUCHI and Hiroh IKEZAWA Faculty of Pharmaceutical...
385KB Sizes 0 Downloads 0 Views