JOURNAL OF BACTERIOLOGY, July 1979, p. 132-136
Vol. 139, No. 1
0021-9193/79/07-0132/05$02.00/0
Distribution of Glycerophospholipid-Cholesterol Acyltransferase in Selected Bacterial Species S. MAcINTYRE, T. J. TRUST, AND J. T. BUCKLEY* Department of Biochemistry and Microbiology, University of Victoria, Victoria, British Columbia, Canada V8W2Y2 Received for publication 12 February 1979
The distribution of glycerophospholipid-cholesterol acyltransferase in selected bacterial species was examined. Enzyme activity was demonstrated in cell-free growth media from all members of the family Vibrionaceae which were tested except Plesiomonas shigelloides. In each case, enzyme was produced in exponential to early stationary phase and was excluded from Sepharose 6B, indicating a complex of high molecular weight. In a limited survey of other families, Staphylococcus aureus was the only organism outside the Vibrionaceae which was shown to produce the enzyme. In this case, however, the enzyme exhibited much less activity against erythrocyte membranes and appeared to have a lower molecular weight. The reasons for these differences and the importance of the acyltransferase as a biochemical identification tool are discussed.
Although the biochemical modes of action of only a few bacterial toxins are known with certainty, indications are that a wide range of different attack mechanisms may be utilized. Members of the Vibrionaceae produce several extracellular proteins which serve as virulence factors (11). Until recently most interest has been focussed on choleragen, the first of these toxins to be reported. The toxins produced by Vibrio species other than Vibrio cholerae are less well characterized, even though species such as Vibrioparahaemolyticus are significant pathogens of humans (1). Toxins secreted by members of the genus Aeromonas have received even less attention. Yet Aeromonas salmonicida has long been recognized as a significant pathogen of fish, causing large mortalities due to furunculosis (13), and Aeromonas hydrophila, which causes septicemic infections in warm-water fishes (8), is being increasingly implicated in a variety of clinical infections of humans (10). The extracellular toxins produced by these aeromonads are obviously important virulence factors. One of the extracellular proteins produced by A. hydrophila, the hemolytic toxin aerolysin, has been partially purified (4). However, its mode of action is unknown. The properties of the other toxins have not been elucidated, although the properties of the enterotoxin may be similar to those of choleragen (17). Indeed, the number of other toxins has not been established. A. hydrophila has, however, recently been reported to produce a glycerophospholipid-choles-
terol acyltr&nsferase-lysophospholipase enzyme complex of high apparent molecular weight (12). Because of the possibility of a contribution of this activity to the pathogenicity of A. hydrophila and perhaps other aeromonads, or even all members of the Vibrionaceae, the extent of its distribution in selected bacterial species was determined. In this paper we describe its occurrence in various bacterial genera and species and compare the levels of enzyme activity in some species which produce it. MATERIALS AND METHODS Bacterial strains. The organisms used and their sources are listed in Table 1. The identity of all organisms was confirmed before use by means of staining characteristics and biochemical testing (6). Strains were maintained at 4°C on Trypticase soy agar which contained 3.5% sodium chloride for V. parahaemolyticus. Growth and sampling conditions. All organisms were grown in yeast extract-Casamino Acid broth (12) with 0.25% glucose. The culture medium for V. parahaemolyticus also contained 3.5% sodium chloride. Bacterial cultivation was normally at 30°C with gyratory shaking at 300 rpm. Cell growth was monitored by measuring culture turbidity in a Klett-Summerson photoelectric colorimeter, using the green filter. Samples were withdrawn at the time intervals indicated below, centrifuged at 14,000 x g for 20 min at 4°C, dialyzed against three changes of 20 mM Tris (pH 7.4) for 48 h, and frozen until required. In some cases dialysis and freezing were omitted and cholesteryl ester production was measured directly. Strains of Staphylococcus aureus were also grown
132
BACTERIAL CHOLESTEROL ACYLTRANSFERASE
VOL. 139, 1979 TABLE 1. Bacterial strains Strain
Source'
Achromobacter xerosis 14780 ...... ATCC Aeromonas hydrophila Ah65 ........ UVIC UVIC A. hydrophila AhllO ............... A. hydrophila 9071 ................. ATCC A. proteolytica 15338b ........... ATCC A. punctata 11163 .................. ATCC A. salmonicida 14174 .............. ATCC CIP A. sobria 7433 ................ ATCC Alcaligenes faecalis 8750 ....... Bacillus cereus 11778 ....... ....... ATCC UVIC .... Escherichia coli ........... Plesiomonas shigelloides 9242 ...... NCMB Pseudomonas aeruginosa .......... ATCC ATCC P. fluorescens 13525 ............... P. putida 12633 .................... ATCC UVIC Salmonella typhimurium ........... UVIC Serratia marcescens ................ UVIC Staphylococcus aureus 1 .. S. aureus 2 ..................... UVIC ATCC Vibrio anguillarum 19264 .......... UVIC V. parahaemolyticus a ATCC, American Type Culture Collection; UVIC, University of Victoria culture collection; CIP, Collection Institut Pasteur; NCMB, National Collection of Marine Bacteria. b This species may not belong to the genus Aeromonas (3). in the presence of CO2 by using the methods of Owens (14). Culture supernatants were prepared as outlined above. Enzyme assay with human erythrocyte membranes. Methods for preparation of hemoglobin-free membranes and conditions for enzyme assay have been described previously (7, 12). Lipid breakdown products were separated by thin-layer chromatography in chloroform-methanol (20:1, vol/vol) or petroleum ether-ether-acetic acid (90:10:1, vol/vol/vol) on Silica Gel G (Merck). Authentic lipid standards (Supelco, Bellefonte, Pa.) were used in all separations. Enzyme assay with egg yolk emulsion. The incubation medium of Owens (14) was used with a 2.5% egg yolk (Oxoid) suspension pre-equilibrated with [7(n)-3H]cholesterol. Each milliliter of suspension contained 8,000 dpm of cholesterol at a concentration of 0.15,uM. Enzyme preparations (0.2 ml) were incubated at 37°C with 10 ml of this suspension for the time intervals indicated below. Cholesterol, cholesteryl ester, and fatty acids were separated by chromatography in petroleum ether-ether-acetic acid (90:10:1, vol/vol/ vol) on silica gel thin-layer plates after incubation. Areas of silica gel containing compounds of interest were scraped into small columns and eluted with chloroform-methanol (4:1, vol/vol). Eluates were taken to dryness in scintillation vials and counted in PCS (New England Nuclear Corp.) scintillation fluid. Ammonium sulfate precipitation and gel filtrations. Resuspended ammonium sulfate precipitates (a 50-fold concentration of the culture medium in 20 mM Tris-hydrochloride, pH 7.4) were prepared and concentrated as described earlier (12) or by the
133
method of Owens (14). Samples were applied to a Sepharose 6B (Pharmacia, Montreal, Quebec, Canada) column (2.5 by 60 cm) equilibrated in 20 mM Tris-hydrochloride (pH 7.4)-0.02% sodium azide. The column was eluted at 12 to 15 ml/h by using the same buffer, and 4-ml fractions were collected. Fractions were tested against either erythrocyte membranes or egg yolk emulsion, and lipid breakdown products were further characterized by thin-layer chromatography. Analytical methods. Protein was assayed by the method of Bradford (5) and water-soluble phosphate was measured by the Bartlett procedure (2).
RESULTS Lipid breakdown products observed when culture media from individual organisms were tested against erythrocyte membranes are given in Table 2. The results indicate that, although several bacteria in other families were capable of producing enzymes with phospholipase A or phospholipase C activity, only members of the family Vibrionaceae produced an enzyme capable of generating cholesteryl ester. All tested members of this family except Plesiomonas shigelloides showed production of acyltransferase as well as lysophospholipase, as fatty acids and water-soluble phosphate were invariably produced together with cholesteryl ester. None of TABLE 2. Breakdown of erythrocyte lipids by cellfree growth medium Breakdown productsb
Straina CE FA WSP DG
Aeromonas hydrophila Ah65 + A. hydrophila AhllO ...... + A. hydrophila ATCC 9071 .... + + A. punctata ... ....... + A. sobria .... .. + A. proteolytica. + A. salmonicida .. + Vibrio anguillarum ........ V. parahaemolyticus ...... Plesiomonas shigelloides .. Pseudomonas aeruginosa .... P. fluorescens .... P. putida. Escherichia coli ... Salmonella typhimurium ... Serratia marcescens. +
_
Alcaligenes faecalis. Staphylococcus aureus.
_
+ + +
+ + +
-
+ +
ND ND
-
+
+
_
+ +
+ +
-
+
+
-
+
+
+
_
+
_
-
ND
-
+
+
+
_ _
Bacillus cereus. Achromobacter xerosis 14780 a Culture supernatants were obtained as described in the text at 8, 12, 16, 24, 36, 48, and 72 h. b After incubation breakdown products were separated by thin-layer chromatography. CE, Cholesteryl ester; FA, fatty acid; WSP, water-soluble phosphate; DG, diglyceride; ND, not determined; +, appearance of a breakdown product at any of the designated times. +
134
MACINTYRE, TRUST, AND BUCKLEY
"
'
J. BACTERIOL.
J
c~~~~ .4
440~ ~ ~ ~
~
~
~
~
4
~
~
20
TIME
IN
*',
~
~
~~~~0
40
40
HOURS
FIG. 1. Bacterial growth and acyltransferase production. Organisms were grown and enzyme levels were determined in cell-free growth media as described in the text. (a) A. salmonicida. (b) A. proteolytica. (c) V. anguillarum. (d) V. parahaemolyticus. Symbols: 0, bacterial growth; V, enzyme activity. these organisms seemed to produce a phospho-
lipase C, as no diglyceride breakdown products were observed. Representative curves for growth and enzyme production are given in Fig. 1. In all but one case, maximal enzyme activity was obtained at late exponential to early stationary phase, and activity decreased thereafter. A. salmonicida medium, however, showed high levels of activity from exponential phase to the last time tested (72 h), indicating either continuous production through stationary phase or increased enzyme durability in culture media (Fig. 1). Because of the high activity of the enzyme obtained from this organism and because of its stability with time, A. salmonicida should be the organism of choice in future studies. Other representatives of the Enterobacteriaceae and Neisseriaceae were not tested, as none of them are known to produce a positive egg yolk reaction. However, cell-free culture medium from S. aureus, which is egg yolk positive and was reported by Owens (14) to contain acyltransferase, was also found to be negative against erythrocyte membranes (Table 2) unless concentrated by ammonium sulfate precipitation. In addition, Owens reported no acyltransferase activity in Aeromonas sp.; however, this author used cultures 24 h or more old and may have missed activity due to the drop in activity observed for all Aeromonas species except A. salmonicida. Furthermore, Owens reported diglyceride production by Aeromonas sp. (14), whereas we have not observed phospholipase C activity released from any strain at any time. Some of the differences between our obser-
vations in assays with erythrocyte membranes and those of Owens in assays with egg yolk emulsions might be accounted for by differences in enzyme substrate specificity. A quantitative comparison of positive bacterial species was therefore carried out by using tritiated cholesterol in egg yolk as substrate. Table 3 shows that ammonium sulfate precipitates from the members of the Vibrionaceae tested were more active in production of cholesteryl ester than was S. aureus but that S. aureus nevertheless contained significant acyltransferase activity. A. hydrophila, A. salmonicida, and Vibrio anguillarum fractions were all considerably higher than the S. aureus fractions when specific activities were compared at 24 h. Activity responsible for cholesteryl ester, fatty acid, and water-soluble phosphate production eluted in the void volume on chromatography of ammonium sulfate precipitates ofA. hydrophila supernatants on Sepharose 6B, which confirmed our earlier findings (12). Enzyme activity from A. salmonicida and V. parahaemolyticus was similarly excluded (Fig. 2). Interestingly, enzyme activity from S. aureus culture medium was not excluded from Sepharose 6B, indicating that this organism produces a complex of lower molecular weight.
DISCUSSION The results of this limited survey indicate that among the gram-negative rods production of an enzyme capable of producing cholesteryl ester from egg yolk emulsion or human erythrocyte membranes may be a characteristic of the Vibrionaceae alone. P. shigelloides was the only member of this family we tested which did not TABLE 3. Acyltransferase activities of ammonium sulfate precipitates from cell-free growth medium Sp act (24
Sp act (30
1.0 1.2
8.2 5.2
2.4
0.4
5.4
Total acStrain
tivity (,umol of CE per 24
h)a
Aeromonas hydrophila (Ah 65) A. salmonicida Vibrio parahaemolyticus Staphylococcus aureus
h) min) (tmol of (ttmol of CE per mg CE per mg of protein of protein per 24 h) per 30 min)
0.3 1.4 0.2 " Total activity was measured with 0.2-ml samples of resuspended ammonium sulfate precipitates in 10 ml of egg yolk emulsion as described in the text. Each 10 ml of assay suspension contained 1.5 ,umol of cholesterol. Ammonium sulfate precipitates were obtained from 13-h culture supernatants of A. hydrophila Ah65 and V. parahaemolyticus and from 24-h culture supernatants for A. salmonicida. The procedures of Owens (14) were followed for S. aureus. CE, Cholesteryl ester.
BACTERIAL CHOLESTEROL ACYLTRANSFERASE
VOL. 139, 1979
0M Z0 m N
*C' 2 0.0
OA
1.2
10
20
FRACTION
30
40
50
NUMBER
FIG. 2. Separation of ammonium sulfate precipitates on Sepharose 6B. See text for details of isolation
and chromatographic procedures. (a) A. salmonicida. (b) V. parahaemolyticus. (c) S. aureus. In each case the arrow indicates the position of maximum enzyme activity. The void volume was at tube 18 (72 ml).
demonstrate acyltransferase under our conditions. This may not be surprising, as this organism is known to lack a number of characteristics of the Vibrionaceae, including extracellular enzyme activities (16). Of all the organisms examined, S. aureus was the only member of another family which produced a phospholipid-cholesterol acyltransferase (Table 2). Column fractionation of ammonium sulfate precipitates indicated (Fig. 2) that this bacterium produces an enzyme of lower apparent molecular weight than that of the Vibrionaceae and, furthermore, that the enzyme released by the gram-positive coccus is much less active against erythrocyte membranes. Recent evidence not shown here indicates that members of the family Vibrionaceae release the enzyme associated with vesicles derived from the outer cell membrane. This may account for the lower molecular weight of the staphylococcal enzyme, as it may be produced as a much smaller unassociated entity. Furthermore, outer membrane vesicles contain lipopolysaccharide (15),
135
which may facilitate attachment of enzyme-containing vesicles to erythrocyte membranes (9), thereby increasing enzyme activity. The probable absence of such a binding component in the enzyme produced by S. aureus may account for its lower activity against erythrocyte membranes. Table 2 also shows that in the case of every positive organism, fatty acid, cholesteryl ester, and water-soluble phosphate were invariably produced together as breakdown products. This observation indicates the combined action of an acyltransferase and a phospholipase on erythrocyte lipids. As previously reported for A. hydrophila, both activities were produced simultaneously during growth and eluted together upon Sepharose column fractionation. This was also the case with S. aureus. These results suggest either that acyltransferase and phospholipase are associated in a single complex or that a single enzyme is responsible for both activities. The capacity of the family Vibrionaceae to produce cholesteryl ester from egg yolk emulsion and erythrocyte membranes may become useful as an additional biochemical identification tool for this family, and rapid techniques for its demonstratiorn are currently being evaluated. Clearly, the contrasting results of Owens (14) indicate that current microbial procedures to discriminate organisms with acyltransferase capability from those which produce phospholipase C are inadequate. The significance of the production of the acyltransferase in the pathogenicity of members of the family Vibrionaceae is also under investigation. ACKNOWLEDGMENT This work was supported in part by a grant from the National Research Council of Canada.
LITERATURE CITED 1. Barker, W. H., Jr., and E. J. Gangarosa. 1974. Food poisoning due to Vibrio parahaemolyticus. Annu. Rev. Med. 25:75-81. 2. Bartlett, G. R. 1959. Phosphorus analysis in column
chromatography. J. Biol. Chem. 234:466-471. 3. Baumann, P., and L. Baumann. 1977. Biology of the marine enterobacteria: genera Beneckea and Photobacterium. Annu. Rev. Microbiol. 31:39-61. 4. Bernheimer, A. W., and L. S. Avigad. 1974. Partial characterization of aerolysin, a lytic exotoxin from Aeromonas hydrophila. Infect. Immun. 9:1016-1021. 5. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. 6. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 7. Buckley, J. T., and J. N. Hawthorne. 1972. Erythrocyte membrane polyphosphoinositide metabolism and the
136
MAcINTYRE, TRUST, AND BUCKLEY
regulation of calcium binding. J. Biol. Chem. 247:72187223. 8. Bullock, G. L., D. A. Conroy, and S. F. Sniesko. 1971. Septicemic disae caused by motile aeromonads and pseudomonads, p. 348-349. In S. F. Snieszko and H. R. Axehrod (ed.), Diseases of fishes, book 2A. Bacterial diseases of fishes. THF Publications, Jersey City, N.J. 9. Davies, M., D. E. S. Stewart-Tull, and D. M. Jackson. 1978. The binding of lipopolysaccharide from Escherichia coli to mammalian cell membranes and its effect on liposomes. Biochim. Biophys. Acta 508:260-276. 10. Davies, W. A., J. G. Kane, and V. F. Garagusi. 1978. Human aeromonas infections: a review of the literature and a case report of endocarditis. Medicine 57:267-277. 11. Ljungh, A., B. Wretlind, and T. Wadatrom. 1978. Evidence for enterotoxin and two cytolytic toxins in human isolates of Aeromonas hydrophila, p. 947-960. In P. Rosenberg (ed.), Toxins: animal, plant and microbial. Pergammon Press, Oxford. 12. Maclntyre, S., and J. T. Buckley. 1978. Presence of
J. BACTERIOL. glycerophospholipid: cholesterol acyltransferase and phospholipase in culture supematant of Aeromonas hydrophila. J. Bacteriol. 117:402-407. 13. McCarthy, D. J. 1975. Fish furunculosis. J. Inst. Fish. Biol. 6:13-18. 14. Owens, J. J. 1974. The egg yolk reaction produced by several species of bacteria. J. Appl. Bacteriol. 37:137148. 15. Russel, R. R. B. 1976. Free endotoxin. Microbios Lett. 2: 125. 16. Schubert, R. H. W. 1974. Genus III. Plesiomonas Habs and Schubert 1962, 324, p. 348-349. In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 17. Wadstrom, T., A. Ijungh, and B. Wretlind. 1976. Enterotoxin, haemolysin and cytotoxic protein in Aeromonas hydrophila from human infections. Acta Pathol. Microbiol. Scand. Sect. B 84:112-114.
Vol. 139, No. 1
0021-9193/79/07-0132/05$02.00/0
Distribution of Glycerophospholipid-Cholesterol Acyltransferase in Selected Bacterial Species S. MAcINTYRE, T. J. TRUST, AND J. T. BUCKLEY* Department of Biochemistry and Microbiology, University of Victoria, Victoria, British Columbia, Canada V8W2Y2 Received for publication 12 February 1979
The distribution of glycerophospholipid-cholesterol acyltransferase in selected bacterial species was examined. Enzyme activity was demonstrated in cell-free growth media from all members of the family Vibrionaceae which were tested except Plesiomonas shigelloides. In each case, enzyme was produced in exponential to early stationary phase and was excluded from Sepharose 6B, indicating a complex of high molecular weight. In a limited survey of other families, Staphylococcus aureus was the only organism outside the Vibrionaceae which was shown to produce the enzyme. In this case, however, the enzyme exhibited much less activity against erythrocyte membranes and appeared to have a lower molecular weight. The reasons for these differences and the importance of the acyltransferase as a biochemical identification tool are discussed.
Although the biochemical modes of action of only a few bacterial toxins are known with certainty, indications are that a wide range of different attack mechanisms may be utilized. Members of the Vibrionaceae produce several extracellular proteins which serve as virulence factors (11). Until recently most interest has been focussed on choleragen, the first of these toxins to be reported. The toxins produced by Vibrio species other than Vibrio cholerae are less well characterized, even though species such as Vibrioparahaemolyticus are significant pathogens of humans (1). Toxins secreted by members of the genus Aeromonas have received even less attention. Yet Aeromonas salmonicida has long been recognized as a significant pathogen of fish, causing large mortalities due to furunculosis (13), and Aeromonas hydrophila, which causes septicemic infections in warm-water fishes (8), is being increasingly implicated in a variety of clinical infections of humans (10). The extracellular toxins produced by these aeromonads are obviously important virulence factors. One of the extracellular proteins produced by A. hydrophila, the hemolytic toxin aerolysin, has been partially purified (4). However, its mode of action is unknown. The properties of the other toxins have not been elucidated, although the properties of the enterotoxin may be similar to those of choleragen (17). Indeed, the number of other toxins has not been established. A. hydrophila has, however, recently been reported to produce a glycerophospholipid-choles-
terol acyltr&nsferase-lysophospholipase enzyme complex of high apparent molecular weight (12). Because of the possibility of a contribution of this activity to the pathogenicity of A. hydrophila and perhaps other aeromonads, or even all members of the Vibrionaceae, the extent of its distribution in selected bacterial species was determined. In this paper we describe its occurrence in various bacterial genera and species and compare the levels of enzyme activity in some species which produce it. MATERIALS AND METHODS Bacterial strains. The organisms used and their sources are listed in Table 1. The identity of all organisms was confirmed before use by means of staining characteristics and biochemical testing (6). Strains were maintained at 4°C on Trypticase soy agar which contained 3.5% sodium chloride for V. parahaemolyticus. Growth and sampling conditions. All organisms were grown in yeast extract-Casamino Acid broth (12) with 0.25% glucose. The culture medium for V. parahaemolyticus also contained 3.5% sodium chloride. Bacterial cultivation was normally at 30°C with gyratory shaking at 300 rpm. Cell growth was monitored by measuring culture turbidity in a Klett-Summerson photoelectric colorimeter, using the green filter. Samples were withdrawn at the time intervals indicated below, centrifuged at 14,000 x g for 20 min at 4°C, dialyzed against three changes of 20 mM Tris (pH 7.4) for 48 h, and frozen until required. In some cases dialysis and freezing were omitted and cholesteryl ester production was measured directly. Strains of Staphylococcus aureus were also grown
132
BACTERIAL CHOLESTEROL ACYLTRANSFERASE
VOL. 139, 1979 TABLE 1. Bacterial strains Strain
Source'
Achromobacter xerosis 14780 ...... ATCC Aeromonas hydrophila Ah65 ........ UVIC UVIC A. hydrophila AhllO ............... A. hydrophila 9071 ................. ATCC A. proteolytica 15338b ........... ATCC A. punctata 11163 .................. ATCC A. salmonicida 14174 .............. ATCC CIP A. sobria 7433 ................ ATCC Alcaligenes faecalis 8750 ....... Bacillus cereus 11778 ....... ....... ATCC UVIC .... Escherichia coli ........... Plesiomonas shigelloides 9242 ...... NCMB Pseudomonas aeruginosa .......... ATCC ATCC P. fluorescens 13525 ............... P. putida 12633 .................... ATCC UVIC Salmonella typhimurium ........... UVIC Serratia marcescens ................ UVIC Staphylococcus aureus 1 .. S. aureus 2 ..................... UVIC ATCC Vibrio anguillarum 19264 .......... UVIC V. parahaemolyticus a ATCC, American Type Culture Collection; UVIC, University of Victoria culture collection; CIP, Collection Institut Pasteur; NCMB, National Collection of Marine Bacteria. b This species may not belong to the genus Aeromonas (3). in the presence of CO2 by using the methods of Owens (14). Culture supernatants were prepared as outlined above. Enzyme assay with human erythrocyte membranes. Methods for preparation of hemoglobin-free membranes and conditions for enzyme assay have been described previously (7, 12). Lipid breakdown products were separated by thin-layer chromatography in chloroform-methanol (20:1, vol/vol) or petroleum ether-ether-acetic acid (90:10:1, vol/vol/vol) on Silica Gel G (Merck). Authentic lipid standards (Supelco, Bellefonte, Pa.) were used in all separations. Enzyme assay with egg yolk emulsion. The incubation medium of Owens (14) was used with a 2.5% egg yolk (Oxoid) suspension pre-equilibrated with [7(n)-3H]cholesterol. Each milliliter of suspension contained 8,000 dpm of cholesterol at a concentration of 0.15,uM. Enzyme preparations (0.2 ml) were incubated at 37°C with 10 ml of this suspension for the time intervals indicated below. Cholesterol, cholesteryl ester, and fatty acids were separated by chromatography in petroleum ether-ether-acetic acid (90:10:1, vol/vol/ vol) on silica gel thin-layer plates after incubation. Areas of silica gel containing compounds of interest were scraped into small columns and eluted with chloroform-methanol (4:1, vol/vol). Eluates were taken to dryness in scintillation vials and counted in PCS (New England Nuclear Corp.) scintillation fluid. Ammonium sulfate precipitation and gel filtrations. Resuspended ammonium sulfate precipitates (a 50-fold concentration of the culture medium in 20 mM Tris-hydrochloride, pH 7.4) were prepared and concentrated as described earlier (12) or by the
133
method of Owens (14). Samples were applied to a Sepharose 6B (Pharmacia, Montreal, Quebec, Canada) column (2.5 by 60 cm) equilibrated in 20 mM Tris-hydrochloride (pH 7.4)-0.02% sodium azide. The column was eluted at 12 to 15 ml/h by using the same buffer, and 4-ml fractions were collected. Fractions were tested against either erythrocyte membranes or egg yolk emulsion, and lipid breakdown products were further characterized by thin-layer chromatography. Analytical methods. Protein was assayed by the method of Bradford (5) and water-soluble phosphate was measured by the Bartlett procedure (2).
RESULTS Lipid breakdown products observed when culture media from individual organisms were tested against erythrocyte membranes are given in Table 2. The results indicate that, although several bacteria in other families were capable of producing enzymes with phospholipase A or phospholipase C activity, only members of the family Vibrionaceae produced an enzyme capable of generating cholesteryl ester. All tested members of this family except Plesiomonas shigelloides showed production of acyltransferase as well as lysophospholipase, as fatty acids and water-soluble phosphate were invariably produced together with cholesteryl ester. None of TABLE 2. Breakdown of erythrocyte lipids by cellfree growth medium Breakdown productsb
Straina CE FA WSP DG
Aeromonas hydrophila Ah65 + A. hydrophila AhllO ...... + A. hydrophila ATCC 9071 .... + + A. punctata ... ....... + A. sobria .... .. + A. proteolytica. + A. salmonicida .. + Vibrio anguillarum ........ V. parahaemolyticus ...... Plesiomonas shigelloides .. Pseudomonas aeruginosa .... P. fluorescens .... P. putida. Escherichia coli ... Salmonella typhimurium ... Serratia marcescens. +
_
Alcaligenes faecalis. Staphylococcus aureus.
_
+ + +
+ + +
-
+ +
ND ND
-
+
+
_
+ +
+ +
-
+
+
-
+
+
+
_
+
_
-
ND
-
+
+
+
_ _
Bacillus cereus. Achromobacter xerosis 14780 a Culture supernatants were obtained as described in the text at 8, 12, 16, 24, 36, 48, and 72 h. b After incubation breakdown products were separated by thin-layer chromatography. CE, Cholesteryl ester; FA, fatty acid; WSP, water-soluble phosphate; DG, diglyceride; ND, not determined; +, appearance of a breakdown product at any of the designated times. +
134
MACINTYRE, TRUST, AND BUCKLEY
"
'
J. BACTERIOL.
J
c~~~~ .4
440~ ~ ~ ~
~
~
~
~
4
~
~
20
TIME
IN
*',
~
~
~~~~0
40
40
HOURS
FIG. 1. Bacterial growth and acyltransferase production. Organisms were grown and enzyme levels were determined in cell-free growth media as described in the text. (a) A. salmonicida. (b) A. proteolytica. (c) V. anguillarum. (d) V. parahaemolyticus. Symbols: 0, bacterial growth; V, enzyme activity. these organisms seemed to produce a phospho-
lipase C, as no diglyceride breakdown products were observed. Representative curves for growth and enzyme production are given in Fig. 1. In all but one case, maximal enzyme activity was obtained at late exponential to early stationary phase, and activity decreased thereafter. A. salmonicida medium, however, showed high levels of activity from exponential phase to the last time tested (72 h), indicating either continuous production through stationary phase or increased enzyme durability in culture media (Fig. 1). Because of the high activity of the enzyme obtained from this organism and because of its stability with time, A. salmonicida should be the organism of choice in future studies. Other representatives of the Enterobacteriaceae and Neisseriaceae were not tested, as none of them are known to produce a positive egg yolk reaction. However, cell-free culture medium from S. aureus, which is egg yolk positive and was reported by Owens (14) to contain acyltransferase, was also found to be negative against erythrocyte membranes (Table 2) unless concentrated by ammonium sulfate precipitation. In addition, Owens reported no acyltransferase activity in Aeromonas sp.; however, this author used cultures 24 h or more old and may have missed activity due to the drop in activity observed for all Aeromonas species except A. salmonicida. Furthermore, Owens reported diglyceride production by Aeromonas sp. (14), whereas we have not observed phospholipase C activity released from any strain at any time. Some of the differences between our obser-
vations in assays with erythrocyte membranes and those of Owens in assays with egg yolk emulsions might be accounted for by differences in enzyme substrate specificity. A quantitative comparison of positive bacterial species was therefore carried out by using tritiated cholesterol in egg yolk as substrate. Table 3 shows that ammonium sulfate precipitates from the members of the Vibrionaceae tested were more active in production of cholesteryl ester than was S. aureus but that S. aureus nevertheless contained significant acyltransferase activity. A. hydrophila, A. salmonicida, and Vibrio anguillarum fractions were all considerably higher than the S. aureus fractions when specific activities were compared at 24 h. Activity responsible for cholesteryl ester, fatty acid, and water-soluble phosphate production eluted in the void volume on chromatography of ammonium sulfate precipitates ofA. hydrophila supernatants on Sepharose 6B, which confirmed our earlier findings (12). Enzyme activity from A. salmonicida and V. parahaemolyticus was similarly excluded (Fig. 2). Interestingly, enzyme activity from S. aureus culture medium was not excluded from Sepharose 6B, indicating that this organism produces a complex of lower molecular weight.
DISCUSSION The results of this limited survey indicate that among the gram-negative rods production of an enzyme capable of producing cholesteryl ester from egg yolk emulsion or human erythrocyte membranes may be a characteristic of the Vibrionaceae alone. P. shigelloides was the only member of this family we tested which did not TABLE 3. Acyltransferase activities of ammonium sulfate precipitates from cell-free growth medium Sp act (24
Sp act (30
1.0 1.2
8.2 5.2
2.4
0.4
5.4
Total acStrain
tivity (,umol of CE per 24
h)a
Aeromonas hydrophila (Ah 65) A. salmonicida Vibrio parahaemolyticus Staphylococcus aureus
h) min) (tmol of (ttmol of CE per mg CE per mg of protein of protein per 24 h) per 30 min)
0.3 1.4 0.2 " Total activity was measured with 0.2-ml samples of resuspended ammonium sulfate precipitates in 10 ml of egg yolk emulsion as described in the text. Each 10 ml of assay suspension contained 1.5 ,umol of cholesterol. Ammonium sulfate precipitates were obtained from 13-h culture supernatants of A. hydrophila Ah65 and V. parahaemolyticus and from 24-h culture supernatants for A. salmonicida. The procedures of Owens (14) were followed for S. aureus. CE, Cholesteryl ester.
BACTERIAL CHOLESTEROL ACYLTRANSFERASE
VOL. 139, 1979
0M Z0 m N
*C' 2 0.0
OA
1.2
10
20
FRACTION
30
40
50
NUMBER
FIG. 2. Separation of ammonium sulfate precipitates on Sepharose 6B. See text for details of isolation
and chromatographic procedures. (a) A. salmonicida. (b) V. parahaemolyticus. (c) S. aureus. In each case the arrow indicates the position of maximum enzyme activity. The void volume was at tube 18 (72 ml).
demonstrate acyltransferase under our conditions. This may not be surprising, as this organism is known to lack a number of characteristics of the Vibrionaceae, including extracellular enzyme activities (16). Of all the organisms examined, S. aureus was the only member of another family which produced a phospholipid-cholesterol acyltransferase (Table 2). Column fractionation of ammonium sulfate precipitates indicated (Fig. 2) that this bacterium produces an enzyme of lower apparent molecular weight than that of the Vibrionaceae and, furthermore, that the enzyme released by the gram-positive coccus is much less active against erythrocyte membranes. Recent evidence not shown here indicates that members of the family Vibrionaceae release the enzyme associated with vesicles derived from the outer cell membrane. This may account for the lower molecular weight of the staphylococcal enzyme, as it may be produced as a much smaller unassociated entity. Furthermore, outer membrane vesicles contain lipopolysaccharide (15),
135
which may facilitate attachment of enzyme-containing vesicles to erythrocyte membranes (9), thereby increasing enzyme activity. The probable absence of such a binding component in the enzyme produced by S. aureus may account for its lower activity against erythrocyte membranes. Table 2 also shows that in the case of every positive organism, fatty acid, cholesteryl ester, and water-soluble phosphate were invariably produced together as breakdown products. This observation indicates the combined action of an acyltransferase and a phospholipase on erythrocyte lipids. As previously reported for A. hydrophila, both activities were produced simultaneously during growth and eluted together upon Sepharose column fractionation. This was also the case with S. aureus. These results suggest either that acyltransferase and phospholipase are associated in a single complex or that a single enzyme is responsible for both activities. The capacity of the family Vibrionaceae to produce cholesteryl ester from egg yolk emulsion and erythrocyte membranes may become useful as an additional biochemical identification tool for this family, and rapid techniques for its demonstratiorn are currently being evaluated. Clearly, the contrasting results of Owens (14) indicate that current microbial procedures to discriminate organisms with acyltransferase capability from those which produce phospholipase C are inadequate. The significance of the production of the acyltransferase in the pathogenicity of members of the family Vibrionaceae is also under investigation. ACKNOWLEDGMENT This work was supported in part by a grant from the National Research Council of Canada.
LITERATURE CITED 1. Barker, W. H., Jr., and E. J. Gangarosa. 1974. Food poisoning due to Vibrio parahaemolyticus. Annu. Rev. Med. 25:75-81. 2. Bartlett, G. R. 1959. Phosphorus analysis in column
chromatography. J. Biol. Chem. 234:466-471. 3. Baumann, P., and L. Baumann. 1977. Biology of the marine enterobacteria: genera Beneckea and Photobacterium. Annu. Rev. Microbiol. 31:39-61. 4. Bernheimer, A. W., and L. S. Avigad. 1974. Partial characterization of aerolysin, a lytic exotoxin from Aeromonas hydrophila. Infect. Immun. 9:1016-1021. 5. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. 6. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 7. Buckley, J. T., and J. N. Hawthorne. 1972. Erythrocyte membrane polyphosphoinositide metabolism and the
136
MAcINTYRE, TRUST, AND BUCKLEY
regulation of calcium binding. J. Biol. Chem. 247:72187223. 8. Bullock, G. L., D. A. Conroy, and S. F. Sniesko. 1971. Septicemic disae caused by motile aeromonads and pseudomonads, p. 348-349. In S. F. Snieszko and H. R. Axehrod (ed.), Diseases of fishes, book 2A. Bacterial diseases of fishes. THF Publications, Jersey City, N.J. 9. Davies, M., D. E. S. Stewart-Tull, and D. M. Jackson. 1978. The binding of lipopolysaccharide from Escherichia coli to mammalian cell membranes and its effect on liposomes. Biochim. Biophys. Acta 508:260-276. 10. Davies, W. A., J. G. Kane, and V. F. Garagusi. 1978. Human aeromonas infections: a review of the literature and a case report of endocarditis. Medicine 57:267-277. 11. Ljungh, A., B. Wretlind, and T. Wadatrom. 1978. Evidence for enterotoxin and two cytolytic toxins in human isolates of Aeromonas hydrophila, p. 947-960. In P. Rosenberg (ed.), Toxins: animal, plant and microbial. Pergammon Press, Oxford. 12. Maclntyre, S., and J. T. Buckley. 1978. Presence of
J. BACTERIOL. glycerophospholipid: cholesterol acyltransferase and phospholipase in culture supematant of Aeromonas hydrophila. J. Bacteriol. 117:402-407. 13. McCarthy, D. J. 1975. Fish furunculosis. J. Inst. Fish. Biol. 6:13-18. 14. Owens, J. J. 1974. The egg yolk reaction produced by several species of bacteria. J. Appl. Bacteriol. 37:137148. 15. Russel, R. R. B. 1976. Free endotoxin. Microbios Lett. 2: 125. 16. Schubert, R. H. W. 1974. Genus III. Plesiomonas Habs and Schubert 1962, 324, p. 348-349. In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 17. Wadstrom, T., A. Ijungh, and B. Wretlind. 1976. Enterotoxin, haemolysin and cytotoxic protein in Aeromonas hydrophila from human infections. Acta Pathol. Microbiol. Scand. Sect. B 84:112-114.
