Staphylococcus aureus

JOURNAL OF BACTERIOLOGY, Nov. 1973, p. 571-576 Copyright 0 1973 American Society for Microbiology Vol. 116, No. 2 Printed in U.S.A. Protein and Fatty Acid Composition of Mesosomal Vesicles and Plasma Membranes of Staphylococcus aureus T. S. THEODORE AND C. PANOS Laboratory of Microbiology, National Institute ofallergy and Infectious Diseases, Bethesda, Maryland 20014, and Department of Microbiology, Jefferson Medical College, Philadelphia, Pennsylvania 19107 Received for publication 25 July 1973 Qualitatively, the protein and fatty acid composition of purified mesosomal vesicles and the plasma membrane isolated from Staphylococcus aureus ATCC 6538P are identical, the major difference between these two cellular components being only quantitative in nature. Mesosomal vesicles and plasma membranes, when subjected to acidic or neutral disk gel electrophoresis, exhibited more than 22 bands of protein. With urea-acetic acid gels, the plasma membrane had a higher concentration of "slower-migrating proteins" whereas "faster-migrating proteins" predominated in the mesosomal vesicles. With neutral disk gel electrophoresis, mesosomal vesicles exhibited one prominent protein band with an approximate molecular weight of 35,000 and which was four times greater than that found in the corresponding region on gels of the plasma membrane. Finally, fatty acid analyses by capillary column gas chromatography showed that although the fatty acid composition is the same, the fatty acid content in mesosomal vesicles is 48% greater than that of the plasma membrane. The dominant fatty acids in both of these cellular components are the iso and anteiso branched methyl C15, C17, and C19 fatty acids and comprise at least 85% of the total fatty acids extracted. These results show that distinct chemical differences exist between the mesosomal vesicles and the plasma membrane of Staphylococcus aureus. At present, a great deal of information is available Qn the chemical composition of membranes isolated from gram-positive bacteria (3, 16-18, 22, 23). Most of these analyses refer to the composition of the total membrane fraction, which would include both the plasma and mesosomal membranes. However, due to the lack of suitable fractionation and isolation techniques, only a few studies have been made on the chemical composition of isolated mesosomes and membranes. Most of these have utilized bacilli and pertain to their gross chemical composition (1, 2, 11, 13, 15). Recently, we fractionated the membrane system of Staphylococcus aureus (12, 20). The mesosomes were purified from the protoplast supernatant medium by differential and gradient centrifugation and shown to be of the vesicular type and homogeneous in size, ranging from 35 to 50 nm in diameter. The plasma membranes, produced by osmotic shock of the protoplasts and similarly concentration on gradients, showed the typical profile widely associated with the unit membrane structure. In overall chemical composition, the mesosomal vesicles do differ from the plasma membranes (20). On a dry weight basis, the plasma membrane fraction is composed of 56% protein, 25% lipid, and 14% ribonucleic acid (RNA, detected as orcinol-positive material), while the mesosomal vesicles contain only 41% protein, 34% lipid, and 8% RNA (orcinol-positive material). The ratio of protein to lipid in the plasma membranes is 2.24/1, whereas in the mesosomal vesicles it is 1.21/1. Since the present status of the mesosome with respect to its functional role in the cell and its relationship to the plasma membrane is still a matter of speculation, it was of interest to define more precisely the chemical differences (if any) between the two types of membrane. In the present study, mesosomal vesicles and plasma membranes of S. aureus ATCC 6538P were isolated and characterized with respect to 571

572 THEODORE AND PANOS J. BACTERIOL. their protein composition and fatty acid content. MATERIALS AND METHODS Preparation of mesosomal vesicles and plasma membranes. Mesosomal vesicles and plasma membranes were isolated from S. aureus ATCC 6538P after growth in Difco AOAC synthetic medium. After late log-phase growth, cells were washed once, suspended in 0.01 M phosphate buffer (ph 6.5), containing 20% NaCl and 0.02 M MgSO4, and treated with lytic enzyme for 2 h at 37 C. The enzyme used to degrade staphylococcal cell walls was isolated by (NH4)2SO4 fractionation of culture fluids of S. aureus strain LS (obtained from A. Scott, University of Dundee, Scotland) after growth on tryptose phosphate broth (Difco). Procedures for the preparation of LS lytic enzyme, protoplasting conditions, and purification and isolation of mesosomal vesicles and plasma membranes by differential and sucrose density gradient centrifugation are detailed elsewhere (20). Protein content of these preparations was estimated by the method of Lowry et al. (6) with bovine serum albumin as the standard. Polyacrylamide gel electrophoresis. Mesosomal vesicles and plasma membranes collected from sucrose gradients were washed once, suspended in 0.01 M phosphate buffer (ph 7.0), and analyzed for their protein composition by use of two gel systems commonly employed for this purpose with membranous materials. For the urea-acetic acid gel system, the procedure of Theodore et al. (19) was followed. Mesosomal vesicles and plasma membranes (250 jug of protein) were solubilized in phenol-acetic acid (2: 1, wt/vol) containing 2M urea and 1% mercaptoethanol for 2 h at 37 C. Any insoluble material was removed by centrifugation at 4,500 x g for 20 min at room temperature (24 C). Electrophoresis was carried out in glass columns containing 1.0 ml of separating gel (7.5% acrylamide, 8 M urea [ph 4.5]) and 0.2 ml of stacking gel (2.5% acrylamide, 8 M urea [ph 6.7]). The lower electrode served as the cathode, and both upper and lower reservoirs contained 0.07 M,8-alanine buffer adjusted to ph 4.5 with acetic acid. Pyronin Y was used as the tracking dye, and the samples were subjected to electrophoresis at a constant current of 4 ma per tube for 75 min at room temperature. Gels were stained for 1 h with 1% amido black and destained by simple diffusion in 7.5% acetic acid. The preparation of sodium dodecyl sulfate (SDS) gels followed the method described by Weber and Osbom (21). Mesosomal vesicles and plasma membranes (100 Mg of protein) were dissolved in 1% SDS and 1% mercaptoethanol for 2 h at 37 C and subjected to electrophoresis on 7.5% acrylamide gels containing 0.1% SDS. The lower electrode served as the anode, and both upper and lower reservoirs contained 0.1 M phosphate buffer (ph 7.0) and 0.1% SDS. Bromophenol blue in 40% sucrose served as the marker dye. Samples were run at a constant current of 5 ma per tube for 5 to 5.5 h at 24 C. Gels were prefixed overnight in methanol-acetic acid-water (5:1:4), stained 30 min in 0.25% Coomassie blue made up in methanol-acetic acid-water (4.5:1:4.5) and destained by simple diffusion in methanol-acetic acid-water (5:7.5:87.5). All gels were scanned with a Gilford gel scanner at a speed of 2 cm/min. Lipid from the mesosomal vesicles and plasma membranes was extracted with chloroform-methanol (2:1) solvent. The lipid-depleted preparations were washed several times in phosphate buffer and prepared from electrophoresis as described above. Fatty acid extraction and analyses. Mesosomal vesicles and plasma membranes collected by differential and sucrose gradient centrifugation were washed once in phosphate buffer, twice in distilled water, and lyophilized. The alkaline hydrolysis-fatty acid extraction methods of Hofmann, Henis, and Panos (4), modified by the omission of bicarbonate washings, were used throughout. Methylation of the free fatty acids was performed with perchloric acid in methanol at 55 C for 15 min. Hydrogenations of methylated fatty acids for confirmatory identification were performed as described elsewhere (9). The relative retention times of all fatty acids was confirmed with available standards and from construction of a James plot (5) of the logarithms of the relative retention times versus the number of carbon atoms of the major acids detected. Mixtures of fatty acid methyl esters were resolved by capillary column gas chromatography, using a polar column (150 ft by 0.01 inch) coated with Carbowax K20-M plus V-930 (99/1) as has been described (8, 10). The column temperature was 186 C, and the detector was at 240 C. The column was obtained from the Perkin-Elmer Corporation, Norwalk, Conn. The proportions of fatty acids were determined by computation of the area from the peak height times the width at half-height with an eye ocular. The average deviation of the weight percent for all peaks in two randomly selected runs was -0.40. RESULTS AND DISCUSSION Protein patterns of mesosomal vesicles and the plasma membrane. The protein patterns derived from mesosomal vesicles and plasma membranes extracted with acidified phenol and run on urea-acetic acid gels are shown in Fig. 1. Qualitatively, both gel patterns were similar and each could be resolved into approximately 26 bands of protein (Fig. 2). The principal difference between the two membrane fractions was one of a quantitative nature. On the basis of stain intensity, which is related to protein concentration, plasma membranes consistently had a higher concentration of "slower-migrating proteins" (upper third of separating gel, Fig. 1), whereas mesosomal vesicles, which stained poorly in that region, had a higher concentration of "faster-migrating proteins" (mid-portion of separating gel, Fig. 1). Although the urea-acetic acid gel system was quite reproducible and gave clear patterns, it was never possible to solubilize more than 75%

::.,... :! X,. s.4 r.... t.*..,... q8!. i : i VOL. 116,1973 CHEMICAL COMPOSITION OF MESOSOMES AND MEMBRANES 573 I I: _ XF..: 3 N_...}.4.rsag.=......}r. :F j.. A B > ;r.:: 4... 4<.:,*,. FIG. 1. Urea-acetic acid polyacrylamide gel patterns of Staphylococcus aureus mesosomal vesicles FIG. 2. Densitometric tracings of gel patterns (A) and plasma membranes (B). shown in Fig. 1. of the mesosomal vesicles or plasma membranes with acidified phenol. Also, a significant portion of the proteins never entered the gel as evidenced by the intensity of staining at the top of both the stacking and separating gels (Fig. 1). Since the possibility existed that the differences..i_ i9& iz noted between the two membrane fractions might be due to incomplete solubilization and.w...l.e.. the inability of some proteins to migrate into T0';!Sn'Vw the separating gel, we tested other solubilizing agents and gel systems commonly used for the analysis of membrane proteins. Of all the reagents tested (Triton X-100, Nonidet P-40, sodium deoxycholate, and SDS), only the SDS completely solubilized both membrane fractions. The results of SDS solubilization and electrophoresis on SDS gels are shown in Fig. 3. As shown with urea-acetic acid gels, only a concentration difference was apparent between the two membrane fractions and each could be *, / resolved into 22 protein bands (Fig. 4). The results in Fig. 3 also show that the SDS gel FIG. 3. SDS polyacrylamide gel patterns of Staphylococcus aureus mesosomal vesicles (A), plasma patterns of lipid-extracted mesosomal vesicles and plasma membranes were identical to those membranes (B), lipid-depleted mesosomal vesicles not so extracted. (Urea-acetic acid gels of similarly extracted mesosomal and plasma mem- arrow shows the region of major (C), and lipid-depleted plasma membranes (D). The difference.

574 THEODORE AND PANOS J. BACTERIOL. Top of separating gel FIG. 4. Densitometric tracings of SDS gel patterns of mesosomal vesicles (A) and plasma membranes (B) shown in Fig. 3. The arrow shows the region of major difference. brane fractions were not run due to their insolubility in the phenol-acetic acid mixture after chloroform-methanol extraction.) One distinct feature of the SDS gel pattern of mesosomal vesicles is the single prominent protein band present in the mid-region of the gel (see arrows, Fig. 3 and 4). This protein, which is also present in plasma membranes but is much less pronounced, has an approximate molecular weight of 35,000 as determined by the method of Weber and Osborn (21), using bovine serum albumin, pepsin, trypsin and cytochrome c as protein standards. Based on the densitometric tracing data (Fig. 4), the concentration of this protein is four times greater in the mesosomal vesicles than in the plasma membranes. In urea-acetic acid gels a similar pattern was seen; however, not one, but several more concentrated faster-migrating proteins were evident in the gels of mesosomal vesicles (Fig. 1). This difference observed between the two gel systems may be due to either the effect of the particular solubilizing agent used or the large amounts of protein needed to produce stainable bands in the urea-acetic gels (250 jig of protein compared to 100,ug for SDS gels). Attempts to electrophorese 250 gg of membrane protein on SDS gels resulted in overstaining and obscurity of the gel patterns. Regardless of which solvent-gel system was used to analyze mesosomal vesicles and plasma membrane proteins, the results were consistent and demonstrated the following: (i) heterogeneity in protein composition of mesosomal vesicles and plasma membranes, (ii) considerable overall similarity in their gel patterns, and (iii) distinct quantitative differences between these two components, especially in the lower region of the gels. In a similar study using mesosomes and membranes prepared from Bacillus licheniformis (13), Reaveley also showed overall similarities in gel patterns with significant quantitative differences in the fastermigrating proteins. Apart from these quantitative differences, Owen and Freer have also reported qualitative differences between mesosomes and membranes of Micrococcus lysodeikticus (7); however, it should be noted that these differences may be due to a variation in the amount of each membrane fraction subjected to electrophoresis on polyacrylamide gels. In accordance with these studies, the question arises whether the quantitative differences noted in protein patterns between the mesosomal vesicles and the plasma membranes is a reflection of their purity. In a previous study, we showed that the mesosomal vesicles were quite pure and free of contaminating ribosomes and membranes (12, 20). Similarly prepared plasma membrane fractions did contain free and attached ribosomes and a few mesosomal vesicles; however, when plasma membranes were prepared in the absence of Mg2+ and treated with ribonuclease, only a few visible or membraneassociated ribosomes or vesicles were present (12). Plasma membranes prepared in this manner and run on polyacrylamide gels gave results identical to those shown in Fig. 1 and 3. Also, our results with respect to the faster-migrating proteins are the reverse of what would be expected if differences between the protein

VOL. 116, 1973 CHEMICAL COMPOSITION OF MESOSOMES AND MEMBRANES patterns of mesosomal vesicles and the plasma membrane were due solely to the contamination of these components with ribosomes and vesicles. Fatty acid analyses of mesosomal vesicles and the plasma membrane. As expected, the greater fatty acid content of the mesosomal vesicles (by 48%) over that of the plasma membrane from this organism (Table 1) was similar to that reported by us earlier (20) for the gross lipid content from these same components. However, although this difference remained reproducible, the fatty acid composition from these two components was almost identical. Others had reported a total of 64 fatty acids in S. aureus as determined by gas chromatography (23). However, since our study was concerned primarily with detecting major changes, only those fatty acids with a concentration greater than 0.2% are tabulated (Table 1). These fatty acid results are reminiscent of a similar finding with B. licheniformis (C. Panos, D. Reaveley and J. J. Rogers, unpublished data) in which its mesosomal fatty acid content was also found to be greater (by more than twofold) than that of the bacillary membrane. Here again, however, the fatty acid composition of these two cellular components was the same. Therefore, these collective results indicate that although the fatty acid content of mesosomal vesicles is significantly greater than that of the bacterial membrane, no major differences are apparent in the distribution of these long-chain fatty acids between these two subcellular components. As known for the intact staphylococci (23), the iso and anteiso branched methyl C15, C17, and C19 fatty acids were also dominant in the mesosomal vesicles and the plasma membrane from this staphylococcus, comprising at least 85% of their total fatty acids, with anteiso C15 being the predominating fatty acid. Similarly, the anteiso isomers predominated over their homologous iso fatty acids in both of these cellular preparations. Finally, although others had reported a high content of arachidic (eicosanoic) acid in an intact S. aureus when grown aerobically or anaerobically (23), this fatty acid comprised less than 4% of those from the mesosomal vesicles and membrane of the coccus used in the present study. This quantitative difference, however, may be due to a difference in the growth phase from which this bacterium was harvested in the respective studies. Because mesosomal vesicles are a part of the bacterial membrane system and presumably TABLE 1. Major fatty acid composition of mesosomal vesicles and plasma membranes from Staphylococcus aureus Total fatty acids (%) Fatty acida Mesosomal Plasma vesicles membranes C14 i 0.73 0.66 C14 n 0.30 0.20 C15 i 10.65 8.88 C15 a 39.57 41.80 C16 i 1.74 1.73 C16n 1.24 1.28 C17 i 8.10 7.88 C17 a 17.73 18.33 C18 i 1.10 1.06 C18 n 4.77 4.42 Oleic 0.44 0.64 C19 i 3.68 3.58 C19 a 5.55 5.70 C19 n 0.55 0.49 C20 n 3.81 3.32 Total fatty acids" 27.5 18.5 575 a Fatty acid composition as determined by capillary column gas chromatography. Only fatty acids with a concentration greater than 0.2% are tabulated. Fatty acid methyl esters are designated as the number of carbon atoms, with (a) for anteiso and (i) for iso branched methyl fatty acids and (n) for normal straight chain fatty acids. bindicates the total fatty acids obtained from each source and expressed as percent of the dry weight of the starting cellular component. arise by an invagination of the plasma membrane, it is not surprising that only quantitative (rather than qualitative) differences in protein and fatty acid composition exist between the two membrane fractions. These results are compatible not only with our earlier studies relating to their gross chemical composition (20), but also to the numerous specific functions that have been proposed for the mesosome (17). Qualitatively, the same biochemical activities associated with the mesosomal membrane are also found in the plasma membrane (14), and, except for one instance relating to a purportedly increased level of lytic activity by mesosomes of M. lysodeikticus (7), most of these activities are present to a much lesser extent in the mesosome. In this study, we show not only increased levels of fatty acids in mesosomal vesicles, but

576 THEODORE AND PANOS J. BACTERIOL. also distinct quantitative differences in proteins of certain sizes. The most pronounced difference is seen on SDS gels, in which preparations from mesosomal vesicles contain a protein (approximately 35,000 molecular weight) whose concentration is four times greater than that found in the corresponding membrane fraction. These findings suggest that mesosomal vesicles may be active sites of fatty acid and lipid biosynthesis. However, several studies with bacilli have shown no preferential incorporation of labeled acetate into the mesosome (14). The relation of these specific chemical differences to the functional role of the mesosome is not yet clear. However, the fact that only quantitative differences have been demonstrated between mesosomal vesicles and plasma membranes should not negate the importance of this unique organelle. Present studies in our laboratory on the lipid composition have shown that mesosomal vesicles contain four times more glycerol than that found in the corresponding plasma membrane fraction, and that a 1:1 ratio of glycerol to phosphate exists. Whether this "glycerol-phosphate" is associated with a neutral phospholipid fraction or the membrane teichoic acid component is presently under investigation (T. S. Theodore and E. Huff, unpublished results). ACKNOWLEDGMENTS This investigation was supported, in part, by Public Health Service research grant AI 11170-01 to one of us (C. Panos) from the National Institute of Allergy and Infectious Diseases. LITERATURE CITED 1. Ellar, D. J., T. D. Thomas, and J. A. Postgate. 1971. Properties of mesosomal membranes isolated from Micrococcus lysodeikticus and Bacillus megaterium. Biochem. J. 122:44p-45p. 2. Ghosh, B. K., and R. G. E. Murray. 1969. Fractionation and characterization of the plasma and mesosome membrane of Listeria monocytogenes. J. Bacteriol. 96:426-440. 3. Grula, E. A., and C. F. Savoy. 1971. A detergent-polyacrylamide gel system for electrophoretic resolution of membrane and wall proteins. Biochem. Biophys. Res. Commun. 43:325-332. 4. Hofmann, K., D. B. Henis, and C. Panos. 1957. Fatty acid interconversions in lactobacilli. J. Biol. Chem. 228:349-355. 5. James, A. T. 1960. Qualitative and quantitative determination of fatty acids by gas-liquid chromatography, p. 1-59. In D. Glick (ed.), Methods of biochemical analysis, vol. 8. Interscience Publishers Inc., New York. 6. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurements with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 7. Owen, P., and J. H. Freer. 1972. Isolation and properties of mesosomal membrane fractions from Micrococcus lysodeikticus. Biochem. J. 129:907-917. 8. Panos, C. 1965. Separation and identification of positional isomers of bacterial long chain monoethenoid fatty acids by Golay column chromatography. J. Gas Chromatogr. 3:278-281. 9. Panos, C., M. Cohen, and G. Fagan. 1966. Lipid alterations after cell wall inhibition. Fatty acid content of Streptococcus pyogenes and derived L-form. Biochemistry 5:1461-1468. 10. Panos, C., and C. V. Henrickson. 1969. Fatty acid interconversions in Mycoplasma sp. KHS. Biochemistry 8:652-658. 11. Patch, C. T., and 0. E. Landman. 1971. Comparison of the biochemistry and rates of synthesis of mesosomal and peripheral membranes in Bacillus subtilis. J. Bacteriol. 107:345-357. 12. Popkin, T. J., T. S. Theodore, and R. M. Cole. 1971. Electron microscopy during the release and purification of mesosomal vesicles and protoplast membranes from Staphylococcus aureus. J. Bacteriol. 107:907-917. 13. Reaveley, D. A. 1968. The isolation and characterization of cytoplasmic membranes and mesosomes of Bacillus licheniformis 6346. Biochem. Biophys. Res. Commun. 30:649-655. 14. Reusch, V. M., and M. M. Burger. 1973. The bacterial mesosome. Biochim. Biophys. Acta 300:79-104. 15. Rogers, H. J., D. A. Reaveley, and I. D. J. Burdett. 1967. The membrane systems of Bacillus licheniformis, p. 303-313. In H. Peeters (ed.), Protides of biological fluids, vol. 15. Elsevier Publishing Co., Amsterdam. 16. Salton, M. R. J. 1967. Structure and composition of bacterial membranes, p. 279-288. In H. Peeters (ed.), Protides of biological fluids, vol. 15. Elsevier Publishing Co., Amsterdam. 17. Salton, M. R. J. 1967. Structure and function of bacterial cell membranes. Annu. Rev. Microbiol. 21:417-442. 18. Salton, M. R. J., M. D. Schmitt, and P. E. Trefts. 1967. Fractionation of isolated bacterial membranes. Biochem. Biophys. Res. Commun. 29:728-733. 19. Theodore, T. S., J. R. King, J. G. Tully, and R. M. Cole. 1970. Polyacrylamide gel patterns of microorganisms, p. 122-138. In C. J. Corum (ed.), Developments in industrial microbiology, vol. 11. Garamond/Pridemark Press, Baltimore. 20. Theodore, T. S., T. J. Popkin, and R. M. Cole. 1971. The separation and isolation of plasma membranes and mesosomal vesicles from Staphylococcus aureus. Prep. Biochem. 1:233-248. 21. Weber, K., and M. Osborn. 1969. The reliability of molecular weight determinations by dodecyl sulfatepolyacrylamide gel electrophoresis. J. Biol. Chem. 244:4406-4412. 22. White, D. C., and F. E. Frerman. 1967. Extraction, characterization, and cellular localization of the lipids of Staphylococcus aureus. J. Bacteriol. 94:1854-1867. 23. White, D. C., and F. E. Frerman. 1968. Fatty acid composition of the complex lipids of Staphylococcus aureus during the formation of the membrane-bound electron transport system. J. Bacteriol. 95:2198-2209.