JOURNAL OF BACTERIOLOGY, Sept. 1975, p. 972-977 Copyright 0 1975 American Society for Microbiology
Vol. 123, No. 3 Printed in U.S.A.
Properties of a Membrane-Bound Cardiolipin Synthetase from Lactobacillus plantarum MARY F. BURRIIT' AND THOMAS 0. HENDERSON* Department of Biological Chemistry, University of Illinois at the Medical Center, Chicago, Illinois 60612 Received for publication 16 April 1975
Cardiolipin (CL) synthetase of Lactobacillus plantarum 17-5 catalyzed the stoichiometric conversion of 2 mol of phosphatidylglycerol to 1 mol of CL. The enzyme activity was linear with time for 30 min at 37 C and with protein concentration between 20 and 200 ytg of protein per ml. The enzyme was membrane associated, had a pH optimum of 5.1 in phosphate buffer, and was not stimulated by Mg2+, and the activity was not affected by the addition of ethylenediaminetetraacetic acid, cytidine diphosphate diglyceride, or cytidine triphosphate. The reaction was inhibited about 95% by Triton X-100 (0.5% final concentration) and by CL, the end product of the reaction. The activity of this enzyme was studied as a function of growth. The CL synthetase specific activity was highest during the early and midexponential growth phases, as was the cellular content of CL. The results demonstrate a correlation between enzymespecific activity and lipid content of the cells.
Kiyasu et al. (7) postulated in 1963 that cardiolipin (CL) was synthesized by the reaction: PG + CDP diglyceride _ CL + CMP However, more recent studies have indicated that in Micrococcus lysodeikticus (2), Staphylococcus aureus (15), and Escherichia coli (5, 19), one molecule of CL is synthesized from two molecules of phosphatidylglycerol (PG) by an enzyme found in a particulate fraction by the following reaction: 2 PG - CL + glycerol Abbreviations: CDP, cytidine diphosphate; CMP, cytidine monophosphate. The CL synthetase from S. aureus has been well characterized (15) and found to have a pH optimum of 4.4, no metal requirement, and a high affinity for substrate. When ["C ]CDP-diglyceride (CDP-DG) was incubated with CL synthetase at pH 4.4 in the presence of [32P ]PG, the CL isolated from the reaction mixture was labeled only with 32P. In all cases studied, the increase in CL concentration was balanced by a loss of PG. In a companion report (1) we present the results of studies on polar lipid metabolism in Lactobacillus plantarum under conditions of 'Present address: Department Clinic, Rochester, Minn. 55901.
inhibited protein synthesis. Despite the earlier interest in the lipid metabolism of L. plantarum and related lactobacilli, there have been no reports on the enzymes of polar lipid metabolism under any conditions (except for those studies by us [T. 0. Henderson and J. J. McNeill, Bacteriol. Proc., p. 34, 1967] and others  on the fatty acid synthetase of L. plantarum). In the present report, we describe our studies on the changes in the polar lipid composition of L. plantarum as a function of culture age, as well as some observations on the membrane-associated enzyme that catalyzes the conversion of PG to CL. (Portions of this work were presented at the 74th Annual Meeting of the American Society for Microbiology, 12-17 May 1974. This work is based on a dissertation submitted by M. F. B. to the Graduate College of the University of Illinois at the Medical Center, Chicago, in partial fulfillment of the requirements for the Ph.D. degree.)
of Immunology, The Mayo 972
MATERIALS AND METHODS Growth of cells. L. plantarum 17-5 (ATCC 8014) was grown in a complex medium of the following composition: 1 liter of medium, pH 6.8, contained 5 g of Difco yeast extract; 10 g of Difco tryptone; 10 g of glucose; 0.3 g of sodium acetate; and 5 ml of a salt solution containing 10 mg of MgSO, * 7H,O, 0.28 g of MnSO4 H,O, 1 g of KCl, 0.27 g of FeSO4, and 23 ml of 85% HSP04 per 500 ml. One liter of medium was -
VOL. 123, 1975
CARDIOLIPIN SYNTHETASE FROM L. PLANTARUM
inoculated with 1.0 ml of a 1:100 dilution of a 24-h broth culture. The cells were grown for 17 h at 33 C and harvested by centrifugation at 4 C. Preparation of enzyme. The bacteria were harvested from mid exponential-phase cultures (absorbance at 440 nm of a 1:4 dilution = 0.5) by centrifugation at 16,000 x g for 15 min. The cells were washed twice with 0.9% NaCl and suspended in 50 mM phosphate buffer, pH 6.0, containing 10 mM MgCl,, in a ratio of 1 g of wet cells to 4 ml of buffer. The cells were broken with a French pressure cell (Aminco) (three passes, 15,000 to 20,000 lb/in2) and centrifuged at 100,000 x g for 2 h in a Spinco model L ultracentrifuge, and the pellet was suspended in 12 ml of 50 mM phosphate buffer, pH 6.0, containing 10 mM MgCl, unless otherwise indicated. Protein was measured by the method of Lowry et al. (8), using bovine serum albumin as the standard. Preparation of radioactive PG. L. plantarum was grown for 20 h in the presence of [I4C,Jsodium acetate (0.2 MCi/ml; specific activity, 100 mCi/nmol), ['4CJoleic acid (0.1 uCi/ml; specific activity, 1.4 mCi/nmol), or ['H,,,0]oleic acid (2.5 MCi/ml; specific activity, 35.3 mCi/nmol), and the cells were harvested and washed as described above. All isotopes were purchased from New England Nuclear Corp. Lipids were extracted from the whole cells as described by Thorne and Kodicek (18), and the lipid extract was taken to dryness in vacuo and then dissolved in 2 ml of benzene. The lipids were streaked on analytical Silica Gel G thin-layer plates (20 by 20 cm; E. Merck), and the plates were developed with chloroform-methanolwater (65:25:4) (21). The radioactive lipids were visualized by autoradiography (1) and the appropriate phospholipid-containing area (consisting of CL and PG) was scraped into a flask. Lipids were eluted from the silica gel by standing overnight in 1% ammonium formate in methanol, and the silica gel was removed by filtration. The lipid solution was evaporated to dryness, dissolved in chloroform, and placed on a small Unisil (Clarkson Chemical Co.) column that had been packed in chloroform. The column was eluted with 50 ml of chloroform followed by 50 ml of methanol; this procedure effectively removed the ammonium formate from the lipids. The methanol fraction (containing PG and CL) was collected and the solvent was removed in vacuo. The lipids were again dissolved in a small volume of benzene. The phospholipids (PG and CL) in benzene were then applied to analytical Silica Gel G thin-layer plates (20 by 20 cm) and developed with chloroformmethanol-acetic acid-water (65:25:8:4) (17). The radioactive lipids were again visualized by autoradiography. The area containing the PG was scraped and eluted as described above. The PG thus isolated was at least 95% pure as deterrnined by thin-layer chromatography in a variety of one- and two-dimensional thin-layer chromatographic solvent systems (described below). Phosphorus determinations were carried out as described by Kirkpatrick and Bishop (6). Enzyme assay. CL synthetase activity was assayed in 0.5 ml of reaction mixtures (except where noted) containing: 35 mM phosphate buffer, pH 6.0; 7 mM
MgCl2; 100 to 200 ug of protein; and sonicated Triton X-100 dispersions of PG (final concentration of Triton X-100, 0.2%). The reaction mixtures were incubated at 37 C for 30 min unless otherwise indicated. The reactions were terminated by the addition of 3.75 ml of 2:1 methanol-chloroform, along with nonradioactive carrier lipid (L. plantarum total lipid, 400 Mg) and 0.5 ml of buffer. The mixtures were allowed to stand for 1 h on ice, and then 1.25 ml of water and 1.25 ml of chloroform were added to each tube. The tubes were refrigerated overnight and then centrifuged to separate the layers. The water layer was aspirated, the lower chloroform layer was evaporated under a stream of nitrogen, and the lipids were dissolved in 0.3 ml of benzene. Samples (150 ul) were spotted on commercially prepared Silica Gel G analytical plates and run in two dimensions. The first-dimension solvent system was chloroform-methanol-water (65:25:4) (21) and the second-dimension system was chloroformmethanol-acetic acid-water (65:25:8:4) (17). Lipids were located by exposure to iodine vapor and quantitated by scraping the appropriate areas of the silica gel into scintillation vials and counting the samples as described elsewhere (4). In addition, two other onedimensional chromatographic systems were used to identify the reaction products: chloroform-methanolammonia (65:35:5) (16) and chloroform-acetonemethanol-acetic acid-water (5:2:1:1:0.5) (13). The identity of CL and PG in these systems was further substantiated by chromatographic behavior of the products of mild alkaline methanolysis as described by Maruo and Benson (10) (glycerylphosphorylglycerylphosphorylglycerol and grycergylphosphorylglycerol, respectively.) Measurement of complex lipid synthesis. Complex lipid synthesis was measured in cultures of L. plantarum grown in the presence of [3H Joleic acid (2.5 MCi/ml; specific activity, 8 Mg/MCi). The radiolabeled oleic acid was added as a dispersion of 20 mg of oleic acid and 100 mg of Triton X-100 in 10 ml of water. Samples (20 ml) were removed at various time intervals and the cells were collected on solvent-resistant membrane filters (Millipore Corp.). The filters were placed in Teflon-capped test tubes containing 7.5 ml of methanol, 3.75 ml of chloroform, and 3.75 ml of water, mixed on a Vortex stirrer, and allowed to stand overnight. Chloroform and water (3.75 ml of each) were added to each tube, the contents were mixed, and the layers were allowed to separate. The top aqueous layer was removed by aspiration, the remaining chloroform layer was evaporated to dryness under a stream of nitrogen, and the residue was dissolved in 0.5 ml of benzene. Samples (200 Ml) were spotted on Silica Gel G thin-layer plates, chromatographed, and quantitated as described above.
RESULTS An enzyme located in the membrane fragments of L. plantarum catalyzes the nucleotideindependent conversion of PG to CL with the stoichiometry of the reaction being: 2 PG - 1 CL + (glycerol)
BURRrIT AND HENDERSON
All of this enzyme activity, initially found in the 40,000 x g supernatant solution, was found in the pellet after centrifugation at 100,000 x g for 2 h. Triton X-100, at concentrations above 0.3% final concentration, inhibited the enzyme, and the enzyme was inactivated by heating at 100 C for 5 min. The enzyme did not appear to have a divalent metal requirement as demonstrated by its insensitivity to added ethylenediaminetetraacetic acid or the addition of MgCl2. It was not affected by the addition of dithiothreitol, although there appeared to be a slight inhibition by mercaptoethanol. p-Hydroxymercuribenzoate, at two concentrations, was inhibitory. These results are summarized in Table 1. The pH optimum for L. plantarum CL synthetase was 5.1 (Fig. 1). At both lower and higher pH values, the amount of PG converted to CL decreased markedly. At the lower pH values, there was an increase in radioactivity in the neutral lipids; this was not true above pH 6. Therefore, except where noted, the reactions were carried out at pH 6. (The radioactive neutral lipids were co-chromatographed in the TAE 1. Some properties of cardiolipin synthetase from Lactobacillus plantarum 17-5G Addition to assay mixtureb
CL formed (nmol/mg of protein)
None (control) ..................... MgCl3 (7 mM) ...................... EDTA (50 mM) ....................
85.0 ± 2.0 81.5 ± 8.1 90.0 0.1
Triton X-100 (0.3%, final conc) ................. 81.7 ± 15.7 (0.4%, final conc) ................. 30.05 ± 1.2 (0.5%, final conc) ................. 4.9 ± 4.0 Mercaptoethanol (2 mM) ............ 65.0 ± 7.3 Dithiothreitol (4 mM) .............. 82.9 ± 0.3
p-Hydroxymercuribenzoate 0.1 mM .......................... (0.25 mM) ....................... CTP (300 nmol) .................... CDP-DG (300 mmol) ............... CL (30 nmol) (0.3% Triton X-100, final conc) ....................... None (boiled enzyme preparation,
29.7 1.1 23.6 i 0 76.9 ± 4.2 66.7 ± 9.5 10.8 ± 1.5
aAssay tubes contained 300 nmol of ["C IPG (50,000 counts/min), Triton X-100 (0.2%, final concentration), and 100 to 150 gg of membrane protein. The mixtures were brought to a final volume of 0.5 ml with 50 mM potassium phosphate buffer, pH 6.0, and incubated at 37 C for 30 min. The reaction was terminated by the addition of methanol and chloroform as described in the text. b Abbreviations: EDTA, ethylenediamenetetraacetic acid; CTP, cytidine triphosphate.
I I.6 . I
2 'S& -a--6r0 I- 4
pH FIG. 1. pH optimum curve for cardiolipin synthetase from L. plantarum. The enzyme activity was measured in 50 mM buffers, containing 10 mM MgCI2, 4 mg of membrane protein, and 40 nmol of [9ljphosphatidylglycerol (71,000 counts/min) as described in the text. Solid line represents cardiolipin synthetase activity; dashed line represents phospholipase activity. The buffers used were: glycine-hydrochloride (A); sodium acetate (0); potassium phosphate (0); and tris(hydroxymethyl)aminomethanehydrochloride (A).
solvent systems of Wagrier et al.  and of Malins and Mangold  and shown to migrate with synthetic 1,2-diacylglycerol, indicating the action of either a phospholipase C-like enzyme with PG as the substrate, or the sequential action of a membrane-bound phospholipase D on PG and phosphatidic acid phosphatase on the resulting phosphatidic acid. The membrane-bound enzyme(s) has a pH optimum of 4.6 [Fig. 1 ] and does not hydrolyze CL under the conditions investigated thus far.) CL was the only phosphate-containing product of the synthetase reaction, as determined by co-chromatography of the products of deacylation with authentic CL that had been carried through the same procedure (10). The synthesis of CL proceeded rapidly and was linear for the first 30 min of reaction time, with the increase in radioactive CL being balanced by the stoichiometric decrease in PG (Fig. 2). The enzyme rate increased linearly with respect to protein concentration (Fig. 3) in the range of 10 to 100 Mg of membrane protein
VOL. 123, 1975
CARDIOLIPIN SYNTHETASE FROM L. PLANTARUM
E c) 0
FIG. 2. Time course for cardiolipin synthetase. The activity was measured in 50 mM potassium phosphate buffer, pH 6.0, containing 10 mM MgCl2. Each assay contained 175 jsg of membrane protein and 300 nmol of [14CIPG (32,600 counts/min) in 0.5 ml. Symbols: (A) cardiolipin formed; (0) phosphatidylglycerol remaining. Each point represents the average of duplicate assays and the bars represent the range of the assay. enzyme
but was not linear with respect to substrate concentration at either low or high substrate concentrations (Fig. 4). Because CDP-DG had been postulated to be a substrate in the CL synthetase reaction (7), the effect of this compound on enzyme activity was investigated. When either CDP-DG (300 nmol) or cytidine triphosphate (300 nmol) was added to the reaction mixture, no increase or stimulation of CL formation was seen. In fact, if anything, there was a slight inhibition by CDP-DG (Table 1). The product of the reaction, CL, was tested for its ability to inhibit enzyme activity. When 30 nmol of CL (maximum amount formed under the experimental conditions used) was incubated with CL synthetase, the amount of CL formed was decreased by 53% relative to controls (Table 1). The changes in cell mass, lipid composition, and CL synthetase specific activity as a function of culture age are shown in Fig. 5. As the cells shifted from exponential growth to the stationary phase, the radioactivity of both CL and PG decreased abruptly, whereas the radioactivity of the glycolipids leveled off and then slowly decreased. The specific activity of CL synthetase, determined at four different times during the growth cycle, was highest during the
PROTEIN (tAg) FIG. 3. Conversion of phosphatidylglycerol to cardiolipin as a function of protein concentration. Each assay contained 192 nmol of [8H]PG (82,800 counts! min). The assays were carried out as described for Fig. 2; other conditions were as stated in the text.