INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Cardiolipin (CL)¹ is a structurally unique acidic phospholipid which carries four acyl groups and two negative charges. In eucaryotic cells, CL is found primarily in mitochondrial membranes (reviewed in Ref. 1). The unique structure and localization suggest an important functional role for CL in the mitochondria (reviewed in Ref. 2). In vitro studies have shown that CL is required for cytochrome c oxidase activity (3, 4) and ADP/ATP carrier activity (5). Evidence also suggests that CL may be involved in import of proteins into the mitochondria (6-10). Characterization of the enzymes which are required for CL synthesis would greatly facilitate our understanding of the biosynthesis and function of this lipid.

Regulation of CL biosynthesis has been studied in the model eucaryotes Saccharomyces cerevisiae and Schizosaccharomyces pombe, in crude cell and mitochondrial extracts (11-13). In both yeasts, the first step of the CL biosynthetic pathway, catalyzed by the enzyme CDP-diacylglycerol:sn-glycerol-3-phosphate 3-phosphatidyltransferase (phosphatidylglycerolphosphate synthase; PGPS; EC 2.7.8.5), appears to be highly regulated. In S. cerevisiae and S. pombe, PGPS is regulated by the phospholipid precursors inositol and choline (11, 12). In S. cerevisiae, this enzyme is also regulated by factors which affect mitochondrial development, such as carbon source, growth phase, and the presence of a mitochondrial genome (13). The fact that regulation of PGPS has been conserved in two yeasts that are only distantly related evolutionarily indicates that this enzyme plays a key role in the regulation of CL biosynthesis. Therefore, to further understand how CL is synthesized, we purified PGPS from the yeast S. pombe and characterized its enzymological properties.

A major obstacle to the purification of membrane-associated enzymes is that, following solubilization, the enzyme molecules usually exist in heterogeneous protein-detergent-lipid micellar complexes, which render many classical purification techniques ineffective. A second problem is that membrane enzymes frequently require the association of intrinsic phospholipids for either stability or activity, while extensive purification procedures result in removal of lipids from the enzyme, leading to loss of enzyme activity. We were able to circumvent these problems using a variety of approaches. The enzyme was solubilized with the mild nonionic detergent Triton X-100, which did not interfere with ion exchange chromatography. Furthermore, identification of phospholipid cofactors in experiments with partially purified enzyme was useful in overcoming the loss of activity. In addition, salt-induced phase separation followed by gel filtration greatly enriched and concentrated the enzyme without significant loss of activity. Finally, we utilized CDP-DG affinity chromatography (14), which has been effectively used in the purification of PGP synthase from Bacillus licheniformis (14) and Escherichia coli (15), phosphatidylserine synthase (PSS) from E. coli (16), Clostridium perfringes (17), S. cerevisiae (18), and B. licheniformis (19), phosphatidylinositol (PI) synthase from S. cerevisiae (20) and human placenta (21), and CDP-DG synthase from S. cerevisiae (22). In this paper, we report details of the purification of S. pombe PGPS to homogeneity, and describe its kinetic and enzymological properties.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Materials-- All chemicals used were reagent grade or better. Phenylmethylsulfonyl Fluoride, leupeptin, aprotinin, and glycerol kinase were obtained from Boehringer. Triton X-100 and Silica Gel 60H were purchased from EM Science. Sodium dodecyl sulfate (electrophoresis grade) and Florisil (60-100 mesh) were from Fisher Scientific. Peptone and yeast extract were obtained from Difco Laboratories. Centricon 10 microconcentrators were made by Amicon. The following chromatographic materials were from Pharmacia: Superose 12 HR 10/30 gel filtration column, Sephacryl H-200 HR resin, S-Sepharose Fast Flow resin, agarose-adipic acid hydrazide, GelBond film as well as the gel filtration calibration kit. The Mini-Protean II electrophoresis system, molecular mass standards for electrophoresis, silver stain kit, Bradford protein determination kit, hydroxylapatite, Bio-Lyte 3/10 carrier ampholyte, and Agarose Zero-M_r were purchased from Bio-Rad. CDP-diacylglycerol (CDP-DG) was purchased from Life Science Resources; all other phospholipids were from Sigma. Biosafe II liquid scintillant was purchased from Research Products International Co. [2-³H]Glycerol was purchased from NEN Life Science Products. [2-³H]Glycerol 3-phosphate was synthesized as described previously (23, 24).

Synthesis of CDP-DG Affinity Resin-- NaIO4-oxidized CDP-DG was attached to an agarose matrix by incubating with agarose-adipic acid hydrazide (14). The oxidization and coupling reactions were optimized for efficient binding of PGPS (20).

Yeast Strain and Growth Conditions-- Wild type S. pombe strain 972 (h) was used for enzyme purification. Cultures were maintained in 15% glycerol at 80 °C for long term storage and on YE (0.5% yeast extract, 3% glucose, and 2.3% agar) plates at 4 °C for short-term storage. Cells were grown in YE media at 30 °C to late exponential phase in a 200-liter fermentor, harvested by centrifugation, and stored at 80 °C.

Preparation of Mitochondria-- 400 g (wet weight) of cells obtained from a 200-liter fermentor culture were stored frozen at 80 °C. To prepare mitochondria, 50 g (wet weight) aliquots of cells were thawed, added to 300 g of chilled, acid-washed glass beads, and suspended in buffer A (50 mM Tris-HCl, 0.6 M sorbitol, 1 mM EDTA, 1 mM -mercaptoethanol, 1 µg/ml aprotinin, 2 µg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride). Cells were broken using a glass bead beater (Biospec Products) for four 1-min bursts, with a 5-min cooling interval on ice between each burst. This cell extract was centrifuged at 4,000 × g for 5 min in a GS3 rotor to remove unbroken cells and cell debris. The supernatant was then centrifuged at 17,000 × g for 20 min in an SS34 rotor. The pellet was resuspended in 60 ml of buffer A and centrifuged again at 3,000 × g for 5 min in an SS34 rotor. Mitochondria were collected from the supernatant by spinning at 17,000 × g for 20 min, and the pellet was resuspended in 1 ml of buffer A and routinely frozen at 80 °C for future purification and analysis.

Solubilization of PGPS-- Prior to solubilization, mitochondria were diluted with buffer A containing 250 mM KCl to a final protein concentration of 5 mg/ml, and the mixture was agitated on ice for 30 min to remove loosely bound extrinsic proteins. Mitochondria were then collected by centrifugation at 17,000 × g for 20 min. The pellets were resuspended in solubilization buffer (buffer B) (50 mM Tris-HCl (pH 7.5), 1% Triton X-100 (w/v), 250 mM KC1, 1 mM -mercaptoethanol, 1 µg/ml aprotinin, 2 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 20% glycerol), and agitated on a rototorque at medium speed for 1 h. The solubilization mixture was then overlayered with one-tenth volume of 50 mM Tris-HCl (pH 7.5) buffer and subjected to centrifugation at 100,000 × g for 1 h. The clear supernatant under the overlayer was collected as the Triton X-100 extract. The enzyme was stable at this stage in the presence of protease inhibitors and could be stored at 80 °C for months.

Ammonium Sulfate Phase Separation-- A two-step phase separation procedure was carried out according to Parish et al. (25) with modifications optimized for the enrichment of PGPS activity. Briefly, the ammonium sulfate (AS) concentration of the Triton X-100 extract was adjusted to 30% saturation by adding saturated AS solution in Tris-HCl buffer. The mixture was vortexed and centrifuged immediately to separate the phases. About 70% of the total PGPS activity and less than 30% of the total protein was present in the aqueous phase. The resultant lipid phase was discarded, and solid AS was then added to a final concentration of 55% saturation. The mixture was stirred for 1 h on ice and then centrifuged at 100,000 × g to pellet the protein. The pellet was dissolved in a small volume of solubilization buffer (buffer B). This step resulted in a 5-6-fold purification and 70% yield of PGPS. The enzyme had a highly uniform micellar structure, as evidenced by a sharp, symmetric peak during Superose 12 HR 10/30 chromatography (data not shown).

Gel Filtration-- A Sephacryl H-200 HR column (1 × 70 cm) was packed and equilibrated with Buffer B. The column was calibrated by determining the elution positions of proteins of known molecular weight (26, 27). Enzyme sample following phase separation was applied to the column and eluted with buffer B at a flow rate of 50 ml/h. Eluates containing peak PGPS activity were pooled and used for further purification. The yield of PGPS over the previous step was 74%, with a 4-fold increase in specific activity.

S-Sepharose Chromatography-- A 25-ml S-Sepharose column was packed and equilibrated with buffer C (50 mM Tris-HCl (pH 7.5), 0.1% Triton X-100, 20% glycerol, 1 µg/ml aprotinin, 2 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 1 mM -mercaptoethanol) containing 180 mM KCl. The pooled peak fractions from gel filtration were diluted 1:1 with buffer C without KCl, and applied to the column at a flow rate of 40 ml/h using a peristaltic pump. 75% of the applied protein and less than 10% of the applied PGPS activity were present in the flow-through fraction. The column was then washed with 3 column volumes of buffer C containing 250 mM KCl. PGPS was eluted with 6 volumes of buffer C using a linear gradient of KCl from 250 to 500 mM. Typically, 60% of the applied PGPS activity was recovered with 6% of the applied protein in the 350 mM KCl eluates. A 12-fold enrichment of the enzyme over the previous step was achieved. Since PGPS was labile at this stage, the fractions containing peak PGPS activity were identified by the quick enzyme assay described below and immediately applied to the next column.

Hydroxylapatite Chromatography-- A hydroxylapatite column (2.5 × 5 cm) was packed and equilibrated with buffer D (10% glycerol, 0.1% Triton X-100, 1 µg/ml aprotinin, 2 µg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, and 1 mM -mercaptoethanol) containing 10 mM potassium phosphate (pH 7.5). Active fractions from S-Sepharose were diluted 1:1 in buffer B without KCl and applied to the column at a flow rate of 20 ml/h using a peristaltic pump. The column was washed with 5 column volumes of equilibration buffer followed by elution of the enzyme using 8 column volumes of a linear gradient of potassium phosphate (300-500 mM) in buffer D. The majority of bound PGPS activity was eluted from the column at a phosphate concentration of 400 mM. A 5.8-fold increase in specific activity was achieved, with a 46% yield of the applied activity. Because the enzyme was highly sensitive to denaturation at this stage and could not tolerate freeze-thaw cycles, it was applied to the next column immediately.

CDP-DG Affinity Chromatography-- The hydroxylapatite eluate containing PGPS activity was first pumped through a 1.0 × 16-cm G-25 gel filtration column for buffer exchange. The running buffer contained 50 mM Tris-HCl (pH 7.5), 20% glycerol, 60 mM KCl, 5 mM MgCl₂, and 0.5% Triton X-100, which was optimal for the binding of PGPS to the CDP-DG resin. The enzyme sample was then applied to a 1.5 × 5-cm CDP-DG affinity column pre-equilibrated with running buffer at a flow rate of 30 ml/h. Typically, over 85% of the total PGPS activity bound to the resin. The column was then washed with 3 bed volumes of running buffer containing 0.1% Triton X-100. The bound synthase activity was eluted from the resin using 20 volumes of running buffer with a linear gradient of NaCl from 0.2 to 1.0 M. Fractions were assayed for PGPS activity in the presence of phospholipid cofactors. The peak PGPS activity was eluted at a NaCl concentration of 0.3 M. Alternatively, enzyme activity could be eluted by resuspending the resin in 1 bed volume of binding buffer containing 1.5 mM CDP-DG and agitating the mixture for 1 h. Eluting with high salt was used in the final purification scheme, since excess CDP-DG complicated further analysis of the enzyme. Both methods failed to elute all the bound activity from the column, as evidenced by the remaining PGPS activity on the eluted resin. To concentrate the enzyme sample, fractions containing peak PGPS activity were diluted 1:1 with binding buffer, and the activity was reabsorbed to a 1.0-ml CDP-DG affinity column. The enzyme was eluted with 2.0 ml of buffer C containing 0.5 M NaCl. The eluates were then further concentrated and desalted using Centricon-10 microconcentrator in a nitrogen atmosphere.

Enzyme Assay-- PGPS activity was assayed at 30 °C by quantitating the incorporation of 0.5 mM [2-³H]glycerol 3-phosphate (3500-5500 dpm/nmol) into chloroform-soluble material as described by Karkhoff-Schweizer et al. (11). The reaction mixture contained 200 mM Tris-HCl (pH 7.5), 0.6 mM CDP-DG, 6 mM Triton X-100, and 0.1 mM MgCl₂ in a total volume of 0.1 ml. The reaction was started by addition of labeled substrate to the reaction mixture, which was incubated at 30 °C for 20 min. The reaction was stopped by addition of 0.5 ml of 0.1 N HCl in methanol. To separate the labeled substrate from the labeled reaction product, 1 ml of chloroform and 1.5 ml of 1 M MgCl₂ were added to each assay tube. The tubes were vortexed and the phases separated by brief centrifugation. Aliquots of 0.5 ml of the chloroform phase were dried and counted in a 1600 TR liquid scintillation analyzer (Packard Instrument Co.).

Preparation of Detergent/Phospholipid Mixed Micelles-- Phospholipids in chloroform or alcohol were transferred to glass tubes and dried under a nitrogen current. The residual solvent was removed by keeping the dried samples in vacuo for 20 min at room temperature. Triton X-100/phospholipid micelles were prepared by adding Triton X-100 solution to the dried phospholipids and vortexing at 40 °C until a homogeneous suspension was obtained.

Protein Assay-- Protein concentration was determined by the method of Bradford (29) with bovine serum albumin as the standard. Buffers identical to those containing the protein samples were used as blanks to overcome the interference of reagents such as Triton X-100.

Electrophoresis-- SDS-polyacrylamide gel electrophoresis (SDS-PAGE) was performed on 10 × 7-cm slab minigels with a thickness of 0.75 or 1.5 mm at an acrylamide concentration of 12.5%. Electrophoresis was carried out using the procedure described by Laemmli (30). Nondenaturing polyacrylamide gel electrophoresis in the presence of Triton X-100 was performed according to Warlow and Bernard (31). Nondenaturing polyacrylamide-agarose combination gel electrophoresis in the presence of CHAPS was performed as described by Aledo et al. (32). Agarose-IEF was performed on a 70 × 115 × 0.5-mm gel cast on GelBond film using the procedure of Righetti et al. (33). The gel contained 0.5% agarose, 0.8% (w/v) Triton X-100, 12% sorbitol, and 1/20 volume of Bio-Lyte 3/10. Samples were applied using 0.5 × 1.0-cm filter applicators. The duration of the run was 4,000 volt hours, with maximum voltage of 1,500 volts. After electrophoresis, the gel was fixed with 20% trichloroacetic acid for 30 min, washed with 10% acetic acid, 25% methanol, and dried directly in a warm air current. All gels were stained with either Coomassie Brilliant Blue R-250 or silver stain (Bio-Rad).

Determination of PGPS Activity in Gel Slices-- After electrophoresis, the gel was cut into 0.5-cm slices. Each slice was then minced in 0.2 ml of assay mixture (with lipid cofactor if necessary). The mixture was kept at 30 °C for 4-12 h with agitation. The gel debris was then sedimented by centrifugation, and a 0.15-ml aliquot was withdrawn from the supernatant. [³H]PGP was extracted from this aliquot, isolated, and quantitated as described.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Purification of PGPS

A summary of the purification of PGPS is shown in Table I. The overall purification of the enzyme over the KCl-washed mitochondria was 1577-fold, with an activity yield of 2%. However, although the CDP-DG affinity chromatography purified PGPS to homogeneity, the enzyme had no detectable activity under standard assay conditions without lipid activator, and was also significantly denatured. Therefore, the specific activity determinations of purified enzyme could not accurately reflect the purification fold.

The purified PGPS was subjected to SDS-PAGE. A single band with an apparent molecular mass of about 60 kDa was visualized by silver stain (Fig. 1A). Efforts to reconstitute enzyme activity after SDS-PAGE and Western blot in the presence of phospholipid cofactors were unsuccessful, probably due to irreversible denaturation by SDS. Polyacrylamide gel electrophoresis under nondenaturing conditions was carried out using both Triton X-100 and CHAPS systems as described (31). In both cases, although PGPS activity was steadily recovered after electrophoresis for 30 h, the proteins formed precipitates in the stacking gel and thus could not be resolved. While the reason for this was not clear, hydrophobic interactions between protein/micelle and gel matrix were possibly involved. To solve this problem, we employed agarose-isoelectrofocusing (IEF) as described (32), using a 0.8% horizontal agarose gel. A pH gradient was formed using Bio-Lyte 3/10 carrier Ampholyte (Bio-Rad). As shown in Fig. 1B, the sample migrated as a single spot which, in the presence of CL, was associated with PGPS activity (Fig. 1C).

View larger version (46K): [in this window] [in a new window]   Fig. 1.   Electrophoretic analysis of purified PGPS. A, SDS-PAGE analysis of purified PGPS. Purified PGPS (approximately 5 µg) resolved on SDS-PAGE and stained with Silver Stain Plus (Bio-Rad). B, purified PGPS (approximately 6 mg) resolved on a horizontal agarose-IEF gel (pH 3-10) in the presence of 0.5% Triton X-100. The gel was air dried and silver stained. C, PGPS activity recovered from gel slices containing the protein spot after nondenaturing agarose-IEF. A parallel lane on the agarose gel shown in B was cut into slices of 0.5 cm and assayed for PGPS activity as described under "Experimental Procedures."

Properties of PGPS

The purified PGPS was analyzed, together with either a partially purified sample or crude mitochondrial membrane when necessary, to elucidate the properties of this enzyme. The partially purified enzyme was obtained by incorporating Triton X-100 solubilization, hydroxylapatite and ionic exchange chromatographies. The preparation retained substantial PGPS activity without the addition of lipid activators. To assay the purified enzyme, CL was added to the reaction mixture.

Effect of pH, Divalent Cations, Detergents, and Temperature on PGPS Activity-- PGPS activity was measured at pH 6.5-7.5 using 200 mM Tris maleate buffer, and at pH 7.5-9.5 using 200 mM Tris-HCl buffer. All other components in the reaction mixture were kept constant, and assays were performed using standard conditions as described above. Results are shown in Fig. 2. The optimal pH for crude mitochondrial extract was 8.0, and activity decreased slowly as the pH was raised or lowered. The partially purified sample had an optimal pH of 7.5. In contrast, the purified enzyme had an optimal pH of 7.0, and a very narrow pH range of activity. It is possible that the sensitivity to pH change was due to removal of intrinsic lipids from the enzyme.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Effect of pH on PGPS activity. PGPS activity of the purified () and partially purified () enzyme, and the mitochondrial crude extract () were measured at the indicated pH values with 200 mM Tris malate buffer (pH 6.5-7.5) or 200 mM Tris-HCl buffer (pH 7.5-9.5). Activity is normalized relative to those obtained at the optimal pH values (100%).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Effect of divalent cations on PGPS activity. Partially purified PGPS (1170 units/mg) was measured at the indicated concentrations of MgCl2 () or MnCl2 (). Activity is relative to that obtained in the absence of cations (100%).

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Effect of Triton X-100 on PGPS activity. Partially purified PGPS (1170 units/mg) was measured at the indicated concentrations of Triton X-100. Activity is relative to that obtained at the optimal [Triton X-100] (100%).

Enzyme Kinetics-- Purified and partially purified PGPS exhibited typical saturation kinetics when glycerol 3-phosphate concentrations were varied at fixed concentrations of CDP-DG (data not shown). The same saturation kinetics were observed in crude mitochondrial membrane (11). For analysis of PGPS kinetics with respect to CDP-DG, a mixed micelle system with Triton X-100 was used. PGPS activity was measured as a function of the surface concentration of CDP-DG in the micelle. The bulk CDP-DG concentration was held constant at 0.2 mM. The molar ratio of CDP-DG in the micelle was varied by changing the concentration of Triton X-100. The results are shown in Figs. 5 and 6B. When the molar ratio of CDP-DG was varied at fixed concentrations of glycerol 3-phosphate, the kinetic curve of the purified and the partially purified enzyme was sigmoidal and appeared positively cooperative, in contrast to the typical saturation kinetics observed in crude mitochondrial membrane (11). In addition, the purified enzyme reached its maximal velocity at a higher molar concentration of CDP-DG in comparison with the partially purified sample (Fig. 6B).

View larger version (21K): [in this window] [in a new window]   Fig. 5.   CDP-DG dependence of partially purified PGPS in Triton X-100/CDP-DG mixed micelles. PGPS activity of the partially purified sample was measured as a function of the mole % CDP-DG in the micelles at fixed concentrations of glycerol 3-phosphate. The bulk concentration of CDP-DG was maintained at 0.2 mM; the concentration of Triton X-100 was varied to adjust the molar concentration of CDP-DG.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   CDP-DG dependence of the purified PGPS in micelles with and without CL. A, PGPS activity of the affinity purified sample was measured as a function of the mole % CDP-DG in the micelles at fixed concentration of glycerol 3-phosphate. The bulk concentration of CDP-DG was maintained at 0.2 mM; 10 mol % CL was added to the +CL group, and total micelle concentration was constant. B, re-plot of PGPS activity versus the CDP-DG concentrations on a larger scale.

Effect of Phospholipids on PGPS Activity in Vitro-- The sigmoidal nature of the kinetic curve for CDP-DG dependence suggested that PGPS activity is affected by phospholipids. We therefore investigated the effect of individual phospholipids on the purified and partially purified enzymes.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Effect of CL and PI on CDP-DG dependence of partially purified PGPS. PGPS activity was measured in a partially purified sample as a function of mole % CDP-DG in the mixed micelles at fixed concentration of glycerol 3-phosphate. The bulk concentration of CDP-DG was maintained at 0.2 mM; concentrations of Triton X-100, CL, or PI were varied. , 2% CL; , 4% CL; , 0% CL, PI; , 2% PI; , 4% PI.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Effect of phospholipids or lipid precursors on PGPS activity. PGPS activity of the affinity purified sample was measured as a function of the mole % of lipids in the mixed micelles. The bulk concentrations of glycerol 3-phosphate and CDP-DG were fixed. Lipids were incorporated into the mixed micelles by sonication, and the total micelle concentration was maintained constant by varying the concentration of Triton X-100. , E. coli CL; , bovine CL; ×, PA; , PI; , PS; , DG; , PG.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

In this report we describe the first purification to homogeneity of a eucaryotic PGPS. The S. pombe protein has a molecular mass of 60 kDa, and its activity is strongly affected by phospholipids.

Purification of PGPS-- By incorporating salt-induced phase separation into our protocol (25), we enriched and concentrated the enzyme without significant loss of activity. It has been observed that, during phase separation, some membrane proteins fail to enter the lipid phase, while some soluble proteins do (25, 35). In our procedure, the majority of PGPS activity remained in the aqueous solution during the first phase separation. This was advantageous since many hydrophobic proteins were removed, together with most detergent and membrane lipids, avoiding further interference with the purification. A portion of the PGPS activity (about 30%) entered the lipid layer, possibly due to the heterogeneous nature of the crude solubilization mixture. Since a high concentration of AS tends to stabilize most enzymes, AS-induced phase separation followed by gel filtration may be useful as a general procedure in membrane protein purification.

Activation of PGPS by Phospholipids-- Activation of membrane proteins by phospholipids is a common phenomenon (38). Activation of phospholipid biosynthetic enzymes is of particular interest because of the possibility that phospholipids may play dual roles as both substrates and cofactors (39). Two classes of activation by phospholipids have been proposed, and examples of each class have been found among phospholipid biosynthetic enzymes (39). The first class includes enzymes that require a large number of phospholipid molecules or lipid boundary for activation (40). Examples include glycerolphosphate acyltransferase from E. coli, which required 20 mol % CL or 40 mol % PG for half-maximal stimulation of activity (41), and rat liver phosphatidylethanolamine N-methyltransferase, which required 30-40 mol % phosphatidylcholine to convert the Hill coefficient to near one (40). The second class of activation involves a few activator molecules, probably binding to specific sites (39). An example of this class is diglyceride kinase from E. coli. Diacylglycerol served as both substrate and activator of this enzyme, and CL activated the enzyme half-maximally at less than 1 mol % (42). The yeast ethanolamine and choline phosphotransferases also exhibited absolute dependence on phospholipids for activity, with half-maximal activation at 2.5 mol % phosphatidylcholine and 0.5 mol % PC, respectively (43).

Implications for Regulation of PGPS in Vivo-- The inhibition of PGPS activity by PI is intriguing in light of the regulation studies reported by Gaynor and Greenberg (44). They showed that both phosphatidylinositol synthase and PGPS from S. pombe are regulated by inositol, and examined the kinetics of inositol regulation for both enzymes. Expression of PI synthase decreased when cells were starved for inositol; 30 min after addition of inositol to the culture medium, PI synthase expression increased. In a similar experiment, PGPS expression from inositol-starved cells was increased, after addition of inositol to the culture medium, a decrease in PGPS expression was not detected for 2 h. These results may be explained by the effect of PI on PGPS. It is possible that the derepression of PI synthase leads to increases in PI levels in mitochondrial membranes. Increased PI levels could, in turn, lead to decreased activity of PGPS. Thus the regulation of PGPS by PI could maintain a balance of charged phospholipids in mitochondrial membranes. Experiments to determine the effect of exogenous inositol on the composition of mitochondrial membranes are necessary to determine the validity of this model.