Neurochemical Research, Vol. 23. No. 5, 1998, pp. 589-599

Kinetic Analysis in Mixed Micelles of Partially Purified Rat Brain Phospholipase D Activity and its Activation by Phosphatidylinositol 4,5-Bisphosphate* Vered Chalifa-Caspi,1 Yona Eli,1 and Mordechai Liscovitch1'2 (Accepted October 16, 1997)

A partially purified rat brain membrane phospholipase D (PLD) activity was characterized in a mixed micellar system consisting of l-palmitoyl-2-[6-n-(7-nitrobenzo-2-oxa-l,3-diazol-4-yl)amino]caproyl-phosphatidylcholine (NBD-PC) and Triton X-100, under conditions where Triton X-100 has a surface dilution effect on PLD activity and the catalytic rate is dependent on the surface concentration (expressed in terms of molar ratio) of NBD-PC. PLD activity was specifically activated by phosphatidylinositol 4,5-bisphosphate (PIP2), and the curve of activation versus PIP2 molar ratio fitted a Michaelis-Menten equation with a Kact value between molar ratios of 0.0010.002. Maximal activation was observed at a PIP2 molar ratio of 0.01. Similar values were obtained when activities of partially purified PLD as well as membrane-bound PLD were determined towards pure NBD-PC micelles. In the mixed micellar system PIP2 was shown to elevate by 6-22 fold the specificity constant of PLD towards NBD-PC (KA, which is proportional to Vmax /K m ). Kinetic analysis of PLD trans-phosphatidylation activity towards ethanol, 1-propanol and 1-butanol revealed a Michaelis-Menten type dependence on alcohol concentration up to 1000, 200 and 80 mM, respectively. While Vmax values were similar towards all three alcohols, enzyme affinity increased as the alcohol was longer, and Km values for ethanol, 1-propanol and 1-butanol were 291, 75 and 16 mM (respectively). PLD specificity constants (KA) towards ethanol, 1-propanol and 1-butanol were shown to be respectively 260, 940 and 5,920 times higher than to water, the competing substrate. 1-Propanol and 1-butanol inhibited PLD activity above 400 and 100 mM, respectively. The present results indicate that partially purified PLD obeys surface dilution kinetics with regard to its phospholipid substrate PC and its cofactor PIP2, and that in the presence of alcohols, its transphosphatidylation activity may be analyzed as a competitive reaction to the hydrolysis reaction.

KEY WORDS: Phospholipase D; rat brain; surface dilution kinetics; signal transduction; alcohol.

thus producing phosphatidic acid (PA), and releasing the free polar head group. In the presence of a primary alcohol, the enzyme can also catalyze a trans-phosphati-

INTRODUCTION Phospholipase D (PLD1) catalyzes the hydrolysis of phospholipids at their terminal phosphodiester bond, Department of Biological Regulation, The Weizmann Institute of Science, Rehovot 76100, Israel. 2 Address reprint requests to: Dr. M. Liscovitch, Department of Biological Regulation, Weizmann Institute of Science, P.O. Box 26, Rehovot 76100, Israel. Tel: +972 8 934 2773, Fax: +972 8 934 4116, E-mail: [email protected]. * Special issue dedicated to Dr. Richard J. Wurtman. 1

' The abbreviations used are: ARF, ADP-ribosylation factor; CMC, critical micellar concentration; FU, fluorescence units; PA, phosphatidic acid; PC, phosphatidylcholine; NBD-PC, l-palmitoyl-2-[6-A' (7-nitrobenzo-2-oxa-l,3-diazol-4-yl)amino] caproyl-phosphatidylcholine; PIP2, phosphatidylinositol 4,5-bisphosphate; PLD, phospholipase D; PPr, phosphatidylpropanol; TLC, thin layer chromatography.

589 0364-3190/98/0500-0589$15.00/0 C 1998 Plenum Publishing Corporation

590 dylation reaction in which the primary alcohol serves as a phosphatidyl group acceptor (1). PLD is nearly ubiquitous in the mammalian body and, depending on cell type, is activated following cell surface receptor stimulation by agonists such as neurotransmitters, hormones, growth factors, cytokines, antigens and cell adhesion molecules (2,3). PLD activation involves multiple mechanisms, including guanine nucleotide-binding (G) protein(s), elevated Ca2+ levels, protein kinases C and tyrosine phosphorylation (4,5). Activation of PLD leads to elevation of two important signal molecules: PA, the primary hydrolytic product of PLD, and diacylglycerol, which is formed from PA by the action of phosphatidic acid phosphohydrolase (6). The cellular response to these signals depends on the cell type and may include, for example, mitogenesis and hormone release (3). Multiple forms of eukaryotic PLDs have been molecularly cloned recently, including a plant enzyme (7), two mammalian PLDs (8,9), and a yeast PLD (10-12). We have previously described in vitro conditions for assaying synaptic plasma membrane PLD activity at a neutral pH (13), utilizing exogenous [3H]dipalmitoylPC as a substrate. This PLD was specific for the cholinephosphoglycerides PC and 1-alkyl-PC, while activity towards other phospholipids such as phosphatidylserine, phosphatidylinositol, phosphatidylethanolamine and phosphatidylethanolamine-plasmalogen was barely detectable (14). Recently, we and others have found that phosphatidylinositol 4,5-bisphosphate (PIP2) stimulates PLD activity and is required for PLD activation by G proteins (15-18). Stimulation of brain membrane PLD activity is specific to PIP2, as phosphatidylinositol 4phosphate, phosphatidylinositol, phosphatidylserine and PA were completely or nearly ineffective (16). In addition, PIP2 is required for PLD activation by the small G proteins ADP-ribosylation factor (ARF) (15) and RhoA (18), and PIP2 synthesis was found to be essential for PLD activation by guanine nucleotides (17). Previously we have shown that neomycin (which binds polyphosphoinositides with high affinity) inhibits PLD activity in brain membranes, as well as guanine nucleotide-induced PLD activation in permeabilized neuroblastoma-glioma hybrid cells (19). This effect likely results from the binding of neomycin to endogenous membrane PIP2, because the inhibitory action of neomycin is lost upon purification of PLD, but can be restored by including PIP2 in the assay (16). Collectively, these data suggest that PIP2 is a cofactor required for PLD activation in vitro and in vivo. A major limitation of previous studies in which membrane PC-PLD activity was measured in vitro was the use of poorly defined reaction systems. PLD, its

Chalifa-Caspi, Eli, and Liscovitch phospholipid substrate and its detergent-like activator(s) all reside in amphiphilic aggregates (e.g. vesicles, liposomes, micelles), on the surface of which the enzymatic reaction is believed to take place. Kinetic models that describe enzyme activity at amphiphilic surfaces (see below) require that the composition of the aggregates, their physical properties and the interaction between them are known, or at least that simplified assumptions can be made about them. The adaptation of Michaelis-Menten kinetics to the analysis of these reactions has been a challenge for a number of research groups. Detailed theoretical models have been developed to describe the hydrolytic activity of phospholipase A2 against phospholipid substrates (20-23) and kinetic analyses of other enzymes acting at amphiphilic surfaces were performed as well (e.g. protein kinase C (24), phospholipase C (25), diacylglycerol kinase (26), phosphatidic acid phosphohydrolase (27), phosphatidylethanolamine 7V-methyltransferase (28), phosphatidylserine synthase and phosphatidylinositol synthase (29). The basic principles that underlie kinetic studies of enzymes acting at amphiphilic surfaces have been reviewed recently (30). Therefore, a necessary condition for further characterization of PLD activity and its regulation in vitro was the application of a well-defined reaction system and consideration of the special features of enzyme kinetics at amphiphilic surfaces. The aim of the present work was to carry out a kinetic analysis of partially purified PLD activity in a physically defined mixed phospholipid-detergent micellar system, and to determine the kinetic parameters of PLD activity toward its phospholipid and alcohol substrates as well as its activation by PIP2.

EXPERIMENTAL PROCEDURE Materials. 1 -Palmitoyl-2-[6-N-(7-nitrobenzo-2-oxa-1,3-diazol-4yl)amino]caproyl-phosphatidylcholine (NBD-PC) was obtained from Avanti Polar Lipids. Phosphatidylinositol 4,5-bisphosphate, Reactive Green 19-agarose and sodium oleate were obtained from Sigma. Hepes (sodium salt) was from Research Organics, Inc. Triton X-100 was from Aldrich. Silica gel LK6 thin layer chromatography glass plates were purchased from Whatman International. Q-Sepharose was from Pharmacia Biotech Inc. and Bio-Gel HTP hydroxyl apatite was from Bio-Rad. Preparation of Total Brain Membranes. Rat brain membranes were prepared essentially as described (31). Briefly, a homogenate of 36 brains from 30-day old male rats (in 0.25 M sucrose, 10 mM Hepes pH 7.2 and 1 mM EDTA) was centrifuged 10 min at 1,000 g. The pellet was washed once by resuspension and centrifugation and the pooled supernatants were ultracentrifuged for 1 h at 100,000 g. The pellet was homogenized again in a hypotonic lysis buffer (10 mM Hepes pH 7.2, 1 mM EDTA) and incubated for 30 min at 4°C. To remove peripheral membrane proteins, NaCl was added to a final con-

Kinetic Analysis of Brain Phospholipase D in Mixed Micelles

591

Table I. Partial Purification of Rat Brain Membrane PLD Protein Chromatography step

Triton X-100 extract Q-Sepharose Green 19 Agarose Hydroxyl-apatite Dialysis

Fraction volume

ml 390 40 17 6 6

PLD activity

Concentration

Total /fraction

Yield

mg/ml

mg

%

1.290 0.848 0.111 0.114 0.124

503.2 33.9 1.90 0.68

100.00 6.74 0.38 0.14

Activity measured

Total /fraction

Yield

Specific activity

Purification factor

FU 266

FU

%

FU/mg

fold

103,935 18,960 10,702 4,113 3,207

100.0 18.2 10.3 4.0 3.1

474 629 685 534

207 559 5,647 6,009 4,316

1 2.7 27.3 29.1 20.9

Triton X-100-solubilized total brain membrane proteins were separated by chromatography on a sequence of Q-Sepharose, Reactive Green 19Agarose and Hydroxyl Apatite columns as described under Experimental Procedures. PLD activity was determined by measuring PPr production in the presence of PIP2 as follows: Q-Sepharose, 20 uM PIP2; Reactive Green 19-Agarose, 10 uM PIP2; Hydroxyl Apatite, 10 u,M PIP2.

centration of 1 M and the membranes were further incubated for 30 min at 4°C. The membrane suspension was then ultracentrifuged for 1 h at 100,000 g, and the pellet suspended with a small volume of 50 mM Hepes pH 7.2. The membrane suspension was divided into portions, snap-frozen in liquid nitrogen and stored at —70°C. Protein concentration was determined by a modified Lowry procedure (32). Each such preparation yielded an average of 790 ± 150 mg protein. Triton X-100 Solubilization of Total Brain Membranes. Total brain membranes were brought to 5 mg protein/ml in solubilization buffer containing (final concentrations): 1% Triton X-100, 10 mM NaCl and 50 mM Hepes, pH 7.2. The mixture was gently stirred for 45 min at 4°C, and then ultracentrifuged for 90 min at 100,000 g. The supernatant was collected, frozen in liquid nitrogen and stored at -70°C. Partial Purification of PLD. A Triton X-100 extract of total brain membranes (140 or 190 ml) was mixed with an equal volume of 10 mM NaCl/50 mM Hepes pH 7.2, and loaded on a Q-Sepharose column (2.6 X 24 cm), which had been equilibrated with buffer A (0.5% Triton X-100, 10 mM NaCl, 50 mM Hepes, pH 7.2). The column was washed with 600 ml of buffer A, followed by 600 ml of buffer A containing an NaCl gradient of 0.01-0.6 M. The flow rate was 200 ml/h, and 8 ml fractions were collected. Active fractions were pooled, and 40 ml of the pool were loaded on a Reactive Green 19-agarose column (1.6 X 4.5 cm), which had been equilibrated with buffer A containing 0.5 M NaCl. The column was then successively eluted with buffer A containing: 0.45 M NaCl, 0.5% Triton X-100 (55 ml); 1 M NaCl, 0.5% Triton X-100 (55 ml); 1 M NaCl, 0.1% Triton X-100 (55 ml); 4 M NaCl, 0.1% Triton X-100 (24 ml). Flow rate was 28 ml/h during load, and 45 ml/h during the elution steps. The 4 M NaCl eluate was loaded on a Bio-Gel HTP hydroxyl apatite column containing 3 ml gel, which had been equilibrated with 0.01 M NaPO4 pH 6.8/0.1% Triton X-100 (Buffer B). The column was successively washed with buffer B (24 ml) and buffer B containing 0.1 M NaPO4 (15 ml), then transferred to room temperature and, after 10 min, PLD activity was eluted with 6 ml of buffer B (at room temperature) containing 0.3 M NaPO4. Flow rate was approximately 1 ml/min. The final eluate was dialyzed against 50 mM Hepes pH 7.2, 40 mM NaCl and 0.1% Triton X-100, frozen in liquid nitrogen and stored at -70°C. Kinetic Analysis of PLD Activity in Mixed Lipid-Detergent Micelles. Kinetic studies of Triton X-100-solubilized, partially purified PLD were carried out in a reaction mixture buffered with Hepes (50 mM, pH 7.2), containing 50 mM NaCl, and NBD-PC, Triton X-100 and PIP2, the amounts of which are indicated in the relevant figure legends. The final volume of the reaction mixture was 120 ul, and incubations were for 5 min at 37°C (unless otherwise indicated). The

reactions were terminated and products determined as described (31). Normally, NBD-PA production was measured. Where indicated, NBDPPr production was determined, in which case 150 mM of 1-propanol were included in the reaction mixture. Total mixed micellar content denotes the sum of the molar concentrations of the micellar components after subtracting their CMC values (0.3 mM and 32 nM for Triton X-100 and NBD-PC, respectively) when they are non-negligible. Molar ratio denotes the molar concentration of the micellar component under consideration (minus its CMC) divided by the total mixed micellar content. When calculating the Triton X-100 concentration, we took into account the amount donated by the PLD-containing fraction, calculating the molecular weight of Triton X-100 as 650. Fitness of the experimental results to Michaelis-Menten and Hill equations was carried out by nonlinear regression utilizing Kaleidagraph™ (Macintosh Version 2.1.1).

RESULTS Partial Purification of PLD. PLD activity was extracted from salt-washed, total rat brain membranes by solubilization with 1% Triton X-100. The partial purification of PLD on Q-Sepharose, Reactive Green 19Agarose and hydroxylapatite columns is summarized in Table I. The hydroxylapatite chromatography served mainly to remove the high NaCl concentration that accompanied PLD after the Reactive Green 19 chromatography. PLD activity after the hydroxylapatite chromatography was enriched 29 fold compared to the activity in the Triton X-100 extract, and the yield was 4%. Dependence of PLD Activity on NBD-PC Concentration. The standard mixed-micellar system used in the present study consisted of Triton X-100-dominated micelles, which contained the enzyme, the short chain micelle-forming substrate NBD-PC and, sometimes, PIP2 as an activator. Usually, PLD activity was measured by monitoring PA production by TLC and fluorescence measurement of the extracted lipid. PA was not converted to diacylglycerol under these reaction conditions

592

Chalifa-Caspi, Eli, and Liscovitch

Fig. 1. Dependence of PLD activity on NBD-PC concentration (Triton X-100 below its CMC). Reaction mixtures contained 0.01% (0.15 mM) Triton X-100 and the indicated concentrations of NBD-PC. The curve was fitted to Michaelis-Menten equation.

(data not shown). The PLD-catalyzed transphosphatidylation reaction was studied in several experiments, where production of PPr was determined in the presence of 1-propanol. In the absence of PIP2, PA accumulation was absolutely linear up to 30 minutes, and continued to increase up to 90 minutes (longer times were not measured; data not shown). The percentage of NBD-PC molecules converted to PA throughout the experiment was calculated to be less than 1%, indicating that the criterion for measuring "initial velocity" required for Michaelis-Menten analysis is met under these reaction conditions. Initially, the dependence of PLD activity on NBD-PC molar concentration was examined, keeping Triton X-100 below its CMC. As shown in Fig. 1, PLD activity toward these NBD-PC-dominated micelles exhibited a classical "Michaelis-Menten" behavior, with an apparent K^ of 0.02 mM and reaching saturation at 0.1 mM NBD-PC. In the next experiment, the dependence of PLD activity on the surface concentration of NBD-PC was examined in three mixed NBD-PC-Triton X-100 micellar systems, each containing a different total number of micelles (i.e., a different sum of Triton X100 and NBD-PC molar concentrations after subtracting CMCTriton x_100). As shown in Fig. 2A, enzyme activity increased with increasing NBD-PC molar ratio, and was independent on the total number of micelles. The inhibitory "surface dilution" effect of Triton X-100 on PLD activity in this experiment is evident when PLD activities are plotted against the molar concentration of NBDPC (Fig. 2B). Addition of Triton X-100 (included to reach the desired micellar concentration) caused a shift of the NBD-PC molar concentration-response curve to the right, without affecting maximal PLD activity. How-

Fig. 2. Dependence of PLD activity on NBD-PC concentration at different values of total mixed micellar content. Reaction mixtures contained variable amounts of NBD-PC and Triton X-100. Total mixed micellar content denotes [NBD-PC] + [Triton X-100]-CMCTnlon x-l00. Incubation time was 10 min. Panel A: NBD-PC Molar ratios examined were 0.05 (solid squares), 0.1 (solid circles), 0.2 (solid triangles), 0.4 (open squares), 0.6 (open circles) and 0.8 (open triangles). Panels B and C: total micellar content was either 0.5 (open circles), 1 (open squares) or 2 mM (open triangles).

ever, when the curves are replotted to show the dependence of PLD on NBD-PC surface concentration, the three curves are superimposable (Fig. 2C), indicating that at a total micellar content above 0.5 mM, PLD ac-

Kinetic Analysis of Brain Phospholipase D in Mixed Micelles

Fig. 3. Time course for PPr production in the presence and absence of PIP2. Reaction mixtures contained 0.3 mM NBD-PC (molar ratio 0.15), 1.66 mM Triton X-100, with (solid circles) or without (open circles) 40 uM PIP2 (molar ratio 0.02). Total mixed micellar content was 2 mM. PLD activity was determined by measurement of PPr production.

tivity solely depends on the surface concentration of NBD-PC2. For determination of the kinetic parameters of NBD-PC hydrolysis by PLD, NBD-PC molar ratios had to be kept below 0.15, so that the micelles would be dominated by Triton X-100 and their physical properties would not be significantly influenced by variations in NBD-PC molar ratio. However, at these NBD-PC molar ratios PLD activity increased linearly and was far from saturation, prohibiting a reliable evaluation of Km and Vmax (cf. Fig 5). Activation of PLD by PIP2. The effect of PIP2 (at a molar ratio of 0.02) on the time course of PLD activity in the Triton X-100-NBD-PC micellar system is shown in Fig. 3. PPr accumulation in the presence and absence of PIP2 was linear throughout the experiment (45 and 60 minutes, respectively), and the degree of PLD activation by PIP2, as calculated from the ratio of the slopes of the two lines, was 34. To determine the kinetic parameters of PLD activation by PIP2 in the mixed Triton X-100NBD-PC micellar system, PIP2 at increasing molar ratios was introduced to the micelles and PA production was monitored under standard reaction conditions (Fig. 4, top panel). Fitting the results to the Michaelis-Menten equation revealed maximal PLD activation already at a PIP2 2

Lower micellar content values were not examined, since then Triton X-100 concentrations approached its CMC, and the amount of Triton X-100 molecules that partition to the micelles could not be estimated.

593

molar ratio of 0.01, and JC,,, between 0.001-0.002. Fitting the data to the Hill equation (not shown) gave a Hill coefficient around 1, suggesting lack of a cooperative interaction. When PLD activation by PIP2 was examined in micelles that were composed only of NBD-PC and PIP2 (keeping Triton X-100 below its CMC), saturation was also attained around PIP2 molar ratio of 0.01, and similar "Michaelis-Menten" constants and a Hill coefficient close to 1 were observed (Fig. 4, middle panel), again excluding the possibility of a cooperative interaction with PIP2. Activation of membrane-bound PLD by PIP2 was examined next. PLD activity was plotted against PIP2 molar ratio (taking into account endogenous brain membrane phospholipids), and the measurements were fitted to Michaelis-Menten equation (Fig. 4, bottom panel). PLD activation reached a maximum level at a PIP2 molar ratio of 0.02, and Kact (0.0039) was comparable to the previous cases. Although this experiment was performed in a reaction system whose physical aggregation state was undefined, the activation of PLD by PIP2 had similar kinetic characteristics as in the more defined, mixed micellar system. It should be noted that in all three cases, PLD activation level decreased at higher PIP2 molar ratios (data not shown). Effect of PIP 2 on the Kinetic Properties of PLD Towards NBD-PC. In order to investigate how PIP2 influences the kinetic properties of PLD towards its phospholipid substrate, partially purified enzyme was incubated with increasing surface concentrations of NBD-PC in the presence and absence of PIP2. In both cases, when PA production was monitored, PLD activity fitted a straight line (Fig. 5). Such a behavior may represent the first part of a Michaelis-Menten curve in which substrate concentration is much lower than the Km and the Michaelis-Menten equation is reduced to the form V = (V max /K m [A]. Therefore, only the effect of PIP2 on the specificity constant KA (V max /K m = KA[E0]) towards NBD-PC could be determined. In three separate experiments in which PA production was monitored, PIP2 elevated the KA value of PLD towards NBD-PC by 6- to 22-fold; however, whether PIP2 affected the Vmax or Km of PLD cannot be determined from these data. The Kinetics of PLD Transphosphatidylation Activity Towards Primary Alcohols. To study the kinetic behavior of PLD transphosphatidylation activity towards primary alcohols under the standard reaction conditions of the mixed micellar system, partially purified PLD was incubated with increasing concentrations of either ethanol, 1-propanol or 1-butanol and the production of PA as well as of the respective phosphatidylalcohol was determined. As seen in Fig. 6, the increase in phosphatidylalcohol production with increasing alcohol

594

Chalifa-Caspi, Eli, and Liscovitch concentrations was accompanied by a decrease in production of PA, indicating that water and alcohol compete for PLD as phosphatidyl group acceptor substrates. At propanol and butanol concentrations above 400 and 100 mM, respectively, production of phosphatidylalcohol started to decrease (Fig. 6B,C). This seems to result from an inhibitory effect of the alcohol on PLD activity. The results suggest that longer-chain alcohols are more effective inhibitors of PLD activity. This mode of inhibition by alcohols appeared also in membrane-bound PLD (13) and cholate-solubilized PLD (31). It is possible that the inhibitory effect of the alcohols is influenced by their partitioning to the hydrophobic phase of the micelles or vesicles, which is higher the longer the alcohol chain is. To determine the kinetic parameters of PLD activity towards the individual alcohols, phosphatidylalcohol formation at increasing alcohol concentrations (before onset of inhibition) was plotted, and the data was fitted to the Michaelis-Menten equation. The obtained kinetic parameters (Vmax and apparent Km are shown in Table II. It appears that the calculated maximal rates of PLD activity (Vmax) were approximately the same with all three alcohols, while the enzyme affinity increased (Km decreased) as a function of the alcohol's aliphatic chain length. Analysis of the Competition Between Water and Alcohols as PLD Substrates. When two competing substrates, A and A', are simultaneously present in the reaction mixture, the ratio between the formation rates (V, V) of their respective products is (Ref. 33):

Fig. 4. Dependence of PLD activity on PIP2 in Triton X-100 mixed micelles, NBD-PC micelles and brain membrane vesicles. Top panel, Triton X-100 micelles: Reaction mixtures contained 0.3 mM NBD-PC (molar ratio 0.15) and variable amounts of Triton X-100 and PIP2 calculated to give the indicated PIP2 concentrations. Total mixed micellar content was 2 mM. Middle panel, NBD-PC micelles: Reaction mixtures contained 0.3 mM NBD-PC, 0.008% (0.12 mM) Triton X100 and the indicated concentrations of PIP2. Bottom panel, brain membranes: Reaction mixtures contained 0.3 mM NBD-PC, 1 mM MgCl2, total brain membranes at 66 u,g protein/ml (containing approximately 100 uM endogenous phospholipids) and the indicated concentrations of PIP2. Incubation time was 10 min. In all panels the data were fitted to the Michaelis-Menten equation from which the indicated parameters were derived.

This means that the ability of an enzyme to distinguish between two competing substrates, i.e. its specificity, is determined by their concentrations and the ratio of their specificity constants (KA values). Thus, although all of the individual kinetic parameters for water and for the alcohol present in the reaction mixture cannot be determined in this system, the ratio between the specificities of PLD towards each of them can be calculated. Furthermore, this allows comparison between specificities of PLD towards the different alcohols although they are not simultaneously present in the reaction mixture. The analysis was carried out as follows: The ratio between the specificity values towards water and towards alcohol (Kalcohol/Kwater) is:

Kinetic Analysis of Brain Phospholipase D in Mixed Micelles

595

Fig. 5. Dependence of PLD activity on NBD-PC molar ratio in the presence and absence of PIP2. Reaction mixture contained increasing molar ratios of NBD-PC, in a total mixed micellar content of 2 mM. Where indicated, PIP2 was included at a molar ratio of 0.005.

Since the experimental values for Vtransphosphatidylation and VhydroiysiS for each alcohol concentration are available, and considering a water concentration of 55.37 M in all reaction tubes, the specificity ratios were calculated to be hKethanol/Kwater = 260 ± 13; Kpropanol/Kwater = 939 ± 152; Kbutanol/Kwater = 5,920 ± 480. Hence, the specificity values of PLD to ethanol, propanol and butanol (respectively) were 260, 940 and 5,920 times higher than to water. The existence of such high ratios could be anticipated intuitively, since PLD produces significant amounts of phosphatidylalcohols in the presence of alcohols, despite the fact that the concentration of water in the reaction mixture is several orders of magnitude higher than that of the alcohols. From the obtained ratios, the proportion between PLD specificity values towards ethanol, propanol and butanol is 1:3.6:22.8. When this proportion is calculated directly from the apparent V max/Km values shown in Table II, a very similar result is achieved (1:3.4:19.4). Because PLD has similar maximal rates towards all the alcohols (see above), it may be concluded that this proportion is mainly due to differences in the affinity of PLD towards the alcohols. When the sum of PA and phosphatidylalcohol production was drawn as a function of alcohol concentration (Fig. 6A-C, dashed line), a nonlinear curve was observed. In other words, the sum of the hydrolysis and transphosphatidylation reactions was not constant. This finding is explained by the higher specificity constant of

Fig. 6. Dependence of PLD activity on primary alcohol concentration. Reaction mixtures contained 0.3 mM NBD-PC, 1.71 mM Triton X-100,40 uM PIP2 (molar ratio 0.02), and the indicated concentrations of either ethanol (A), 1-propanol (B) or 1-butanol (C). Total mixed micellar content was 2 mM. Panels A-C show production of PA (open circles) and phosphatidylalcohols (PEt, PPr and phosphatidylbutanol, respectively; solid circles) with each of the alcohols. The calculated sum of PA and phosphatidylalcohol amounts produced at each alcohol concentration is shown as a dashed line.

PLD towards alcohols than towards water (i.e., k0(alcohol)/Km(alcohol) > k0(Water)/Km(water)) ShOWH abOVB.

Chalifa-Caspi, Eli, and Liscovitch

596 Table II. The Kinetic Parameters of PLD Transphosphatidylation Reaction Towards Primary Alcohols Apparent Km

Alcohol

Vmax

FU

mM

Ethanol 1-Propanol 1-Butanol

211 178 228

290.8 74.6 16.2

The Vmax and apparent Km values were derived by fitting the MichaelisMenten equation to the experimental data shown in Fig. 6 (before onset of alcohol-induced inhibition).

DISCUSSION Basic Principles of Interfacial Catalysis. The classical Michaelis-Menten kinetic analysis deals with a reaction mixture in which all the molecules that participate in the enzymatic reaction (enzyme, substrates, inhibitors, activators) are homogeneously dispersed in an aqueous solution. In phospholipid hydrolysis reactions, however, the substrate (and sometimes the enzyme as well as other regulatory compounds) is embedded in the membrane bilayer, such that its lateral diffusion in the bilayer is equivalent to the spatial diffusion of water soluble compounds in an aqueous reaction mixture. Thus, if the rate of a reaction catalyzed by a phospholipase is measured as a function of the surface concentration of the phospholipid substrate, one would expect to observe a kinetic behavior that would fit the Michaelis-Menten equation. The "dual phospholipid model" (21) describes the kinetics of water-soluble phospholipases, where a step of enzyme adsorption to the membrane surface (either nonspecifically or via a specific interaction with a particular phospholipid) precedes the catalysis step. In this case, the total reaction rate will be influenced by the molar concentration of the membranes (or of the specific component in the membrane to which the enzyme binds) as well as by the dissociation constant of the enzyme from the membrane. The rate equation for product formation is the same as for a two substrate reaction (21,30). Mixed Lipid-Detergent Micellar Systems. A number of experimental systems have been applied for kinetic analysis of phospholipase-catalyzed reactions in vitro, e.g. monolayers (22), bilayer-vesicles (23,34) and mixed lipid-detergent micelles (21, 25-29). In a mixed lipiddetergent micellar system, a micelle-forming detergent is used to solubilize the hydrophobic and amphipatic components of the reaction mixture. If a fast exchange of components occurs between the micelles, then the total micelle population may mimic one large surface,

where the sum of the spherical surfaces of the micelles represents the planar surface(s) of the bilayer. When a soluble cytosolic enzyme is studied, its mode of interaction with the micelle surface is regarded as analogous to the interaction with the bilayer surface that the enzyme faces in vivo. In the case of a membrane-bound enzyme, the enzyme molecules are themselves solubilized by the detergent and thus embedded in mixed lipiddetergent micelles whose components are supposed to continuously exchange with the other mixed micelles present in the solution. Thus, the position of a detergentsolubilized enzyme either may be analogous to that of an enzyme embedded in (or persistently bound to) a membrane bilayer, or to that of a cytosolic enzyme, depending on its mode of interaction with micellar components. These two possibilities may be distinguished experimentally, because in the former case enzyme activity will depend only on the surface concentration of the substrate whereas in the latter, an additional dependence on the total micellar content should also be evident. Surface concentrations of the micelle-located elements are approximated by using their molar ratios, after subtraction of their CMC values (which represent an upper limit for the monomer concentration), when they are quantitatively significant. Dependence of PLD Activity on Substrate Surface Concentration in Mixed Micelles. The present paper describes a kinetic study in which PLD activity was characterized in mixed NBD-PC/Triton X-100 micelles. When the total micellar concentration ranged between 0.5-2 mM, PLD activity was dependent only on the surface concentration of NBD-PC and not on the total number of micelles, thus opening the way to interfacial kinetic analysis of the enzyme. A major limitation of the analysis was that we had to restrict ourselves to low surface concentrations of NBD-PC, in which the physical state of the micelles is not influenced by NBD-PC. At the NBD-PC molar ratios used, PLD activity continuously increased and was far from saturation. A linear dependence of PA production on NBD-PC molar ratio was observed also in the presence of activating concentrations of PIP2, and therefore only the effect of PIP2 on the KA towards NBD-PC could be determined. Similar results were obtained also with other enzymes, where the entire activity vs. substrate concentration curve could not be explored [e.g. phospholipase A2 activity towards Triton X-100/thio-PC mixed micelles (21); phospholipase C-P1 activity towards dodecyl maltoside/PIP2 mixed micelles (35); yeast phosphatidate phosphatase towards Triton X-100/phosphatidate mixed micelles (27)]. A linear dependence of PA production on NBD-PC molar ratio was observed also in the presence of acti-

Kinetic Analysis of Brain Phospholipase D in Mixed Micelles vating concentrations of PIP2, and therefore only the effect of PIP2 on the KA towards NBD-PC could be determined. PIP2 elevated the KA value of PLD towards NBD-PC by 6 to 22 fold in three separate experiments. Dual Dependence of PLD Activity on Both the Molar and Surface Substrate Concentration in NBD-PC Micelles. If PLD activity in micellar systems depends only on the surface concentration of NBD-PC, it could have been expected that in a reaction mixture containing pure NBD-PC micelles (where substrate surface concentration is always 100%), PLD activity will be constant at any molar NBD-PC concentration applied. Since Triton X100 is always present in the PLD preparation, the examination of this question was carried out while keeping Triton X-100 below its CMC. We have found that PLD activity was dependent on the molar concentration of NBD-PC, showing a classical "Michaelis-Menten" behavior with a half maximal velocity at 0.02 mM NBD-PC, and reaching saturation at 0.1 mM NBD-PC. There are several possible explanations for this behavior at the molecular level, depending on the mode of interaction of PLD with the substrate. PLD may act on NBD-PC molecules that are associated with the protein regions that were originally embedded in the membrane bilayer. Assuming fast exchange of NBD-PC molecules between protein-bound and protein-free micelles, the enzyme constantly sees the same averaged environment, and its rate of activity depends on the molar ratio of NBD-PC in the micellar phase. Alternatively, PLD does not interact with NBD-PC molecules bound to it, but with NBD-PC present in other micelles, where its activity depends on NBD-PC surface concentration. In case (i), PLD activity is expected to be independent of NBDPC molar concentration as long as the micellar phase is composed of 100% NBD-PC. The "contradictory" finding of dual dependence of PLD activity on both molar and surface substrate concentration may be explained by nonexistence of the fast exchange assumption at low numbers of NBD-PC micelles; alternatively, it may be due to the fact that at low NBD-PC concentrations the amount of Triton X-100 molecules that do partition into the micelles is no longer negligible and thus have a surface diluting effect. In case (ii), PLD acts in a manner analogous to that described for cytosolic phospholipase A2 in the "dual phospholipid model" (21). According to this model, there is an initial micelle-association step whose rate depends on the molar concentration of the micellar component to which the enzyme binds. This step precedes the catalytic step, whose rate in turn depends on the substrate surface concentration in the micelles. Our results thus fit case (ii) without the need to accept nonexistence of the assumptions underlying the

597

mixed micelles/interfacial kinetics model, and the obtained molar Km (or S0.5) values represent the affinity of PLD for the initial interaction with the NBD-PC micellar surface. In practice, though, if measurements are conducted under conditions in which PLD activity is independent on micelles number (i.e., above 0.1 mM NBD-PC) then the kinetic analysis of the interfacial catalytic step will be the same whatever case (i or ii) is the one occurring in the reaction system. Activation of PLD by PIP2. As a phospholipid, PIP2 was presumed to interact with PLD at the micellar phase. Indeed, we showed that PLD activity depended on the surface concentration of PIP2, and that activating concentrations of PIP2 were very low. Accordingly, a complete Michaelis-Menten type curve could be achieved for PLD activation by PIP2. Kact values were between 0.0010.002 (molar ratio units), and maximal activation was observed at a PIP2 molar ratio of 0.01. Similar values were obtained when the effect of PIP2 on PLD activity was determined utilizing NBD-PC-dominated micelles, and when activity of membrane-bound PLD was determined in the presence of a mixture of brain membrane vesicles and NBD-PC. This suggests that the characteristics of PLD activation by PIP2 have a general relevance. Moreover, the activating concentrations of PIP2 found here correspond to its physiological concentration which in erythrocyte membranes is ~1% of total phospholipids (36). Assuming that PIP2 is asymmetrically distributed and is present predominantly in the inner leaflet of the membrane (37), the surface concentration of PIP2 in that leaflet will approach a molar ratio of 0.02. PLD Transphosphatidylation Activity Towards Primary Alcohols. The specificity of mammalian PLDs for their alcohol substrates have received relatively little attention. Ethanol, 1-propanol and 1-butanol have been used to monitor the transphosphatidylation activity of PLD in intact cells (38-40). Secondary alcohols are not substrates for plant PLD and this seems also to be the case for mammalian PLD (41). Among the primary alcohols examined, methanol was the best substrate for synaptic plasma membrane PLD, and as the length of the alcohol was longer, less alcohol was formed (13). In contrast, in HL-60 cells, the optimal length of alcohol was 3 carbons (1-propanol) (42), whereas in lymphocytes, the optimal length was 2 carbons (ethanol) (43). This difference may be explained by the assumption that the availability of the various alcohols to PLD in intact cells is influenced by additional factor(s), e.g. their solubility in the lipid bilayer. Such factor(s) would affect the rate of penetration of alcohols through the plasma membrane, but could also affect their accessibility to the enzyme active site.

Chalifa-Caspi, Eli, and Liscovitch

598 In the present study we aimed to explore the dependence of PLD transphosphatidylation activity on the alcohols in the more defined, mixed micellar system. Since alcohols are water soluble, we assumed that they interact with PLD at the aqueous phase. Thus, a classical Michaelis-Menten kinetic analysis was performed, taking into account the molar concentration of the alcohols. The calculated maximal rates of PLD activity (Vmax) were approximately the same with all alcohols examined, while the enzyme affinity increased as the alcohol was longer. Thus, ethanol, 1-propanol and 1-butanol revealed apparent Km values of 291, 75 and 16 mM, respectively. The Km for ethanol is similar to apparent Km values for ethanol found in human amniotic microsomes (270 mM) (44) and in rat brain synaptosomes (200 mM) (45). The apparent Km is high compared with alcohol levels that can be reached in human alcoholics. The blood alcohol concentration which causes euphoria and beginning of sensory-motor impairment is 8.5 mM and a concentration of above 100 mM is lethal. This range fits the initial range of the PLD dose response curve to ethanol, in which phosphatidylethanol production is linearly dependent on ethanol concentration (phosphatidylethanol accumulation has been observed already at 5 mM ethanol). The dependence of PLD activity on methanol concentration was difficult to assess, since phosphatidylmethanol was hardly separated from phosphatidic acid on TLC. Alcohols longer than 1-butanol were not examined due to their low solubility in the reaction mixture. Competition Between the Hydrolysis and Transphosphatidylation Reactions. In the presence of alcohols PLD produces significant amounts of phosphatidylalcohols, despite the presence of several orders of magnitude higher concentration of water. By analyzing the transphosphatidylation as a competitive reaction to the hydrolysis reaction, and considering alcohol and water as competing substrates, we could quantify the ratio of PLD specificity values (KA) to water and to alcohol. PLD was shown to be 260, 940 and 5,920 times more specific to ethanol, 1-propanol and 1-butanol (respectively) than to water. The higher specificity of PLD to alcohols than to water explains the observation that the sum of phosphatidylalcohol and PA production with increasing alcohol concentrations was not constant. It can be demonstrated mathematically that a constant sum of the transphosphatidylation and hydrolysis products (i.e. that the sum of phosphatidylalcohol and PA will be equal to PA production in the absence of alcohol), could be expected if the kinetic constants of PLD towards the competing substrates water and alcohol were equal, i.e. that k0(water) = k0(alcohol) and Km(waler) = Km(alcohol).

PLD Inhibition by Alcohols. At high concentrations, propanol and butanol exhibited an inhibitory effect on both the transphosphatidylation and hydrolysis activities of PLD. Longer chain alcohols inhibited PLD activity at a lower alcohol concentration. This may be related to the ability of alcohols to partition into the hydrophobic phase of the reaction mixture. The same result was obtained with cholate-solubilized PLD (31). Ethanol did not inhibit the detergent-solubilized PLD at concentrations up to 1 M (Ref. 31 and the present work), but it inhibited membrane bound PLD [from rat brain synaptosomes (13,45), rat liver (46) and human amnion (44)] at concentrations above 400 mM. It remains to be established whether the alcohols inhibit PLD activity by acting as inhibitors at the active site, or by causing a non specific change in the physicochemical properties of the lipophilic surface.

ACKNOWLEDGMENTS This work was supported by a grant from the Israel Science Foundation. ML. is the incumbent of the Harold L. Korda Professorial Chair in Biology.

REFERENCES 1. Liscovitch, M., and Chalifa-Caspi, V. 1996. Enzymology of mammalian phospholipases D: in vitro studies. Chem. Phys. Lipids 80: 37^4 2. Billah, M. M. 1993. Phospholipase D and cell signaling. Curr. Opin. Immunol. 5:114-123 3. Klein, J., Chalifa, V., Liscovitch, M., and Loffelholz, K. 1995. Role of phospholipase D activation in nervous system physiology and pathophysiology. J. Neurochem. 65:1445-1455 4. Kiss, Z. 1996. Regulation of phospholipase D by protein kinase C. Chem. Phys. Lipids 80:81-102 5. Natarajan, V., Scribner, W. M., and Vepa, S. 1996. Regulation of phospholipase D by tyrosine kinases. Chem. Phys. Lipids 80:103116 6. Brindley, D. N., and Waggoner, D. W. 1996. Phosphatidate phosphohydrolase and signal transduction. Chem. Phys. Lipids 80:4557 7. Wang, X., Xu, L., and Zheng, L. 1994. Cloning and expression of phosphatidylcholine-hydrolyzing phospholipase D from Ricinus communis L. J. Biol. Chem. 269:20312-20317 8. Hammond, S. M., Altshuller, Y. M., Sung, T.-C., Rudge, S. A., Rose, K., Engebrecht, J., Morris, A. J., and Frohman, M. A. 1995. Human ADP-ribosylation factor-activated phosphatidylcholinespecific phospholipase D defines a new and highly conserved gene family. J. Biol. Chem. 270:29640-29643 9. Colley, W. C., Sung, T.-C., Roll, R., Jenco, J., Hammond, S. M., Altshuller, Y., Bar-Sagi, D., Morris, A. J., and Frohmann, M. A. 1997. Phospholipase D2, a distinct phospholipase D isoform with novel regulatory properties that provoke cytoskeletal reorganization. Curr. Biol. 7:191-201 10. Rose, K., Rudge, S. A., Frohman, M. A., Morris, A. J., and Engebrecht, J. 1995. Phospholipase D siganling is essential for meiosis. Proc. Natl. Acad. Sci. U.S.A. 92:12151-12155

Kinetic Analysis of Brain Phospholipase D in Mixed Micelles 11. Waksman, M., Eli, Y., Liscovitch, M., and Gerst, J. E. 1996. Identification and characterization of a gene encoding phospholipase D activity in yeast. J. Biol. Chem. 271:2361-2364 12. Ella, K., Dolan, J. W., Qi, C, and Meier, K. E. 1996. Characterization of Saccharomyces cerevisiae deficient in expression of phospholipase D. Biochem. J. 314:15-19 13. Chalifa, V., Mohn, H., and Liscovitch, M. 1990. A neutral phospholipase D activity from rat brain synaptic plasma membranes. Identification and partial characterization. J. Biol. Chem. 265: 17512-17519 14. Mohn, H., Chalifa, V., and Liscovitch, M. 1992. Substrate specificity of neutral phospholipase D from rat brain studied by selective labeling of endogenous synaptic membrane phospholipids in vitro. J. Biol. Chem. 267:11131-11136 15. Brown, H. A., Gutowski, S., Moomaw, C. R., Slaughter, C., and Sternweis, P. C. 1993. ADP-ribosylation factor, a small GTP-dependent regulatory protein, stimulates phospholipase D activity. Cell 75:1137-1144 16. Liscovitch, M., Chalifa, V., Fertile, P., Chen, C.-S., and Cantley, L. C. 1994. Novel function of phosphatidylinositol 4,5-bisphosphate as a cofactor for brain membrane phospholipase D. J. Biol. Chem. 269:21403-21406 17. Pertile, P., Liscovitch, M., Chalifa, V., and Cantley, L. C. 1995. Phosphatidylinositol 4,5-bisphosphate synthesis is required for activation of phospholipase D in U937 cells. J. Biol. Chem. 270: 5130-5135 18. Yokozeki, T., Kuribara, H., Katada, T., Touhara, K., and Kanaho, Y. 1996. Partially purified RhoA-stimulated phospholipase D activity specifically binds to phosphatidylinositol 4,5-bisphosphate. J. Neurochem. 66:1234-1239 19. Liscovitch, M., Chalifa, V., Danin, M., and Eli, Y. 1991. Inhibition of neural phospholipase D activity by aminoglycoside antibiotics. Biochem. J. 279:319-321 20. Deems, R. A., Eaton, B. R., and Dennis, E. A. 1975. Kinetic analysis of phospholipase A2 activity toward mixed micelles and its implications for the study of lipolytic enzymes. J. Biol. Chem. 250:9013-9020 21. Hendrickson, H. S., and Dennis, E. A. 1984. Kinetic analysis of the dual phospholipid model for phospholipase A2 action. J. Biol. Chem. 259:5734-5739 22. Ransac, S., Riviere, C., Soulie, J. M., Gancet, C., Verger, R., and de Haas, G. H. 1990. Competitive inhibition of lipolytic enzymes. I. A kinetic model applicable to water-insoluble competitive inhibitors. Biochim. Biophys. Acta 1043:57-66 23. Berg, 0. G., Yu, B., Rogers, J., and Jain, M. K. 1991. Interfacial catalysis by phospholipase A2: Determination of the interfacial kinetic rate constants. Biochemistry 30:7283-7297 24. Hannun, Y. A., Loomis, C. R., and Bell, R. M. 1985. Activation of protein kinase C by Triton X-100 mixed micelles containing diacylyglycerol and phosphatidylserine. J. Biol. Chem. 260: 10039-10043 25. Jones, G. A., and Carpenter, G. 1993. The regulation of phospholipase C-"y by phosphatidic acid. Assessment of kinetic parameters. J. Biol. Chem. 268:20845-20850 26. Walsh, J. P., and Bell, R. M. 1986. sn- 1,2-Diacylglycerol kinase of Escherichia coli. Mixed micellar analysis of the phospholipid cofactor requirement and divalent cation dependence. J. Biol. Chem. 261:6239-6247 27. Lin, Y., and Carman, G. M. 1990. Kinetic analysis of yeast phosphatidate phosphatase toward Triton X-100/phosphatidate mixed micelles. J. Biol. Chem. 265:166-170 28. Ridgway, R. D., and Vance, D. E. 1988. Kinetic mechanism of phosphatidylethanolamine N-methyltransferase. J. Biol. Chem. 263:16864-16871

599

29. Bae-Lee, M., and Carman, G. M. 1990. Regulation of yeast phosphatidylserine synthase and phosphatidylinositol synthase activities by phospholipids in Triton X-100 /phospholipids mixed micelles. J. Biol. Chem. 265:7221-7226 30. Carman, G. M., Deems, R. A., and Dennis, E. A. 1995. Lipid signaling enzymes and surface dilution kinetics. J. Biol. Chem. 270:18711-18714 31. Danin, M., Chalifa, V., Mohn, H., Schmidt, U. S., and Liscovitch, M. 1993. Rat brain membrane-bound phospholipase D. In: Fain J.N. (eds), Lipid Metabolism in Signaling Systems, Vol. 18, Academic Press, San Diego, pp. 14-24. 32. Markwell, M. A., Haas, S. M., Bieber, L., and Tolbert, L. L. 1978. A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal. Biochem. 87:206-211 33. Cornish-Bowden, A., and Wharton, C. W. 1988. Enzyme Kinetics. IRL Press, Oxford. 34. Diez, E., Louis-Flamberg, P., Hall, R. H., and Mayer, R. J. 1992. Substrate specificity and properties of human phospholipase A2 in a mixed vesicle model. J. Biol. Chem. 267:18342-18348 35. Cifuentes, M. E., Honkanen, L., and Rebecchi, M. J. 1993. proteolytic fragments of phosphoinositide-specific phospholipase C5,. Catalytic and membrane-binding properties. J. Biol. Chem. 268:11586-11593 36. Sulpice, J.-C., Zachowski, A., Devaux, P. F., and Giraud, F. 1994. Requirement for phosphatidylinositol 4,5-bisphosphate in the Ca2+-induced phospholipid redistribution in the human erythrocyte membrane. J. Biol. Chem. 269:6347-6354 37. Zachowski, A. 1993. Phospholipids in eukaryotic membranes: Transverse assymetry and movement. Biochem. J. 294:1-14 38. Moehren, G., Gustavsson, L., and Hoek, J. B. 1994. Activation and desensitization of phospholipase D in intact rat hepatocytes. J. Biol. Chem. 269:838-848. 39. Ben-Av, P., Eli, Y., Schmidt, U. S., Tobias, K., and Liscovitch, M. 1993. Distinct mechanisms of phospholipase D activation and attenuation utilized by different mitogens in NIH-3T3 fibroblasts. Eur. J. Biochem. 215:455-463. 40. Kaszkin, M., Seidler, L., Kast, R., and Kinzel, V. 1992. Epidermal growth factor-induced production of phosphatidylalcohols by HeLa cells and A431 cells through activation of phospholipase D. Biochem. J. 287:51-57. 41. Liscovitch, M., and Eli, Y. 1991. Ca2+ inhibits guanine nucleotideactivated phospholipase D in neural-derived NG108-15 cells. Cell Regul. 2:1011-1019. 42. Tettenborn, C. S., and Mueller, G. C. 1987. Phorbol esters activate the pathway for phosphatidylethanoi synthesis in differentiating HL-60 cells. Biochim. Biophys. Acta 931:242-250. 43. Pai, J., Liebl, E. C., Tettenborn, C. S., Ikegwuonu, F. I., and Mueller, G. C. 1987. 12-O-tetradecanoylphorbol-13-acetate activates the synthesis of phosphatidylethanoi in animal cells exposed to ethanol. Carcinogenesis 8:173-178. 44. Inamori, K., Sagawa, N., Hasegawa, M., Itoh, H., Ueda, H., Kobayashi, F., Ihara, Y., and Mori, T. 1993. Identification and partial characterization of phospholipase D in the human amniotic membrane. Biochem. Biophys. Res. Commun. 191:1270-1277. 45. Kobayashi, M., and Kanfer, J. N. 1987. Phosphatidylethanoi formation via transphosphatidylation by rat brain synaptosomal phospholipase D. J. Neurochem. 48:1597-1603. 46. Bocckino, S. B., Wilson, P. B., and Exton, J. H. 1987. Ca2+-mobilizing hormones elicit phosphatidylethanoi accumulation via phospholipase D activation. FEBS Lett. 225:201-204.