Kinetic Analysis of Arabidopsis Phospholipase Dδ

Phospholipase D (PLD) is a major plant phospholipase family involved in many cellular processes such as signal transduction, membrane remodeling, and lipid degradation. Five classes of PLDs have been identified in Arabidopsis thaliana, and Ca2+ and polyphosphoinositides have been suggested as key regulators for these enzymes. To investigate the catalysis and regulation mechanism of individual PLDs, surface-dilution kinetics studies were carried out on the newly identified PLDδ fromArabidopsis. PLDδ activity was dependent on both bulk concentration and surface concentration of substrate phospholipids in the Triton X-100/phospholipid mixed micelles.V max, K s A , andK m B values for PLDδ toward phosphatidylcholine or phosphatidylethanolamine were determined; phosphatidylethanolamine was the preferred substrate. PLDδ activity was stimulated greatly by phosphatidylinositol 4,5-bisphosphate (PIP2). Maximal activation was observed at a PIP2 molar ratio around 0.01. Kinetic analysis indicates that PIP2 activates PLD by promoting substrate binding to the enzyme, without altering the bulk binding of the enzyme to the micelle surface. Ca2+ is required for PLDδ activity, and it significantly decreased the interfacial Michaelis constantK m B . This indicates that Ca2+activates PLD by promoting the binding of phospholipid substrate to the catalytic site of the enzyme.

Expression and Purification of the PLDs-Full-length cDNAs of Arabidopsis PLD␦ and PLD␤ were cloned into pGEX-4T-1 and pGEX-2T-1 vectors (Amersham Biosciences) (19,22), respectively, which produced a glutathione S-transferase (GST) fusion at the N terminus. The recombinant plasmids were transformed into Escherichia coli JM109 and then into BL21 (Promega). All cell cultures were grown in Luria-Bertani medium with 50 mg/liter ampicillin. The transformed cells were grown at 37°C to an absorbance of ϳ1.0 at 600 nm. Five milliliters of the cells was transferred into a 1-liter flask with 200 ml of LB medium containing 0.1 mM isopropyl-1-thio-␤-galactopyranoside and incubated overnight at room temperature. Then the cells were harvested and lysed by sonication in phosphate-buffered saline buffer containing 140 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , and 1.8 mM KH 2 PO 4 (pH 7.3), with addition of 5 mM dithiothreitol and 0.1 mM phenylmethylsulfonyl fluoride.
The GST fusion proteins were purified with glutathione-Sepharose 4B beads (Amersham Biosciences), following the manufacturer's instructions. The GST tag was removed by thrombin digestion at room temperature overnight with gentle shaking. The protein purity was checked on SDS-PAGE gel stained with Coomassie Blue, and protein concentration was estimated by using a dye-binding protein assay kit (Bio-Rad) with bovine serum albumin as a standard. Purified protein (0.1-0.5 g) was used per reaction, depending on different enzymes and substrates, so that less than 5% of the substrate was hydrolyzed.
Preparation of Triton X-100/Phospholipid Mixed Micelles-Two Ci of [ 3 H]PC or 0.5 Ci of [ 14 C]PE was added to 1 mol of cold PC or 1 mol of cold PE in chloroform, respectively. The chloroform solutions were dried under a stream of nitrogen, and de-ionized water was added to give a final phospholipid concentration of 1 mM. The phospholipid suspension was probe-sonicated on ice until clear. To obtain the desired substrate concentration at the desired mole fraction, the phospholipid solution was diluted with Triton X-100 stock solution (40 mM) using the following formula: mole fraction phospholipid ϭ [phospholipid]/([phospholipid] ϩ [Triton X-100 (free)]); [Triton X-100 (free)] ϭ [Triton X-100 (total)] Ϫ critical micelle concentration of Triton X-100 (0.24 mM) (23). When the effect of PIP 2 mole concentration was to be tested, PIP 2 was added at this point, using the following formula: mole fraction PIP2 ϭ [PIP 2 ]/([total phospholipid] ϩ [Triton X-100 (free)]). The Triton X-100/ phospholipid mixture was vortexed briefly and let stand at room temperature for half an hour. The total phospholipid concentration in Triton X-100/phospholipid mixed micelles did not exceed 25 mol % to ensure that the structure of the mixed micelles was similar to the structure of pure Triton X-100.
PLD Assays-PLD activity was measured by following the release of [ 3 H]choline from dipalmitoylglycero-3-phospho-[methyl-3 H]choline or [ 14 C]ethanolamine from dioleoyl-glycero-3-P-[ethanolamine-2-14 C]ethanolamine. The basic assay mixture contained 100 mM Tris-HCl (pH 7.0), 80 mM KCl, 200 M CaCl 2 , 0.5 mM MgCl 2 , the indicated concentrations of Triton X-100/phospholipid vesicles, and purified PLD protein, in a total volume of 100 l. The reactions were initiated by adding enzyme proteins and were stopped by adding 1 ml of chloroform/methanol (2:1) and then 100 l of 2 M KCl. The release of the radiolabeled head groups into the aqueous phase was quantitated by scintillation counting. All assays were conducted for 15 min at 30°C. The enzyme reactions were linear with time and protein concentration.

Kinetic Analysis of PLD␦ toward PC in Triton X-100/PC
Micelles-To study the kinetic behavior of PLD␦, we used a detergent/phospholipid mixed micelle assay system, following the surface dilution model of enzyme kinetics (10 -13, 24). This model takes into account both two-dimensional surface interaction and three-dimensional bulk interaction between an enzyme and lipid substrate. Such consideration is critical in determining the kinetic parameters of these enzymes because lipid-dependent enzymes, such as PLD, catalyze reactions at a water-lipid interface (10). The principle of the surface dilution model is presented in Equation 1 (24).

Bulk step
Surface step According to this model, the action of enzyme (E) consists of two consecutive steps. First, the enzyme interacts non-catalyt- ically with the surface of detergent/lipid mixed micelles (A). Subsequently, the enzyme-mixed micelle complex (EA) binds to an individual lipid substrate (B) at its catalytic site, which leads to the hydrolysis of substrate to product (Q). The first association depends on the bulk concentrations of both E and A, whereas the second step depends on the surface concentrations of EA and B. Here A represents the sum of the molar concentrations of detergent and lipid substrate, and B is the mole fraction of lipid substrate in the mixed micelle. Equation 2 is the rate expression for surface dilution kinetic model (10,12,24).
Three kinetic parameters, V max , K s A , and K m B can be determined from this equation. V max is the true V max when both the bulk concentration and the surface concentration of the lipid substrate approach infinity. K s A , which is equal to k Ϫ1 /k 1 and expressed in bulk concentration terms, is the dissociation con-stant describing the interaction of the enzyme with the mixed micelles in the first binding step. K m B , equal to (k Ϫ2 ϩ k 3 )/k 2 , defines the interfacial Michaelis constant for the second binding step and is expressed in surface concentration units, such as mole fraction.
Triton X-100 is one of the most frequently employed detergents for surface dilution kinetics. It has been shown to form uniform mixed micelles with a variety of lipid molecules including PC (23,25), PE (26), phosphatidylinositol (27), phosphatidate (16), and phosphatidylserine (28). By using purified Arabidopsis PLD␦ (Fig. 1A), the data show that Triton X-100 serves as a neutral dilutor at the concentration range (0.5-2.5 mM) employed in this study. Arabidopsis PLD␦ was inactive toward vesicles composed of only PC or only PE, and it became active when Triton X-100 was included. When the molar concentration of substrate phospholipid was held at 0.1 mM, the optimal activity was achieved at around 0.5 mM Triton X-100 (Fig. 1B). Afterward, PLD␦ activity exhibited typical surface dilution effects in the presence of increasing concentrations of Triton X-100 in Triton X-100/phospholipids mixed micelles   1B). The enzyme activity decreased when the surface concentration of the lipid substrate in the mixed micelle was diluted by the addition of detergent as the bulk concentration of the substrate was kept the same.
PLD␦ activity was determined as a function of the sum of the molar concentrations of Triton X-100 and PC (A of Equation 2) at a series of mole fractions of PC (B of Equation 2) (Fig. 2A). The activity was dependent on the sum of molar concentrations of Triton X-100 and PC at each surface concentration. As the surface concentration of PC in the mixed micelles decreased, the apparent V max decreased. A double-reciprocal plot of the results in Fig. 2A indicates that PLD␦ exhibits saturation kinetics when the bulk concentration of Triton X-100 and PC was varied at each fixed mole fraction of PC (Fig. 2B). According to Equation 2, the replot of the 1/V intercepts versus 1/B from Fig.  2B should be linear; the intercept of the 1/V intercept axis is equal to 1/V max , and the intercept of the 1/B axis is equal to Ϫ1/K m B . A replot of the data is linear (Fig. 2C), and the V max and K m B are 1.0 mol/min/mg and 0.13 mole fraction, respectively (Table I). Equation 2 also predicts that a replot of the slopes versus 1/B from Fig. 2B should be linear and cross the origin, and the slope of the replot is equal to K s A K m B /V max . Fig.  2D shows that such a replot was linear and passed the origin. By using the slope of the line in Fig. 2D and the values for V max and K m B obtained in Fig. 2C, the dissociation constant K s A for the Triton X-100/PC mixed micelle was calculated to be 1.5 mM (Table I).
PIP 2 Stimulates PLD␦ Activity by Promoting Substrate Binding-PIP 2 is an important co-factor and regulator of various PLD activities (1)(2)(3). It is required for Arabidopsis PLD␤ and -␥ activities but not for PLD␣ (6). In the above experiments, we have shown that PLD␦ did not require PIP 2 when assayed in the Triton X-100/PC mixed micelles. To test whether the previously reported PIP 2 -requiring PLDs are active toward the Triton X-100/PC mixed micelles, we measured and compared the activities of PLD␤ and -␦ as functions of PC mole fraction in the absence of PIP 2 (Fig. 3A). No activity was detected for PLD␤, showing that its PIP 2 requirement is not affected by the assay conditions. PLD␤ became active when PIP 2 was included in the detergent micelles (Fig. 3B). PIP 2 greatly stimulated PLD␦ activity (Fig. 3B). When PLD activity was measured as a function of PIP 2 mole fraction in the Triton X-100/PC/PIP 2 mixed micelles, the molar concentration of PC was set at 0.2 mM, and the PC mole fraction was set at 0.2. Dramatic increases of both PLD␤ and PLD␦ activities were observed when the PIP 2 mole fraction increased, with the maximal activation achieved at PIP 2 mole fraction around 0.01. The K act for PLD␤ and -␦ were similar, around 0.002 to 0.003 mole fraction of PIP 2 in the mixed micelles.
To study the possible mechanism by which PIP 2 stimulates the PLD␦ activity, we compared the two constants, K s A and K m B , in the absence or presence of PIP 2 . For determining the effect of PIP 2 on K s A , PLD␦ activity was measured as a function of the sum of the molar concentrations of Triton X-100 and PC at three set mole fractions of PIP 2 (0, 0.002, and 0.02) (Fig. 4A) with the PC mole fraction fixed at 0.1. Although increasing molar ratio of PIP 2 increased the apparent V max , it did not significantly change the apparent K s A (Fig. 4A). This suggests that PIP 2 does not affect the bulk binding of PLD␦ to the mixed micelles. For examining the effect of PIP 2 on K m B , PLD␦ activity was measured as a function of the mole fraction of PC at the same three set PIP 2 mole fractions (Fig. 4B). The apparent interfacial Michaelis constant, K m B , decreased about 2-3-fold in the presence of PIP 2 , and the apparent V max increased with the increase of PIP 2 mole fraction (Fig. 4B). Overall, the specificity constant (apparent V max /K m B ) increased up to 10-fold in the presence of 0.02 mole fraction of PIP 2 . These results suggest that PIP 2 may activate PLD␦ by increasing the affinity of its catalytic site to the phospholipid substrate but may not affect its binding to the Triton X-100/phoshpolipid mixed micelle surface.
Ca 2ϩ Activates PLD␦ by Decreasing the Interfacial Michaelis Constant K m B -Ca 2ϩ is another important regulator for plant PLD activity. The optimal Ca 2ϩ concentration for PLD␦ activity in the Triton X-100/PC mixed micelle system was around 1 mM, and this optimum was not affected by the presence of PIP 2 in the mixed micelles (Fig. 5A). To determine the effect of Ca 2ϩ on K m B , PLD␦ activity was measured as a function of the PC mole fraction in the mixed micelles at two Ca 2ϩ concentrations, 0.2 and 1 mM (Fig. 5B). Apparent K m B at 1 mM of Ca 2ϩ was nearly 3-fold lower than the apparent K m B at 0.2 mM of Ca 2ϩ (Fig. 5B). These results indicate that Ca 2ϩ may activate PLD␦ by promoting the binding of the phospholipid substrate to the catalytic site of the enzyme.
PE Serves as a Better Substrate for PLD␦ than PC-The substrate preference of PLD␦ was studied by substituting PC with PE in the Triton X-100 mixed micelles and measuring the activity of PLD␦ as a function of the sum of the molar concentrations of Triton X-100 plus PE (A) at a series of set mole fractions of PE (B). The PLD␦ activity was dependent both on the sum of the bulk concentration of Triton X-100 and PE and on the surface concentration of PE in the mixed micelles (Fig.  6A). As predicted by Equation 2, both the replot of 1/V intercepts (from Fig. 6B) versus 1/B and the replot of the slopes versus 1/B were linear, with the latter going through the origin (Fig. 6, C and D). By using the two intercepts obtained from Fig. 6C and the line slope from Fig. 6D, we calculated the V max , K s A , and K m B for PE to be 1.5 mol/min/mg, 3.5 mM, and 0.03 mole fraction, respectively (Table I).
The above results indicate that PLD␦ follows the surface dilution kinetic model using two common phospholipid substrates, PC and PE, but the catalytic constants of these two phospholipids are quite different (Table I). The V max of PE is about 1.5-fold higher than that of PC, whereas the interfacial Michaelis constant, K m B , of PE is about 4 -5-fold lower. Therefore, the specific constant (V max /K m B ) of PE is about 7-fold higher than that of PC, suggesting that PE has a higher affinity to the catalytic site of the enzyme. On the other hand, the dissociation constant, K s A , of PE is similar to or slightly higher than that of PC. This suggests that PLD␦ has similar affinities for binding sites to Triton X-100/PC and Triton X-100/PE mixed micelles.
To confirm that PE indeed serves as a better substrate for PLD␦, we carried out side-by-side comparison toward the two substrates. In this analysis, PLD␦ activity was measured as a function of the sum of the molar concentrations of Triton X-100 plus phospholipid, at a set phospholipid mole fraction of 0.1 (Fig. 7A). The apparent dissociation constants, K s A , of PC and PE were very similar, whereas the apparent V max of PE was about 3-fold higher than that of PC (Fig. 7A). In Fig. 7B, PLD␦ activity was measured as a function of the phospholipid mole fraction at a set phospholipid molar concentration of 0.1 mM. The apparent V max of PE was about 2-fold higher than that of PC, whereas the apparent interfacial Michaelis constant, K m B , of PE was much lower, around 4 -5-fold smaller than that of PC. Therefore, the specificity constant (apparent V max /K m B ) of PE was almost 9-fold higher than the specificity constant of PC. Consistent with the results shown in Fig. 2 and Fig. 6, this indicates that PE is a preferred substrate for PLD␦, and it may have higher affinity to the catalytic site of the enzyme than PC. On the other hand, similar dissociation constants of PC and PE may suggest that PLD␦ had similar affinities for Triton X-100/PC and Triton X-100/PE mixed micelles; hence the first binding step of PLD␦ to the mixed micelle surface was nonspecific.

DISCUSSION
The present study provides the first detailed kinetic description of a PLD on its interaction with lipids. The results help understand the mechanism behind the substrate selectivity and activity regulation of PLD. Activation of PLD in the cells is achieved by a number of regulators, which can vary from individual PLDs (1-3). Among those, PIP 2 is a universal activator of PLD. It is required by animal and yeast PLDs, and Arabidopsis PLD␤, -␥, and -. Though not required, it stimulates greatly the activity of PLD␣ and -␦. Mutagenesis results have identified specific amino acid regions on mammalian (29) and plant (22) PLDs that are involved in PIP 2 binding, but how PIP 2 affects PLD catalysis is unclear. A previous study on a partially purified rat brain PLD fraction showed maximal PIP 2 activation was observed at PIP 2 molar ratio of 0.01, and the K act was calculated to be between 0.001 and 0.002 (30), which are similar to the results obtained in this study (Fig. 3B). However, it was unknown which one of the kinetic constants was affected by PIP 2 .
Results from this study show that PIP 2 has no effect on K s A ( Fig. 4A) but significantly decreases K m B (Fig. 4B). These results suggest that PIP 2 promotes lipid substrate binding to the active site of the enzyme, without affecting the bulk binding of PLD␦ to the micelle surface. These results are consistent with the previous mutagenesis work showing that PIP 2 binding enhances PC binding to the catalytic site of PLD␤ (22,31). Those studies show that PIP 2 binds to PLD␤ at the C2 domain and the catalytic region, but Ca 2ϩ modulates the binding differentially. In the absence or at low levels of Ca 2ϩ , PLD␤ binds to PIP 2 at the C2 domain. Higher concentrations of Ca 2ϩ inhibit this binding but enhance the binding of PIP 2 to the catalytic region, which promotes PC binding and catalysis (31). One PIP 2 -binding site near the catalytic region has been identified in PLD␤ (22). It lies between the two HKDs and contains many basic side chains (Lys, Arg, and His), and these critical basic residues are conserved in PLD␦. This region is essential for PLD␤ activity, and the deletion of this region abolished the phospholipid stimulation effect on PC binding to the enzyme (22). Although PIP 2 binding itself is not required for PLD␦ activity, this conserved region may also be responsible for the PIP 2 stimulation effect observed for PLD␦. The different PIP 2 requirements suggest that other unique factor(s) may help activate PLD␦. Oleic acid has been shown to be a specific stimulator for PLD␦ but not for other Arabidopsis PLDs (19). The present data show that the oleic acid requirement by PLD␦ can be substituted by Triton X-100, consistent with the hypothesis that oleic acid may affect substrate presentation and modulating PLD-substrate interaction.
PLD␦, like most other plant PLDs, requires Ca 2ϩ for activity. The only exception for the cation requirement to date is Arabidopsis PLD1 that is Ca 2ϩ -independent and contains putative phox homology/pleckstrin homology domains (5). Instead, the majority of plant PLDs contains the Ca 2ϩ /phospholipidbinding C2 domain at their C terminus (5). The ability of the PLD C2 domain to bind Ca 2ϩ has been demonstrated in the binding and structural studies on Arabidopsis PLD␣ C2 and PLD␤ C2 (31). However, how Ca 2ϩ affects PLD catalysis is unknown. The present data show that Ca 2ϩ decreases the apparent interfacial Michaelis constant, K m B (Fig. 5B), suggesting that it promotes lipid substrate binding to the active site of the enzyme.
Another significant finding of this analysis is that PLD␦ prefers PE to PC as substrate and the kinetic description of this preference. This PE selectivity of PLD␦ is in contrast to the preference of many other PLDs for PC as substrate. For example, the cloned mammalian PLD1 and -2 and Arabidopsis PLD1 use only PC as substrate (3). The most prevalent Arabidopsis PLD, PLD␣, selectively hydrolyzes PC, as shown by lipid analysis in freezing-stressed Arabidopsis (9). Only Arabidopsis PLD␥1 was previously indicated to have a higher activity toward PE than PC, but the kinetics and the extent of preference are not known (7). The present results show that PLD␦ has similar dissociation constants, K s A , for the first binding step of PE and PC, but its specificity constant (V max /K m B ) for PE is about 7-9-fold higher than that of PC. Thus, it is conceivable that when PLD␦ is activated in the cell, its higher affinity to PE may result in selective hydrolysis of PE and PE-enriched membranes. As two major lipids in eukaryotic membranes, PC and PE differ greatly in their membraneforming properties and possibly cellular functions. Unlike PC, PE is a non-laminar lipid and prefers hexagonal phase formation. PE concentrations in vesicles and membranes have been shown to affect the activities of various proteins, including G-protein (32), phospholipase C (33), mammalian PLD (34), and Arabidopsis PLD␤ and -␥ (7). Thus, the cellular function and the mode of action of PLD␦ could be distinctively different from the PC-selective PLDs.