Chemistry and Physics of Lipids 164 (2011) 196–204
Contents lists available at ScienceDirect
Chemistry and Physics of Lipids journal homepage: www.elsevier.com/locate/chemphyslip
Discrimination between the regioisomeric 1,2- and 1,3-diacylglycerophosphocholines by phospholipases Johanna Mansfeld a,∗ , Wolfgang Brandt b , Regine Haftendorn a,1 , Regina Schöps a,2 , Renate Ulbrich-Hofmann a a b
Institute of Biochemistry and Biotechnology, Martin-Luther University Halle-Wittenberg, D-06120 Halle (Saale), Kurt-Mothes-Straße 3, Germany Department of Bioorganic Chemistry, Leibniz Institute of Plant Biochemistry, D-06120, Weinberg 3, Halle (Saale), Germany
a r t i c l e
i n f o
Article history: Received 22 October 2010 Received in revised form 18 December 2010 Accepted 22 December 2010 Available online 30 December 2010 Keywords: Secretory phospholipases A2 Phospholipase C Phospholipase D 1,3-Diacylglycero-2-phosphocholines
a b s t r a c t The artificial 1,3-diacyl-glycero-2-phosphocholines (1,3-PCs), which form similar aggregate structures as the naturally occurring 1,2-diacyl-sn-glycero-3-phosphocholines (1,2-PCs), were tested as substrates for different classes of phospholipases such as phospholipase A2 (PLA2 ) from porcine pancreas, bee and snake venom, and Arabidopsis thaliana, phospholipase C (PLC) from Bacillus cereus, and phospholipase D (PLD) from cabbage and Streptomyces species. The regioisomers of the natural phospholipids were shown to bind to all investigated phospholipases with an affinity similar to the corresponding naturally occurring phospholipids, however their hydrolysis was reduced to different degrees (PLA2 s and PLC) or even abolished (PLDs belonging to the PLD superfamily). The results are in accordance with binding models obtained by docking the substrates to the crystal structures or homology models of the phospholipases. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Phospholipases are ubiquitous enzymes that act on membrane phospholipids and play important roles in several signaling processes by generating mediators and second messengers for signal transduction, e.g. lysophospholipids, arachidonic acid, phosphatidic acid, inositol phosphates, and diacylglycerol. Depending on the site of attack, the enzymes are classified as phospholipases A, B, C or D. Phospholipases A (PLA) hydrolyze the acyl ester bond in the sn-1 (PLA1 ) or the sn-2 position (PLA2 ) of 1,2-diacylsn-glycero-3-phospholipids (1,2-PCs). Phospholipases C (PLCs) are phosphodiesterases that cleave the glycerophosphate bond, while
Abbreviations: atPLA2, secreted PLA2 from Arabidopsis thaliana; bvPLA2, PLA2 from Apis mellifera venom; 1,2-diC12PC, 1,2-dilauroyl-sn-glycero-3phosphocholine; 1,2-diC14PC, 1,2-dimyristoyl-sn-glycero-3-phosphocholine; 1,3-diC12PC, 1,2-dilauroyl-sn-glycero-3-phosphocholine; 1,3-diC14PC, 1,3dimyristoyl-sn-glycero-2-phosphocholine; NaDOC, sodium deoxycholate; nmPLA2, PLA2 from Naja mossambica venom; PC, phosphatidylcholine; 1,2-PC, 1,2-diacylsn-glycero-3-phosphocholine; 1,3-PC, 1,3-diacyl-glycero-2-phosphocholine; PLC, phospholipase C; PLA2, phospholipase A2, PLD, phospholipase D; ppPLA2, PLA2 from porcine pancreas. ∗ Corresponding author. Tel.: +49 345 5524865; fax: +49 345 5527303. E-mail address:
[email protected] (J. Mansfeld). 1 Present address: Genzyme GmbH, Siemensstr. 5b, D-63263 Neu-Isenburg, Germany. 2 Present address: Department of Chemistry, Martin-Luther-University HalleWittenberg, D-06120 Halle, Germany. 0009-3084/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.chemphyslip.2010.12.009
phospholipases D (PLDs) cleave the terminal phosphodiester ester bond (Fig. 1). Moreover, within one class of phospholipase, the enzymes from different sources possess distinct preferences for certain types of substrates and structures of the membrane surface (Ulbrich-Hofmann, 2000). These preferences can differ dramatically even in the case of isoenzymes with a high degree of sequence identity (Burke and Dennis, 2009; Dippe and Ulbrich-Hofmann, 2009). Presumably, the differences in substrate preferences coincide with the biological functions of the corresponding enzymes. For several reasons, such as the development of drugs, gene and drug delivery systems or the design of analytical tools, phospholipid analogs are important research objects. Thus, ether phospholipids or alkylphosphate esters are known to show anticancer effects (Wieder et al., 1999). Another group of noteworthy phospholipid analogs – the 1,3-diacylglycero-2-phospholipids such as the 1,3diacylglycero-2-phosphocholines (1,3-PCs; Fig. 2) – have hardly been investigated so far, although they promise highly interesting properties as analogs of the natural phospholipase substrates. One of the main reasons for this lack of information might be that these compounds are not commercially available and their chemical synthesis in regio-isomerically pure states is difficult. This problem, however, could be overcome by a chemoenzymatic strategy (Haftendorn and Ulbrich-Hofmann, 1995), which allows the synthesis of 1,3-PCs with acyl chain lengths of C8 –C18 (Haftendorn et al., 2000). As shown in our and other laboratories, the 1,2and 1,3-PCs are characterized by a similar aggregation behavior in aqueous (Haftendorn et al., 2000) as well as in organic
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
Fig. 1. Cleavage sites of phospholipases in 1,2-PC.
media (Frense et al., 1995). From NMR studies on bilayer structures of the regioisomers, differences in the bent conformation of the hydrocarbon regions with a higher disorder in the interiors of 1,3-PCs than in the corresponding 1,2-PCs were derived, whereas almost identical head group conformations were detected (Seelig et al., 1980). Infrared spectroscopy revealed small differences in the packing of the low-temperature lamellar phase (Dluhy et al., 1985). Only a few data are available on the acceptance of 1,3-PCs by phospholipases. Very early papers report that 1,3-PCs can be hydrolyzed by PLA from snake venom (de Haas and van Deenen, 1964) but activities or kinetic constants were not determined. The only comparison of a kinetic characterization of the hydrolysis of 1,3-PCs was performed for PLA2 from porcine pancreas (ppPLA2 ) with 1,3-PCs containing acyl moieties of different chain length (Slotboom et al., 1976). Similarly, the cleavability of 1,3-PCs by PLC from Bacillus cereus was mentioned (de Haas and van Deenen, 1965) but quantitative data are missing. In contrast, no cleavage of 1,3PCs by PLD (from Brassica oleracea) was observed. Instead, these compounds were shown to act as competitive inhibitors (Dittrich
197
et al., 1998; Haftendorn et al., 2000) and could be used as substrate mimics in PLD binding studies on monolayers (Kuppe et al., 2008). The lack of kinetic data on the hydrolysis of 1,3-PCs by PLA2 and PLC as well as the non-cleavability of these compounds by PLD prompted us to perform comparative kinetic experiments and docking studies, the results of which are presented in this paper. With respect to PLD, we show that the non-acceptance of 1,3-PCs is not restricted to plant PLD but also to microbial PLDs. Insofar as they belong to the PLD superfamily, the enzymes are unable to cleave these substrate analogs. Kinetic constants determined for the hydrolysis of 1,2- and 1,3-PCs with different chain lengths (C10 –C16 ) by PLC from B. cereus and ppPLA2 show larger differences for PLA2 than for PLC. Finally, the kinetics of the cleavage of 1,2- and 1,3-dilauroyl-PC (1,2-diC12 PC and 1,3-diC12 PC) has been compared for different secretory PLA2 s from several other sources including the recently described PLA2 from Arabidopsis thaliana (atPLA2 ). 2. Materials and methods 2.1. Enzymes ppPLA2 , PLA2 from honey bee venom (bvPLA2 ), PLA2 from Naja mossambica venom (nmPLA2 ), PLC from B. cereus, PLD from Streptomyces sp. and the enzyme from Streptomyces chromofuscus were supplied by Sigma (Taufkirchen, Germany). Production of atPLA2 in E. coli BL21(DE3) cells containing the atPLA2 gene in the vector pET-26b(+) and its purification to homogeneity were performed as described (Mansfeld et al., 2006). PLD used in this paper is PLD2 (␣type) from white cabbage and was prepared as described previously (Schäffner et al., 2002). 2.2. Substrates 1,3-PCs were synthesized as described in Haftendorn et al. (2000). 1,2-PCs were a gift of Lipoid GmbH (Ludwigshafen, Germany) or purchased from Sigma (Taufkirchen, Germany). 2.3. Other materials 1,4-Dithiothreitol (DTT), sodium deoxycholate, the Bradford protein assay reagent, and all buffer substances were from Sigma (Taufkirchen, Germany). The BCA protein assay kit was from Pierce (Bonn, Germany). The NEFA C Kit was obtained from WAKO Chemicals (Neuss, Germany). All other reagents were of the purest quality available. 2.4. Protein determination Protein concentrations were determined by the Bradford or the BCA protein assay with bovine serum albumin as standard according to the instructions of the suppliers. 2.5. Activity measurements of PLD
Fig. 2. Structural formula of 1,2-PC (A) and 1,3-PC (B).
The hydrolytic and transphosphatidylation activities of PLD were determined in a two-phase system according to Hirche et al. (1997). 570 L of diethylether (hydrolysis) or 560 L of diethyl ether containing 10 L of glycerol (transphosphatidylation) were placed in 1.5-mL vials. After addition of 80 L of buffer (100 mM sodium acetate buffer, pH 5.6, 40 mM CaCl2 in the case of PLD from cabbage and Streptomyces sp. or 100 mM Tris/HCl buffer, pH 8.0, 10 mM CaCl2 , in the case of the enzyme from S. chromofuscus) containing the corresponding enzyme, the reaction was started by injection of 50 L of 1,2- or 1,3-dimyristoyl-PC (1,2- or 1,3-diC14 PC; 1,48 mol) dissolved in dichloromethane. The conversions were performed on a horizontal shaker at 30 ◦ C and 300 min−1 . Aliquots
198
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
of the organic phase were withdrawn at different times and analyzed for the phospholipid content by quantitative HPTLC with densitometric evaluation using standards of the corresponding substrates and the expected products. The aqueous phase did not contain detectable amounts of phospholipids. All results are the means of duplicates of at least two independent experiments. 2.6. Determination of kinetic constants of PLC in aqueous and two-phase systems For the preparation of the substrate stock solutions, 50 mol of 1,2- or 1,3-PCs with acyl chain lengths of C10 –C16 were dissolved in 500 L of chloroform. The solvent was evaporated in vacuum, and the lipid film was dissolved in 10 mL of 5 mM sodium deoxycholate containing 100 mM NaCl by vortexing for 15 min. The solution was heated to 40 ◦ C in a water bath and dissolved by ultrasonic treatment with 300 Ws mL−1 using a 60 W Vibra Cell ultrasonic disintegrator (Avantec, France) with a microtip probe of 3 mm. The assay solution (1.5 mL) contained 50–1000 L of the substrate stock solution, 500 L of 2 mM Hepes buffer, pH 8.1, with 100 mM NaCl, and 950–0 L H2 O. After preincubation at 30 ◦ C, the pH of the reaction mixture was adjusted to 8.1. The reaction was started by addition of 5 L of enzyme (0.16 U) dissolved in a 1 mM solution of ZnSO4 . The amount of liberated product was determined by continuous titration (3–4 min) with 2 mM NaOH using the Autotitrator VIT 90 (Radiometer, Denmark). The corresponding initial reaction rates were determined from the slopes of the titration curves at substrate concentrations between 0.17 and 5 mM, and were expressed as mol of liberated phosphocholine per minute and mg protein. KM,app and Vmax,app values were obtained by non-linear regression using the Michaelis–Menten function in Sigma-Plot 8.0. All data are the means of at least two independent experimental series. Kinetic parameters of PLC in the two-phase system (chloroform/50 mM Hepes buffer, pH 7.0) were determined as described in Haftendorn and Ulbrich-Hofmann (2002). 2.7. Determination of the kinetic constants of ppPLA2 by pH-stat titration For the preparation of the substrate stock solution, 50 mol of 1,2- or 1,3-PCs with acyl chain lengths of C10 –C16 were dissolved in 500 L of chloroform. The solvent was evaporated under vacuum, and the lipid film was dissolved in 10 mL of 5 mM sodium deoxycholate containing 100 mM NaCl by vortexing for 15 min as described above. The assay solution (1.5 mL) contained 20–1000 L of the substrate stock solution, 500 L of 1 mM Tris/HCl buffer, pH 8.0 with 100 mM NaCl and 16 mM CaCl2 , and 980–0 L H2 O. The pH of the reaction mixture was adjusted to 8.0, and the reaction was started at 37 ◦ C by addition of 5 L (62.5 mg mL−1 ) of enzyme solution. The amount of liberated fatty acid was determined by continuous titration (3–6 min) with 2 mM NaOH. For the determination of the kinetic constants, initial reaction rates were determined at substrate concentrations between 0.17 and 5 mM and expressed as mol of liberated fatty acid per minute and mg protein. KM,app and Vmax,app were obtained by non-linear regression using the Michaelis–Menten function in Sigma-Plot 8.0. All data are the means of two independent experimental series. 2.8. Determination of the kinetic constants of different PLA2 s by the NEFA C Kit 60 mM stock solutions of 1,2- or 1,3-diC12 PC in chloroform/methanol (2:1), were diluted with the same organic solvent mixture to yield solutions (300 L) with substrate concentrations between 0.2 and 60 mM. These solutions were dried under vacuum, and the lipid films were dissolved in 600 L of 0.5 M Tris/HCl buffer,
pH 8.5, containing 83 mM Triton X-100 and 42 mM CaCl2 by vortexing for 15 min. The enzyme (0.1 nmol) dissolved in 0.1 M Tris/HCl buffer, pH 8.5 was incubated with the corresponding solutions of 1,2 or 1,3-diC12 PC in a total volume of 100 L for 5–30 min at 37 ◦ C. Aliquots (10 L) were removed after defined times of incubation, and the reaction was stopped by addition of 10 L of 0.2 M EDTA. Initial rates were determined from the increase of released fatty acids using the NEFA C Kit according to Hoffmann et al. (Hoffmann et al., 1986). KM,app and Vmax,app were obtained by nonlinear regression using the Michaelis–Menten or Hill function in Sigma-Plot 8.0. All data are the means of at least two independent experimental series. 2.9. Homology modeling of atPLA2 , nmPLA2 , and PLD (Streptomyces sp.) The crystal structure of PLA2 from rice (2wg8) (Guy et al., 2009) was used as template for homology modeling of atPLA2 , the sequence of which (Mansfeld et al., 2006) is 56.1% identical. The crystal structure of PLA2 from the venom of Naja naja sagittifera (1YXH) (Jabeen et al., 2005) was used as template for homology modeling of nmPLA2 (89% sequence identity). For PLD from Streptomyces sp. (Hatanaka et al., 2002), the X-ray structure of PLD from Streptomyces sp. strain PMF (1v0y) (Leiros et al., 2004) with a sequence identity of 70.6% served as template to model the tertiary structure. The homology models were generated using MOE (Molecular Operating Environment, Chemical Computing Group Inc., Montreal, Canada), and the structures were energy-minimized using the CHARMM22 force field (gradient below 0.05) (MacKerell et al., 1998) and Born solvation (Pellegrini and Field, 2002). The resulting stereochemical quality of the structural models of these enzymes was examined by PROCHECK (Laskowski et al., 1993), and native folding was evaluated by PROSAII (Sippl, 1993). In the models, 86.0% (atPLA2 ), 95.5% (nmPLA2 ), and 85.0% (PLD from Streptomyces sp.) of all residues were found in most favored areas of the Ramachandran plot, one outlier appeared in atPLA2 and two outliers in PLD from Streptomyces sp. All other criteria such as peptide bond planarity, bad contacts, H-bond energies were inside allowed regions for a vir˚ The graphical analysis of the energy tual X-ray resolution of 2 A. plot of PROSA II showed all amino acid residues in the negative energy range. For all three models the resulting combined energy z scores are in the expected ranges for natively folded proteins. 2.10. Docking studies For the docking studies, available X-ray structures of the proteins were downloaded from the protein database (http://www.rcsb.org/pdb/home/home.do) (Berman et al., 2000) or the homology models created in this study were used. Using the molecular graphics program SYBYL (Tripos Inc., 2007) all water molecules and co-crystallized ligands were removed and hydrogen atoms were added. The ligands 1,2-di12 PC and 1,3-di12 PC were also constructed with SYBYL and subsequently energy minimized with the Tripos Force field (Clark, 2005). Gasteiger charges were assigned to both the proteins and the ligands (Gasteiger and Marsili, 1978, 1980). Thirty docking arrangements were calculated for each ligand using the docking program GOLD (Genetic Optimized Ligand Docking, Cambridge Crystallographic Data Center; Jones et al., 1997; Nissink et al., 2002; Verdonk et al., 2003) with standard settings and GOLD score as fitness function. The active site was defined by choosing a distance of 15 A˚ from an appropriate atom in the centre of the active site of each protein. The docking arrangements with best fitness scores and appropriate orientation to allow catalysis were energy optimized with the YASARA2 force field embedded in the YASARA modeling program
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
(http://www.yasara.org/). YASARA automatically creates a periodic boundary water box and neutralizes the system by adding chlorine and sodium contour atoms at the surface of the proteins but also adds water molecules or – where appropriate (PLA2 s) – chlorine atoms close to the catalytically essential calcium atom in the active site. 3. Results 3.1. Examination of the cleavage of 1,3-PCs by PLD PLDs are known to catalyze not only the hydrolysis of the terminal phosphodiester bond in 1,2-phospholipids but also – if an appropriate acceptor alcohol is present – the transphosphatidylation. This reaction in competition to hydrolysis is typical of most PLDs belonging to the PLD superfamily (Sung et al., 1997). Transphosphatidylation is particularly pronounced in some microbial PLDs such as the commercially available PLD from Streptomyces sp. (Type VII, Sigma), which is frequently exploited in biocatalysis (Ulbrich-Hofmann, 2000). To verify the non-acceptance of 1,3-PCs by PLD as described by our group for PLD from white cabbage (B. oleracea) in hydrolysis (Haftendorn et al., 2000), we also checked the transphosphatidylation propensity of this plant enzyme in the reaction of 1,3-diC14 PCs with glycerol in comparison to the same reaction of 1,2-diC14 PCs. In contrast to the aqueous reaction systems used in the previous studies (Haftendorn et al., 2000), this reaction was performed in a two-phase system in which the phospholipids are present in the organic phase (diethyl ether). Neither hydrolysis nor transphosphatidylation was observed with 1,3-diC14 PCs, whereas the conversion of 1,2-diC14 PCs resulted in the corresponding products of hydrolysis (phosphatidic acid, 2.3 mol min−1 mg−1 ) and transphosphatidylation (phosphatidylglycerol, 39.7 mol min−1 mg−1 ). This check was also extended to the action of PLD from Streptomyces sp. (Type VII, Sigma). Although the identity of this enzyme with the eukaryotic PLDs on the amino acid sequence level is low, it also belongs to the PLD superfamily, which is characterized by two conserved so-called HKD motifs forming the active site (Ponting and Kerr, 1996). Even this enzyme did not show any reaction with 1,3-diC14 PCs, whereas the transphosphatidylation rate with 1,2-diC14 PCs was 59.1 mol min−1 mg−1 without any significant hydrolysis. Different results were obtained with an enzyme from S. chromofuscus, also called PLD but not belonging to the PLD superfamily (Ulbrich-Hofmann, 2003). This enzyme was able to hydrolyze 1,2as well as 1,3-diC14 PC with a certain preference for the natural regioisomer (16.1 mol min−1 mg−1 ) compared to the 1,3-PC (3.9 mol min−1 mg−1 ). No transphosphatidylation was detected with either substrate. 3.2. Cleavage of 1,3-PCs by PLC The hydrolysis of 1,3-PCs with different chain lengths of acyl moieties (C10 –C16 ) was examined with PLC from B. cereus and compared with the cleavage of the corresponding 1,2-PCs. In addition to the determination of kinetic constants in conventional aqueous system using mixed micelles of the substrate and sodium deoxycholate by pH-stat titration, a two-phase system containing chloroform to solubilize the phospholipid substrates and product quantification by HPTLC was used. In all systems and with all substrates, 1,3-PCs could be cleaved by PLC, even though this occurred with a lower rate than their natural counterparts. The initial rates showed hyperbolic curves as a function of the substrate concentration, which allowed the determination of apparent Michaelis–Menten parame-
199
Table 1 Kinetic constants of PLC-catalyzed hydrolysis of 1,2- and 1,3-PCs with varying acyl chain length in the presence of sodium deoxycholate. Chain length
10 12 14 16 *
Vmax,app (mol min−1 mg−1 )
KM,app (mM)
1,2-PC*
1,2-PC*
1698 854 622 895
± ± ± ±
1,3-PC 280 40 18 58
765 404 232 170
± ± ± ±
258 45 11 6
1.15 0.29 0.76 2.36
± ± ± ±
1,3-PC 0.53 0.07 0.09 0.32
2.07 1.13 0.19 0.05
± ± ± ±
1.40 0.27 0.05 0.01
Data from Haftendorn and Ulbrich-Hofmann (2002).
ters KM,app and Vmax,app . As demonstrated in Table 1 for the reactions in the aqueous system, Vmax,app values of 1,2- and 1,3-PCs differ by a factor of 2–5, while no trend in binding differences of the substrate can be derived from the KM,app values. The same trends were also observed in the two-phase system (data not shown). Interestingly, the maximum rates in the aqueous system are highest with the shortest acyl chain length in 1,3-PCs as well as in 1,2PCs. This effect is even more pronounced for the 1,3-PCs where at substrate saturation the PC with acyl chains of 10 carbon atoms is hydrolysed 4.5 times faster than that with acyl chains of 16 carbon atoms. KM ,app values are in the micro- to millimolar range. For 1,3PCs, these values decrease significantly with increasing acyl chain length, indicating an increasing binding capability. 3.3. Cleavage of 1,3-PCs by PLA2 The hydrolytic cleavage of 1,3-PCs in comparison to 1,2-PCs by PLA2 was first studied with the most common enzyme of the PLA2 superfamily, ppPLA2 . Again the pairs of regioisomers with different chain lengths of acyl moieties (C10 –C16 ), solubilized by sodium deoxycholate, were used in a titrimetric assay, and Vmax,app and KM ,app values (Table 2) were determined from the initial rates measured as a function of the substrate concentration (0.067–3.3 mM). The maximum rates were significantly lower (by the factor of 13–20) for the 1,3-PCs compared to the rates for the 1,2-PCs, whereas the differences in the KM ,app values were not significant within the generally large error of these measurements. Also no significant differences in either parameter could be derived if different chain lengths of the acyl moieties within one class of substrate are compared (Table 2). As PLA2 s are the most studied phospholipases but quantitative data on their activity with 1,3-PCs are rare, we extended our studies to PLA2 s from other sources such as bvPLA2 , nmPLA2 and atPLA2 , which was produced in pure form only recently (Mansfeld et al., 2006; Mansfeld and Ulbrich-Hofmann, 2007), and followed the enzyme kinetics with 1,3- and 1,2-diC12 PC as substrates. In these studies, not the anionic detergent deoxycholate but the neutral Triton X-100 was used for solubilization and the initial rates of the reaction were determined by the enzymatic determination of the released fatty acid. In all cases, the enzyme kinetics revealed sigmoidal curves (Fig. 3), which could be fitted well to the Hill model (Table 3). For all enzymes, sigmoidicity was more pronounced with 1,3-diC12 PC than with the natural substrate, which Table 2 Kinetic constants of ppPLA2 -catalyzed hydrolysis of 1,2- and 1,3-PCs with varying acyl chain length in the presence of sodium deoxycholate. Chain length
10 12 14 16
Vmax,app (mol min−1 mg−1 )
KM,app (mM)
1,2-PC
1,2-PC
119 72 95 55
± ± ± ±
24 7 13 9
1,3-PC 8.97 3.20 6.70 3.72
± ± ± ±
1.70 0.20 1.10 0.30
0.47 0.23 0.47 0.48
± ± ± ±
1,3-PC 0.33 0.11 0.27 0.36
0.71 0.29 1.92 0.12
± ± ± ±
0.40 0.09 0.73 0.13
200
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
Fig. 3. Initial rates as a function of the substrate concentration in the hydrolysis of 1,3-diC12 PC (䊉) and 1,2-diC13 PC () by atPLA2 (A), bvPLA2 (B), nmPLA2 (C), and ppPLA2 (D).
is also reflected in the larger Hill coefficients. In the comparison of the KS,app values (Table 3) no dramatic differences could be found, either between the different enzymes or between the 1,3- and 1,2diC12 PCs. In contrast, PLA2 s from different sources drastically differ in the maximum rates. ppPLA2 showed the highest Vmax,app values toward 1,3- as well as 1,2-diC12 PC, which are 375 and 1630 times larger than those of atPLA2 which had the lowest Vmax,app value. All enzymes distinctly prefer 1,2- over 1,3-diC12 PC. The differentiation is strongest with the plant enzyme atPLA2 and weakest with the snake venom enzyme nmPLA2 .
2004), respectively, were created for the docking studies. As exemplified in Figs. 4–6 for PLD, PLC, and ppPLA2 , and shown for atPLA2, bvPLA2 , and nmPLA2 in the Supplementary material (Figs. S1—S3), 1,3-PCs fit into the active sites of all the enzymes in a manner similar to the natural 1,2-PCs. Some differences, however, were observed in the distances of the catalytically essential residues in the active sites to the critical bonds in the phospholipid structure (Tables 4–6). Thus, His448 which is one of the key residues in the catalytic mechanism of PLD (Fig. S4 in the Supplementary material) is too remote from the respective P–O bond in the 1,3-PC (Table 4, Fig. 4A and B). In contrast, the differences in the distances between the catalytic Zn2+ ions (Fig. S5 in the Supplementary material) and the P–O bond, which is cleaved by PLC (Fig. 5A and B), are not dramatically changed with the 1,2- and 1,3-PCs (Table 5). For the cleavage of the fatty acid ester bond in 1,2-PCs by PLA2 such as ppPLA2 , a catalytic dyad together with an aligned Ca2+ ion and one water molecule is essential (Fig. S6 in the Supplementary material). As concluded from the corresponding critical molecular distances (Table 6, Fig. 6A and B), there are certain deviations for the 1,3-PCs compared to the distances for the 1,2-PCs.
3.4. Docking of 1,2- and 1,3-PCs onto tertiary structures of phospholipases Starting from the available tertiary structures of ppPLA2 , bvPLA2, and PLC from B. cereus, docking experiments with 1,2- and 1,3diC12 PCs were performed. For atPLA2 , nmPLA2 , and PLD from Streptomyces sp. homology models, based on the tertiary structures of PLA2 from rice (Guy et al., 2009), Andaman cobra venom (Jabeen et al., 2005), and PLD from Streptomyces sp. strain PMF (Leiros et al.,
Table 3 Kinetic constants of the hydrolysis of 1,2- and 1,3-diC12 PC by different PLA2 s in the presence of Triton X-100. Enzyme
Vmax,app (mol min−1 mg−1 ) 1,2-PC
atPLA2 bvPLA2 nmPLA2 ppPLA2
15.7 401 200 5890
± ± ± ±
KS,app (mM)
1,3-PC 0.6 11 20 151
0.40 11.0 124 652
± ± ± ±
1,2-PC 0.01 1.1 5 12
1.9 7.8 5.2 2.5
± ± ± ±
0.2 0.4 0.9 0.1
Hill coefficient 1,3-PC 5.7 7.9 2.4 4.7
± ± ± ±
0.3 1.2 0.2 0.1
1,2-PC 1.4 1.4 1.4 2.7
± ± ± ±
0.2 0.1 0.2 0.3
1,3-PC 2.6 1.9 2.6 5.1
± ± ± ±
0.3 0.2 0.4 0.5
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
201
Fig. 4. Active site of PLD from Streptomyces sp. docked with 1,2-diC12 PC (A) and 1,3-diC12 PC (B). Blue colour indicates the His residues (H170, H448) of the two HKD motifs, black colour the two Lys residues (K172, K450) of the HKD motifs. The amino acids shown in red represent the two catalytically important Asp residues (D202, D473) (Fig. S4), the amino acid in orange is Y390 which is involved in substrate binding. The molecules drawn in yellow represent the corresponding PCs (blue marks nitrogen, yellow carbon, orange phosphorus, red oxygen). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Table 4 Distances of selected atoms involved in catalysis of the cleavage of the phosphodiester bond by PLD (numbering of residues according to PLD from Streptomyces sp.). Phospholipid
1,2-diC12 PC 1,3-diC12 PC
Distances (Å) N(H170)–P
N(H448)–O (POR)
H(K172)–O (PO)
H(H448)–O (D202)
H(H170)–O (D473)
3.64 3.32
2.91 5.51
2.26 2.82
1.69 1.69
1.78 1.78
N(H170)–P represents the distance between His170N␦1 and the phosphorus atom of the phospholipid being attacked. N(H448)–O (POR) is the distance between His448N␦1 and the oxygen atom of the phosphodiester bond between the phosphate and the choline moiety of the phospholipid to be cleaved (Figs. 4 and S4).
Table 5 Distances of selected atoms being involved in catalysis of the cleavage of the phosphodiester bond by PLC. Phospholipid
1,2-diC12 PC 1,3-diC12 PC
Distances (Å) O(D55)–H (H2 O)
P–O (H2 O)
Zn1 –O (PO)
Zn2 –O (PO)
Zn3 –O (PO)
3.27 3.42
3.29 3.37
3.05 3.48
1.75 1.77
1.79 1.78
O(D55)–H (H2 O) represents the distance between D55, acting as general base, and the water which attacks the phosphodiester bond. P–O (H2 O) is the distance between the activated water and the attacked phosphodiester bond (Figs. 5 and S5).
Fig. 5. Active site of PLC from B. cereus docked with 1,2-diC12 PC (A) and 1,3-diC12 PC (B). The green spheres indicate the three Zn2+ ions. The amino acids coloured in green (H118, D122, E146) are involved in complexation of the three Zn2+ ions; the amino acids shown in grey (Y56, F66) are involved in interactions with the choline moiety. D55, also shown in grey, acts as general base in deprotonation of the attacking water molecule in the catalytic mechanism (Fig. S5). The amino acid drawn in pale cyan stabilizes the positive charge of the choline moiety. The yellow-coloured molecules are the corresponding PCs as described in Fig. 4. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
202
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
Fig. 6. Active site of ppPLA2 docked with 1,2-diC12 PC (A) and 1,3-diC12 PC (B). The magenta sphere indicates the catalytic Ca2+ ion. The yellow-coloured molecules are the corresponding PCs as described in Fig. 4. The amino acid residues presented in grey are the amino acids being involved in catalysis according to the mechanism shown in Fig. S6. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
4. Discussion The evaluation of kinetic data for phospholipases is problematic because their substrates are poorly water-soluble and are, therefore, mostly used in the presence of detergents or organic solvents. The enzymes, in turn, are activated by interfaces and, thus, the reaction rates are strongly influenced by the type and size of the substrate aggregates. For this reason, kinetic constants of phospholipases for different substrates can be compared only if comparable interfaces can be assumed, and the ascertained constants must be treated as apparent constants. 1,3- and 1,2-PCs were previously shown to form very similar aggregate structures (Frense et al., 1995; Haftendorn et al., 2000; Kunitake et al., 1984) allowing a direct comparison of their cleavage by the same phospholipase. Although the enzymes considered in this study do not represent all types of PLD, PLC and PLA2 and the results, therefore, do not allow general conclusions striking differences between the studied members of the three subclasses of phospholipases in the acceptance of 1,3PCs are observed. While the differences are moderate with PLC (B. cereus) (Table 1), they are larger with PLA2 s from several sources (Tables 2 and 3). Interestingly, common PLDs from plant or microbial origin are not able to convert 1,3-PCs. The non-acceptance of 1,3-PC by PLDs, in contrast to PLC and PLA2 s, is remarkable, because the potential cleavage site in the com-
Table 6 Distances of selected atoms being involved in the catalytic mechanism of PLA2 s. Enzyme
Distances (Å) O–Ca
atPLA2 bvPLA2 nmPLA2 ppPLA2
O(H2)–C O
N–C
1,2-PC
1,3-PC
1,2-PC
1,3-PC
1,2-PC
1,3-PC
2.51 2.30 2.46 2.16
2.54 2.41 2.47 2.39
3.56 3.09 3.01 3.05
4.06 3.50 2.81 3.95
5.95 3.70 5.44 4.18
6.83 3.53 4.03 4.17
O–Ca represents the distance of the carbonyl oxygen atom of the ester bond to be hydrolyzed to the Ca2+ ion in the active site; O(H2)–C O is the distance of the reacting water oxygen atom to the carbonyl carbon atom being attacked by the water; and N–C denotes the distance of the HisN␦1 atom and the above-mentioned carbonyl carbon atom (Figs. 6, S1–S3 and S6).
pound is more remote from the 1,3-positioned acyl chains than the cleavage sites for PLC or PLA2 . On the other hand, 1,3-PCs were found to be competitive inhibitors of PLD (Dittrich et al., 1998; Haftendorn et al., 2000) suggesting that the compounds are bound in the active site. Obviously, however, they are bound in a way that does not allow cleavage of the terminal phosphate ester bond. This interpretation is supported by the docking studies performed with a homology model for PLD from Streptomyces sp. (Fig. 4A and B, Table 4). Accordingly, 1,2- as well as 1,3-PC fit to the active site but the head group of 1,3-PC is positioned in a way that does not allow the cleavage of the phosphatidylcholine bond. On the other hand, alkylphosphocholines, which lack the glycerol backbone, can be cleaved by PLD (Dittrich et al., 1998). The finding that the commercial enzyme from S. chromofuscus behaves differently and is able to hydrolyze 1,2- as well as 1,3-PCs, but does not transesterify these compounds, is consistent with the completely different properties of this enzyme which should be assigned to the group of alkaline phosphatases rather than to PLDs (Ulbrich-Hofmann, 2003; Mansfeld and Ulbrich-Hofmann, 2009). Differences in the kinetics of the hydrolysis of 1,3- and 1,2-PCs by PLC as well as by PLA2 s become more manifest in the Vmax,app than in the KM,app values (Tables 1–3). Obviously, steric differences play a smaller role in substrate binding than in the catalytic cleavage step. PLC (B. cereus) proved to be relatively unspecific with respect to the position of the head group at the glycerol backbone. The maximum rates in the hydrolysis of 1,2-PCs with different acyl chain lengths (C10 –C16 ) by PLC (B. cereus) are only 2.1–5.2 times higher than those of 1,3-PCs (Table 1). Obviously, the active site cleft, which is 8 A˚ deep and 5 A˚ wide (Hough et al., 1989) is large enough to accommodate the 1,3-PCs as well as the 1,2-PCs, and the polar and hydrophobic interactions essential for catalysis (Hough and Hansen, 1994) are similar to those of the natural substrate isomer. These results are in accordance with the docking studies (Fig. 5A and B, Table 5), which reveal only marginal differences between 1,2- and 1,3-PCs. A certain non-specificity of this enzyme was also previously reported with respect to other substrate analogs such as short-chain lysophosphatidylcholines (El-Sayed et al., 1985) or asymmetric short-chain phosphatidylcholines (Lewis et al., 1990). The slight decrease of the maximum rate with increasing chain length of the acyl moieties in 1,3-PCs
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
is similar to that observed in 1,2-PCs with acyl chain lengths of C10 –C16 (Table 1). A similar trend was previously reported for the same enzyme with 1,2-PCs with acyl chain lengths of C12 –C16 (Roberts et al., 1978). The differences in the maximum rates of PLA2 s vary considerably with their source, although all four enzymes studied belong to one superfamily and possess very similar tertiary structures (Burke and Dennis, 2009; Mansfeld, 2009). They are most pronounced in atPLA2 and bvPLA2 where 1,2-diC12 PC is 39 and 36 times faster cleaved in comparison to 1,3-diC12 PC, whereas these factors are only 9 and 1.6 with ppPLA2 and nmPLA2 (Table 3). Presumably, small changes in the architecture of the active sites, as also derived from the docking studies (Fig. 6A and B and Figs. S1–S3 in the Supplements, Table 6), are responsible for the differences. Recent modeling studies in the literature have shown that too bulky substituents in the sn-1 position of 1,2-PCs (Linderoth et al., 2008) or the presence of the head group in position 2 of 1-alkyl, 3-acylglycerophospholipids (Peters et al., 2007) prevent the catalytically important water molecule from getting into the correct position to take part in cleavage of the substrate. Compared to the position of the polar head group (1,2- or 1,3-position of the glycerol backbone), the chain length of the acyl moieties (C10 –C16 ) in 1,3-PCs as well as in 1,2-PCs in the case of ppPLA2 plays a minor role (Table 2). The slight tendency of decreasing maximum rate with increasing chain length is consistent with the observations for atPLA2 (Mansfeld and Ulbrich-Hofmann, 2007) and PLA2 from Naja naja naja (Roberts et al., 1978) with 1,2-PCs of varying chain length solubilized by Triton X-100. Interestingly, the kinetic functions of all the different PLA2 s in Fig. 3 show sigmoidicity, which is almost negligible with the 1,2-PCs but distinct with the 1,3-substrates. In these cases, experimental data can be fitted well according to the Hill model (Table 3), indicating cooperative substrate binding to at least two binding sites. Although the kinetics of selected PLA2 s have been carefully studied and these enzymes serve as models of interfacial kinetics (Berg and Jain, 2002), cooperativity in substrate binding was seldom reported. Hendrickson and Dennis (1984) described cooperativity with a Hill coefficient of 1.4 in the hydrolysis of the synthetic substrate 1,2-bis(decanoylthio)-1,2-dideoxy-snglycero-3-phosphocholine in the presence of Triton X-100 by PLA2 from Naja naja naja. Obviously, cooperativity in substrate binding becomes more efficient with poor substrates. These effects may be similar to the carefully studied effects of cooperative binding of anionic amphiphiles to the interfacial binding surface of ppPLA2 (Berg et al., 2004, 2009). In summary, the present study shows that the 1,3-PCs are bound to phospholipases with similar affinity as the natural 1,2PCs, whereas their hydrolysis is decelerated to different degrees (PLC, PLA2 ) or even abolished (PLD). These properties make them to interesting substitutes for natural phospholipids such as in liposomal drug carriers and gene delivery systems.
Acknowledgements The financial support of the Kultusministerium des Landes Sachsen-Anhalt (Magdeburg, Germany) is gratefully acknowledged. The authors thank Gary Sawers (Martin-Luther University Halle-Wittenberg, Germany) for reading the manuscript and Mrs. Sonntag for expert technical assistance.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.chemphyslip.2010.12.009.
203
References Berg, O.G., Jain, M.K., 2002. In: Burgoyne, R.D., Morgan, A. (Eds.), Interfacial Enzyme Kinetics. John Wiley, Chichester, UK, 301 pp. Berg, O.G., Yu, B.Z., Chang, C., Koehler, K.A., Jain, M.K., 2004. Cooperative binding of monodisperse anionic amphiphiles to the i-face: phospholipase A2 -paradigm for interfacial binding. Biochemistry 43, 7999–8013. Berg, O.G., Yu, B.Z., Jain, M.K., 2009. Thermodynamic reciprocity of the inhibitor binding to the active site and the interface binding region of IB phospholipase A2 . Biochemistry 48, 3209–3218. Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, H., Shindyalov, I.N., Bourne, P.E., 2000. The Protein Data Bank. Nucleic Acids Res. 28, 235–242. Burke, J.E., Dennis, E.A., 2009. Phospholipase A2 structure/function, mechanism, and signaling. J. Lipid Res. 50 (Suppl.), S237–242. Clark, M., 2005. Generalized fragment-substructure based property prediction method. J. Chem. Inf. Model. 45, 30–38. de Haas, G.H., van Deenen, L.L., 1964. The site of action of phospholipase A on lecithins. Biochim. Biophys. Acta 84, 469–471. de Haas, G.H., van Deenen, L.L., 1965. Structural identification of isomeric lysolecithins. Biochim. Biophys. Acta 106, 315–325. Dippe, M., Ulbrich-Hofmann, R., 2009. Substrate specificity in phospholipid transformations by plant phospholipase D isoenzymes. Phytochemistry 70, 361–365. Dittrich, N., Haftendorn, R., Ulbrich-Hofmann, R., 1998. Hexadecylphosphocholine and 2-modified 1,3-diacylglycerols as effectors of phospholipase D. Biochim. Biophys. Acta 1391, 265–272. Dluhy, R.A., Chowdhry, B.Z., Cameron, D.G., 1985. Infrared characterization of conformational differences in the lamellar phases of 1,3-dipalmitoyl-sn-glycero2-phosphocholine. Biochim. Biophys. Acta 821, 437–444. El-Sayed, M.Y., DeBose, C.D., Coury, L.A., Roberts, M.F., 1985. Sensitivity of phospholipase C (Bacillus cereus) activity to phosphatidylcholine structural modifications. Biochim. Biophys. Acta 837, 325–335. Frense, D., Haftendorn, R., Ulbrich-Hofmann, R., 1995. 2-modified 1,3diacylglycerols as new surfactants for the formation of reverse micelles. Chem. Phys. Lipids 78, 81–87. Gasteiger, J., Marsili, M., 1978. New model for calculating atomic charges in molecules. Tetrahedron Lett., 3181–3184. Gasteiger, J., Marsili, M., 1980. Iterative partial equalization of orbital electronegativity—a rapid access to atomic charges. Tetrahedron 36, 3219–3228. Guy, J.E., Stahl, U., Lindqvist, Y., 2009. Crystal structure of a class XIB phospholipase A2 (PLA2 ): rice (oryza sativa) isoform-2 pla2 and an octanoate complex. J. Biol. Chem. 284, 19371–19379. Haftendorn, R., Schwarze, G., Ulbrich-Hofmann, R., 2000. 1,3-Diacylglycero-2phosphocholines—synthesis, aggregation behaviour and properties as inhibitors of phospholipase D. Chem. Phys. Lipids 104, 57–66. Haftendorn, R., Ulbrich-Hofmann, R., 1995. Synthesis of 2-modified 1,3diacylglycerols. Tetrahedron 51, 1177–1186. Haftendorn, R., Ulbrich-Hofmann, R., 2002. Activity of phospholipase C in two-phase systems. Anal. Biochem. 306, 144–147. Hatanaka, T., Negishi, T., Kubota-Akizawa, M., Hagishita, T., 2002. Purification, characterization, cloning and sequencing of phospholipase D from Streptomyces septatus TH-2. Enzyme Microb. Technol. 31, 233–241. Hendrickson, H.S., Dennis, E.A., 1984. Kinetic analysis of the dual phospholipid model for phospholipase A2 action. J. Biol. Chem. 259, 5734–5739. Hirche, F., Koch, M.H.H., König, S., Wadewitz, T., Ulbrich-Hofmann, R., 1997. The influence of organic solvents on phospholipid transformations by phospholipase D in emulsion systems. Enzyme Microb. Technol. 20, 453–461. Hoffmann, G.E., Schmidt, D., Bastian, B., Guder, W.G., 1986. Photometric determination of phospholipase A. J. Clin. Chem. Clin. Biochem. 24, 871–875. Hough, E., Hansen, L.K., Birknes, B., Jynge, K., Hansen, S., Hordvik, A., Little, C., Dodson, ˚ crystal structure of phospholiE., Derewenda, Z., 1989. High-resolution (1.5 A) pase C from Bacillus cereus. Nature 338, 357–360. Hough, E., Hansen, S., 1994. Structural aspects of phospholipase C from Bacillus cereus and its reaction mechanism. In: Woolley, P., Petersen, S.B. (Eds.), Lipases. Cambridge University Press, Cambridge, pp. 95–118. Jabeen, T., Singh, N., Singh, R.K., Ethayathulla, A.S., Sharma, S., Srinivasan, A., Singh, T.P., 2005. Crystal structure of a novel phospholipase A2 from Naja naja sagittifera with a strong anticoagulant activity. Toxicon 46, 865–875. Jones, G., Willett, P., Glen, R.C., Leach, A.R., Taylor, R., 1997. Development and validation of a genetic algorithm for flexible docking. J. Mol. Biol. 267, 727–748. Kunitake, T., Okahata, Y., Tawaki, S.I., 1984. Bilayer characteristics of 1,3-dialkyl- and 1,3-diacyl-rac-glycero-2-phosphocholines. J. Colloid Interface Sci. 103, 190–201. Kuppe, K., Kerth, A., Blume, A., Ulbrich-Hofmann, R., 2008. Calcium-induced membrane microdomains trigger plant phospholipase D activity. ChemBiochem 9, 2853–2859. Laskowski, R.A., MacArthur, M.W., Moss, D.S., Thornton, J.M., 1993. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, 283–291. Leiros, I., McSweeney, S., Hough, E., 2004. The reaction mechanism of phospholipase D from Streptomyces sp. strain PMF. Snapshots along the reaction pathway reveal a pentacoordinate reaction intermediate and an unexpected final product. J. Mol. Biol. 339, 805–820. Lewis, K.A., Bian, J.R., Sweeney, A., Roberts, M.F., 1990. Asymmetric short-chain phosphatidylcholines: defining chain binding constraints in phospholipases. Biochemistry 29, 9962–9970.
204
J. Mansfeld et al. / Chemistry and Physics of Lipids 164 (2011) 196–204
Linderoth, L., Andresen, T.L., Jorgensen, K., Madsen, R., Peters, G.H., 2008. Molecular basis of phospholipase A2 activity toward phospholipids with sn-1 substitutions. Biophys. J. 94, 14–26. MacKerell Jr., A.D., Bashford, D., Bellott, M., Dunbrack Jr., R.L., Evanseck, J., Field, M.J., Fischer, S., Gao, J., Guo, H., Ha, S., Joseph, D., Kuchnir, L., Kuczera, K., Lau, F.T.K., Mattos, C., Michnick, S., Ngo, T., Nguyen, D.T., Prodhom, B., Reiher, W.E., Roux, B., Schlenkrich, M., Smith, J., Stote, R., Straub, J., Watanabe, M., WiorkiewiczKuczera, J., Yin, D., Karplus, M., 1998. All-atom empirical potential for molecular modeling and dynamics studies of protein. J. Phys. Chem. B 102, 3586–3616. Mansfeld, J., 2009. Plant phospholipases A2 : perspectives on biotechnological applications. Biotechnol. Lett. 31, 1373–1380. Mansfeld, J., Gebauer, S., Dathe, K., Ulbrich-Hofmann, R., 2006. Secretory phospholipase A2 from Arabidopsis thaliana: insights into the three-dimensional structure and the amino acids involved in catalysis. Biochemistry 45, 5687–5694. Mansfeld, J., Ulbrich-Hofmann, R., 2009. The modulation of phospholipase D activity in vitro. Biochim. Biophys. Acta 1791, 913–926. Mansfeld, J., Ulbrich-Hofmann, R., 2007. Secretory phospholipase A2 -␣ from Arabidopsis thaliana: functional parameters and substrate preference. Chem. Phys. Lipids 150, 156–166. Nissink, J.W., Murray, C., Hartshorn, M., Verdonk, M.L., Cole, J.C., Taylor, R., 2002. A new test set for validating predictions of protein–ligand interaction. Proteins 49, 457–471. Pellegrini, E., Field, M.J., 2002. A generalized-born solvation model for macromolecular hybrid-potential calculations. J. Phys. Chem. A 106, 1316–1326. Peters, G.H., Moller, M.S., Jorgensen, K., Ronnholm, P., Mikkelsen, M., Andresen, T.L., 2007. Secretory phospholipase A2 hydrolysis of phospholipid analogues is dependent on water accessibility to the active site. J. Am. Chem. Soc. 129, 5451–5461. Ponting, C.P., Kerr, I.D., 1996. A novel family of phospholipase D homologues that includes phospholipid synthases and putative endonucleases: identifica-
tion of duplicated repeats and potential active site residues. Protein Sci. 5, 914–922. Roberts, M.F., Otnaess, A.B., Kensil, C.A., Dennis, E.A., 1978. The specificity of phospholipase A2 and phospholipase C in a mixed micellar system. J. Biol. Chem. 253, 1252–1257. Schäffner, I., Rücknagel, K.-P., Mansfeld, J., Ulbrich-Hofmann, R., 2002. Genomic structure, cloning and expression of two phospholipase D isoenzymes from white cabbage. Eur. J. Lipid Sci. Technol. 104, 79–87. Seelig, J., Dijkman, R., de Haas, G.H., 1980. Thermodynamic and conformational studies on sn-2-phosphatidylcholines in monolayers and bilayers. Biochemistry 19, 2215–2219. Sippl, M.J., 1993. Recognition of errors in three-dimensional structures of proteins. Proteins 17, 355–362. Slotboom, A.J., Verger, R., Verheij, H.M., Baartmans, P.H., Deenen, L.L., Haas, G.H., 1976. Application of enantiomeric 2-sn-phosphatidylcholines in interfacial enzyme kinetics of lipolysis. Chem. Phys. Lipids 17, 128–147. Sung, T.C., Roper, R.L., Zhang, Y., Rudge, S.A., Temel, R., Hammond, S.M., Morris, A.J., Moss, B., Engebrecht, J., Frohman, M.A., 1997. Mutagenesis of phospholipase D defines a superfamily including a trans-Golgi viral protein required for poxvirus pathogenicity. EMBO J. 16, 4519–4530. Tripos Inc., S.H.R.S.L., MO, USA, 2007. SYBYL Computer Program. Version 7.0. Ulbrich-Hofmann, R., 2000. Phospholipases used in lipid transformations. In: Bornscheuer, U.T. (Ed.), Enzymes in Lipid Modification. Wiley-VCH, Weinheim, pp. 219–262. Ulbrich-Hofmann, R., 2003. Enzyme-catalysed transphosphatidylation. Eur. J. Lipid Sci. Technol. 105, 305–307. Verdonk, M.L., Cole, J.C., Hartshorn, M.J., Murray, C.W., Taylor, R.D., 2003. Improved protein–ligand docking using GOLD. Proteins 52, 609–623. Wieder, T., Reutter, W., Orfanos, C.E., Geilen, C.C., 1999. Mechanisms of action of phospholipid analogs as anticancer compounds. Prog. Lipid Res. 38, 249–259.