Purification of lysophospholipases. Application of a continuous spectrophotometric assay using thioester substrate analogs.

Purification of lysophospholipases was monitored with four analogs of the natural lysophosphatidylcholine substrate, including two analogs with an acylthioester bond. In all chromatographic procedures employed, peaks of enzymatic activity towards each of the substrates were coincidental; moreover, the ratio of thioester to oxyester hydrolysis rates remained essentially constant over a more than 500-fold purification. These findings strongly support the conclusion that the hydrolysis of the thioester substrates truly reflects the specificity of lysophospholipases, thus allowing the use of a convenient spectrophotometric assay for these enzymes.

Phase properties of the aqueous dispersions of n-octadecylphosphocholine.

Properties of the aqueous dispersions of n-octadecylphosphocholine are examined by differential scanning calorimetry, fluorescence depolarization, light scattering, 31P-NMR, pig pancreatic phospholipase A2 binding, and X-ray diffraction. On heating, these dispersions exhibit a sharp lamellar to micelle transition at 20.5 degrees C. The lamellar phase consists of frozen (gel-state) alkyl chains which do not bind phospholipase A2. The kinetics of the transition are asymmetric: the micelle to lamellar transition is very slow and the lamellar to micelle transition is fast. It is suggested that the lamellar phase is a frozen chain bilayer in which the chains interdigitate.

Kinetic behavior of porcine pancreatic phospholipase A2 on zwitterionic and negatively charged single-chain substrates.

The kinetic properties of porcine pancreatic phospholipase A2 were studied on a series of n-acylglycollecithins and n-acylglycol sulfates containing acyloxy or acylthio ester bonds at substrate concentrations below and above the critical micelle concentration. These single-chain detergents containing a primary (thio) ester bond are hydrolyzed rather slowly by the pancreatic enzyme, and maximal activity was found always for the n-octanoyl derivatives. The acylthio ester group is split 4-5 times faster than the corresponding acyloxy ester function. The kinetic behavior of the enzyme acting on zwitterionic glycollecithins or on anionic glycol sulfates is quite different and provides an explanation for the differences in pH optimum. Both for glycollecithins and for glycol sulfates, maximal enzyme activities are found in high molecular weight aggregates consisting of several enzyme molecules and detergent monomers. Their pathway of formation, however, is not the same.

Kinetic behavior of porcine pancreatic phospholipase A2 on zwitterionic and negatively charged double-chain substrates.

A number of isomeric diacylglycerophosphocholines and diacylglycero sulfates containing O-acyl and/or S-acyl ester bonds were investigated as substrates for porcine pancreatic phospholipase A2 and its zymogen. A comparison is made with the kinetic properties of the enzyme toward the corresponding glycol detergents previously described [van Oort, M. G., Dijkman, R., Hille, J. D. R., & de Haas, G. H. (1985) Biochemistry (preceding paper in this issue)]. Hydrolysis of the secondary ester bond in the 1,2-diacylglycero-3-type lipids proceeds much faster than the splitting of the primary ester function present in the isomeric 1,3-diacylglycerol and 1-acylglycol derivatives. In sharp contrast to the glycol detergents, the substitution of the cleavable oxygen ester by a thio ester bond in the glycerol lipids results in 5 times lower kcat values. At alkaline pH and above the critical micelle concentration, the anionic sulfates are much better substrates than the corresponding phosphocholine-containing detergents. At very low detergent concentrations, below the critical micelle concentration, the anionic sulfates induce protein aggregation such that phospholipase A2, as well as its zymogen, is present in high molecular weight complexes containing several protein molecules. In these aggregates, protein-protein and/or lipid-protein interactions strongly activate phospholipase but not the zymogen.

Aggregation of porcine pancreatic phospholipase A2 and Its zymogen induced by submicellar concentrations of negatively charged detergents.

The interaction of two sodium n-alkyl sulfates (C10 and C12) with porcine pancreatic phospholipase A2 and its zymogen was studied by various spectroscopic techniques, equilibrium gel filtration, calorimetry, and photochemically induced dynamic nuclear polarization 1H NMR. At very low concentrations of n-alkyl sulfate [up to 0.07 X critical micelle concentration (cmc)] the enzyme (Mr 14 000) is able to build up a complex with the detergent molecules having a molecular weight of about 90 000. This complex consists of 6 enzyme molecules and about 40 n-alkyl sulfate monomers. The formation of the detergent-protein aggregate occurs in a two-step process: First, two detergent molecules strongly bind to a hydrophobic surface region of the protein, previously called interface recognition site [Pieterson, W. A., Vidal, J. C., Volwerk, J. J., & de Haas, G. H. (1974) Biochemistry 13, 1455-1460]. Subsequently, at higher detergent concentrations suddenly 6 enzyme molecules aggregate, probably including about 30 additional detergent monomers. Although the zymogen of the pancreatic phospholipase A2 seems to form comparable high molecular weight aggregates with these detergents, there are spectroscopic differences, and higher detergent concentrations are required. Moreover, as will be shown in the following paper [Hille, J. D. R., Egmond, M. R., Dijkman, R., van Oort, M. G., Sauve, P., & de Haas, G. H. (1983) Biochemistry (following paper in this issue)], only the phospholipase A2 becomes superactivated in these complexes.

Unusual kinetic behavior of porcine pancreatic (pro)phospholipase A2 on negatively charged substrates at submicellar concentrations.

The negatively charged detergents S-n-alka-noylthioglycol sulfates (C8, C9, and C10) are substrates for porcine pancreatic phospholipase A2 and its zymogen. At pH 6.0 and detergent concentrations up to 0.08 X critical micelle concentration (cmc), the activities of active enzyme and zymogen are similar and very low. From 0.08 X cmc to 0.12 X cmc a tremendous increase in activity is observed for phospholipase A2, but not for the zymogen. Concomitant with this increase in activity there is a sharp rise in molecular weight of the substrate-enzyme complex, from 15 000 to 95 000, and in detergent to protein molar ratio of 1:1 to about 7:1. This indicates both substrate and enzyme aggregation. Most probably a lipid-water interface is formed inside the aggregated protein particle by which the enzyme is activated. Although the zymogen also forms high molecular weight complexes with similar molar ratios, no activation is observed probably because of distortion of its lipid binding domain.

Physicochemical studies on the interaction of pancreatic phospholipase A2 with a micellar substrate analogue.

Porcine pancreatic phospholipase A2 interacts with micelles of the substrate analogue n-octadecylphosphocholine to form a specific complex over considerably wide concentration ranges of both lipid and protein. UV absorption difference spectroscopy measurements indicate that the ratio of lipid to protein molecules in the complex is approximately 50. This number is confirmed by using other techniques to study the composition of the complex, namely, ultracentrifugation experiments and light scattering. The latter techniques furthermore demonstrate that the lipid--protein complex consists of 100 lipid and 2 enzyme molecules. Thus, the number of lipid molecules in the free micelle (200) is halved when the complex with phospholipase is being formed. The consequences of the results are discussed in relation to a theoretical model of the lipid--protein interaction.

Binding of porcine pancreatic phospholipase A2 to various micellar substrate analogues. Involvement of histidine-48 and aspartic acid-49 in the binding process.

The interaction of porcine pancreatic phospholipase A2 (PA2) with micelles of various single-chain phospholipid analogues was studied by ultraviolet absorption difference spectroscopy and light-scattering measurements. The phospholipids used were either substrate analogues or products, varying in hydrocarbon chain lengths and polar head groups. The results indicate that the enzyme forms a stable complex over a wide range of enzyme and lipid concentrations. From the equivalent "molecular weight" and from the lipid to enzyme molar ratio (N) of the micelle--enzyme complex, it can be calculated that complexes containing saturated hydrocarbon chain lipids generally consist of two enzyme molecules and half of the number of lipid monomers present in free micelles. The interaction forces between the enzyme and lipid monomers bound in the complex are mainly hydrophobic. Stronger binding is found when the essential cofactor Ca2+ is bound to the enzyme. pH-titration studies on the binding of native PA2 to aggregated lipid structures showed that at least one group with a pKA value of 6.25 is involved in the interaction with lipid micelles. At acidic pH, micelle binding is stronger than at neutral or alkaline pH. Alkylation of the active site residue His48 resulted in a shift of the pKA value to 4.6, while addition of Ca2+ appears to stabilize the micelle-binding conformation of both native and modified enzymes over a broad pH range (pH 4--9.5). From these observations it is suggested that both the Ca2+ binding residue Asp49 [Fleer, E. A. M., Verheij, H. M., & de Haas, G. H. (1980) Eur. J. Biochem. 113, 283--288] and His48 control micelle binding of the native enzyme. For optimal binding in the absence of Ca2+, a long-distance hydrogen bond between these two residues is required; this can be established via a water molecule. It is assumed that it is a proton of this "H bond" which is titrated with a pKA value of 6.25. When the "H bond" is absent, as in the alkylated enzymes, Asp49 alone controls micelle binding with a pKA of 4.6. These results, together with the effect of Ca2+ on micelle binding, indicate that it is not the "hydrogen bridge" between His48 and Asp49 which is of main importance for an optimum binding conformation of the enzyme but the effective charge in the microenvironment of Asp49. It is proposed that a negative charge on this carboxylate causes a conformational change of the enzyme which leads to a protein conformation lacking an active micelle binding site. Binding of Ca2+ or reprotonation neutralizes this negative charge and restores the enzyme's ability to bind micelles.