Distinction between esterases and lipases: comparative biochemical properties of sequence-related carboxylesterases.

Carboxylesterases (Carboxyl ester hydrolase) include two groups of enzymes, namely non-specific esterases (EC 3.1.1.1) and lipases (EC 3.1.1.3) which have been early differentiated on the basis of their substrate specificity. Esterases hydrolyse solutions of water-soluble short acyl chain esters and are inactive against water-insoluble long chain triacylglycerols which, in turn, are specifically hydrolyzed by lipases. Based on the comparison of the primary structures, three families of sequence-related carboxylesterases, namely the lipoprotein lipase family (L-family), the hormonesensitive lipase family (H-family) and the cholinesterase family (C-family) have been identified. Using solutions and emulsions of vinyl, glyceryl and p-nitrophenyl esters, we have reinvestigated the kinetic properties of some esterases and lipases of the H- and C-families. Results indicate that esterases and lipases, which are both active on soluble esters, can be differentiated by their value of Km. Moreover, esterase, unlike lipases, are inactive against water-insoluble esters as vinyl laurate and trioctanoylglycerol. From the the comparison of structural features of sequence-related esterases and lipases, it appears that lipases, unlike esterases, display a significant difference in the distribution of hydrophobic amino acid residues at vicinity of their active site. This observation supports the hypothesis of the existence in lipases of a particular surface domain that specifically interacts with lipid-water interfaces and contributes to the transfer a single substrate molecule from the organized lipid-water interface (supersubstrate) to the catalytic site of the enzyme.

Lipase-catalysed hydrolysis of short-chain substrates in solution and in emulsion: a kinetic study.

We have studied the enzymatic hydrolysis of solutions and emulsions of vinyl propionate, vinyl butyrate and tripropionin by lipases of various origin and specificity. Kinetic studies of the hydrolysis of short-chain substrates by microbial triacylglycerol lipases from Rhizopus oryzae, Mucor miehei, Candida rugosa, Candida antarctica A and by (phospho)lipase from guinea-pig pancreas show that these lipolytic enzymes follow the Michaelis-Menten model. Surprisingly, the activity against solutions of tripropionin and vinyl esters ranges from 70% to 90% of that determined against emulsions. In contrast, a non-hyperbolic (sigmoidal) dependence of enzyme activity on ester concentration is found with human pancreatic lipase, triacylglycerol lipase from Humicola lanuginosa (Thermomyces lanuginosa) and partial acylglycerol lipase from Penicillium camembertii and the same substrates. In all cases, no abrupt jump in activity (interfacial activation) is observed at substrate concentration corresponding to the solubility limit of the esters. Maximal lipolytic activity is always obtained in the presence of emulsified ester. Despite progress in the understanding of structure-function of lipases, interpretation of the mode of action of lipases active against solutions of short-chain substrates remains difficult. Actually, it is not known whether these enzymes, which possess a lid structure, are in open or/and closed conformation in the bulk phase and whether the opening of the lid that gives access to the catalytic triad is triggered by interaction of the enzyme molecule with monomeric substrates or/and multimolecular aggregates (micelles) both present in the bulk phase. From the comparison of the behaviour of lipases used in this study which, in some cases, follow the Michaelis-Menten model and, in others, deviate from classical kinetics, it appears that the activity of classical lipases against soluble short-chain vinyl esters and tripropionin depends not only on specific interaction with single substrate molecules at the catalytic site of the enzyme but also on physico-chemical parameters related to the state of association of the substrate dispersed in the aqueous phase. It is assumed that the interaction of lipase with soluble multimolecular aggregates of tripropionin or short-chain vinyl esters or the formation of enzyme-substrate mixed micelles with ester bound to lipase, might represent a crucial step that triggers the structural transition to the open enzyme conformation by displacement of the lid.

Kinetic properties of Penicillium cyclopium lipases studied with vinyl esters.

Penicillium cyclopium produces two lipases with different substrate specificities. Lipase I is predominantly active on triacylglycerols whereas lipase II hydrolyzes mono- and diacylglycerols but not triacylglycerols. In this study, we compared the kinetic properties of P. cyclopium lipases and human pancreatic lipase, a classical triacylglycerol lipase, by using vinyl esters as substrates. Results indicate that P. cyclopium lipases I and II and human pancreatic lipase hydrolyze solutions of vinyl propionate or vinyl butyrate at high relative rates compared with emulsions of the same esters, although, in all cases, maximal activity is reached in the presence of emulsified particles, at substrate concentrations above the solubility limit. It appears that partially water-soluble short-chain vinyl esters are suitable substrates for comparing the activity of lipolytic enzymes of different origin and specificity toward esters in solution and in emulsion.

Production of extracellular lipases by Penicillium cyclopium purification and characterization of a partial acylglycerol lipase.

Penicillium cyclopium, grown in stationary culture, produces a type I lipase specific for triacylglycerols while, in shaken culture, it produces a type II lipase only active on partial acylglycerols. Lipase II has been purified by ammonium sulfate precipitation and chromatographies on Sephadex G-75 and DEAE-Sephadex. The enzyme exists in several glycosylated forms of 40-43 kDa, which can be converted to a single protein of 37 kDa by enzymatic deglycosylation. Activity of lipase II is maximal at pH 7.0 and 40 degrees C. The enzyme is stable from pH 4.5 to 7.0. Activity is rapidly lost at temperatures above 50 degrees C. The enzyme specifically hydrolyzes monoacylglycerols and diacylglycerols, especially of medium chain fatty acids. The sequence of the 20 first amino acid residues is similar to the N-terminal region of P. camembertii lipase and partially similar to lipases from Humicola lanuginosa and Aspergillus oryzae, but is different from Penicillium cyclopium lipase I. However, it can be observed that residues of valine and serine at positions 2 and 5 in Penicillium cyclopium lipase II are conserved in Penicillium expansum lipase, of which 16 out of the 20 first amino acid residues are similar to Penicillium cyclopium lipase I.

The C-terminal domain of pancreatic lipase: functional and structural analogies with c2 domains.

The 3D structure of pancreatic lipase (PL) consists of two functional domains. The N-terminal domain belongs to the alpha/beta hydrolase fold and contains the active site, which involves a catalytic triad analogous to that present in serine proteases. The beta-sandwich C-terminal domain of PL plays an important part in the binding process between the lipase and colipase, the specific PL cofactor. Recent structure-function studies have suggested that the PL C-terminal domain may have an extra role apart from that of binding colipase. This domain contains an exposed hydrophobic loop (beta5') which was found to be located on the same side as the hydrophobic loops surrounding the active site, and it may be involved in the lipid binding process. Indirect evidence for this new function of the PL C-terminal domain has been provided by studies with monoclonal antibodies directed against the beta5' loop. The catalytic activity of the PL-antibody complexes on water insoluble substrates decreased drastically, whereas their esterase activity on a soluble substrate remained unchanged. During the last few years, a number of protein structures (15-lipoxygenase, alpha-toxin from Clostridium perfringens) have been determined that contain domains with close structural homologies with the beta-sandwich C-terminal domain of PL. Generally speaking, these domains show structural homologies with the C2 domains occurring in a wide range of proteins involved in signal transduction (e.g. phosphoinositide-specific phospholipase C, protein kinase C, cytosolic phospholipase A2), membrane traffic (e.g. synaptotagmin I, rabphilin) and membrane disruption (e.g. perforin). Here it is proposed to review the structure and function of the C2 domains, based on the recent 3D structures and improved sequence alignments.

Biochemical and structural characterization of triacylglycerol lipase from Penicillium cyclopium.

An extracellular lipase, active on water-insoluble triacylglycerols, has been isolated from Penicillium cyclopium. The purified enzyme has a molecular mass of 29 kDa by gel filtration and SDS-polyacrylamide gel electrophoresis. It hydrolyzes emulsions of tributyrin, trioctanoin, and olive oil at the same rate as pancreatic lipase and shows very low activity against partial acylglycerols (monooctanoin and dioctanoin) and methyl esters. It is stable at 35 degrees C for 60 min and has maximal activity in a pH range of 8-10. Hydrolysis of triacylglycerols by P. cyclopium lipase is inhibited by detergents such as Triton X-100. Comparison of the sequence of the 20 first amino acid residues of P. cyclopium triacylglycerol lipase with other Penicillium lipases indicates a high homology with previously characterized lipases produced by P. expansum and P. solitum which are enzymes of comparable size and substrate specificity. Conversely, homology between P. cyclopium lipase and P. simplicissimum lipase, a nonspecific lipolytic enzyme, is low. Penicillium cyclopium triacylglycerol lipase shows no homology with P. camembertii lipase which is specific to monoacylglycerol and diacylglycerol.

Inactivation of pancreatic and gastric lipases by THL and C12:0-TNB: a kinetic study with emulsified tributyrin.

THL is a potent inhibitor of pancreatic (PPL) and gastric (HGL, RGL) lipases. Inactivation occurs preferentially at the oil/water interface (method B, C). In the aqueous phase (method A), the inhibition of HGL was accelerated by the presence of bile salts. C12:0-TNB, a disulfide reagent, specifically inactivates gastric lipases and had no effect on the pancreatic lipase (in the presence of bile salts) whatever the method used. The capacity of THL and C12:0-TNB to inactivate lipases using Methods B and C was found to depend directly upon the interfacial area of the system used. Consequently, inactivation can be reduced or prevented by further addition of a water-insoluble substrate which reduces the surface density of inactivator molecules. With a heterogeneous system of this kind, typical of lipolysis, the use of a classical Michaelis-Menten model is irrelevant and hence the traditional kinetic parameters (Km, KI, Vmax) are only apparent values.

Characterization of the cyclic AMP-dependent protein kinase from rat pancreas, further purification of the catalytic subunit, substrate specificity, effect of the pancreatic heat stable inhibitor.

In order to investigate the sequence of events triggered by cyclic AMP and cyclic GMP in exocrine pancreatic cells, the identification of the various protein kinases possibly present in this tissue is of major interest. Further analysis of the two cyclic AMP-dependent protein kinases previously reported [11] suggests that KI is a degraded form of KII. It is therefore likely that a single holoenzyme is present in exocrine cells. In addition no protein kinase, specifically stimulated by cyclic GMP, has been detected in any fraction obtained in the course of purification of the cyclic AMP-dependent protein kinase. A faster and more efficient method than the one previously described [11] allows the purification (5000 times) of the protein kinase catalytic subunit. Analysis of the subunit by sodium dodecyl sulphate polyacrylamide gel electrophoresis indicates a molecular weight of 40 000 +/- 1 000. The enzyme phosphorylates specifically histone H2B (Vm = 236 min(-1), Km = 1.15 10(-5) M) and to a lesser extent H2A, H5 and H1 (Vm = 55--77 min(-1), Km 5--25 10(-5) M). Histones H3 and H4 are not phosphorylated. The effect of the heat stable inhibitor, extracted from rat pancreas, on the phosphorylation of H2B has been investigated. The inhibition is of the non competitive type with respect to ATP. The inhibition at various histone concentrations cannot be described by the Michaelis-Menten equation.