The present project constitutes a further qualitative jump in the research project previously carried out in the framework of the BRIDGE T-Lipase programme (BIOT-CT91-0274). The aim of this project is to characterize the interfacial binding step and the monomeric lipid binding step at the molecular level during the enzymatic lipolysis of glycerides by lipases the 3-D structure of which was either determined within the previous BRIDGE project (Pancreatic Lipase/Colipase complex) or will be determined within the present project: Pancreatic Lipase Related Proteins (PLRP), Gastric Lipase (GL), Bile-Salt Stimulated Lipase (BSSL), Lipoprotein Lipase (LPL) and Hormone Sensitive Lipase (HSL). The expected results of this research project will be a detailled molecular and atomic description of how lipases specifically and very efficiently bind to and hydrolyse water-insoluble substrates. We hope to be able to explain more fully the processes involved in dietary fat digestion in Humans as well as the overall triglyceride homeostasis. Structure/function studies may therefore contribute to designing specific drugs for counteracting these important human enzymes.
Studies on Pancreatic Lipase, PLRPs and Gastric Lipase at Dr. Verger's laboratory:
We designed chimeric mutants by exchanging the lid domains of the classical human pancreatic lipase (HPL) and the guinea pig pancreatic lipase related protein 2 (GPLRP2). This later enzyme possesses naturally a large deletion within the lid domain and is not activated by lipid-water interfaces. Furthermore, GPLRP2 exhibits phospholipase A1 and lipase activities in the same order of magnitude whereas HPL has no significant phospholipase activity and displays a clear interfacial activation. An HPL mutant [HPL(-lid)] with GPLRP2 mini-lid domain does not display interfacial activation. Its specific activity towards triglycerides is however dramatically reduced. A GPLRP2 mutant [GPLRP2(+lid)] with HPL full-length lid domain is not interfacially activated and its lid domain exists probably under a permanent open conformation. Therefore, the phenomenon of interfacial activation in HPL is not only due to the presence of a full-length lid domain but also to other structural elements which probably allow the existence of stabilized closed and open conformations of the lid. GPLRP2(+lid) phospholipase activity is significantly reduced as compared to GPLRP2, whereas its lipase activity remains at the same level. Therefore, the lid domain plays a major role in substrate selectivity and can be considered as part of the active site. However, the presence of a full-length lid domain is not sufficient to explain the absence of phospholipase activity in HPL since HPL(-lid) does not display any phospholipase activity. We also produced a chimeric GPLRP2 mutant in which the C-terminal domain was substituted by the HPL C-terminal domain (N-GPLRP2/C-HPL chimera). The colipase effects i.e. anchoring and stabilisation of the lipase at the interface, are clearly observed with the chimera, whereas GPLRP2 is insensitive to colipase. The kinetic characterization of this chimera reveals for the first time that the interfacial stability of pancreatic lipases depends on the structure of the C-terminal domain.
The chimeric mutants of pancreatic lipases have been used to discriminate four monoclonal antibodies specific to HPL (mAbs 81, 146, 315, 320) and raised in the mouse. The mAb 146 is specific for the N-terminal domain of HPL, whereas the three mAbs 81, 315, 320 are specific for the C-terminal domain of HPL. Using the mAb 146 and rabbit anti-HPL polyclonal antibodies, an ELISA was developped to assay specifically HPL in biological fluids.
A catalytically active recombinant human gastric lipase (HGL) was produced in the baculovirus insect cell system. The level of HGL secretion was about 32 mg/l of culture medium. Replacing the HGL signal peptide by the human pancreatic lipase signal peptide resulted in a 2-fold increase in the amount of lipase secreted from the insect cells. The protein was purified. It showed an apparent molecular mass of 45 kDa under SDS/PAGE analysis, which indicates that the level of glycosylation was lower in the insect cells than in the human gastric chief cells. The maximum specific activities of the recombinant lipase were 434, 730 and 562 units/mg using long-chain (Intralipide TM), medium-chain (trioctanoyl glycerol) and short-chain (tributyroyl glycerol) triacyl glycerols, respectively. Experiments were conducted to mutate each of the four putative glycosylation sites in order to further reduce the heterogeneity of HGL preparations. The Ser or Thr codon in the potential N-glycosylation sites were changed to Ala codon. When all four N-glycosylation sites were eliminated, the production of rHGL was unchanged, relative to that of the wild-type rHGL. This suggests that N-glycosylation has no major effect on biosynthesis and secretion. N-deglycosylated rHGLs display the same specific activity as the wild-type rHPL.
Using the oil drop technique, we studied the effects of colipase and bile salts on the rate of hydrolysis of soybean oil by human pancreatic lipase (HPL) as well as on the interfacial binding. Upon continuously recording the decrease in the interfacial tension with time, a 10-15 fold increase in the HPL activity was found to occur in the presence of colipase. The catalytic rate constants of hydrolysis measured at the oil drop surface were found to be of the same order of magnitude as those obtained with monomolecular films spread at the air-water interface. Biotin labelled HPL (HPL*) was used to determine the amount of adsorbed enzyme using an ELISA test. Less than 1% of the total amount of injected HPL* molecules was found to have adsorbed to the oil-water interface, and no significant effects of colipase on HPL* binding were observed. No significant changes in the hydrolysis rates or the binding of HPL* were detected in the presence of bile salts at concentrations ranging from below their critical micellar concentration (CMC), up to 100 uM. At the oil/water interface, in the absence or presence of bile salts below their CMC, it can be concluded that the colipase is a true lipase cofactor, i.e, it increases the enzyme turnover and does not affect the interfacial lipase adsorption.
Studies on Hormone-sensitive lipase (HSL) from the group of Dr. Cecilia Holm at Prof. Belfrage's laboratory:
The activation mechanism of hormone-sensitive lipase (HSL) upon phosphorylation has been studied Lipolytic activity in the adipose tissue is dramatically increased by the action of catecholamines. The mechanism underlying such activation is believed to be the phosphorylation of HSL by cAMP-dependent protein kinase. This phosphorylation triggers a 2-3 fold activation of the enzyme activity as measured in vitro with phospholipid-stabilized emulsions of triolein or cholesterol oleate. However, in vivo stimulation of adipose tissue with catecholamines results in a 20-50 fold activation of lipolysis. The reason for this discrepancy is not due to differences in the maximal lipolytic rate (which is similar in both systems), but rather in the artificially high basal activity (that of unphosphorylated HSL) obtained in vitro compared with the virtually zero lipolysis observed in the unstimulated adipocyte in vivo. One possible explanation for the discrepancy in the fold activation of HSL observed in vitro and in vivo is the disparate physical properties of the artificial and natural substrates. Based on this hypothesis, we have performed several experiments using the lipidic monomolecular film technique at Dr. Verger's laboratory, in order to investigate the behaviour of HSL towards different water-lipid interfaces. The results from the monolayer measurements can be summarized in the following conclusions: a) the interaction of HSL with the water/lipid interphase is greatly affected by the nature of this interphase. b) Phosphorylation of HSL enhances the interaction of the enzyme with less favored interphases (e.g., pure diglyceride monolayers in the lipidic monomolecular film), but no significant differences are observed between phosphorylated and non-phosphorylated HSL when interphases that are favorable for the binding of HSL (e.g. a mixed monolayer of dicaprin and phosphatidylcholine) are used to present the substrate.
The second aspect of our collaboration in the BIOTECH project is the production of recombinant human HSL, and fragments thereof, in order to fully characterize the enzyme and to elucidate its 3D structure by crystallographic means, in collaboration with the Dr. Cambillau and his colleagues in Marseille. A baculovirus/insect cell expression system for intact human HSL has been established, as well as a purification scheme, yielding homogenous recombinant HSL in milligram amounts. Several batches of protein have been purified, in different types of detergents. Attempts to crystallize HSL have been made in Dr. Cambillau's laboratory. No crystals have been obtained. One major problem is the limited solubility of HSL and its strict detergent requirement. In order to overcome the solubility problems of HSL, recombinant baculovirus for the testicular isoform of human HSL, containing an additional 300 amino acids (mostly hydrophilic) compared to the adipocyte isoform, has been generated, as well as recombinant baculovirus for the catalytic domain of human HSL (encoded by exons 4-9). The testicular isoform of HSL is expressed in a stable form and exhibits enzyme activity. The catalytic domain of HSL is expressed in a stable form, but exhibits neither lipase nor esterase activity. The reason for the latter is not known. Both the testicular isoform of HSL and the catalytic domain have been expressed with a His-tag, to facilitate the purification of these proteins, and purification schemes are presently being worked out.
Based on secondary structure comparisons with lipases of known crystal structure, we have recently constructed a three-dimensional model for the catalytic domain of HSL. The model reveals the topological organization of the catalytic domain, predicts the components of the catalytic triad and suggests a localization of the regulatory module, and provides an excellent tool for the design of molecular biology and crystallographic approaches to study the structural-function relationships of HSL. The proposed catalytic triad from the model has been confirmed by site-directed mutagenesis.
In the general context of the project, following the line of structural studies of lipases of medical interest, we have extended our original plans to a new enzyme, the monoglyceride lipase (MGL). MGL works in coordination with HSL and is required to obtain a complete degradation of adipocyte triacylglycerol. From the point of view of understanding the structural basis for the substrate specificity for lipases, MGL is very interesting, since it has an absolute specificity for monoglycerides. MGL has been purified from rat adipose tissue and the sequence of several tryptic peptides obtained. Oligonucleotides were synthesized based on these tryptic sequences and used to obtain a partial MGL cDNA probe by RT-PCR from mouse adipose tissue. The partial cDNA probe was used as a probe to screen a mouse adipose tissue cDNA library. Positive clones containing the entire coding sequence for MGL were obtained. The cDNA encodes a protein of 303 amino acids, corresponding to molecular mass of about 33 kDa. All tryptic sequences were identified in the amino acid sequence, thus confirming the identity of the cDNA. The sequence is not closely related to any other known lipase, but contains the consensus sequence for an active site serine (GXSXG). The serine of this motif has shown to be essential for lipase and esterase activity, and is therefore, together with an identified essential histidine residue, proposed to be part of the catalytic triad. Recombinant baculovirus for MGL, with and without His-tag, have been generated and purification schemes for expressed MGL are presently being worked out.
Studies on Bile salt-stimulated lipase (BSSL) at Dr Blädackberg's laboratory:
Our main aim was to obtain variants of bile salt-stimulated lipase (BSSL) suitable for determination of the 3-D structure by X-ray crystallography. Human BSSL is highly glycosylated with a number of O-linked sugars localized to the C-terminal repeat part, 16 near identical repeats of 11 amino acid residues, of the molecule and with a single N-linked sugar chain. We have, using site-directed mutagenesis, constructed new variants of the enzyme lacking the N-linked sugar and all, or all but two of the repeats. These variants, as well as corresponding, previously constructed variants where the N-linked sugar remained were expressed in a murine fibroblast cell-line (C-127) using a bovine papilloma-based vector and purified to homogeneity. We have worked out conditions to get optimal solubility of the enzyme variants although this proved difficult due to the removal of hydrophilic carbohydrate moities. We have thus prepared in excess of 10-20 mg of the respective four variants which has been forwarded to Dr Cambillau's laboratory. All variants produced has been carefully characterized in order to make sure that they were catalytically active with properties similar to the native enzyme purified from human milk.
Studies on Lipoprotein lipase (LPL) at Dr Olivecrona's laboratory:
We have concentrated on production of bovine lipoprotein lipase for crystallization attempts in the laboratory of Dr Christian Cambillau in Marseille. Many approaches have been tried (reviewed by Cambillau), but have this far not been successful. The outcome of the project is that we now know what is not possible to do with LPL with regard to crystallization. As a development towards more soluble forms of LPL we have started production and characterization of monoclonal antibodies to be used for co-crystallization.
In other projects we have initiated a collaboration with Prof. Kokotos, Athens, on production of better active site inhibitors for LPL than those which are now available. We have also studied site-directed mutants of LPL affecting interactions with lipoproteins and with specific cell surface receptors. Creation of more mutants are in progress using expression in eucaryotic cells.
Much effort have been devoted to studies on the interaction of LPL with heparin, with heparan sulfate proteoglycans and with well characterized fragments of heparin using the surface plasmon resonance technique. We have also used this system for direct studies of the binding of LPL to different classes of lipoproteins.
Crystallogenesis, X-ray crystallography, Molecular modeling at Dr. Cambillau's and Dr van Tilbeurgh's laboratory: A novel structure of human pancreatic lipase/colipase complexed with a C11 alkyl phosphonate inhibitor has been obtained at 2.46 Å resolution.
The crystal structure of the N-GPLRP2/C-HPL chimera has been solved by molecular replacement and refined at 2.1Å resolution to a crystallographic R factor of 0.188. This lipase belongs to the a/b hydrolase fold and shows a high structural homology with classical pancreatic lipases, with the exception of the lid domain which is made of 5 amino acid residues (mini-lid) instead of 23. The model reveals a fully accessible active site, closely related to those of serine esterases, except for an unusual geometry of the catalytic triad. The b5 loop is found in a conformation superimposable with the one observed in the open form of the human pancreatic lipase (HPL). Part of the b9 loop, which stabilizes the lid domain in the closed conformation of the classical HPL, is totally exposed to the solvent and is not visible in the electronic density map. A sequence comparison reveals important deletions in the lid and the b9 loop of the structurally related hornet phospholipase A1 (DolmI). We built up a model of DolmI PLA1 based on the 3D structure of the N-GPLRP2/C-HPL chimera. A set of models of phospholipids and galactolipids at the active site of GPLRP2 have also been constructed. These investigations highlight the roles of three surface loops (the lid domain, the b5 loop and the b9 loop) in substrate selectivity towards triglycerides, phospholipids and galactolipids.
The crystallizations of LPL and BSSL resulted only in micro-crystals not suitable for X-ray diffraction. Crystallographic trials with HSL have been started. The first crystals of recombinant HGL have been obtained and their characterization is in progress.
Synthesis and assay of lipase inhibitors at Prof. Buono's laboratory, in collaboration with Dr. Verger:
In an attempt to further characterise the active site and catalytic mechanism, we synthesised a C11 alkyl phosphonate compound. This compound is an effective inhibitor of pancreatic lipase. The crystal structure of the pancreatic lipase-colipase complex inhibited by this compound was determined at a resolution of 2.46 Å. In order to understand the stereochemical course of the nucleophilic substitution on the phosphorus atom by the Og of the active serine, it is important to possess optically pure alkyl phosphonate compounds with well known absolute configuration. In this way, we have recently developed a new diastereoselective synthetic method of chiral phosphonates with p-nitrophenoxy as leaving group. Using this method we have obtained methyl p-nitrophenyl alkyl phosphonates in 95% ee. The absolute configuration of the phosphorus atom is established by chemical correlation methods.
We also have synthesised optically pure phosphonate compounds analogous to triglycerides and tested each of the four diastereomers obtained on inhibition studies with HPL and HGL by using the monolayer technique. The results obtained shown an important influence of the alkyl chain length on the phosphorus atom upon the inhibitory effects. This research claims, however, some support to the interfacial chiral recognition hypothesis at a monomolecular homochiral film, but further experiments are required to substantiate this attractive theory.
In order to improve the molecular fit in the two hydrophobic clefts of the active site, we have designed new chiral phosphonate inhibitors which will soon be tested with HPL and HGL by using the monolayer technique.
Synthesis and assay of lipase inhibitors at Prof. Kokotos' laboratory, in collaboration with Dr. Verger:
The group has developed methods for: (a) the selective one pot conversion of carboxylic acids - including saturated and polyunsaturated fatty acids - into alcohols through the reduction of the corresponding fluorides with sodium borohydride, (b) the synthesis of lipidic 1,4-diols starting from 1,2-diols, (c) the synthesis of chiral triamines and diamines from natural and lipidic alpha-aminoacids. The methods have been applied for the synthesis of chiral pseudoglycerides. Two of them have been sent to Prof. Buono's lab as starting material for the preparation of the corresponding phosphonate inhibitors of lipase. Chiral sulfonates, derivatives of lipidic taurines, have ben prepared and will be tested as inhibitors of lipases (analogs of deacylation step).
A series of lipid mimetics (amides of lipidic a-amino acids, lipidic dipeptides, lipidic diamines, amino alcohols, 3-amino-1,2-diols) have been synthesized. These compounds have been tested in London (at W. Gibbons' lab) as inhibitors of pancreatic lipase and phospholipase A2 and in Marseilles (at R. Verger's lab) using the lipid monomolecular film technique.
A series of lipophilic alpha-keto amides have been prepared in Athens and tested in Marseilles (at R. Verger's lab) as inhibitors of pancreatic lipase and cutinase by the pH-stat method. Two representative compounds have been sent to Dr. Olivecrona's lab for studies on lipoprotein lipase.
Synthesis and assay of lipase inhibitors at Prof. Gibbons'laboratory, in collaboration with Dr. Verger:
Synthetic molecules which could potentially inhibit lipases and/or phospholipases were tested using radiometric, pH titration and monolayer assays. Thus 40 lipidic molecules, including lipidic amino alcohols, lipidic diamines, derivatised lipidic amino acids, and dipeptides were compared in two or more of these assays. Two thirds of these had IC50 - 200-300 uM in the first two assays but when these were further tested in monolayer assays none exhibited significant inhibitory lipase activity. A few compounds however, exhibited low inhibitory power for pancreatic pla2. Several compounds which stimulated porcine pancreatic phospholipase in the first two assays were inactive in the monolayer assays. A natural lipase inhibitor was found in sunflower, pine, peanut and soyabean seeds. The sunflower inhibitor was purified by gel permeation, ion exchange and preparative IEF. Its molecular mass was Ca 70 kd and pl 4.6 +/- 0.4. These inhibitors, once their inhibitor mechanisms, constants, and structural parameters are determined, may prove of industrial and technical significance.
MAJOR SCIENTIFIC BREAKTHROUGHS:
Structure at 2.46 Å resolution of the human pancreatic lipase-colipase complex inhibited by a C11 alkyl phosphonate inhibitor.
Expression in insect cells of several pancreatic lipase mutants
Structure at 2.1 Å resolution of the GPLRP2/HPL chimera and modeling of the structurally related hornet PLA1.
Expression in insect cells of a catalytically active recombinant human gastric lipase and several mutants.
Studies on the activation mechanism of hormone-sensitive lipase (HSL) upon phosphorylation.
Overexpression of human HSL in a baculovirus system
Three-dimensional model for the catalytic domain of HSL.
Molecular cloning and large-scale expression of monoacylglycerol lipase.
Construction of BSSL and LPL mutants to improve cristallogenesis
Synthesis of new phosphonate and a-keto amide inhibitors.
Funding SchemeCSC - Cost-sharing contracts
221 00 Lund
15771 Zographos, Athens
WC1N 1AX London
901 87 Umee