Controlling Lipase Stereoselectivity via the Surface Pressure*

In the present study, the stereoselectivity of Rhizo-mucor miehei lipase, lipoprotein lipase, Candida antarctica B lipase, and human gastric lipase towards racemic dicaprin spread as a monolayer at the air-water interface was investigated. For this purpose we have developed a method with which the enantiomeric excess of the residual substrate can be measured in monomolecular films. The stereoselectivity, which is one of the main aspects of enzymic catalysis, was found to depend on the surface pressure of the substrate. With all four lipases tested, low surface pressures enhanced the stereoselectivity while decreasing the enzymes’ catalytic activity. Lipases, which lipolytic enzymes lipid-water interface, lipids, self-organize and orientate interfaces, are chiral molecules, chirality the molecular interactions between proteins and biomembranes. Membrane-like attractive model lipolytic ac-tivities the chirality of the lipid-water

In the present study, the stereoselectivity of Rhizomucor miehei lipase, lipoprotein lipase, Candida antarctica B lipase, and human gastric lipase towards racemic dicaprin spread as a monolayer at the airwater interface was investigated. For this purpose we have developed a method with which the enantiomeric excess of the residual substrate can be measured in monomolecular films. The stereoselectivity, which is one of the main aspects of enzymic catalysis, was found to depend on the surface pressure of the substrate. With all four lipases tested, low surface pressures enhanced the stereoselectivity while decreasing the enzymes' catalytic activity.
Lipases, which are lipolytic enzymes acting at the lipidwater interface, display stereoselectivity towards glycerides and other esters (1-3). Biological lipids, which self-organize and orientate at interfaces, are chiral molecules, and their chirality is expected to play an important role in the molecular interactions between proteins and biomembranes. Membranelike lipid structures, such as monolayers, provide attractive model systems for investigating to what extent lipolytic activities depend upon the chirality and other physicochemical characteristics of the lipid-water interface.
The mechanism whereby an enzyme differentiates between two antipodes of a chiral substrate may be influenced by physicochemical factors such as temperature ( 4 5 ) or solvent hydrophobicity (6), which can affect the reaction stereoselectivity. In the present study, we investigated the assumption that the stereoselectivity, which is one of the basic factors involved in enzymic catalysis, may be pressure-dependent. When working with bulk solutions, the external pressure is not a practical variable because liquids are highly incompressible, whereas the monolayer surface pressure is easy to manipulate. To establish the effects (if any) of the surface pressure on the stereochemical course of the enzyme action, during which optical activity is generated in a racemic substrate insoluble in water, we have developed a method with which the enantiomeric excess of the residual substrate can be measured in monomolecular films.
Here we present the results of a study on the influence of the surface pressure on the enantioselectivity of lipases in lipid monolayers during kinetic resolution of racemic dicaprin. With all four lipases tested, low surface pressures enhanced the stereoselectivity while decreasing the catalytic activity. This finding which to our knowledge is unprecedented, should help to elucidate the mode of action of water-soluble enzymes on water-insoluble substrates.
Kinetic Experiments on Monolayers-Before each utilization, the Teflon trough used to form the monomolecular film was cleaned with water, then gently brushed in the presence of distilled ethanol, washed again with tap water, and finally rinsed with double-distilled water. The aqueous subphase was composed of 10 mM Tris/HCl, pH 8.0 (or pH 5.0 in case of HGL), 100 mM NaC1, 21 mM CaC12, and 1 mM EDTA. The buffers were prepared with double-distilled water and filtered through a 0.45-wm Millipore filter. Any residual surface-active impurities were removed before each assay by sweeping and suction of the surface. Kinetic experiments presented in Fig. 2 were performed with a KSV-2200 barostat (KSV, Helsinki) and a 29-cm long, 17.5cm wide, 350-ml total volume single compartment Teflon trough with four 2.5-cm magnetic stirrers operating at 250 rpm. The experiments presented in Figs. 1 and 3 were performed in a two-compartment trough, as explained in the legends. The trough was equipped with a mobile Teflon barrier, which was used to compensate for the substrate molecules removed from the film by enzyme hydrolysis (monocaprin and capric acid are soluble in water), thereby keeping the surface pressure (n) constant. The latter was measured using a Wilhelmy plate (perimeter 3.94 cm) attached to an electromicrobalance, which was connected in turn to a microprocessor controlling the movement of the mobile barrier. The reactions were performed at ambient temperature (25 "C). The enzyme solution (5-25 ~l ) was injected through the film with a Hamilton syringe over the 4 magnetic stirrers. Reactions were stopped when 50% of the substrate had been hydrolyzed. At this point the film of the residual, nonhydrolyzed substrate was aspirated using a water pump into a conical 50-ml flask, the cone of which had been drawn out into a tube 1.5 cm in length and 0.5 cm in diameter. Before aspiration, 2 ml of CHCL and 60 pi of 0.12 M HC1 (in order to reach a final pH of around 2) were placed in the flasks, which were kept on ice. These steps were taken to inactivate the enzymes immediately on aspiration. A further 2 ml of cold CHCI, was aspired after harvesting the film to recover the whole quantity of dicaprin from the small glass tube used as an aspirator.
Deriuatization of Diglycerides-The flasks containing the aspired water phase, dicaprin, and CHCI, were gently agitated and left on ice for a few minutes. Water phase was then entirely removed with a Pasteur pipette and the CHC13 phase was rinsed with 2 ml of doubledistilled H20. The water phase was discarded again and CHCL was evaporated on a water pump. The flasks were rinsed with 0.5 ml of CHC13 to ensure that all the dicaprin was collected in the tube at the base of the flask and the CHCI, was evaporated again. This was repeated for the second time with 0.1 ml of CHCl,, CHCla was evaporated on a water pump and the dessicator containing the flasks was then connected to an oil pump to dry any residual traces of water. Teflon stirring bars (1 X 1-mm) were placed in the tubes of the flasks, and 25 pl of distilled heptane and 10 p1 of (+)-(R)-phenylethyliso- The abbreviations used are: HGL, human gastric lipase; HPLC, high performance liquid chromatography; N, newton; ee%, enantiomeric excess percent. cyanate were then added. All flasks were sealed with Teflon film and ground stoppers. The flasks were left under stirring for 48 h at room temperature and the excess (+)-(R)-phenylethylisocyanate was then evaporated overnight under a vacuum at 0.05 mm Hg. The diasteromeric dicaprin-phenylethylcarbamates thus obtained were taken up in 60 pl of distilled heptane, which sufficed to perform three or four separations of the sample on HPLC.
HPLC Separations-The carbamates were injected on HPLC to separate the pairs of diastereoisomers on a Beckman Ultrasphere 5pm column (10 x 25 cm) with 0.4% ethyl alcohol/heptane as eluent at a flow rate of 3.5 ml/min flow rate. Both heptane and ethyl alcohol were distilled prior to use. A light scattering Cunow DDL 11 (12) detector was used to monitor the separations. The enantiomeric excess percentages were calcdated after automatic integration of the corresponding peak areas and are given in Figs. 1 and 2 as a function of interfacial pressure.

RESULTS AND DISCUSSION
Chiral, optically pure (7-9), or racemic (10) compounds, spread as monolayers at the air-water interface, have been used previously to study enzyme kinetics under biomembrane simulating conditions. To our knowledge, however, the problem of the steric course of enzyme action on racemic lipid monolayers has never previously been approached experimentally. The data on enzymic stereopreferences obtained with enantiomerically pure monomolecular films (7-9) cannot be directly extrapolated to racemates, since the physicochemical properties of optically pure enantiomers and racemates in monolayers may be very different (11,12) and this in turn may influence the enzyme behavior, which is controlled by molecular recognition processes.
The chromatographic method for resolving enantiomeric diglycerides developed recently a t our laboratory (13) have now been scaled down to yield insights on the interfacial enzymic stereochemistry of monomolecular films. T h e ee% (enantiomeric excess percent) measurements were adapted to working with very small quantities (about 5 pg) of the initially racemic substrate forming the monolayer. The substrate chosen here was 1,2-rac-dicaprin, since it forms stable monomolecular films at surface pressures as high as 40 mN. m" and the products of lipolysis, monocaprin, and capric acid, are soluble in the water subphase, which means that they do not disturb the monolayer during the reaction. 1,2-Dicaprin shows a liquid-expanded phase behavior (no phase transition) within the K range of 5-40 mN.m" a t enantiomer compositions varying between 100% 1,2-sn-dicaprin and racemate (results not shown). The four enzymes used, R. miehei lipase, bovine milk lipoprotein lipase (LPL), Candida antarctica B lipase, and human gastric lipase (HGL) were chosen because they specifically catalyze the hydrolysis of primary ester bonds in glycerides, in positions sn-1 or srz-3.
The R. miehei lipase ee% measurements are given in Fig. 1 as a function of the reaction yield. With this lipase, the residual substrate recovered from the surface was found to have an optical purity as high as 80% ee at 50% yield and 90% ee at 75% yield. The results of this experiment, performed at the arbitrarily chosen surface pressure of 30 mN. m" confirmed the accuracy and reproducibility of the method.
The main advantage of the monolayer technique, however, is that by changing the surface pressure it becomes possible t o modulate the conformation and interactions of the filmforming molecules (12). We expected the enzyme-substrate chiral recognition to be affected by these changes and to result in surface pressure-dependent ee% variations.
The experimental results confirmed our expectations (Fig.   2). We noted that in all the lipases studied, the reaction stereoselectivity was governed by the surface pressure and dicaprin as substrate in a two-compartment zero order trough (8). The surface coverage by the enzymes was assumed to range between 20 pg/cm2 and 80 ng/cm2 (10). and lo%, respectively. Unlike the stereoselectivity, the enzyme activity showed the highest values at high surface pressures (Fig. 3).
Any interpretations of the observed phenomenon can only be speculative in the absence of analytical methods with which to determine the changes occurring at the molecular level in the lipid film and in the associated water layer and enzyme as a function of the surface pressure. Enzyme-substrate molecular recognition is a critical step in enzyme-catalyzed reactions. Formation or cleavage of covalent bonds in the substrate is possible only if a sufficient number of non-bonding enzyme-substrate interactions have occurred for the steric control of the reaction to be possible (14). Lipases are usually capable of catalyzing the hydrolysis of monolayers of both of the optical antipodes of their substrates, but at different rates (7,8). This means that the efficiency of the molecular recognition is defined by the absolute configuration at the chiral center of the substrate.
On the basis of crystallographic data (15,16) we speculate that the aliphatic chain of the fatty acid to be cleaved from glyceride position 1 (or alternatively, position 3) occupies a hydrophobic channel within the enzyme interior. Furthermore, we hypothesize that the ester group and its aliphatic chain in position 2 (chiral carbon) in both enantiomers may fit a specific non-polar pocket within the molecular recognition site (the fact that the ester group in position 2 is rarely hydrolyzed supports this idea). In the better fitting enantiomer, the carbinol group of the chiral center may also be stabilized by some specific interactions lacking in the other enantiomer.
The results of the crystallographic studies cited above (15), as well as NMR, UV, and fluorescence spectroscopy data (17-19), show the ability of lipolytic enzymes to undergo conformational changes upon lipid binding. Assuming that the first step in interfacial catalysis, i.e. lipid binding, is not stereoselective, then the anticorrelation observed between the cata-lytic activity and the ee% depending on the surface pressure can be explained as follows. The presumed surface pressuredependent enzyme conformational changes (20) may result in a deterioration of the molecular recognition of both enantiomers due to the progressive loss of the residue-specific interactions and the concomitant decrease in catalytic activity at low surface pressures. The deteriorating molecular recognition will, however, have a relatively stronger destabilizing effect on the already less well fitted enantiomer, which may then lose the catalytically efficient orientation before its antipode does so, thus enhancing the reaction stereoselectivity.
The differences in the catalytic activity and ee% patterns between the four enzymes probably reflect their individual susceptibility to undergo different conformational changes due to their flexibility and/or interfacial binding efficiency, i.e. a flexible enzyme sticking strongly to the interface would undergo changes more easily than a rigid enzyme with a limited capacity to interact with the interface.
The preliminary results on several other lipases tested at our laboratory' suggest that the dependence of lipase stereoselectivity on surface pressure is a general phenomenon and may therefore be applicable to enantioselective biocatalysis as well as being relevant in general to the enzyme reactions associated with biological membranes.