Open and closed states of Candida antarctica lipase B: protonation and the mechanism of interfacial activation

Lipases (EC 3.1.1.3) are ubiquitous hydrolases for the carboxyl ester bond of water-insoluble substrates, such as triacylglycerols, phospholipids, and other insoluble substrates, acting in aqueous as well as in low-water media, thus being of considerable physiological significance with high interest also for their industrial applications. The hydrolysis reaction follows a two-step mechanism, or “interfacial activation,” with adsorption of the enzyme to a heterogeneous interface and subsequent enhancement of the lipolytic activity. Among lipases, Candida antarctica lipase B (CALB) has never shown any significant interfacial activation, and a closed conformation of CALB has never been reported, leading to the conclusion that its behavior was due to the absence of a lid regulating the access to the active site. The lid open and closed conformations and their protonation states are observed in the crystal structure of CALB at 0.91 Å resolution. Having the open and closed states at atomic resolution allows relating protonation to the conformation, indicating the role of Asp145 and Lys290 in the conformation alteration. The findings explain the lack of interfacial activation of CALB and offer new elements to elucidate this mechanism, with the consequent implications for the catalytic properties and classification of lipases.

regulating processes such as secretion, endocytosis, and signal transduction. Lipids are one of the major sources of energy, or serve as biological modulators and signal transducers (e.g., steroids, phosphatidylinositol derivatives, etc.) and as vehicles for carrying fat-soluble vitamins. The functional interaction between lipids and proteins is essential for such activities.
Lipases (EC 3.1.1.3) are ubiquitous hydrolases for the carboxyl ester bond of water-insoluble substrates such as triacylglycerols, phospholipids, and other insoluble substrates, acting in aqueous as well as in low-water media . For these reasons lipases are enzymes of considerable physiological signifi cance ( 1 ) with high interest also for their industrial applications ( 2 ).
The hydrolysis reaction, catalyzed by lipases, does not follow Michaelis-Menten kinetics, and presents a two-step mechanism, defi ned as interfacial activation, with adsorption of the enzyme to a heterogeneous interface and subsequent enhancement of the lipolytic activity ( 3 ). The crystal structures of human pancreatic lipase ( 4 ) and of lipase-procolipase complex ( 5,6 ), a fungal enzyme from Geontrichum candium ( 7 ), and a lipase from Pseudomonas glumae ( 8 ) showed a lid domain, either a loop or a helix, shielding the active site from the solvent and thus regulating the exposition of the substrate to the catalytic triad. It was postulated that a conformational change would occur in the presence of a lipid-water interface inducing interfacial activation. Thus the presence of a lid domain and interfacial activation Abstract Lipases (EC 3.1.1.3) are ubiquitous hydrolases for the carboxyl ester bond of water-insoluble substrates, such as triacylglycerols, phospholipids, and other insoluble substrates, acting in aqueous as well as in low-water media, thus being of considerable physiological signifi cance with high interest also for their industrial applications. The hydrolysis reaction follows a two-step mechanism, or "interfacial activation," with adsorption of the enzyme to a heterogeneous interface and subsequent enhancement of the lipolytic activity. Among lipases, Candida antarctica lipase B (CALB) has never shown any signifi cant interfacial activation, and a closed conformation of CALB has never been reported, leading to the conclusion that its behavior was due to the absence of a Supplementary key words fatty acid/metabolism • lipids/chemistry • enzymology/enzyme regulation • X-ray crystallography Lipids are, together with proteins and carbohydrates, the basic constituents of the architecture of a living organism. Lipids form membrane systems, which function as physical and chemical barriers separating aqueous compartments, activation ( 16 ). Taken together, these observations have led to the conclusion that this behavior was due to the absence of a lid regulating the access to the active site ( 16 ).
Whether or not CALB lacks a lid structure is matter of debate, according to several molecular dynamics simulation studies (17)(18)(19)(20). CALB has two ␣ -helixes surrounding the active site, namely ␣ 5 and ␣ 10 ( 14,15 ), which have been shown to be very fl exible regions and could work as a lipase lid by a relative motion between them either by increasing the working temperature ( 19 ) or by working in organic solvents ( 17 ). Such helix motions, in aqueous media, could not be large enough to produce variation in the hydrophobic surface and in the solvent-accessible area or large enough to prevent access to the active site ( 20 ). However, the CALB sequence stretch around the ␣ -helix 5 has been shown to signifi cantly infl uence the catalytic properties of the enzyme, including the enantioselectivity, which, consequently, seems not to be exclusively directed by the fi t of the substrate into the enzyme's active site pocket ( 18 ). The issue have long been used to distinguish lipases from esterases, where the interfacial activation phenomenon was not observed ( 9 ). More recently, the observation of a novel lid domain allowed the formulation of a hypothesis for a likely mode of interfacial activation of Candida antarctica lipase A ( 10 ), while a cutinase from Trichoderma reesei , showing interfacial activation behavior and the presence of the lid domain, has been reported as having the kinetic and structural features of a "true lipase" ( 11 ).
Among lipases, Candida antarctica lipase B (CALB) is the one that found the widest application in many industrial processes because of its high enantioselectivity, wide range of substrates, thermal stability, and stability in organic solvents ( 12,13 ). CALB belongs to the ␣ / ␤ hydrolase fold family with a conserved catalytic triad consisting of Ser, His, and Asp/Glu within an acyl binding pocket and a binding pocket for the moiety of secondary alcohols ( 14,15 ). However, a closed conformation of CALB has never been reported and CALB has never shown any signifi cant interfacial , where N atoms is the number of the atoms included in the refi nement, N refl is the number of the refl ections included in the refi nement, D max is the maximum resolution of refl ections included in the refi nement , compl is the completeness of the observed data, and for isotropic refi nement, N params ≈ 4N atoms ( 31 ). e Taken from BAVERAGE ( 28 ). f Taken from PROCHECK ( 30 ).
conserved Asp145 and Lys290 residues returns fi ve hits, potentially leading to a new sub-classifi cation of lipases.

Crystallization conditions and xenon derivatization
CALB was purchased from Hampton Research and crystallized without further purifi cation. Crystallization trials were performed at 293 K using the hanging-drop method using a Qia-gen™ EasyXtal 15-well plate. One microliter of a 15 mg/ml CALB solution in 20 mM Na(CH 3 COO) (pH 4.8) was diluted with 1 l of the precipitant solution, made of 200 mM Na(CH 3 COO) (pH 4.8), 20% (w/v) PEG4000, and 10-13% (v/v) 2-propanol. The drop was equilibrated by vapor diffusion against 500 ml of the precipitant solution. Protein crystals of native of the open and closed conformations of CALB is then not only related to the interfacial activation, but it has implications for the catalytic properties of the enzyme ( 18 ).
We report here the crystal structure of CALB in open and closed conformation refi ned at 0.91 Å resolution, representing the highest resolution structure of CALB to date, and of all currently available triacylglycerol lipase X-ray structures (EC 3.  For the xenon complex of native CALB, data collection was carried on at a wavelength of 2.25 Å, chosen to enhance the anomalous signal from the expected xenon atoms and warrant the necessary resolution for the experiments. At = 2.25 Å, close to the L I edge of Xe, the imaginary component of the anomalous dispersion f" is of 13.14 e Ϫ . Crystals of the xenon complex diffracted to 2.27 Å resolution with unit cell dimensions of a = 39.6 Å, b = 48.8 Å, and c = 71.6 Å and space group P1. The resolution of data for the xenon complex was limited by the geometry of the camera. The structures of native CALB and of the xenon complex were solved by molecular replacement using MOLREP ( 24 ) starting from the deposited structure of CALB as search model [Protein Data Bank (PDB): 1TCA ( 14 )]. The models were subjected to rigid-body minimization and subsequently to refi nement steps with REFMAC ( 25 ). For the native structure, anisotropic temperature factors refi nement was used and all the hydrogen generated, with automatic weighting scheme, given the value of 53 for ratio refl ection/atom. For the xenon complex, isotropic temperature factors refi nement was used. Manual rebuilding of the models were performed using the COOT graphic interface ( 26 ) by inspecting the electron density map, calculated with 2F obs Ϫ F calc or F obs Ϫ F calc coeffi cients and calculated phases from the model . Models were validated using the PDB_REDO web server ( 27 ). The native structure and its xenon complex were refi ned to an R factor and CALB appeared within 1 week and grew to a size of 0.2 × 0.4 × 0.5 mm 3 in 3 weeks. The xenon complex of native CALB was prepared by exposing the crystals to xenon pressure (10 atm) by using the Hampton Research xenon chamber available at the European Molecular Biology Laboratory (EMBL). Xenon incubation times were kept to around 15-20 min. After pressure release, the loop with the crystal was quickly plunged into liquid nitrogen for fl ash cooling and then transferred to the beam line. The crystal loop was maintained at 100 K with a cold nitrogen stream.

X-ray diffraction data collection and structure analysis
Diffraction data were collected at 100 K using synchrotron radiation at the EMBL P13 beamline at the Petra III storage ring, Deutsches Elektronen-Synchrotron (DESY), Hamburg (Germany ). The beam line was equipped with a Dectris Pilatus 6M detector and a MD2 goniometer (MAATEL-EMBL) with horizontal spindle axis. Crystals were cooled at 100 K with a cold nitrogen stream. The wavelength was set to 0.826 Å, using a Si(III) crystal monochromator (FMB-Oxford). Data were collected from four crystals oriented with different kappa angles while performing helical scans. The data were integrated using the program XDS and scaled with XSCALE ( 23 ). Crystals of the native enzyme diffracted to 0.91 Å resolution with unit cell dimensions of a = 39.67 Å, b = 48.94 Å, and c = 71.61 Å and space group P1. viation of 0.005 Å. The analysis reveals that, in chain A, the carboxylic group of Asp145 is protonated with bond lengths of C-O ␥ 2 ≈ 1.3 Å and of C-O ␥ 1 ≈ 1.2 Å, while in chain B it is deprotonated due to salt bridge formation, with bond lengths of the Asp145 carboxylic group of C-O ␥ 2 < 1.2 Å and of C-O ␥ 1 > 1.4 Å outside standard values. In the open conformation, Asp145 is placed at a distance of 3.8 Å above the indole group of Trp155, within hydrogen bond distance from the carbonyl group of the main chain of Ala141, the hydroxyl group of Ser150, and Thr158 side chains ( Fig. 4A ). The side chain of Asp145 shifts by 7.4 Å, R free (5% of data) of 11.1 and 13.2%, and of 13.9 and 19.4%, respectively. Data collection and refi nement statistics are reported in Table 1 . The protonation-state determination of the amino acid side chains followed the protocol described by Fisher et al. ( 22 ).

Open and closed conformations
The root-mean-square (rms) deviations of the C ␣ atoms of CALB monomers A and B, reported here, against the monomer A of the structure reported by Uppenberg et al. ( 14 ) (PDB code 1TCA) are 0.34 Å and 0.54 Å, respectively, as calculated using SUPERPOSE ( 28 ) over 310 residues ( Fig. 2A ). The rms deviation values of C ␣ atoms of monomer A against monomer B, presented here, are 0.53 Å . The rms deviation values ( Fig. 2A ) indicate that the secondary structure of monomer A is maintained and consistent with the classical open conformation previously observed ( 14,15,29 ), while the rms deviation values of C ␣ atoms for monomer B show a dramatic change between residues 140 and 147. The residue range Leu140-Ala141-Gly142-Pro143-Leu144-Asp145-Ala146-Leu147, corresponding to ␣ -helix 5, undergoes a conformational change from an ␣ -helix structure to a loop, as clearly shown by the assignment of the secondary structure computed with PROCHECK ( 30 ) ( Fig. 2B ). The electron density maps at 0.91 Å resolution clearly describe the residue range 140-147 of monomer A as an ␣ -helix, corresponding to an open conformation, and of monomer B as an unfolded loop, corresponding to a closed conformation ( Fig. 2C ). The residue range 140-147 presents Asp145 as the only amino acid with a charged side chain.
The fi nal model of CALB has a standard uncertainty for the positional parameters of all atoms of 0.015 Å, as calculated using Cruickshanck's dispersion precision indicator ( DPI ) factor ( 31 ) ( Table 1 ), which is six times lower than the values of the best currently reported CALB structures . Determination of the protonation state by bond length analysis ( 21,22 ) of the Asp and Glu amino acid side chains ( Fig. 3 ) resulted in a fi nal computed average standard de-  from the different C-O ␦ 1 and C-O ␦ 2 distances, where the C-O ␦ 1 bond, hydrogen bonded to the N-H of the backbone of His224, is 0.05 Å shorter than the C-O ␦ 2 facing toward a water molecule. When looking at the maps and the structures of His224, ND1 proton makes a hydrogen bond to the deprotonated Asp187. Ser105 proton hydrogen bonds to His224 NE2. In both monomer A and monomer B, a water molecule is shared between His224 and Ser105 ( Fig. 5A, B ).

DISCUSSION
The presence of a lid covering the active site in CALB has long been a matter of debate. The lack of structural evidence ( 14,15 ) and the lack of obvious interfacial activation during catalysis ( 16 ) have been the main argument against the existence of a lid in CALB. The ␣ -helix 5 and ␣ -helix 10 are shown here to be responsible for the closed conformation of CALB, and describe a completely unforeseen closing mechanism ( Fig. 6 ) at atomic level. In the open conformation, ␣ -helix 5 shows a series of aliphatic residues that line the channel leading to the active site ( 14 ) with the only polar residue, Asp145, making hydrogen reaching the side chain of Lys290 at a distance of 2.7 Å measured between Lys290 NZ atom and Asp145 O ␥ 1 atom, thus generating a salt bridge ( Fig. 4B, C ). The consequent large conformational change of the Leu140-Leu147 region results in the closing of the CALB active site, as shown by the space-fi ll representation ( Fig. 4D, E ).
Pressurizing the crystal with xenon probed the tightness of the closed lipase conformation. Three xenon atoms, with anomalous signal bigger than 4 , were located in monomer A in the catalytic site in close proximity to the catalytic triad. In particular, the three xenon atoms (8.3 , 6.3 , and 6 map level, respectively, refi ned with fractional occupancies) in monomer A are located above His224, in the hydrophobic region of the catalytic site defi ned by Leu144, Val154, Ile189, Leu219, Val221, Ala281, Ala282, Ile285, and Leu287 ( Fig. 4F ). No xenon atoms were observed in the active site of monomer B, thus confi rming that the conformational rearrangement of the residue range 140-147 prevents the access of any substrate to the active site ( Fig. 4G ). Xenon is also found in the hydrophobic cavity defi ned by Met83, Val101, Leu115, Ile121, Val125, Val37, and Ile187 in both monomer A (10.2 ) and monomer B (10.5 ). A further site is found next to Trp155 in monomer B (7.2 ).
The rms deviation values of C ␣ atoms ( Fig. 2A ) confi rm a reorientation of the loop residues between ␣ -helixes 8 and 9, localized on Leu219 and around ␣ -helix 10. These shifts are compatible with the protein-protein interaction and with the crystal packing. In particular, Leu219 of monomer A, in open conformation, is pointing toward ␣helix 10 of the symmetry-related molecule in closed conformation, in a proximal position to Lys290. Leu219 of monomer B is found above the active site of monomer A.

Active sites
In the active sites of both monomer A and monomer B, Ser105 is refi ned with multiple conformations, with the hydroxyl group swinging toward and away from NE2 of His224 and one water molecule. In both active sites, molecules of isopropanol are found tethered by the hydrogen bonding with the hydroxyl group of the side chain of Thr40, with the alkane chain positioned in the void between Asp134, Gln157, and Ile189. A network of waters is found within the active site and extending from the catalytic triad toward ␣ -helix 5 and ␣ -helix 10.
The bond length analysis ( 22 ) shows that, in monomer A ( Fig. 5A ), the side chains of Asp187 and Glu188 are deprotonated, with bond lengths of C-O ␥ 1 = 1.225(9) Å and C-O ␥ 2 = 1.246(8) Å for the Asp187 residue and of C-O ␦ 1 = 1.233(12) Å and C-O ␦ 2 = 1.225 (14) Å for the Glu188. The protonation state in His224 can be rationalized by analyzing the chemical environment as suggested by Fisher et al. ( 22 ). In monomer A, Asp187 forms hydrogen bond via O ␥ 2 with the His224 ND1 proton (2.79 Å), while a water molecule donates protons to NE2 (2.73 Å). In monomer B ( Fig. 5B ), the side chain of Asp187 is also deprotonated with bond lengths of C-O ␥ 1 = 1.277(9) Å and C-O ␥ 2 = 1.254(8) Å. The side chain of Glu188 appears to be protonated or partially protonated, as evident bonds with the side chains of Ser150 and Thr158. When closing the active site, the ␣ -helix 5 undergoes a dramatic conformational change to an unstructured loop while bringing the carboxylic group of Asp145 close to the side chain of Lys290, thus forming a salt bridge. The spatial rearrangement of Lys290 brings ␣ -helix 10 closer to new lid region, completing the closure of the catalytic cavity. The unfolding of ␣ -helix 5 results in the closing of the lid over the cavity, preventing the access of any substrate to the catalytic site, as shown by probing binding sites with xenon.
Xenon is able to bind to a protein in closed intramolecular hydrophobic cavities, accessible active sites, intermolecular cavities, and channel pores, as the result of noncovalent weak-energy van der Waals forces ( 32 ), by diffusing rapidly toward potential interaction sites through the solvent channels of a protein crystal ( 32 ). The relatively high atomic number and its anomalous dispersion of xenon, together with data collection at long wavelengths, make it possible to locate it within protein structures in crystallography ( 33 ).
The location of three xenon atoms in monomer A, in close proximity to the catalytic triad, and none in monomer B, confi rms that the conformational rearrangement of ␣helix 5 seals the active site. In our structure the lid region in monomer B is held in place by a close contact with the ␣ -helix 9' region of the symmetry related monomer A. No evident hydrogen bond pattern is present, and the interaction happens via hydrophobic surface patches from each monomer.
Comparison with previously reported CALB structures reveals reasons for the absence, so far, of structural evidence for this closed lid conformation. In the structures reported by Uppenberg et al. ( 14 ) (PDB identifi ers 1TCA, 1TCB, and 1TCC), the range Asp145-Pro143 would clash with residue Leu199 of a symmetry-related CALB monomer that is part of a face-to-face (i.e., binding site to binding site) dimer, if the lid segment would adopt the more extended closed conformation ( Fig. 2D ). Instead, in these structures, the lid segment adopts the more compact open ␣ -helical conformation. In our structure, however, the monomers are shifted with respect to each other such that for the monomer in closed conformation, this clash does not occur ( Fig. 2C ). A similar clash between residues Asp145-Pro143 in one monomer and Leu199 in the other monomer would occur for a closed lid conformation in an  disordered structure, manifested by a lack of continuous electron density ( 14 ). The lid region itself, with its primary (amino acid composition and number of residues) and secondary structure, can modulate the "interfacial activation behavior" of CALB, but also of lipases in general. Chimeras of CALB, generated by lid swapping, displayed large variations in kinetic constants and enantiomeric ratio for hydrolysis of p-nitrophenyl esters ( 18 ). So, while the previous studies pointed toward an activation of lipases based on an interfacial oil-water mechanism, in the case of CALB, the environmental pH could also play a role. At pH 7, CALB in solution is likely to oscillate between the open and the closed conformation; hence, the reason for the lack of interfacial activation of CALB must be searched for elsewhere, rather than in the lack of a lid region in the structure. It is noteworthy that the removal of the Asp145 from the ␣ -helix 5 region in the chimeras of CALB removed the activity pH dependence ( 18 ), clearly indicating a relationship between optimal activity pH and primary sequence of the lid region. The two variants, where the lid region was cloned from Neurospora crassa and from Gibberella zeae , on the other hand, yielded surprisingly low specifi c activities in the assay, with no clear pH dependency ( 18 ). In these two variants, the Asp145 is missing from the ␣ -helix 5 sequence.
A BlastP search ( 36 ) using CALB sequence as a search motif returns more than 50 results with sequence identity higher than 20% (results not shown). However, among these, only fi ve hits are found, which conserve Asp145 in the ␣ -helix 5 region and Lys290 after the ␣ -helix 10 region ( Fig. 7 ). Candida antarctica (LIPB_CANAR), Pseudozyma antarctica (M9MDK9_PSEA3), and Pseudozyma aphidis (W3VNS6_9BASI) share a sequence identity of more than 99%, while Ophiocordyceps sinensis (T5ACW5_OPHSC), Gibberella moniliformis (W7N1B5_GIBM7), and Fusarium oxysporum (W9LLN1_FUSOX) have a sequence identity of about 25% with Candida antarctica (LIPB_CANAR). The fact that Asp145 is rarely conserved in the sequence of lipases, would explain why, among lipases, CALB is the only lipase reported so far, which does not show the interfacial activation behavior typical of other lipases, but rather shows a "pH activation" behavior, as shown by Skjøt et al. ( 18 ). Our structure has shown that a salt bridge is formed between Asp145 and Lys290 upon lid closure . In the absence of Asp145, such a salt bridge cannot form, thus a pH-dependent behavior is not possible. Because of the presence of Asp145 in the ␣ -helix 5 region, CALB does represent a class of its own among lipases on the basis of the sequence homology and its characteristic behavior as lipase.
Several molecular dynamics simulation studies (17)(18)(19)(20) have shown that ␣ -helix 5 is a highly dynamic region and works as a lipase lid depending on the working temperature ( 19 ) or the working organic solvents ( 17 ). Derewenda et al. ( 37 ) suggested that the interfacial activation has both enzyme and substrate components, and that conformational changes in the enzyme are essential, but not suffi cient, for its maximum activity. Schrag et al. ( 38 ) proposed that the open or closed conformation of lipases was determined by the solution conditions, and perhaps additional CALB structure [PDB identifi er 1LBS ( 15 )] where the open conformation of ␣ -helix 5 has been modeled ( 15 ). Interestingly, in the same study, the authors obtain an alternative crystal form of CALB in the presence of Tween 80 at pH 4.0 that is unusually loosely packed and has a solvent content of over 60% ( 15 ). The hydrophobic surface around the active site is fully exposed to the solvent, and the lid segment is found in the open conformation. The absence of the closed conformation, in this case, cannot be attributed to steric reasons alone, as the crystal packing would allow for the lid to adopt the more extended closed conformation. However, it is tempting to speculate that the predominantly hydrophobic lid segment, ␣ -helix 5, upon exposure to polar solvents and/or high pH, prefers a more compact helical conformation, minimizing unfavorable solvent contacts. In our structure, however, the lid segment is better protected against solvent exposure by tighter crystal packing (solvent content 38%).
CALB was crystallized in conditions very similar to those reported by Uppenberg et al. ( 29 ). We buffered the precipitation solution to pH 4.8, a value that is just above the pK a of the side chains Asp (pK a = 3.86) and Glu (pK a = 4.07) residues . At pH 4.8, it is reasonable to expect that most of the carboxylic groups would be in the deprotonated state, subject to the local protein environments. Figure 3 , reporting the bond lengths of C-O ␥ 1 and C-O ␥ 2 of the carboxylic group of aspartic residues, and of C-O ␦ 1 and C-O ␦ 2 of the carboxylic group of glutamic residues, clearly shows that most Glu and Asp residues are, in fact, in a deprotonated state, but exceptions in a protonated state are also present. However, the protonation state of the charged residues in the active site of the two monomers ( Table 2 ) indicates that the charge inside the cavity is conserved. Residue Asp187 maintains its deprotonated state, Glu188 is deprotonated in monomer A and protonated in monomer B, while Asp145 is protonated in monomer A and deprotonated in monomer B ( Table 2 ) as it forms a salt bridge.
The bond lengths of Asp187 within the catalytic triad in open and closed conformations are consistent with the bond lengths of Asp102 in trypsin and of Asp32 in subtilisin reported by Fisher et al. ( 22 ), and further confi rm previous NMR ( 34 ) and crystallographic evidence ( 35 ) that the key catalytic proton is positioned between the two key catalytic residues, His and Asp, rather than being fi xed on the histidine, thus supporting the theory that the proton may, in fact, be involved in a low-barrier hydrogen bond ( 34 ).
Overall there is no change of the net charge on the protein in the open and closed conformations ( Table 2 ) following the destructuring of ␣ -helix 5 and the subsequent ion pairing of Asp145 with Lys290 from ␣ -helix 10.
The optimal pH for catalysis is 7, and the enzyme is stable in aqueous media in the range of pH 3.5-9.5 ( 12 ), values close to the pK a of the side chains of Asp145 and Lys290 residues discussed before. It was already noted that, in the monoclinic form P2 1 , raising the pH from 3.6 to 5.5 would cause dramatic changes to ␣ -helix 10, whose residues would change from an ordered structure to a the dielectric constants, on the basis of the crystallization conditions of various microbial lipases and their observed conformations. Vice versa, in nonaqueous media, the lack of interfacial activation suggested that the lid was predominantly closed, thus resulting in reduced enzymatic activity ( 39 ).
Here we have shown that the basis of the interfacial activation for lipases should not, therefore, be discussed on the basis of the presence or the absence of a lid region in an enzyme structure, which could go unobserved, but should rather be related to both the amino acid composition of the lid regions and how and why these regions react to changes in the dielectric value or ionic strength of the media . The interfacial activation phenomenon is then the convolution of the properties of the substrate, the media, and the enzyme with its fold and with its amino acid sequence.