Fluorescence and FTIR study of pressure-induced structural modifications of horse liver alcohol dehydrogenase (HLADH) Marie Trovaslet 1 , Sandrine Dallet-Choisy 1 , Filip Meersman 2 , Karel Heremans 2 , Claude Balny 3 and Marie Dominique Legoy 1 1 Laboratoire de Ge ´ nie Prote ´ ique et Cellulaire, Universite ´ de La Rochelle, France; 2 Department of Chemistry, Katholieke Universiteit Leuven, Belgium; 3 INSERM U 128, Montpellier Cedex 5, France The process of pressure-induced modification of horse liver alcohol dehydrogenase (HLADH) was followed by measuring in situ catalytic activity (up to 250 MPa), intrinsic fluorescence (0.1–600 MPa) and modifications of FTIR spectra (up to 1000 MPa). The tryptophan fluor- escence measurements and the kinetic data indicated that the pressure-induced denaturation of HLADH was a process involving several transitions and that the observed transient states have characteristic properties of molten globules. Low pressure (< 100 MPa) induced no important modification in the catalytic efficiency of the enzyme and slight conformational changes, characterized by a small decrease in the centre of spectral mass of the enzyme’s intrinsic fluorescence: a native-like state was assumed. Higher pressures (100–400 MPa) induced a strong decrease of HLADH catalytic efficiency and fur- ther conformational changes. At 400 MPa, a dimeric molten globule-like state was proposed. Further increase of pressure (400–600 MPa) seemed to induce the dissoci- ation of the dimer leading to a transition from the first dimeric molten globule state to a second monomeric molten globule. The existence of two independent struc- tural domains in HLADH was assumed to explain this transition: these domains were supposed to have different stabilities against high pressure-induced denaturation. FTIR spectroscopy was used to follow the changes in HLADH secondary structures. This technique confirmed that the intermediate states have a low degree of unfold- ing and that no completely denatured form seemed to be reached, even up to 1000 MPa. Keywords: alcohol dehydrogenase; FTIR spectroscopy; high hydrostatic pressure; molten globule; tryptophan fluores- cence. Several papers have reported on the effects of high pressure on protein structures [1–6] or on enzyme catalytic activities [7–12]. These works have shown that high hydrostatic pressure can modify the structure or the function of enzymes by altering intra- or intermolecular interactions involved in protein stability [13,14]. However, relatively few studies have been performed to correlate conformational modifications of an enzyme to changes in its catalytic activity; although it is generally recognized that conform- ational integrity is important for preserving the activity of an enzyme [15–17]. Horse liver alcohol dehydrogenase (HLADH) is a metal protein containing two zinc ions. This enzyme is mesostable and dimeric, consisting of two identical subunits. The reasons for investigating HLADH in the present work are several. This enzyme is believed to be represen- tative of a group of proteins: dehydrogenases. Its kinetics and its three-dimensional structure are well known at ambient pressure (0.1 MPa). Its catalytic behaviour under pressure has already been studied [7,8]: Morild has shown that HLADH presented a complicated behaviour at high pressure (which was believed to be due to the pressure- induced modifications of the substrate inhibition pheno- menon occurring at ethanol concentrations > 10 m M ). Moreover, the conformational changes of this enzyme have already been monitored as a function of pressure (up to 300 MPa) by means of the intrinsic tryptophan fluores- cence, phosphorescence emission and binding of ANS fluorophore [18,19]. These studies have revealed unequivo- cal perturbations of HLADH structure in the region of the chromophores. Analysis of these structural modifications seemed to lead to the conclusion that, at about 300 MPa, the pressure induced a dimeric molten globule-like state rather than HLADH subunit dissociation. In the present work, pressure effects on HLADH structure also have been studied by intrinsic fluorescence. In fact, the intrinsic fluorescence of a protein is due mainly to the tryptophan residues [20,21] with some exceptions where the emission could be dominated by tyrosine [22]. The wavelength at maximum fluorescence, thus the centre of the spectral mass (CSM), depends on the polarity of the environment around these residues; for example, the CSM decreases as the polarity of the environment increases [4]. Based on this characteristic, the environment of the tryptophans can be monitored upon pressure-induced Correspondence to S. Dallet-Choisy, Laboratoire de Ge ´ nie Prote ´ ique et Cellulaire, Universite ´ de La Rochelle, Avenue Michel Cre ´ peau, La Rochelle, France. Fax: +33 5 46 45 82 47, Tel.: +33 5 46 45 82 77, E-mail: sdallet@univ-lr.fr Abbreviations: HLADH, horse liver alcohol dehydrogenase; CSM, center of spectral mass; FTIR, Fourier transform infrared; MG,moltenglobule;DAC,diamondanvilcell. Enzyme: horse liver alcohol dehydrogenase (EC 1.1.1.1). (Received 20 June 2002, revised 30 September 2002, accepted 18 November 2002) Eur. J. Biochem. 270, 119–128 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03370.x protein denaturation [23]. Moreover, tryptophan residues have specific positions in the HLADH molecule. The enzyme has two tryptophans per subunit: tryptophan 15 is always exposed to the solvent whereas tryptophan 314 lies right at the subunit interface [24] and upon dissociation will become exposed to the solvent. So any modifications of the enzyme’s intrinsic fluorescence would be a proof of pressure-induced unfolding, denaturation and/or dissoci- ation of the dimeric enzyme. However, because of technical limitations, the pressure in the fluorescence experiments did not exceed 600–700 MPa. In order to follow possible changes at higher pressures, Fourier Transform Infrared Spectroscopy (FTIR) was used. With this technique, in combination with the diamond anvil cell (DAC), it was possible to determine the evolution of secondary structures of a protein, up to pressures of 1000 MPa [25,26]. More- over, kinetic modifications of HLADH have been followed under pressure (up to 250 MPa), in a reactor described previously [27]. Experimental conditions were as those used for structural studies (30 °C, pH 8). This allowed us to correlate changes in HLADH catalytic activity to the pressure-induced conformational changes of this enzyme. Materials and methods Reagents HLADH (alcohol:NAD + oxidoreductase), NAD + , NADH, Tris and Mes were from Sigma. Ethanol was from Fluka. All chemicals were of analytical grade. Enzyme assays The enzyme reactions were carried out in 50 m M Tris/HCl pH 8 at 30 °C. Assays of HLADH activity for ethanol oxidation were followed by NADH absorption increase at 340 nm, directly under pressure, in a reactor described previously [27]. The initial reaction velocity was expressed as lmoles NADHÆmin )1 Æmg protein )1 using the molar absorption coefficient of NADH calculated at all pres- sures. All reactant solutions were prepared prior to use. Assays were performed for all substrates (coenzyme and ethanol) by keeping the concentration of one substrate constant (always saturating but never inhibitory for ethanol concentration) whilst varying the concentration of the other, as shown in Table 1. The enzyme concentration was 0.625 lgÆmL )1 . The apparent Michaelis constant (K m ) and the maximum velocity (V m ) were determined using STATISTICA Ò software from the least-squares fit at all given pressures. Fluorescence spectroscopy Experimental procedures. Intrinsic fluorescence measure- ments were carried out on an AB2 fluorospectrophotometer (US SLM Co.), modified in the Montpellier laboratory to measure fluorescence in the pressure range 0.1–700 MPa, through a thermostated cell [4]. The maximum pressure applied was 600 MPa and the temperature was constant at 30 °C. The enzyme concentration was 0.5 mgÆmL )1 ,1mgÆmL )1 and 5 mgÆmL )1 (in 50 m M Mes buffer pH 8), depending on the experiments. Samples were placed in a 0.5 mL quartz cuvette (5 mm path length) and were allowed to reach equilibrium for 5 min before data collection. In order to investigate selectively tryptophan fluorescence, the excita- tion wavelength was 295 nm (4 nm slit, 1 nm step size). The emission spectra were recorded between 300 nm and 370 nm (8 nm slit, 1 nm step size). Each spectrum was the result of three accumulations. Fluorescence intensities were first corrected by subtract- ing the fluorescence spectra of the buffer at each pressure and then they were corrected for volume contraction under high pressure. The CSM, defined below (Eqn 1), was used to quantify the spectra [21,28,29]. CSM ¼ X vi  Fi= X Fi ð1Þ where mi is the wavenumber and Fi the fluorescence intensity at mi. Determination of thermodynamic parameters. Thermody- namic parameters DG 0 and DV of the pressure-induced spectral transitions were determined by analysing the CSM. A two state-transition (n for native and d for denatured) could be assumed. Generally, if CSM was assumed to take different values in the two states (CSMn and CSMd) and an increase of pressure to cause a transition from one state to the other, then an equilibrium constant K(p) can be defined at a given pressure (Eqn 2): KðpÞ¼ CSMd À CMSp½ CSMp À CSMn½ ð2Þ where CSMp is the intermediate value of the CSM at a given pressure (P). Table 1. Substrate concentrations used for activity measurements under pressure. Pressure (MPa) < 100 > 100 > 200 Varying concentration of coenzyme Ethanol concentration (m M ) 20 200 NAD + concentration (m M ) Varying from 0 to 3 Varying concentration of alcohol Ethanol concentration (m M ) Varying from 0 to 700 NAD + concentration (m M ) 0.62 12 120 M. Trovaslet et al. (Eur. J. Biochem. 270) Ó FEBS 2003 The pressure dependence of K(p)canberelatedtoDG 0 (at atmospheric pressure) and DV (3) [30]: ÀRT ln½KðpÞ ¼ DG 0 þ PDV ð3Þ Combining Eqns (2) and (3): CSM ¼ CSMd þ CSMn À CSMd 1 þ e ÀDG 0 þPDV RT ÀÁ ð4Þ Eqn (4) was used to determine HLADH thermodynamic parameters upon pressure-induced structural modifications. The pressure of half denaturation (P 1/2 )isgivenbyEqn(5): P 1=2 ¼ÀDG 0 =DV ð5Þ FTIR spectroscopy The pressure-induced structural modifications of HLADH was also investigated by FTIR, in the Leuven laboratory, by using the DAC (Diacell Products, Leicester, UK). The gasket between the diamonds had an original thickness of 50 lm. The DAC was connected to a thermostat and the temperature inside the cell was measured by means of a thermocouple localized near the diamonds [5,25]. This allowed measurements at stable temperature of 30 °C. Barium sulfate was used as an internal pressure standard [31]. The infrared spectra were obtained with a Bruker IFS66 FTIR spectrometer equipped with a liquid nitrogen cooled broad-band mercury–cadmiun–telluride detector. The sample compartment was continuously purged with dry air. A total of 250 interferograms were coadded after registration at a resolution of 2 cm –1 . Spectral noises originating from water vapour were removed through the subtraction of the water vapour spectrum using the BRUKER software. Spectra shown were smoothed 9 points after vapour correction. The second-derivative spec- tra containing the amide I¢ band of HLADH were obtained through application of the 9-data-point Savitzky–Golay function available from Grams Research software. HLADH concentration was 50 mgÆmL )1 . The lyophi- lized protein was dissolved in 10 m M deuterated Tris/DCl pD 8. Due to its high concentration, the protein itself could buffer the solution [5]. In this case, the pH value of the buffer was the same as the pH obtained by measuring that of the protein solution. After solubilization, the sample was stored overnight at 25 °C to ensure that all the solvent accessible protons are exchanged for deuterons. Results and discussion Pressure-induced kinetic modifications In our preliminary report on HLADH reaction under pressure [32], kinetic parameters and thermodynamic acti- vation volumes of HLADH oxidation of ethanol with the coenzyme NAD + as oxidizing agent were determined. Based on this previous study, kinetic parameters of this reaction (V m and K m ) were now determined always using noninhibitory ethanol concentrations. Fig. 1 shows the pressure dependence of these kinetic parameters and of HLADH catalytic efficiency (inset), respectively, at saturating NAD + concentration and varying ethanol concentration (Fig. 1A), and at saturating (but not inhibi- tory) ethanol concentration and varying NAD + concentra- tion (Fig. 1B). Whatever the varying substrate concentration, up to 200 MPa, the V m values increased with an increase in pressure: at 200 MPa, V m was about 10 times higher than at 0.1 MPa. In the same range of pressure, HLADH affinities for both substrates (NAD + and ethanol) seemed to be strongly decreased. The K m values for NAD + and ethanol, at atmospheric pressure and at 225 MPa, were 18.92 ± 1.97 l M and 561.82 ± 68.93 l M ,0.56±0.03m M and 24.51 ± 3.17 m M , respectively (Fig. 1A and B). Conse- quently, these variations could influence HLADH catalytic efficiency (k cat /K m ). Up to 75–100 MPa (depending on the varying substrate concentration), k cat /K m values were almostthesameasat0.1MPa:thepressuredidnotseem to influence HLADH catalytic efficiency. Above 100 MPa, further elevation of pressure gave a lower catalytic effi- ciency. At 225 MPa, the k cat /K m value was reduced by % 70% compared with the value obtained at 0.1 MPa. As shown in Fig. 1A and B (inset), the plot of k cat /K m vs. pressure seemed to be ÔbiphasicÕ. This behaviour could be related to a multistep process of pressure-induced unfolding, Fig. 1. Pressure-induced modifications in HLADH kinetics parameters (V m and K m ), during the oxidation of ethanol with the coenzyme NAD + as oxidizing agent (30 °C, Tris/HCl pH 8). (A) At saturating NAD + concentration and varying ethanol concentration. Inset: k cat /K m vs. pressure. (B) At saturating (but not inhibitory) ethanol concentration and varying NAD + concentration. Inset: k cat /K m vs. pressure. Ó FEBS 2003 HLADH under pressure (Eur. J. Biochem. 270) 121 denaturation and/or dissociation of HLADH molecule. So, a better knowledge of the pressure-induced structural modifications of the dimeric enzyme seemed to be essential to understand this kinetic behaviour. Fluorescence spectroscopy To monitor the effects of high pressure on HLADH tertiary and quaternary structures (pressure-induced denaturation or dissociation of the dimeric enzyme), fluorescence spec- troscopy under pressure was used. The intrinsic fluorescence spectra of HLADH excited at 295 nm, under different pressures from 0.1 to 600 MPa, are shown in Fig. 2. Fluorescence intensity of aromatic residues seems to vary in somewhat unpredictable manner, but the wavelength of the emitted light seems to be a better indication and can be used to follow the environment of the fluorophores [20]. At atmospheric pressure, the enzyme displayed a typical fluorescence emission spectrum with a maximum at 332 nm. This fluorescence emission maximum is characteristic of tryptophans placed in a relatively hydrophobic environment, buried in protein [20]. The three-dimensional structure of HLADH is well character- ized [24]. The dimeric enzyme has two tryptophans per subunit. Tryptophan 15 is always exposed to the solvent whereas tryptophan 314 is buried at the subunit interface within a large b-sheet that extends from one subunit to the other (Fig. 3). Tryptophan 314 is thus buried in an extremely hydrophobic environment in contrast with tryp- tophan 15 [24]. So, any change in the fluorescence charac- teristics of HLADH with pressure could be attributed exclusively to alterations in the environment of tryptophan 314. When pressure was raised to 600 MPa, the fluorescence intensity decreased by 60% and the maximum emission wavelength had a red shift of 8 nm (Fig. 2), indicating the pressure induced unfolding of HLADH molecule. In fact, in high pressure studies of proteins, red shifts in tryptophan fluorescence have invariably been attributed to the hydra- tion of fluorophores as a result of either subunit dissociation and/or penetration of water molecules to interior sites of the protein globule [21]. To monitor tryptophan hydration during the high pressure treatment, the CSM, which reflects global changes of the population of fluorophores, was calculated for each spectrum and then was plotted against pressure in Fig. 4. Between 0.1 and 600 MPa, a gradual decrease in the CSM occurred with increasing pressure. In this range of pressure, the transition curve did not follow a simple two-state transition. The pressure-induced modifications of HLADH seemed to occur through a multi-step process, of at least three transitions and four different states, although the final state did not seem to be completely reached. The first Fig. 2. Fluorescence spectra of tryptophan in HLADH at 30 °Cand pH 8, under different pressures. The spectra from top to bottom cor- respond to the pressures: 0.1, 100, 400 and 600 MPa. The concentra- tion of HLADH was 1 mgÆmL )1 in 50 m M Mes buffer, excitation wavelength: 295 nm. Fig. 3. Strands diagram of HLADH, based on X-ray crystallographic data of Eklund et al. [33] (6ADH) obtained from the Swiss Prot data- base. The two different domains of one subunit are coloured orange (coenzyme-binding domain) and yellow (catalytic domain); trypto- phans 15 and 314 are shown in red and NAD + isshowninwhite. Fig. 4. Pressure dependence of the CSM of HLADH intrinsic fluores- cence. d, Compression; s, decompression. Conditions: 30 °C, pH 8, concentration of HLADH, 1 mgÆmL )1 in 50 m M Mes buffer, excita- tion wavelength: 295 nm. Dotted lines show the half of the second transition. 122 M. Trovaslet et al. (Eur. J. Biochem. 270) Ó FEBS 2003 transition was over at about 100 MPa, the second one occurred between 100 and 400 MPa and the last one began at higher pressures (% 500 MPa). Relatively low pressure (< 100 MPa) seemed to induce small conformational changes of HLADH, because the amplitude of the change in the CSM was moderate. Since pressure promotes structural rearrangement of the protein/ solvent interactions, the application of low pressure may provide pathways for water to penetrate into the protein and probably between subunits favouring the protein hydration. It could explain the decrease of CSM for this range of pressure and hence the increase of polarity of the environment of Trp314. Moreover, pressure could also induce a reduction of size of internal cavities, voids that result from imperfect packing of amino acids and a change in the length of chemical bonds [34]. The protein could reach a new conformation state that could be assumed to a native- like state (N¢) and could correspond to the first intermediate state of pressure-induced structural modification. These conformational changes do not alter the molecular folding which is confirmed by the weak effect of pressure on the catalytic efficiency of HLADH, as shown in the previous section (Fig. 1A and B inset): more precisely, pressure favours the catalytic step whereas the substrate binding steps are influenced by pressure only slightly. At atmo- spheric pressure and noninhibitory primary alcohol con- centration, the dissociation rate of the enzyme–NADH complex is the rate-limiting step [35,36]. Hence, the positive effect of pressure on HLADH catalysis could be due to a new conformation of the enzyme which enhances the dissociation rate of NADH from the enzyme [32]. Under pressure, hydration and decrease of volume have antagon- istic effects on the flexibility of proteins. In the range of 75–100 MPa, these antagonisms have probably induced an optimal conformation of HLADH as far as the kinetics are concerned. As the environment of Trp314 is only slightly more polar in this pressure range, we deduce that the hydrophobic core of HLADH is not much affected. As a consequence, the dissociation of the dimeric form of HLADH cannot take place. Higher pressures, between 100 and 400 MPa, induced greater changes in the intrinsic fluorescence of HLADH. In this pressure range, the enzyme also progressively lost its catalytic efficiency. Then, the increase of catalytic activity (V m ) was strongly counteracted by the decrease of HLADH affinities for its substrates (NAD + and alcohol) (Fig. 1A and B). Therefore, conformational changes observed prob- ably corresponded to alterations of the dimeric enzyme active sites, where both NAD + and ethanol were bound. These suggestions could be in good agreement with Tsou’s conclusions [37,38]. Based on studies of inactivation/dena- turation of several enzymes (including alcohol dehydroge- nase from baker’s yeast), Tsou has proved that enzyme active sites are formed by relatively weak molecular interactions and may be conformationally more flexible than the whole molecule. In the HLADH molecule, the position of the active site is well known: the enzyme subunits are divided into two different domains (the coenzyme binding domain and the catalytic domain). These domains are separated by a crevice that contains a wide and deep pocket which is the binding site for the substrate and the nicotinamide moiety of the coenzyme [24]. This specific position of the active site could explain why it was more sensitive to pressure than the molecule as a whole. Because of the sigmoidal shape of the transition observed between 100 and 400 MPa, a two-state transition was assumed and pseudo-thermodynamic parameters were calculated. The apparent molecular standard volume change (DV app ) and the apparent free energy ðDG app 0 Þ of HLADH upon pressure-induced structural modifications (between 100 and 400 MPa), were )41.5 ± 4.6 mLÆmol )1 and +10.7 ± 1.2 kJÆmol )1 , respectively. The pressure of half-denaturation was 260 MPa. The value of the apparent molecular standard volume change (DV app ¼ )41.5 ± 4.6 mLÆmol )1 )wasnotvery important and did not suggest HLADH dissociation. In fact, molecular standard volume changes upon oligomeric protein dissociation usually vary between )100 and )500 mLÆmol )1 , depending on proteins and experimental conditions [39,40]. So, our result seemed to confirm the conclusions of Cioni and Strambini [18]. Based on a HLADH phosphorescence study, they have shown that the dimeric molecule is not dissociated up to 300 MPa, even if the tryptophan 314 becomes more exposed to the solvent. It is generally admitted that oligomeric proteins dissociate under moderate pressures [41] with numerous exceptions [42]. For example, butyrylcholinesterase studies have shown that pressure-induced modifications of this tetrameric enzyme is not a dissociation (up to 350 MPa), but a multi-step process of denaturation and that the observed transient pressure-denatured state has characteristics of the molten globule state [43]. Furthermore, work on RNase A [6] or on the 7-kDa protein P2 from Sulfolobus solfataricus [44] has shown a major role of hydrophobic residues and interactions among aromatics in barostability. So, the large hydrophobic interaction area present between the HLADH subunits could explain why it was difficult to separate the enzyme subunits without denaturation [24]. The modifications of the enzyme’s tertiary structure, the loss of catalytic activity and the capacity to bind ANS [18] are characteristic properties of a molten globule-like state [43]. Hence the existence of a pressure-induced dimeric molten globule like state (MG1) could be assumed at 400 MPa. The higher pressures applied (> 400 MPa) showed an additional transition in the CSM variation vs. pressure, although the final state did not seem to be completely reached. Two different questions were then addressed: did this new transition correspond to a conversion from a dimeric state to a monomeric state? Did it correspond to conversion from a molten globule to a second molten globule like state or to a completely denatured state? While there was no direct evidence that the final state corresponds to a second molten globule-like state, the incomplete fluorescence red shift strongly supported this conclusion. In fact, as water-exposed indole side chains fluoresce with a kmax between 350 and 355 nm [20] and for HLADH kmax reached only at 340 nm at 600 MPa, the HLADH fluorophores seemed to be partly shielded from the solvent. This incomplete fluorescence red shift could be characteristic of a dimeric intermediate state at 600 MPa. However, it did not exclude monomer formation because structural rearrangements in the monomer may partly shield tryptophan 314 from the solvent. Ó FEBS 2003 HLADH under pressure (Eur. J. Biochem. 270) 123 In order to characterize better the new state obtained at 600 MPa, reversibility of pressure-induced modifications of HLADH structure was followed. The fluorescence spectra and the CSM values were followed upon enzyme decom- pression. When the pressure was released from 600 MPa to atmospheric pressure, the fluorescence spectrum of HLADH did not return to the original (data not shown) and the CSM did not exactly come back to the original value although the sample was stored more than 30 min at 0.1 MPa after decompression (the recovery of the signal was % 80% of the value at 0.1 MPa) (Fig. 4). So, under these conditions, the pressure-induced modifications of HLADH structure seemed to be partly reversible, supporting that at 600 MPa, no completely denatured state was reached: a new intermediate state seemed to be obtained. Interestingly, the recovery of the CSM showed a strong hysteresis when the pressure was gradually released: the CSM at a pressure in the decompression direction was much lower than at the same pressure in the compression direction. This Ôconform- ational driftÕ could be explained by assuming that the last transition represented the dissociation of HLADH subunits, the modifications of these subunits’ structure and that the subunit reassociation started before complete refolding of individual subunits was completed [45]. These suggestions could explain why the pressure-induced modifications of HLADH were only partly reversible: a possible aggregation of the (partly) unfolded monomer could occur. It is known that the renaturation of proteins in native and aggregated form is different. So, all our results suggested that a second intermediate state was observed at 600 MPa: it could be a monomeric molten globule-like state. Thus, the native like state (N¢)seemedtobefurther unfolded into a first molten globule (MG1) and a second molten globule (MG2) as pressure increased. Different models could be proposed to explain this process: (a) each of the transitions observed in the CSM changes with pressure concerned the whole molecule. Then, the whole HLADH molecule was melted into a first molten globule, which was melted into a second molten globule state. In such cases, both intermediate states still have partially secondary structure but their compactness is different [46]; (b) the protein molecule was assumed to be composed of at least two different domains each of which unfolded independ- ently. Then, MG1 corresponded to the melting of one domain into a molten globule-like state, whereas the second domain preserved its native conformation, and MG2 corresponded to the melting of the second domain into a molten globule-like state. As shown in Fig. 3, each subunit of HLADH is divided into two separate domains: the coenzyme binding domain and the catalytic domain. These two domains are unequal in size and in amount of secondary structure [24]. It is possible that they underwent conformational changes nonsimulta- neously, showing that they are independent and different in stability against high pressure-induced structural modifica- tions. The coenzyme-binding domain could be less pressure sensitive than the catalytic domain because of its amount of secondary structure. In the catalytic domain, a large number of residues – 32% – have no regular secondary structure, whereas in the coenzyme-binding domain only % 10% of the residues have no regular secondary structure, and no continuous stretch of irregular structure is longer than four residues [24]. According to the second model proposed, at moderate pressures alterations of the catalytic domain could explain the decrease of HLADH affinity for its substrates and thus the modifications of the enzyme catalytic efficiency without dissociation of the dimeric enzyme; the dissociation happened at higher pressures, when modifications of the coenzyme-binding domain start. A similar behaviour has already been observed with creatine kinase. Zhou et al.[16] have shown that creatine kinase inactivation, at low pressure, may precede the enzyme dissociation and the unfolding of the hydrophobic core, occurring at higher pressure. They have postulated that the multi-state transi- tions induced both by pressure and guanidine denaturation were in direct relationship with the existence of hydrogen bonds which maintain the dimeric structure of the enzyme. At last, in order to determine whether the observed effects of pressure were due to HLADH dissociation or not, the same experiment was performed at different enzyme concentrations. Usually dissociation of oligomeric enzymes exhibits a strong concentration dependence as expected from the law of mass action [28,29,47]. Fig. 5 shows that the half-transition pressure values of the two intermediate states (N¢fiMG1 and MG1 fi MG2)remainthesamewhat- ever the enzyme concentration. The concentration inde- pendence of pressure effects suggests that unfolding (and inactivation) of the enzyme is not a result of the dissociation of the dimer. It seems that our results are in agreement with Ruan and Weber [29] who have shown that the degree of dissociation of an oligomeric protein could be insensitive to protein concentration. However our results do not com- pletely exclude the formation of a monomer. Recent investigations of the reversible subunit dissociation of several oligomers (ranging from dimer to viral particles) by hydrostatic pressure has revealed significant deviations from the law of mass action and hence an anomalous or complete lack of protein concentration dependence [47]. HLADH could fit to the previous case as the dissociation of the dimers seems to follow different rates and therefore the dimer population is heterogeneous in terms of thermody- namic stability. In the case of the dimeric triosephosphate Fig. 5. Pressure variation of the CSM of HLADH intrinsic fluores- cence, at 30 °C and pH 8. The concentration of HLADH was 0.5 mgÆmL )1 (d)and5mgmL )1 (n)in50 m M Mes buffer. Excitation wavelength: 295 nm. 124 M. Trovaslet et al. (Eur. J. Biochem. 270) Ó FEBS 2003 isomerase from rabbit muscle it was shown that the activation free-energy barriers for dimer dissociation are quite high corresponding to a characteristic time of dissociation of 15 h or longer [47]. Perhaps, it should be considered that the HLADH dimer dissociation is a slow process and that the reaction could be controlled by kinetic rather than thermodynamic factors [47,48]. FTIR studies of pressure-induced structural modifications of HLADH Mozhaev et al. [49] have suggested that the data obtained by fluorescence spectroscopy could be complementary to those obtained by other techniques such as FTIR. In fact, in the few instances when these techniques have been applied to the same protein (trypsin, gliadin, staphylococcal nucle- ase, Trp apo-repressor [2,50–52] for example), almost identical denaturation pressures have often been found, except the case of Trp apo-repressor of Escherichia coli for which the different behaviour could be explained by the use of different solvent conditions. But, in contrast with the tryptophan fluorescence experiments, where local pressure effects were observed, FTIR spectroscopy was used to look at the secondary structures of the whole molecule. In fact, the secondary structures of the protein may be determined from the analysis of the amide I¢ bandshape of the infrared spectrum. The amide I¢ band of proteins occurs due to the in-plane C ¼ O stretching vibrations which are weakly coupled with the C–N stretching and in-plane N–H bending vibrations. It is located in the frequency band of 1600– 1700Æcm )1 and usually consists of many overlapping com- ponent bands that represent different structural elements such as a-helices, b-sheets, turns, nonordered or irregular structures [53,54]. Several different approaches can be used to determine these structures: the spectrum can be compared with a database of the amide I¢ bands of several proteins with known secondary structures, curve fitting after Fourier self-deconvolution can be observed or second derivative of the spectrum can be calculated. For studies of pressure- induced structural changes of proteins, the first approach cannot be easily used because of the need of a data set of reference proteins as a function of pressure. In pressure- induced structural modifications of HLADH, the second derivative of the absorption spectrum was used. Then, a qualitative analysis can be carried out in order to identify the secondary structures present in the protein and to detect theirs pressure-induced modifications. Furthermore, the use of the DAC made it possible to extend the pressure range up to 1000 MPa and thus allowed us to follow possible changes in the protein conformation beyond the reach of fluores- cence conditions. Fig. 6 shows the absorbance spectra of HLADH in D 2 O in the frequency region 1350–1750Æcm )1 for pressures of 0.1–1000 MPa. The amide I¢ bands is centred at 1640Æcm )1 and the amide II¢ band at % 1455Æcm )1 . The band at % 1557Æcm )1 is attributed to the amide II mode which is a mixed vibration involving N–H in plane bending and the CN stretching [55]. This band shifts to % 1455Æcm )1 as a result of deuteration of labile protons on the amide groups. On this graph we observe a simultaneous decrease of the band at 1557Æcm )1 and an increase of the band at 1450Æcm )1 with pressure increase up to 800 MPa. This phenomenon could be attributed to H/D exchange during the increase of pressure for hydrogens buried in the core of the protein. Nevertheless this H/D exchange is not total which is in good concordance with the partial unfolding of the protein. Fig. 7 shows the original spectra of amide I¢ band (1600– 1700Æcm )1 ) and Fig. 8 shows the second derivative of nondeconvoluted spectra for the same range. At atmo- spheric pressure, the amide I¢ band consists of several frequency regions, summarized in Table 2, showing that different secondary structures could be observed in the HLADH molecule. Up to relative low pressure (90– 100 MPa), when the native-like state (N¢) was assumed, no change in the infrared spectrum of HLADH seemed to take place. At higher pressures (between 100 and 600 MPa), when molten globule like states MG1 and MG2 were assumed, only a few changes in the secondary structures of HLADH were observed. These changes were rather small, as can be seen in Fig. 8. The band at 1610Æcm )1 ,which corresponds to the amino acid side chains, exhibits the main change suggesting that pressure induced modifications of HLADH tertiary structure. The bands at 1658Æcm )1 , 1666Æcm )1 and 1672Æcm )1 have been assigned to a-struc- tures, b-turns and b-sheets [53,54]. Pressurization of HLADH seemed to cause a small modification of the amount of these structures. These trends continued as the Fig. 6. Original spectra of the amide I¢ and II¢ areas of HLADH (50 mgÆmL -1 ) in Tris/DCl 10 m M at varying pressures 0.1–1000 MPa (pH 8, 30 °C). Fig. 7. Original spectra of the amide I¢ areas of HLADH (50 mgÆmL -1 ) in 10 m M Tris/DCl at varying pressures 0.1–1000 MPa (pH 8, 30 °C). Ó FEBS 2003 HLADH under pressure (Eur. J. Biochem. 270) 125 pressure was increased up to 1000 MPa; but no further modifications were observed up to the maximum pressures. These results were consistent with the molten globule like states observed at 400 MPa (MG1) and 600 MPa (MG2) in fluorescence experiments. In fact, one of the characteristic properties of the molten globules is the partially secondary structures present in these intermediate states [30,43]. Fig. 9 gives a more quantitative representation of the pressure-induced spectral changes of HLADH, by plotting the bandwidth at half height of the amide I¢ band of the enzyme vs. pressure (up to 1000 MPa). It is of interest to note that no cooperative change in the spectrum was observed in that pressure range. Moreover, the shape of the amide I¢ band of the protein (Fig. 7) contained distinct features up to 1000 MPa, which suggested that HLADH was not completely unfolded at this pressure. In fact, whatever the applied pressure – up to 1000 MPa – no plateau was obtained (Fig. 9), showing that HLADH was not transformed into a stable pressure denatured state. All of these FTIR results were in agreement with the fluorescence data. They allowed us to confirm the existence of at least two molten globule like states upon pressure- induced unfolding of HLADH, while no completely dena- turedstatewasobtainedupto1000MPa. Fig. 9 also led to the conclusion that the pressure-induced modifications of HLADH were irreversible, which was in contrast with the fluorescence results. However, several authors have shown that the high concentrations of proteins used in FTIR studies often lead to irreversible inactivation or denaturation [5,49]. In fact, in FTIR studies, the intermolecular interactions (protein–protein interactions) became more pronounced and this could explain the findings on the irreversibility [56]. Conclusion It is well known that the activity of an enzyme is strongly dependent on its conformational integrity [57,58]. So, the aim of this study was to characterize the pressure-induced conformational changes of the dimeric HLADH and to correlate them with the modifications of the enzyme’s catalytic activity. To this end, the pressure-induced modi- fications of the enzyme’s activity were studied (up to 250 MPa) as well as the pressure effects on protein fluorescence (up to 600 MPa) and the changes in HLADH secondary structures (studied by FTIR technique up to 1000 MPa). All of these studies indicated that the pressure- induced modifications of HLADH represents a multi-step process leading to several different intermediate states of denaturation. A native like state (N¢) and two different molten globules (at 400 and 600 MPa) were assumed, while no completely denatured state seemed to be reached up to 1000 MPa (Figs 4 and 9). Although it is widely accepted that oligomeric enzymes are dissociated under moderate pressures [28,29,59,60], our fluorescence and FTIR studies have confirmed that pressure (up to 400 MPa) did not induce HLADH dissociation: our structural studies sup- ported the existence of a dimeric molten globule at 400 MPa. Moreover, at 600 MPa, the intermediate seemed to be monomeric because of the hysteresis observed when the pressure was released from 600 to 0.1 MPa. Further experiments (like electrophoresis under high hydrostatic pressure) could be necessary to confirm that the last Table 2. Amide I¢ frequencies (cm -1 ) characteristic of the amide bond in various conformations [51,52]. Conformation Band position (cm )1 ) Amino acid side chains % 1610 Unordered structures % 1643 % 1656–1660 a-Helices % 1648–1655 b-Turns % 1662–1667 b-sheets % 1628–1638 % 1672–1678 % 1690–1693 Fig. 9. Pressure effects on the bandwidth at half height of the amide I¢ band of HLADH (in Tris/DCl 10 m M pH 8), at 30 °C, upon compres- sion (d)anddecompression(s). Fig. 8. Second derivative IR spectra of the amide I¢ area of HLADH (50 mgÆmL -1 )in10m M Tris/DCl at various pressures (pH 8, 30 °C). 126 M. Trovaslet et al. (Eur. J. Biochem. 270) Ó FEBS 2003 transition observed in the fluorescence experiments (Fig. 4) really corresponded to conversion from a dimeric state to a monomeric state. At last, in order to test the contribution of the pressure insensitive interactions involved in HLADH subunit contacts, to the pressure insensitivity of the molecule, specific hydrophobic residues (present at the subunit interface) could be replaced with nonhydrophobic amino acids. Then, high pressure studies of HLADH mutants could be considered to observe their pressure- induced unfolding, denaturation and/or dissociation. This could allow us to understand the molecular basis of the pressure stability of this enzyme and perhaps of the other proteins of the alcohol dehydrogenase family. Acknowledgements We thank N. Bec and R. Lange for stimulating discussions and fruitful advice. References 1. Silva, J.L., Foguel, D. & Royer, C.A. (2001) Pressure provides new insights into protein folding, dynamics and structure. Trends Biochem. Sci. 26, 612–618. 2. Balny, C., Masson, P. & Heremans, K., eds. (2002) Frontiers in High Pressure Biochemistry and Biophysics. Elsevier, Amsterdam. 3. Ikeuchi, Y., Suzuki, A., Oota, T., Hagiwara, K., Tatsumi, R., Ito, T. & Balny, C. (2002) Fluorescence study of the high pressure- induced denaturation of skeletal muscle actin. Eur. J. Biochem. 269, 364–371. 4. Ruan, K., Lange, R., Bec, N. & Balny, C. (1997) A stable partly denatured state of trypsin induced by high hydrostatic pressure. Biochem. Biophys. Res. Comm. 239, 150–154. 5. Ruan, K., Lange, R., Meersman, F., Heremans, K. & Balny, C. (1999) Fluorescence and FTIR study of the pressure-induced denaturation of bovine pancreas trypsin. Eur. J. Biochem. 265, 79–85. 6. Torrent, J., Rubens, P., Ribo, M., Heremans, K. & Vilanova, M. (2001) Pressure versus temperature unfolding of ribonuclease A: An FTIR spectroscopic characterization of 10 variants at the carboxy-terminal site. Protein Sci. 10, 725–734. 7. Morild, E. (1977) Pressure neutralization of substrate inhibition in the alcohol dehydrogenase reaction. J. Phys. Chem. 81, 1162– 1166. 8. Morild, E. (1977) Pressure variation of enzymatic reaction rates: yeast and liver alcohol dehydrogenase. Biophys. Chem. 6, 351–362. 9. Hamon, V., Dallet, S. & Legoy, M.D. (1996) The pressure- dependence of two b-glucosidases with respect to their thermo- stability. Biochim. Biophys. Acta 1294, 195–203. 10. Ciardiello, M.A., Schmitt, B., di Prisco, G. & Herve ´ , G. (1999) Influence of hydrostatic pressure on 1-glutamate dehydrogenase from the Antarctic fish Chaenocephalus aceratus. Marine Biol. 134, 631–636. 11. Northrop, D.B. & Cho, Y K. (2000) Effects of high pressure on solvent isotope effects of yeast alcohol dehydrogenase. Biophys. J. 79, 1621–1628. 12. Northrop, D.B. (2002) Effects of high pressure on enzymatic activity. Biochim. Biophys. Acta 1595, 71–79. 13. Deville-Bonne, D. & Else, A.J. (1991) Reversible high hydrostatic pressure inactivation of phosphofructokinase from Escherichia coli. Eur. J. Biochem. 200, 747–750. 14. Masson, P., Froment, M T., Fort, S., Ribes, F., Bec, N., Balny, C. & Schopfer, L.M. (2002) Butyrylcholinesterase-catalyzed hydro- lysis of N-methylindoxyl acetate: analysis of volume changes upon reaction and hysteresis behavior. Biochim. Biophys. Acta 1597, 229–243. 15. Kornblatt, M.J., Lange, R. & Balny, C. (1998) Can monomers of yeast enolase have enzymatic activity? Eur. J. Biochem. 251, 775–780. 16. Zhou, J M., Zhu, L. & Balny, C. (2000) Inactivation of creatine kinase by high pressure may precede dimer dissociation. Eur. J. Biochem. 267, 1247–1253. 17. Roitel, O., Bec, N., Lange, R., Balny, C. & Branlant, G. (2001) Pressure denaturation of phosphorylating glyceraldehyde- 3-phosphate dehydrogenase from Bacillus stearothermophilus. Biochem. Biophys. Res. Comm. 283, 347–350. 18. Cioni, P. & Strambini, G.B. (1996) Pressure effects on the structure of oligomeric proteins prior to subunit dissociation. J. Mol. Biol. 263, 789–799. 19. Cioni, P. & Strambini, G.B. (2002) Tryptophan phosphorescence and pressure effects on protein structure. Biochim. Biophys. Acta 1595, 116–130. 20. Schmid, F.X. (1989) Spectral methods of characterizing protein conformation and conformational changes. In Protein Structure: a Practical Approach (Creighton, T. E., ed.), pp. 251–285. IRL Press, Oxford. 21. Ruan, K. & Balny, C. (2002) High pressure static fluorescence to study macromolecular structure-function. Biochim. Biophys. Acta 1595, 94–102. 22. Ruan, K., Li, J., Liang, R. & Xu, C., YuY., Lange, R. & Balny, C. (2002) A rare protein fluorescence behavior where the emission is dominated by tyrosine: case of the 33-kDa protein from spinach photosystem II. Biochem. Biophys. Res. Comm. 293, 593–597. 23. Ruan, K., Tian, S., Lange, R. & Balny, C. (2000) Pressure effects on tryptophan and its derivatives. Biochem. Biophys. Res. Comm. 269, 681–686. 24. Eklund, H., Nordstro ¨ m, B., Zeppezauer, E., So ¨ derlund, G., Ohlsson, I., Boiwe, T., So ¨ derberg, B.O., Tapia, O. & Bra ¨ nde ´ n, C.I. (1976) Three-dimensional structure of horse liver alcohol dehy- drogenase at 2.4 A ˚ resolution. J. Mol. Biol. 102, 27–59. 25. Smeller, L. & Heremans, K. (1997) In situ observation of the pressure-induced changes in proteins with infrared spectroscopy in the diamond anvil cell. In High Pressure Research in the Bios- ciences and Biotechnology (Heremans, K., ed.), Leuven University Press, Leuven. 131–134. 26. Dzwolak, W., Kato, M. & Taniguchi, Y. (2002) Fourier transform infrared spectroscopy in high-pressure studies on proteins. Biochim. Biophys. Acta 1595, 131–144. 27. Dallet, S. & Legoy, M.D. (1996) Hydrostatic pressure induces conformational and catalytic changes on two alcohol dehy- drogenases but not oligomeric dissociation. Biochim. Biophys. Acta 1294, 15–24. 28. Silva, J.L., Miles, E.W. & Weber, G. (1986) Pressure dissociation and conformational drift of the dimer of tryptophan synthase. Biochemistry 25, 5780–5786. 29. Ruan, K. & Weber, G. (1989) Hysteresis and conformational drift of pressure dissociated glyceraldehyde phosphate dehydrogenase. Biochemistry 28, 2144–2153. 30. Dumoulin, M., Ueno, H., Hayashi, R. & Balny, C. (1999) Con- tribution of the carbohydrate moiety to conformational stability of the carboxypeptidase Y. High pressure study. Eur. J. Biochem. 262, 475–483. 31. Wong, P.T.T. & Moffat, D.J. (1989) A new internal pressure calibrant for high-pressure infrared spectroscopy of aqueous systems. Appl. Spectrosc. 43, 1279–1281. 32. Trovaslet, M., Dallet, S. & Legoy, M.D. (2000) Effects of hydrostatic pressure and temperature on catalytic activity of horse liver alcohol dehydrogenase (HLADH). High Pressure Res. 19, 303–309. Ó FEBS 2003 HLADH under pressure (Eur. J. Biochem. 270) 127 33. Eklund, H., Samma, J.P., Wallen, L., Branden, C.I., Akeson, A. & Jones, T.A. (1981) Structure of a triclinic ternary complex of horse liver alcohol dehydrogenase at 2.9 A ˚ resolution. J. Mol. Biol. 146, 561–587. 34. Heremans, K. & Smeller, L. (1988) Protein structure and dynamics at high pressure. Bioch. Biophys. Acta 1386, 353–370. 35. Dalziel, K. & Dickinson, F.M. (1966) The kinetics and mechanism of liver alcohol dehydrogenase with primary and secondary alco- hols as substrates. Biochem. J. 100, 34–46. 36. Bra ¨ nde ´ n, C.I., Jo ¨ rnvall, H., Eklund, H. & Furugen, B. (1975) Alcohol dehydrogenases. In The Enzymes,Vol.XI(BoyerP.D, eds), 3rd edn, pp. 103–190. Academic Press, New York. 37. Tsou, C.L. (1986) Location of the active sites of some enzymes in limited and flexible molecular regions. Trends Biochem. Sci. 11, 427–429. 38. Tsou, C.L. (1993) Conformational flexibility of enzyme active sites. Science 262, 380–381. 39. Erijman, L., Lorimer, G.H. & Weber, G. (1993) Reversible dissociation and conformational stability of dimeric ribulose bisphosphate carboxylase. Biochemistry 32, 5187–5195. 40. Schade, B.C., Rudolph, R., Ludemann, H.D. & Jaenicke, R. (1980) Reversible high-pressure dissociation of lactic dehy- drogenase from pig muscle. Biochemistry 19, 1121–1126. 41. Mozhaev, V.V., Heremans, K., Frank, J., Masson, P. & Balny, C. (1994) Exploiting the effects of high hydrostatic pressure in bio- technological applications. Trends Biotechnol. 12, 493–501. 42. Balny, C. (2002) High pressure and protein oligomeric dissoci- ation. High Press Res. 22, 737–741. 43. Cle ´ ry, C., Renault, F. & Masson, P. (1995) Pressure- induced molten globule state of cholinesterase. FEBS Lett. 370, 212–214. 44. Fusi, P., Goossens, K., Consonni, R., Grisa, M., Puricelli, P., Vecchio,G.,Vanoni,M.,Zetta,L.,Heremans,K.&Tortora,P. (1997) Extreme heat- and pressure-resistant 7-kDa protein P2 from the archaeon Sulfolobus solfataricus is dramatically destabi- lized by a single-point amino acid substitution. Proteins 29, 381– 390. 45. Weber, G. (1986) Phenomenological description of the associ- ation of protein subunits subjected to conformational drift. Effects of dilution and of hydrostatic pressure. Biochemistry 25, 3626– 3631. 46. Levitt,M.,Gerstein,M.,Huang,E.,Subbiah,S.&Tsai,J.(1997) Protein folding: the endgame. Annu. Rev. Biochem. 66, 549–579. 47. Ferreira, S.T. & De Felice, F.G. (2001) Protein dynamics, folding and misfolding: from basic physical chemistry to human con- formational diseases. FEBS Lett. 498, 129–134. 48. Cioni, P. & Strambini, G.B. (1997) Pressure-induced dissoci- ation of yeast glyceraldehyde-3-phosphate dehydrogenase: heterogeneous kinetics and perturbations of subunit structure. Biochemistry 36, 8586–8593. 49. Mozhaev, V.V., Heremans, K., Frank, J., Masson, P. & Balny, C. (1996) High pressure effects on protein structure and function. Proteins: Struct., Funct. Genet. 24, 81–91. 50. Lullien-Pellerin, V., Popineau, Y., Meersman, F., Morel, M.H., Heremans, K., Lange, R. & Balny, C. (2001) Reversible changes of the wheat gamma46 gliadin conformation submitted to high pressures and temperatures. Eur. J. Biochem. 268, 5705–5712. 51. Royer, C.A., Hinck, A.P., Loh, S.N., Prehoda, K.E., Peng, X., Jonas, J. & Markley, J.L. (1993) Effects of amino acid substitu- tions on the pressure denaturation of staphylococcal nuclease as monitored by fluorescence and nuclear magnetic resonance spec- troscopy. Biochemistry 32, 5222–5232. 52. Desai,G.,Panick,G.,Zein,M.,Winter,R.&Royer,C.A.(1999) Pressure-jump studies of the folding/unfolding of trp repressor. J. Mol. Biol. 288, 461–475. 53. Jackson, M. & Mantsch, H.H. (1995) The use and misuse of FTIR spectroscopy in the determination of protein structure. Crit. Rev. Biochem. Mol. 30, 95–120. 54. Arrondo, J.L.R., Muga, A., Castresana, J. & Goni, F.M. (1993) Quantitative studies of the structure of proteins in solution by Fourier-transform infrared spectroscopy. Prog. Biophys. Mol. Biol. 59, 23–56. 55. Wong, P.T.T. &.Heremans, K. (1988) Pressure effects on protein secondary structure and hydrogen deuterium exchange in chy- motrypsinogen: a fourier transform infrared spectroscopic study. Biochim. Biophys. Acta 956, 1–9. 56. Goossens, K., Smeller, L., Heremans, K. & Frank, J. (1992) Pressure effect on lipoxygenase: a FTIR study with the diamond anvil cell. High Pressure and Biotechnology. Colloque INSERM (Balny, C., Hayashi, R., Heremans, K. & Masson, P., eds), pp. 171–173. John Libbey Eurotext Ltd, Paris. 57. Ruan, Q., Ruan, K., Balny, C., Glaser, M. & Mantulin, W.W. (2001) Protein folding pathways of adenylate kinase from E. coli: hydrostatic pressure and stopped-flow studies. Biochemistry 40, 14706–14714. 58. Chatani, E., Nonomura, K., Hayashi, R., Balny, C. & Lange, R. (2002) Comparison of heat- and pressure-induced unfolding of ribonuclease A: the critical role of Phe-46 which appears to belong to a new hydrophobic chain-folding initiation site. Biochemistry 41, 4567–4574. 59. King, L. & Weber, G. (1986) Conformational drift of dissociated lactate dehydrogenase. Biochemistry 25, 3632–3637. 60. Ruan, K. & Weber, G. (1993) Physical heterogeneity of muscle glycogen phosphorylase revealed by hydrostatic pressure dis- sociation. Biochemistry 32, 6295–6301. 128 M. Trovaslet et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . Fluorescence and FTIR study of pressure-induced structural modifications of horse liver alcohol dehydrogenase (HLADH) Marie Trovaslet 1 , Sandrine. 128, Montpellier Cedex 5, France The process of pressure-induced modification of horse liver alcohol dehydrogenase (HLADH) was followed by measuring in situ