Microscopic viscosity and rotational diffusion of proteins in a macromolecular environment

The Stokes-Einstein-Debye equation is currently used to obtain information on protein size or on local viscosity from the measurement of the rotational correlation time. However, the implicit assumptions of a continuous and homogeneous solvent do not hold either in vivo, because of the high density of macromolecules, or in vitro, where viscosity is adjusted by adding viscous cosolvents of various size. To quantify the consequence of nonhomogeneity, we have measured the rotational Brownian motion of three globular proteins with molecular mass from 66 to 4000 kD in presence of 1.5 to 2000 kD dextrans as viscous cosolvents. Our results indicate that the linear viscosity dependence of the Stokes-Einstein relation must be replaced by a power law to describe the rotational Brownian motion of proteins in a macromolecular environment. The exponent of the power law expresses the fact that the protein experiences only a fraction of the hydrodynamic interactions of macromolecular cosolvents. An explicit expression of the exponent in terms of protein size and cosolvent's mass is obtained, permitting definition of a microscopic viscosity. Experimental data suggest that a similar effective microviscosity should be introduced in Kramers' equation describing protein reaction rates.     

Protein dynamics and electrostatics in the function of p-hydroxybenzoate hydroxylase

para-Hydroxybenzoate hydroxylase is a flavoprotein monooxygenase that catalyzes a reaction in two parts: reduction of the enzyme cofactor, FAD, by NADPH in response to binding p-hydroxybenzoate to the enzyme, then oxidation of reduced FAD by oxygen to form a hydroperoxide, which oxygenates p-hydroxybenzoate to form 3,4-dihydroxybenzoate. These diverse reactions all occur within a single polypeptide and are achieved through conformational rearrangements of the isoalloxazine ring and protein residues within the protein structure. In this review, we examine the complex dynamic behavior of the protein that enables regulated fast and specific catalysis to occur. Original research papers (principally from the past 15 years) provide the information that is used to develop a comprehensive overview of the catalytic process. Much of this information has come from detailed analysis of many specific mutants of the enzyme using rapid reaction technology, biophysical measurements, and high-resolution structures obtained by X-ray crystallography. We describe how three conformations of the enzyme provide a foundation for the catalytic cycle. One conformation has a closed active site for the conduct of the oxygen reactions, which must occur in the absence of solvent. The second conformation has a partly open active site for exchange of substrate and product, and the third conformation has a closed protein structure with the isoalloxazine ring rotated out to the surface for reaction with NADPH, which binds in a surface cleft. A fundamental feature of the enzyme is a H-bond network that connects the phenolic group of the substrate in the buried active site to the surface of the protein. This network serves to protonate and deprotonate the substrate and product in the active site to promote catalysis and regulate the coordination of conformational states for efficient catalysis.     

Kinetics of the lactate dehydrogenase reaction in high-viscosity media

The effect of the medium viscosity on kinetics parameters of lactate dehydrogenase reaction was studied. The viscosity increase results in a sharp decline in the catalytic rate for both the pyruvate reduction and lactate oxidation reactions. It is shown that the catalytic step and its associated conformational motions is the only step which is considerably retarded when the viscosity increases. The reaction is not sensitive to changes in the dielectric properties of the medium. An inverse power function observed between the rate constant and viscosity cannot be explained by the theory of absolute reaction rates. However, it can easily be interpreted on the basis of the Kramers theory dealing with the transition over the activation barrier as a diffusional motion in the field of random forces. The influence of the medium's viscosity on the kinetic parameters indicates the existence of strong coupling between the dynamics of the solvent and the conformational motions of the protein molecule, which are correlated with the catalytic step.     

Viscosity and transient solvent accessibility of Trp-63 in the native conformation of lysozyme

We have measured the rates of isotope exchange at the nitrogen of the indole ring of Trp-63 of lysozyme and of L-tryptophan as a function of solution viscosity. We have used two cosolvents, glycerol and ethylene glycol, to modify the relative viscosity. We have derived the appropriate kinetic equations for the alternative possibilities that the exchange takes place either in solution or in the intact protein matrix. Because we chose to study the proton-catalyzed exchange reaction, the rate of it is not expected to be diffusion-limited. We confirmed this by measuring the exchange from tryptophan. These results and the known effects of glycerol and ethylene glycol on the solvation of indole allow us to predict that if the exchange reaction takes place in a protein matrix the effects of the two cosolvents when compared under isoviscous conditions should be identical. This is what we find for Trp-63 in lysozyme at 15, 20 and 26 degrees C. The slope of the linear plot of log k vs. log relative viscosity is 0.6. This strongly supports a model for conformational fluctuations where transient solvation takes place without major changes in protein folding. The most interesting feature of our findings is the fact that a slow reaction admittedly not diffusion-limited shows, when taking place in a protein matrix, a linear dependence on solution viscosity. We suggest that what we observe is the effect of damping of movement of the side chain expressed as a change in the friction along the reaction coordinate in the corresponding phase space. The presence of such effects stresses the validity and usefulness of Kramers model of rate processes for reactions taking place in a protein matrix. Such behavior is predicted by several of the recently proposed general mechanisms of enzyme catalysis.     

A flexible interface between DNA ligase and PCNA supports conformational switching and efficient ligation of DNA

DNA sliding clamps encircle DNA and provide binding sites for many DNA-processing enzymes. However, it is largely unknown how sliding clamps like proliferating cell nuclear antigen (PCNA) coordinate multistep DNA transactions. We have determined structures of Sulfolobus solfataricus DNA ligase and heterotrimeric PCNA separately by X-ray diffraction and in complex by small-angle X-ray scattering (SAXS). Three distinct PCNA subunits assemble into a protein ring resembling the homotrimeric PCNA of humans but with three unique protein-binding sites. In the absence of nicked DNA, the Sulfolobus solfataricus DNA ligase has an open, extended conformation. When complexed with heterotrimeric PCNA, the DNA ligase binds to the PCNA3 subunit and ligase retains an open, extended conformation. A closed, ring-shaped conformation of ligase catalyzes a DNA end-joining reaction that is strongly stimulated by PCNA. This open-to-closed switch in the conformation of DNA ligase is accommodated by a malleable interface with PCNA that serves as an efficient platform for DNA ligation.     

Effect of viscosity on the deamidation rate of a model Asn-hexapeptide.

The effect of viscosity on the deamidation rate of a model Asn-containing hexapeptide (l-Val-l-Tyr-Pro-l-Asn-Gly-l-Ala) was assessed in aqueous solution and in solids containing varying amounts of poly(vinyl pyrrolidone) (PVP) and water. Stability studies were conducted at 0.1 mg/mL peptide and 0-50% PVP (w/w) in aqueous solution, and at 5% (w/w) peptide and different relative humidities (31.6, 53.1, 74.4 and 96%) in the solid state. The parent peptide and its deamidation products were analysed by reverse-phase high-performance liquid chromatography. Deamidation rates decreased with increasing solvent viscosity in a manner described by a semi-empirical mathematical model developed to describe this relationship. The results suggest that the motion of the Asn side-chain along the reaction coordinate is a function of the macroscopic solvent viscosity. However, the apparent energy barrier for the diffusive movement of the side-chain appears to be less than the energy barrier for that associated with macroscopic viscosity. The dependence of the deamidation rate on viscosity in both viscous solution and hydrated solids further demonstrates the importance of mobility in peptide deamidation.     

Visualization of energetics and conformations from molecular computer simulations

The quantity of data generated from molecular dynamics simulations and energy minimizations of macromolecules is overwhelming. It is an arduous task to extract the relevant and interesting information from the numerous coordinate sets produced. To help solve this problem, the authors have developed a method to aid the visualization of the relevant information from the simulations. This approach combines animation of the results on a high performance graphics device, such as the PS300, with colour-coded atoms based on changes in energy or conformation. The method will be illustrated using as examples: the molecular mechanics minimization of a nonapeptide, the molecular dynamics simulation of the protein myoglobin, including the analysis of the motion of helices during a 300ps trajectory, and changes in sugar puckering that occur during the molecular dynamics simulation of a DNA oligomer. The method is also applicable for analysing energy components and conformational properties of a fixed conformation.     

Enzymatic transition states: thermodynamics, dynamics and analogue design

Kinetic isotope effects and computational chemistry have defined the transition state structures for several members of the N-ribosyltransferase family. Transition state analogues designed to mimic their cognate transition state structures are among the most powerful enzyme inhibitors. In complexes of N-ribosyltransferases with their transition state analogues, the dynamic nature of the transition state is converted to an ordered, thermodynamic structure closely related to the transition state. This phenomenon is documented by peptide bond H/D exchange, crystallography and computational chemistry. Complexes with substrate, transition state and product analogues reveal reaction coordinate motion and catalytic interactions. Isotope-edited spectroscopic analysis and binding specificity of these complexes provides information about specific enzyme-transition state contacts. In combination with protein dynamic QM/MM models, it is proposed that the transition state is reached by stochastic dynamic excursions of the protein groups near the substrates in the closed conformation. Examples from fully dissociated (D(N) *A(N)), hybrid (D(N)A(N)) and symmetric nucleophilic displacement (A(N)D(N)) transition states are found in the N-ribosyltransferases. The success of transition state analogue inhibitor design based on kinetic isotope effects validates this approach to understanding enzymatic transition states.     

Activation of C55-isoprenoid alcohol phosphokinase from Staphylococcus aureus. II. Biophysical studies.

Physicochemical techniques were used to investigate the activation of C55-isoprenoid alcohol phosphokinase by synthetic lecithins. Complexes of the enzyme with phospholipids were prepared using a method employing sodium dodecyl sulfate as a protein-solubilizing agent. Circular dichorism and the intrinsic fluorescence of the kinase were used as optical probes of protein conformation with these complexes. No evidence for a major lipid-dependent conformational change in the protein was observed when these complexes were studied under conditions where the lipid mesomorphic transitions occurred. EPR studies of mixtures of synthetic lecithins and the C55-isoprenoid alcohol indicated a correlation between kinase activity and the rotational diffusion rate within the hydrophobic phase. It is concluded that the lipid physical state probably does not affect the enzyme activity by altering the protein conformation but more likely does so by affecting the motion of the molecular participants in the reaction.     

Trapping conformational intermediate states in the reaction center protein from photosynthetic bacteria

In protein, conformational changes are often crucial for function but not easy to observe. Two functionally relevant conformational intermediate states of photosynthetic reaction center protein (RCs) are trapped and characterized at low temperature. RCs frozen in the dark do not allow electron transfer from the reduced primary quinone, Q(A)(-), to the secondary quinone, Q(B). In contrast, RCs frozen under illumination in the product (P(+)Q(A)Q(B)(-)) state, with the oxidized electron donor, P(+), and reduced Q(B)(-), return to the ground state at cryogenic temperature in a conformation that allows a high yield of Q(B) reduction. Thus, RCs frozen under illumination are found to be trapped above the ground state in a conformation that allows product formation. When the temperature is raised above 120 K, the protein relaxes to an inactive conformation which is different from the RCs frozen in the dark. The activation energy for this change is 87 +/- 8 meV, and the active and inactive states differ in energy by only 16 +/- 3 meV. Thus, there are several conformational substates along the reaction coordinate with different transition temperatures. The ground state spectra of the RCs in active and inactive conformations report differences in the intraprotein electrostatic field, demonstrating that the dipole or charge distribution has changed. In addition, the electrochromic shift associated with the Q(A)(-) to Q(B) electron transfer at low temperature was characterized. The electron-transfer rate from Q(B)(-) to P(+) was measured at cryogenic temperature and is similar to the rate at room temperature, as expected for an exothermic, electron tunneling reaction in RCs.     

The effect of viscosity on the accessibility of the single tryptophan in human serum albumin.

When reactions take place with one of the reactants tied to protein matrix, movements along the reaction coordinate towards the transition state can become coupled to structural fluctuations of the protein matrix. This investigation aims to test the assumptions underlying the arguments supporting such a coupling. A coupling is allowed only if the activation barrier is high and broad enough as shown to be the case for the proton catalyzed isotope exchange at Trp-63 of lysozyme. In the present investigation the activation barrier for the same reaction has been lowered radically in an effort to show that the coupling, as measured by the dependence of rate on solution viscosity, will diminish and ideally vanish, despite the unchanged effects of cosolvents on the chemical activities of all the reactants. The isotope exchange rate at the indole nitrogen of the single tryptophan residue of human serum albumin was measured with UV. This residue is rigidly held to the protein surface and the solvent access, although restricted, corresponds to a partially exposed residue. As a consequence, the isotope exchange rates and the bimolecular quenching rate of fluorescence by acrylamide, also measured, are high. The experiments were carried out at pH 5.2 where the molecule is in the N-form and the exchange is catalyzed by OH- ions. The activation energy of the hydroxyl catalyzed reaction is 22 kJ lower than for the proton catalyzed process. Under these conditions the exchange rate is viscosity independent both in the case of glycerol and in ethylene glycol.(ABSTRACT TRUNCATED AT 250 WORDS)     

Acylation of human hemoglobin with polyoxyethylene derivatives

Monomethoxypolyoxyethylene (Mw = 5000) was covalently linked to human hemoglobin via an amide bond formed between amino groups of the protein and a carboxylic group introduced onto the polymer. The conjugates thus obtained have a molecular size corresponding to that of a globular protein with a molecular weight of about 190 000. Their oxygen-binding properties depend upon the initial conformation of the hemoglobin and reaction pH: hemoglobin modified in the deoxy state exhibited a lower oxygen affinity than that modified in the oxy state, and the lower the reaction pH, the lower the oxygen affinity of polymer-linked hemoglobin. However, the affinity of modified hemoglobin is always higher than that of native hemoglobin. On the other hand, when deoxyHb was complexed with organic phosphates during the condensation reaction, the resulting conjugates exhibited oxygen-binding characteristics quite similar to those of native hemoglobin, i.e., the same oxygen affinity, modified cooperativity and the same alkaline Bohr effect. Finally, in order to decrease the oxygen affinity of hemoglobin conjugates, the polymer was coupled to deoxy hemoglobin previously covalently modified with pyridoxal phosphate. The oxygen affinity of such conjugates was in fact as low as that of the initial pyridoxylated hemoglobin.     

Folding pathway dependence on energetic frustration and interaction heterogeneity for a three-dimensional hydrophobic protein model

Monte Carlo simulations of a hydrophobic protein model of 40 monomers in the cubic lattice are used to explore the effect of energetic frustration and interaction heterogeneity on its folding pathway. The folding pathway is described by the dependence of relevant conformational averages on an appropriate reaction coordinate, pfold, defined as the probability for a given conformation to reach the native structure before unfolding. We compare the energetically frustrated and heterogeneous hydrophobic potential, according to which individual monomers have a higher or lower tendency to form contacts unspecifically depending on their hydrophobicities, to an unfrustrated homogeneous Go-type potential with uniformly attractive native interactions and neutral non-native interactions (called Go1 in this study), and to an unfrustrated heterogeneous potential with neutral non-native interactions and native interactions having the same energy as the hydrophobic potential (called Go2 in this study). Folding kinetics are slowed down dramatically when energetic frustration increases, as expected and previously observed in a two-dimensional model. Contrary to our previous results in two dimensions, however, it appears that the folding pathway and transition state ensemble can be significantly dependent on the energy function used to stabilize the native structure. The sequence of events along the reaction coordinate, or the order along this coordinate in which different regions of the native conformation become structured, turns out to be similar for the hydrophobic and Go2 potentials, but with analogous events tending to occur at lower pfold values in the first case. In particular, the transition state obtained from the ensemble around pfold = 0.5 is more structured for the hydrophobic potential. For Go1, not only the transition state ensemble but the order of events itself is modified, suggesting that interaction heterogeneity, in addition to energetic frustration, can have significant effects on the folding mechanism, most likely by modifying the probability of different contacts in the unfolded state, the starting point for the folding reaction. Although based on a simple model, these results provide interesting insight into how sequence-dependent switching between folding pathways might occur in real proteins.     

Evidence for an RNA-based catalytic mechanism in eukaryotic nuclear ribonuclease P

Ribonuclease P is the enzyme responsible for removing the 5'-leader segment of precursor transfer RNAs in all organisms. All eukaryotic nuclear RNase Ps are ribonucleoproteins in which multiple protein components and a single RNA species are required for activity in vitro as well as in vivo. It is not known, however, which subunits participate directly in phosphodiester-bond hydrolysis. The RNA subunit of nuclear RNase P is evolutionarily related to its catalytically active bacterial counterpart, prompting speculation that in eukaryotes the RNA may be the catalytic component. In the bacterial RNase P reaction, Mg(II) is required to coordinate the nonbridging phosphodiester oxygen(s) of the scissile bond. As a consequence, bacterial RNase P cannot cleave pre-tRNA in which the pro-Rp nonbridging oxygen of the scissile bond is replaced by sulfur. In contrast, the RNase P reaction in plant chloroplasts is catalyzed by a protein enzyme whose mechanism does not involve Mg(II) coordinated by the pro-Rp oxygen. To determine whether the mechanism of nuclear RNase P resembles more closely an RNA- or a protein-catalyzed reaction, we analyzed the ability of Saccharomyces cerevisiae nuclear RNase P to cleave pre-tRNA containing a sulfur substitution of the pro-Rp oxygen at the cleavage site. Sulfur substitution at this position prohibits correct cleavage of pre-tRNA. Cleavage by eukaryotic RNase P thus depends on the presence of a thio-sensitive ligand to the pro-Rp oxygen of the scissile bond, and is consistent with a common, RNA-based mechanism for the bacterial and eukaryal enzymes.     

The sites of cleavage on Escherichia coli glutamine synthetase by dithiothreitol, Fe(III) and O2.

Native conformation of an enzyme molecule is required for the specific non-enzymatic cleavage of Escherichia coli glutamine synthetase by a metal-catalyzing oxidation system comprised of dithiothreitol, Fe(III) and O2. The cleavage reaction is greatly inhibited by the addition of Mg(II). Two major cleavage sites are identified between amino acid residues 264 and 268, and roughly between amino acid residues 31 and 34, which are located on the protein segments forming the active site of the enzyme. These results suggest that the cleavage reaction is a largely site-specific process involving active oxygen species generated at the divalent cation binding sites on glutamine synthetase.     

Intrinsic viscosity of ovomucoid in random coil conformation.

Ovomucoid is denatured by concentrated solutions of guanidine hydrochloride. The intrinsic viscosities of the glycoprotein in 6 M guanidine hydrochloride in the absence and presence of beta-mercaptoethanol were found to be 8.1 and 16.0 ml/g, respectively. Ovomucoid with disulphide bonds reduced exists in linear random coil conformation. However, the intrinsic viscosity of the randomly coiled protein was less than that predicted from the empirical equations describing the molecular weight dependence of intrinsic viscosities of random coil proteins in 6 M guanidine hydrochloride. On excluding the carbohydrate content of the protein, which is theoretically justified, the calculated intrinsic viscosity interestingly became closer to the measured one. The temperature dependence of the intrinsic viscosity of ovomucoid in linear random coil conformation was studied in the temperature range, 25-55 degrees. The features of the intrinsic viscosity-temperature profile are not comparable with those exhibited by other linear random coil proteins in 6 M guanidine hydrochloride.     

Dynamics involved in catalysis by single-component and two-component flavin-dependent aromatic hydroxylases

Flavoprotein monooxygenases are involved in a wide variety of biological processes including drug detoxification, biodegradation of aromatic compounds in the environment, biosynthesis of antibiotics and siderophores, and many others. The reactions use NAD(P)H and O2 as co-substrates and insert one atom of oxygen into the substrate. The flavin-dependent monooxygenases utilize a general cycle in which NAD(P)H reduces the flavin, and the reduced flavin reacts with O2 to form a C4a-(hydro)peroxyflavin intermediate, which is the oxygenating agent. This complicated catalytic process has diverse requirements that are difficult to be satisfied by a single site. Two general strategies have evolved to satisfy these requirements. para-Hydroxybenzoate hydroxylase, the paradigm for the single-component flavoprotein monooxygenases, is one of the most thoroughly studied of all enzymes. This enzyme undergoes significant protein and flavin dynamics during catalysis. There is an open conformation that gives access of substrate and product to solvent, and a closed or in conformation for the reaction with oxygen and the hydroxylation to occur. This closed form prevents solvent from destabilizing the hydroperoxyflavin intermediate. Finally, there is an out conformation achieved by movement of the isoalloxazine toward the solvent, which exposes its N5 for hydride delivery from NAD(P)H. The protein coordinates these dynamic events during catalysis. The second strategy uses a reductase to catalyze the reduction of the flavin and an oxygenase that uses the reduced flavin as a substrate to react with oxygen and hydroxylate the organic substrate. These two-component systems must be able to transfer reduced flavin from the reductase to the oxygenase and stabilize a C4a-peroxyflavin until a substrate binds to be hydroxylated, all before flavin oxidation and release of H2O2. Again, protein dynamics are important for catalytic success.     

A comparative study on the hydrodynamic shape, conformation, and stability of E. coli ribosomal subunits in reconstitution buffer.

We have compared the hydrodynamic shape, conformation, and stabilities of active, unwashed ribosomal subunits, as well as their susceptibilities to changes in temperature and ionic strength. Both intrinsic viscosity and sedimentation velocity measurements indicate that the 30 S subunit has a more asymmetric hydrodynamic shape. The intrinsic viscosity of this subunit in reconstitution buffer has been found to be significantly larger than the value reported previously. While the RNA conformation in both subunits may be very similar as suggested by the near uv CD spectra, the average conformation of the protein in the two subunits is drastically different. The 30 S subunit has a lower T_m. The 50 S subunit is rather stable toward changes of ionic strength, whereas the 30 S subunit is quite susceptible to changes in ionic strength.     

Effects of solvent viscosity and different guanidine salts on the kinetics of ribonuclease A chain folding

The kinetics of folding of the two forms of unfolded ribonuclease A have been measured as a function of solvent viscosity by adding either glycerol or sucrose. The aim is to find out if either reaction is rate limited by segmental motion whose rate depends on external friction. The fast folding reaction (U₂ ? N) is known to be the direct folding process, and the slow folding reaction (U₁ ? N) is known to be rate limited by an interconversion between two forms (U₁ ? U₂) which are present after unfolding in strongly denaturing conditions. No dependence on solvent viscosity is found, either for the direct folding reaction or for the interconversion reaction. Each folding reaction has also been tested to see if its rate depends on the concentration of one or more partly folded intermediates, by adding denaturants destabilize any partly folded structures. Different guanidine salts are used as denaturants to vary the denaturing effectiveness of the salt while holding the guanidinium ion concentration constant. The rates both of the direct folding reaction and of the interconversion reaction decrease in relation to the denaturing effectiveness of the salt. However, there is a basic difference between the responses of the fast and slow folding reactions to low concentrations of denaturants. Although each folding reaction produces native protein, there is an 800-fold decrease in the rate of the fast folding reaction in 1M guanidine thiocyanate and only a 13-fold decrease in the rate of the slow folding reaction. This is consistent with the fast reaction being the direct folding process and the slow reaction being rate limited by the initial conversion of the slowrefolding to the fast-refolding form. Both the lack of viscosity dependence and the effects of denaturants indicate that the formation of structure is rate limiting in the direct folding reaction, U₂ ? N. The failure to find a viscosity dependence for the interconversion reaction, U₁ ? U₂, indicates that in this reaction also friction-limited segmental motion is not the rate-limiting process. Since the U₁ ? U₂ interconversion still occurs when the polypeptide chain is completely unfolded, the surprising result is that its rate in refolding conditions depends significantly on a reaction intermediate which is “denatured” by guanidine salts.     

The role of hydration on the mechanism of allosteric regulation: in situ measurements of the oxygen-linked kinetics of water binding to hemoglobin

We report here the first direct measurements of changes in protein hydration triggered by a functional binding. This task is achieved by weighing hemoglobin (Hb) and myoglobin films exposed to an atmosphere of 98% relative humidity during oxygenation. The binding of the first oxygen molecules to Hb tetramer triggers a change in protein conformation, which increases binding affinity to the remaining empty sites giving rise to the appearance of cooperative phenomena. Although crystallographic data have evidenced that this structural change increases the protein water-accessible surface area, isobaric osmotic stress experiments in aqueous cosolutions have shown that water binding is linked to Hb oxygenation. Now we show that the differential hydration between fully oxygenated and fully deoxygenated states of these proteins, determined by weighing protein films with a quartz crystal microbalance, agree with the ones determined by osmotic stress in aqueous cosolutions, from the linkage between protein oxygen affinity and water activity. The agreements prove that the changes in water activity brought about by adding osmolytes to the buffer solution shift biochemical equilibrium in proportion to the number of water molecules associated with the reaction. The concomitant kinetics of oxygen and of water binding to Hb have been also determined. The data show that the binding of water molecules to the extra protein surface exposed on the transition from the low-affinity T to the high-affinity R conformations of hemoglobin is the rate-limiting step of Hb cooperative reaction. This evidences that water binding is a crucial step on the allosteric mechanism regulating cooperative interactions, and suggests the possibility that environmental water activity might be engaged in the kinetic control of some important reactions in vivo.