Electrostatic interactions in protein structures: The influence of water on enzyme fluctuations

Journal of Electrostatics, 21 (1988) 245-256

245

Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands

E L E C T R O S T A T I C I N T E R A C T I O N S IN P R O T E I N S T R U C T U R E S : T H E I N F L U E N C E OF W A T E R ON E N Z Y M E F L U C T U A T I O N S

S. BONE

Institute of Molecular and Biomolecular Electronics, University College of North Wales, Bangor (Great Britain) (Received February 9, 1988; accepted in revised form April 15, 1988)

Summary A discussion is presented of the close relationships that exist between protein-water interactions, protein structural fluctuations and enzyme activity. This discussion is based on a review of the various X-ray crystallographic, dielectric and NMR investigations that have addressed the problem of the nature and extent of protein hydration.

1. Introduction

In the past water was regarded as little more than a suspending medium for the active biological molecules with little or no significant contribution to biological processes. It has only recently become recognised that water plays a critical role in determining the physical and chemical properties of proteins which are ultimately responsible for the unique biological activity of these molecules. The property which, more than any other, determines the way in which water interacts with a protein is its polar character and the drive to maximise the number of electrostatic hydrogen bond interactions with the protein and with other water molecules to achieve the lowest energetic state. Also, as a consequence of its polar character, water has a high dielectric constant (79 at room temperature) and has the ability to weaken other electrostatic interactions, including, most importantly for the discussion here, those between polar groups in the protein structure. Protein molecules have, until recently, been regarded as static, rigid molecules having, in the case of enzymes, the rather surprising ability to accelerate reactions by factors of at least a million. The fluctuating nature and the dynamic and the thermodynamic properties of protein structures are only now being investigated with a view to understanding the physical processes which must underpin such activity [1-5 ]. Again, electrostatic interactions between polar components of the protein structure and the way that these are influenced by water must be crucial in determining the dynamic behaviour of the 0304-3886/88/$03.50

© 1988 Elsevier Science Publishers B.V.

246 protein structure. In this article the close relationship between protein-water interaction, protein structural fluctuations and protein activity are discussed. 2. P r o t e i n - w a t e r interactions

2.1 Water-binding sites Infrared [6,7 ] and adsorption isotherm [8 ] hydration studies involving the incremental addition of water to the protein from a dry state have been instrumental in elucidating the sequence of group hydration in the protein. The ionisable groups ( C O 0 - and NH + ) appear to be the first hydrated. The existence of soluble, ionisable residues allows the initial condensation of water molecules onto the protein and the hydration and ionisation of these are the primary processes in protein hydration. As described by Baker and Hubbard [9], the COO- groups of aspartic (Asp) and glumatic acid (Glu) are able to bind, on average, two water molecules each, which constitutes some 34% of the total water bound by the protein. Some Glu and Asp groups which make fewer hydrogen bonds with the protein are able to bind three or even four water molecules. Water is bound to the polar side chains and polar protein backbone at higher hydration levels and this process is competitive. Analysis of the water component of 13 protein crystal structures [9] suggests that 44% of side-chain oxygens and nitrogens are hydrated compared with 42% of peptide carbonyl oxygen atoms. In contrast, only 14% of the main-chain amides are hydrated, presumably since they are, on average, sterically less accessible than the peptide carbonyl oxygens. Most polar amino acid residues tend to bind an average of one water molecule per side chain. Solvent accessibility may be a determining factor in the extent to which particular side groups are hydrated. By virtue of their lengths, lysine, arginine and glutamic acid are most likely to protrude into the solvent. By contrast, histidine and aspartic acid often have low solvent accessibilities as a result of a bulky aromatic ring and short side chain, respectively. The hydrophilic groups are not randomly arranged in the protein structure. This is due to the fact that there is a favourable decrease in the protein's free energy when non-polar groups are withdrawn from contact with an aqueous environment to the hydrophobic protein interior and it is then energetically unfavourable for polar groups to be buried in a dehydrated state in the nonpolar interior of the protein. So non-polar groups tend to lie in the protein interior and polar ones near to, or on, the protein surface. In general, therefore, protein hydration can be said to take place predominantly on the protein surface. There are notable exceptions to this rule. Even small proteins can contain buried water molecules, which may occupy internal cavities. Lysozyme, for example, incorporates four internal water molecules [10] and cytochrome c

247 has three [ 11 ]. These internal water molecules nearly always electrostatically compensate internal polar or charged groups by forming hydrogen bonds with them, thus stabilising the local structure. For example, actinidin, a plant protein, contains 11 internal peptide carbonyl groups and 10 internal amide groups which are compensated by internal water molecules [12]. A number of proteins contain extensive internal water networks, for instance 6 water molecules form an internal cluster in trypsin [13], whilst in actinidin 8 of the 16 internal waters are arranged in one network [12]. Also where two or more internal polar groups are unable to make direct hydrogen bonding contact, water molecules often form a bridge between them [ 14 ]. This phenomenon is evident in the structure of carboxypeptidase-A [ 15 ] and papain [ 16 ]. Although this water is situated in the interior of the protein, solvent exchange experiments have shown that it is still accessible to the bulk water environment, although the exchange rates are much reduced compared with those of surface-bound water [17]. In addition to buried internal water, there are other water molecules which, while being tightly bound to the protein, are also accessible to the bulk aqueous environment. These are often located in surface pockets and crevices. Although there is a tendency for hydrophobic groups to be internal and hydrophilic groups to be on the protein surface, this is a generalisation. Hydrophobic groups are found on the surface of the great majority of proteins. In fact many proteins possess considerable areas of surface hydrophobicity since there is also tendency for clusters to form, that is for non-polar groups to be preferentially surrounded by other non-polar groups and likewise for polar groups [18,19]. These areas of surface hydrophobicity are hydrated last, with this hydration taking the form of the spreading of water clusters from the ionised and polar groups to cover the hydrophobic regions. It seems most likely that water bridges form over the non-polar region [ 14 ].

2.2 Water and protein conformation Water plays a crucial role in the conversion of the unfolded nascent polypeptide chain into the active, native protein. Primarily the native structure is under the direction of the amino acid sequence itself [ 20 ]. A variety of intramolecular interactions influence the final protein conformation; electrostatic and torsional interactions, dispersion forces and most importantly solvent interaction. Polar water molecules compete with possible intramolecular interactions between polar groups of the protein and also take part in hydrophobic interactions with non-polar amino acid side groups, making the protein interior a more energetically favourable environment for these species. Although the peptide backbone is composed of polar amide and carbonyl groups, interaction with the aqueous environment is sometimes not the most energetically favourable option. Hydrogen bonding between the amide nitrogen and carbonyl oxygen in a helical or sheet structure provides for electrostatic compensation and is relatively strong in a non-aqueous environment. Also, as we have

248 seen, the interaction of water molecules with the peptide components, particularly the main-chain amides, is, for geometric reasons, relatively weak. In homopolypeptides, where the side chains do not impose structural irregularity, this results in helical conformations where there is maximisation of the hydrogen bonding between the peptide components. 2.3 The state of protein-bound water An important aspect of protein-water interaction is the state of the bound water in terms of its translational and rotational properties compared with those of normal bulk water. Some information regarding this has come from X-ray crystallographic studies. In protein crystals approximately half the total volume is occupied by water, the protein molecules being surrounded largely by water. Although there can be difficulties resolving the water molecules even in high-resolution studies, a number of well-resolved solvent peaks have been reported for several proteins. Water molecules located inside the protein have been those most easily defined, being hydrogen bonded to a number of buried polar side groups or to other water molecules and therefore relatively rotationally immobile. Solvent peaks allocated to water molecules bound by two or more hydrogen bonds have also been reported located in the active sites of a number enzymes [21,22], and also on the surface of a number of different proteins where up to 82 water molecules per protein have been detected [10,21,23-25]. The majority of the solvent in the protein crystal, however, appears to be relatively unperturbed by the protein, indicated by the absence of well-resolved electron density peaks over most regions not occupied by the protein molecule. Solvent peaks which might be attributed to water molecules bound by single hydrogen bonds to the protein are much less well defined and it appears that water molecules appear "fixed", as far as X-ray studies are concerned, only if they are anchored by multiple hydrogen bonding to the protein. The extent to which these water molecules are "fixed" has been elucidated to some extent by dielectric studies on hydrated proteins. In a study on hen egg-white lysozyme [26], 36 water molecules were found to be dielectrically invisible, implying that they were irrotationally bound to the enzyme. This value compared well with the X-ray crystallographic data on human lysozyme [10] which detected 34 well-resolved solvent peaks attributed to water molecules making two or more hydrogen bonds with the enzyme. The dielectric studies seem to imply that these water molecules are tightly incorporated into the vibrating protein structure with relaxation times equivalent to those of the protein. Similar dielectric studies [27,28 ] on hydrated proteins both in the frequency and time domains have also detected populations of water molecules which have no influence on the permittivity of the protein, and are therefore assumed to be rotationally immobile. Good correlation has been found [27 ] between the

249

number of water molecules constituting this population and the B.E.T. monolayer hydration value describing the number of primary sorption sites in gravimetric measurements on hydrated ~-chymotrypsin. Dielectric studies [2630 ] have also indicated that water bound in addition to this population is only slightly rotationally hindered, having a room temperature relaxation frequency of 10 GHz compared with 17 GHz for normal bulk water. The apparent existence of a group of tightly bound water molecules seems to be at variance with the conclusions reached in a number of N M R studies [31-33 ]. In contrast with the dielectric and X-ray techniques, however, NMR provides both rotational and translational data and it is apparent from the translational component that all protein-bound water, including internal water, is able to exchange rapidly with the aqueous phase. In addition, as pointed out by Grigera et al. [30], unless line-width measurements are carried out over a wide frequency range, N M R is unable to distinguish between different rotational correlation times. The derived rotational correlation times in fact represent weighted averages where there are more than one species of water molecules. In X-ray crystallographic and dielectric studies where translational effects are not observed, the terms "fixed" and "strongly bound" refer to sites which are generally occupied by water molecules which are relatively rotationally immobile. So far the nature of the water bound to polar regions of the protein has been discussed. However, at high hydrations water clusters spread over the protein surface and interact with exposed apolar hydrophobic groups. The question then arises as to the nature and extent of this type of interaction. In the past it has been suggested that the water molecules surrounding such hydrophobic patches may be partially stabilised in cage-like structures such as those known to exist around non-polar atoms in the clathrate hydrates [34]. These hydrate cage-like water structures are formed when non-polar solutes are dissolved in water. None are able to readily hydrogen bond to water and all are sparingly soluble. The process of dissolving such substances in water is accompanied by a significant decrease in entropy and this has been interpreted as a "structure making" phenomenon. Since the non-polar solutes occupy space, the random hydrogen bond network must reorganise around it in such a way that sufficient room is available to accomodate the guest molecule while at the same time maximising the number of hydrogen bonds and avoiding too much damage to the hydrogen bond network. The cage is usually far from perfect and although there is some strengthening of hydrogen bonds compared to those present in bulk water, substantial strain and disorder remain. The resulting structures are by no means solid or long-lived but exhibit increased correlation times and lessened irregularity. A computer simulation study [35] of the structure of liquid water at an extended hydrophobic surface has shown density oscillations that extend at least 10 A into the liquid and molecular orientational preferences that extend at

250 least 7 ~,. Also dielectric studies employing very sensitive dielectric difference measurements in the range 8 to 25 GHz have investigated the influence of hydrophobic solutes on the dynamic behaviour of water [ 36 ]. It was considered that the dielectric relaxation rate of water was significantly slowed down by the dissolution of hydrophobic solutes indicating increased hydrogen bonding between water molecules in the immediate vicinity of the hydrophobic solute. What are the implications of these studies for water surrounding hydrophobic patches on proteins? The positions of water molecules surrounding nonpolar sites on the protein surface are not observed in X-ray crystallography but this is to be expected, since it is likely that the mobility of the water molecules in the partial clathrate-type of structure would be only slightly reduced compared with that of the bulk aqueous environment. In fact there is little experimental evidence to confirm or refute a scheme involving a greater structuring of water surrounding clusters of hydrophobic groups on the protein surface. NMR, X-ray crystallography and dielectric studies have indicated that there is a relatively large group of water bound to the protein which exhibits very similar properties to normal bulk water. Inspite of this, the freezing properties of this water are not normal in that the so-called non-freezing fraction of protein-bound water exhibits a considerable freezing point depression. For instance, for lysozyme approximately 0.32 g per g protein remains in the liquid state down to temperatures below 213 K. This fraction was considered to be equivalent to the amount of water required to satisfy all the exposed polar groups [37]. The non-freezing phenomenon has been observed with a variety of experimental techniques but the reasons for it are poorly understood. Evidently this water is unable to form a three-dimensional ice crystal structure and, being bound by single hydrogen bonds to the protein, is able to retain similar rotational and translational properties to bulk water at temperatures well below normal freezing.

2.4 Water and biological activity The most important hydration-induced property, and of central interest to hydration studies, is the onset of biological activity. A dry protein is completely inactive but activity is achieved at a relatively low hydration, and certainly well below solution conditions. For example, lysozyme begins to show signs of activity at about 20 weight percent water [38]. There may be a number of reasons for the onset of activity at this particular hydration level. The enzyme may require this amount of water to finally achieve the correct conformational state, the substrate might only find access to the active site possible, or the enzyme might only achieve the required kind of internal vibrational motion above this level of hydration. The rate of activity of lysozyme at 0.2 g protein per g water is very low but increases rapidly with hydration up to 0.5 g/g after which it follows a logarithmic rise to solution conditions [39].

251

3. Protein dynamics and biological function

3.1 Protein fluctuations In the past proteins were thought of as having static, rigid structures with enzymes and substrates interacting in a simple "lock and key" or jigsaw configuration. Even allosteric proteins, where large conformational changes were known to be induced by ligand binding, were treated as proteins in which there were two possible static conformational states. In more recent years there has been a gradual acceptance of a more dynamic picture of protein structure. Evidence of considerable structural motions has come from X-ray crystallography where poorly-resolved parts of the protein structure have indicated motional freedom, and from solvent accessibility studies where solvent exchange has been observed in parts of the internal protein structure which, from static packing considerations, should be solvent inaccessible. High-resolution X-ray structural studies on myoglobin [40,41] have shown that no static pathways exist by which ligands such as oxygen or carbon monoxide could gain access to the heme-binding site from the protein surface. Fluctuations of the protein structure must therefore be involved in the entrance and exit of these ligands from the protein interior. It is interesting to note that substrate binding often has a marked effect on protein structural motions. Solvent exchange studies carried out on such protein-ligand associations as trypsin/trypsin inhibitor [42] and lysozyme/nacetyl glucosamine [43] have indicated a significant decrease in the extent of structural fluctuations compared with the protein and ligand as free entities. In this cases the effect has not been confined to the immediate area of the protein-ligand association but has extended over a large proportion of the protein. It appears, therefore, that the bound substrate elicits an influence which is not limited to simply stabilising the region of the active site. Since X-ray studies are unable to provide direct information about protein motions, the protein structures described by X-ray crystallography can only be regarded as time-averaged structures. It is likely that there are a range of conformational states with similar energies determined by cooperative and correlated fluctuations about that average. This follows from the fact that protein folding is stabilised by weak interactions which are not much more energetic than thermal motions, so that the various interactions determining conformation may, in some proteins, be so finely balanced that a number of similar conformations may coexist. Evidence for the existence of alternative protein conformations comes from X-ray crystallographic studies where a number of alternative protein structures have been observed for the same molecular unit in the same crystal structure [44,45] and also for different crystalizing conditions [23]. Since comparisons between crystal and solution properties have, on the whole, shown a high degree of similarity, these different structures observed in crystal studies may be examples of variations which

252 exist in solution. A more concise description of the extent of conformational fluctuations has come from an X-ray diffraction study on ferrimyoglobin. Frauenfelder et al. [46] have investigated changes in the conformation of this protein over the temperature range 250 to 300 K and have produced a picture of the extent to which each part of the protein structure is able to fluctuate from the static average. Other X-ray studies on ribonuclease [47] and lysozyme [ 10 ] and denaturation studies on a number of enzymes [48 ] have shown that for these enzymes the active site region is the area which supports the largest amplitude structural fluctuations. The nature of the atomic bonding is crucial in determining the motional freedom of individual atoms. For instance the ~-bonding in the aromatic rings of a number of side chains and in the peptide backbone will restrict the rotational freedom of the groups involved. In addition there will often be steric considerations. For instance, steric restrictions to side-chain motion will result from collisions with other side chains and with the peptide backbone. In fact from packing considerations, in many cases it is likely that concerted sidechain motion is the rule rather than the exception. The scales of protein internal motions in terms of amplitude, energy and time vary over many orders of magnitude indicating a range of possible individual, collective and cooperative types of motions. For instance, small, closely packed groups of atoms may experience very short-time, small-amplitude motions in the order of 10-:3 s and 0.2 A. In less closely packed regions of the protein, much larger displacements may occur, possibly involving a partial unfolding of a region of the protein. The types of motion involved may include rotational orientation of groups about individual bonds, the concerted movement of a number of side chains or section of the main peptide chain relative to the rest of the protein structure, or the fluctuating motion of counterions or solvent molecules associated with the protein.

3.2 The influence of protein-bound water One of the thermodynamic considerations of protein dynamics is the flexibility of the protein structure and implicit in a discussion of protein structural flexibility is the influence of protein-bound water. In recent dielectric studies on hydrated lysozyme [28] and chymotrypsin [27] the influence of the electrostatic interaction between the enzyme and bound solvent was discussed. It was suggested that loosely bound water molecules, that is those assumed to be bound by single hydrogen bonds, might be able to act as a plasticizer to the protein structure. This plasticizing action involved the screening of electrostatic interactions between the protein's polar and charged groups. Loosely bound water molecules exhibited an effective dipole moment only slightly less than that observed for normal bulk water and were expected to exert a relatively large dielectric screening effect. It was envisaged that this type of hydration-dependent screening might be allowing the protein specific vibrational

253 modes discussed above, these being important in the protein's biological activity. This scheme was supported by the observation that a significant increase in the motional freedom of the protein structure occurred at hydration levels approaching that at which the onset of the enzyme activity of lysozyme and chymotrypsin has been observed. This was also consistent with ESR spin probe measurements on lysozyme which have indicated a direct correlation between the motional freedom of the lysozyme structure and the extent of enzymatic activity [39 ].

3.3 Enzyme activity As discussed above, one of the most important aspects of protein dynamics is the possibility that certain vibrational modes may be involved in biological activity. The possibility that such fluctuations play a part in, for instance, enzyme action may go some way to explaining a long-standing fundamental question regarding the relative smallness of the enzyme's active site compared with its total volume. The previously held view that the total enzyme structure is required to form and maintain the three-dimensional conformation of the active site is now in doubt. An alternative explanation might be that large enzyme structures are required to transmit, in a very specific way, the translational energy of solute molecules to the enzyme's active site. Of the many possible fluctuations of the enzyme structure, some may be of a correlated, directional nature which serve to provide the optimum energetic conditions at the active site required to produce the transition state and enzyme activity. Careri et al. [49] have suggested a model of enzyme action which involves dynamic coupling of the thermal energy of the ambient medium to the catalytic events at the active site. They have assumed that this type of coupling should be apparent in the fluctuational behaviour of the protein and have estimated that coupling of the solvent's thermal energy to the catalytic process requires events at the protein/solvent interface on a 10 -s s time scale. In fact quite a number of different dynamic processes associated with the enzyme surface such as proton transfer reactions, local conformational motion, counterion fluctuations and bound water relaxations, take place around this time scale. This is thought to indicate the likelihood that the solvent/protein interface and the active site are statistically coupled and that free energy can be exchanged between the environment and the enzyme through coupling events at the protein surface. This model has been extended [39] to consider enzymes as being composed of domains which, on the time scale of interest (about 10 -s s), are themselves internally relatively rigid structures. It is envisaged that each of the domains have some degree of freedom and are able to interact with the substrate with one or two critical residues per domain at the active site. In this model domain rigidity is important in reducing the number of degrees of freedom of the enzyme, so reducing the search time for the optimum active site conformation. Thus, a large number of small solvent interactions at

254 t h e p r o t e i n surface are e n v i s a g e d as c o n t r o l l i n g t h e d i s p l a c e m e n t of t h e c a t a lytic residues at the active site. In this s c h e m e i n t e r n a l a n d s u r f a c e - b o u n d w a t e r are likely to be i m p o r t a n t in d e t e r m i n i n g t h e p o s i t i o n a n d b o u n d a r i e s of t h e rigid d o m a i n s .

4. Concluding remarks W e h a v e s e e n h o w biological a c t i v i t y is i n t i m a t e l y r e l a t e d to specific p r o t e i n p r o p e r t i e s w h i c h m a y be critically d e p e n d e n t u p o n p r o t e i n - w a t e r i n t e r a c t i o n . O f t h e s e t h e d i r e c t i o n of t h e p r o t e i n s t r u c t u r e t o w a r d s a n active c o n f o r m a t i o n is c e r t a i n l y a critical process. P e r h a p s equally i m p o r t a n t t h o u g h is t h e c o u p l i n g of free e n e r g y p r o v i d e d b y t h e a q u e o u s m e d i u m to t h e p r o t e i n s t r u c t u r e a t t h e surface of t h e p r o t e i n a n d t h e t r a n s m i s s i o n of t h i s e n e r g y to t h e active site. I t s e e m s likely t h a t w a t e r is i n v o l v e d in b o t h of t h e s e processes. A l t h o u g h it m a y be p r e m a t u r e to r e a c h a n y definite conclusions, t h e p r e p o n d e r a n c e of experim e n t a l a n d t h e o r e t i c a l e v i d e n c e s e e m s to be c o n s i s t e n t w i t h t h e view of a n e n z y m e as a n e n e r g y t r a n s d u c i n g s y s t e m w i t h w a t e r a c t i n g as t h e m e d i a t o r .

References 1 F.R.N. Gurd and T.M. Rothgeb, Motions in proteins, Adv. Protein Chem., 33 (1979) 73165. 2 M. Karplus and J.A. McCammon, Dynamics of proteins: elements and function, Annu. Rev. Biochem., 53 (1983) 263-300. 3 G.R. Welch, B. Somogyi and S. Damjanovich, The role of protein fluctuations in enzyme action: a review, Prog. Biophys. Mol. Biol., 39 (1982) 109-146. 4 G.R. Welch, (Ed.), The Fluctuating Enzyme, Wiley, New York, NY, 1986. 5 R. Pethig and D.B. Kell, The passive electrical properties of biological systems: their significance in physiology, biophysics andbiotechnology, Phys. Med. Biol., 32(8) (1987) 933-970. 6 P.L. Poole and J.L. Finney, Sequential hydration of a dry globular protein, Biopolymers, 22(1) {1983) 255-260. 7 G. Careri, E. Gratton, P.-H. Yang and J.A. Rupley, Correlation of IR spectroscopic, heat capacity, diamagnetic susceptibility and enzymatic measurements on lysozyme powder, Nature, 284 (1980) 572-573. 8 C.H. Rochester and A.V. Westerman, Sorption of water vapour by some derivatives of bovine serum albumin, J. Chem. Soc. Farad. Trans. 1, 73 (1977) 33-43. 9 E.N. Baker and R.E. Hubbard, Hydrogen bonding in globular proteins, Prog. Biophys. Mol. Biol., 44 (1984) 97-179. 10 C.C.F. Blake, W.C.A. Pulford and P.J. Artimiuk, X-ray studies of water in crystals of lysozyme, J. Mol. Biol. 167 (1983) 693-723. 11 T. Takano and R.E. Dickerson, Conformational change of cytochrome c, J. Mol. Biol., 153 (1982) 79-94. 12 E.N. Baker, Structure of actinidin, after refinement at 1.7 A resolution, J. Mol. Biol., 141 ( 1980 ) 441-484. 13 W. Bode and P. Schwager, Refined crystal structure of bovine fl-trypsin at 1.8 A resolution, J. Mol. Biol., 98(4) (1975) 693-717.

255 14 15 16 17 18 19 20 21 22 23

24 25 26 27 28 29 30 31 32 33 34 35 36 37

38

J.L. Finney, The organisation and function of water in protein crystals, Philos. Trans. R. Soc. London, Ser. B, 278 (1977) 3-32. J.A. Hartsuck and W.N. Lipscomb, Carboxypeptidase A, in: P.D. Boyer (Ed.), The Enzymes, Academic Press, New York, NY, 1972, Vol. 3, pp. 1-56. H.J.C. Berendsen, Specific interactions of water with biopolymers, in: F. Franks (Ed.), Water: A Comprehensive Treatise, Plenum, New York, NY, 1975, pp. 293-330. A.A. Kossiakoff, Protein dynamics investigated by the neutron diffraction-hydrogen exchange technique, Nature, 296 (1982) 713-721. W.R. Krigbanm and A. Komoriya, Local interactions as a structure determinant for protein molecules: II, Biochim. Biophys. Acta, 576 (1979) 204-228. I.D. Kuntz, Tertiary structure of carboxypeptidase, J. Am. Chem. Soc., 94 (1972) 8568-8572. C.B. Anfinsen, Principles that govern the folding of protein chains, Science, 181 (1973) 223230. J. Moult, A. Yonath, W. Traub, A. Smilansky, A. Podjarny, D. Rabinovich and A. Saya, The structure of triclinic lysozyme at 2.5 A resolution, J. Mol. Biol., 100 {1976) 179-195. F.A. Quiocho and W.N. Lipscomb, Carboxypeptidase A: a protein and an enzyme, Adv. Protein Chem., 25 (1971) 1-78. T. Imoto, L.N. Johnson, A.C.T. North, D.C. Phillips and J.A. Rupley, Vertibrate lysozymes, in: P.D. Boyer (Ed.), The Enzymes, Academic Press, New York, NY, 1972, Vol. 7, pp. 665868. J.J. Birktoft and D.M. Blow, Structure of crystalline ~-chymotrypsin, J. Mol. Biol., 68 (1972) 187-240. K.D. Watenpaugh, T.N. Margulis, L.C. Sieker and L.H. Jensen, Water structure in a protein crystal: rubredoxin at 1.2 ]k resolution, J. Mol. Biol., 122 (1978) 175-190. S. Bone and R. Pethig, Dielectric studies of the binding of water to lysozyme, J. Mol. Biol., 157 (1982) 571-575. S. Bone, Time-domain reflectrometry studies of water binding and structural flexibility in chymotrypsin, Biochim. Biophys. Acta, 916 (1987) 128-134. S. Bone and R. Pethig, Dielectric studies of protein hydration and hydration-induced flexibility, J. Mol. Biol., 181 (1985) 323-326. S.C. Harvey and P. Hoekstra, Dielectric relaxation spectra of water absorbed on lysozyme, J. Phys. Chem., 76 {1972) 2987-2994. J.R. Grigera, F. Vericat, K. Hallenga and H.J.C. Berendsen, Dielectric properties of hydrated collagen, Biopolymers, 18 (1979) 25-45. R.G. Bryant, N.M.R. relaxation studies of solute-solvent interactions, Annu. Rev. Phys. Chem., 29 (1978) 167-188. L. Grosch and F. Noack, N.M.R. relaxation investigation of water mobility in aqueous bovine serum albumin solutions, Biochim. Biophys. Acta, 453 (1976) 218-232. S.H. Koenig, K. Hallenga and M. Shporer, Protein water interaction studied by solvent 1H, 2H and 170 magnetic relaxation, Proc. Natl. Acad. Sci. U.S.A., 72 {1975) 2667-2671. I.M. Klotz, Water: its fitness as a molecular environment, in: E.E. Bittar (Ed.), Membranes and Ion Transport, Wiley Interscience, New York, NY, 1970, Vol. 1, pp. 116-129. C.Y. Lee, J.A. McCammon and P.J. Rossky, The structure of liquid water at an extended hydrophobic surface, J. Chem. Phys, 80 (9) (1984) 4448-4455. K. Hallenga, J.R. Grigera and H.J.C. Berendsen, Influence of hydrophobic solutes on the dynamic behaviour of water, J. Phys. Chem., 84 (1980) 2381-2390. J.L. Finney and P.L. Poole, Protein hydration and enzyme activity: the role of hydrationinduced conformation and dynamic changes in the activity of lysozyme, Comments Mol. Cell. Biophys., 2 (1984) 129-151. P.-H Yang and J.A. Rupley, Protein-water interactions. Heat capacity of the lysozyme water system, Biochemistry, 18 (1979) 2654-2661.

256 39 40 41 42 43 44 45 46 47 48 49

J.A. Rupley, E. Gratton and G. Careri, Water and globular proteins, Trends Biochem. Sci., 8(1) (1983) 18-22. M.F. Perutz and F.S. Mathews, An X-ray study of azide methaemoglobin, J. Mol. Biol., 21 (1966) 199-202. S.E. Philips, Structure ofoxymyoglobin, Nature, 273 (1978) 247-248. C.K. Woodward, Dynamic solvent accessibility in soybean trypsin inhibitor-trypsin complex, J. Mol. Biol., 111 (1977) 509-515. R.R Wicket, G.J. Ide and A. Rosenberg, A hydrogen exchange study of lysozyme conformational changes induced by inhibitor binding, Biochemistry, 13 (1974) 3273-3277. A. Tulinsky, R.L. Vandlen, C.N. Morimoto, N.V. Mani and L.H. Wright, Variability in the tertiary structure of a-chymotrypsin at 2.8 A resolution, Biochemistry, 12 ( 1973 ) 4185-4192. A. Mavridis, A. Tulinsky and M.N. Liebman, Asymmetrical changes in the tertiary structure of ~-chymotrypsin with change in pH, Biochemistry, 13 (1974) 3661-3666. H. Frauenfelder, G.A. Petsko and D. Tsernoglou, Temperature-dependent X-ray diffraction as a probe of protein structural dynamics, Nature, 280 (1979) 558-563. F.M. Richards, The interpretation of protein structures: total volume, group volume distributions and packing density, J. Mol. Biol., 82 (1974) 1-14. C.-L. Tsou, Location of the active sites of some enzymes in limited and flexible molecular regions, Trends Biochem. Sci., 11 (1986) 427-429. G. Careri, P. Fasella and E. Gratton, Enzyme dynamics: the statistical physics approach, Annu. Rev. Biophys. Bioeng, 8 (1979) 69-97.