摘要:

DISCUSSIONS OF THE FARADAY SOCIETY No. 20, 1955 THE PHYSICAL CHEMISTRY OF ENZYMES THE FARADAY SOCIETY Agents f o r the Society's Publications : The Aberdeen University Press Ltd. 6 Upper Kirkgate A berdeenThe Faraday Society reserves the copyright of all Communications published in the ‘‘ Discussions ” PUBLISHED 1956 PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS A B E R D E E NA GENERAL DISCUSSION ON THE PHYSICAL CHEMISTRY OF ENZYMES A GENERAL DISCUSSION on the Physical Chemistry of Enzymes was held in the University Laboratory of Physiology, Parks Road, Oxford (by kind permission of the Vice-Chancellor) on the 10, 11 and 12th August, 1955. The President, Prof. R. G. W. Norrish, F.R.S., was in the Chair and over 250 members and visitors were present. Among the distinguished members and guests welcomed by the President Dr.S. J. Adelstein (Boston, Mass.), Dr. R. L. Baldwin (Wisconsin), Dr. H. J. Berg (Eindhoven), Dr. and Mrs. F. Bergmann (Jerusalem), Dr. S. A. Bernhard (Bethesda, Md.), Dr. and Mrs. A. Bloch (Brunswick, N.J.), Dr. J. J. Blum (Bethesda, Md.), Prof. C. J. F. Bottcher (Leiden), Dr. D. J. Botts (Bethesda, Md.), Prof. and Mrs. P. D. Boyer (Minnesota), Dr. W. Brackman (Amsterdam), Prof. and Mrs. Britton Chance (Pennsylvania), Dr. J. A. Cohen (Holland), Prof. Harold Edelhoch (Kansas), Dr. and Mrs. Eric Ellenbogen (Pittsburgh), Dr. F. G. M. Elliott (Belgium), Dr. F. L. Feitelson (Jerusalem), Dr. Seymour L. Friess (Bethesda, Md.), Dr. Maria Fuld (Pittsburgh), Dr. Alfred Gierer (Tubingen), Mr. L. Ginsaar (Holland), Prof.and Mrs. S. F. Gomes da Costa (Lisbon),Prof. B. C. Guha (Calcutta), Dr. W. F. Harrington (Copenhagen), Prof. F. Haurowitz (Indiana), Dr. Edith Heilbronn (Sweden), Dr. F. L. Hoch, Dr. 0. Hoffman-Ostenhof (Vienna), Dr. B. J. Jandorf (Army Chemical Centre, Md.), Dr. G. M. Kellerman (Australia), Dr. D. Kertesz (Tunis), Dr. T. E. King (Oregon), Prof. J. G. Kirkwood (New Haven, Conn.), Dr. and Mrs. D. E. Koshland, Jr. (New York), Prof. K. J. Laidler (Ottawa), Dr. F. Lindner (Frankfurt), Dr. and Mrs. J. W. Lorimer (Leiden), Prof. and Mrs. J. Murray Luck (Stanford), Prof. R. Lumry (Minnesota), Dr. J. Lundin (Sweden), Dr. A. K. Mills (Dublin), Dr. Manuel Morales (Bethesda, Md.), Prof. and Mrs. H. Neurath (Seattle), Dr. S. 0. Nielsen (Copenhagen), Dr. Z. Nikuni (Japan), Miss Ursula Olpp (Germany), Mr.A. J. J. Ooms (Holland), Dr. R. A. Oosterbaan (Holland), Dr. Sven Paleus (Stockholm), Prof. S. R. Palit (Calcutta), Miss A. M. Paolucci (Rome), Dr. J.-F. Pechere (Brussels), Dr. D. Perlmann (Princeton, N.J.), Dr. Gertrude Perlmann (New York), Dr. D. Peters (Hamburg), Dr. Renzo Rendi (Rome), Dr. and Mrs. D. Rittenberg (New York), Dr. and Mrs. L. Robert (Paris), Prof. R. A. Robinson (Singapore), Ir. M. Sangster (Amsterdam), Prof. R. L. Scott (California), Dr. and Mrs. I. W. Sizer (Cambridge, Mass.), Prof. E. C. Slater (Amsterdam), Dr. L. A. Sluyterman (Eindhoven), Prof. E. L. Smith (Utah), Dr. Bo Sorbo (Stockholm), Prof. and Mrs. M. A. Stahmann (Wisconsin), Mr. B. R. Stein (Gottingen), Dr. G. Stein (Jerusalem), Dr. R. F.Steiner (Bethesda, Md.), Prof. and Mrs. J. M. Sturtevant (New Haven, Conn.), Dr. P. Talalay (Chicago), Prof. and Mrs. H. Theorell (Stockholm), Dr. Fred Vaslow (Copenhagen), Dr. Birgit Vennesland (Chicago), Dr. Mary C . Voorneman (The Hague), Dr. and Mrs. J. N. Walop (Holland), Miss M. G. P. J. Warringa (Holland), Dr. and Mrs. Wetlaufer (Copenhagen), Dr. M. B. Williamson (Chicago), Prof. I. B. Wilson (New York). were the following : 3A GENERAL DISCUSSION ON THE PHYSICAL CHEMISTRY OF ENZYMES A GENERAL DISCUSSION on the Physical Chemistry of Enzymes was held in the University Laboratory of Physiology, Parks Road, Oxford (by kind permission of the Vice-Chancellor) on the 10, 11 and 12th August, 1955. The President, Prof. R. G. W. Norrish, F.R.S., was in the Chair and over 250 members and visitors were present.Among the distinguished members and guests welcomed by the President Dr. S. J. Adelstein (Boston, Mass.), Dr. R. L. Baldwin (Wisconsin), Dr. H. J. Berg (Eindhoven), Dr. and Mrs. F. Bergmann (Jerusalem), Dr. S. A. Bernhard (Bethesda, Md.), Dr. and Mrs. A. Bloch (Brunswick, N.J.), Dr. J. J. Blum (Bethesda, Md.), Prof. C. J. F. Bottcher (Leiden), Dr. D. J. Botts (Bethesda, Md.), Prof. and Mrs. P. D. Boyer (Minnesota), Dr. W. Brackman (Amsterdam), Prof. and Mrs. Britton Chance (Pennsylvania), Dr. J. A. Cohen (Holland), Prof. Harold Edelhoch (Kansas), Dr. and Mrs. Eric Ellenbogen (Pittsburgh), Dr. F. G. M. Elliott (Belgium), Dr. F. L. Feitelson (Jerusalem), Dr. Seymour L. Friess (Bethesda, Md.), Dr. Maria Fuld (Pittsburgh), Dr.Alfred Gierer (Tubingen), Mr. L. Ginsaar (Holland), Prof. and Mrs. S. F. Gomes da Costa (Lisbon),Prof. B. C. Guha (Calcutta), Dr. W. F. Harrington (Copenhagen), Prof. F. Haurowitz (Indiana), Dr. Edith Heilbronn (Sweden), Dr. F. L. Hoch, Dr. 0. Hoffman-Ostenhof (Vienna), Dr. B. J. Jandorf (Army Chemical Centre, Md.), Dr. G. M. Kellerman (Australia), Dr. D. Kertesz (Tunis), Dr. T. E. King (Oregon), Prof. J. G. Kirkwood (New Haven, Conn.), Dr. and Mrs. D. E. Koshland, Jr. (New York), Prof. K. J. Laidler (Ottawa), Dr. F. Lindner (Frankfurt), Dr. and Mrs. J. W. Lorimer (Leiden), Prof. and Mrs. J. Murray Luck (Stanford), Prof. R. Lumry (Minnesota), Dr. J. Lundin (Sweden), Dr. A. K. Mills (Dublin), Dr. Manuel Morales (Bethesda, Md.), Prof. and Mrs. H. Neurath (Seattle), Dr.S. 0. Nielsen (Copenhagen), Dr. Z. Nikuni (Japan), Miss Ursula Olpp (Germany), Mr. A. J. J. Ooms (Holland), Dr. R. A. Oosterbaan (Holland), Dr. Sven Paleus (Stockholm), Prof. S. R. Palit (Calcutta), Miss A. M. Paolucci (Rome), Dr. J.-F. Pechere (Brussels), Dr. D. Perlmann (Princeton, N.J.), Dr. Gertrude Perlmann (New York), Dr. D. Peters (Hamburg), Dr. Renzo Rendi (Rome), Dr. and Mrs. D. Rittenberg (New York), Dr. and Mrs. L. Robert (Paris), Prof. R. A. Robinson (Singapore), Ir. M. Sangster (Amsterdam), Prof. R. L. Scott (California), Dr. and Mrs. I. W. Sizer (Cambridge, Mass.), Prof. E. C. Slater (Amsterdam), Dr. L. A. Sluyterman (Eindhoven), Prof. E. L. Smith (Utah), Dr. Bo Sorbo (Stockholm), Prof. and Mrs. M. A. Stahmann (Wisconsin), Mr. B. R.Stein (Gottingen), Dr. G. Stein (Jerusalem), Dr. R. F. Steiner (Bethesda, Md.), Prof. and Mrs. J. M. Sturtevant (New Haven, Conn.), Dr. P. Talalay (Chicago), Prof. and Mrs. H. Theorell (Stockholm), Dr. Fred Vaslow (Copenhagen), Dr. Birgit Vennesland (Chicago), Dr. Mary C . Voorneman (The Hague), Dr. and Mrs. J. N. Walop (Holland), Miss M. G. P. J. Warringa (Holland), Dr. and Mrs. Wetlaufer (Copenhagen), Dr. M. B. Williamson (Chicago), Prof. I. B. Wilson (New York). were the following : 3CONTENTS PAGE General Introduction. By M. Dixon . . 9 I. CHARACTERIZATION AND PHYSICAL PROPERTIES- Transport Processes and the Heterogeneity of Proteins. By R. L. Baldwin, L. J. Gosting, J. W. Williams and R. A. Alberty . . 13 Certain Physical Properties of Chymotrypsin and Chymotrypsinogen using the Depolarization of Fluorescence Technique.By V. Massey, W. F. Harrington and B. S. Hartley . . 24 Mechanism of Activation of Trypsinogen and Chymotrypsinogen. By H. Neurath and W. J. Dreyer . . 32 Activity of Catalase-Lipid Complexes at Oil/Water Interfaces. By M. J. Fraser, J. G. Kaplan and J. H. Schulman . . 44 Activation and Inactivation of Milk Xanthineoxidase by Physico- chemical Means. By L. Robert and J. Polonovski . . 54 GENERAL DIscussIoN.-Dr. D. B. Wetlaufer, Prof. H. Neurath, Dr. Hartley, Dr. V. Massey, Dr. G. A. Gilbert, Prof. F. Haurowitz, Prof. J. G. Kaplan, Dr. L. Robert, Dr. R. K. Morton, Dr. Tsoo E. King, Dr. V. H. Cheldelin, Prof. D. D. Eley . 65 11. KINETICS AND MECHANISMS- The Influence of Fluctuations in Protein Charge and Charge Con- figuration on the Rates of Enzymatic Reactions. By J.G. Kirkwood 78 Some Kinetic and Mechanistic Aspects of Hydrolytic Enzyme Action. By K. J. Laidler . . 83 Some Chemical and Kinetic Studies of Crystalline Papain. By E. L. Smith, B. J. Finkle and A. Stockell . . 96 The Chemical Kinetics of Some Reactions Catalyzed by Pancreatic Carboxypeptidase. By R. Lumry and E. L. Smith . . 105 The Chemical Structure of the Reactive Group of Esterases. By J. A. Cohen, R. A. Oosterbaan, M. G. P. Warringa and H. S. Jansz . 114 Promotion of Acetylcholinesterase Activity by the Anionic Site. By I. B. Wilson . 119 Fine Structure of the Active Surface of Cholinesterases and the 5 Mechanism of Enzymatic Ester Hydrolysis. By F. Bergmann . 1266 CONTENTS The Mechanism of Reaction Between Esterases and Phosphorus- Containing Anti-Esterases. By B.J. Jandorf, H. 0. Michel, N. K. Schaffer, R. Egan and W. H. Summerson. . PAGE Isotopic Exchange Criteria for Enzyme Mechanisms. By D. E. Koshland, Jr. . The Group-Transfer Activity of Certain Hydrolytic Enzymes. By R. K. Morton Enzyme Action at a Distance. An Examination of its Kinetics and Physical Possibilities. By G. Weber . . Activation and Inhibition of Enzymes. By A. G. Ogston . Steps in the Formation and Decomposition of Some Enzyme-Substrate Complexes. By H. Gutfreund . Hydrogen-Activating Enzymes of Bacteria. By A. Couper, D. D. Eley and A. Hayward . Interaction of Hydrogenase with Hydrogen. By D. Rittenberg and A. I. Krasna . The Action of Fluorocitric Acid on Aconitase. By R.A. Peters . The Kinetics of Haemoglobin and Haem Compounds as Models for . Enzyme Action. By Q. H. Gibson and F. J. W. Roughton Intracellular Reaction Kinetics. By B. Chance . The Haem-Linked Ionizing Group in Myoglobin. By P. George and G. I. H. Hanania . On the Interaction Between Coenzymes and Enzymes. By H. Theorell Calculation of the Rate Constants of the Reaction Between an Enzyme and its Substrate from the Overall Kinetics of the Reaction Catalyzed by the Enzyme. By E. C. Slater , Some Applications of Deuterium to the Study of Enzyme Mechanisms. By B. Vennesland . The Hydrogen Ion in the Enzymatic Oxidation and Reduction of DPN. By I. W. Sizer and A. Gierer . GENERAL D~scussIoN.-Prof. K. J. Laidler, Prof. J. M. Sturtevant, Dr. D. J. R. Laurence, Dr. H. Gutfreund, Prof. H. Neurath, Prof. R. Lumry, Dr. B. R. Rabin, Prof. E. L. Smith, Dr. R. J. P. Williams and Dr. B. L. Vallee, Dr. L. Robert, Dr. J. J. Blum, Dr. S . A. Bernhard, Dr. F. Bergmann, Dr. P. D. Boyer, Dr. D. R. Davies and Dr. A. L. Green, Dr. R. K. Morton, Dr. D. E. Koshland, Jr., Mr. P. A. T. Swoboda, Dr. 0. Hoffmann-Ostenhof, Dr. D. D. Eley, Dr. D. Bunn, Prof. J. A. V. Butler, Dr. M. Sangster, Dr. A. Klug, and Prof. F. J. W. Roughton, Mr. J. G. Beetlestone and Dr. A. Couper, Prof. H. N. Rydon, Dr. H. Kacser, Prof. F. Haurowitz, 134 142 149 156 161 167 174 185 189 195 205 216 224 23 1 240 248CONTENTS 7 PAGE Dr. K. Dalziel, Dr. P. George, Prof. Q. H. Gibson, Mr. J. S. Griffith, Dr. G. I. H. Hanania, Dr. S . J. Adelstein and Dr. J. A. Olson, Dr. F. L. Hoch, Mr. W. E. Trevelyan, Dr. M. Dixon, Dr. L. L. Ingraham, Dr. E. C. Slater, Dr. B. Vennesland, Dr. P. Talalay, Dr. I. W. Sizer and Dr. A. Gierer . . 254 Author Index . . 317

ISSN:0366-9033    

DOI:10.1039/DF9552000001

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

THE PHYSICAL CHEMISTRY OF ENZYMES GENERAL INTRODUCTION BY M. DIXON Enzymology is a subject which has a special interest because it lies just on the borderline where the biological and the physical sciences meet. On the one hand, it is the enzymes which bring about nearly all biological processes; on the other hand, many of these catalysts are now available in the pure state for exact study by physicochemical methods in vitro. All professional enzymologists will be glad that the Faraday Society has arranged this Discussion, at which people from both sides can meet together to consider questions of common interest. Since the subject is a rapidly developing one, which has made great progress in recent years, it seemed to me that the most useful thing I could do by way of introduction would be to take a quick look at the main ways in which the subject has been developing and at some of the changes which have taken place in our views on enzymes in consequence.Many of the modern developments could not have taken place without the availability of pure enzymes. It is less than 30 years since the first isolation of an enzyme; even 20 years ago the number of purified enzymes was very small. We now know over 500 different well-defined enzymes, most of which have been purified to a greater or less degree, and nearly 100 of which have been obtained pure and crystalline. This I think is something of an achievement when we consider their great fragility, and the fact that each has to be isolated from a complex mixture of very similar substances in which it is only present in minute amounts.There are many other enzymes which have not yet been isolated, and I should be inclined to put the total number as well over 1OOO. All those which have been obtained pure have turned out to be proteins; and the intactness of the protein structure is essential for the catalytic activity, which is lost on denaturation of the protein. Thus physicochemical studies on enzymes may relate either to their protein properties (molecular weight and so on) or to their catalytic properties. We are here mainly concerned with the latter ; the former properties really come under the heading of protein chemistry (although indeed one might almost say that protein chemistry is a branch of enzyme chem- istry, since the great majority of the known proteins are enzymes !).The study of purified enzymes has brought out clearly what is perhaps their most important characteristic, namely, their high specificity. The great majority of enzymes catalyze effectively only their own particular chemical reaction. Even a very small change in the chemical structure of the substrate molecule usually reduces the velocity considerably, and anything more than a small change will reduce it to zero. It is quite true that there are some rather unspecific enzymes, especially among the hydrolases, but they are few in number. The multiplicity of specific enzymes, however, does not mean a multiplicity of different types of chemical reactions. The number of different types of re- actions catalyzed by enzymes is relatively small; but the same kind of reaction may be brought about in a number of different molecules and a different enzyme 1 is required for each.For example, over 50 different enzymes oxidize the CHOH I I group to the CO group in different molecules, 33 different enzymes are known 9 I10 GENERAL INTRODUCTION which transfer the terminal phosphate group of adenosinetriphosphate to other molecules, over 50 different enzymes hydrolyze -CO-NH- links in different chemical structures, and so on. The multiplicity of enzymes is therefore due to the fact that their specificity is not limited to that part of the substrate molecule which is involved in the reaction, but extends to the other parts of the molecule as well. Now that we have an idea of the types of reactions which are catalyzed by enzymes, it has become clear that nearly all enzyme reactions are transfer reactions, in which some group is transferred from one molecule to another, without at any time being liberated in the free state.We may write the typical reaction as AB + C = A + BC, where B is the group transferred, which may be a phos- phate group, an acyl group, ammonia, hydrogen, a glycosyl group, or a number of others. Even the hydrolyzing enzymes may be regarded as transferring one part of the substrate from the other part to a hydroxyl from the water (C = OH). In fact an increasing number have been shown to transfer the part B to other molecules besides water, so that the distinction between hydrolyzing and trans- ferring enzymes has largely broken down.Several papers in our discussion refer to this. The non-transferring enzymes are represented by about 25 enzymes which add groups of various kinds to double bonds, e.g. fumarase. The specificity data show that there is generally a close fit between enzyme and substrate which may extend over a considerable part of the substrate mole- cule. This means that the substrate is probably combined with the enzyme not by one link only but by many. The active centre of the enzyme, i.e. that part which combines with the substrate and at which the catalysis takes place, is there- fore probably a structure of some complexity, containing a number of different chemical groups arranged in such a pattern as to fit the substrate and all combining with different parts of the substrate molecule, some possibly by electrostatic attraction, others by hydrogen bonding.A very rough estimate of the area over which this fit may extend in several cases suggests that over a patch of up to 15-20 A diameter it may be fairly close, but that the specificity falls off rapidly beyond this. Of course, many substrates are smaller than this and then the specificity may cover the whole molecule. As would be expected if there is a close fit, enzymes usually act only on one optical isomer of an asymmetric substrate, and this is very important. But enzymes go further than this, as was first pointed out by Dr. Ogston, for even in substrate molecules of the type Ca2bc, possessing a plane of symmetry, they can distinguish between the two identical u groups. For instance, glycerol kinase always phosphorylates the same end of the symmetrical glycerol molecule, and as Dr.Vennesland's paper shows, the two hydrogen atoms of a CH2 group are not at all enzymically equivalent. There has been a considerable change of views about the number of active centres in an enzyme molecule. Enzyme catalysis used to be regarded, by analogy with inorganic catalysts, as occurring at a surface dotted all over with a large number of active centres. The availability of pure enzymes and of several methods for estimating active centres has shown on the contrary that the number is very small. Many enzymes, perhaps the majority, have one active centre per enzyme protein molecule; some have two; a few of high molecular weight have four, but there is some evidence that they may be associations of smaller units. It is important to bear this fact in mind when formulating theories of enzyme catalysis in terms of protein structure; apart from this one active spot in the molecule, the protein structure is cat a1 y tically inactive.In a large number of enzymes, however, this active spot seems to be part of the protein structure and not to contain any non-protein chemical groups. Its special properties seem to be due rather to the arrangement of the protein side- chain groups in a pattern fitting the substrate. My own idea of the active centre is somewhat like the sketch, which represents 4 peptide chains or sections of aM. DIXON 11 folded chain lying side by side in the protein, with the substrate held by groups situated in different chains combining with different parts of its molecule.The activity is therefore dependent on what Eyrhg has called the tertiary protein struc- ture. This will explain (i) the loss of activity when enzymes are denatured, because in the unfolding the chains separate and the active centre no longer exists as such, though it can be re- formed when the denaturation is reversed; (ii) the stabilizing action of the substrate against denaturation, since it would tend to hold the chains together; (iii) the retention of some power to combine with the substrate after denaturation, for the separate groups still exist and the substrate might combine with any one separately although no reaction would follow. Many lines of evidence are now beginning to give indications of the chemical nature of some of the groups in the active centres ; at the moment emphasis is on sulphydryl and imidazol groups, which seem to be concerned in the activity of many enzymes.As several of the following papers show, considerable progress is being made in this field. Some enzymes, particularly oxidizing enzymes, contain in addition to an active centre of this kind a 3 f 0 L ‘ S special non-protein prosthetic group such as flavin or haem, which, although attached to the enzyme, forms essentially one of the reactants. Some enzymes indeed may contain more than one such group, for instance, yeast lactate dehydro- genase, a crystalline enzyme whose purification was begun by us and finished recently by Morton. This contains both a flavin and a haem group and may be represented as follows : S + F H cyt.-c I I I I 1 .Here the substrate (lactate) reduces the flavh, which in turn reduces the haem, which reduces cytochrome-c in solution, giving three successive reactions with one crystalline enzyme. In these cases the substrate combines with the active centre in the protein, which is distinct from the prosthetic group, but there is a small group of enzymes catalyzing relatively simple reactions in which the active group with which the substrate combines is in the prosthetic group itself. For example, in catalase, which decomposes H202, the peroxide unites with the iron atom of the haem group. This, however, is a rather special case. In enzyme kinetics considerable advances have been made, especially in analyzing the overall process in terms of the rate constants of the separate steps. In certain cases it is possible to measure the different rate constants separately, either by direct observations on the enzyme-substrate complex using the spectro- scopic methods of Prof.Chance or the fluorescence method of Prof. Theorell, or by the pre-steady-state kinetics of Dr. Gutfreund, or by the extrapolation methods of Prof. Slater. It is now clear that in different cases the rate-determining step may be either the combination of enzyme and substrate, or the transformation of one form of the complex into another, or the dissociation of the product from the enzyme. The Michaelis kinetics have been replaced by the Briggs-Haldane kinetics, which, however, were practically limited to unimolecular reactions and have in their turn been replaced by a number of more complex expressions.12 GENERAL INTRODUCTION The use of pure enzymes has permitted the determination of velocities in ab- solute units and the application of the absolute rate theory has given the free energies and entropies of activation of some of the reactions, but the available data are still all too scanty and there is no single enzyme for which complete kinetic and energetic data can be given.There has been quite a spate of papers in the last two years on the effect of pH on enzyme kinetics. I cannot myself escape some share of the responsibility for this, having suggested methods for the interpretation of the effects in terms of the dissociation constants of groups in the active centre, the substrate and the complex.There is some hope that the pK values may give clues to the nature of the groups concerned in the union of enzyme and substrate, but due caution is necessary on this point. Much progress has been made recently in elucidating the mechanism of enzyme reactions. One result is that, whereas we used to regard the reaction as a process occurring “ in presence of the enzyme ”, we now tend increasingly to write the enzyme into the reaction as one of the reactants. For example, in the transfer reaction already mentioned, if we add no C, but instead add an equivalent amount of the enzyme (which the availability of pure enzymes allows us to do) in many cases the following reaction occurs : AB + E = A + BE. In fact in many cases, though not in all, the group transferred (B) becomes attached to the enzyme by a firm link quite unlike an ordinary enzyme-substrate combination, though it can be removed by transfer from the enzyme to another molecule (C), so com- pleting the reaction.In some cases it is known to be attached to an -SH group in the enzyme in the form of a thioester. Dr. Koshland has developed some interesting criteria or tests for this mechanism, depending partly on optical in- version and partly on isotopic exchanges, some of which are discussed in his paper. The application of isotope techniques has been very fruitful in throwing light on enzyme mechanisms. They have been used, for example, to show which part of the substrate, if any, becomes attached to the enzyme, even when only catalytic amounts of enzyme are available.In the above case, if one adds AB and A* one should obtain the reaction AB + A* = A*B + A, an exchange reaction of the kind called by Rittenberg a “ virtual reaction ” since no chemical reaction occurs. With B*, on the other hand, no exchange occurs; it is the group which is not attached to the enzyme which exchanges. This method has revealed the very interesting triple-transfer mechanism of the synthetases, by means of which the linking together of two molecules is coupled with the breakdown of a third. Isotopes have also been used to identify the actual bond broken in the enzymic hydrolysis of molecules of the type R-0-R’, which proves to be the bond nearest the part for which the enzyme is most specific. They have also been used in the very elegant work of Dr. Vennesland to show the nature of the transfer reaction ; in this case by the use of deuterium in studying hydrogen-transfer reactions, which will be discussed later. These are some of the ways in which the subject has developed. Our discussion includes papers on many different aspects, most of which fall under the headings of structure, kinetics or mechanism. Kinetics and structure, I believe, should be studied not separately, but in relation to one another ; they are both means to an end, which is the elucidation of the mechanism of enzyme catalysis, the most interesting problem in the field.

ISSN:0366-9033    

DOI:10.1039/DF9552000009

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

I. CHARACTERIZATION AND PHYSICAL PROPERTIES TRANSPORT PROCESSES AND THE HETEROGENEITY OF PROTEINS BY ROBERT L. BALDWIN, LOUIS J. GOSTING, J. W. WILLIAMS AND ROBERT A. ALBERTY* Depts. of Biochemistry, Dairy and Food Industries, and Chemistry, University of Wisconsin, Madison Received 4th July, 1955 Many methods are now available to provide certain molecular characteristic data for protein molecules. Some of them have a sound foundation in thermo- dynamics and provide solute molecular weights and activity coefficients. Others involve transport phenomena and serve, among other things, in the description of solute heterogeneity. It is with three of these transport processes that we shall be concerned : diffusion, electrophoresis, and sedimentation. We shall endeavour to show how the heterogeneity of enzymes and other proteins can be quantitatively described by an analysis of the shape of the boundary gradient curves and its variation with time.Ordinary qualitative inspection of these gradient curves in electrophoresis and in velocity sedimentation can be quite misleading. Although it is a necessary criterion of homogeneity that only a single moving boundary be observed, this is not a sufficient criterion even when the gradient curve is symmetrical. For example, some of the gradient curves obtained in the sedimentation velocity experiments with dextran and gelatin samples, materials which possess a broad distribution of molecular weights, are not unlike the corresponding curves which have been cited as evidence for the homogeneity of a protein. Furthermore, because of the spreading effects of diffusion, the boundary gradient curves formed by two components may remain unresolved throughout an experiment, showing but one continuous curve with a single maximum.Both theoretical and experimental problems are encountered in making quantitative tests of heterogeneity by the methods we are to consider. The place of diffusion is unique because, in addition to being important in itself, it is involved in the analysis of electrophoresis and sedimentation velocity experiments. One description of these two popular approaches of the biophysical chemist might be that they, too, are diffusion experiments, but with externally applied fields, one electrical and the other mechanical. In the three separate sections which follow, limitations of space have made it necessary for us to restrict ourselves to a discussion and description of procedures in use in this laboratory.1 . DIFFUSION An understanding of the process of diffusion is important in molecular-kinetic studies of the heterogeneity of protein preparations. Boundary spreading measurements during electrophoresis or ultracentrifugation are complicated by * This research was supported in part by the National Institutes of Health and by the Research Committee of the Graduate School from funds supplied by the Wisconsin Alumni Research Foundation. 1314 HETEROGENEITY OF PROTEINS diffusion, which contributes to the spreading of the refractive index gradient curves; an observer must know the differences in laws governing spreading by diffusion and spreading by the applied field to sort out properly these different effects.On the other hand, it is possible from measurements of free diffusion alone to obtain some information about protein heterogeneity. The basic methods 1-5 for carrying out such analyses were developed before the techniques for analyzing boundary spreading in an electrical or centrifugal field became available, but their application has been handicapped because of insufficient experimental accuracy.5 Much more accuracy is required to analyze a sum of Gaussian curves with all their centres at the same position, as in diffusion, than to obtain information about the displacement of their centres by an applied electric or centrifugal field. This inherent insensitivity of diffusion measurements to sample heterogeneity has been partially overcome by the application during the last decade of interferometric optical systems to diffusion studies.6-13 These systems have increased the accuracy of diffusion measurements by at least ten-fold allowing, in favourable cases (such as a sample with only one impurity), a measure- ment of inhomogeneity with respect to diffusion coefficient which approaches the sensitivity of corresponding measurements with either electrophoresis apparatus or the ultracentrifuge.Methods of analyzing Gouy fringe data for non-homogeneous solutes have been described by Ogston,l4~ 15 Charlwood 16 and Akeley and Gosting.17 In this discussion we use the notation and approach of Akeley and Gosting and we assume that Fick's first law for one-dimensional diffusion,l* (1 * 1) may be applied to each diffusing solute separately.* Here 4 denotes the number of grams of component i crossing a square cm per sec as a result of the con- centration gradient ( S C J S X ) ~ at height x and time t.Providing that each diffusion coefficient, Di, and each specific refractive increment is independent of the q solute concentrations, an expression for the refractive index gradient for free dif- fusion can be derived in the form Ji = - Di (8 C i P X h , The positive direction of x is here taken as downward, with its origin at the position of the sharp initial boundary formed at time t = 0 between solution and solvent. The refractive index difference between the lower and upper starting phases is An, while cci denotes the fraction of this difference contributed by component i ; therefore, Ri denotes a fraction of the total solute on the basis of refractive index.If all the specific refractive increments are equal, a1 reduces to the weight fraction of component i. It has been found convenient to analyze fringe photographs from a Gouy diffusiometer experiment so as to obtain two kinds of data : (i) the " height-area " (average) diffusion coefficient and (ii) a fringe deviation graph which is independent of time and which contains all of the information about deviations of the refractive index gradient curve from Gaussian form. To obtain these data, the total number, jm, of fringes is first measured with the aid of a Rayleigh double slit which brackets the initially sharp boundary, This measurement gives the fractional part of j,, the integral part being determined directly from the Gouy photographs which are * Thus it is assumed that each concentration difference within the cell is sufficiently small that the partial specific volume of each component can be considered constant; otherwise, a term in bulk flow must be added to eqn.(1.1) (ref. (19)). Furthermore, it is assumed that the solute flows do not interact, i.e. that the flow of a given solute depends only on the concentration gradient of that solute and not on other solute concentration gradients. A method for treating systems in which the solute flows interact is described in ref. (20) and (21).R . L. BALDWIN, L . J . GOSTING, J. w.WILLIAMS, R. A . ALBERTY 15 taken during the diffusion process, after removing the Rayleigh double slit. For each photograph several yI, the downward displacements from the undeviated slit image of representative fringes (intensity minima numbered j = 0, 1, 2, etc.), are measured throughout the interference pattern starting with the lowest minimum, j = 0. From j , and each fringe number, j , a reduced fringe number, is calculated. The approximations, 21, to the series ( j + 314 + . . .) have been tabulated 22 as has the reduced fringe displacement, exp (- &2), as a function of f ( 6 ) . 8 Hence the quotient yj/exp (- &2) is readily computed for the lower fringes and extrapolated 17 to 2’4 = 0 to obtain Cry which is the deflection cor- responding to (Sn/Gx),, for that photograph.Then, knowing the wavelength h of the monochromatic light and the optical lever arm b of the apparatus, the “ height-area ” diffusion coefficient is easily obtained : Quensel3 has shown that this value is an average of the form Because the experimental starting boundary is not perfectly sharp, the values of DA computed from several photographs are extrapolated 7 to (llt) = 0. 0.5 0- 0.5 f ( F ) f(r) FIG. 1 .-Illustrative fringe deviations calculated for four values of Dl/D2 : For thejth fringe the reduced fringe deviation is defined by , a2 = 0.01 ; - - - -, cc2 = 0.005. Qj = [exp (- &2) - yj/Ct]. (1.6) A plot of Qj against f(&) constitutes the desired fringe deviation graph; for a Gaussian curve, .iz/ = 0 for all j . The subscripts j will now be omitted for sim- plicity.It has been shown 17 that when solute 1 is the major solute, so that q, . . ., aq are all small, the fringe deviations are given to a good approximation by the series16 HETEROGENEITY OF PROTEINS in which terms of order (apk), etc., and higher terms are omitted here. The functions F(5, Dl/Di) have been tabulated as functions of f(5) for several values of the diffusion coefficient ratios, D1/Dj, and they provide useful information for interpreting heterogeneity. For simplicity we will restrict such interpretations to cases in which only two solutes are present, solute 2 being present in small amount as an impurity in solute 1. The shape and magnitude of representative fringe deviation graphs for 1 % impurity in the sample are shown as solid lines in fig.1. It will be noted that the maxima of these graphs lie on the left of f(5) = 0.43 when the impurity diffuses slower than the main solute, and on the right when the impurity diffuses I NUB No I 2 5 I * 0 X C 0 X C * 0 X C w 0 X C X " * 1 X ; "oL 05 I EPA NO 2 8 4 - 8 t ~ B P I No R370295A I 0 0.5 t EPA No 212113 I f (CI FIG. 2.-Gouy fringe deviation graphs for three different samples of bovine plasma albumin (BPA) and one sample of normal human albumin (NHA). faster than the main solute. For this small amount of impurity the fringe deviations are proportional to a2. This is illustrated on the graph for D1/D2 = 4 where the dashed line corresponds to 0.5 % impurity. Because fringe deviations of i2 = 2 x 10-4 are de- tectable experimentally, as little as 0-1 % of impurity could be detected in this case.When a2 is small the position of the maximum de- pends on 0 1 / 0 2 , while the maximum height depends on this ratio and is proportional to a2. It should be mentioned that from ex- perimental fringe deviation graphs and the value of DA it is possible 17 to calculate the two higher average diffusion coefficients,2 i = 1 a i = 1 with reasonable accuracy. To illustrate the use of the Gouy diffusio- meter in studying protein inhomogeneity, measurements were made on three different samples of crystallized bovine plasma albu- min* @PA) and one sample of normal human albumin t (NHA). Each experiment was performed within f 0.004" of 25" C in a buffer which was 0.0100, 0.0100 and 0-1500 molar in acetic acid, potassium acetate and potassium chloride, respectively ; 17 at 25" this buffer had a density of 1.00470 g/ml, a vis- cosity relative to water of 1.0037 and a pH of 4.60 & 0.02.About 0.85 g of the dry protein was dissolved in 75 ml of the buffer, dialyzed at least twice against 400 mi samples of the buffer for 10 h or more at lo, and then allowed to stand for at least 9 h at 25" for attaining final dialysis equilib- rium against the last buffer sample. Other details of the experimental procedure are described elsewhere.17 Fringe deviation graphs for the experiments are shown in fig. 2, in which dots at a given value of f(5) represent experimental values of 52 at different times while crosses represent averages of these values for each * supplied by Armour & Co., Chicago, Illinois.t supplied as " Normal Human Albumin " by the Michigan Dept. of Health Labora- tories, Rework no. 125 of Fraction V (electrophoresis analysis reported as Alb. 99 %, a-Glob. 1 x).R . L . BALDWIN, L . J . GOSTING, J . w. WILLIAMS, R . A . ALBERTY 17 fringe. Since the maxima lie to the left of f(5) = 0.43, at least one slower diffusing impurity is present in each sample ; the sample of NHA is seen to possess more heterogeneity than the BPA samples. Assuming that the same impurity is present in the several samples, the amount of impurity is seen to be different in each sample as indicated by the different maximum heights of the fringe deviation graphs. The absence of any appreciable fringe deviations near f(5) = 1 proves that no appreciable buffer gradients were present to contribute significantly to the values of 0 ~ .Measured values of DA for these four experiments, corrected to water at 25" by multiplying by the relative viscosity of the buffer, are reported in column 4 of table 1. Since there is a small dependence of the diffusion coefficients on TABLE l.-Gou~ DIFFUSIOMETER DATA FOR FOUR PROTEIN SAMPLES Corrected to water at 25" C 1 2 3 4 5 6 7 8 sample NHAno. 125 90.52~ 0.519 6.460 6.552 6.603 0.068 6.75 BPA no. 284-8 87.226 0.498 6.696 6.729 6.742 0.025 6.80 BPA no. R370,295A 91.05~ 0.515 6.670 6-732 6.791 0,039 6-84 BPA no. 212,113 89.40~ 0.507 6.661 6.706 6.745 0.031 6-79 (a) cell dimension, a = 2.5062 cm, (b) 9, ,, a = 2.4862 cm, (c) ,, ), a = 2.5103 cm. Values of C were computed from the equation I = Ajn,/[2a(An/Ac)] in which X = 5460.7 x 10-8 cm and (Anlac) is the specific refractive increment of the protein.concentration,23 these values correspond to the mean protein concentration c of each experiment. Values for 7, column 3, were calculated from measured values ofj,, column 2, taking 1.922 x 10-3 (g/100 ml)-1 and 1.900 x 10-3 as the specific refractive increments * of BPA and HSA, respectively.24 Values of 3 and (D2)t in columns 5 and 6 were computed 17 from DA and the fringe deviation graphs. Estimates of the diffusion coefficient of" pure " albumin were obtained, column 8, by assuming that the fringe deviations were due to a single impurity. Taking 01/02 = 1-75, all four fringe deviation graphs were fitted 17 within experimental error by selecting the values of a2 in column 7.Using eqn. (1.5)) the numbers in column 8 for the diffusion coefficient of "pure" albumin were then computed from the values of DA, and cc2. It is seen that the diffusion coefficient of NHA is thus brought into reasonable agreement with the values for BPA. Because only two solutes are assumed to be present these values are not necessarily the correct values for the diffusion coefficient of albumin. They should, however, be nearer to the correct value than the values of DA now in common use. It is hoped that more reliable values for the diffusion coefficient of albumin will eventu- ally be obtained by applying this procedure in connection with independent measurements of the amount of impurity in the sample by means of sedimentation and electrophoresis experiments.* I t is assumed that the data of Perlmann and Longsworth for these proteins in solutions containing sodium chloride also apply to this buffer. 1. Difusion 1 Lamm, Nova Acta Regiae SOC. Sci. Upsaliensis, 1937, series IV, 10, no. 6. 2 Gralen Kolloid-Z., 1941, 95, 188 ; Diss. (Uppsala, 1944). 3 Quensel, Dim. (Uppsala, 1942). 4Neurath, Chem. Rev., 1942,30, 357. 5 Bevilacqua, Bevilacqua, Bender and Williams, Ann. N. Y. Acad. Sci., 1945, 46, 309. 6 Calvet, Conzpt. rend., 1945, 220, 597 ; 1945, 221, 403 ; Rev. opt., 1950, 29, 35.18 HETEROGENEITY OF PROTEINS 7 Longsworth, J. Amer. Chem. Soc., 1947, 69,2510, 8 Kegeles and Gosting, J. Amer. Chem. SOC., 1947, 69, 2516. 9 Coulson, Cox, Ogston and Philpot, Proc.Roy. SOC. A, 1948, 192,382. 10 Philpot and Cook, Research, 1948,1,234. 11 Svenson, Acta Chem. Scand., 1949,3, 1170 ; 1950,4, 399 ; 1951,5, 72. 12 Longsworth, J. Amer. Chem. SOC., 1952,74, 4155. 13 Scheibling, J. chim. phys., 1950, 47, 689 ; 1951, 48, 559. 14 Ogston, Proc. Roy, SOC. A, 1949, 196,272. 15 Ogston, Biochem. J., 1949, 45, 189. 16 Charlwood, J. Physic. Chem., 1953,57, 125. 17 Akeley and Gosting, J. Amer. Chem. Soc., 1953, 75, 5685. 18 Fick, Pogg. Ann., 1855, 94, 59. 19 Onsager, A m . N. Y. Acad. Sci., 1945, 46, 241. 20 Baldwin, Dunlop and Gosting, J. Amer. Chem. SUC., 1955, 77, in press. 21 Dunlop and Gosting, J. Amer. Chem. SOC., 1955, 77, in press. 22 Gosting and Morris, J . Amer. Chem. Suc., 1949, 71, 1998. 23 Creeth, Biuchem. J., 1952, 51, 10. 24 Perlmann and Longsworth, J.Amer. Chem. SOC., 1948, 70,2719. 2. ELECTROPHORESIS The moving boundary method of electrophoresis offers a means for testing one aspect of the homogeneity of a protein preparation, that is, the homogeneity with respect to mobility. The mobility of a protein depends not only upon the number of ionized acidic and basic groups but also upon the size and shape of the protein molecule and the interaction with buffer ions other than hydrogen and hydroxyl ions. The sensitivity of the mobility to structural differences is indicated by the fact 1 that for serum albumin a difference of one electronic charge per molecule corresponds to a difference in mobility of 0.2 x 10-5 cm2 volt-1 sec-1. Moving boundary electrophoresis experiments offer two types of tests for the homogeneity of a protein preparation.The first test consists in carrying out electrophoretic experiments over a wide range of pH and ionic strength to deter- mine whether more than one protein boundary is obtained. The problems encountered in obtaining a quantitative analysis of a protein mixture by the moving boundary method have been discussed elsewhere.% 3 The fact that a protein preparation yields a single moving boundary, even over a range of pH and ionic strength, does not prove that all the molecules have identical electrophoretic mobilities4 A family of closely related proteins may yield a sbgle boundary but this boundary will spread at a faster rate than if a single protein were present. Thus the second test for electrophoretic heterogeneity consists in studying the spreading of the protein boundary during electrophoresis under conditions such that convection and pH and conductivity gradients are negligible.Convection caused by the temperature gradient set up in the electro- phoretic cell or caused by electro-osmosis along the cell wall may also spread the protein boundary, but such convection effects would not be reversed by reversing the field, and so it is possible to test the electrophoretic experiments for their presence. Spreading and sharpening caused by the gradients in conductivity and pH may be minimized by performing the experiment at the isoelectric point of the protein and at a low protein concentration. If the various molecules in the protein mixture have different mobilities the protein gradient curve will spread faster in the electric field than in the absence of the field, and upon reversal of the field the boundary will spread less rapidly than expected for diffusion and may even become sharper.A protein preparation containing molecules with a continuous range of mobilities may be characterized by a distribution function such that the value of the function, g(ui), is the relative frequency on a refractive index basis of molecules with mobility ui. To the extent to which it is possible to study the spreadingR . L . BALDWIN, L . J . GOSTING, J. w. WILLIAMS, R . A . ALBERTY 19 of a protein boundary under conditions such that the electric field strength E and mobility ui are constant, it is possible to evaluate g(u). The integral equation for the refractive index gradient of a boundary which has been spread both by diffusion and electrophoretic migration was given by Sharp, Hebb, Taylor and Beard.5 The refractive index gradient, 6n/8x, as a function of height x in the electrophoretic cell at time after the formation of the sharp initial boundary and time tE after application of the electric field is given by The diffusion coefficient D is assumed to be the same for all protein molecules in the solution and independent of concentration.It is apparent that eqn. (2.1) is not unrelated to eqn. (1.2) of the preceding section. Eqn. (2.1) gives only the contribution of the protein to the refractive index gradient. It is assumed that in the actual experiments this approximation does not introduce an appreciable error.If the mobility distribution function g(u) has some simple form, eqn. (2.1) may be integrated in closed form. For example, if the mobility distribution may be represented by the Gaussian error function, (2.2) 1 h d2n g(u) = -exp (- u2/2h2), Alberty 6 has shown that the heterogeneity constant h (the standard deviation of the mobility distribution) may be calculated from the variation with time of a2, the second moment about the mean of the refractive index gradient curve. The necessary equation is where D* is the apparent diffusion coefficient and go2 is the value of 02 at the time the electric field is applied ( t ~ = 0). Actually eqn. (2.3) may be derived in more general ways which show that h is the standard deviation of the mobility distribution whether or not this distribution is of the Gaussian form,7~ 8 and also that D is equal to 5, the mean diffusion coefficient, when the sample contains molecules of different diffusion coefficients.Since the plot of D* against t~ should extrapolate to the value of the ordinary diffusion coefficient D at t~ = 0 and also decrease to this same value after the current has been reversed for an equal period, these tests may be used to determine whether irreversible effects other than free diffusion have occurred. A completely general method for the determination of the mobility distribution for a heterogeneous protein has been worked out by Baldwin, Laughton, and Alberty.8 It was used in an investigation of the heterogeneity of chemically reacted albumins.The method depends upon an extrapolation of the apparent mobility distribution, g*(u), defined by g*(zl) = (sn/sx),Ef/an, (2.4) to give the actual distribution of mobilities, g(u). This is accomplished by plotting values of g*(ui), the apparent distribution value for the mobility ui, against l/t and extrapolating to (l/t) = 0 to obtain g(ui). Such a plot is given in fig. 3 for a methyl ester of bovine serum albumin. The value of the heterogeneity constant obtained from these data, 0.54 x 10-5cmzV-1sec-1, is the same as that cal- culated from the plot of D* against t. The advantage of the extrapolation method is that it gives more information as to the shape of the distribution. Gosting9 has shown that for long periods of time the extrapolation should be linear and has computed the expected slope.20 HETEROGENEITY OF PROTEINS The heterogeneity constants of a number of purified proteins have been determined [table xx, ref.(3)]. All preparations which have been studied have shown some reversible spreading. The experiments with some gamma globulins are of particular interest. The heterogeneity constants of human y1- and 72- globulins from plasma pools are 0.26 x 10-5 and 0.52 x 10-5cm2V-1sec-1 at 0.1 ionic strength.109 11 Cam, Brown and Kirkwood 12 have fractionated bovine y-globulin by the electrophoresis convection method, and although the fractions had about the same heterogeneity constant as the starting material (Armour fraction 11, h = 0.65 x 10-4 V-1 sec-I), they varied considerably in mean iso- electric point.The sum of the mobility distributions for the three fractions gave the mobility distribution of the starting material. Recently, reversible spreading experi- ments have been made with the y-globins which appear in large quantities in the FIG. 3(a).- Measurement of g(u) FIG. 3(b).- The mobility distri- for a methyl ester of bovine plasma bution ( t = CO) and g*(u) at time of g*(ui) at representative positive values of ui. albumin by extrapolation to infinite t = 7.3 x 103. plasma of patients having certain pathologies. Deutsch and Kratochvil13 found that of five such globulins studied, three had heterogeneity constants at 0.1 ionic strength of less than 0.1 x 10-5 cm2 V-1 sec-1 and the others had values of 0.13 x 10-5 and 0.16 x 10-5. Putnam and Udin 14 studied a cryoglobulin with a heterogeneity constant of 0.1 x 10-5 cm2 V-1 sec-1.A Bence-Jones protein studied by Deutsch 15 had a value of h less than 0.1 x 10-5 cm2 V-1 sec-1. The principal limitation in accurary of reversible boundary spreading experi- ments results from deviations from the assumed constancy of electric field strength and mobility through the boundary region. If the protein is not electrophoretically homogeneous and exactly at its isoelectric point, there will be changes in buffer concentration and pH in the boundary region. One means for investigating the magnitude of these disturbances would be to study reversible spreading over a range of protein concentration. For a given rate of electrical heating, higher field strengths may be used-and thus higher resolving power obtained-when buffers of low ionic strength are employed; however, the effects of the electro- phoretic " anomalies " are also increased, for a given protein concentration.The accuracy of reversible boundary spreading experiments would be greatly increased by use of interferometric optical systems. 2. Electrophoresis 1 Longsworth and Jacobsen, J . Physic. Chem., 1949, 53, 126. 2 Longsworth, in Corcoran, Methods in Medical Research (Year Book Publishers, 3 Alberty, in Neurath and Bailey, The Proteins, vol. 1, part A (Academic Press, N.Y., 4 Tiselius and Horsfall, Arkiv Kemi, Miizer., Ceol. A, 1939, 13, no. 18. Inc., Chicago, Ill., 1952). 1953).R . L . BALDWIN, L. J . GOSTING, J . W. WILLIAMS, R . A . ALBERTY 21 5 Sharp, Hebb, Taylor and Beard, J.Biol. Chem., 1942, 142,217. 6 Alberty, J. Amer. Chem. SOC., 1948, 70, 1675. 7 Brown and Cann, J. Physic. Chem., 1950,54,364. 8 Baldwin, Laughton and Alberty, J. Physic. Chem., 1951, 55, 11 1. 9 Gosting, J. Amer. Chem. SOC., 1952, 74, 1548. 10 Anderson and Alberty, J. Physic. Chem., 1948, 52, 1345. 11 Alberty, Anderson and Williams, J. Physic. Chem., 1948, 52, 217. 12 Cann, Brown and Kirkwood, J . Amer. Chem. SOC., 1949,71,2687. 13 Deutsch and Kratochvil, personal communication. 14 Putnam and Udin, J. Biol. Chem., 1953, 202,727. 15 Deutsch, J . Biol. Chem., 1955, 216, 103. 3 . SEDIMENTATION The analysis of heterogeneity in sedimentation velocity experiments is similar to the analysis for electrophoresis. In both cases, the applied field produces resolution of different components and diffusion tends to obscure this resolution. In comparison with the equations for electrophoresis, those for sedimentation are complicated by the necessity for using a sector-shaped cell, in which the cross- sectional area is proportional to the distance from the centre of rotation, x, and by the proportionality to x of the centrifugal field strength, w2x; w is the angular speed of revolution.Furthermore, effects of the dependence of sedimentation coefficient s on concentration generally are significant and equations derived on the assumption that s is constant rarely are adequate. If a sample contains material with a continuous distribution of sedimentation coefficient, or if it contains a finite number of components whose boundary gradient curves are not resolved from one another by the end of the experiment, then analysis of the heterogeneity must take into account at least three factors: the heterogeneity itself, the diffusion, and the dependence of the sedimentation coefficients on concentration.When the boundary gradient curves of separate components are resolved from one another, diffusion need not be considered; the analysis for this case 1 9 2 will not be discussed here. Also, since we will not consider the dependence of s and D on pressure, systems in non-aqueous solvents are implicitly excluded. The case of a continuous distribution of s will be considered first. The dis- tribution can be represented on a refractive index basis by g(s), where In this equation An0 is the total refractive increment, or the difference in refractive index between solvent and solution before sedimentation begins; Ans is the refractive increment of material with sedimentation coefficient < s.Thus the value of g(sJ is the relative frequency within a sample, on a refractive index basis, of material with sedimentation coefficient q. When all the species have the same specific refractive increment, g(s) is also the relative frequency on a weight basis.3 If the sedimentation coefficients vary with concentration then, by this definition (3.1), g(s) also varies with concentration. However, at present it is possible to obtain g(s) for such systems only by correcting the measurements to infinite dilution. In the simplest analysis, when diffusion and concentration dependence are treated as negligible,4 the distribution curve may be obtained by a simple trans- formation of co-ordinates for any one of the boundary gradient curves of (Siz/Gx) against x (after subtraction of a reference baseline).To plot the new curve the ordinate becomes (3.2a)22 HETEROGENEITY OF PROTEINS where xo is the distance of the meniscus from the centre of rotation and the value of s corresponding to this value of g(s) is given by s = In (x/xo)/w2t. (3.26) This analysis was used by Bridgman 5 in his study of the sedimentation behaviour of some glycogens. In these and in the following equations it is assumed that L) and temperature are constant and also that the specific refractive increment of each species is constant. Eqn. (3.2) were used by Baldwin and Williams 6 to define an apparent distribu- tion of sedimentation coefficient for the case in which diffusion is not negligible, but in which the effects of concentration dependence are negligible.They pointed out that, for this case, extrapolation to infinite time of the apparent distribution should give g(s). In order to study the extrapolation to infinite time, Gosting 3 solved the boundary spreading equation, for the case when Faxen’s equation 7 may be used to give the boundary gradient curve formed by a single sedimenting component (for higher terms, cf. Gosting 3) : + . . .I. (4.3a) D SU2 y2 = - [exp ( 2 ~ 0 2 t ) - 11. (4.4b) This equation applies when s and D are constants, when there is a region ahead of the boundary in which c does not vary with x, and when the initial period of restricted diffusion-caused by the proximity of the boundary to the meniscus- is negligible, From tXm solution to eqn.(3.3), the correct variable was found for use in the extrapolation to infinite time and also the region of time in which the extrapolation should be carried out.3 Since the resolving power of the ultracentrifuge is limited by the length of the cell and the maximum permissible speed of rotation, it is not possible to study certain systems in the region of time indicated (for a given w) by Gosting’s theory. Consequently a solution to the boundary spreading equation was also sought 8 for the case that g(s) is Gaussian. The purpose of doing this was to estimate the error in g(s) when the extrapolation procedure is used in a range of time where the extrapolation is not strictly linear.It was found that a satisfactory representation of g(s) could be obtained only when the boundary spreading caused by heterogeneity was fairly large relative to that caused by diffusion. This type of analysis, in which the effects of diffusion are removed by extra- polation to infinite time, was found to be very useful in connection with size distribution analysis for plasma extender materials such as dextran 9*10 and gelatin.11 The effects of concentration dependence of sedimentation coefficient were sufficiently severe that a fwther correction to infinite dilution was required, either by extrapolation 99 11 or calculation.loD 12 Thus the apparent distribution defined by the following equations, (3.5a) 5 = In (x/xo)/w2t, (3.56) is still a function of the initial total concentration CO after extrapolation to infinite time.The new quantity 8 is a reduced co-ordinate with the units of sedimentationR. L . BALDWIN, L. J . GOSTING, J. w. WILLIAMS, R. A. ALBERTY 23 coefficient. We now label as g*@, CO) the quantity obtained by extrapolating g*(S, Co, t ) against l/t at fixed values of G. It can be extrapolated to infinite dilution, and thus to g(so), the distribution of sedimentation coefficient at infinite dilution, in various ways. One method which seems to work rather well is as follows. After obtaining the curve of g*@, CO) against 3 from each of several experiments with different initial concentrations, Coy one plots against CO the value of 5 which is found at a particular ratio (e.g.0.1, 0.2, . . ., 0.9, 1.0, 0.9, . . ., 0.2, 0.1) of g*(% Co)/g*(3, C O ) ~ ~ . It would be helpful to have a theory to guide this extrapolation. The method of correcting g*(P, CO) to infinite dilution 12 from just one experi- ment (and a knowledge of the concentration dependence of s) involves certain assumptions, the most serious being that the sedimentation coefficients are a function only of the total concentration of the mixture. Consequently the cal- culation is applicable only when the effects of concentration dependence are mild -roughly speaking, when s/so > 0.9, where s is a sedimentation coefficient at the concentration of the experiment and so is the corresponding value at infinite dilution. The method is essentially that of Jullander 13 but has been modified to take into account, approximately, the Johnston-Ogston effect? Using this correction to infinite dilution gave essentially the same distribution curve 12 for a dextran sample as that found by extrapolation to infinite dilution.9 When s/so > 0.9, there is little difference in the distribution curves 14 obtained with and without correction for the Johnston-Ogston effect; if no account is taken of the Johnston-Ogston effect, then very little labour is required to correct g*(S, CO) to infinite dilution : (3.6a) (3.66) In eqn.(3.6b), C is the total concentration at the co-ordinate 3 (midway through an experiment : because of the dilution effect, the curve of C against 3 changes slowly with time during an experiment) and soi is the sedimentation coefficient at infinite dilution of species i.It is assumed that the sedimentation coefficient of i can be represented as soi (1 - kC). The derivative dP/dsoi) is obtained from the plot of 3 against the corresponding value of soi (eqn. 3.6b) by graphical or numerical differentiation. With the development of interferometric optics and a double cell, it should become possible to measure accurately the boundary re- fractive index curves for many substances at concentrations sufficiently dilute that s/so > 0.9. A different kind of analysis is required for systems composed of a small number of components as well as for systems where, because the resolving power of the ultracentrifuge is insufficient,g g(s) cannot be obtained by the procedure involving extrapolation to infinite time.The standard deviation p of the distribution of s can be found from c9, the second moment about the mean G) of the boundary gradient curve. When the sedimentation and diffusion coefficients are constant and the period of restricted diffusion is negligible, 02 is given by 8 ~ 1 5 (3.7) where s and Td are the mean sedimentation and diffusion coefficients. An equation for u2 has recently been derived 16 which takes account of the dependence of sedi- mentation coefficients on concentration and of the initial period of restricted diffusion. Its use cannot be described here, but it can be said from the magnitude of the concentration-dependence term that eqn. (3.7) is adequate to describe relatively few experimental cases. A commonly applied test for heterogeneity 17 is the comparison of an experi- mental boundary gradient curve with the curve predicted by Faxen's eqn. (3.4) for a single solute. This test does not take account of the dependence of s on c 0 2 = (pW23)2(1 + . . .> + 2 E ( l + Sw2t + . . . },24 CHYMOTRYPSIN AND CHYMOTRYPSINOGEN and, as is shown by the size of the concentration dependence terms in the equa- tion 16 for 02, such a test is valid only for concentrations far more dilute than those customarily used, even in the case of the “globular” proteins. Conse- quently, it is of great interest that Fujita has obtained a solution 18 to the differential equation 19 for the case that s is a linear function of c. 3. Sedimentation 1 Johnston and Ogston, Trans. Faraday SOC., 1946, 42, 789. 2 Trautman, Schumaker, Harrington and Schachman, J. Clzem. Physics, 1954,22, 555. 3 Gosting, J. Amer. Chem. SOC., 1952, 74, 1548. 4 Signer and Gross, Helv. chim. Acra, 1934, 17, 726. 5 Bridgman, J. Amer. Chem. SOC., 1942, 64,2349. 6 Baldwin and Williams, J. Amer. Chem. Soc., 1950, 72, 4325. 7 Faxen, Arkiv Mat., Astron., Fysik B, 1929, 21, no. 3. 8 Baldwin, J. Physic. Chem., 1954, 58, 1081. 9 Williams, Saunders and Cicirelli, J. Physic. Chem., 1954, 58, 774. 10 Ogston and Woods, Trans. Faraday SUC., 1954, 50, 635. 11 Williams and Saunders, J. Physic. Chem., 1954, 58, 854. 12 Baldwin, J. Amer. Chem. SOC., 1954, 76,402. 13 Jullander, Arkiv Kemi, Miner. Geol. A , 1945, 21, no. 8. 14 Baldwin, paper presented at the Sept. 1954, Amer. Chem. Sac. meeting in New York. 15 Williams, Baldwin, Saunders and Squire, J . Amer. Chem. SOC., 1952, 74, 1542. 16 Baldwin, Biochem. J., to be published. 17 Svedberg and Pedersen, The Ultracenfrdfuge (Oxford University Press, 1940), p. 286. 18 Fujita, J. Chem. Physics, 1955, in press. 19 Lamm, Arkiv Mat., Astron., Fysik B, 1929, 21, no. 2.

ISSN:0366-9033    

DOI:10.1039/DF9552000013

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

24 CHYMOTRYPSIN AND CHYMOTRYPSINOGEN CERTAIN PHYSICAL PROPERTIES OF CHYMOTRYPSIN AND CHYMOTRYPSINOGEN USING THE DEPOLARIZA- TION OF FLUORESCENCE TECHNIQUE BY V. MASSEY,*$ W. F. HARRINGTON t AND B. S. HARTLEY * Departments of Biochemistry * and Colloid Science,? University of Cambridge Received 31st May, 1955 The work reported here originated in an attempt to determine whether any changes in physical state accompanying the conversion of chymotrypsinogen to chymotrypsin would help in our understanding of the catalytic properties of this enzyme. The fluorescence polarization method of Weber 1 was used in this study because of its sensitivity in dealing with low-molecular-weight proteins. In fact few physical differences between chymotrypsinogen and chymotrypsin were found. The chief difference found in physical properties between the two proteins is a hitherto unreported polymerization reaction of chymotrypsin which does not occur with the precursor under the experimental conditions studied.However, the most important conclusion from the work, which must necessarily remain tentative until values of fluorescent lifetime are determined, is that a group is free in chymotrypshogen which is identical with a part of the active centre of chymotrypsin. * Imperial Chemical Industries Research Fellows, Dept. of Biochemistry, Cambridge University. -f Fellow of the National Foundation for Infantile Paralysis, U.S.A. (1953-4). $ Present address : Edsel B. Forn, Institute, Detroit 2, Michigan, U.S.A.V. MASSEY, W. F. HARRINGTON A N D B .S . HARTLEY 25 EXPERIMENTAL CHYMoTRYPs1NoC;EN.-The chymotrypsinogen used was alcohol-crystallized material obtained from Worthington Biochemicals, New Jersey, U.S.A. CC-CHYMOTRYPSIN was prepared from chymotrypsinogen by activation with trypsin according to the method of Kunitz and Northrop.2 The enzyme was crystallized four times from ammonium sulphate solution at pH 4. DIP-CHYMOTRYPSIN was prepared from a-chymotrypsin by reaction with di-isopropyl fluorophosphonate (DFP), and crystallized four times, according to the method of Jansen et aZ.3 the method of Weber.1 PREPARATION OF FLUORESCENT coNJuGAm.-Protein samples were dialyzed against cold 0.1 M Na2HP04 solution. One-tenth volume of a cold acetone solution of the dye (1-dimethylaminonapthhalene-5-sulphonyl chloride) was added, and the reaction allowed to proceed in the cold, generally for a period of 18 h.The labelled protein was then dialyzed in the cold with changes of the buffer required for the particular experi- ment until the dialysate was free from fluorescence. jugated per molecule of protein was calculated by the following method. The protein concentration was determined by its 280 mp absorption before reaction, corrected for the volume change during dialysis. The conjugated dye concentration was calculated from the extinction coefficient of the dye on these proteins.4 This extinction coefficient has been determined as 3-3 x 106 cmz/mole and is independent of the degree of labelling. l-DIMETHYLAMINONAPHTHALENE-5-SULPHONYL CHLORIDE Was prepared according to CALCULATION OF THE DEGREE OF LABELLING.-The number Of mOleCUkS Of dye Con- CALCULATION OF RESULTS FROM FLUORESCENCE-POLARIZATION MEASUREMENTS be calculated from the equation 1 From the variation of fluorescence polarization with temperature the ratio ph/70 may where 7 0 is the lifetime of excited state of the fluorescence, ph is the mean harmonic rotational relaxation time at the temperature T, B is a term characteristic of the conjugate and determined by the variation of T is the absolute temperature r ) is the viscosity of the solvent.polarization with temperature, If TO is known then ph may be determined. Weber1 has found with conjugates of both serum albumin and ovalbumin that his results are consistent with a value of TO of 1.4 x 10-8 sec. However, our results suggest that with chymotrypsin conjugates the value of 70 can vary with the degree of labelling.Hence all the results are reported in the form ph. 20/70, where ph. 20 is the value of ph corrected to 20" C. A NOTE ON THE INSTABILITY OF CHYMOTRYPSIN-DYE CONJUGATES Most of the work reported here was carried out under pH conditions which favour the autolysis of chymotrypsin. Fortunately the enzyme is inhibited by conjugation. This inhibition is proportional to the degree of labelling and is due to the reaction of the dye with the active centre of chymotrypsin.4 However, at degrees of labelling below 1 molecule dye/molecule protein there is still appreciable proteolytic activity. Hence in measurements involving lightly-labelled chymotrypsin the temperature range over which reliable information can be obtained is restricted to 0"-25".At higher degrees of labelling where the inhibition is virtually complete, higher temperatures can be tolerated. However, as a matter of routine all fluorescence polarization readings were restricted to 0-25" as it has been found with these proteins that the dye is hydrolyzed from the protein at an appreciable rate at pH 7-9 when the temperature exceeds 30". This instability is markedly different from that found with other protein conjugates,l andis presumably due to an unstable conjugate of the dye with part of the active centre. Hence, to avoid spurious fluorescence polarization readings the following conditions have to be observed with the labelled proteins used in this study : (i) The conjugate used should be freshly prepared and the fluorescence polarization readings made as soon as possible.26 CHYMOTRYPSIN AND CHYMOTRYPSINOGEN (ii) The conditions used must avoid proteolysis. (iii) The conditions used must avoid hydrolysis of the dye from the protein. As an experimental check that these conditions were fulfilled the lowest temperature reading of polarization was repeated at the end of the experiment after the highest tem- perature reading.Any liberation of low-molecular weight fluorescent compounds due to proteolysis or dye hydrolysis is then easily detected by a lowered polarization reading. RESULTS VARIATION OF ph. 20/70 WITH PROTEIN CONCENTRATION Fig. 1 shows the effect on P~.~O/TO of varying the concentration of 0.50 labelled (i.e. 0.5 mole dyelmole protein) or-chymotrypsin in 0.01 M phosphate buffer pH 7-9.The value of ph.20/70 extrapolated to zero protein concentration is 5. Thus over the range of I 5 10 Protein concentration (rnq/ml) FIG. 1.-The variation of ph.20/70 with protein concentration in 0.01 M phosphate pH 7.9. x 0.50 labelled a-chymotrypsin ; 0 1-56 labelled chymotrypsinogen. protein concentration 0-10 mg/ml there- is a fourfold increase in ph. 20/70. Assuming that TO does not change with protein concentra- tion this would indicate a considerable in- crease in average molecular volume over this concentration range. pH-Dependent dimerization of chymotrypsin has been ob- served by Schwert and Kaufman.5 Although the highest pH used by these authors was pH 6.2, the trend of dimerization with pH which they observed indicates that at pH 7.9 considerable dimerization would not be ac- hieved until higher protein concentrations than those used here.Even if dimerization were occurring under these conditions it would hardly account for such large varia- tions in ph as indicated here. Sedimentation studies in the Spinco ultracentrifuge confirmed that these results were in fact due to considerable polymeri- zation. Fig. 2 shows the Schlieren diagrams of a-chymotrypsin in 0.01 M phosphate buffer, pH 7.9 at several different protein concentrations. It will be noticed that the polymer separates partially from a slower- moving component, although it never sep- arates completely into two symmetrical peaks. The leading edge of the slower component and the trailing edge of the faster component tend to merge together, indicating the interconversion of the two forms during the course of sedimentation. Hence the rate of attainment of equilibrium between the two forms must be rather slow.However, equilibrium must be attained within an hour, since no differences in sedimentation constants or proportions of polymer are found when a concentrated solution of protein is diluted and centrifuged immediately, and when it is allowed to stand for 24 h before centrifuging. It will be noted from fig. 2 that the proportions of the two components vary with protein concentration, so that at low concentration there is only a small proportion of polymer, increasing as the protein concentration increases. The proportion of polymer is not influenced by conjugation with the fluorescent dye, as is shown in fig.3 where the proportion of polymer (obtained by measuring the Schlieren peak areas) is plotted against protein concentration. Fig. 4 shows the effect of protein concentration on the sedimentation constants of the two components. It will be observed that the apparent average size of the polymer0 b C d - - FIG. 2.-Schlieren patterns of a-chymotrypsin in 0.01 M phosphate pH 7.9. (a) concentration 3.5 (mg/ml) ; (c) concentration 10-4 (mg/ml). (b) Y Y 5.2 YY (4 Y Y 20.5 7 9 [To face 26V . MASSEY, W. F . HARRINGTON A N D B. S . HARTLEY 27 as judged by its sedimentation constant, is dependent on protein concentration up to a concentration of 12 mg/ml. At higher protein concentrations the sedimentation constant decreases with increasing protein concentration, suggesting that in this range the polymer is a single molecular species which exhibits the usual sedimentation-concentration de- pendence.The slower component has an extrapolated sedimentation constant of 2.9 S Pro t in con cen t rat ion ( mq/m I ) FIG. 3.-The variation of the proportion of polymer with protein concentration in 0.01 M phosphate pH 7.9. A 2.6 labelled a-chymotrypsin. X native or-chymotrypsin ; 0 0.60 labelled or-chymotrypsin ; Protein concentqation ( m q / m l ) FIG. 4.-The sedimentation constants of chymotrypsin in 0.01 M phosphate pH 7.9. x native or-chymotrypsin ; A 2-6 labelled a-chymotrypsin ; 0 0.60 labelled a-chymotrypsin ; 0 DIP chymotrypsin. at infinite dilution, while the extrapolated value of the polymer is 8-2 S.These values are possibly maximal and minimal values respectively, owing to interconversion during28 CHYMOTRYPSIN AND CHYMOTRYPSINOGEN the course of sedimentation. This interconversion should cause least error in the sedi- mentation constant of the monomer at low protein concentrations (when most of the protein is in the form of monomer) and least error in the sedimentation constant of the polymer at high protein concentrations (when most of the protein is in the form of polymer). Hence the extrapolated values shown in fig. 4 are not likely to be significantly different from the true values. The value of 2.9 S would correspond to a minimum molecular weight of 23,000 (for a sphere, no hydration) and the value of 8.2 S to a minimum molecular weight of 110,000.Hence the minimum value of polymer unit size at high protein con- centrations is 4.8 times that of the monomer. Fig. 1 also shows the variation of ph ~ O / T O with concentration of 1.56 labelled chymo- trypsinogen in 0.01 M phosphate buffer pH 7.9. Again ph. ~ O / T O is dependent on protein concentration, increasing a little over twofold over the concentration range 0-10 mg/ml. It is evident that the variation is much smaller than with chymotrypsin under the same conditions. The variation is in fact consistent with a dimerization also involving an increase in axial ratio. Sedimentation studies confirmed that chymotrypsinogen dimerized under these con- ditions. Fig. 5 shows the variation of sedimentation constant with protein concentration of native chymotrypsinogen and at two different degrees of labelling. The value of S20, ~~0 extrapolated to zero protein concentration, of 2.70 S, agrees well with values found under different conditions by other workers.6 It should be noted that under these conditions .0 - 5 10 15 2 0 25 Chymofryptinoqrn Concentration (mq/mI) FIG. 5.-Sedimentation constants of chymotrypsinogen in 0.01 M phosphate pH 7.9. 1-56 labelled chymotrypsinogen. 0 native chymotrypsinogen ; A 0.89 labelled chymotrypsinogen ; the chymotrypsinogen sediments to give a single symmetrical Schlieren pattern, indicating that the equilibrium between monomer and dimer is established very rapidly. The extrapolated value of 3.3 S of the high protein concentration sedimentation constants indicates that the dimerization is accompanied by an increase in axial ratio, i.e. the dimerization is probably an end-to-end association.The conditions in these experiments differ in two ways from those employed by pre- vious workers,s where dimerization of chymotrypsin was observed. Not only is the pH higher but the ionic strength is considerably lower. Schwert and Kaufman 5 concluded that the dimerization they observed was independent of ionic strength, but only two ionic strengths were used, 0.2 and 0.5. Here the ionic strength was 0.03. In order to see whether the polymerization of chymotrypsin reported here was due to the high pH or the low ionic strength, the sedimentation studies were repeated with 0.1 M phosphate buffer pH 7-9 (ionic strength = 0-2).A representative Schlieren pattern for chymotrypsin sedimented under these conditions is shown in fig. 6. Now the protein sediments as a single symmetrical component at all concentrations studied. Fig. 7 shows the variation of the sedimentation constant with concentration under these conditions. At high ionic strength dimerization only occurs, compared to the extensive polymerization at low ionic strength.FIG. 6.-Schlieren diagram of a-chymotrypsin in 0.1 M phosphate pH 7.9. (concentration 18 mg/ml). [To face 28V. MASSEY, W. F . HARRINGTON AND B . S. HARTLEY 29 VARIATION OF Ph.20/70 WITH THE DEGREE OF LABELLING It has been shown in the previous section that the variation in ph. 20170 with protein concentration is due to aggregation of the particular proteins studied. Weber 1 has found with fluorescent conjugates of ovalbumin and serum albumin that p h / ~ o is independent of the degree of labelling.With chymotrypsin, however, at any constant protein con- \ O X \ ~ 3.5 - 3.0 - (Sved bergs) I 1 5 10 15 2 0 2 5 Protein concentration (mq/ml) FIG. 7.-Sedimentation constants of chymotrypsin in 0.1 M phosphate pH 7.9. 0 a-chymotrypsin, x DIP chymotrypsin. Deqrec of labelling (molt dye/mol protein) FIG. 8.-Variation of ph. 2 0 / ~ ~ with degree of Iabelling (0.01 M phosphate pH 7.9, protein concentration 1-0 mglml). X cx-chymotrypsin ; 0 DIP chymotrypsin ; A chymotrypsinogen. centration ph. 20/70 varies quite markedly when the degree of labelling is greater than 1 molecule of dye per molecule of protein.Results for a protein concentration of 1 rng/ml are shown in fig. 8. Also shown in this figure are results for labelled chymotrypsinogen and label Izd DIP chy mot rypsin (i .e. crystal line di-isopropylfluorophosphonate-reacted30 CHYMOTRYPSIN AND CHYMOTRYPSINOGEN chymotrypsin which has then been conjugated with the fluorescent dye). It is evident that there is no significant difference between these three proteins in the variation of ph. 20/-ro with degree of labelling. The possible significance of this observation concerning the active centre of chymotrypsin will be considered in the discussion. A possible explanation for this variation was that the coupling of the dye to the protein resulted in physical changes, which became apparent only at high degrees of labelling.However, such an explanation is unlikely in view of the fact that the sedimentation characteristics are unchanged on coupling with the dye, even at high degrees of labelling. This can be seen from an inspection of fig. 3,4 and 5, which show that the proportion of polymer and the sedimentation constants are unchanged on conjugation. It thus seems unlikely that any major physical changes have accompanied conjugation. The most feasible explanation for this variation is that ph. 20 remains unchanged, and that the in- crease in ph. 20/70 found as the degree of labelling increases is due to a decrease in 70. Reaction of 1-dimethylaminonaphthalene-5-sulphonyl chloride with amino acids and amino acid derivatives has shown,4 that as well as reacting with free amino groups, this dye will also react with imidazole, sulphydryl and phenolic hydroxyl groups.If the lifetime of excited state of the dye depends on the group with which it reacts, and if different groups have reacted at different degrees of labelling, then such variations in 70 with degree of labelling would be expected. DISCUSSION THE VALUE OF 70 OF FLUORESCENT CONJUGATES OF CHYMOTRYPSINOGEN, CHYMO- TRYPSIN AND DIP CHYMOTRYPSIN The value of ph.201~0 extrapolated to zero protein concentration of 0.50 labelled chymotrypsin is approximately 5 (fig. 1). From previous sedimentation and diffusion data 5 s 7 it is known that the molecular weight of chymotrypsin is 21,500-24,000 and that the axial ratio is approximately 4. The value of p0.20 for a sphere of molecular weight 23,000 can be calculated from the equation1 po = 37 V/RT, where Vis the molecular volume.For a sphere of molecular weight For an ellipsoid of axial ratio 4, molecular weight 23,000 and 30 % hydration, the mean harmonic rotational relaxation time at 20" C, ph. 20, would be 5 x 10-8. As the zero concentration value of ph. &O is 5, then if the ph. 20 were 5 x 10-8, TO would be 1.0 x 10-8. Assuming a constant value of ph. 20 of 5 x 10-8 the variation of TO with degree of labelling is shown in fig. 9. Such a variation, occurring to appreciable extent above degrees of labelling of 1 molecule of dye per molecule of protein would be consistent with the preferential reaction of the dye with one specific group giving a higher TO than a non-specific reaction with other groups which give a lower 70.Weber 1b has found with conjugates of serum albumin and ovalbumin that a value O f To of 1.4 x 10-8 is consistent with his polarization data and independent determinations of rotational relaxation time.8 The considerably lower values of TO suggested above may be due to an error of extrapolation of ph. 20/70 to infinite dilution in fig. 1. The value of 70 calculated from this extrapolation is likely to be minimal owing to the curvature of the ph. 201~0 plot. Furthermore the cal- culation of theoretical ph. 20 from sedimentation and diffusion data could yield a lower ph. 20 value than the true value. In this calculation a perfect ellipsoidal shape was assumed. Deviations in shape from a perfect ellipsoid would be ex- pected to increase ph.Hence the values of 70 shown in fig. 9 probably represent minimal values. However, the above assumptions do not affect the relative values of 70 shown in fig. 9. 23,000, p0.20 = 2.14 x 10-8. THE POSSIBLE SIGNIFICANCE OF To CHANGES CONCERNING THE ACTIVE CENTRE OF There is good evidence for believing that the primary reaction of the dye with chymotrypsin is a specific labelling of a group in the active centre. Fig. 10 shows the inhibition of enzymic activity of chymotrypsin produced by varying CHY MOTRYPSINV. MASSEY, W. F . HARRINGTON AND B . S . HARTLEY 31 degrees of labelling with the fluorescent dye, showing stoichiometric inhibition by the first molecule of dye. That this inhibition is caused by a specific reaction with a group in the active centre has been shown 4 by reacting the dye with chymo- trypsin in the presence of a competitive inhibitor.Under these conditions the inhibition caused by any particular degree of labelling is considerably reduced. It is clear from fig. 3 and 4 that the active centre cannot be concerned in the polymerization of chymotrypsin at low ionic strength, since no differences in sedi- mentation constants or proportions of polymer are found between labelled and unlabelled chymotrypsin. If the mechanism postulated in the previous section to explain the variation in ph. ~ O / T O with degree of labelling is correct, then the first group to which the dye conjugates is probably the same in chymotrypsinogen, DIP chymotrypsin Deqrce of labelling (molr dye/mol protein) FIG 9.-The postulated variation of TO with degree of labelling.The values of 70 shown on this curve are calculated from fig. 1 and 8, to keep the value of ph. 20 6xed at 5 X 10-8 sec. I00 c 0 .- & E 5 0 .- x c .- 0-0 Deqree of labelling (mots d y e / mot protein) FIG, 10.-The inhibition of esterase activity of a-chymotrypsin produced by conjugation with dye. and a-chymotrypsin. This is clear from fig. 8 where no significant differences in ph. 20/70 between the three proteins can be detected. Hence it appears likely that the enzymically-inactive precursor of this enzyme, chymotrypsinogen, as well as the chemically-inhibited enzyme, DIP chymotrypsin, contain a specific group (capable of chemical reaction) which is essential for the catalytic activity of the active enzyme. Hence the tryptic activation of chymotrypsinogen to a-chymo- trypsin must involve the liberation of another group also essential to activity. This hypothesis finds support in the experiments of Doherty and Vaslow 9 who showed that chymotrypsinogen, as well as chymotrypsin, will bind one molecule32 ACTIVATION OF TRYPSINOGEN of a substrate per molecule of protein. Thus one might conciude that the primary site of reaction of the dye with these proteins is a group concerned with binding the substrate to the active centre, and that another group (the DFP-reacting site) 10 is necessary for the activation of this enzyme substrate complex. Our thanks are due to Mr. Boon for carrying out the sedimentation experiments. 1 Weber, (a) Biochem. J., 1952, 51, 145 ; (b) Biochem. J., 1952, 51, 155. 2 Kunitz and Northrop, J. Gen. Physiol., 1935, 18, 433. 3 Jansen, Nutting, Jang and Balls, J. Biol. Chem., 1949, 179, 189. 4 Hartley and Massey, unpublished. 5 Schwert and Kaufman, J. Biol. Chem., 1951,190,807. 6 Schwert, J. Biol. Chem., 1951, 190, 799. 7 Smith, Brown and Laskowski, J. Biol. Chem., 1951, 191, 639. Hartley and Kilby, 8 Oncley, Chem. Rev., 1942, 30, 443. 9 Doherty and Vaslow, J. Amer. Chem. SOC., 1953, 75, 928. Vaslow and Doherty, 10 Schaffer, May and Summerson, J. Biol. Chem., 1953, 202, 67. Oosterbaan, Kunst Biochem. J., 1954, 56, 288. J. Amer. Chem. SOC., 1953, 74, 931. and Cohen, Biochim. Biophys. Acta, 1955,16,299.

ISSN:0366-9033    

DOI:10.1039/DF9552000024

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

32 ACTIVATION OF TRYPSINOGEN MECHANISM OF ACTIVATION OF TRYPSINOGEN AND CHYMOTRYPSINOGEN BY HANS NEURATH AND WILLIAM J. DREYER Dept. of Biochemistry, University of Washington, Seattle 5, Washington, U.S.A. Received 6th June, 1955 In an attempt to elucidate the mechanism of the tryptic activation of trypsinogen and chymotrypsinogen, respectively, the molecular changes occurring during the conversion of these zymogens to their respective active forms have been determined by physical and chemical means. These have included the methods of peptide and end-group analysis, electrophoresis and sedimentation, and enzymatic activity toward synthetic substrates. The tryptic activation of trypsinogen appears to be a one-step process accompanied by the release of an acidic hexapeptide from the N-terminal portion of the molecule.In contrast, the activation of chymotrypsinogen involves several enzymatically active inter- mediates, which have been characterized by end-group analysis and electrophoresis, and the release of a basic dipeptide, serylarginine. The active enzymes, unlike the zymogens, are subject to reversible dimerization. It is concluded that the primary chemical event requisite for activation is the hydrolysis of a peptide bond between a basic amino acid residue (lysine in the case of trypsinogen, and arginine in the case of chymotrypsinogen) and an isoleucylvaline sequence. It remains to be determined whether concomitant intramolecular rearrangements are also involved in the transformation of zymogens to the active form. It is a well-established fact, recognized by physiologists and chemists many decades ago, that the proteolytic enzymes of the digestive tract are secreted by the cells in an inactive form and through extracellular transformations become active enzymes, capable of hydrolyzing proteins and peptides.These transforma- tions of zymogens to the active catalysts present features of unusual interest : 1 9 2 they are enzyme catalyzed and it has not yet been possible to duplicate these transformations by nonenzymatic means. The activating enzymes are proteolytic enzymes themselves which operate with a high degree of specificity and selectivity. In all known cases is the conversion of zymogen to the active form an irreversible process. The activation process represents the final steps in the formation of aH .NEURATH AND W. J . DREYER 33 biologically active protein and thus offers a close view of the relation of chemical structure to biological activity of the proteins. As the nature of the changes occurring during activation becomes recognized, inferences may be drawn of the nature of the catalytically active centre. Since it seems now well-established that both chymotrypsin and trypsin have only one active centre per molecule,3 the question arises whether the structural configuration characteristic of the centre is preexistent in the zymogen molecule or whether it is created during activation by a process of intramolecular rearrangement. Tine present discussion will deal with two activation systems, i.e., trypsinogen and chymotrypsinogen ; some of the previously reported findings 4.5969 7 will be integrated with data which have been more recently obtained in this laboratory and elsewhere.Since during enzymatic activation of the zymogens peptide bonds are opened, new terminal groups are formed. Although no large molecular weight changes occur, the release of small peptide fragments may cause changes in the electrical charge of the protein components. On the basis of these considerations, the following experimental methods were employed conjointly in the study of the activation process : (i) end-group analysis, employing Sanger's 1 : 2 : 4-fluorodinitrobenzene (FDNB) reagent for amino terminal groups and carboxypeptidase 49 9 for carboxyl terminal groups; (ii) peptide isolation by paper or ion exchange chromatography or by paper electrophoresis, followed by amino acid and end-group analysis of the isolated peptides ; (iii) electrophoretic analysis of the protein components of the activation mixture; (iv) sedi- mentation analysis in the ultracentrifuge ; and (v) measurement of enzymatic activity with the use of synthetic substrates (benzoyl-L-arginine ethyl ester for trypsin, acetyl-L-tyrosine ethyl ester for chymotrypsin).THE ACTIVATION OF TRYPSINOGEN Of the two activating systems considered herein the conversion of trypsinogen to trypsin will be considered first since, apparently, it presents a one-step reaction which does not involve the formation of intermediate products. The activation of trypsinogen can be independently catalyzed by three proteolytic enzymes,l i.e.(i) enterokinase, (ii) trypsin, and (iii) penicillium kinase. While the first of these is probably physiologically the most important, the autocatalytic conversion under the influence of trypsin has the operational advantage that no new protein is introduced into the system. The activation by penicillium kinase has the practical merit that the reaction occurs at an acid pH (PH 3) where autolytic degradation of trypsin is held to a minimum. Maximum autocatalytic conversion of trypsinogen to trypsin at pH 7.8-8 requires the presence of calcium ions to suppress the formation of enzymatically " inert protein ". From a kinetic view- point, the reaction can be formulated in terms of a bimolecular product-catalyzed chemical reaction.1 The molecular-kinetic properties of trypsinogen 10 and trypsin 11 have been considered at a previous Discussion of the Faraday Society.12 Some of the chemical and physical properties of trypsinogen and trypsin (or DIP-trypsin)," pertinent to considerations of the present problems, are summarized in table 1.Within the limits of the experimental error, the zymogen and the active enzyme have identical molecular weights of approximately 23,800, this value being in good agreement with the minimum molecular weight determined from amino acid analysis 13 (see below). Although no significant change in molecular weight accom- panies the tryptic activation of trypsinogen, the data summarized in table 1 clearly point toward the liberation of an acidic peptide from the N-terminal sequence * Here and elsewhere, DIP refers to di-isopropylphosphoryl, DFP to di-isopropyl- phosphofluoridate, and TCA to trichloracetic acid.B34 ACTIVATION OF TRYPSINOGEN of the trypsinogen molecule. Thus, according to the work of Desnuelles and co-workers, the N-terminal valine of the single polypeptide chain of trypsinogen is replaced by an isoleucine group, and the iso-ionic point of the protein shifts toward more alkaline regions. It is of interest that even after denaturation by 6 M urea, both proteins remain unreactive toward carboxypeptidase? As shown by representative data in table 2, the amino acid composition of trypsinogen and DIP-trypsin is almost identical. The observed differences in aspartic acid and lysine are in accord with the composition of the activation peptide (see below) and the calculated minimum molecular weights agree favourably with those determined by physical methods. TABLE 1 .-SOME PHYSICAL AND CHEMICAL PROPERTIES OF TRYPSINOGEN AND TRYPSIN (OR DIP TRYPSIN) trypsinogen trypsin molecular weight (sedimentation, diffusion) 23,800 10 23,800 11 E;g 13.9 14.4 N-terminal5 1 valine 1 isoleucine C-terminal (carboxypeptidase) 7 none none iso-ionic point (mixed bed ion exchange resin) 9.3 10.1 TABLE 2.-sOME COMPARATIVE DATA OF THE AMINO ACID COMPOSITION* OF TRYPSINOGEN AND DIP-TRYPSIN 13 amino acid aspartic acid glutamic acid arginine lysine histidine phenylalanine tyrosine proline 112 cystine trypsinogen 25.4 11.0 2.0 14-5 3.1 4.0 9.4 8.0 6.5 total N minimum molecular weight (calculated) 23,320 f 280 * given in terms of amino acid residue per mole.j- determined as cysteic acid. DIP-try pin 21-5 11.1 2-0 14.0 3.0 4.1 9.9 8.0 - 16.7 % 23,020 f 340 The activation peptide has been isolated and identified. In these experiments aliquots were removed from activation mixtures at various stages of activation, free amino acids and peptides adsorbed on to and eluted from Dowex-50 ion- exchange resin and separated on Dowex-50 ion-exchange columns (2 % cross- linked, sodium form). A representative chromatogram is shown in fig. 1. The amount of the peptide, represented by the major peak, was found to be proportional to the extent of activation, whereas the other, minor peaks failed to reveal any quantitative relation to the appearance of enzymatic activity and may, therefore, be ascribed to non-specific autolysis of the proteins.Quantitative amino acid analysis of the isolated peptide fraction yielded valine, aspartic acid and lysine in mole ratios of 1 : 4 : 1 in addition to fractional amounts of ammonia, probably introduced during the operation. The N-terminal position of valine was verified by Sanger's FDNB technique 8 and no other ether soluble DNP amino acid could be found; and since no valine appeared in the aqueous layer, the presence of only one valine per peptide was demonstrated. (E-DNP lysine was the only detectable DNP amino acid in the aqueous layer.) The C-terminal position of lysine was inferred from the specificity requirements of the activating enzyme, trypsin, but could not be demonstrated experimentally since the isolated peptide was unreactive toward carboxypeptidase.H .NEURATH AND W. J . DREYER 35 Accordingly, the structure val-(asp)4-lys has been assigned to this hexapeptide. The predominantly acidic properties, resulting from the presence of four aspartic acid residues, are in accord with the alkaline shift in iso-ionic point of trypsinogen during activation and was further confirmed by paper electrophoresis. In the presence of a citrate buffer, ionic strength 0.05, the isoelectric point was found to be 34-35, which is within the range expected for a polyvalent compound of the proposed structure. It is of interest to note that the peptide had no inhibitory effect on the esterase activity of trypsin when the enzyme was incubated with a 25-fold molar excess of the peptide.0 . 5 0 0 5 ACTIVATION PEPTIDE 0.400 0.300 0.200 0 100 I- pH 4.0 BUFFER I-PH 5.0 0.300- 0.200- 0.100- pH 5.0 BUFFER pH 6.0 BUFFER- ml. EFFLUENT FIG. 1.-Chromatogram (Dowex-50 ion exchange resin) of the peptide fraction of an activated sample of trypsinogen.7 I I Val - (Asp),-Lys Ileu ' 7 2 ; Trypsinogen + Val - (Asp), - L y s + Ileu '-c7 Peptide Trypsin FIG. 2.-Schematic representation of the tryptic activation of trypsinogen. The absence of a C-terminal amino acid, reactive toward carboxypeptidase, is indicated by the looped chains of the trypshogen and DIP-trypsh molecules. On the basis of all of these findings, the autocatalytic activation of trypsinogen may be depicted as shown in fig. 2. According to this scheme the single autolytic event accompanying activation is the splitting of the lysine-isoleucine bond, giving rise to the hexapeptide and the protein with an N-terminal, isoleucyl-valine sequence.* The general significance of these findings in relation to the mechanism of activation will be considered more fully in the discussi0n.t * The resulting decrease in molecular weight (705) is in accord with the change in ex- tinction coefficient (table 1).-f It is important to recognize that the experiments just described do not exclude the possibility that other peptides may have been liberated as well. All that can be said at this time is that no other ninhydrin-reactive peptide has yet been found which bears any quantitative relation to the amount of trypsin formed and which is subject to quantitative elution from Dowex-50 ion exchange resin.36 ACTIVATION OF TRYPSINOGEN THE ACTIVATION OF CHYMOTRYPSINOGEN The activation of this zymogen represents a more complex process since it is catalyzed by two enzymes, trypsin and chymotrypsin, and involves several en- zymatically active intermediates.The nature of the products of activation depends, therefore, on the concentration of both chymotrypsinogen and trypsin, a phenomenon which was clearly recognized by Jacobsen 14 by kinetic studies of the activation process. Under conditions of " rapid " activation (chymotryp- sinogen : trypsin ratio approximately 30 : 1) 6-chymotrypsin is the major end- product, whereas under conditions of " slow " activation (chymotrypsinogen : trypsin ratio 10,000: 1) the proteins described by Kunitz as a, and y-chymo- trypsins seem to be the major products of the reaction.Much of the published work on chymotrypsin was carried out with the crystalline product described by Kunitz 1 (a-chymotrypsin) without the realization that it represents electro- phoretically a heterogeneous mixture (see below). For this reason it is difficult ACTIVATION TIM E ELECTROPHORESIS 3.3 min. 6.0 min. 2 0 min. 55 min. 90 min. 1262 min. A 1389 min. I017 min. 1723 min. CHTG. 7- 8- FIG. 3.-Electrophoretic diagrams (ascending) of rapid activation mixtures. The patterns have been aligned horizontally to facilitate comparison of the three electro- phoretic components, chymotrypsinogen 7r and 8 chymotrypsins.23 to infer from a comparison of the chemical and physical properties of chymotryp- sinogen and a-chymotrypsin the nature of the changes which occur during the activation process.It is clear, however, that the two proteins have similar molecular weights 15~16 and that during the process several peptide bonds are opened ; thus, during the conversion to a-chymotrypsin four new end-groups are formed, i.e., N-terminal isoleucine and alanine,l7 and C-terminal leucine and tyrosine.9 Chymotrypsinogen is a homogeneous protein, as judged by solubility,l sedimentation,*s ion-exchange chromatography 18 and electrophoresis. Accord- ing to recent analytical data,lg the zymogen contains 18 basic amino acids (12 lysines, 2 histidines and 4 arginines) but only 12 acidic groups; there are 8 half- cystine residues (assumed molecular weight of 23,000).The protein is devoid of any reactive C-terminal groups 9 but contains one N-terminal half cystine.20H . NEURATH A N D W. J . DREYER 37 Rapid activation of chymotrypsinogen (chymotrypsinogen, 40 mg/d ; trypsin, 1.2 mg/ml, sodium phosphate buffer pH 7.8,0.05 M, 0") is accompanied by changes in electrophoretic patterns and mobilities and by the appearance of terminal groups. Characteristic electrophoretic patterns obtained in the presence of DFP (acetate buffer pH 4.98, ionic strength 0.1) are shown in fig. 3. It will be noted that during activation, the boundary corresponding to chymotrypsinogen is replaced by components of decreasing electrophoretic mobility. The mobility differences between chymotrypsinogen and the first activation product, rrchymo- trypsin are so small6 that they require prolonged electrophoresis for a visible resolution of the components.?-r-Chymotrypsin has been obtained in nearly pure form, as judged by electrophoresis, if activation was carried out in the presence of 8-phenylpropionic acid. Much larger mobility changes occur when rapid activation is permitted to proceed to the formation of 8-chymotrypsin. Evidence has already been presented to show that the conversion of rr-chymotrypsin to the &form is catalyzed by chymotrypsin rather than by trypsin.6 Thus the appear- ance of the electrophoretic peak characteristic of 8-chymotrypsin is greatly retarded by the addition of 8-phenylpropionate, an effective inhibitor of chymotrypsin, even though the activity of the activation mixtures remained unchanged.In contrast, soybean trypsin inhibitor had no effect on the rate of formation of 8-chy mo tr ypsin. The protein components resulting from rapid activation have been character- ized by end-group analysis; the results are summarized in table 3. Thus T- TABLE 3 .-DISTRIBUTION OF TERMINAL GROUPS * IN CHYMOTRYPSINOGEN, ACTIVATION MIXTURES AND CRYSTALLINE CHYMOTRYPSINS protein activation N-terminal C-terminal - CYSP chymotrypsinogen 20 - n-chymotrypsin 6 rapid cys/2 iIeu-vaI (a%> -1- 8-chymotrypsin 6 . 2 1 , * 2 rapid cys/2 ileu-Val leu chymotrypsin 6 extended, rapid cys/2 ileu-val leu tyr a-chymotrypsin 6 . 9 , 17 slow cys/2 ileu-Val ala leu tyr ,f3 : y-chymotrypsin 9,17 extended, slow (cys/2?) ileu-Val ala leu tyr * N-terminal groups were determined as DNP-derivatives, C-terminal groups with t inferred from composition of activation peptide (see below).the use of carboxypeptidase. chymotrypsin differs from chymotrypsinogen by the presence of an additional N-terminal group, i.e. isoleucine, whereas a-chymotrypsin possesses, in addition, a C-terminal leucine group. All three proteins contain the N-terminal half- cystine group. While rr-chymotrypsin has not yet been obtained in a crystalline form, the crystallization of a protein from an activation mixture corresponding to 6-chymotrypsin has been accomplished after inactivation with DFP.6s 22 However, electrophoretic analysis showed that no purification was achieved by this procedure; 6 on the contrary a solution of the crystals appeared electro- phoretically more heterogeneous than the original activation mixture. The appearance of a C-terminal leucine residue, together with a large shift in electrophoretic mobility during the T-8 conversion,6.23 provided presumptive evidence for the liberation of a basic peptide, and hence, attempts were made to determine the presence of this peptide, to isolate it, and to determine its chemical structure .23 To this end, activation mixtures were prepared and when the formation of 8-chymotrypsin was complete, the reaction was stopped by the addition of DFP.The fraction soluble in 10 % TCA was subjected to column chromatography on XE-64 ion-exchange resin. A representative chromatogram is shown in fig. 4. The first major peak appearing was found to consist of a mixture of acidic and neutral peptides which were subsequently resolved on a Dowex-50 column in the38 ACTIVATION OF TRYPSINOGEN sodium cycle.These are shown in the lower half of fig. 4. These peptides, none of which occurred in stoichiometrically significant quantities, have probably arisen from secondary proteolytic degradations. The centre peak was found to be primarily due to ammonia whereas the peak on the right represents the activation peptide. It appears at the same effluent volume as free arginine and, upon hydro- lysis, was found to yield only serine and arginine in a mole ratio of approximately 1 : 1. The presence of arginine was further c o w e d by a positive Sakaguchi reaction. The N-terminal position of serine was confirmed by a positive Nessler test for ammonia after periodate oxidation as well as by reaction with FDNB; thus, following acid hydrolysis, DNP-serine was found.Analysis for the peptide in the TCA-soluble portion, using paper electrophoresis. indicated that the peptide occurred only during the T-8 conversion. It was not AND n _ _ 10 20 30 40 50 60 - 7 0 8 0 I- pH 5.0 +-pH 6.5-1 EFFLUENT MILLILITERS FIG. 4.-Chromatogram of peptide fraction derived from the rapid activation of chymo- trypsinogen.23 The lower chromatogram @owex-SO ion exchange resin) represents the resolution of the fractions collected from tubes 4 to 12 of the upper chromatogram (XE-64 ion exchange resin). seen in activation mixtures containing only n-chymotrypsin. Other experiments indicated its absence when a-chymotrypsin was allowed to act on chymotrypsinogen.The conclusion that the peptide is liberated during the conversion of n- chymotrypsin to the &form is further strengthened by the quantitative results shown in fig. 5, in which the rates of formation and disappearance of the electro- phoretic protein components of the activation mixtures are shown as a function of time. Also plotted in this figure are the amounts of peptide liberated during activation. In these experiments the entire activation mixture, after treatment with DFP, was applied to the XE-64 column and the peptide determined quanti- tatively by elution with 0-3 M citrate buffer. It is evident from these results that the yield of peptide follows rather closely the curve describing the formation of a-chymotrypsin. Activity measurements on a similar activation mixture indicated that the per cent of maximally attainable esterase activity corresponded approxim- ately to the amount of chymotrypsinogen disappearing.H.NEURATH A N D W. J . DREYER 39 On the basis of the data just presented it is possible to identify the segment of the peptide chain of chymotrypsinogen which is primarily affected by the rapid activation process. It will be remembered that w-chymotrypsin differs from chymotrypsinogen in possessing an N-terminal isoleucyl-valine sequence and that the conversion of 7 ~ - to the 6-form yields no new N-terminal group, a Gterminal leucine group and the dipeptide serylarginine. Furthermore, since the release of the dipeptide appears to be exclusively associated with the m-8 conversion it is unlikely that serylarginine is a C-terminal sequence in chymotrypsinogen. The present data, therefore, suggest that tryptic hydrolysis of the argin yl-isoleucine bond in the sequence leucyl-seryl-arginyl-isoleucyl-valine is the single chemical event accompanying the formation of w-chymotrypsin.The subsequent chymo- trypsin-catalyzed conversion of w to the 8 form involves the hydrolysis of the leucyl- seryl bond, giving rise to the dipeptide, serylarginine, and a Gterminal leucine group and an N-terminal isoleucyl-valine sequence in the protein. It is to be anticipated that the action of the proteolytic enzymes present in activation mixtures will be more profound the slower the rate and the longer the time of activation. This is borne out by end-group analysis which has shown that upon prolonged incubation of a rapid activation mixture S-chymotrypsin is converted to an active enzyme which contains an additional C-terminal tyrosine RAPID ACTIVATION of CHY MOTRY PS I N OG E N 0 8-CHT 0 PEPTIDE A T-CHT tf 40f Q CHTG 20 0 10 20 30 40 50 60 70 80 90 Minutes of Activation FIG. 5.-Distribution of protein and peptide components of rapid activation mixtures of chymotrypsinogen as a function of time of activation.23 RELATION TO SLOW ACTIVATION group (table 3).The number of terminal groups is further increased, to four, when slow activation mixtures are permitted to stand under conditions leading to the formation of a-chymotrypsin. No additional changes in end-groups occur during the subsequent stages of slow activation required for the isolation of /3 and y-chymotrypsin; however, this type of analysis is not conclusive since these proteins are electrophoretically heterogeneous and since by happenstance partial degradation of the proteins may have led to a moiety having qualitatively identical terminal groups.The increasing complexity of the systems resulting from prolonged activation is clearly demonstrated by electrophoresis in the moving boundary apparatus for extended periods of time.24 Representative patterns of slow activation mixtures and of crystalline chymotrypsins are shown in fig. 6. Although these patterns are more heterogeneous than those of rapid activation mixtures, there is no detectable change in the mobility of the major component; indeed, it has not been possible to resolve the major peak of a mixture of rapid and slow40 ACTIVATION OF TRYPSINOGEN activation mixtures.Electrophoretic patterns of the crystalline chymotrypsins resemble in complexity those of slow activation mixtures, All of these crystal- line proteins contain more than one electrophoretic component and certain rela- tionships among the mobiIities of the various components exist. For instance, the electrophoretic mobilities of the major components of rapid and slow activa- tion mixtures and of a-chymotrypsin are identical under the conditions of these measurements (sodium acetate buffer, pH 4.97, 0.1 ionic strength) and are some- what lower than those of the major components of B and y-chymotrypsins, which are also identical. The mobilities of the slowest moving peaks of /3- and y- chymotrypsins are identical to that of the major peak of a-chymotrypsin.It is likely that the shift to somewhat higher mobilities attending the formation of /3- and y-chymotrypsin can be related to the release of a peptide having pre- dominantly acidic properties.4 I 2 . 3 4 1 5 1 6 FIG. 6.-Electrophoretic diagrams (as- cending) of slow activation mixtures and crystalline chymotrypsins in sodium acetate buffer, pH 4-97, 0.1 ionic strength.24 (1) Rapid activation mixture, chymotrypsinogen : trypsin 33 : 1, (DIP-&chymotrypsin). (2) Slow activation mixture, chymotrypsino- gen : trypsin 5000 : 1, 28 h activation. (3) Slow activation mixture, chymo- trypsinogen: trypsin 10,OOO: 1, 87 h of activation. (4) 2 x crystallized DIP-a-chymotrypsin.(5) 2 x crystal- lized DIP-y-chymotrypsin. (6) DIP derivative of 2 x crystallized p-chy- mo trypsin . THE EFFECT OF ACTIVATION ON SEDIMENTATION BEHAVIOUR It has been shown in previous studies of a-chymotrypsin that the concentration dependence of the sedimentation constant,lS, 1 6 2 5 diffusion constant,l6 viscosity and light scattering 26 is in many ways explainable in terms of a reversible dimer- ization. The extent of dimerization was reported to decrease with increasing pH, within the range of pH 3-86 to 6.2, in contrast to solutions of chymotrypsinogen which followed approximately the course expected for a monomer. In view of the electrophoretic heterogeneity of a-chymotrypsin it was deemed of interest to re-investigate the sedimentation behaviour of the products of rapid and slow activation. The results are summarized in fig.7. In contrast to crystallhe a-chymotrypsin, the DIP-derivatives of rapid as well as slow activation mixtures (chymotrypsinogen : trypsin 5000 : 1) revealed, at pH 3-86, a sedimentation behaviour characteristic of a pure monomer. Except for DIP-/3-chymotrypsin, only one sedimenting peak was seen in all patterns of the series represented by fig. 7. However, it was of considerable interest to note that activation mixtures corresponding to DIP-, T- and 6-chymotrypsins, while monomeric at pH 3-86 revealed at pH 7.5 a concentration dependence characteristic of a monomer- dimer equilibrium. Dimerization was found to be reversible both with respect3.0- e e A 9 0 e B m 0 ra m 0 0 0 ' 2.0' pH 3.0 GLYCINE r$= 0.1 40 0 9 , - A 0 .I I I I 1 25.0 2.0-t 0 5.0 10.0 15.0 20.0 Protein Conc. mg./ mi. FIG. 7.--Concentration dependence of sedimentation constants of chymotrypsinogen, activation mixtures and crystalline chymotrypsins. The symbols refer to the following preparations : A chymotrypsinogen ; (j rapid activation mixture containing /3-phenyl- propionic acid (DIP-n-chymotrypsin) ; 0 rapid activation mixture (DIP-8-chymotrypsin) ; 0 slow activation mixture, chymotrypsinogen : trypsin, 5000 : 1, 28 h of incubation ; 0 slow activation mixture (according to Kunitz and Northrop; 2 x recrystallized DIP-cc-chymotrypsin ; <) DIP derivatives of 2 x crystallized pzhymotrypsin. activation, appears to exist only as a monomer, regardless of pH, whereas n- and 8-chymotrypsin reveal a sedimentation behaviour at pH 7.5 characteristic of a monomer-dimer equilibrium.(ii) The relatively homogeneous products of rapid activation, n- and 8-chymotrypsin, differ from the crystalline proteins in the pH dependence of dimerization. Since, in terms of end-group analysis and electro- phoresis, n- and 8-chymotrypsin are more nearly homogeneous than the crystalline enzymes, the difference in sedimentation behaviour between chymotrypsinogen and 71- and 8-chymotrypsin appears to be more pertinent to the study of the activation process. 2.5- wQ DISCUSSION Comparison of the activation of trypsinogen and chymotrypsinogen shows certain common features of unusual interest: the splitting of a singIe peptide bond suffices to convert the zymogen to the active form.This step is in both cases catalyzed by trypsin and a bond between a basic amino acid (lysine and arginine, respectively) and an isoleucyl-valine sequence is broken. It is apparent Q rI2' 0.2 - El .y42 ACTIVATION OF TRYPSINOGEN that the activating enzyme operates with a high degree of selectivity and restraint. Thus trypsinogen has sixteen peptide bonds which might conform to the specificity of trypsin (contributed by 14 lysines and 2 arginines) 13 and chymotrypsinogen has an equivalent number (contributed by 12 lysines and 4 arginines).lg Nevertheless, only one of these bonds is broken during activation. Before considering the rela- tion of this single hydrolytic event to the appearance of enzymatic activity, it might be of interest to focus attention to those features which set these two activating systems apart.In trypsinogen, the release of a peptide is an obligatory accompaniment of the activation process. It is possible that the predominantly acidic properties constitute a unique feature of this activation peptide. Thus the N-terminal portion of the polypeptide chain of trypsinogen may not fit into the helical configuration of the molecule because of the electrostatic repulsion between the four adjacent aspartic acid residues which would tend to keep this segment in a fully extended configuration. The lysyl-isoleucyl bond may thus be exposed to the action of trypsin, whereas the other 15 vulnerable bonds remain protected within the helical configuration of the molecule.When activation is carried out in the absence of calcium ions, the helical structure might become sufficiently distorted to expose these additional bonds to tryptic hydrolysis, yielding " inert protein " or more profoundly degraded products. Alternatively, it might be suggested that the N-terminal hexapeptide sequence of the trypsinogen molecule, by virtue of its acidic properties, shields electrostatically basic groups of the potentially active centre and that the removal of the peptide is the structural event responsible for activation. However, this mechanistic interpretation appears less plausible if a common fundamental mechanism is to be ascribed to the activation of the two zymogens, since the release of a peptide is not involved in the primary step of the activation of chymotrypsinogen. As already mentioned, the rapid activation of chymotrypsinogen is a stepwise process mediated by the two proteolytic enzymes trypsin and chymotrypsin ; each of these opens, in sequence, a single peptide bond.Since the two products of activation, n- and S-chymotrypsin, have identical specific enzymatic activities the molecular changes responsible for the appearance of catalytic activity must be associated with the trypsin-catalyzed hydrolysis of the arginyl-isoleucine bond. Several related interpretations may be considered. From a chemical viewpoint it may be significant that trypsin and all active forms of chymotrypsin possess the N-terminal isoleucyl-valine sequence. If this configuration were part of the " active centre ", removal or chemical modification of these amino acid residues should have a profound effect on enzymatic activity. Alternatively it may be suggested that cleavage of the strategically located peptide bond may unmask a preexisting structural region responsible for catalytic activity; or else, more or iess profound intramolecular changes attending the cleavage of the arginyl-iso- leucine bond may create a configuration endowed with enzymatic functions.Such structural differences as may exist between the zymogen and the active enzyme might be revealed by X-ray diffraction analysis, and are perhaps also reflected in the tendency of the T- and 8-chymotrypsins, in contrast to chymotrypsinogen, to undergo reversible dimerization at pH 7.5.* It is worthy of note that active trypsin, unlike trypsinogen, is also subject to molecular association.*2 * Recent experiments in this laboratory indicate that the activation of both chymo- trypsinogen and trypsinogen is associated with a sigmficant decrease in optical laevo- rotation.29 In terms of the work of Kauzmann and co-workers27 these observations may be interpreted as indicative of structural rearrangements of the protein molecules.Thus the zymogens could be considered to be held in somewhat constrained configurations which are free to rearrange toward more highly folded configurations when the appropriate peptide bonds are broken, However, the greater sensitivity of the optical rotation of the active enzyme (S-chymotrypsin) as compared to the zymogen (chymotrypsinogen), to change in pH might hply, at the same time, a greater flexibility in the structure of the enzyme.H .NEURATH AND W. J . DREYER 43 As previously mentioned, trypsinogen can be activated also by enterokinase and by a pencillium kinase and, as recently reported, chymotrypsinogen can also be activated by a protease from B. subtiZis.28 It is to be expected that investigation of these activation processes by methods similar to those reported herein, will aid in the evaluation of the present conclusions and expand on their significance. The unpublished work described in this paper has been supported in part by the United States Public Health Service, research grant C-2286 and by funds made available by the people of the State of Washington, Initiative 171. 1 Northrop, Kunitz and Herriott, Crystalline Enzymes (Columbia University Press, 2 Green and Neurath, in The Proteins, ed.Neurath and Bailey (Academic Press, 3 Balls and Jansen, Adv. in Enzymol., 1952, 13, 321. 4 Neurath, Gladner and Davie, in The Mechanism of Enzyme Action, ed. McElroy 5 Desnuelle and Rovery, in The Chemical Structure of Proteins, ed. Wolstenholme 6 Bettelheim and Neurath, J. Biol. Chem., 1955, 212, 241. 7 Davie and Weurath, J. Biol. Chem., 1955, 212, 515. 8 Sanger, Biochem. J., 1945, 39, 507 ; 1949, 45, 126, 563. 9 Gladner and Neurath, J. Biol. Chem., 1953,205, 345 ; 1954,206,911. loTietze, J. Biol. Chem., 1953, 204, 1. 11 Cunningham, J. Biol. Chem., 1954, 211, 13. 12 Cunningham, Tietze, Green and Neurath, Faraday Soc. Discussions, 1953, 13, 58. 13 Cohen and Neurath, unpublished experiments. 14 Jacobsen, Compt. rend. trm. Lab. Carlsberg, Serie Chim., 1947, 25, 325. 15 Schwert, J. Biol. Chem., 1949, 179, 655. 16 Schwert and Kaufman, J . Biol. Chem., 1951, 190, 799, 807. 17 Rovery, Fabre and Desnuelle, Biochim. Biophys. Acta, 1953, 10,481 ; 12, 547. 18 Hirs, J. Biol. Chem., 1953, 205, 93. 19 Wilcox and Cohen, unpublished experiments ; see also ref. (2), table 3. 20 Bettelheim, J. Biol. Chem., 1955, 212, 235. 21 Rovery and Desnuelle, Biochim. Biophys. Acta, 1954, 13, 300. 22 Rovery, Poilroux and Desnuelle, Biochim. Biophys. Acta, 1954, 14, 145. 23 Dreyer and Neurath, J . Amer. Chem. Soc., 1955, 77, 814 ; J . Biol. Chem., in press. 24 Dreyer, Wade and Neurath, Archiv. Biochem. Biophys., in press. 25 Smith and Brown, J. Biol. Chem., 1952, 195, 525. 26 Steiner, Arch. Biochem. Biophys., 1954, 53, 457. 27 Kauzmann, in The Mechanism of Enzyme Action, ed. McElroy and Glass (Johns 28 Abrams and Jacobsen, Compt. rend. trav. Lab. Carlsberg, Serie Chim., 1951,27, 447. 29 Rupley, Dreyer and Neurath, Biochem. Biophys. Acta, in press, New York, 1948), 2nd ed., chap. 5 and 6. New York, 1954), vol. IIB, chap. 25. and Glass (Johns Hopkins Press, Baltimore, 1954), p. 50. and Cameron (Little, Brown & Co., Boston, 1953), p. 58. Hopkins Press, Baltimore, 1954), p. 70.

ISSN:0366-9033    

DOI:10.1039/DF9552000032

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

ACTIVITY OF CATALASE-LIPID COMPLEXES AT OILIWATER INTERFACES BY M. J. FRASER,* J. G. KAPLAN AND J. H. SCHULMAN Dept. of Colloid Science, The University, Cambridge and Physiology Dept., Dalhousie University, Halifax, N.S., Canada Received 25th May, 1955 The adsorption of the haem-enzyme, catalase, at the oil/water interface has been studied as a function of the nature of the interface. By the appropriate choice of stabilizers it has been found possible to obtain this protein adsorbed on emulsion droplets in the form of monolayers and multilayers varying in state from fully active to completely in- active enzyme. The activity of the enzyme may be correlated with the degree of un- folding of the protein, which is dependent upon the interfacial energy, charge on the protein and interface, pH, polar group specificity, balance of hydrophobic to hydrophilic groups, and steric factors.Provided the unfolding process has not gone to completion, desorption of the enzyme from the interface by charge reversal or displacement with a chemical agent, results in partial to complete restoration of activity. These results have been confirmed by a spectro- scopic study of adsorbed and desorbed oxyhaemoglobin. The action of the stabilizers at the oil/water interface on catalase activity may be correlated with the action of their water-soluble homologues in bulk solution. There is little need to re-emphasize the importance of interfaces in biological systems. Cheesman and Davies 1 have recently reviewed the physicochemical and biological aspects of proteins at interfaces.It is now generally agreed that the spreading of an enzyme at the airlwater (A/W) interface results in a complete loss of activity when the unfolding of the protein is permitted to proceed to com- pletion such that a homogeneous monolayer 8-lOA thick is formed. The forma- tion of such a film from a globular protein molecule approximately 50 A in diameter by adsorption at a clean A/W or oilfwater (O/W) interface involves a drastic reorientation of polypeptide chains, with the breaking of many intramolecular bonds. The final arrangement of the hydrophobic and hydrophilic side chains in a protein monolayer on clean A/W and O/W interfaces has been elucidated in earlier investigations of protein monolayers and recently in studies of synthetic polymers and co-polymers, especially of the polyamino acids.2-* The loss of structural specificity is simulated in bulk by the action of various physical and chemical agents such as heat, ultra-violet light, acid, alkai and urea.9 When proteins adsorb at interfaces at which amphipathic molecules sit, further complexities are introduced.At clean A/W and O/W interfaces the adsorption and unfolding processes were influenced largely by the interfacial energy, whereas when the interfacial energy has been reduced by the presence of other molecular species one has to consider chiefly the interaction of the protein with these molecules. Adsorption at such interfaces need not lead necessarily to a complete unfolding of a protein monolayer so that an adsorbed enzyme might retain some of its biological activity.At the mobile A/W and O/W interfaces (as opposed to the immobile solid/liquid interface where the solid surface greatly restricts movement in the adsorbed film 19) the techniques employed by Doty and Schulman,lo and Matalon and Schulman,ll in studying the adsorption of proteins on lipid mono- layers are useful, though no work has yet been done at the O/W interface. It * 1851 Exhibition Scholar. 44M. J . FRASER, J. G. KAPLAN AND J. H. SCHULMAN 45 proves more convenient to study the activity of enzymes at the O/W interface by employing O/W emulsions stabilized with various lipids. In addition to the factors that one must consider in a study of monolayer formation on clean surfaces, one has now to take into account such things as charge on the interface and protein, pH, specific interactions between the polar groups in the lipid and protein, association of the polar groups of the two molecular species through hydrogen bonding, ion-dipole interaction and of the non-polar groups by van der Waals' forces.In other words, one should recognize all the interactions that occur in bulk between lipids and proteins together with the special steric restrictions imposed by the alignment of the lipid at the interface. This paper reports the investigation of catalase adsorbed at O/W interfaces stabilized with various oil-soluble lipids. It is shown that if the unfolding process does not go too far then enzymatic activity may be partially or wholly retained. Desorption of the enzyme in this case leads to partial recovery of activity. * EXPERIMENTAL PROTEIN soLuT1oNs.-Three samples of catalase were used in this work, a 1 % horse kidney catalase in 1 % saline, a 2.5 % horse liver catalase in 1 % saline, and 1 % saline solutions of twice crystallized and lyophilized beef liver catalase obtained from Worthington Biochemicals, Freehold, New Jersey, U.S.A.A sample of freshly prepared 7-2 X 10-3 M beef oxyhaemoglobin was used in studying the adsorption of haemoglobin at the O/W interface. All protein solutions were stored at 4" C and not used longer than a three-week period, during which little or no loss of catalase activity occurred. Suitable dilutions in buffer were made when required. PREPARATION OF EMULSIONS AND PROCEDURE.-wateT-inSOlUble lipid stabilizers were chosen as far as possible to minimize the escape of material into the bulk phase and hence limit the interaction with the protein to the interface.The stabilizers were dissolved in the oil phase, warming slightly to aid solution when necessary. Usually emulsions were prepared by forcing one part oil phase to three parts distilled water through a hand- operated emulsifier three times. Appropriate amounts of freshly prepared emulsion were then added to buffered protein solutions. Activity measurements were made by the standard Warburg manometric technique on the whole mixture, and after centrifuga- tion, on the oil-free subnatant, and on control catalase solutions. This permitted cal- culation of the amount of adsorption at the interface and the activity of the adsorbed enzyme.In many cases the latter was measured directly also on emulsion droplets washed in buffer. Most activity measurements were made in a constant temperature Warburg respiro- meter at 25°C and were calculated from the linear portion of the oxygen evolution curves.16 Activities were directly proportional to enzyme and substrate concentrations over the ranges studied. The final concentration of hydrogen peroxide in the Warburg vessels varied from 0.1 5-1-5 % depending upon the particular experiment involved. Several temperature studies were undertaken to obtain activation energies for the bulk, interfacial and desorbed catalases. The " cephalin " series (see below) was studied over the range 5"-30" C in a refrigerated Warburg apparatus and the other series from 15"-30" C (winter room temperature and above).Controls were always run for pressure fluctu- ations (thermobarometers) and for activity of the emulsion without enzyme, which was vanishingly small to zero in all cases studied. The various systems studied will be outlined individually with reference to the stabilizers. (i) Sulphoiiate (Petvornor)-a commercial preparation (from Newton Chambm & Co. Ltd., Thorncliffe, nr. Sheffield) of a C21 anthracene sulphonate, molecular weight approxim- ately 420, containing 40 % Nujol, 50 % active material, 2 % inorganic salts was diluted 1 : 5 in Nujol. An emulsion 24 % in oil and 76 % in distilled water prepared from this material was very stable even when diluted 100-fold in 0.1 M buffer. The droplet size was determined by photomicrography.As this O/W interface bore a negative charge the adsorption of catalase was carried out on the acid side of the isoelectric point (I.P.) 12 at pH = 4.0 in 0.1 M standard McIlvaine citrate + phosphate buffer or at pH = 4.6 in 0.1 M acetate buffer. The same procedure held for oxyhaemoglobin. Adsorption of more than 10 mg protein per m2 of interfacial area resulted in flocculation of the emulsion. Desorption of the protein from the washed floccules was brought about by changing46 CATALASE-LIPID COMPLEXES the pH of the medium with sodium hydroxide to pH = 8.5-9.5 or with small amounts of dodecyltrimethyl ammonium bromide @TAB). (ii) CephaZin-a crude commercial preparation (Delta Chemicals, New York, N.Y., U.S.A.), a mixture of phosphatidyl serine and phosphatidyl ethanolamine.50 ml 0-01 % catalase, 20 ml 0.1 M acetate buffer, pH = 4.6 and 30 ml pure olive oil (0.05 % in cephalin) was put through the emulsifier. The oil droplets formed flocculated while the oil-free subnatant had a very slight milky appearance, possibly due to the precipitation of very small amounts of the cephalin. The droplet size was measured microscopically. The floccules were washed as many as six times in buffer without appreciable breaking. Desorption of the catalase from the interface was brought about by shifting the pH of the medium to pH = 9 or with small amounts of various alcohols (see results). (iii) Mixed sodium ZauryZ phosphates-a commercial preparation (Albright and Wilson, London, England). An emulsion of Nujol (1 % in sodium lauryl phosphates) in distilled water, one part oil + three parts distilled water, was added to 2.5 x 10-3 % catalase solutions in 0.1 M McIlvaine buffer, pH = 4.0, whereupon the emulsion flocculated, No change in pH occurred on preparation of the emulsion in distilled water indicating the absence of free acids.Desorption of the catalase from this interface was brought about by changing the pH of the medium to pH = 8-9 with sodium hydroxide. (iv) Hydrocarbons-adsorption of catalase on a Cg fraction of petroleum hydrocarbons (b.p. 125-155" C) was carried out by emulsifying 30 ml001 % catalase in 0.1 M McIlvaine buffer, pH = 5.6 with 10 ml oil phase according to the method of Cockbain.14 The oil droplets were washed four times in buffer and the activity of the washes and droplets tested.(v) OctadecyZtrimethyl ammonium bromide (CISTAB) and alcohol mixtures-a pure sample prepared in the laboratory. Adsorption of catalase on the alkaline side of the isoelectric point at these positively charged interfaces was brought about with 0-1 % c18TAB in the oil phase and CISTAB plus approximately equholar : (a) n-lauryl alcohol, (6) oleyl alcohol, (c) methyl nonyl alcohol, (d) trimethyl cyclohexanol and (e) cholesterol. Since polyvalent ions such as phosphate discharged these interfaces and possibly reverse the sign of their charge 11 preventing any appreciable adsorption of protein, 0.01 M and 0-1 M ammonium chloride + ammonia buffers at pH = 8-3 were used. The emulsions (10 ml Nujol + stabilizers and 30 ml distilled water) were added to 40 mlO.001 % catalase in buffer.Desorption from the interface was obtained by changing the pH of the medium to the acid side of the isoelectric point,12 pH = 4-0. The adsorption and desorption of oxyhaemoglobin at the C18TAB + n-lauryl alcohol stabilized interface was studied with the same procedures, following the techniques of Ekes, Frazer, Schulman and Stewart.12 RESULTS ADSORPTION AT THE SULPHONATE-STABILIZEJ3 INTERFACE At pH = 4.0 catalase is very strongly adsorbed on the negative sulphonate-stabilized interface. Measurement of the droplet size gave a diameter of 0.76 &- 0.2 p permitting estimation of the amount of protein adsorbed per unit area of surface. Well over 100 mg of catalase, and 150 mg of oxyhaemoglobin per m2 could be adsorbed. The adsorption of up to 20 mg/m2 of catalase may be examined in table 1. In all cases the activity remaining TABLE 1 .-ADSORPTION AND DESORPTION OF CATALASE AT THE SULPHONATE (PETROMOR) STABILIZED O/W INTERFACE.ACTIVITIES IN ARBITRARY UNITS activity in DTAB-treated solution (desorption) wt. protein activity * (mg/m2) phase expt. I expt. I1 a p t . I expt. I1 activity at surface P:~,","' in aqueous 1-8 0 0 0 0 0 9 0 66 59 418 376 10 0 60 60 290 290 12 0 75 858 1030 104 1892 1351 112 2494 1560 134 2466 1370 134 3420 1710 14 16 18 0 0 0 i:' 20 0M. J . FRASER, J . G. KAPLAN AND J . H. SCHULMAN 41 in the aqeuous phase was zero showing that all the enzyme was adsorbed (and no absorption spectrum or colour could be detected in the aqueous phase with oxyhaemoglobin). De- sorption was brought about in the presence of - 10-4 M DTAB.The results show that all activity was lost irreversibly in the first 8 mg/m2 and increasingly larger amounts of activity are recoverable beyond 9mg/m2. The activity at the interface in this region increases slightly but not in proportion, probably because the substrate does not have easy access to the inner layers of enzyme. That desorption occurs even where no activity is recovered in the aqueous phase after treatment with alkali (changing the pH to 9*0), was shown by nitrogen determinations on the aqueous phase. Desorption may not be complete, for only 80 % of the nitrogen was recovered. Desorption of the catalase by changing the pH to 9 from an interface at which 15 mg/m2 is adsorbed results in a 7-fold increase in activity, due correction being made for the effect of pH on enzyme activity.This is a slightly smaller increase than that obtained with DTAB (see table 1). Several experiments over a long period of time indicated a slow diminution in catalase activity at interfaces at which 10-20mg/m2 were present. The measurements here pre- sented were made at the same time interval after adsorption which was usually less than 1 h. Only " denatured " haemoglobin (parahaematin 12 or globin haemochrome 13) can be recovered on desorption by pH change (or treatment with DTAB) from 9 mg/mz or less, while a mixture of " denatured " and " native " oxyhaemoglobin is recovered from 150 mg/m2. The strong oxyhaemoglobin absorption bands at 535-550 mp and 565-580 mp fused into a weaker continuous band from 545-590 mp on reduction with sodium hydro- sulphite at alkaline pH (characteristic of oxyhaemoglobin) with a large absorption at 552-565 mp and a faint diffuse band at 525-530 mp (characteristic of parahaematin). The absorption by unreduced parahaematin is negligible.12 ADSORPTION ON THE CEPHALIN-STABILIZED INTERFACE Catalase is adsorbed at the cephalin-stabilized olive oil/water interface at pH = 4.6 with an 18 f 9-fold reduction in activity, measured at 25" C.In a sample experiment the following activities corrected for dilution factors, were found : activity of bulk catalase initial activity of adsorbed catalase activity of oil-free aqueous phase (unadsorbed enzyme) activity of 1st washing activity of adsorbed catalase after 1st washing activity of 2nd, 3rd, 4th washings activity of adsorbed catalase after 2nd, 3rd, 4th washings activity of desorbed catalase 783 32 30 0 30 0 30 120 These activities are averages of 2-3 determinations.From the results one may calculate that the reduction of activity on adsorption was (783 - 30)/32 = 23-5-f01d, while the % adsorption was 100 (783 - 32)/783 = 96. The recovery of activity on desorption with octanol amounted to about (120 x 100)/783 = -15 %. A study of the variation of the activities at the O/W interface and in the bulk phase with catalase concentration under the same conditions (see experimental) showed that the interface is " saturated " at relatively low concentrations (see fig. 1). At 0.008 % catalase the activity at the interface has become constant.Adding more catalase did not raise this but greatly increased the activity in the aqueous phase in contrast to ad- sorption at the sulphonate-stabilized interface. The size of the oil droplets determined microscopically was 6.89 p in diameter. From this one may calculate the surface area of the emulsion to be - 25 m2 (see experimental). This area apparently took up only 4 mg of catalase (50 ml of 0.008 % catalase) giving a specific area of approximately 6 m2/mg. Taking the density of catalase to be 1.37g/ml this leads to an apparent film thickness The activities of catalase in bulk, at the O/W interface, and after desorption (run under the same conditions) were measured over the temperature range 5"-30" C. From these measurements the experimental activation energy p, and heat AH+, entropy AS*, and free energy AF* of activation for the enzyme-substrate reaction were calculated (see table 2), assuming the velocity constant k" of Chance holds for the bulk enzyme,ls while a value 1/18th of this is taken for the adsorbed enzyme.16 With the exception of the p values these figures cannot be considered to have any more than comparative value.The adsorbed enzyme is seen to have a p value 10 kcal/mole higher than the enzyme in bulk with of only 1.2A.48 CATALASE-LIPLD COMPLEXES correspondingly different values for AH*, AS* and AF*, while the desorbed catalase is seen to have the same characteristics as the bulk enzyme before adsorption. Desorption of the enzyme from the interface was brought about by altering the pH of the medium to 9 with sodium hydroxide or treating with any of the following surface active agents, the C7-Cll aliphatic alcohols, n-octanol and n-octanone.The maximum increase in activity observed was 4-fold (usually 2-3-fold) but the active enzyme that could be desorbed gave the same p as that before adsorption (see table 2). The CIS alcohol and c16TAB gave no increases in activity under the conditions employed. TABLE 2.-THERMODYNAMIC CONSTANTS FOR ENZYME + SUBSTRATE REACTION (T= 25" C, pH = 4.0) tL (kcal/mole) A@ AS* (kcal/mole) (cal/moie deg.) system bulk catalase (n = 6) + 12.0 + 11.5 + 14 desorbed catalase (pH, 11.5 11.0 13 adsorbed catalase (n = 7) 22.2 21.7 43 octanol treated) (n = 2) L 480 - C 5 360- t? 1 2 4 0 - * .- .- v r .: .- 120- & U d Concentration of CataIase (% x 1900) A* + 7.3 7.1 8.9 (kcal/mole) FIG.1. Penetration 1% 11 of catalase into a film of cephalin spread at the A/W interface at pH = 4-6 on 0.1 M acetate buffer and held at 20 dyneslcm was demonstrated. There was a 32 % increase in film area in 3 h and a sudden decrease to less than half (20 %) this value in 9 h when sodium hydroxide was injected underneath, the a m changing the pH to 9.5. The control cephalin film was not appreciably affected by these operations, its area remaining constant. Very delicate fibres could be removed from this mixed film after partial desorption of the catalase. These results again show that the desorption process is not complete.10 ADSORPTION AT THE LAURYL PHOSPHATE-STABILIZED INTERFACE Adsorption of catalase at pH = 4.0 on sodium lauryl phosphate interfaces under the conditions employed is complete with a 15-20-fold reduction in activity.In three experi- ments changing the pH to 7-5-85 resulted in desorption of the enzyme with a 4-0-fold increase in activity or a recovery of 25 % of the original activity, corrections being made for the variation in bulk of activity with pH. At pH = 4.0, treatment with DTAB gave a 4-fold increase in activity. The emulsion droplets had a diameter of approximately 5 ,u (not accurately determined) which leads to a " saturated " surface of - 20 m2/mg. It was noticed that the activity of the adsorbed enzyme disappeared with time, only 47 % remaining after 34 h and no activity after 6 h. In later experiments the disappear- ance seems to have been more rapid, there being no activity at the end of 2 h.This factor prevented a convenient measurement of the activation energy for substrate decomposition by the adsorbed enzyme.M. J . FRASER, J . G . KAPLAN AND J . H. SCHULMAN 49 ADSORPTION AT THE HYDROCARBON/WATEF INTWACE Cockbain 14 has reported obtaining n-decane emulsions stabilized with bovine serum albumin at pH = 5.6 in which the protein is in the form of monolayers having a specific area of 0.7 m2/mg. In two experiments with catalase on nonane there was 90 % adsorp- tion with, respectively, a 31-fold and a 19-fold reduction in activity. The residual 2-5 % activity could not be washed off. Treatment of the emulsion with small amounts of DTAB did not affect the activity. The emulsion showed breaking if submitted to even moderate centrifugation. ADSORPTION AT INTERFACES STABILIZED WITH c18TAB AND ALCOHOL MIXTURES The percentage adsorption and reduction in catalase activity for the systems cl8TAB and CISTAB + n-lauryl alcohol, oleyl alcohol, methyl nonyl alcohol, trimethyl cyclo- hexanol and cholesterol at two salt concentrations are presented in table 3.The ad- sorption at the 618TAB and c18TAB + oleyl alcohol stabilized interfaces is seen to be relatively poor and variable, and the values for reduction in activity are correspondingly inaccurate. There is little or no reduction of activity at the cl8TAB + n-lauryl alcohol stabilized interface in high salt concentration, while the reduction is appreciable when cholesterol is present. At low salt concentration where over 90 % adsorption occurs the reduction in activity follows the order methyl nonyl alcohol m Iauryl alcohol < cyclohexanol < cholesterol.TABLE 3.-ADSORPTION OF CATALASE AT C18TAB STABILlZED INTERFACES (PH = 8.3) molar buffer system adsorption of reduction in activity concentration (stabilizers) catalase (yo) (-fold) 0 1 M ClsTAB 93 2.7 cl8TAB + n-lauryl alcohol c18TAB + oleyl alcohol 88 2.8 c18TAB + cholesterol 98 (n = 3) 4.6, 3.5 98 (n = 5) 1.0, 1.1, 1.2, 1.0, 1.1 0.01 M c18TAB 68, 100 54, co CISTAB + oleyl alcohol 20, 69 1.0, 1-1 CISTAB + n-lauryl alcohol 98,99 5.2, 11 c18TAB + methyl nonyl alcohol c18TAB + trimethyl cyclo- c18TAB + cholesterol 96, 100 15, co 86, 93, 96 2.2, 2.2, 13 hexanol 86, 92, 100, 100 3-0, 13, 13, 22 The CISTAB + n-lauryl alcohol system especially, and the c18TAB + cholesterol system at 0.1 M buffer were investigated more fully than the others.Catalase at these interfaces lost activity only very slowly. In two experiments with the n-lauryl alcohol stabilizer 50 % activity was lost in 6 h, while with cholesterol there was no appreciable loss over a 3-h period. All emulsions were more stable at the higher salt concentration. At low salt concentration breaking occurred with moderate centrifugation. The average diameter of the droplets in the CISTAB + n-lauryl alcohol system at 0.1 M buffer was approximately 2 p, leading to a specific area of nearly 40 mz/mg. Desorption from this interface occurred readily by shifting the pH of the medium to the acid side of the iso- electric point or on addition at alkaline pH of very small amounts of sodium dodecyl sulphate (SDS) so that the final concentration was - 10-4 M.The recovery of activity in the aqueous phase was 90-100 %. Activities of the c18TAB + n-lauryl alcohol and C18TAB + cholesterol systems were measured at pH = 8-3 over the temperature range 15"-30" C permitting of calculation of activation energies for the adsorbed catalase + H202 reaction. The bulk values in these experiments were 3100, 3800 and 5400 cal/mole, while the value for catalase at the cl8TAB + n-lauryl alcohol stabilized interface was 6900 cal/mole and for the C18TAB + cholesterol stabilized interface was 8200 cal/mole. When oxyhaemoglobin is adsorbed at pH = 8.3 in 0-1 M buffer on the C18TAB + n- lauryl alcohol stabilized interface, then desorbed by changing the pH to 4.0, the absorption bands of oxyhaemoglobin were still evident.On reduction with sodium hydrosulphite these fused to give a continuous weaker band from 540-590mp, with a slightly heavier absorption at 560mp and a very faint band around 530mp showing the presence of a small amount of parahaematin.50 CATALASE-LIPID COMPLEXES The results on adsorption of 1 mg of catalase or less per m2 of surface (monolayers of enzyme) are collected in table 4. It will be noted that as the reduction in activity on TABLE 4.-hSORPTION OF MONOLAYERS OF CATALASE AT THE OIL/WATER INTERFACE (1 mg/m2 or less) reduction % activity desorption stabilizer of interface in activity ,,:$kgle, (g$zOf,, recovered on PH 5.6 (nonane/HzO) 19-3 1 - - 2-5 % 4.6 lauryl phosphate 18 8.3 c18TAB + cholesterol 4 f 1 3 8 8.2 65-70 % 8-3 c18TAB + n-lauryl alcohol 1 6.9 90-100 % (-fold) 4.0,4-6 sulphonate a3 12.0 co 0 4.6 cephalin 18 12.0 22.2 20 % 20 % I adsorption becomes less the activation energy for substrate decomposition by the adsorbed enzyme approaches the value for the bulk enzyme.Also, the amount of activity that may be recovered on adsorption becomes greater in this direction. OBSERVATIONS IN BULK.-The effects of a soluble sulphate (SDS) at pH = 4.0 and of soluble TAB (DTAB) at pH = 8.3 in 0.1 M buffers on catalase activity and the absorption spectrum of oxyhaemoglobin were explored for comparative purposes. The activity of dilute solutions of catalase (-2 x 10-8 M) at pH = 4.0 disappears very rapidly in the region of 10-4 M SDS, a plot of catalase activity in the region of 10-5-10-3 M SDS having the appearance of a titration curve with the equivalence point at 2 x lO-4M SDS.Sodium lauryl phosphate under the same conditions (concentration, pH) does not affect catalase activity, nor do concentrations of SDS up to 10-3 M at pH = 8.1. In this region also the absorption spectrum of oxyhaemoglobin is converted into that of parahaematin. On the other hand, DTAB in the same concentration regions at pH = 8-3 is quite ineffective in bringing about the activity loss and spectral change and only does so when the concentration is about 0.5 % (-2 x 10-2M DTAB). At these con- centrations SDS will bring about these changes even at alkaline pH.DISCUSSION It is reasonable to correlate the loss of activity that occurs in the adsorption process with the degree of unfolding of the protein molecule. As the structural specificity of an enzyme in bulk is lost by changing the pH of the medium from that of optimum activity or through the action of " denaturing " agents, there is a corresponding loss in enzymatic activity accompanied by an increase in the activation energy for substrate decomposition. The breakage of many weak intramolecular bonds that occurs is also accompanied by a lowering of the energy barrier for thermal inactivation of the enzyme.17 Whether the structural loosening occurs in bulk or on adsorption at an interface, the same results are expected, provided that the substrate has equal access to the active centres in both cases.16 If the loosening of the polypeptide chains has not gone too far it might be expected that desorption from the interface would restore the enzyme to the original bulk condition, just as activity may be restored by bringing acid or alkaline solutions of enzymes back to the optimum pH if these conditions have not been too extreme. The results presented in table 4 show clearly that catalase may be adsorbed at the O/W interface in any state from fully active to completely inactive enzyme.The loss in activity, which may be taken as a measure of the degree of unfolding of the molecule, is dependent upon the nature of the interface. Despite the opposite charges on the TAB-stabilized interface and the protein at pH 8.3 the adsorption of catalase is very weak and leads to little or no unfolding of the enzyme.This could be accounted for by the fact that these interfaces may be partially discharged by traces of polyvalent anions.11 The influence of salt concentration on the stability of these emulsions and the amount of adsorption at the interface makes it difficult to draw any clear generalizations from the data presented in table 3. However, Hocking 18 has found that the presence of alcohols mixed with C16TAB stabilizer serves to raise the zeta potential of the emulsionM. J. FRASER, J. G . KAPLAN AND J . H. SCHULMAN 51 droplets markedly, indicating an increase in charge density at the interface ap- parently due to the condensation of the TAB film through van der Waals’ inter- action between the hydrocarbon tails of the molecules.In the series of alcohols studied, from n-Iauryl alcohol to cholesterol, the diminution in the activity of catalase on adsorption is greater as the size and hydrophobic character of the alcohol increase. The alcohol may thus be thought of as a “ spacer ” molecule acting to separate the positive TAB groups at the interface and thus stretching the interacting negative groups on the protein farther apart, resulting in a partial unfolding of the molecule. Or, again, as the hydrophobic character of the alcohol increases the stronger interaction with the non-polar groups of the protein could result in further unfolding. Adsorption of catalase at acid pH on the negative cephalin, lauryl phosphate and sulphonate interfaces results in considerable reduction in activity, complete in the latter case where the adsorption is very strong.The mechanism of adsorp- tion and unfolding likely involves not only association of the charged groups of the stabilizer and catalase, but also association of the non-polar side chains of the protein and hydrocarbon tails of the lipids. This would result in partial or complete penetration 1% 11 of the enzyme into the lipid film at the interface, as demonstrated for cephalin at the A/W interface. In this particular case also, there is considerable chance for hydrogen bonding and ion-dipole interaction between the oleic acid of the oil phase, the polar phosphate, serine and ethanol- amine residues of the cephalin, and the polar groups of the protein.The steric restrictions imposed by all these interactions are apparently enough to prevent a complete unfolding of the catalase. At the sulphonate interface the strong adsorption indicates a specific interaction with the positive groups on the protein resulting in a complete unfolding of the molecule. The small energies of the oil/lipid/water interfaces (interfacial tension-2 dynes/cm for the olive oil/cephalin/water interface, for example) are unlikely to contribute much to the unfolding of the protein on adsorption. This is probably a more important factor for catalase at the hydrocarbon/water interface. In the presence of excess protein 19 some of the adsorbed enzyme apparently does not have the opportunity to unfold and a residual activity is detected.Weak adsorption of the enzyme at the interfaces studied (with the exceptions of the nonane/H20 and that stabilized by sulphonate), indicated by the high apparent specific areas, suggests that the enzyme is adsorbed in patches. The reason that all the interface is not covered may be due possibly to a partial dis- charge of the interface by polyvalent ions or to preferential adsorption 12 of in- active catalase or other protein impurity which would not be detected by the methods employed here. The observed disappearance of activity of the adsorbed enzyme may be thought of as a continuing penetration with further unfolding as has been observed at the A/W interface.lo.11 Penetration of the enzyme into the lipid film at the interface may result in the formation of such strong complexes that desorption of the enzyme by pH change (charge reversal) or detergent action will be incomplete.This is indicated at the sulphonate interface by the fact that only 80 % of the nitrogen was recovered on desorption and that desorption of catalase from cephalin films at the A/W interface was incomplete. Whether this “ bound protein ” retains some activity is not yet known. Adsorption of catalase on the sulphonate-stabilized interface is in striking contrast to that on the other interfaces studied here. The unfolding process goes far enough to reduce the activity of 8mg/m2 to zero irreversibly, while the unfolding beyond this is less, as more activity can be recovered from the outer layers (see table 1). A less strong adsorption has been noted for oxyhaemogfobin on sodium hexadecyl sulphate stabilized interfaces 12 where 20 mg of protein were taken up per m2 from 2 % haemoglobin solutions.If flocculation of the emu& sion may be taken as an indication that the oil droplets have been discharged then it becomes difficult to account for the adsorption of any further protein.52 CATALASE-LIPID COMPLEXES Whatever the forces responsible, they apparently weaken when further layers are adsorbed, as the amount of unfolding becomes less drastic. When oxyhaemoglobin is adsorbed on the sulphonate interface only " de- natured " protein can be recovered from the inner layers whereas a large amount of undenatured oxyhaemoglobin may be desorbed from the outer layers where the unfolding has not gone too far.Also oxyhaemoglobin can be recovered by desorption from the ClgTAB + n-lauryl alcohol stabilized interface showing that little unfolding has taken place there. Finally, it is perhaps not surprising to note that the action of these oil-soluble stabilizers at the O/W interface on enzymatic activity is paralleled by the action I 2. or1 phase 0s-phase 5% activity, partlaltywversibk On deSWDtk2l 4. oil phav 3 F 3 y actlvp d a y e r . activity rotaiwd on desorptlon 5. 25% activity. portlally reverslbk on demrpm 6 addsaption d excess cmtcin an ckon hydrocarbon Interface - Inactive irnvcrsibk on -tbn FIG. 2.-Adsorption of catalase at various lipid-stabilized oil/water interfaces. of their water-soluble homologues in bulk solution. Few, Ottewill and Parreira 20 have found that DTAB binds to bovine serum albumin at relatively high protein and detergent concentrations giving a sudden increase in molecular asymmetry (as measured by viscosity changes) after 6 molecules of DTAB per protein molecule have been bound.The binding takes place through the charged carboxyl groups and the protein-detergent complex dissociates with decreasing protein concentration. This is likely the reason that DTAB does not affect the activity of the very dilute catalase solutions used here. The water-soluble disodium lauryl phosphate does not affect the activity of dilute catalase solutions either. The corresponding oil- soluble interfacial compounds do, on the other hand, reduce the activity of the adsorbed enzyme much less than the sulphonate. The binding of SDS to bovine serum albumin is much stronger than the binding of DTAB 20.24 and takes place through the available ionized cationic groups 24M.J . FRASER, J . G. KAPLAN AND J . H . SCHULMAN 53 -NH3+). The binding at neutral pH and high protein and detergent concentra- tions results in (i) a release of -SH groups, (ii) decreased solubility, (iii) loss of crystallizability, (iv) a marked increase in molecular asymmetry, (v) a rearrange- ment of the polypeptide chains, as determined by X-ray analysis.21 All these facts indicate a considerable amount of unfolding of the molecule and this would result in the loss of enzymatic activity. The effect of SDS on the activity of very dilute catalase solutions at acid pH shows that the binding to this protein is quite strong also.Schulman 22 has previously noted that long chain sulphates and sulphonates were very strong inhibitors of pancreatic lipase activity and that amines and quaternary bases did not inhibit, but even enhanced its activity. Wills23 has recently noted the great inhibitory effect of long chain and cyclic sulphates and sulphonates at low concentrations (0.001 M) on such a variety of enzymes including lipase, D-amino acid oxidase, invertase, urease, papain, rennin and salivary amylase. The action of the corresponding interfacial sulphonate on the activity of adsorbed catalase is drastic. The idea of the unfolding of catalase molecules adsorbed at the various lipid- stabilized oillwater interfaces is presented in diagrammatic form in fig. 2. At the sulphonate-stabilized interface 1 all grades from fully unfolded to globular enzyme are to be found in the form of multilayers.Partial unfolding of adsorbed monolayers of catalase occurs at the negative cephalin,5 4 the cholesterol-Clg TAB-stabilized interfaces, while at the n-lauryl alcohol418 TAB-stabilized inter- face 3 adsorption occurs without unfolding. Just as the interaction between cephalin and proteins is reflected in the penetration of monolayers of this lipid at the air-water interface and the partial loss of catalase activity at the cephalin- stabilized oil-water interface, so too is the lack of interaction between lecithin and proteins reflected in the fact that they do not penetrate monolayers of this lipid 10911 at the air-water interface or adsorb on to lecithin particles suspended in bulk.5 Adsorption of catalase on to the clean oil-water interface6 is a very labile process, the enzyme continually unfolding and spinning off the surface of the hydrocarbon droplets in the inactive form.Where excess oil surface is present the enzyme unfolds to give complete and irreversible loss in activity. BIOLOGICAL IMPLICATIONS In the investigations of Kaplan 25.26 and Fraser and Kaplan 16 it was postulated that the catalase of normal yeast cells exists partially unfolded at some interface of the oil/water type within the cell. Treatment of the cells with various physical and chemical agents or extraction of the enzyme was accompanied by an 18-fold increase in activity. This was envisaged as resulting from a desorption of the enzyme from the interface with consequent changes in the kinetic and thermo- dynamic properties of the catalase, as were observed.The catalase of these treated cells (called “ altered ” yeast catalase) came to resemble the extracted yeast catalase and crystalline beef liver catalase in bulk. It becomes immediately apparent that the systems of intermediate catalase activity studied here are striking models for the intracellular yeast catalase. The results presented herein thus add great force to this interfacial hypothesis. It is now evident that where dissimilarities in intracellular and extracted enzymes occur that yet another factor to account for this may be the difference in physical state of the two enzymes. The organization may differ from that in bulk, either because of adsorption of the intracellular enzyme on some interface similar to the “simple” O/W type studied here, or by association of the enzyme with the formed elements of the cell such as the microsomes and mitochondria.The inter- actions between enzymes and cellular lipids is likely important in many cases, and the role of interfaces in the life of the cell is becoming more evident.]. 2754 ACTIVATION OF MILK XANTHINEOXIDASE The authors wish to acknowledge the gifts of catalase from Dr. E. F. Hartree of the Molten0 Institute, from G. H. F. Fraser, for the oxyhaemoglobin from Dr. G. Hanania, for the lauryl phosphates from Dr. B. A. Pethica, and for the octadecyl TAB from Dr. A. V. Few, Colloid Science Department, Cambridge. Thanks are due to Mr. G. S. Coleman, Biochemistry Department, Cambridge, for the nitrogen determinations, and to Dr. Few and Dr. Pethica for many helpful discussions. We also wish to acknowledge the financial support of the National Research Council of Canada and the National Cancer Institute of Canada in part of this work. 1 Cheesman and Davies, Advances in Protein Chemistry, 1954, 9, 439. 2 Hughes and Rideal, Pruc. Roy. SOC. A , 1932, 137, 70. 3 Neurath, J. Physic. Chem., 1940,44,296. 4 Flory, J. Chem. Physics, 1942, 10, 51. 5 Huggins, J. Physic. Chem., 1942, 46, 151. 6 Singer, J. Chern. Physics, 1948, 16, 872. 7 Cumper and Alexander, Trm. Furuduy Suc., 1950, 46, 235. 8 Davies, Biochim. Biophys. Acfa, 1953, 11, 165 ; Biuchem. J., 1954, 56, 509. 9 Neurath, Greenstein, Putnam and Erickson, Chem. Rev., 1944, 34, 157. 10 Doty and Schulman, Furuhy SOC. Discussions, 1949, 6, 21. 11 Matalon and Schulman, Furuduy SOC. Discussions, 1949, 6, 27. 12Elkes, Frazer, Schulman and Stewart, Proc. Roy. SOC. A , 1945, 184, 102. 13 Lemberg and Legge, Hemutin Compounds and Bile Pigments (Interscience Publishers 14 Cockbain, personal communication. 15 Chance, in Sumner and Myrback, The Enzymes (Academic Press, New York, 1951), 16 Fraser and Kaplan, J. Gen. Physiol., 1955, 38, 515. 17 Glasstone, Laidler and Eyring, The Theory of Race Processes (McGraw-Hill, New 18 Hocking, Ph.D. Diss. (University of Cambridge, 1953). 19 Cumper and Alexander, Rev. Pure Appl. Chem., 1951, 1, 121. 20 Few, Ottewill and Parreira, to be published. 21 Putnam, Advances in Protein Chemistry, 1948, 4, 79. 22 Schulman, Truns. Furuduy Suc., 1941,37, 134. 23 Wills, Biuchem. J., 1954, 57, 109. 24 Pallansch and Briggs, J. Amer. Chem. Suc., 1954, 76, 1396. 25 Kaplan, Expt. Cell. Res., 1955, 8, 305. 26 Kaplan, J. Gen. Physiol., 1954, 38, 197. 27 Danielli, Nature, 1945, 156, 468. Inc., New York, 1949). vol. 2, part 1. York, 1941).

ISSN:0366-9033    

DOI:10.1039/DF9552000044

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

54 ACTIVATION OF MILK XANTHINEOXIDASE ACTIVATION AND INACTIVATION OF MILK XANTHINEOXIDASE BY PHYSICOCHEMICAL MEANS BY L. ROBERT AND J. POLONOVSKI Service de Chimie Biologique, FacultC de MCdicine, Paris Received 7th June, 1955 Xanthineoxidase (X.O.) is found in fresh milk in a particular physicochemical state, attached to the surface of fat globules as lipoprotein cenapses. These cenapses may be broken by a variety of physicochemical means : heat, pressure, ultrasonic waves and detergents, which, before denaturing the enzyme, cause a steep rise in xanthineoxidase activity. This activation is a characteristic feature of milk X.O. We studied the kinetics and the mechanism of activation and of inactivation of X.0. at different stages of purity, by means of heat, high pressures and ultrasonic waves, andL.ROBERT AND J. POLONOVSKI 55 calculated the activation energy of its heat denaturation. The inactivation of X.O. by ultrasonic waves has been compared with the destruction of riboflavine in the same conditions. These studies throw some light on the mechanism of the action of the enzyme and on the nature of its lipoprotein complex in milk. Milk X.O. is one of the most intensively studied enzymes, its structure and mechanism of action being now relatively well understood.109129 249 25 Recent progress in cell fractionation techniques has furthered studies on the variation of enzyme activity in relation to changes of structural elements. A good under- standing of such phenomena is essential for elucidating the mechanism of multi- enzyme action, e.g.in mitochondria or in microsomes.~* 149 29 Milk X.O. is integrated into morphological elements in the fat globules, its activity being largely conditioned by the state of a lipoprotein complex. A characteristic feature of milk X.O. is that its activity is low in very fresh milk and increases considerably the longer the milk stands. This phenomenon, observed by Dixon,4 described by Wieland 30 and studied by Polonovski, Neuzil, Baudu 15-19 and by q 2 6 - 2 7 seems to be a consequence of a change in the physico- chemical state of X.O. produced by a variety of chemical and physical factors. In the present paper we give a brief description of our experiments on this topic and its interpretation in terms of structural organization. EXPERIMENTAL MATERLQLS.-Fresh miIk was directly milked into a Dewar and brought to the labora- tory ; its temperature varied slightly around 35" C.Purification of X.O. was achieved by different methods. For the high pressure and ultrasonic (U.S.) studies we used partially purified preparations obtained by a method described previously,26* 27 using organic solvents (ether + propanol) for the delipidation. In other experiments preparations according to Dixon and Kodama,4 to Ball,2 and to Green et aZ.12 were used, as indicated. A11 chemicals used were of analytical grade. Aldehydes were redistilled before use. Purines were supplied by Roche and riboflavine came from the Lederle laboratories. METHODS.-(i) Enzyme activity was measured by the Thunberg method using con- ventional tubes which had been evacuated for 10 min at the water pump and filled with nitrogen.Methylene Blue was placed in the stopcock, and the decoloration time (d.t.) measured at 37" C. For the experiments where rapid successive activity determinations were needed (e.g. kinetics of the thermoactivation), narrow (8 mm diam.) haemolysis tubes, sealed with a paraffin layer, were used. Using whole milk as the enzyme source this method gives reproducible results for decoloration times between 0.5 and 15 min.l9. * (ii) In other experiments, spectrophotometric assays were used (Kalckar 7) with 0 2 , Methylene Blue or triphenyl-tetrazolium chloride (TPT) as electron acceptors (Unicam quartz spectrophotometer). With the first, uric acid formation was followed at 290 or 300 mp; with the second, decoloration was measured at 675 mp.With TPT, formazan formation was followed at 500 mp. I t seemed interesting to compare reduction kinetics of 2 acceptors reduced simul- taneously by the X.O. + hypoxanthine (H.X.) system. Fig. 1 gives the results of a typical experiment. The 300mp plot shows that pro- duction of uric acid with 0 2 as acceptor is of zero order, at least over the first 10 min. If Methylene Blue is added, its rate of reduction measured at 675 mp is not constant but gets faster as the 0 2 is exhausted. The same is true for TPT, where reduction is preceded by a rather long time-lag followed by a linear rise of optical density at 500p.m. The affinity of this dye for the enzyme system is much smaller than that of Methylene Blue. If the X.O.+ H.X. system contains both dyes, Methylene Blue is preferentially reduced and the time lag for TPT reduction is prolonged (see fig. 2). * Enzyme activity is expressed as the reciprocal of decoloration time (10yd.t. min) in minutes per unit enzyme (ml of milk, mg of purified enzyme). In activation experi- ments the activation coefficient, (a.c.) gives the ratio : enzyme activity after activation/ enzyme activity before activation.56 ACTIVATION OF MILK XANTHINEOXIDASE Thus we used TPT only occasionally for qualitative demonstration of enzyme activity and Methylene Blue or 0 2 for quantitative studies. With the substrate/acceptor ratios used (0.05-0.25 pM Methylene Blue for 5 pM of H.X.), the reciprocal of the decoloration time is proportional to the initial rate of oxidation of H.X.Reproducibility of results is within f 10 % with Methylene Blue in presence of air: parallel determination of de- coloration time of 0-25 pM of Methylene Blue by 2 ml of milk (ha1 volume 3 ml, pH 8.2) yielded 19 a mean of 4-6 min with a standard deviation of 0.43. FIG. 1.-Kinetics of oxidation of hypoxanthine (5 pM) by purified X.O. (according to Green et aZ.12) and different electron acceptors. Abscisses : time in minutes ; ordinates : optical density. 1 ml of enzyme is mixed with 0.5 ml 0.1 M phosphate buffer. pH 7-2, 5 p M hypoxanthine, the ac- ceptor, and 2.8 ml dist. water. Acceptors : 675 mp plot 0-05 pM Methylene Blue; 500mp plot: 7 pM TPT; 300 mp plot : oxygen, 1 cm quartz cell, at apptox. 20" C. Reduction of Methylene Blue and of TPT in the presence of air.k J 1 U Minutes FIG. 2.-Effect of Methylene Blue on the reduction of TPT by X.O. The same condition as in fig. 1. 675 mp plot: reduction of Methylene Blue in the TPT + Methylene Blue mix- ture; 500 mp plot no. 1 : reduction of 7pM TPT in presence of air; 500 m p plot no. 2 : reduction of TPT in presence of air and 0.05 pM Methylene Blue. I5 10 2 0 Minutes (iii) In the Warburg experiments the enzyme and buffer were placed in the main corn- partment, and 0-2 ml 5 % KOH in the centre well ; the substrate is added after a 5-min equilibration period. As acetaldehyde distils into the KOH, its final concentration is uncertain.L . ROBERT AND J . POLONOVSKI 57 PHYSICAL TREATmNTS.-(i) Experiments with U.S. waves were performed with a S.C.A.M.type of instrument 15 having a quartz plate of 6-cm diam. connected to a 2 kW generator. The frequencies used varied between 176 and 960 kc/s, and the power emitted was approximately 2-9-3-9 W/cm2 (82-102 acoustic watts). The solution was placed in a glass cylinder having a Cellophane membrane at the bottom, and this was cooled during irradiation by tap water at about 15" C. (ii) The high pressure work was done using the presses constructed by J. and J. Basset? The enzyme was submitted at constant temperature (37" C) to pressures varying from 5,000 to 12,000 kg/cm2. Compressions and expansions were sufficiently slow as to avoid changing the temperature of the sample. The enzyme solution was placed in a cylindrical glass container, closed by non-vulcanized rubber tubing by which the pressure of the oil chamber was transmitted.(iii) High speed centrifugations were performed on a Spinco preparative ultracentri- fuge * at 0". RESULTS TEMPERATURE-ACTIVATION AND INACTIVATION OF MILK x.0. As was emphasized by Wieland and the late M. Polonovski and colleagues,ls-l7 the dehydrogenase activity of fresh milk is very low and is strongly activated by a brief ex- posure to low temperatures (< 15" C). If immediately after milking, fresh milk is centrifuged at 37" C, enzyme activity is entirely confined to the globules. If milk is refrigerated before separation of the fat globules, the separated liquid is rich in X.O. Fig. 3 shows manometric experiments with refrigerated and fresh whey and with refrigerated cream suspension separated by sentrifugation at 5,000 rev/30 min.The low residual activity of fresh (non-refrigerated) whey is due to small fat globules and lipo- sornes,13 it disappears after centrifugation at 25,000 revlmin (45,000 g). M inutc s FIG. 3.-X.O. activity of different fractions of fresh and refrigerated milk, as determined by the manometric technique at 37" C. Abscisses : time in minutes, ordinates : pl. oxygen absorbed per mg dry weight. (A) 0-0 whey of milk centrifuged after cooling 2 h to - 5" c. x - x whey of fresh milk ; centrifuged at 37" c..- -a the same after centri- fugation 45 min at 45,000 g, 0" C. (B) .-a cream of fresh milk refrigerated in phosphate buffer, x - x cream of refrigerated milk. 5 pM hypoxanthine and 0.1 mM phosphate buffer, pH 7-2 in a total volume of 3.2 ml.We have studied the kinetics of this activation to find out whether it is of a physical or a chemical nature. Fig. 4 gives the results of such experiments as a three-dimensional plot, showing the activation coefficient as a function of time and temperature in the first 10 min. Between 0 and 20" C, the activity (lOz/d.t. min per ml milk) rises to 3-5 times its initial value and remains at this level. Between 20 and 40"C, fresh milk retains its original low activity (A.C. = 1) (though not indefinitely) but a slow activation appears after a few hours or days, probably due to the action of bacterial or milk enzymes (lipases 15). * Thanks are due to Dr. Rebeyrotte for the ultracentrifugations.58 ACTIVATION OF MILK XANTHINEOXIDASE At high temperatures, activation and thermal inactivation of X.O.proceed simultaneously, giving rise to a curve with a maximum. Both degree and rate of activation and rate of inactivation depend on temperature. At 0" C and above 80" C activation seems to be nearly linear with time, but the exact kinetics cannot be obtained by this method. Between 0" and 20" C, and 50" and 80" C, activation curves are S-shaped. 7 6 Activation 5 A.C.=I A.C. Inactivation 4 3 2 I 0 FIG. 4.--Kinetics of thermal activation and inactivation of milk X.O. 5pM hypoxan- thine, 0.25 pM Methylene Blue, 2 ml milk, final volume 3 mI. Methylene Blue decolor- ation time determined in haemolysis-tubes in presence of air at 37" C. Table 1 gives the times of attainment of half maximal activity and maximal activity, as a function of temperature (half times are given because they are proportional to zero order and to fist order rate constants).These data enabled us to estimate apparent activation energies of the thermal activation process. TABLE MAXIMAL ACTIVATION COEFFICIENT (d.t.)o/(d.t.)t AND HALF-TIMES AND TIMES OF ITS ATTAINMENT (t+ AND tm& AS A FUNCTION OF TEMPERATURE (determined by Methylene Blue decoloration in haemolysis tubes with 2 ml milk, 5 p M hypoxanthine, 0.25 pM Methylene Blue ; final volume 3 ml, p H 8-2 at 37" C ) temp. "C - min . 0 14 17 50 60 70 78 86 97 t4 1 2 3.5 6.5 4.4 1 0.5 0.25 0.10 tmax 4 22 16 6 2 1 0.5 0.7 - activation coefficient max 5 4.9 3.4 2 4.5 5.1 4.5 4.4 4.6 Fig. 5 shows the Arrhenius plots (log l/t+ or log l/tmax against reciprocal of absolute temperature, where t+ and tmax are the times required to attain half of the maximal activity and maximal activity respectively.Satisfactory straight lines are obtained with both t+ and tmax between 50 and 100" C. For the high-temperature activation, the activation energy is about 21 kcal/mole. The activation of the enzyme is followed by its heat de- naturation above 60" C.L. ROBERT AND J . POLONOVSKI 59 Table 2 gives the graphically computed first-order constants of the heat inactivation. Fig. 5B shows the corresponding Arrhenius plot with a slope corresponding to an activation energy of about 40 kcal and a frequency factor, log P Z E 22. The enthalw, FUNCTION OF TEMPERATURE (same method as for table 1) TABLE 2.-FIRST-ORDER CONSTANTS FOR THERMAL INACTIVATION OF MILK x.0.AS A t "C k, Sec-1 x 104 70 11.65 78 38-8 86 85-5 97 332 entropy and free energy of activation of the heat denaturation are : AH* = 39.4 kcal/mole ; AS* = +405 cal/mole deg. ; AF* = 26-7 kcal/mole. The degree of activation of X.O. by chilling or heating varies with the substrate. It is more important with hypoxanthine than with the aldehydes. '*"----- 'h I . 5 1.0- 0 . 5 - O r a I '--. rmax L , \ 28 29 3 0 31 - I o4 T I I I 20 29 3 0 FIG. 5.-Arrhenius diagrams for the thermal activation and inactivation of milk X.O. Abscisses : reciprocal of absolute temperature ; ordinates : (A) log of reciprocal of the time necessary to attain the maximal and half-maximal activity. (B) log of the first-order rate constant of thermal inactivation of X.O.Rates determined as indicated in fig. 4. Table 3 gives rates of oxidation of H.X. and different aldehydes by fresh and activated milk-enzyme (1 h at - 10" C). Measurements were performed in the Warburg ap- paratus, using 1 ml milk, 1 ml phosphate buffer in a final volume of 3.2 ml at 37" C. It is seen that the thermal activation coefficient is near unity for the lower aldehydes and increases with molecular length and volume. TABLE 3.-hLATIVE RATES OF OXIDATION OF ALDEHYDES BY FRESH AND COLD-ACTIVATED MlLK; MANOMETRIC METHOD, 37" c, IN PRESENCE OF 0.1 mM PHOSPHATE BUFFER, FINAL VOLUME 3 . 2 m l rate of oxidation, pl.. 02/60 min per ml of milk fresh milk cold activated milk activation coefficient PH. acetaldehyde 0.5 86 71 0.83 7.1 propylaldehyde 0.5 105 94 0.89 7.1 butylaldehyde 0.5 87 146 1.68 6.3 heptylaldehyde 1.0 75 115 1-5 7.4 salicylaldehyde 0.5 18 31 1-72 6.9 hypoxanthine 0.005 46 152 3-3 8 0.010 42 102 2.4 7.8 substrate mM60 ACTIVATION OF MILK XANTHINEOXIDASE HIGH PRESSURES As the thermal activation experiments suggested a physical mechanism, it seemed interesting to submit milk X.O. to other physicochemical treatments.Activity against time curves with a maximum would substantiate this hypothesis. Mechanical agitation was shown to provoke a similar activation to that of chilling.l% 30 So we submitted milk to hydrostatic pressures varying from 5 to 12,000 kg/cm2. Variation of physico- chemical properties of proteins and enzymes submitted to high pressures (5-20,000 kg/cm2) has been intensively investigated by the Macheboeuf school.Barbu and Joly 3 found evidence for modification in the physical state of proteins above 4000 kg/cm2. Fig. 6 gives activation coefficients as a function of time at different pressures for fresh milk. As the pressure-chamber was thermostatted at 37" C, thermal activation was avoided. These curves are similar to those of fig. 4, with the exception that inactivation does not proceed by a first-order mechanism. At 7000 kg/cm2 this rise, if taking place, is followed so rapidly by inactivation that we could not measure it (after 2 min the Mi nutcs FIG. 6.-High pressure activation and inactivation of milk X.O. Abscisses: time of application of pressure ; ordinates : activation coefficient. Rate of hypoxanthine de- hydrogenation determined by the Thunberg method, 2 ml of milk, 01 mM phosphate buffer, pH 7.4, 5 pM, hypoxanthine 0.1 pM ; Methylene Blue, final volume 3.6 ml ; enzyme is practically inactivated).Similar results can be obtained with acetaldehyde as substrate,27 but the ratio of activities with both substrates decoloration time determined at 37" C. 1 t.d. min. (acetaldehyde) [ t.d. min. (hypoxanthine) varies with time and pressure (from 1.0 to 22). With washed fat globules and with purified enzyme (by the organic solvent method 26) there is no activation, enzyme activity disappearing at 7000 kg/cm2 in 30 min. Purified enzyme and fat globules show about the same sensitivity to pressure (table 4). ENZYME AFTER 30 MIN AT DIFFERENT PRESSURES (ENZYME ACTIVITY ESTIMATED BY THE THUNBERG METHOD: 2 ml ENZYME SOLN., 2 ml PHOSPHATE BUFFER (0.1 M, pH 7.4) TABLE 4.-INACTIVATION OF x.0.IN WASHED FAT GLOBULES AND IN THE PURIFIED HYPOXANTHINE 5pM, METHYLENE BLUE, 0.1 pM, FINAL VOLUME 4 . 6 d) % activity remaining after 30 min pressure kg/cm2 _. fat globules purified enzyme 5000 100 88 6000 25 28 0 7000 8000 0 - -L. ROBERT AND J . POLONOVSKI 61 ULTRASONIC WAVES Effect of U.S. waves on fresh milk is shown in fig. 7, with both H.X. and acetaldehyde as substrate. Activation and inactivation follow a similar course as that at high pressure, and are more important with hypoxanthine as substrate, than with acetaldehyde. Activation by U.S. waves was sometimes observed even with washed fat globules (d.t. of Methylene Blue diminished in the ratio 3 : l), or with partially purified enzymes (organic solvent method).The preparations obtained by the use of propanol or butanol were only partially soluble in the phosphate buffer media. U.S. waves by their depolymerizing and dispersing A \\ocetald. ‘1 2 0 4 0 6 0 8 0 100 B I Minutes FIG. 7.-Activation and inactivation of milk X.O. by ultrasonic waves. Abscisses : time of irradiation ; ordinates : (A) activation coefficient, (B) reciprocal of decoloration time. Dehydrogenase activity determined in Thunberg tubes at 37” C with 2 ml milk or enzyme solution and 0.2 mM phosphate buffer, pH 7.4 ; 5 pM hypoxanthine acetal- dehyde 0.68 mM, 0-1 pM Methylene Blue ; final volume 4.6 ml. (A) fresh milk ; (B) fat globule suspension (about 10 % w/v) and partially purified enzyme (organic solvent method 26) ; 8.9 mg/ml) ; final pH 7.1.U.S. waves of 960 kc/s, 100 W, 20” C. FIG. 8.-Oxidation of riboflavine by ultrasonic waves. 9.7 x 10-6 M solution in 0.075 M borate buffer, pH 8-42. U.S. waves of 960 kc/s, 130 W, refrigerated in ice. (A) U.-v. and visible spectrum of riboflavine as a function of time of irradiation. (B) log of ratio of the 222 and 267 mp bands against time. effect greatly solubilized the preparation.6 This caused the apparent activation, which was followed by a rapid inactivation of the fat globule suspension and the purified preparation.26 This inactivation of enzyme is approximately fist-order, in the first 10 to 20 min. K, varies between 7.7 and 19.1 x 10-4 sec-1 for the purified preparation and is about 24 to 42 x 10-4 for fat globules.62 ACTIVATION OF MILK XANTHINEOXIDASE Prudhomme23 showed that the effect of U.S.waves is diminished by increasing the protein concentration and ionic strength. The mechanism of U.S. inactivation may be due to at least two factors : mechanical degradation and free-radical reactions induced by cavitation. It is difEcult to decide upon the exact role played by these two factors in X.O. inactivation. An approximate idea about free radical participation in the in- activation mechanism may be obtained by comparing enzyme inactivation to coenzyme destruction. We carried out some model experiments with riboflavine, part of the X.O. coenzyme. U.V. and visible spectra of riboflavine solutions were obtained as a function of irradiation time. Fig. 8 gives the results of a typical experiment.The intensity of all but one band decreases with time of irradiation, The intensity of the 222 mp band increases and the log of the ratio of the two absorption coefficients at 222 and 267 mp (log €222/€267) increases with time according to a first-order mechanism, yielding a rate constant k, = 9.5 x 10-4 sec-1 (fig. 8). Comparison of these data suggests that the rate-determining step in enzyme inactiva- tion is not the same as in riboflavine oxidation. The low rate constant for the milk enzyme, may be explained by the protective effect of proteins. This is substantiated by the much higher rate of inactivation of X.O. in fat globule suspensions than in milk, and by protection of riboflavine by casein against U.S. oxidation: irradiation of ribo- flavine, 1-3 x 10-5 M in a 120 mg casein solution at pH 6.5, does not eliminate the fluorescence nor the yellow colour ; both disappear in the control solution without casein after 20 min (960 kc/s, 80 W).[A new substance appears in the irradiated riboflavine .. 40.0 mq/ m I M i n u t e r FIG. 9.-Activation of milk X.O. by lauryl sulphate. Manometric method, 1 ml milk (after 8 h at 15" C), 5pM hypoxanthine, 0.1 mM phosphate [buf€er, pH 7.2, final volume 3.2 ml, final pH 7.0. Lauryl sul- phate as indicated on the figure, 37" C. solution, which migrates very slowly (Rf- 001) in butanol + acetic acid + water (4 : 1 : 5 ) on Whatman no. 1 paper (ascending chromatography) ; it has a blue fluorescence on the paper and is not iden- tical with lumitlavine or lumichrome.28] After having been left a few hours, milk contains reductases of bacterial origin. These enzymes (as measured by the d.t.of Methylene Blue without any substrate added) are much more sensitive to U.S. waves and to high pressure than X.0.%27 and may be selectively eliminated from contaminated milk. ACTIVATION AND INACTIVATION OF MILK x.0. BY DETERGENTS Detergents activate the X.O. of fresh milk.15~ 17918 Fig. 9 gives the results of a typical experiment. Fresh milk was par- tially activated by standing 8 h at 15" C, then mixed in Warburg vessels with increas- ing amounts of lauryl sulphate, and the rate of H.X. oxidation determined. Acti- vation increases with concentration up to 8 mg/d of milk; even 40 mg/ml did not inhibit the enzyme. This experiment shows that detergents are more effective as activators than is refrigeration.If 20 ml of fresh milk (at 37" C) are mixed with 8 mg of lauryl sulphate, oxygen consump- tion is doubled (with 5 p M H.X. as sub- strate at pH 7.4). Non-ionic detergents such as Tweens have a similar effect. Fig. 10 gives activation coefiicients as a function of time with Tween 60 (as determined by the haemolysis tube method). Similar results are obtained with cationic detergents? The resistance of milk enzyme is relatively great even to this class of substances which have very strong denaturing action. 50 mg of alkyltrimethylammonh.un chloride (zephirol) per ml of a 5-times washed fat globule suspension-this avoids activation by zephirol- equivalent to 270mg of detergent per gram of dry weight, effects only 43 % inhibition; 1 mg/ml cream suspension (5.4 mg/g dry weight) gives 10 % inhibition.L .ROBERT AND J . POLONOVSKI 63 FIG. 10.-Activation of fresh A.C. milk X.O. by Tween 60 (mono- stearate of polyoxymethylene- s - sorbitane). Abscisses: time of incubation with the detergent at 37" C ; ordinates : activation 3 - coefficient. Enzyme activity de- termined by the haemolysis tube 2 - method, as in fig. 4. Final con- centration of Tween, 5 mg/ml. DISTRIBUTION OF X.O. IN MILK, WHEY AND FAT GLOBULE SUSPENSION BETWEEN THE PARTICULATE FRACTION AND THE SOLUTION Morton's interesting work 13 suggested the following experiments, the aim being to decide whether all the X.O. obtained from the fat globules in the activation process is contained in the microsomal or liposomal fraction or if it goes immediately in " true " solution.Earlier work on X.O. purification 12 and its recent work on crystallization 1 leaves little doubt that X.O. can be obtained in true solution free of microsomal particles. On the other hand, even 5-times washed fat globules (0.182 g/ml suspension of cream, * ~ , _ _ _ - _ _ _ _ - - - - - - - - - a / / ' 0 /' ,/ 5 I b TABLE DISTRIBUTION OF ENZYME ACTIVITY AFTER 45,OOOg CENTRIFUGATION (45 min at 0" C) (Enzyme activity measured by the manometric technique with 5 pM hypoxanthine, 0.1 mM of phosphate buffer at 37" C, pH 7.3 in final vol. of 3 ml). activity in p l . 0 2 absorbed per hour per mg dry weight refrigerated ~ ~ ~ r " y ; d separated s'?s- sus- milk pension in pension in sulphate H20 phosphate top layer (fat globules) 1.2 1-5 - 0 0.7 liquid layer 0.44 0.6 1.1 2.8 8.2 precipitate 0.16 0.56 1.2 24.0 17.0 precipitate in presence of 0.05 pM Methylene Blue 100 y- cytochrome C, 1 pM Moo3 0.55 - - - 0.9 * Refrigerated milk (2 h at - 5" C) centrifuged at 5000 rev/min for 30 min to separate the cream layer.t Fresh cream suspension, twice washed in water, suspended in water, refrigerated at 0" overnight, churned by shaking 15 min, centrifuged at 5000 rev/min at -7°C for 30 min, and the liquid separated ; the cream suspended in 01 M phosphate buffer, pH 7.3, left at - 2" C for 2 h, centrifuged and separated as before ; the top layer activity, in the 5th column refers to this 4-times extracted cream suspension. It can be seen that the distribution of enzyme activity is different for milk and fat globule suspension.A non-negligible fraction of total activity appears in solution (at a degree of dispersity not sedimented by the gravitational field used). Only in refrigerated cream suspension does the greatest part of activity in the liposomes appear. Lauryl sulphate increases enzyme activity in all 3 fractions. The relatively low figures in table 5 are due to the high lipid and protein content of milk. If this is related to protein64 ACTIVATION OF MILK XANTHINEOXIDASE concentration, we find 720-655 pl. 0 2 per hour per mg protein for the typical liposomal fractions, and 0.6 pl. 0 2 for the “ casein ” particles of separated milk (see Morton.13) So we can confirm all but one of Morton’s statements, viz., X.O.seems not to be wholly present in Iiposomes. There is a non-negligible fraction firmly attached to the fat globules and another fraction which passes directly into solution following activation by cold or other physicochemical means. DISCUSSION The related experiments together with earlier work from this laboratory l5 enable us to formulate an hypothesis about the mechanism of the activation process. X.O., in its natural state in milk, seems to be wholly attached to the fat globules by lipoprotein cenapses of varying strength (see scheme). The globule is pro- tected by a peripheral structure, permeable to short-chain molecules (no activation for aldehydes of 2-3 carbon units). - Activation I iposomes FIG. 11. On cooling, this protective layer is disrupted, a fraction of the low-melting lipids goes into solution (“ oiling off ”) and the enzyme is partially liberated as liposome complexes which dissociate to yield soluble enzyme.The relatively high energy barrier (20 kcal) of the activation process may be explained by the fission of the lipoprotein complex. The fraction of enzyme which cannot be detached from the fat globule proves the solid character of these cenapses, which could not be due only to adsorption. Activation by detergents substantiates the analogy between these cenapses and serum lipoproteins (Macheboeuf 11). We may assign a functional role to the lipoid constituent of the enzyme complex : during Thunberg experiments the reduced Methylene Blue penetrates the fat globules, the leucobasis being much more lipo-soluble than the oxidized dye.This lipo-penetration of the reduced acceptor may displace the oxidation-reduction equilibrium in favour of complete reduction.2ob21 Whether liposomes are pre- formed elements existing before activation in the fat globule or adsorbed on it cannot be answered yet. The dependence of the activation on the substrate may be explained by two alternative hypotheses : (i) a steric hindrance limited to molecules greater than C3-C4 aldehyde : (ii) a change in the structure of the active site during the activation process. In favour of the second hypothesis we note the experiments of Knobloch9 on the change in the relative rates of oxidation of acetaldehyde and hypoxan- thine 26.27 during pressure and U.S. activation. Knobloch finds evidence for a different mechanism of oxidation of aldehydes than of purines by milk X.O.Though Knobloch claims the existence of two enzymes, these experiments do not prove this. The probability of the existence of a different mechanism of oxidation for aldehydes may be taken into consideration as an alternative hypothesis. The reorganization of the structure of fat globules after refrigeration and compression is shown by the studies of King.8 This author demonstrated the transformation of fat globules into spherical shell-crystals after such treatment. On cooling and compressing fat molecules previously distributed at random undergo a radial orientation. This may be considered as the last stage of struc- tural change before the rupture of the globule, being moreover accompanied byL.ROBERT AND J . POLONOVSKI 65 the separation of “ butter oil ” (low melting lipids). The thermal activation process may interfere with calculations of yield in enzyme purification 4.19 and should not be neglected, in such experiments. It is quite reasonable to assume that milk phosphatase and lipase have physico- chemical structures similar to X.O. and that similar activation of this enzyme might also be expected.13~ 15 Comparative studies on the physical state of X.O. in the mammary gland may throw light on the mechanism of synthesis of this particular enzyme - lipoprotein complex. 1 Avis, Bergel and Bray, J . Chenz. SOC., 1955, 1100. ZBall, J. Biol. Chem., 1939, 128, 51. 3 Barbu and Joly, Faraday SOC. Discussions, 1953, 77. 4 Dixon and Kodama, Biochenz. J., 1936, 20, 1104. 5 Edwards and Ball, J. Biol. Chem., 1954, 209, 619. 6 Grabar and Prudhomme, Collogue siir les hauts polynibres du C.N.R.S. (Strassburg, 7 Kalckar, J. Biol. Chem., 1947, 167, 429. 8 King, The Netherlands Milk and Dairy, 1950, 4, 30. 9 Knobloch, Col. Czech. Chem. Comm., 1947, 12, 581. 1946), p. 145. 10 Linen, in Bamann and Myrb2ck’s Handbuch der Enzymforschung (Springer, 1944), 11 Macheboeuf, Etat des Lipides dans la Matibre Vivante (Hermann, 1936). 12 Mackler, Mahler and Green, J . Biol. Chem., 1954, 210, 149. 13 Morton, Biochem. J., 1954, 57, 231. 14 Nygaard, Dianzani and Baler, Expt. Cell. Res., 1954, 6, 453. 15 Polonovski, M., Baudu and Neuzil, Le Lait, 1949, 29, 1. 16 Polonovski, M., Baudu and Neuzil, Bull. SOC. Clzim. Biol., 1947, 29, 958. 17 Polonovski, M., Neuzil, Baudu and Polonovski, J., C.R. SOC. Biol., 1947, 141, 460. 18 Polonovski, M., Baudu, Robert, M., Robert, L., Bull. SOC. Chim. Biol., 1950,32, 855. 19 Polonovski, M., Robert, M. and Robert, L., Bull. SOC. Chim. Biol., 1950, 32, 862. 20 Polonovski, M. and Robert, L., BUN. SOC. Chim. Biol., 1951, 33, 1139. 21 Polonovski, J., Exposb Ann. Bioch. Med., 1955, 17. 22 Polonovski, J., Ann. Chim., 1950, 671. 23 Prudhomme, Thise de Science (Paris, 1954). 24 Richert, Vanderlinde and Westerfeld, J . Biol. Chem., 1950, 186, 261. 25 Richert and Westerfeld, J. Biol. Chem., 1954, 209, 179. 26 Robert, L. and Nolla, Bull. SOC. Clzim. Biol., 1953, 35, 1363. 27 Robert, L. and Basset, B d . SOC. Chim. Biol., 1953, 35, 1375. 28 Robert, L. and Kallos, to be published. 29Tyler, J. Biol. Chem., 1954, 209, 893. 30 Wieland and Macrae, Liebigs Ann., 1930, 483, 217. vol. 3, p. 2347

ISSN:0366-9033    

DOI:10.1039/DF9552000054

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

L. ROBERT AND J . POLONOVSKI 65 GENERAL DISCUSSION Dr. D. B. Wetlaufer (Copenhagen) (communicated) : In reference to the paper by Massey, Harrington and Hartley, since they find that chymotrypsinogen labelled with the fluorescent dye can be activated to chymotryptic activity, and the reaction between chymotrypsin and the dye chloride leads to an inactivation, it would appear that a re-location of the dye molecule occurs in the activation of dye- chymotrypsinogen. This raises two questions that can be answered experimentally : (i) does activated dye-chymotrypsinogen have the same activity (as chymotrypsin) toward substrates that are strongly bound to native chymotrypsin; and (ii) does the function (ph20/7-0) change as a result of the activation of dye-chymotrypsinogen. These experiments should help show whether there actually has been a re-location of the dye during the enzymatic activation of this dye-labelled zymogen.I should also like to comment on the low ionic-strength polymerization of chymotrypsin (fig. 3 and 4 of their paper). The persistence of about 10 % monomer C66 GENERAL DISCUSSION even at the highest protein concentrations seems quite inconsistent in a system undergoing polymerization to a tetramer on higher polymer. These results might be explained if one postulated the presence of two species of chymotrypsin, one of which does not polymerize under their conditions. Do they have any information on the homogeneity of their chymotrypsin preparations which bears on this point ? Prof. H. Neurath (University of Washington) said: While it is interesting to note that under the special conditions of the experiments, Dr.Massey and co- workers were able to demonstrate what appears to be a polymerization of chymo- trypsinogen, I should like to raise the question whether in solutions of 0-03 ionic strength primary charge effects have been sufficiently suppressed to render the numerical values of the sedimentation constants meaningful. One might expect a retardation of sedimentation rates, and hence, sedimentation constants lower than those observed in the absence of charge effects. Actually, the authors report a sedimentation constant of 2.9 S for the monomer of chymotrypsin, as compared to 2-5-2-6 found by others at 0.1 ionic strength. The question arises, therefore, whether what the present authors consider to be the monomer is actually a dimer and whether the component with a sedimentation constant of 8-2 S does not cor- respond to a higher degree of polymerization than has been suggested in the paper.Dr. Hartley (Cambridge) said: In reply to Dr. Neurath’s suggestion that inhibition of enzymic activity in fluorescent conjugates of ChTr might be due to denaturation of the enzyme, we think that this does not occur for the following reasons. In the first place the mild conditions used in preparing the conjugates cause no inactivation of the enzyme. The conjugates have exactly the same sedimentation properties as the native enzyme. Dye-labelled chymotrypsin can be reactivated by incubation at pH 9.7, which causes the dye to hydrolyze from the protein, and finally a fully-active dye-labelled chymotrypsin can be prepared by trypsin activation of labelled chymotrypsinogen.Dr. V. Massey (Cambridge) said : In connection with the suggestion by Dr. Neurath and Dr. Dreyer that the N-terminal isdencine might be important in the catalytic activity of chymotrypsin, I would like to point out that we have found that isolencine does not react with FDNB when the enzyme is in the native state. This presumably means that the N-terminal isdencine is sterically blocked in native chymotrypsin, and it is therefore unlikely that it can be concerned in the catalytic activity of the enzyme. Concerning the argument presented in the Discussion on our paper that chymo- trypsinogen and DIP-chymotrypsin contain a chemically reactive grouping which is identical with part of the active centre of chymotrypsin, fig.10, which shows the degree of inhibition against the extent of reaction of the dye, clearly shows that the first molecule in the enzyme with which the dye reacts also destroys the enzymic activity. Fig. 8 shows that the ratio ph20/70 increases markedly when more than one molecule of dye has reacted per molecule of enzyme. As no changes in molecular size of the enzyme could be detected over the degrees of labelling studied, we have interpreted this change to a decrease in TO when more than one molecule of dye has reacted per molecule of enzyme. A change in PO would be expected if the first group labelled had different fluorescent character- istics from the groups subsequently labelled.Now it will be noticed from fig. 8 that the same changes of ph&0 also occur with chymotrypsinogen and DIP-chymotrypsin. Hence it seems very likely that the first group to react with the dye is the same in these three proteins, and as it has been found that the first group to react with the dye in chymotrypsin is a group necessary for activity, then it would follow that this active-centre group is also free to react in chymotrypsinogen and DIP-chymotrypsin. Now this argument rests upon the assumption that the low ph&0 values found at low degrees of labelling is due to reaction of the dye with the active centre. That this is in fact so has beenGENERAL DISCUSSION 67 shown by reacting with dye in the presence of the competitive inhibitor, sodium 8-phenylpropionate.Under these conditions labelled enzyme is obtained which has very much greater activity than a similar preparation labelled in the absence of any inhibitor. Such labelled chymotrypsin derivatives have been prepared at various degrees of labelling below 1 mole per mole, and found to have ph20/70 values of 14-15, compared to the normal value of 7. Thus there is little doubt that the specific reaction of the dye with part of the active centre is responsible for the low ph20/70 values found at low degrees of labelling, and that this active centre grouping is also free to react in chymotrypsinogen and DIP-chymotrypsin. As Vaslow and Doherty have shown that chymotrypsinogen will also bind one molecule of a substrate with much the same affinity as it is bound by chymo- trypsin, it is quite possible that this group reacting specifically with the dye is the substrate-combining site.From a comparison of the stability properties of the labelled chymotrypsin with various model systems we believe that this site is prob- ably an imidazole group, although we have as yet no conclusive evidence that this is the case. However, in support of this possibility we find that one, and only one, of the two histidine residues in these three proteins is free to react with fluorodinitro benzene. Against this concept of a common active centre site in the active and inactive proteins is the following work. Labelled chymotrypsinogen, when treated with trypsin, gives active labelled chymotrypsin. This evidence seems to be in direct contradiction to that presented above, as the most obvious interpretation is that the dye was not on a group in chymotrypsinogen which is essential to enzymic activity.We do not want to minimize the seriousness of this contradiction, but as usual several explanations are possible, of which we would like to mention two. (i) During the activation process the dye may have been transferred from the histidine (or whatever the group may be) to some other group not essential for catalytic activity. (ii) Chymotrypsin and chymotrypsinogen each possess one free imidazole group, but it is possible that it is not the same one in each protein. In other words, while chymotrypsinogen might possess a free imidazole group which would react preferentially with the dye and confer special fluorescent characteristics to it, it may not be the same imidazole residue which is necessary for catalytic activity in chymotrypsin.Further information about the substrate-combining site has been obtained by reaction with fluorodinitrobenzene under mild conditions. This work was re- ported at the Third International Congress of Biochemistry, but we would just like to emphasize one point here, as it is undoubtedly important in our under- standing of the mode of action of this enzyme. When chymotrypsin was reacted with fluorodinitrobenzene under mild conditions it was shown that only the K, increased with increasing time of reaction but that the maximum velocity remained unchanged. Thus one can say that none of the FDNB-reacting groups (imidazole, phenolic, €-amino groups and cc-amino groups) contribute to the activation process, but that one or more of them is concerned in the formation of the enzyme-substrate complex.If the group concerned here is an imidazole group, as we believe from our fluorescent studies, one might expect an effect of pH on K , in the pH range 5-8. According to Gutfreund and to Laidler, K, is independent of pH in this range. Although we have made no detailed study we must mention in this connection that we have found the K of j3-phenyl propionate to be very different at pH 7.2 and pH 7.9. Dr. Hartley (Cambridge) said : I should like to confirm Dr. Jandorff's observa- tion that imidazole rings catalyze the hydrolysis of p-nitrophenyl-acetate, which is a substrate for chymotrypsin. We have demonstrated that ct-benzoyl histidine methyl ester-which is a reasonable analogue of a histidine residue in a peptide chain-will catalyze the hydrolysis of this compound quite appreciably.We believe that some earlier observations on the catalytic properties of other proteins68 GENERAL DISCUSSION such as insulin for this compound are due to the histidine residues which they contain. Dr. G. A. Gilbert (Birmingham) said: The papers by Massey, Harrington and Hartley,l and Neurath and Dreyer 2 provide further data on the sedimentation of associating molecules, and also draw attention to the absence of a quantitative theory for the sedimentation of complexes in a state of reversible equilibrium. I should like to make some suggestions towards the development of such a theory, based on an analogy between the sedimentation (or electrophoresis) of associating molecules and chromatography.I pointed out this analogy in a previous Dis- cussion of the Faraday Society,3 without being aware that the theory of chromato- graphy predicts that when, as happens only rarely, the adsorption of a substance depends to a higher power than the first on its concentration in solution, a diffuse instead of a sharp leading boundary is formed. To make the analogy clear, we may consider the sedimentation of a monomer, M, in instantaneous equilibrium with a polymer, P, composed of n molecules of M. If the centrifuge cell is imagined to be given a velocity equal and opposite to the sedimentation velocity of P, P is brought to rest and is analogous to the adsorbed species in chromatography, while M streams past P and takes the place of the free solute carried along by the developing solvent.Since P is proportional to Mn, with II > 1, the boundary is diffuse. Its shape and mobility are given (in the idealized case when diflision is neglected and no account is taken of other factors such as cell shape, etc., dis- cussed in the paper by Baldwin, Gosting, Williams and Alberty 4), by the equations for chromatography deduced by De Vault 5 (see also Wilson 6 and Weiss 7). A brief indication of the method of treatment is given below for the general case of a polymer P sedimenting in the presence of its monomer M. Comparison is then made with the experimental results in ref. (1) and (2) for chymotrypsin obtained under two sets of conditions, (i) when P is a hexamer, and (ii) when P is a dimer. No new principle is involved when a group of complexes (each of different n) is present, but the algebra is much more complicated.It is convenient to imagine M halted by giving the system an equal and opposite velocity to M. Let P then sediment with a velocity v relative to the stationary monomer and consider a point at a distance x from the initial sharp boundary, after a time of sedimentation t. Then, within the diffuse boundary, the discussion of De Vault 5 shows that where the concentrations M and P respectively. eqn. (1) can be integrated conditions corresponding t = 0, If the following partial differential equation holds by weight of monomer and polymer are expressed by M"= KP, (2) within the region of the boundary to give for boundary to an initially sharp boundary at the meniscus at time If x/vt is replaced by the parameter 8 we obtain 1 Massey, Harrington and Hartley, this Discussion.2 Neurath and Dreyer, this Discussion. 3 Gilbert, Discussions Faraday SOC., 1953, 13, 159. 4 Baldwin, Gosting, Williams and Alberty, this Discussion. 5 De Vault, J. Amer. Chem. SOC., 1943, 65, 532. 6 Wilson, J. Amer. Chem. SOC., 1940, 62, 1583. 7 Weiss, J . Chern. SOC., 1943, 297.GENERAL DISCUSSION On substituting for M from (2) it also follows that 1 ~ giving by addition 69 (5) This equation describes the variation of the concentration by weight (P + M ) of the substance through the boundary. However, the observed Schlieren pattern in the cell is a measure of the gradient of (P + M ) with respect to x.This is obtained by differentiating (6) to give Thus, if diffusion could be neglected, and the individual sedimentation constants of the polymer and monomer were independent of concentration, etc., the Schlieren pattern would have the shape given by this equation. Although these ideal con- ditions are very far from being realized, it is interesting to draw the patterns for different values of n and compare them with actual cases. Since the above treat- ment was the result of an attempt to explain fig. 4 in the paper of Massey, Harrington and Hartley,l T will take as the first case polymer P of sedimentation constant 8.2 Svedberg units in equilibrium with a monomer M of sedimentation constant 2.5 units corresponding to their values for cc-chymotrypsin in 0.01 M phosphate, pH 7.9.A reasonable value for n is then 6. Inspection and differentiation of eqn. (7) shows that if n is 2, the Schlieren pattern consists of a single peak. For values of n greater than 2, it is a duzrble peak n - 2 with the valley lying at 6 = ____ 3(n - 1)' For n = 6 the minimum will occur at 6 = 0.267. Only one further item of data is needed to solve eqn. (6) and (7). It is provided by the observation in fig. 3 of ref. (1) that the two peaks of the pattern are equal in area when (P + M ) is 3.5 mg/ml (corresponding to 50 % polymer according to ref. (1)). The area of the first, slower peak, which terminates at 8 = 0.267, is therefore equivalent to 1-75 mg/ml, and this fact enables K to be cal- culated from eqn.(6), leaving no further unknown quantities. Using equation (7) and this data I have drawn in fig. la the " ideal " I I tb, O i m e r I t Monomer Hexamer i C I Monomer I 1 1 1 1 1 I ! 2 3 4 S b 4 8 S (Svedberq u n i t s ) (common scale) I FIG. 1 .-Idealized Schlieren patterns for sedimenting complexes. Schfieren pattern for a protein concentration (P + M ) of 3.5 mg/ml. The sharp peaks in the pattern would, of course, be rounded off in practice by diffusion, 1 Massey, Harrington and Hartley, this Discussion.70 GENERAL DISCUSSION and the pattern would then be more comparable to the experimental pattern in fig. 2a of ref. (1). It is a simple matter to transform the abscissa of the pattern in fig. 1 from 8 into equivalent values of sedimentation constant to give the pattern that would result from unit field acting for unit time.An alternative scale in Svedberg units is therefore added. The slow peak (8 = 0 to 0.267) has a constant area for all protein concentrations above 1.75 mg/ml, and it is only for concentrations above this value that the fast Protein concentration (mq/ml) FIG. 2.-Sedimentation of hexamer-monomer equilibrium mixture. Theoretical ratio of peak is seen at all. Thus the ratio of the area of the fast peak to the total area isIgiven by ((P + M) - 1*75)/(P + M). This ratio is plotted as a function of area of fast peak to total area. P r o t e i n c on c e n t r a t I on (m q/m I ) FIG. 3.-Theoretical variation of sedimentation velocity with concentration for hexamer-mono- mer equilibrium mixture.protein concentration in fig. 2. Comparison should be made with fig. 3 of ref. (I), which gives the experimentally found ratio, ex- pressed as " % polymer ". If the sedimentation velocity of a peak is taken as measured by the velocity of the vertical bisector of the area of the peak, the variation of this velocity with concentration can be calculated using eqn. (6). In this way the velocity of the slow peak is found to be practically in- dependent of concentration, and approximately equal to that of monomer, whereas the velocity of the fast peak rises very steeply with concentration. The calculated sedimentation constants are shown in fig. 3, which can be seen to re- semble fig. 4 of ref. (1) even to the extent of close numerical agree- ment.The somewhat higher ex- perimental value S = 2-9 for the slow peak may be the result of the simultaneous presence of dimer. The ratio P/M can be calculated and it is of interest that whereas the immediate intuitive interpretation of the Schlieren pattern when it has two peaks of equal area is that about equal weights of polymer and monomer are present in solution,GENERAL DISCUSSION 71 calculation shows that there is then actually 2.3 times as much monomer by weight as hexamer. Equal weights are not present until the fast peak has 2.3 times the area of the slow peak. As a second example, I will consider the dimerization of chymotrypsin,l accepting values of 2.5 and 3.5 Svedberg units for the respective sedimentation constants, SM and SD, of monomer and dimer.Putting n = 2 in eqn. (7), I have drawn in fig. l b the idealized Schlieren pattern for a solution containing equal weights of monomer and dimer. The peak is asymmetrical with a trailing edge towards the meniscus (see ref. (2) and (8)), and this asymmetry would only be greatly reduced, not eliminated, by taking into account diffusion. The peak for this pattern is bisected by the vertical line that corresponds to a sedimentation constant of 3.05 units. Schwert 1 has shown that the equilibrium constant of dimerization depends upon pH and ionic strength. In the specific case of DIP + 6-chymotrypsin at pH 7.5, ionic strength 0.1, it appears from fig. 7 of ref. (2) that the sedimentation constant has this particular value of 3.05 units for a protein concentration of 3.5 mg/ml.K in eqn. (2) is therefore (3.5) mg/ml. Using this value in equation (6) I have calculated the theoretical variation of the sedimentation constant with protein concentration and plotted the result in fig. 4. A close resemblance to the corresponding experimental curve in fig. 7 of ref. (2) will be noted. Similar experimental curves in fig. 5 and 7 of ref. (1) show an initial rise of the expected kind, but a subsequent decrease, due presumably to a dependence of SD on concentration not taken into account in this treatment. 3 5 1 P r o t e i n concentration (mq/ m I ) FIG. 4.Theoretical variation of sedimentation constant with concentration for dimer- monomer equilibrium mixture. Curve (a) K = 1-75 mg/ml, curve (6) K = 100 mg/ml. Even a very weak tendency to dimerize may have a measurable effect on the sedimentation constant.To illustrate this, I have drawn as a dotted line in fig. 4 the variation of S with concentration for a protein of dimerization constant 100 mg/ml and SM and SD as for chymotrypsin. Thus in many cases where S falls less rapidly with concentration than the increase in viscosity would lead one to expect, dimerization may be accelerating the sedimentation. To conclude, it is obvious that the above treatment remains unreal as long as diffusion is neglected, but it still seems to account for the main features of the sedimentation of complexes. Prof. H. Neurath (University of Washington) said : Let me state at the outset that I strongly believe in experimental facts; and since there is no adequate theory for the sedimentation behaviour of macro-molecules at finite concentrations, I believe that we should be guided primarily by our experimental observations, The sedimentation patterns shown in the accompanying slide 2 indicate that the sedimenting peaks are skewed whenever the sedimentation rate-concentration relation suggests reversible polymerization ; and this skewness is absent whenever we are dealing with what appears to be a single molecular species.Thus, it may be noted that with DIP-6-chymotrypsin, at pH 3.86, where it does not dimerize, the sedimenting boundary is symmetrical, whereas at pH 7.5 where dimerization 1 Schwert, J. Biol. Chem., 1949, 179, 655. 2 Dreyer, Wade and Neurath, Arch. Biochim. Biophys., 1955, 59, 145.72 GENERAL DISCUSSION occurs, the boundary is skewed.When sedimentation at pH 7.5 is carried out in a synthetic boundary cell, no asymmetry is discernible in the synthetic boundary where protein concentration is relatively high on both sides of the boundary, and is in a region where the slope of the concentration dependent curve is relatively small; in contrast, the boundary formed between the protein solution and the pure buffer is slightly asymmetric and would probably have been more so had the experiment been extended for a longer period of time. I was most interested to hear from Dr. Hartley that under mild conditions, the x-amino group of isoleucine in cr-chymotrypsin does not react with FDNB. However, this does not preclude the involvement of the hydrocarbon side-chains of the isoleucyl-valine sequence in the configuration of the active centre; and I believe that more direct chemical evidence is required before the idea of an involve- ment of this sequence in the configuration of the active site can be discounted.As indicated in the introduction to our paper, I am inclined to ascribe more significance to the structural change in the molecule which occurs upon activation of chymotrypsinogen, as evidenced by the decrease in laevo-rotation and by the fact that the rate of change of optical rotation goes parallel to the rate of ap- pearance of enzymatic activity. Since no significant change in either rotation or activity occurs during the conversion of n-chymotrypsin to the 6 form, it is clear that the structural changes towards a more nearly helical configuration must have occurred incidental to the cleavage of the first peptide bond.1 Referring to Dr.Hartley’s comments on his own paper I am glad to note that, following removal of the dye, the enzyme regained activity ; I am sure that he, too, would feel happier about this evidence if reactivation were complete or nearly so. Prof. Felix Haurowitz (Indiana University) said : Is the protein component of catalase essential in Schulman’s experiments ? Would not protein-free haemin, which has weak catalase activity, give the same results? It is well known that the lipoxidase activity of haemin is the same as that of haemoglobin and that both haemin and haemoglobin act as lipoxidases in heterogeneous emulsions of linseed oil or of linoleic acid, but not in the clear solutions of these lipids in organic solvents or in ox bile.? If such clear solutions are converted into emulsions by the addition of water, haemin begins to act again as a lipoxidase.Prof. J. G. Kaplan (Halifax N.S.) (cornmiinicated) : The biological implications of our paper have been made somewhat less clear-cut as a result of some of our recent work on the induced biosynthesis of yeast catalase (a phenomenon first reported by Chantrenne). Our data now show that this enzyme can exist in no less than 3 discrete “states”’, the first 2 of which having been previously described by Fraser and myself. These “ states ” are : (i) that of the enzyme in the normal, aerobic cell, in which there is a high catalase content but in which one can demonstrate in the intact cell only a small fraction of the total activity, the latter being characterized by a comparatively high activation energy (approx.8 kcal/mole) ; (ii) that of the enzyme within the lysed cell, or in aqueous solution, in which the enzyme has undergone alteration, shown by an increase in activity of approx. 17-fold, and a decrease in activation energy (to approx. 3 kcal/ mole), as well as by change in other enzymatic properties; (iii) that of the enzyme within the anaerobic cell, exhibiting a low, basal level of catalase activity which cannot be increased by lysis of the cell, the enzyme nevertheless being characterized by the same high ,u as the catalase in the aerobic cell. State (iii) in the “ unadapted ’’ cell, is intermediate between states (i) and (ii), since the enzyme possesses the high p and other enzymatic properties of state (i), and the full expression of activity of state (ii); although lysis changes but little 1 RupIey, Dreyer and Neurath, Biochiin.Biophys. Acta, 1955, 18, 162. 2 Haurowitz and Schwerin, Enzymologia, 1940, 9, 193.GENERAL DISCUSSION 73 the specific activity of the suspension, it nevertheless causes enzyme alteration, shown by the drop in p values to the level characteristic of state (ii). Work done at Yale in collaboration with Dr. D. M. Bonner shows that these same 3 states are demonstrable in the case of the ,f3-galactosidase of E. coli. Here, state (i) is that of the enzyme in the fully adapted cell (i.e., grown in the presence of lactose) in which only a fraction of total activity is expressed, state (ii) is that following lysis of the cell or extraction of the enzyme with full expression of activity, and state (iii) is that in the unadapted cell, where the basal level of activity per cell is but little affected by lysis or extraction (Rickenberg and Bonner) but in which these procedures cause the p to change from the high level of state (i) (approx.16 kcal/mole) to the low level of state (ii) (approx. 9 kcal/mole). These findings suggest, first, that the interface at which catalase is adsorbed within the cell may be the surface at which it is synthesized, since the activation energy is high (and unchanged) before, during, and after the process of enzyme biosynthesis; additional weight is lent this hypothesis by the evidence from our ultra-violet studies (Kaplan and Paik) that catalase within adapted and unadapted yeast exists in a complex with ribonucleic acid. They suggest further that the most satisfactory type of model interfacial system may be the C18TAB n-lauryl alcohol-stabilized emulsion studied at Cambridge since it is the only one at which the crystalline catalase may be stably adsorbed with some increase in activation energy (from 4-1 to 6.9 kcal after adsorption) but without loss of activity (see table 4), thus constituting a reasonably accurate model of state (iii).(It might be noted that the magnitude of the ratio of activity of desorbed (or altered) enzyme to activity of adsorbed (or unaltered) enzyme-the " factor " of Rickenberg and Bonner and " activation coefficient " of Robert and Polonovski-will change somewhat depending on the temperature of assay, since the p of catalase in the 2 states differs; hence a factor of 1 (table 4) would be valid only at a single tem- perature of assay and would increase slightly at lower temperatures and decrease to a fractional value at higher temperatures.) If to this model interfacial system we could add some hypothetical agent which would block reversibly the activity of a high proportion of the adsorbed catalase molecules without affecting the temperature characteristics of the remaining active enzyme, we would have a work- ing model of state (i).Finally, if desorption of the enzyme were accompanied by release of the inhibition we would have a model of state (ii), and would thus be able to manipulate the model interfacial system as one can the intracellular enzyme.However, even if we never arrive at a perfect model, it is clear that many of the interfacial systems presented in our paper are nonetheless better imitations of the cell enzyme than is a simple aqueous solution of the extracted enzyme. Dr. L. Robert (Paris) (communicated) : The ingenious model experiments of Dr. Frazer, Kaplan and Schulman prove the importance of adsorption-desorp- tion equilibria in organized enzyme systems and parallels our findings on milk xanthine oxidase. It would be tempting to admit that a similar mechanism as proposed by the authors is responsible for the structural localization of X.O. in milk globules (adsorption to the lipid particles, microsomes, in multilayer, in the mammery gland) as well as its " activation " by physicochemical means (desorption of enzyme from the lipid surface).To be able to accept this mechanism, it would be necessary to prove that X.O. in the mammary gland is in " solution " and is incorporated in microsomes during fat globule synthesis, and that multilayer adsorption takes place. Another important feature of this mechanism would be, according to the comment on our paper by Dr. Frazer, a great difference be- tween enthalpies and entropies of activation of the " free " and " adsorbed " enzyme. The only data available are due to Prof. Sizer,l who found the same activation energy for the purified enzyme as for milk. (As milk is a mixture of " soluble " and " adsorbed " enzyme, this measurement needs to be repeated, 1 Enzymologia, 1940, 8, 75.74 GENERAL DISCUSSION with exhaustively washed fat globules, containing only the non-desorbable enzyme, with liposomes and with purified enzyme.) But this interpretation does not explain the irreversible structural changes occurring on " activation " : a certain number of fat globules is destroyed, the size distribution is changed, the average diameter being higher after activation (see ref.(15) of our paper), non- saturated fatty acids appear (" oiling off "), etc. So it seems more probable that the " activation " of X.O. is of a double nature : breakdown of the equilibrium in a binary or more complicated isomorphous mixture with liberation of X.O. from a lipoprotein complex, as was proposed by Neuzil et al.(see ref. in our paper), and perhaps an adsorption-desorption mechanism concerning the enzyme bound to microsomes. We agree entirely with Dr. Fraser, Kaplan and Schulman as to the interpreta- tion of the activation energy of the thermal activation process of X.O. (- 20 kcal, see our paper); it corresponds probably to the rupture of a lipoprotein complex with the transfer of the equivalent of about 280 sq. A hydrocarbon-like chains to the oillwater interface.1 It is interesting to note the relative resistance of X.O. in milk to denaturation by urea, which is probable due to complex formation between fatty acids and urea, diminishing greatly its effective concentration. Resistance to urea denatura- tion diminishes with purification of the enzyme. Dr.R. K. Morton (University of Melbourne) (communicated) : There are several points in the interesting paper by Dr. Robert and Dr. Polonovski on which I would like to comment. The origin of xanthine oxidase of milk and comparative studies of the physical state of the enzyme in mammary gland and milk, have been investigated in my laboratory? Comparison was made of the chemical composition and enzymic (xanthine oxidase and alkaline phosphatase) activities of fractions isolated from the mammary gland of the cow and from the milk of the same cow (obtained immediately before death). The results confirm my earlier suggestion that " milk microsomes "-the lipoprotein particles in cows' milk-are largely derived from microsomes from the secretory epithelian cells of the mammary gland.During incubation in the collecting ducts and cistern of the cow, however, the mammary gland microsomes undergo certain changes, notably: (i) a small but significant loss of lipid, (ii) a considerable loss of nucleic acid, and (iii) a marked increase in xanthine dehydrogenase activity. However, there is little or no change in the alkaline phosphatase activity. Hence, in answer to the question as to whether microsomes (" liposomes ") are performed elements of milk, the answer is certainly " yes ". Whereas the xanthine oxidase of fresh milk is almost entirely associated with the fat fraction as a microsome complex, in the mammary gland the enzyme appears to be distributed between the cytoplasmic particles and the supernatant remain- ing after centrifuging the mammary gland homogenates (in 0.25 M sucrose) at approximately 50,000 g for 2 h.Incubation of mammary gland microsomes at 37" for 10 h, either in milk serum or in phosphate buffer (PH 7.4) leads to an appreciable increase (two to four times) in xanthine oxidase activity. Such " activated " gland microscomes still have considerably less xanthine oxidase activity than do milk microsomes. The xanthine dehydrogenase activity of freshly-isolated milk microsomes may be equal to 5-15 % of that of the most highly purified xanthine oxidase preparations. In contrast to xanthine oxidase, alkaline phosphatase in the mammary gland is almost wholly associated with the microsome fraction, as it is in milk. As yet, no activation of this enzyme by physical means has been observed.1 see Eyring, Lumry and Spikes in McElroy and Glass, Mechanism of Enzyme Action 2 Baillie, M.Sc. Thesis (University of Melbourne, 1955). (Johns Hopkins Press, Baltimore, 1954), p. 129.GENERAL DISCUSSION 75 Dr. Tsoo E. King and Dr. V. H. Cheldelin (Oregon Stare CoZZ.) said: In connection with the report by Robert and Polonoski on the thermal activation and inactivation of xanthine oxidase, we would like to raise a question whether the inactivation or denaturation of enzymes follows a simple order of reaction. This question has been debated for a longtime, cf.,forexample, a review by Putman.1 From the practical view-point of enzymologists, this type of information is very useful. Recently we have studied properties of a new TPN-linked acetaldehyde dehydro- genase from Acetobacter suboxydans.This enzyme catalyzes the following reaction : 2 The inactivation or denaturation rate of the enzyme was determined at tem- peratures between 300 to 310" K. It was found that the rate constants changed with the time at any temperature determined. No simple order of reaction could be observed. The enzyme used in the investigation was a purified preparation with a k3 of 10 m o l e s acetaldehydelmin per g protein. Determinations were performed at an enzyme concentration of 0.0005 % in 0.1 % solution of crystalline bovine serum albumin. On the other hand, the activation energy of the oxidation was successfully determined by measuring initial zero order rates (with respect to either acetalde- hyde or TPN) at temperatures between 273 to 310" K.The activation energy was found to be 14.28 kcal/mole by the conventional Arrhenius plot. The standard error was 0.29 kcal and the confidence limit at a confidence coefficient of 0.95 was 13-53 to 14-96. This result indicates that our methodology was precise enough for such kind of work. However, the fact that rate constants of the thermal inactivation of this enzyme changed with time, suggests the very complex nature of its denaturation. Prof. J. G. Kaplan (Halifax, N.S.) (communicated): There are certain other parallels between the interesting work of Roberts et al. and our own, in addition to the increase in both X .O . and catalase activities following release of these enzymes from the surface of oil droplets. Pre-treatment with heat had been found by von Euler 3 to cause increased catalase activity of the yeast cell, a finding which we were able to confirm 4 but which Fraser and I have found to be a quite variable effect ; 5 the " maximal activation coefficient " was rarely more than one-fifth that obtained after treatment of cells with CHC13, surface-active agents, u.-v., etc., and heat does not alter the enzymatic properties the way these other agents do.I have found that freezing the yeast also caused the increased catalase activity which we have called the Euler effect. Treatment of cells with anionic detergents also caused the Euler effect. We believe the Euler effect follows, directly or indirectly, from desorption of the catalase from some intracellular interface of the oil/water type ; 6 the change in various enzymatic properties which accompanies the Euler effect we have called enzyme alteration.The activation energy for heat " activation " of X. 0. (or what we should call heat-induced alteration, if change in enzyme properties could be demonstrated) was 21 kcal. Fraser and I 5 on the basis of the observed 21 kcal difference between the p values for heat inactivation of altered and of unaltered intracellular yeast catalases, predicted that the process of alteration itself would have an energy of activation of about, but no less than, 21 kcal. The subsequent kinetic study of catalase alteration 7 closely confirmed this prediction, yielding values for butanol-, in Neurath and Bailey, The Proteins, vol. 1 , part B (Academic Press, N.Y., 1955) p. 807.CH3CHO + TPNf + H20 + CH3COOH + TPNH 4- H+. 2 King and Cheldelin, unpublished. 3 von Euler, Z. physiol. Chem., 1919, 105, 83. 4 Expt. Cell. Res., 1955, 0, 305. 5 J. Gen. Physiol., 1955, 38, 515. 6 Physiol. Zool., 1952, 25, 123. 7 Kaplan and Pa&, Can. J. Biochem. Physiol., 1956, in press.76 GENERAL DISCUSSION CHC13- and u.-v.-induced alteration within the narrow range of 20-24 kcal. Evidence that an interfacial process is involved in catalase alteration has been presented.1 The rather striking similarity between our values for catalase alter- ation and that of Robert and Polonovski for X.O. “ activation ” might indicate a common, or similar, rate-limiting step, especially since they conceive this to be a splitting of a lipo-protein complex, a process which might well be only semantically different from a desorption of protein from a lipid/water interface.In this con- nection, I should like to ask Dr. Robert whether his distinction between “ a physical or chemical nature ” of “ activation ” is a meaningful one, since the “ fission of the lipoprotein complex ” must represent rupture of some chemical bond, regard- less of whatever agent is used to cause it (see also Langmuir’s classical paper in this regard).;! I should like also to ask Dr. Robert whether he has studied the effect of tem- perature on the rate of the reaction of X.O. with its various substrates in order to obtain an activation energy for this process, and if so, whether this constant changes following desorption of the enzyme into the aqeuous phase? We should predict on the basis of our work on catalase and ,f3-galactosidase that the p would decrease significantly upon desorption from the oil droplets.May I ask further whether the low “ activation coefficient ” observed with the aliphatic aldehydes could not possibly be due to a desorption of the X.O. in the fresh, uncooled milk during assay of these surface-active substrates? We have shown that the aliphatic aldehydes cause alteration of yeast catalase; 3 this was seen to be related to their surface-activity, and the homologous alcohols of about equal surface-activity, caused desorption of crystalline catalase from the olive oil/cephalin/water interface. Finally, I think that the term “ activation ”, used in the sense of increased en- zyme activity following treatments of various kinds, is rather confusing.I would propose that terms such as “ activation coefficient ”, “ crypticity ”, etc., be replaced by the term “ factor ”, introduced by Bonner 4 and defined as the ratio of activities of maximally active enzyme to that of the enzyme in its original, or less active, state, whatever this may be. Dr. L. Robert (Paris) (communicated) : To the interesting comment by Prof. Kaplan I should like to reply that the distinction between activation of a “ physical ” or “ chemical ” nature concerns the activating agent rather than the process of activation itself. It would be tempting to admit that the nature of “ activation ” or “ alteration ” is the same with physical means (high pressure, heat, ultrasonic vibrations), as with chemical means (surface-active agents, etc.).On the other hand, I do not think that the “ fission ” of a lipoprotein complex must represent a rupture of some chemical bond; such complexes may well be held together by van der Waals’ forces.5 The interesting comment by Dr. Morton confirms the hypothesis that formation and disintegration of milk fat globules and of lipo- somes is very similar in nature to the adsorption-desorption phenomena observed by Prof. Kaplan, Prof. Fraser and Dr. Schulman at oil-water interfaces. We are now studying the activation energy of desorbed and of the highly purified enzyme (kindly provided by Dr. Bray, Prof. Bergel and Dr. Avis) so I cannot yet answer the second question of Prof. Kaplan concerning the experi- ments of Prof.Seizer. The low “ activation coefficient ” observed with aliphatic aldehydes could well be due to desorption of the enzyme by the aldehyde. Direct experimental proof is difficult to obtain, because these aldehydes inactivate the enzyme rather rapidly. 1 J. Gen. Physiol., 1954, 38, 197 ; also Kaplan and Paik.5 2 J. Amer. Chem. SOC., 1917, 39, 1848. 3 J . Gen. Physiol., 1954, 38, 197. 4 Amino Acid Symposium (Johns Hopkins U. Press, 1955). 5 see, for instance, the classical papers of Macheboeuf, in Hughes, Blood Cells and Plasma Proteins.GENERAL DISCUSSION 77 The term activation is in itself perhaps confusing, but it indicates at least that the alteration produced in the enzyme system resulted in a higher activity. Altera- tion is perhaps a more general term, but I think it is too vague. Mr. Polonowski called the phenomenon in French “ revelation ” which is as acceptable as the “ Euler effect ”, if a good definition is given of the reaction which is understood by this phrase. Adsorption and desorption is not acceptable, because for X.O. at least much more happens than is described by these terms. The “ activation coefficient ” is by definition the same as Dr. Bonner’s ‘‘ factor ”. Prof. D. D. Eley (Nottingharn) said : Dr. King has mentioned the change in reaction order observed during enzyme inactivation Coleman, Davies and I have examined this problem for acetylcholinesterase A plot of log activity against time gives two lines connected by an “ elbow ”, corresponding to a rapid initial reaction followed by a slower final process The initial reaction has a high entropy of activation which suggests that the decreasing activity at the active site is due to structural changes in the neighbouring parts of the protein molecule. The final slow process may well concern destruction of the active site itself, and in support of this view we have some evidence that this final process is slowed down in the presence of butyryl choline, which is known to be adsorbed on the active site from its behaviour as an inhibitor in the hydrolysis of acetyl choline by the enzyme. We have also evidence that the initial rapid phase is associated with a decrease in both enzyme-substrate affinity, l/&, and a somewhat less marked decrease in urnax.. It seems clear that this inactivation process is not a simple decrease in enzyme concentration with time, but involves different levels of enzyme activity in one and the same molecule.

ISSN:0366-9033    

DOI:10.1039/DF9552000065

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

11. KINETICS AND MECHANISMS THE INFLUENCE OF FLUCTUATIONS IN PROTEIN CHARGE AND CHARGE CONFIGURATION ON THE RATES OF ENZYMATIC REACTIONS* BY JOHN G. KIRKWOOD Sterling Chemistry Laboratory, Yale University, New Haven, Connecticut Received 6th May, 1955 The Kirkwood-Shumaker theory of fluctuation forces between protein molecules arising from fluctuations in protonic charge and charge configuration is applied to an analysis of the interaction of the protein moiety of an enzyme molecule and a substrate molecule locally bound to an active site. It is established that if there is an increase in dipole moment of the active site + substrate complex in its activation to the transition state, interaction with vicinal basic groups of the protein by the fluctuation mechanism can produce a substantial decrease in the free energy of activation.This interactron is maximal at a pH equal to the pK of the conjugate acids of the effective basic groups. The proposed mechanism accounts in a satisfactory manner for the dependence of the rates of many enzymatic reactions on pH without invoking the simultaneous participation of two or more ionizable groups. An enzyme is a protein possessing an active catalytic site or prosthetic group to which the substrate molecule first associates and subsequently reacts by a mechan- ism determined not only by its local environment but also by the protein moiety of the molecule. The essential role of the protein in enzymatic activity is clearly demonstrated by the catalytic impotence of isolated co-enzymes or prosthetic groups under conditions permitting their dissociation from the protein.For brevity in the following discussion, we shall refer to the protein moiety of the enzyme molecule as the protein and to the attached coenzyme or prosthetic group, whether dissociable or not, as the active site. Abstractly, WE may say that the protein affects the rate of an enzymatic reaction either by forces which groups vicinal to the active site exert cn the associated complex of site and substrate molecule or by serving as a reservoir of mobile ions, for example, hydrogen ions, which exert a local catalytic effect when transferred to the complex. The influence of the protein may manifest itself either in the free energy of binding of the sub- strate to the active site, resulting in a change of the Michaelis-Menten constant K , or in the free energy of activation of the complex, resulting in an increase in the intrinsic rate constant k3 of the enzymatic reaction, which we shall suppose to follow the typical rate law, (1) dF1 - k3LElLSl - _ - _ dt K , + IS]' where [El and [S] are the concentrations of enzyme and substrate, respectively.In principle, any cf the usual types of intermolecular force acting between vicinal parts of the protein and the active site + substrate complex, both in its activated and unactivated states, could produce changes in K , and k3. If the forces are primarily electrostatic, their effect should be sensitive to pH and ionic strength. One of the characteristic features of enzymatic reactions is the sensitivity of their * Contribution no.1299 from the Sterling Chemistry Laboratory, Yale University. 78J . G . KIRKWOOD 79 rates to pH. For most hydrolytic enzymes, though not all, the rate as a function of pH follows a bell-shaped curve, which descends on both sides of an optimum pH corresponding to maximum rate. This behaviour has been ascribed to the simultaneous participation of two types of basic groups, either in the complex or on the protein contiguous to the complex, in the reaction, the conjugate acids of one being uinionized and of the other ionized. We shall presently show that this hypothesis, although possibly correct, is not necessary to account for the effect of pH on the rate of enzymatic reactions. In an earlier article,l we have shown that the mobility of their charge distribu- tions gives rise to a supplementary attractive force between protein molecules which would not exist between static constellations of electric charge.Proteins, considered as ampholytes, contain a large number of neutral and negatively charged basic groups, such as -NH2 and COO-, to which protons are attached. Except in highly acid solutions, the number of basic sites generally exceeds the average number of protons bound to the molecule, so that there exist many possible configurations of the protons, differing little in free energy, among which fluctuations may occur as the result of thermal motion. Similar fluctuations may occur in the configurations of other ions bound to the proteins, when the number of binding sites exceeds the average number of such ions which are bound.Fluctuations in the number and configuration of the mobile ions impart fluctuating charges and fluctuating electric multipole moments to the molecule. The fluctu- ating electric field of each molecule of a protein pair induces fluctuations in the charge constellation of its neighbour in such a manner as to produce an attractive force between them. In a similar manner, an attractive force is established be- tween a protein and a small molecule, of fixed charge distribution. In this instance the field arising from the charge of the small molecule, if ionic, or from its dipole moment, if polar, induces fluctuations in the charge distribution of the protein leading to an average polarization of the macromolecuIe and an attractive force in excess of that due to the interaction of the average charge distribution with the small molecule.Such an attractive force will exist between an enzyme and its substrate, if the substrate is ionic or polar and there is a suitable constellation of basic groups vicinal to the active site to which the substrate molecule is locally bound. As we shall see, this interaction is maximal at a pH equal to the pK of the conjugate acid of the effective basic group. Before proceeding to an analysis of the fluctuation force between the protein and the active site + substrate compIex of an enzyme, we shall briefly describe the origin of this force. According to the general theory of statistical mechanics, the potential of average force W(R) between two molecules separated by a distance R, one or both of which possess several internal configurations, is given by fl = l/kT, where V is the potential energy of the pair of molecules in fixed internal con- figurations and the averages are to be taken over all configurations with the un- correlated canonical distribution functions, appropriate to the isolated molecules separated by an infinite distance.* The function W(R) is the work required to bring the pair of molecules from infinite separation to the separation R.The value of W at contact is related to the standard free energy AFO of formation of * We wish to call attention to an error in eqn. (2) of the article by Kirkwood and Shumaker.1 The second term in the second line should read - (p/2)[< V2)av - ( V)3v] instead of - (p/2)( V2)av.This error has been corrected in eqn. (2) of the present article. Eqn. (5) of ref. (1) should also be corrected by omitting the second term W(W(R) of the first line.80 RATES OF ENZYMATIC REACTIONS a complex, consisting of the molecular pair, from the isolated molecules in the following manner AFOIRT = W/kT, log K = - W/kT, (3) where K is the equilibrium constant for the formation of the complex. If the complex undergoes subsequent chemical reaction through a transition state with free energy of activation AF*, we have AF*/RT = A W/kT AW= W * - W. (4) where W* is the local free energy of the molecular pair in the activated state. We shall now investigate the effect of fluctuations in the charge distribution of the protein on the binding of a dipolar substrate of low molecular weight to an enzyme and on the rate of the decomposition of the substrate.We suppose that the substrate molecule is specifiacally bound to the active site of the enzyme by local forces with a potential WO, in which the protein does not participate, and that in the activated state the local free energy is Wo*. Under the influence of the protein, we suppose the local free energies of the unactivated and activated states to be W' and W*. If KnL and Krno and k3 and k3' are the Michaelis- Menten constants and intrinsic rates of the reaction, with and without protein participation, they are related in the following manner, where for simplicity in the first of eqn. ( 5 ) we have assumed that k3 is small relative to k2, the rate of dissociation of ES the complex into enzyme and intact substrate, a restriction which can easily be removed.We now suppose that the protein possesses v basic groups, the conjugate acids of which have dissociation constants Ki and that the substrate possesses a dipole moment p, equal to p' in the uiiactivated state and p* in the transition state leading to its decomposition. If e is the protonic charge and zie is the permanent charge of basic group i, the potential energy Y arising from the electrostatic interaction of the basic groups, their conjugate acids and the dipole of the substrate molecule, for fixed configuration of the protons is given by the expression, where yf is the angle between the dipole moment of the substrate molecule and the radius vector of length Ri from basic group i, De an effective dielectric constant depending upon geometry and the interior dielectric constant cf the protein DO, and the exterior dielectric constant D of the solvent.The occupation variables Xi specify the proton configuration, X i being equal to unity if a proton is situated on basic group i and zero otherwise. With the use of the methods of Kirkwood and Shumaker 1 for calculating the mean values (Xi), and ( X i X i ) a v , eqn. (2) leads to the following expression for the free energy of interaction of the protein with the substrate, in excess of WO due to its local binding to the active site, [H+ J ] e p cos V i i= 1 [H'] + Ki D,Ri*J . G . KIRKWOOD 81 where [H+] is the hydrogen ion activity, The first term of eqn. (7) arises from the average charges of the basic groups and the second term arises from fluctuations in charge configuration.In the calculation of the mean values over charge configurations, electrostatic interaction between the protons has been neglected, but may be taken into account by numerical calculation if desired. Although the contribution from the average charge distribution may in certain instances play a dominant role, we remark that it will be very sensitive to the geometrical arrangement of the charges, since cos yj can assume both positive and negative values, and appropriately situated pairs of basic groups can mutually nullify their contributions. On the other hand, the fluctuation term contains the factors cos Zvi, which are always positive and permit the effects of a large number of basic groups with randomly oriented radius vectors to accumulate. Since we are particularly interested here in investigating the effect of the fluctuation term, we shall consider onIy instances in which the symmetry of the arrangement of the basic groups relative to the active site causes the contribution from their average charges to vanish.Furthermore, since the fluctuation contribution from each basic group diminishes as the inverse fourth power of its distance from the active site, we may expect only vicinal groups to be effective. If we assume that there are v, vicinal basic groups of a single type a, with acidic dissociation con- stants K,, which make the effective contribution, eqn. (7) reduce to the following expression, The effective distance r, of each basic group from the active site would be equal to the actual distance, if the v, groups were distributed at uniform intervals of arc on the circumference of a circle to the plane of which the substrate dipole moment is parallel. For an actual geometrical arrangement approximating this regular array, r, will be approximately equal to the actual distance.The effecive dielectric constant is difficult to estimate reliably, since it is quite sensitive to geometry 2 near the interface between the protein and the exterior aqueous solvent. If the surface of the protein is regarded as plane in the vicinity of the active site and a pair of charges are situated at a depth b below the interface in the protein, a simple electrostatic calculation for a charge situated at a distance r from the substrate dipole yields where DO is the interior dielectric constant of the protein and D the exterior di- electric constant of the solvent.With an interior dielectric constant cf about 3 and an exterior dielectric constant of 80, D, is estimated with eqn. (9) to be of the order of magnitude 10 for reasonable ratios blr. We are now prepared to estimste the effect of fluctuations in charge configura- tion on the Michaelis-Menten constant K, and the intrinsic rate constant k3 of the enzymatic reaction. Since the effect on K , is of secondary interest, we shall present explicit expressions only for the rate constant k3. By means of eqn. (5) and (8), we obtain at once,82 RATES OF ENZYMATIC REACTIONS where A$ is the increment of the square of the dipole moment of the active site + substrate complex in activation to the transition state leading to decom- position of the substrate.We remark that k3 possesses a maximum when the pH of the solution is equal to the pK, of the conjugate acids of the effective vicinal basic groups. If (10) in the form, we denote by k, the maximum value of k3, we may write eqn. log (k3/km) = - CT tanh2 1 1 , el4 - e - U tanh u = - . eU+ edU According to eqn. (11), log (k3/k,J is represented as a function of pH by a sym- metrical bell-shaped curve with a maximum at p&, of the type commonly observed experimentally. A degree of dissymmetry would be introduced by taking into account electrostatic interactions between the protons. In instances where two or more types of vicinal basic groups were effective in contributing to the fluctu- ation free energy, the actual curve could be broadened by superposing two or more terms of the form of the right-hand side of eqn.(11). In order to make a rough comparison of the theory with experiment, we shall consider briefly the observations of Bergmann and Fruton 3 on the effect of pH on the rate of hydrolysis of carbobenzoxy-L-glutamyl-L-tyrosine by pepsin. The rate constant k3 possesses a maximum at a pH of four. This clearly indicates within the framework of the present theory that the effective basic groups are COO-. The constant k3 diminishes to about half its maximum value in an interval of one unit of pH on either side of the maximum, which requires CT to be ap- proximately unity. At 300" K, with an effective dielectric constant of 10, a value of A p of activation of 3 0 , we estimate cr to be of the order of magnitude unity for about 10 carboxyl groups situated at a distance of 5A from the active site, Although this distance is rather smaller than one might desire, we may reasonably conclude that the predictions of the theory are in semi-quantitative agreement with experiment. In conclusion, we remark that the present theory provides a reasonable mechan- ism for the participation of the protein moiety of an enzyme molecule in the enzymatic reaction. In particular, it provides an unusually simple explanation for the effect of pH on the rate of the reaction. No assertion is made that the interaction of the bound substrate molecule with vicinal basic groups through the fluctuation force is the only mechanism by which the protein may participate in the reaction. The planning of more critical experiments to test the present theory and more detailed applications to the analysis of existing experimental data will be deferred until a later time. 1 Kirkwood and Shumaker, Proc. Nat. Acad. Sci., 1952,38,863. 2 Kirkwood and Westheimer, J. Chem. Physics, 1938, 6, 506. 3 Bergrnann and Fruton, J. Bid. Chem., 1939, 127,627.

ISSN:0366-9033    

DOI:10.1039/DF9552000078

出版商:RSC    

年代:1955

数据来源: RSC

摘要:

SOME KINETIC AND MECHANISTIC ASPECTS OF HYDROLYTIC ENZYME ACTION BY KEITH J. LAIDLER Dept. of Chemistry, The University of Ottawa, Canada Received 16th May, 1955 This paper is mainly concerned with the question of what deductions about mechanisms can be made from the results of kinetic studies with hydrolytic enzymes. Some basic kinetic laws, based on the original Michaelis-Menten mechanism, are referred to, and several methods are described for investigating the nature of the enzyme-substrate constant K = [kl/(k-l + kz)]. It is noted that the pH results suggest a bifunctional type of catalysis for most hydrolytic enzymes. The effects of temperature, hydrostatic pressure, dielectric constant and ionic strength on the kinetic and thermodynamic constants are analysed, and are shown to lead to certain conclusions about reaction mechanisms.The kinetic consequences of structural and solvent effects are considered. Two examples are discussed in further detail ; the a-chymotrypsin-catalyzed hydrolysis of methyl hydro- cinnamate, and the myosin-catalyzed hydrolysis of adenosine triphosphate. Kinetic studies on enzyme reactions can conveniently be divided into two classes. The first comprises investigations of the phenomenological rate laws that are obeyed-of the way in which rates vary with concentrations of substrates, activators, inhibitors and hydrogen ions. Studies of this kind lead to values for certain rate constants and equilibrium constants. Investigations of the effects of pH on the rate, for example, lead to dissociation constants for enzymes and enzyme-substrate complexes, and the numerical values of these frequently allow conclusions to be drawn about the nature of active centres of enzymes, and about the way these centres interact with substrates, activators and inhibitors.Other constants that are obtained, such as the Michaelis constant, relate to the interaction between enzyme and substrate; some aspects of the analysis of constants of this type, which are really composite kinetic constants, are discussed below. The phenomenological rate laws also involve certain purely kinetic rate constants, such as the first-order rate constant for the breakdown of the enzyme-substrate complex into the enzyme and the final products of reaction. The investigations of the second main class are dependent upon the first, and comprise more detailed studies of the thermodynamic and kinetic constants.Up to now most work of this kind has involved studying the effect of temperature on these constants, and of evaluating the corresponding heats and entropies. The magnitudes of these quantities, especially taken in collaboration with other evidence, can throw a good deal of light on the details of the enzyme-substrate interaction and on the mechanism of enzyme action. Other investigations of this type that have been carried out are concerned with the effects of high hydro- static pressures, the dielectric constant and the ionic strength on the kinetic and thermodynamic constants. The main purpose of this paper is to discuss the second class of kinetic studies, and to consider the conclusions to which these studies lead with respect to the mechanism of hydrolytic enzyme action; it will first be necessary, however, to mention certain rate laws in order to provide a background for the later discussion.8384 HYDROLYTIC ENZYME ACTION RATE LAWS AND THEIR INTERPRETATION The basis of most kinetic studies of enzyme reactions is the Michaelis-Menten scheme,l which can be represented as ki k-1 E + S%ES, E S + E + P . If the substrate concentration is much greater than that of the enzyme, application of the steady-state treatment gives rise to the rate equation 2 k2 where [El0 is the total concentration of enzyme, [S] that of substrate and reciprocal of the conventional Michaelis constant K,, is defined by the When [S] > [El0 it can easily be shown that steady-state conditions will prevail throughout the main course of the reaction.* When modifiers (inhibitors and activators) are present-and hydrogen and hydroxide ions can be regarded as such-various types of law are possible. One sometimes finds, for example, that a law of the type of is obeyed; here [TI is the concentration of inhibitor or activator, and Kt is the association equilibrium constant for the reaction E + T +ET.When this is so the inhibition or activation is said to be of the nun-competitive type. In other cases the rate law is k2mlolSl = 1 + K[SJ + &[TI’ and the effect is then said to be competitive. The effects of pH on enzyme reactions, which have recently been treated in some detail 3~4.5, can be either competitive or non-competitive.The general methods for analyzing kinetic results in terms of rate equations of the above forms will not be discussed here, but there is one point that is worthy of some special consideration. It is seen from eqn. (2) that is a composite kinetic constant ; in the event that k-1 > kz it reduces to the equilibrium constant K(= kI/k-l), while if k2 > k-1 it becomes kl/k2. For the further analysis of kinetic data it is important to know something of the relative magnitudes of the constants k-1 and k2. There have been deveIoped several procedures which enable conclusions to be drawn about these relative values, and these will now be discussed. been measured directly, particularly by Chance using the rapid spectrophotometric method.Catalase 6 and peroxidase 7 have been studied in this manner ; an applica- tion of the method to other enzymes would be of great value, but as far as hydrolytic enzymes are concerned certain spectroscopic factors introduce some difficulty. * Actually the condition [ S ] S [El0 is a sufficient but not a necessary one for the fact that steady-state conditions will prevail for the most part. Other sufficient conditions are [El0 S [ S ] ; k-1 + k2 > kl[E]o ; k-1 4- k2 9 kl[S]. The author intends to discuss the conditions for the steady-state, and the kinetics of the transient phase, in later papers.? (Can. J. Chem., 1955, 33, 1614 ; 1956, 34, 150.) (1) DIRECT MEASUREMENT OF CONSTANTS.-In a few cases the constants haveK . J . LAIDLER 85 recently applied an interesting method that depends upon making a study of the kinetics during the transient phase of the reaction, i.e.the very early stages during which the steady-state concentration of complex is being established. The initial kinetic law can be shown to be (2) STUDY OF TRANSIENT-PHASE KINETICS.-Roughton and Gutfreund 9 have so that klk2 can be determined; since k2 and kl/(k-1 + k2) are known from the over-all kinetics the three constants can be separated. author 11 have recently worked out a detailed steady-state treatment of the non- competitive case, corresponding to the scheme of reactions shown in fig. 1. On (3) DEDUCTIONS FROM NON-COMPETITIVE BEHAVIouR.-~Ora~eS 10 and the the basis of the resultant equations they have shown that a rate equation of the form of eqn.(3) above can only arise provided that one of the following three conditions is satisfied : (i) the three complexes ES, ET and EST are essentially at equilibrium with E ; i.e. K = K. (ii) there is steric blockage of one of the two reaction paths by which EST is formed. (iii) certain constants are equal by coincidence. Possibility (iii) may be rendered un- likely if there is non-competitive behaviour with respect to more than one inhibitor or activator. Possibility (ii) can often be shown to be unlikely on the basis of general considerations ; it would seem unreasonable, for example, as far as pH is concerned or for the effects of small ions. - ES ET EST 2 FIG. 1 .-General scheme of reactions for enzyme systems. E represents the enzyme, S the substrate and T a sub- stance that may be a second substrate, an activator, an inhibitor or a hydro- gen ion.In many cases it is therefore possible to deduce from non-competitive behaviour that equilibrium exists. Whenever, for example, is independent of pH over a range in which the rate shows significant variation it seems reasonable to draw this conclusion. which leads to useful conclusions in certain systems involves comparing the results with a homologous series of competitive inhibitors or activators with those for an analogous homologous series of substrates. This method depends upon the fact that the experimentally observed modifier constants [Kt in eqn. (4)] are neces- sarily equilibrium constants. If, therefore, structural changes in the modifiers affect Kt in exactly the same manner as the analogous structural changes in the substrates affect 2, it would seem likely that the substrate constants are also equilibrium constants.This conclusion is not, however, an absolutely firm one, since it is possible that structural changes would affect k2 in the same manner as k-1, and so influence in the same manner as Kt even though (5) STRUCTURAL EFFECTS WITH suBsTuTEs.-In some instances it has been observed that substituent groups in substrates bring about the same change in k2 as they do in 1/2. This is readily understood if l/F is equal to k2/kl, but not (except on the basis of coincidence) if l/?? is k-l/kl. If k2 and 1/K change differently no conclusion can, of course, be drawn. (4) STRUCTURAL EFFECTS WITH COMPETITIVE MODIFIERS.-hlOther method were kl/k2.86 HYDROLYTIC ENZYME ACTION (6) CONSISTENCY.-~ some cases it is possible to show that a particular assumption about the relative magnitudes of k-1 and k2 leads to difficulties in the interpretation of the results.Two examples will illustrate this type of argument. (i) For the carbonic anhydrase reaction an analysis of Kiese’s results 12 leads to the conclusion that the apparent heat of activation corresponding to low substrate concentration is - 11-7 kcal. On the assumption that k2 > k-1 this value would relate to the rate constant k2, and the negative value would be im- possible to interpret. The alternative assumption that k-1 > k2 avoids this diffi- culty by leading to the result that the activation energy relates to the composite constant klk2/k-l.(ii) For carboxypeptidase, Lumry, Smith and Glantz 13 have found that the free energy and heat changes associated with the formation of complex with the competitive inhibitor hydrocinnamic acid are - 5 and - 7 kcal respectively. The assumption that ?? is an equilibrium constant leads to corresponding values of - 3 and 0 kcal for the substrate carbobenzoxy-glycyl-L-phenylalanine. The lack of analogy between the results for substrate and enzyme suggests that this assumption is incorrect and that in reality K = kl/k2. This type of argument is not strong, but is helpful together with other evidence. TABLE 1.-ENZYME-SUBSTRATE SYSTEMS FOR WHICH IT APPEARS THAT k-1 k2, SO THAT K = kl/k-l enzyme trypsin chymotrypsin chymotrypsin chymotrypsin myosin urease sucrase carbonic anhydrase substrate benzo yl-L-arginine ethyl ester met hy 1 hy drocinnama te acetyl-L-phen ylalamine ethyl ester nico tin yl-L- tryp to- phanamide ; acetyl-L- tryptophanamide ATP urea sucrose carbon dioxide and water type of evidence non-competitive pH dependence non-competitive pH dependence transient-phase kinetics structural effects with competitive inhibitors modifier replacement non-competitive pH dependence non-competitive pH dependence consistency ref.14 15 9 16 17 18 19 12 TABLE 2.-ENZYME-SUBSTRATE SYSTEMS FOR WHICH IT APPEARS THAT k2 $ k-1 enzyme substrate type of evidence ref. of constants 6 catalase hydrogen peroxide direct determination peroxidase hydrogen peroxide direct determination carboxypeptidase various (i) structural effects and acceptor of constants 7 with substrates 13 (ii) consistency 13 In tables 1 and 2 are listed some enzyme-substrate systems for which it has proved possible to arrive at a reliable conclusion about the relative magnitudes of k-l and k2. Some non-hydrolytic enzymes are included.SOME DEDUCTIONS ABOUT ENZYMATIC MECHANISMS Brief consideration may now be given to the question of what conclusions with regard to mechanisms can be derived from the phenomenological rate laws. It will be convenient to discuss this here, and later (in the following section) toK. J . LAIDLER 87 see to what extent these mechanisms are supported and supplemented by the magnitudes of certain of the constants that are derived from the kinetic and thermo- dynamical quantities.In the first place, the kinetic studies on enzymes all provide support for the idea that definite complexes are formed between enzyme and substrate-as in the original Michaelis-Menten scheme and in more elaborate schemes such as that of fig. 1. It is true that Medwedew 20 has proposed an alternative mechanism which gives similar kinetic laws but in which complex formation is not essential for reaction. On the whole, however, Medwedew’s scheme is unsatisfactory ; for one thing, it regards enzyme-inhibitor interactions as entirely different from enzyme-substrate interactions.21 Secondly, the study of pH effects on enzyme reactions has provided extremely important information about reaction mechanisms. Since this matter has recently been treated by the author in considerable detail,4 with an analysis of the main experimental results,s only a very brief outline will be given here.Almost in- variably enzyme reaction rates pass through maxima as the pH is varied, and this indicates that at least two groups that change their states of ionization are involved in reaction. The results enable the pKs of these groups to be deduced and also enable one to conclude whether the groups are bound in the enzyme- substrate complex, or merely enter into the subsequent reaction-in the latter case they are bound in the activated complex but not in the stable complex. In the majority of cases for which data are available it appears that only the basic group of the enzyme is involved in the bonding of the stable complex; for an ester hydrolysis, for example, the formation of the stable complex might therefore occur as follows : 0 0- It I R-C-0-R’ R-C-0-R’ 17 + H B A B+ A / + I I I 1 1 enzyme I 1 enzyme The fact that both an acidic and a basic group are involved at least in the formation of the activated complex is of particular interest.There seems little doubt that enzymes act as “ bifunctional ” catalysts, in a similar manner to certain organic catalysts studied by Swain and Brown.22 Such a conclusion,23 as dis- cussed elsewhere,24 may provide an explanation of the very high effectiveness of enzymes as compared with conventional acidic and basic catalysts ; the operation of a “push-pull” mechanism may lead to more effective results than when catalysis occurs by the usual protolytic or prototropic mechanisms that occur in acid-base systems. It should finally be mentioned, in connection with mechanisms, that much valuable evidence comes from non-kinetic sources.Koshland has recently discussed what may be deduced from stereochemistry 25.26 and from isotope studies .26 PROPERTIES OF THE KINETIC AND THERMODYNAMIC VALUES Investigations of the rate laws have been seen to lead to certain kinetic and thermodynamic constants for various types of process. Further information about these constants can be obtained by studying the effect on them of temperature, pressure, dielectric constant and ionic strength. Some of the results of such investigations will now be summarized and discussed.88 HYDROLYTIC ENZYME ACTION (1) TEMPERATURE.-A study of the temperature-dependence of an equilibrium constant gives rise to the corresponding standard entropy and enthalpy ; that of a rate constant gives the entropy and heat or energy of activation, these quanti- ties being related to the rate constant by the relationships 27 (6) (7) kT h k = ~ exp (AS*/R) exp (- AH*/RT), kT h = e - exp (AS*/R> exp (-- E/RT).Generally these entropy and energy terms are found to be temperature-independent although in a few cases (especially fumarase 28) there appears to be a fairly sharp change at a critical temperature. For the systems listed in table 1 it is known that %is an equilibrium constant (K), and for some of them the variation of K with temperature has been studied. Some results are shown in table 3. The significance of the AS values for chymo- trypsin and myosin (adenosine triphosphatase) is discussed later.TABLE 3.-HEAT AND ENTROPY VALUES FOR THE FORMATION OF THE ENZYME-SUBSTRATE COMPLEX enzyme substrate PH AH AS ref. chymotrypsin methyl 7-8 - 5.3 - 11.3 29 myosin ATP 7.0 8.0 49-0 17 urease urea 7.1 - 2.9 0.9 30 hydrocinnamate TABLE 4.--KINETIC VALUES RELATING TO THE BREAKDOWN OF THE ENZYME-SUBSTRATE COMPLEX enzyme pepsin pepsin trypsin trypsin trypsin chymotrypsin chymotrypsin chymotrypsin chymotrypsin chymotrypsin chymotrypsin chymo trypsin chy mo trypsin chymotrypsin carboxy- peptidase carboxy- peptidase carboxy- pep tidase adenosine triphosphatase substrate carbobenzoxy-L-glutamyl- L-tyrosine ethyl ester carbo benzoxy-L-glutamyl- L-tyrosine benzoyl-L-argininamide sturin benzoyl-L-arginine ester methyl hydrocinnamate methyl DL-cx-chloro-/3- methyl-D-/3-phenyllactate methyl-L-/3--phenyllactate acetyl-L-tyrosine ethyl ester benzoyl-L-tyrosine ethyl ester benzo y l-L-p heny lalanine ethyl ester benzoyl-tyrosyl-glycylamide benzoyl-L-tyrosinamide carbobenzoxyglycy1-L- tryptophan carbobenzox ygl ycyl-L- phenylalanine carbobenzoxyglycyl-L- leucine adenosine triphosphate phenylpropionate temp.("C) pH 31.6 4.0 31.6 4.0 25.5 7-8 24.5 7.5 25-0 8.0 25.0 7-8 25.0 7.8 25.0 7.8 25.0 7.8 25.0 7.8 25.0 7.8 25.0 7.8 25.0 7-5 25-0 7.8 25-0 7.5 25.0 7.5 25.0 7-5 25-0 7.0 k2 E2 (sec-1) (kcal) 0.00 108 0.00141 270 13,100 26.7 0.026 0.135 0.139 1.38 193-0 78.0 37.4 37 89 181 106 1 04 0.625 20.7 17.2 14.9 11-8 11.2 16.8 15.4 15-1 11.1 11.5 9.2 12.5 11.5 14.6 9.9 9.6 8.6 13.0 A&* cal/deg.- 6.5 - 21.8 - 8.2 - 6-7 - 16.5 - 11.8 - 13.2 - 14.2 - 23.4 - 10.7 - 21.4 - 11.0 - 19.8 - 13.0 - 19.8 - 18.5 - 20.0 - 8.0 ref. 31 31 32 32 33 29 29 29 29 29 34 29 32 34 13 13 13 17 urease urea 20.8 7.1 20,000 9.7 - 7.2 30K . J . LAIDLER 89 Considerably more extensive are the results relating to the temperature dependence of the first-order rate constant k2 for the breakdown of the enzyme- substrate complex. Table 4 gives a fairly complete list of the results for hydrolytic enzymes that have been investigated in a satisfactory state of purity. It is of interest that all of these are negative in sign, a fact that suggests that there is an increase in polarity of the system during this process, this increase leading to an increase in the extent of binding of the solvent molecules.Values for the second-order rate constants, and the corresponding energies and entropies of activation, are also available for a number of enzyme systems ; a list of values is given in table 5. These second-order rate constants, obtained from the rates in the low substrate concentration region, are actually the com- posite constants k2E, or klk2/(k-l + k2). These second-order constants will not necessarily obey the Arrhenius law, but they do so in the two special cases k-1 > k2 and k2 > k-1. In the former case the heats and entropies of activation that are determined are AH1* + AH2* - AH* and AS,* + AS2* - AS* ; in the latter case they are AH^* and AS,*. The significance of these relationships is illustrated in fig.2. When the situation corresponds to neither of these special cases the plot of log k2K against 1/T will not necessarily be linear, but heat and entropy values may still be calculated from the slope at any temperature. These values are easily shown to be related to the individual values as follows : * - k-l(AHl* + AH2* - AH" 1) + k2AHl* AH - k-i + k2 9 k-l(AS,* + A&* - AS*, ) + k2ASl* k-i + k2 AS* = AF*- AF:+AF:-AF-: AH"- A H y t AH:-AH!, A S * - AS:+ AS:-AS?, ES* A F * - A F: AH* - AH: AS*- AS? FIG. 2.-Free-energy diagram for a simple enzyme reaction, showing the relationships for the t.wo cases (a) k-1 @ k2 and (6) k2 k-1. When k-1 and k2 are of similar mag- nitudes the relationships are shown in eqn. (8) and (9) of the text. They are therefore the weighted means of the values for the extreme cases [(a) and (b) in fig.21, the weighting factors being k-1 and k2. I n order to interpret the magnitudes of the entropies listed in tables 4 and 5 it is necessary to take into account two types of effect, which it is convenient to refer to as solvent and structural effects. By structural effects is meant the re- versible changes in the conformation of the enzyme that occur during the reaction. The existence of such structural effects has previously been discussed with reference to pressure effects on enzyme ~ystems.3~ It was proposed that in some cases the enzyme molecule assumes a more open configuration when it forms a complex,90 HYDROLYTIC ENZYME ACTION and that when this undergoes subsequent reaction there is a refolding of the enzyme molecule.Such a model is consistent with the results for a number of enzyme systems, but cannot be said to have been firmly established. The idea is, however, reasonable in that it brings the enzyme reactions into line with what takes place during enzyme inactivation, in which it seems certain that unfolding occurs. A possible way of separating solvent and structural effects, by varying the dielectric constant, will be considered later. In the meantime we may see how the results in table 5 can be interpreted in a broad way in terms of solvent effects alone.* TABLE 5.--KINETIC VALUES RELATING TO THE FORMATION OF THE ENZYME-SUBSTRATE COMPLEX enzyme pepsin pepsin trypsin chymotrypsin chymotrypsin chymo tr ypsin chymotrypsin chymotrypsin chymotrypsin carboxy- peptidase carboxy- peptidase carboxy- peptidase urease adenosine- triphosphatase substrate carbobenzoxy-L-glutamyl- L-tyrosine ethyl ester carbobenzoxyl-L-glutamyl- L-tyrosine chymotrypsinogen methyl hydrocinnamate me thyl-DL-cc-chloro-P- phenylpropionate methyl-D-P-phenyllactate methyl-L-/3-phenyllactate benzoyl-L-tyrosine ethyl ester benzoyl-L-tyrosinamide carbobenzoxyglycy1-L- tryptophan carbobenzoxyglycy1-L- phenylalanine carbobenzoxyglycy1-L- leucine urea adenosine triphosphate temp 31.6 4.0 31.6 4.0 19.6 7.5 25.0 7.8 25.0 7.8 25.0 7-8 25.0 7.8 25.0 7.8 25.0 7-8 25.0 7.5 25.0 7-5 25.0 7.5 20.8 7.1 25.0 7.0 (“C) pH k,K (mole-1 sec-1) -57 0.79 2,900 6.66 11.2 4.0 138 19,500 14-9 17,444 27,900 3,920 5.0 X 106 8 .2 ~ 106 E AS* 23.1 14.1 20.2 2.6 16.3 8.5 11.5 - 23.2 6.9 - 33.0 3.1 - 47.2 3.8 - 38.5 0.8 - 38.5 3.7 - 43.0 9.9 - 8.5 9.6 - 8.5 11.0 - 7.5 6.8 - 6.8 21.0 44.0 ref.31 31 32 29 29 29 29 34 34 13 13 13 30 17 The entropy values in table 5 may be seen to fall into two main groups : with pepsin, trypsin and adenosine triphosphatase the values are positive, while with chymotrypsin, carboxypeptidase and urease they are all negative. Some light is thrown on these differences by a consideration of the types of substrates that are hydrolyzed by these six enzymes. Chymotrypsin and urease invariably act upon electrically neutral molecules. Adenosine triphosphatase, on the other hand, is itself positively charged, and it acts upon a negatively charged substrate. Pepsin will only act upon substrates that contain a free - COOH group, and this at pH 4 exists at least in part as a negatively charged group.Carboxypeptidase also acts only on substrates containing the - COOH group, which will be ionized. Trypsin’s specificity requirement is for a free - NH2 group, and this will exist as the positively charged - NH3+ group. There is thus a possibility that with ATP-ase, pepsin, carboxypeptidase and trypsin the interaction between enzyme and substrate involves a charge neutralization, whereas with chymotrypsin and urease there can be no such neutralization. During the reaction between an enzyme and an uncharged substrate there will result certain electron shifts as a result of which the activated complex will be * It may be noted that in any bimolecular reaction in solution there is always a small entropy contribution resulting from the fact that a solute species disappears when the activated complex is formed ; there is therefore an entropy of the “ unmixing ” of this with the solvent.For an aqueous solution the entropy loss due to this is 7.9 cal/mole deg.K . J . LAIDLER 91 more polar than the reactants; there is therefore an increase in electrostriction and a corresponding negative entropy of activation. With ATP-ase, pepsin and trypsin, on the other hand, this effect is presumably counteracted by the charge neutralization that occurs when the enzyme and substrate come together. This neutralization will lead to a release of water molecules, and there will be a cor- responding increase of entropy. This explanation of the signs of the AS* values derives some support from the general correlation between the sign of the entropy change and whether the substrate is charged or not.Carboxypeptidase, however, appears to be an exception, in that the substrates are charged but the entropies of activation are negative ; a possible explanation is that the - COO- group on the substrate does not come into contact with a positive group on the enzyme, and is therefore not neutralized. Some of the entropy changes shown in table 5 seem to be too large to be ex- plained in terms of electrostatic effects alone.* For chymotrypsin, where there is no charge neutralization, the entropies of activation are much more negative than can be accounted for on the basis of “unmixing” with solvent. With ATP-ase there is seen to be a large positive entropy change and here, as will be seen, the results indicate an important structural effect.(2) HYDROSTATIC PRESSURE.-BY studying the effect of hydrostatic pressure on the rates of reactions it is possible to determine the volume changes occurring during activation. For both nonenzymatic 36 and enzymatic reactions 37 these volumes changes fall in line with the corresponding entropy changes, and throw much light on possible solvent and structural effects during the course of reaction. The volume values can, in fact, be interpreted satisfactorily in terms of electro- striction effects, along the lines of the discussion above of the entropies of activation. A fairly complete collection of volumes of activation is given in tables 6 and 7, which also give for comparison the entropies of activation for the same or a similar system.There is seen to be a fairly good correlation between the signs and magnitudes of the entropies and volumes. Thus for cases in which the enzyme- substrate interaction involves a charge neutralization (e.g. with pepsin, trypsin and myosin) the AS* and AV* values are generally positive, a fact that is attributed to release of bound water molecules. When, on the other hand, there is probably no charge neutralization (e.g. with chymotrypsin and urease) the results indicate negative values of AS* and of rl V* ; such values are partly due to structural factors (the two molecules coming together) and partly to polarity increases. Inspection of table 6 shows that the AV,* values, like those of A&*, are in- variably zero or negative. This effect is attributed to increases of polarity that occur when the enzyme-substrate complex becomes an activated complex.(3) DIELECTRIC CoNSTANT.-In view of the fact that the characteristics of the rate and equilibrium constants are affected by both structural and solvent factors, it is clearly of importance to develop methods for distinguishing between these two types of effect. Some progress in this direction has been achieved by making kinetic measurements in a series of mixed solvents of different dielectric con- stants.23~ 41 This method is useful in providing semiquantitative information, but unfortunately the situation is too complex to permit an exact theoretical interpretation. The qualitative basis of the method is as follows.A reaction in which the activated complex is more polar than the reactants will be accelerated by an in- crease in dielectric constant; conversely one in which the complex is less polar will be slowed down. Theoretical treatments of these effects have been developed in various ways, and by taking into account the change in dielectric constant with temperature some interpretation can be placed on the entropies of activation. Unfortunately some complications arise from specific solvent-solute interactions, but nevertheless the theory is probably correct in accounting for the main trends. * As a rough rule it can be said that the entropy change due to neutralization on complex formation is equal to - ~OZ~Z,, where z, and Z, are the valencies of the ions.92 HYDROLYTIC ENZYMATIC ACTION The theory has been applied quantitatively to two enzyme-reactions, the a- chymotrypsin-catalyzed hydrolysis of methyl hydrocinnamate,23 and the myosin- catalyzed hydrolysis of adenosine triphosphate.41 In the second of those reactions there is good reason to believe that the enzyme-substrate interaction will involve a charge neutralization, so that increasing the dielectric constant should slow down TABLE 6.-vOLUME AND ENTROPY DATA FOR z[s] 9 1 enzyme pepsin trypsin trypsin trypsin trypsin chymotrypsin chymotrypsin sucrase salivary amylase pancreatic amylase adenosine- triphosphatase enzyme pepsin pepsin t rypsin trypsin chymotrypsin chymotrypsin pancreatic lipase myosin substrate tyrosine benzo yl-L-argininamide benzoyl-L-arginine isopropyl ester L-arginine methyl ester serum albumin L-tyrosine ethyl ester benzoyl-L-tyrosine ethyl ester sucrose starch carbobenzoxy-~-glutamyl-~- A V2* - 0 0 - 5.5 - 13.7 - 13.5 - - 8 - 22 ref.AS2 * - - 21.8 38 - 8.2 38 - 16.5 - 38 39 40 - I - 21.4 - - 35 35 - - starch - 2 8 35 adenosine triphosphate - 32 to - 25 37 8.0 TABLE 7.-vOLUME AND ENTROPY DATA FOR K[S] < 1 substrate A V* ref. AS* - 2.6 carbobenzoxy-L-glutamyl- - L-tyrosine - gelatin 22 35 /I-lactoglobulin - 36 38 chymotrypsinogen - - casein - 13.8 40 tributyrin 13 35 - 6.5 - benzoyl-L-tyrosinamide - - - 43.0 - ATP 8 to 2 3 37 41 ref. 31 32 33 - - - 3 4 - - - 17 ref. 31 - - 32 34 - - 17 the process. In the former, however, there is no such neutralization, and the ionization of the carbonyl group in the substrate would be expected to lead to an increase in polarity as the enzyme and substrate come together ; increasing the dielectric constant should therefore increase the rate of the process.Both of these predictions are supported by experiment, and the results have led to a plausible interpretation of the entropy and volume results. Further details are given later. (4) IONIC STRENGTH.-very little systematic work has been done on the effects of ionic strength on rate constants in enzyme systems. In principle a study of the effect of ionic strength on a reaction should reveal the same kind of information as does changing the dielectric constant. As far as simple reactions between ions are concerned the simple Br~msted-Christiansen-Scatchard treatment is adequate at low ionic strengths, but for ion-dipole and dipole-dipole reactions there appears to be no theory that adequately accounts for the behaviour in terms of the electrical properties of the reacting molecules. Since this is true even for simple reactions it is obvious that little can as yet be learnt of enzyme-substrate interactions from this type of investigation. A fundamental theoretical study of this matter would therefore seem to be required. A second difficulty is that certain ions affect enzyme reactions not only by virtue of the ionic-strength effect but also because they undergo specific interactionsK .J . LAIDLER 93 with the enzyme and perhaps with the substrate. In some cases, indeed, ions are actually part of the structure of the active centre, so that ionic concentrations may then have an important effect on the activity.An example of a study of ionic effects is briefly discussed below in connection with myosin. TWO EXAMPLES The general methods that have been described above will now be discussed with reference to two hydrolytic enzymes whose reactions have been studied kinetically in some considerable detail. a-CHYMoTRYPsIN.-One of the simplest substrates for chymotrypsin, and indeed for any enzyme, is methyl hydrocinnamate, CsH5CH2CH2COOCH3, and attention will here be confined to the work done with this compound. The reaction is not markedly affected by inorganic ions, and the effects of the usual buffer ions (e.g. Na+, HPO42-, B4072-) appear to be ionic-strength effects.The simple Michaelis- Menten law is obeyed accurately.29~ 15.23 A recent study 15 of the reaction over a range of pH values has shown that k2 passes through a maximum, but that i? remains constant. From this can be drawn two important conclusions: (i) the ionizing groups responsible for the variation in rate as the pH is varied are not involved in the formation of the enzyme- substrate complex, and (ii) the Michaelis constant is an equilibrium constant. The first conclusion arises from the fact that invariance of Kcan only arise if the binding of enzyme to substrate does not affect the ionizations of the acidic and basic groups that play a role in the reaction.4 From the variation of k2 with pH the pK of the ionizing groups in the free enzyme are found to be 7.2 and 8.0.There is little evidence regarding the nature of the enzyme-substrate binding, but it is known to be weak; the free energy of binding is only about 3 kcal per mole (see below), so that van der Waals’ effects or hydrogen bonding might be involved. The pK values of 7.2 and 8.0 might correspond to the ring nitrogen of histidine and an a-amino group, so that the following is a possible mechanism for the reaction : 0- / R-C /\ ‘oa+-c€-I3 H--08- I I I II I I R-C O-CHJ I I I / 0 OH H H II I )r’ -4 Y ‘ +H20 I I I 1 I -- R--- C--O-CH3 I I 6+ I -- I I N NH3+ t NS+ NH2 N+ NH2 / \ / \ I enzyme Michaelis complex activated complex products According to this mechanism the catalysis is of the bifunctional type, there being an attack of the basic \N group on the carbonyl carbon atom and an attack of the acidic - NH3’ group at the alcoholic oxygen atom.is an equilibrium constant follows from the discussion given above and is of interest in view of the fact that, on the basis of inhibition studies, Huang and Niemann 16 made the same inference for two other substrates / The conclusion that94 HYDROLYTIC ENZYME ACTION (cf. table 1). It is now possible to re-interpret in a more precise manner certain previous results for the methyl hydrocinnamate system, as follows. The value of 3.9 x 10-3 M for the equilibrium constant Km leads to - 3-2 kcal/mole for the AF of formation of the complex, and from the temperaturc dependence of K, the values AH = - 5.0 kcal and AS = - 6.2 cal/mole deg. are obtained. A splitting of the entropy value into electrostatic (solvent) and non-electrostatic (structural) effects is rendered possible by the dielectric constant work,23 and leads to The actual values obtained in this way are not to be taken seriously, but the signs are quite significant.The loss of entropy due to electrostatic effects indicates an increase of polarity giving rise to electrostriction of solvent; an increase in polarity may occur if the binding of the complex is brought about by hydrogen bonding involving polar groups on the enzyme and the ester grouping in the substrate. The positive value of ASn.e.s. suggests, as appears to occur in other cases, some loosening of the enzyme structure as the complex is formed. As fat- as the breakdown of the enzyme-substrate complex is concerned, the solvent studies led to the following results : ASe.s.= - 18 cal/mole deg., AS,,,. = 12 cal/mole deg. (AS2*),.,. = - 20, (AS*)n.e.s. = 8. The latter value is a normal one for a unimolecular reaction and can be explained in terms of a simple splitting of the complex into enzyme and products. The negative value of (AS2*),, indicates a further charge separation during the forma- tion of the activated complex from the Michaelis complex. Such a separation is consistent with the reaction scheme given above, there being, in particular, an ionization of the carbonyl group during the process. The general picture that emerges for the mechanism of the reaction is therefore as follows. The enzyme and substrate molecules first come together to form (possibly by hydrogen bonding or van der Waals’ forces) a complex, which is essentially in equilibrium with the reactants and decomposes at a rate which is low compared with that of the reformation of the reactants.This complexing appears to be accompanied by some loosening of the enzyme structure. The subsequent breakdown of the enzyme-substrate complex into products involves the attack on the ester by an acidic and a basic group in the enzyme, groups that have played no part in the complexing. There is a simultaneous attack by those groups on the substrate, the basic group (produced by the ionization at pH 7.2) attacking the carbonyl carbon atom and the acidic group (which ionizes at pH 8.0) attacking the alcoholic oxygen atom. The concerted attack of these two groups leads to a flow of electrons through the ester molecule and a consequent falling apart into acid and alcohol.MYosIN.-In 1939 it was shown by Engelhardt and Luibimova42 that the muscle protein myosin has the property of catalyzing the hydrolysis of adenosine triphosphate into the diphosphate and inorganic phosphate. It is now well established 43 that myosin is indeed identical with adenosine-triphosphatase, and a number of kinetic and other studies have been carried out with the pure enzyme. In view of the fact that a detailed review of the kinetic and other aspects has recently been written 44 the present accoupt will merely summarize such points as are particularly relevant to the previous discussion. Ouellet et aZ.17 worked mainly at a KCl concentration of 0.6 M and a CaClz concentration of 0.001 M, and obtained the values for the kinetic and thermo- dynamical constants that are quoted in tables 3,4 and 5.Replacement of calcium ions by magnesium ions gave results that are consistent only with non-competitive behaviour so that it is concluded that the complex is at equilibrium with the reactants, i.e. that k-1 > k2. It is to be seen from the values in the tables that the mutual approach of the enzyme and substrate to form either the MichaelisK. J . LAIDLER 95 complex or the activated complex is associated with a considerable increase of entropy, and that the further reaction of the complex occurs with a small negative entropy of activatian. In order to throw some light on the magnitudes of the entropy changes Laidler and Ethier 41 studied the reaction in solvents of different dielectric constants, while Laidler and Beardell 37 studied the kinetics over a range of hydrostatic pres- sures and at three KCl concentrations.The solvent work showed that there is an electrostatic entropy of activation of 31 cal/mole deg. corresponding to the approach of enzyme and substrate, and this is consistent with there being a charge neutralization. There is also a non-electrostatic entropy increase, consistent with conformational changes in the protein. Direct evidence for such changes has been revealed by the light-scattering experiments of Blum and Morales.45 The pressure studies showed a volume increase for this process, and therefore support the conclusions from the solvent and light-scattering work. For the subsequent reaction of the enzyme-substrate complex the results indicate an increase of electrostatic entropy, a decrease of nonelectrostatic entropy and a decrease of volume during the activation process.These results suggest a separa- tion of like charges during the process, and a change of conformation associated with a volume decrease. A study of the variation of the rate constants with the KCl concentration 37 led to results that are in general consistent with the above conclusions. At low KCl concentrations the slope of a plot of log k2K against the square root of the ionic strength is negative, suggesting again that there is an approach of opposite charges. At higher concentrations, however, the slope is reversed and it would appear that specific effects are now important.The variation of k2 with ionic strength is always consistent with there being a separation of like (presumably negative) charges. The picture of the reaction to which the work leads is therefore as follows, and it may be noted that this picture is entirely consistent with the theory of Morales and Botts 46 of the mechanism of muscular Contraction. Enzyme- substrate complex formation is associated with charge neutralization. Pre- sumably it is this process that is associated with the contraction of a muscle, since the kinetic results indicate that at the same time there is a change of con- formation of such a nature as to lead to an entropy increase. During the subsequent breakdown of the complex the negatively charged phosphate ion separates from the rest of the molecule, which is also negatively charged.At the same time the enzyme returns to its original conformation and there is a loss of entropy due to this cause. The author’s thanks are due to the United States National Science Foundation for making possible his participation in this Discussion. He is also indebted to Dr. M. F. Morales for many valuable discussions and for information about unpublished work. 1 Michaelis and Menten, Biochem. Z., 1913, 49, 333. 2 Briggs and Haldane, Biochem. J., 1925, 19, 338. 3 Laidler, Trans. Faraday SOC., 1955, 51, 528. 4 Laidler, Trans. Faraday SOC., 1955, 51, 540. 5 Laidler, Trans. Faraday SOC., 1955, 51, 550. 6 Chance, Grcenstein and Roughton, Arch. Biockern. Bioplzys., 1952,37, 301.Greenstein, Higgins and Yang, Arch. Biochem. Biophys., 1952, 37, 322. 7 Chance, J . Biol. Chem., 1942, 1, 533 ; Science, 1949, 109, 204. 8 Roughton, Faraday SOC. Discussions, 1954, 17, 1 16. 9 Gutfreund, Faraday SOC. Discussions, 1954, 17, 220. 10 Morales, J. Amer. Chem. SOC., 1955,77,4169 ; cf. Botts and Morales, Truns.Faraday 11 Laidler, to be published. Chance, SOC., 1953, 49, 696.96 CRYSTALLINE PAPAIN 12 Kiese, Biochem. Z., 1941, 307, 400. 13 Lumry, Smith and Glantz, J. Amer. Chem. SOC., 1951, 73,4330; Lumry and Smith, 14 Gutfreund, Trans. Faradny SOC., 1955, 51, 441. 15 Laidler and Barnard, to be published. 16 Huang and Niemann, J, Amer. Chem. SOC., 195 1,73, 1541. 17 Ouellet, Laidler and Morales, Arch. Biochem. BiopJiys., 1952, 39, 37. 18 Myrback, Acta Chem. Scand., 1947, 1, 142. 19 Myrback and Bjorklund, Arkiv Kemi, 1952, 4, 567. 20 Medwedew, Enzymologia, 1937, 2, 1, 17, 53. 21 Hearon and Katzman, Bull. Math. Biophys., 1954, 16, 259. 22 Swain and Brown, J. Amer. Chem. SOC., 1952, 74, 2538. 23 Barnard and Laidler, J. Amer. Chem. SOC., 1952, 74, 6099. 24 Laidler, Introduction to the Chemistry of Enzymes, chap 9 (McGraw-Hill, 1954). 25 Koshland, Biol. Rev., 1953, 28, 414. 26 Koshland, in The Mechanism of Enzyme Action (The Johns Hopkins Press, Balti- 27 Glasstone, Laidler and Eyring, The Theory of Rate Processes (McGraw-Hill, 1941), 28 Massey, Biochem. J., 1953, 53, 72. 29 Snoke and Neurath, J. Biol. Chem., 1950, 182, 577. 30 Wall and Laidler, Arch. Biochem. Biophys., 1953, 43, 299. 31 Casey and Laidler, J. Anrer. Chem. Soc., 1950, 72, 2159. 32 Butler, J. Amer. Chem. Soc., 1941, 63, 2971. 33 Schwert and Eisenberg, J. Biol. Chem., 1949, 179, 665. 34 Kaufman, Neurath and Schwert, J. Biol. Chem., 1949, 177, 792. 35 Laidler, Arch. Biochem., 1951, 30, 226. 36 Burris and Laidler, Trans. Faraday SOC., 1955, 51, 1497. 37 Laidler and Beardell, Arch. Biochem. Biophys., 1951, 31, 285. 38 Werbin and McLaren, Arch. Biochem. Biophys., 1951, 32, 325. 39 Talwar, Macheboeuf and Basset, J. Colloid Sci., 1954, Suppl. 1: 14. 40 Werbin and McLaren, Arch. Biochem. Biophys., 1951, 31, 285. 41 Laidler and Ethier, Arch. Biochem. Biophys., 1953, 44, 338. 42 Engelhardt and Luibimova, Nature, 1939, 144, 668. 43 Engelhardt, Advances in Enzymol., 1946, 6, 147 ; Sur I’Enzymologie de la Myosine, lecture given at Second International Biochemical Conference, Paris, July, 1952. 44 Morales, Botts, Blum and Hill, Physiol. Rev., 1955, 35, 475. 45 Blum and Morales, Arch. Biochem. Biophys., 1953, 43, 208. 46 Morales and Botts, Arch. Biochem. Biophys., 1952, 37, 283 ; Faraday Soc. Discus- this Discussion. more, 1954), p. 608; this Discussion, p. 142. y. 199. sions, 1953, 14, 425.

ISSN:0366-9033    

DOI:10.1039/DF9552000083

出版商:RSC    

年代:1955

数据来源: RSC