18 ENZYME MECHANISMS By M. R. Hollaway (Department of Biochemistry University College London W.C.1) Introduction.-There can be little doubt that the most exciting recent develop- ments relevant to an understanding of enzyme mechanisms have come from X-ray crystallographic studies. The structures of several enzymes and enzyme- inhibitor complexes have now been described at the 2-3 A level; this has made possible a fairly unambiguous identification of protein side-chains which interact with the substrate molecules during catalysis. A review of the structures published up to October 1967 has been given by Stryer’ and some subsequent developments will be referred to in this account. Such a complete knowledge of the three-dimensional structure of an enzyme cannot in itself inform us how the enzyme functions; as Gutfreund and Knowles2 pointed out :‘making a model of a horse from photographs does not necessarily tell us how fast it can run’.Nevertheless,any proposed mechanism must be compatible with the X-ray structures in terms of the geometry of the active site. Furthermore the proposed reactivity of catalytic groups invoked in the mechanism should be consistent with their position in the enzyme-substrate complex i.e. their microenvironment. Significant advances have also been made in the study of enzyme reactions by use of rapid reaction techniques. The approach to mechanism through observation of the nature of intermediates by their physical signals3 and measure- ment of their speed of interconversion is made difficult in enzyme reactions because if the very high rates involved.Thus most enzymes transform between 10 and 1000 substrate molecules per second per active site. This sets an upper limit to half-lives of intermediates to around 100 milliseconds so that special techniques are required for the elucidation of transient steps in enzymic catalysis. Such techniques including stopped-flow. temperature-jump and combined stopped-flow-T-jump are being continually refined so that now Chance et d4report time resolutions below 1millisecond with rapid mixing techniques and Eigen and his co-workers’ are able to study transients in perturbed equili- L. Stryer Ann Rev. Biochem. 1968,37,25. H. Gutfreund and J. R Knowles ‘Essays in Biochemistry’ vol. 3 ed. P.N. Campbell and G. D. Greville Academic Press 1967 p. 25. ’ M. L. Bender in ‘Rates and Mechanisms of Reactions’ pt. 11 ed A. Weissberger Interscience New York 1963 p. 1427. B. Chance D. DeVault V. Legallais L. Mela and T. Yonetani in Nobel Symposium 5 ‘Fast Reactions and Primary Processes in Chemical Kinetics’ ed S. Claesson Interscience New York 1967 p. 437. ’ M. Eigen Quart. Rev. Biophys. 1968 1 3 and refs. therein. M. R. Hollaway bria with half-lives below 1 p second using the temperature-jump technique. The recent development of the combined-flow-T-jump technique’ enables the study of enzyme transients by perturbation of an enzyme-substrate system in the steady state rather than at equilibrium. Excellent accounts of rapid reaction methods as applied to the study of enzymes are given in the books by Gutfreund6 and Bernhard7 and in the reviews by Chance* and Eigen.s*9 A number of very fine reviews have recently been published including the essay on ‘The Foundations of Enzyme Action’ by Gutfreund and Knowles;2 ‘New Looks and Outlooks in Physical Enzymology’ by Eigen ;’an account by Singer” on covalent labelling of the active site enzymes and a review by Kosh- land and Neet on ‘The Catalytic and Regulatory Properties of Enzymes’.Volume XI of the invaluable series ‘Methods in Enzymology”2 contains sections on specific modification reactions and the investigations of confor- mation changes. There is a book on ‘Design of Active-site Irreversible Enzyme Inhibitors’ by Baker.13 Some of the other excellent reviews which are more specific in nature will be referred to in the relevant section.In this report consideration will only be given to those enzymes for which X-ray crystallographic studies are relatively advanced. This limitation may give a somewhat one-sided view of enzymic mechanisms since the enzymes which fall into this category are low molecular-weight hydrolytic proteins com- prising a single-polypeptide chain. Furthermore the molecules usually contain intramolecular disulphide bonds; this seems to be a feature of extra- cellular enzymes whereas as far as the reviewer is aware enzyme molecules found in the interior of cells do not contain disulphide links and usually contain more than one polypeptide chain. (Hartley has pointed out that the disulphide bond is a characteristic structural feature of extracellular proteins e.g.ref. 14.) Nevertheless it is to be hoped that principles discovered for the single chain hydrolytic enzymes will be applicable in more complex systems. The Basis of Enzyme Catalysis.-A great deal of consideration has been given to factors which could explain why enzymes are such good catalysts. The book by Bruice and Benkovic,” the essay by Gutfreund and KnowlesY2 and the review by Koshland and Neet” all give a careful evaluation of the extent to which all or some of the following effects may contribute to enzymic catalysis (u)a proximity effect arising from an enhanced collision frequency for reactant molecules bound to the enzyme; (b) an orientation effect resulting from the H.Gutfreund ‘An Introduction to the Study of Enzymes,’ Blackwell Oxford 1965. ’ S. A. Bernhard ‘The Structure and Function of Enzymes,’ Benjamin New York 1968. B. Chance in ref. 3. M. Eigen ref. 4 p. 333. lo S. J. Singer Adv. Protein Chem. 1967,22 1. l1 D. E.Koshland and K. E. Neet Ann. Rev. Biochern. 1968,37,359. l2 ‘Methods in Enzymology,’ vol. XI ed. C. H. W. Hirs Academic Press New York 1967. l3 B. R Baker ‘Design of Active-Site Directed Irreversible Enzyme Inhibitors,’ Wiley New York 1967. l4 P. B. Sigler D. M. Blow B. W. Matthews and R Henderson J. Mol. Biol. 1968,35 143. T. C. Bruice and J. Benkovic ‘Bioorganic Mechanisms,’ vol. I Benjamin New York 1966. Enzyme Mechanisms optimal positioning of functional groups of the enzyme so as to assist in bond- forming bond-breaking processes ;(c) general base or general acid catalysis including bound metals acting as Lewis acids; (d)covalent catalysis whereby a group on the enzyme is a stronger nucleophile than the final acceptor and a better leaving group than the leaving group portion of the substrate; (e) a ‘strain’ effect in which the substrate molecule binds to the enzyme in a confor- mation that is part of the way towards the transition state of the catalysed reaction (an excellent discussion of this and other types of strain theory has been given by JencksI6); (f)a microenvironment effect whereby the local environ- ment of groups in the enzyme-substrate complex may be such as to facilitate reaction pathways which are energetically unfavourable in aqueous solution ; and (9)concerted catalysis a concept arising largely from the oft-quoted work of Swain and Brown.17 These last authors found that in benzene solution 2- hydroxypyridine was a 7000-fold more effective catalyst of the mutorotation of a-tetra-0-methyl-D-glucosethan a mixture of pyridine and phenol each at the same concentration as the 2-hydroxypyridine.This result cannot be accoun-ted for on the basis of increased acidity of the hydroxy-group or increased basicity of the ring nitrogen in the 2-hydroxypyridine. The proposed inter- pretation was that a concerted acid-base catalysis was operable in the rate- limiting ring-opening step. It may be argued that this system cannot represent a model for enzymic catalysis because the studies were carried out in a non- polar solvent whereas enzyme reactions proceed in aqueous solution.Such an objection may be invalidated by the consideration that the active sites of at least some enzymes contain regions of low polarity (see later). Koshland and Neet l1 have extrapolated from model systems in an attempt to discover what order of rate-enhancement would be expected if the effects (a)-(d) were to operate simultaneously. The results of their calculations showed that a factor of at least lo8 remains to be accounted for in enzymic catalysis by strain concerted catalysis or other effects. Snell Kwok and Kim18 have put forward some suggestions for enzymic mechanisms based on some work on the aminolysis of the methyl esters of salicyclic acid (1) and p-hydroxybenzoic acid (2) in dry dioxan solution.The l6 W. P. Jencks in ‘Current Aspects of Biochemical Energetics,’ ed. N. 0.Kaplan and E. P. Kennedy Academic Press New York 1966,273. C. G. Swain and I. F. Brown J. Amet. Chem. SOC.,1952,74,2538. R. L. Snell W-K. Kwok and Y. Kim J. Amer. Chem. SOC.,1967,89,6728. U M. R.Hollaway reaction of (1) with n-butylamine in dry dioxan at 50 and at 70" was found to be first-order in ester but second-order in amine. Under the same conditions (2) did not react with n-butylamine to any detectable extent. The suggested inter- pretation of the second-order component in amine was that one molecule combined with a molecule of ester to give an ion pair represented simply by (3 in which the butylammonium ion served as an acid catalyst during the nucleo- philic attack of a second amine molecule.This system is considered by the authors as a model for enzyme reactions involving the hydrolysis of esters or amides. A hypothetical enzyme is postulated to contain in its active site a carboxy-group and two amine groups one a stronger base than the other. On binding substrate it is proposed that all water is excluded from the active site and the substrate optimally orientated with respect to the catalytic groups ('microstereochemically oriented'). The carboxy-group and strong base form an ion pair which interacts with the substrate so as to facilitate nucleophilic attack by the weaker amine base. This work is predated by that of Menger,lg who studied the reactions of benzamidine or n-butylamine with p-nitrophenyl acetate in dry chlorobenzene solution.Benzamidine was found to react with the ester some 15,000 times faster than the butylamine monomer and only fourfold slower than hydroxide ions in water. As in the system studied by Smell Kwok and Kim'* the p-nitrophenyl acetate-butylamine reaction was second order in amine but the rate-law for the benzamidine reaction only contained a first-order component in amidine. Menger ' concluded that benzamidine acted as a bifunctional reagent in a manner depicted in (4). OAr ph-I H (4) Such a process would be feasible in chlorobenzene since there is no charge formation in the transition state. This model reaction was considered to support the suggestion that multifunctional catalysis by proteolytic enzymes may occur in a cyclic fashion in non-polar regions of the active sites.As indicated by Menger,Ig the model is reminiscent of the Swain-Brown system." WangZ0 has suggested that facilitated proton transfer along rigidly held hydrogen bonds in the enzyme-substrate complex might account for both the specificity of enzymes and their efficiency as catalysts. It is proposed that this factor enables enzyme systems to reach transition states faster often by a l9 F.M. Menger J. Amer. Chem. SOC. 1966,88,3081. 2o J. H. Wang Science 1968,161,328. Enzyme Mechanisms 605 process which is essentially a pre-transition-state protonation. Mechanisms were suggested for the reactions catalysed by carbonic anhydrase chymotryp- sin ribonuclease and some dehydrogenases.Another suggestion by Braunstein Ivanov and Karpeisky,’ is that enzyme reactions are distinguished from chemical transformations in solution by a high degree of conformational mobility in the enzyme-substrate complexes. This mobility is considered to provide the capacity to meet different sets of requirements in consecutive steps of the catalytic process. A consideration of the chemical and physical properties of metalloenzymes has led Vallee and Williamsz2 to propose that these enzymes are in an entatic state i.e. that they might be ‘poised for catalytic action in the absence of substrate.’ This concept is presented largely on the basis of the atypical spectral properties of metalloenzymes which suggest an unusual geometry in the arrangement of ligands about the bound metal ions.The active site is seen as an area which is ‘closer to a unimolecular transition state than to that of a conventional stable molecule thereby constituting an energetically poised domain’. This concept is reminiscent of the ‘strain’ theory16 but with the enzyme in the deformed state. It should be emphasized that the entatic-state theory cannot account for the specificity of enzymes. Thus if a metallopep- tidase catalyses the hydrolysis of a peptide bond between two amino-acid residues the theory would predict that the enzyme would also readily catalyse the hydrolysis of say acetamide. Such indiscriminate catalysis is not observed in the hydrolytic reactions of metallopeptidases.Nevertheless when taken in conjunction with other factors the unusual environment of bound metals at the active site of metalloenzymes may well be an important factor in catalysis. The role of ligand-induced conformation changes in enzymic reactions i.e. the induced fit theory of Koshland has been extensively discussed in the review by Koshland and Neet.” The evidence for such changes appears convincing especially in the case of carboxypeptidase A,z3 where binding of molecules of similar structure to substrates leads to the movement through about 14 A of one of the enzyme tyrosine hydroxy-groups. An important aspect of the induced-fit theory is its capacity to account for the specificity of enzyme reactions.Wagner and his co-workersz4 have carried out a series of interesting experi- ments on a model system designed to investigate the possible role of apolar bonding in enzymic catalysis. The rate of aqueous hydrolysis of the NN-dimethyl-N-(p-nitrophenyloxycarbonylethyl)dodecylammonium ion (5) was 21 A. E. Braunstein V. I. Ivanov and M. Ya Karpeisky in ‘Pyridoxal Catalysis Enzymes and Model Systems,’ ed E. E. Snell A. E. Braunstein E. S. Severin and Yu M. Torchinsky Interscience New York 1968,291. 22 B. L. Vallee and R J. P. Williams Proc. Nut. Acud. Sci. U.S.A.,1968,59,498. 23 G. N. Reeke J. A. Hartsuck M. L. Ludwig F. A. Quiocho T. A. Steitz and W. N. Lipscomb Proc. Nur. Acud. Sci. U.S.A.,1967 58 2220. 24 T. E. Wagner C-J. Hsu and C. S.Pratt J. Amer. Chem SOC. 1967,89,6366; R G.Shorenstein C.S. Pratt C-J. Hsy and T. E. Wagner ibid. 1968,90 6199. M. R.Holluwuy cNa N H fH2 -0 (6) OOC. CH. NHCO. C17 H35 1 O=C -[CH2I2-i (CH3I2. C12 H25 strongly catalysed by N-stearoylhistidine (6). The latter compound was a 2000-fold better catalyst than N-acetylhistidine. Moreover the reaction catalysed by (6) exhibited a hyperbolic relationship (the so-called Michaelis- Menten kinetics) between rate of hydrolysis and ester concentration at a fixed concentration of (6). It was concluded that the rate enhancement was due to apolar bonding between the lyophobic groups in (5) and (6). The reaction was subject to competitive inhibition by trimethyl(steary1)ammoniumbromide and was also inhibited by urea; these two types of effect are often observed in enzyme reactions.Lysozyme.-Hen eggwhite lysozyme is a small protein of molecular weight CU. 14,600 comprising a single linear polypeptide chain of 129 L-a-amino-acid residue^.^'-^^ There are four intra-chain disulphide bonds in the m~lecule.~’ The bacteriolytic action of lysozyme resides in its abilit~~*-~’ to catalyse the hydrolysis of the p-N-acetylmuraminyl (NAM) glycosidic link in polymers of alternately linked p-(1 4)-N-acetylmuraminyl and N-acetylglucosaminyl (NAG) residues (NAM-NAG),. Chitin,31 the p-(1 + 4)-linked N-acetyl- glucosamine polymer represented (NAG), and derived oligosa~charides~~ are also substrates for lysozyme action. The bond cleaved during catalysis is that between C-1 of the pyranoside ring and the glycosidic oxygen atom and the reaction proceeds with greater than 99 % retention of configuration about the glycosidic carbon atom.32 The elegant X-ray crystallographic studies of Blake et al.33-35and Phillips36 25 R E.Canfield and A. K. Liu J. Biol. Chem. 1963,239 2698. l6 J. Jolles J. Tauregui-Adell and P. Jollbs Biochim. Biophys. Acta 1963 78 668. ’’ R E. Canfield and A. K. Liy J. BioI. Chem. 1965,240 1997. ” M. R J. Salton Ann Rev. Biochem 1965,34,143. 29 J. A. Rupley and V. Gates Proc. Nat. Acad. Sci. U.S.A. 1967,57 496. 30 D. M. Chipman J. J. Pollock and N. Sharon J. Biol. Chem. 1968,243 487. ” L. R. Berger and R. S. Weiser Biochim. W’ophys. Acta 1957,26 517. 32 M. A. Rafferty and T. Rand-Meir Biochemistry 1968 7 3281.’’ C. C. F. Blake D. F. Koenig G.A. Mair A. C. T. North D. C. Phillips and V. R Sarma Nature 1965,206,757. 34 C. C. F. Blake G. A. Mair A. C. T. North D. C. Phillips and V. R. Sarma Proc. Roy. SOC. B 1967,167,365. ” C. C. F. Blake L. N. Johnson G. A. Mair A. C. T. North D. C. Phillips and V. R Sarma Proc. Roy. SOC. B 1967,167 378. 36 D. C. Phillips Proc. Nat. Acad. Sci. U.S.A. 1967,57 484. Enzyme Mechanisms have led to a detailed knowledge of the three-dimensional structure of the hen eggwhite lysozyme molecules and of several enzyme-inhibitor complexes. Key features of the structure are (i) that the shape of the molecule is slightly ellipsoid with a large cleft at its surface and (ii) that the interior of the molecule contains amino-acid residues with non-polar side chains (apart from Gln-57 and Ser-91) a feature in common with other globular proteins of known structure.Furthermore these studies have indicated that the cleft region is the active site of the enzyme. A substrate comprising six sugar residues e.g. (NAG) can be built into the cleft giving an arrangement of protein side-chain and substrate as represented in crude schematic form in (7). Detailed drawings of the active site region containing a bound (NAG) molecule are given in refs. 35 and 36 and an excellent three-dimensional representation produced by the Xograph technique has been published. It should be noted that the proposed conformations and interactions of the substrate residues A B and C are those found in the lysozyme-(NAG) crystal struct~re,~~,~~ whereas the arrangement of residues D.E and F is based on careful model building. Some noteworthy points regarding the lysozyme-(NAG) complex are as follows. s*36 (a) In order to build residue D into the active site it must be distorted towards a half-boat conformation. (b) For steric reasons a sugar with a 3-lactyl substituent as in NAM cannot be placed in position C. This fact in conjunction with the hydrolysis pat- tern2'-,' for (NAM-NAG) and the stability of the lysozyme-(NAG) complex in which the sugars are bound at A B and C suggested that cleavage occurs between residues D and E. (c) The carboxy-groups of Asp-52 and Glu-35 are located at either side of the D-E glycosidic link and are consequently regarded as potential catalytic groups.The environments of these residues are such as to suggest that Glu-35 may have a high pK value whereas Asp-52 being placed in a polar environ- ment may have a low pK,. The foregoing studies have led to a proposal of a mechanism for lysozyme action by the group at The Royal Instit~tion~~~~~ in discussion with Prof. C. A. Vernon.,' It is suggested that the substrate binds to the active site in such a way that sugar residue D is distorted towards the half-chair conforma- tion. In the subsequent bond-breaking step the protonated form of Glu-35 provides general acid catalysis and the ionised Asp-52 serves to stabilise the incipient carbonium ion as shown in (8).The latter postulate serves to explain the retention of configuration at the glycosidic carbon atom.It is significant that the solvolysis of 1,3,4,6-tetra-O-acety1-~~-glucopyranosyl chloride in 37 R A. Harte and J. A. Rupley J. Biol. Chem.,243 1663. C. A. Vernon Proc. Roy. SOC.1967 B 167 389. M.R. Hollaway '/ h c 3 In 3 2-U cy In Enzyme Mechanisms acetic acid is considered to proceed via a carbonium ion with the half-chair conf~rmation,~~ so that the distortion of the D ring towards this arrangement on binding of a substrate to lysozyme could represent an example of catalysis through induced ‘strain’ discussed in an earlier section.’ Lowe and his ~o-workers~~~~ have proposed an alternative mechanism for lysozyme action in which the 2-acetamido-group of the D sugar residue provides anchimeric assistance and Glu-35 acts as a general acid catalyst during the rate limiting bond-breaking step (9).In the most recent com- m~nication~~ no mechanistic role is ascribed to Asp-52. The implication of the 2-acetamido-group was based on the observation that the k, value for the di-N-acetylchitobioside (10) was at least 100 times that of the corresponding glycoside (11) although the CT values for OH ( +0-25) and NHAc ( +028) are (10) R’ = NHAc R2= p-nitrophenyl (11) R’ = OH R2= p-nitrophenyl (12) R’ = NHAc R2= aryl similar. Earlier studies by Lowe et aL41 had. implicated either concerted acid- base or acid-nucleophilic catalysis in the rate-limiting step for lysozyme hydrolysis of some P-aryldi-N-acetylchitobiosides(12) because KM values for these substrates were independent of the aglycone whereas the k,, values varied with the leaving group giving a Hammett p value of + 1-2.Some care must be exercised in extrapolating the mechanistic conclusions derived from these studies to lysozyme action on ‘normal’ substrates e.g. (NAG), since the largest k,, value reported4042 for the derivatives (lo) (ll),and (12) was such that one molecule of lysozyme catalysed the hydrolysis of one glycosidic bond every 40 min Normal substrates are probably hydrolysed at least 4 orders of magnitude faster.43 39 R. U. Lemieux and G. Huber Cad. 1.Res. 1955,33,128. 40 G. Lowe Proc. Roy. SOC.B 1967,167,431. 41 G. Lowe G.Sheppard M. L. Sinnott and A. Williams Biochem. J. 1967,104,893. 42 G. Lowe and G. Sheppard Chem. Comm. 1968,529. 43 J. A. Rupley L. Butler M. Gerring F. J. Hartdegen and R Pecoraro Proc. Nat. Acud. Sci. U.S.A. 1967,57 1088. 610 M. R. Hollaway Piszkiewicz and Br~ice~~ (see also ref. 49) have also discussed the possibility of anchimeric assistance by the 2-acetamido-group in lysozyme action. Raftery and Rand-Meir45 have recently considered several possible mecha- nisms for lysozyme-catalysed reactions. They exclude single displacement reactions on the basis of the greater than 99 % retention of configuration about the glycosidic carbon atom. Anchimeric assistance by the 2-acetamido- (or 2-hydroxy)-group was also shown to be unnecessary by the observation that lysozyme catalysed the rate of release of p-nitrophenol from 2-deoxy-P-~- glucosides (13) at about 16 times the rate for the corresponding glucoside (14).(13) R = H (14) R = OH The authors were careful to state,45 however that ‘. ..although anchimeric assistance is not necessary to explain catalysis by lysozyme the present findings do not of course exclude its occurrence in NAG substrates’. The remaining possibilities are either the carbonium ion intermediate rnechani~m,~ or a double-displacement mechanism involving a glycosyl-enzyme intermediate.46 Vernon38 considers the latter mechanism inoperable on steric grounds. The acid-catalysis carbonium ion mechanism3* is further supported by some work by Rupley Gates and Billbre~.~’ These authors measured the relative rates of hydrolysis and transglycosylation (krel)in the system :lysozyme-(NAG) plus an acceptor nucleophile.Firstly it was found that krelwas essenti- ally constant for a series of alcohols of similar steric requirements but with pK values ranging over 4 units. General-base catalysis by a group on the enzyme was invoked to explain this levelling effect. This group would act as a general acid in the first step of the hydrolytic reaction. Secondly strongly nucleophilic sulphur acceptors were less reactive than their oxygen analogues? a fact which is incompatible with a glycosyl-enzyme inte~mediate.~~ The low reactivity of the thiol analogues cannot be accounted for on steric grounds because the reaction centre is available to a sulphur atom as evidenced by the lysozyme-catalysed hydrolysis of phenylthioglu~oside.~~* 49 44 D.Piszkiewicz and T. C. Bruice J. Amer. Chem. SOC. 1967 89 6237. 45 M. A. Raftery and T. Rand-Meir Biochemistry 1868,7 3281. 46 D. E. Koshland jun. Bid. Rev. 1953,28,416. 47 J. A. Rupley V.Gates and R Billbrey J. Amer. Chem Soc. 1968,90 5633. 48 A. J. Rhind-Tutt and C. A. Vernon J. Chem SOC.,1960,4637. 49 D. Piszkiewicz and T. C. Bruice Biochemistry 1968 7 3037. Enzyme Mechanisms 61 1 Dahlquist and Rafteryso have described some interesting n.m.r. studies on lysozyme-inhibitor interactions. Thus both the glycosidic methyl protons and the acetamido-protons of methyl 2-acetamido-2-deoxy-~-~-glucopyranoside underwent a chemical shift in the enzyme-inhibitor complex.The glycosidic methyl proton shift was pH-independent whereas the extent of the acetamido- proton shift varied with the ionisation of the two groups of pK 4-7 0.1 and 7.0 0.5. Furthermore the pH-dependence of the dissociation constant for the enzyme inhibitor complex implicated a group of pK 6.1. Previous studies5’ on the binding of (NAG) to lysozyme involved two groups of pK values 4-2 and 5.8 in the free enzyme and 3.6 and 6.3 respectively in the complex. A consideration of the crystal structure of the lysozyme<NAG) complex3’ together with the foregoing observations has led Dalhquist and Rafteryso to make the following assignments of pK values Asp-101 pK 4.2(3-6 in the complex); Glu-35 6-1 (7.0 0.5 in the complex); and Asp-103 4-7.The bell- shaped pH profiles with inflections near 4and 6 for the action of lysozyme on low molecular weight substrate^^^*'^ may reflect two or more of these ionisa- tions. An exciting recent discovery is that the amino-acid sequence of bovine a-lactalbumin is largely homologous to that of hen eggwhite lyso~yrne.~~ Furthermore a-lactalbumin is one of the two protein components of lactose synthetase which catalyses the reaction uridine diphosphate galactose + D-glucose + lactose + uridine diphosphate The other protein component A protein catalyses the reaction a-uridine diphosphate galactose + N-acetylglucosamine -+ N-acetyl-lactosamine + uridine diphosphate However addition of a-lactalbumin (designated B protein) to the A protein results in a system which gains ability to catalyse (15) whilst losing the facilitys4 to catalyse (16).The similarity of a-lactalbumin to lysozyme is further accentuated by the fact that it has proved possibless to build a model of a-lactalbumin based upon the main chain backbone of ly~ozyme.~’ In this model the interior 50 F. W. Dahlquist and M. C. Raftery Biochemistry 1968,7 3269 3277; see also M. A. Raftery F. W. Dahlquist S. I. Chan and S. M. Parsons J. Biol. Chem. 1968,243,4175. 51 F. W. Dahlquist L. Jao and M. A. Raftery Proc. Nut. Acad. Sci. U.S.A.,1966,56 26. ’’ T. Osawa and Y.Nakazawa Biochim Biuphys. Acta 1966 130 56. 53 K. Brew T. C. Vanaman and R L. Hill J. Biol. Chern. 1967,242 3747. 54 K. Brew T. C. Vanaman and R L.Hill Proc. Nut. Acad. Sci. U.S.A.,1968,59,491. 55 A. C. T. North Abstracts of the Meeting of The British Biophysical Society Dec. 18th-l9th 1968 p. 6. 612 M. R. Hollaway retains the apolar character seen in the lysozyme molecule and it seems likely that the two proteins have very similar structures. Serine Proteases.-The excellent review by RyleS6on ‘The Endopeptidases of Vertebrates’ contains a description of the relevant literature up to 1966. Two reviews by Bender and KCzdy consider respectively the mechanisms of action of proteolytic enzymess7and of chymotryp~in.~~ Books containing discussionson this group of enzymesare by Gutfreund,6Bernhard,7and Bruice and Benkovic.’ CunninghamS9has also presented an excellent comprehen-sive account of the ‘Structure and Mechanism of Action of Proteolytic En-zymes’.The reader is also referred to the important review by Neurath Walsh and Winter6’ which gives consideration to the ‘Evolution of Structure and Function of Proteases’. The terms ‘analogous proteins’ and ‘homologous proteins’ are used here in the manner defined by these authors. Chyrnotrypsin (CT). There is a considerable body of evidence that the CT-catalysed hydrolysis of specific substrates involves a minimum of three steps (reviewed in refs. 15 and 56-62) k+l k+2 k+3 E+S====ES =EP,eE+P k-1 k-2 ++ P,P k-3 (17) Initial formation of an enzyme-substrate complex through non-covalent bonding is followed by catalytic steps involving acylation (k+ step) and deacylation (k+3 step) of a group on the enzyme.The role of the uniquely reactive Ser-195 residue as the group acylated and the involvement of His-57 in catalysis have been extensively reviewed.”*56-62 Specific substrates for chymotrypsinaction are typically N-acyl-substitutedesters,amides or peptides of L-a-amino-acids with acyl side chains. X-Ray crystallographic studies14* 63 combined with a knowledge of the amino-acid sequence of u-CT~~ have enabled Blow and his collaborators to determine the structure of bovine p-tolylsulphonyl-Ser-195-CTat 2 Aresolu-tion. A difference electron density map between tosylated and native enzyme has also been calculated.’ The a-CT molecule comprises three polypeptide chains designated A B 56 A. P. Ryle Ann Reports 1966,63 614.’’ M. L. Bender and F. J. KCzdy Ann Rev. Biochem. 1965,34 49. ’’ M. L. Bender and F. J. KCzdy J. Amer. Chem SOC.,1964,86 3704. 5g L. Cunningham in ‘Comprehensive Biochemistry,’ ed. M. Florkin and E. H. Stotz Elsevier Amsterdam 1965 VOL 16 p. 85. 6o H. Neurath K. A. Walsh and W. P. Winter Science 1967,158,1638. 61 A. Himoe P. C. Parks and G. P. Hess J. Biol. Chem. 1967,242,919. 62 M. L. Bender M. J. Gibian and D. J. Whelan Proc. Nut. Acad. Sci. U.S.A. 1966,56,833. 63 B. W. Matthews P. B. Sigler R Henderson and D. M. Blow Nature 1967,214 652. 64 B. S. Hartley in ‘The Structure and Activity of Enzymes,’ ed T. W. Goodwin J. 1. Hams and B. S. Hartley Academic Press London 1964 p. 47; B. S. Hartley and D.L. Kauffman,Biochem J.,1966 101 229; B.Meloun I. Kluh V. Kostka L. Moravek Z. Prusik J. VanBkk B. Keil and F. Sorm Biochim. Biophys. Acta 1966 130 543. Enzyme Mechanisms 613 and C; the B chain is covalently joined to A and C by disulphide bonds and the B chain has one and the C chain two intrachain disulphide bonds.64 Apart from eight residues at the N-terminus of the C chain in an a-helical arrangement the polypeptide chains are in an extended conformation and folded so as to form a compact approximately spherical structure with an interior comprised almost entirely of non-polar residues. There is a slight groove at the surface of the molecule in the region of the active site.I4 The most recent reveals that in the uninhibited enzyme the centre of the imidazole ring of His-57 is about 3 %i from the oxygen atom of Ser-195 a distance consistent with hydrogen-bonding between the imidazole 1-nitrogen atom of His-57 and the serine oxygen.However in the tosylated enzyme at low pH values these appears to be a hydrogen bond between imidazole 1-nitrogen of His-57 and the sulphonyl oxygen atom mediated by a water molecule. It is noteworthy that significant differences in the electron density map between native and tosyl-CT are only apparent in the active centre region and the largest side-chain movement that of Met-192 is less than 1A. An exciting recent de~elopment~~ is that residue 102 formerly designated as an asparagine is in fact aspartic acid. This group is located in an essentially non-polar ‘pocket’ generated by residues Ala-55 Ala-56 Cys-58 Tyr- 194 Ile-99 and Ser-214.The entrance to the pocket is sealed by His-57 whose imidazole 3-nitrogen atom is hydrogen-bonded to Asp-102 thereby rendering the latter residue inaccessible to solvent molecules (18). Blow Birktoft and Hartley6’ consider that the species (18) which is only one of several possible mesomeric and tautomeric representations is catalytically inactive but may lose a proton in a process of pK ca. 7 (the pK value observed in acylation and deacylation reactions see previously) to give an active species (19). It is worth mentioning a caueat entered by WallenfelP who emphasised that observed pK values for such a process may not be ascribed to any par- ticular group but are a characteristic of the system as a whole.This arrangement (19) has been christened by Blow Birktoft and Hartle~~~ as a ‘charge-relay system’ for relaying electrons from Asp-102 to the surface of the molecule via His-57 and has led these authors to suggest a mechanism for CT action (20) incorporating many of the features of a mechanism proposed by Wang.,’ This depicts the rate-limiting acylation step in the CT-catalysed hydrolysis of a peptide amide or anilide. Deacylation would involve re- placement of RNH by a water molecule and a reversal of the acylation process. Key features of the mechanism are that (a) The ‘charge relay system’ ensures a significant concentration of the Ser- 195-0 species at neutral pH values ; (b) the ‘charge relay system’ also serves to assist the removal of the leaving group by transfer of the Ser-195 proton; and ‘’ D.M.Blow J. J. Birktoft and 8. S. Hartley Nature 1969,221 337. “ K. Wallenfels personal communication 1964. M. R. Hollaway 614 In z L QrY p\T Enzyme Mechanisms ASP-102 1 95 A (c) The overall process is facilitated by electron transfer uiu pre-formed rigidly held hydrogen bonds.20 It is interesting that an electrophilic component in CT-catalysed reactions has been implicated by various observations. Thus Metzger and Wilson67 found that the rate constant for acylation of CT by diphenylcarbamyl fluoride (3790 M-’ sec-I) was eight times that for the corresponding chloride an inversion of the expected reactivity for these reagents. Such an inversion may occur in situations where electrophilic catalysis can operate e.g.in the reaction of benzoyl halides with Grignard reagents6* where kgCl may act as an acid catalyst. Secondly Inagami York and Patchornik6’ observed that k, (acylation rate constant) for the CT-catalysed hydrolysis of a series of substituted anilides of N-acetyl-L-tyrosine was subject to a large negative p value suggesting the involvement of general acid catalysis. This was supported by a kinetic isotope effect k+2(H20)/k+2(D20) = 3-4. These studies have been extended by Wang and Parker,70 who found that the increase in k, in H20 or D20 for a series of anilides paralleled an increase in the basicity of the substrated as measured by titration in protonated or deuteriated glacial acetic acid.However the kinetic isotope effect diminished with increasing basicity a finding which was interpreted as evidence for protonation of the ” H. P. Metzger and I. B.Wilson Biochemistry 1964,7,926. ‘* C. E. Entemann and J. R Johnson J. Amer. Chem SOC. 1933,55 2900. ‘’ T. Inagami S. S. York and A. Patchornik J. Amer. Chem SOC.,1965,87 126. 70 J. H. Wang and L. Parker Proc. Nut. Acad. Sci. U.S.A.,1967,58,2451; L. Parker and J. H. Wang J. Biol. Chem. 1968 243 3729. 616 M. R. Hollaway anilide nitrogen atom prior to the rate-limiting step in acylation i.e. specific acid catalysis. This argument is subject to the objection that the basicity of the anilide in glacial acetic acid might be associated with protonation of the carbonyl oxygen rather than the nitrogen atom.Nevertheless if this is correct then (17) must be expanded to give (21) k+1 k+2 k+3 k+4 E + S ES /ES' /EP -. . .E + Pz k-1 k-2 k-3 + p (21) It may be of significance in this context that Hess7' has recently described an extra intermediate in the CT hydrolysis of furoylacryloyltryptophanamide. Rapid reaction studies enabled the measurement of the various rate constants k+ = 6-2 x lo6 M-' sec-'; k- = 2.7 x lo3 sec-'; k, = 13 sec-'; k- = 30 sec-' ; and k+4 = 50 sec-'. However whether the k+z step involves a protonation of substrate tetrahedral intermediate or conformation change in the enzyme remains to be seen. Another attempt to account for the exceptional nucleophilicity of Ser-195 in CT is that by Epstein Michel and Mosher7 who consider that Arg-14 may be sufficiently close to Ser-195 to lower the pK value of its hydroxy-group to ca.8. In discussing the nucleophilicity of Ser-195 it is important to note that it is unreactive towards iodoacetic acid. Thus Botvinik and Novodarova7 have shown that reaction of CT with this reagent at pH values between 5 and 9 leads only to substitution of histidine and lysine residues. In view of the ready reaction of alkoxide ion with alkyl halides74 this suggests that in the absence of substrate or substrate analogues there is no significant concentration of the 4form of Ser-195 at neutral pH values. Hydrogen bonding of Ser-95 to His-57 in itself is also insufficient to account for the reactivity of Ser-195 since Piszkiewicz and Bruice7' have shown that the His-15 Thre-89 hydrogen- bonded system in hen eggwhite lysozyme does not possess abnormal esterolytic activity.In (18) an electrostatic interaction is represented between the carboxylate anion of Asp-194 and Ile-16. The importance of this interaction in maintaining an active conformation of CT has been established largely through the elegant work of Hess and his collaborators.61*76*77 Thus the deprotonation of a 71 G. P. Hess Proc. Roy. SOC. December 1968,in preparation; K.G. Brandt A. Himoe and G. P. Hess J. Biol. Chem. 1967,242 3973. 72 J. Epstein H. 0.Michel and W. A. Mosher J. Theor. BWL 1968,19,320. 73 M. M. Botvinik and G. N. Novodarova Biokhimiya 1968,33,296. 74 C. K.Ingold 'Structure and Mechanism in Organic Chemistry,' Bell London 1953.75 D. Piszkiewicz and T.C. Bruice Biochemistry 1968,7,3037. 76 H. L. Oppenheimer B. Labouesse and G. P. Hess J. Biol. Chem 1966 241 2720 and refs. therein. 77 B. H. Havsteen and G. P. Hess Biochem Biophys. Res. Comm. 1964,14,313;k Y.Moon J. M. Sturtevant and G.P. Hess J. Biol. Chem. 1965,240,4204. Enzyme Mechanisms group in CT pK ca. 8.5-9.0 has been shown to be associated with (i) the formation of enzyme unable to bind substrate in a mode leading to catalysis ;61 (ii) an alteration in the 0.r.d. spectrum of the enzyme and a change in the extent of hydrogen ion uptake on binding di-isopropyl phosphofluoridate (DFP).76*77 Evidence has been presented that the ionising group is the a-ammonium group of Ile-16.76* 78 This group cannot be titrated in di-isopropyl-phosphoryl-CT (DIP-CT) which has a pH-independent 0.r.d.spectrum similar to that of the active low-pH form of the enzyme.76 Furthermore chymotrypsinogen in which the a-amino-group of Ile-16 is involved in a peptide bond also has a pH-independent 0.r.d. spectrum but this is similar to the inactive high-pH form of the enzyme. The X-ray crystallographic 63 revealed that the protonated a-amino-group of Ile-16 forms a 'buried' ion-pair with the carboxylate anion of Asp-194 which has led Blow and his co-~orkers'~ to suggest an interpretation of the pH-dependent struc- tural transition. Loss of the proton from the a-ammonium group of Ile-16 is considered to lead to a movement of the carboxylate anion to the surface of the molecule away from the non-polar interior.The resulting catalytically inactive structure is thought to be similar to that of the zymogen and to be completely devoid of capacity to bind sub~trate.'~ In DIP-CT the bulky DIP group hinders the movement of the Asp-194 group thereby explaining the inability to titrate the a-ammonium group. Hess and his collaborator^'^ have observed a proton uptake by 6-CT on binding the specific substrate N-acetyl-L-tryptophanamide.The pH dependence of the number of protons taken up per molecule of CT was interpreted according to the complex ionisation scheme (22) ................................. KSA where E represents an inactive conformation of the enzyme and E* an active conformation.Hence a pK value (pK*,) was determined for the overall process involving ionisation and a conformation change of 9.0 in the free enzyme (Kg = =9.6 in the enzyme-substrate complex (K& = on the binding of proflavin (regarded as a substrate analogue) to a-CT seemingly support a scheme such as (22),since binding appears to proceed in two steps a rapid second-order process followed by a slower first-order isomerisation of the enzyme-'sub- strate' complex. 8o 78 C. Ghelis J. Labouesse and B. Labouesse Biochem Biophys. Res. Comm. 1967,29 101. 79 J. McConn E. Ky C. Odell G.Czerlinski and G. P. Hess Science 1968 161,274. B. H. Havsteen J. Bid. Chem. 1967 242 769. M. R.Hollaway Ligand-induced pK shifts in CT have also been noted by Bender and Wedler," who observed the raising of the pK value of a group (originally 8.8) by at least 2 units on binding the competitive inhibitor benzyl alcohol.Glick82 has investigated the perturbation of prototropic equilibria on ligand binding to CT over a wider pH range than other workers. In particular it was shown that binding of the CT inhibitor,83 chloromethyl L-1-tosylamidophenethyl ketone (23) gave proton release or uptake depending on the starting pH. The results were interpreted as a ligand-induced decrease in the pK value of a group of pK 6-7 and elevation of the pK value of a group with original pK 8-6. The downward shift in the pK value of the group of pK 67 was considered to result from the formation of a hydrogen-bond between His-57 and Ser-195 only on binding of ligand.(24) R' = H,R2 = C0,Me (23) (25) R' = CO,Me R2 = H Several attempts have been made to probe the topography of the active site of CT most often by use of substrates or inhibitors of fixed conformation. There seems to be general agreement (see later) with Hein and Niemann's postulate84 that substrates of the type R'CHR' COR3 bind at three loci p' p2 and p3,and that hydrogen bonding is important in R1-pl interaction and apolar bonding in R2-p2 interaction and that the R3-p3 interaction involves the catalytic groups i.e. Ser-195 and His-57. The original interesting observa- tion by Hein and Niemann84 that ~-3-methoxycarbonyl-3,4-dihydroiso-quinolin-1-one (24)is a 2oO-(kc,) to 4000-(kcaJKd times better CT substrate than the L-analogue (25) and of comparable reactivity to 'normal' substrates has stimulated several authors to consider whether the methoxycarbonyl group in this substrate is required axial or equatorial to the heterocyclic ring for CT-catalysis to occur.Awad Neurath and Hartley8' favoured the axial hypothesis on the grounds that the enzyme would by unable to discriminate between equatorial D-and M. L. Bender and F. C. Wedler J. Amer. Chem. SOC.,89 3052. D. M. Glick Biochemistry 1968,7 3391. G. Schoellman and E Shaw,Biochemistry 1963,2,252 84 G. Hein and C. Niemann Proc. Nat. Acad. Sci. U.S.A. 1961,47 1341. " E. S. Awad H. Neurath and B. S. Hartley. J. Bid. Chem. 1960,235 FT 35. Enzyme Mechanisms L-isomers because they contain the ester group in a similar spatial location.Recent studies by Lawsons6 on the CT-hydrolysis of substrates with highly restricted conformation have also supported the axial hypothesis and some work by Balleau and Chevalliers7 has added further substantiation. The latter authors found that the R-S isomer of the 2,2'-bridged biphenyl analogue of benzoylphenylalanine (26) was not susceptible to hydrolysis by CT but iso- merised in solution to give the conformer (27) which has an axial methoxy- carbonyl group and is a substrate for CT action. Conversely Erlangers8 concluded from studies employing sterically con- strained inhibitors of CT and reactivators of diethylphosphoryl-CT that the equatorial conformer of (24) is the active species. This conclusion is sup- ported by recent work by Silver and Sones9 on the stereospecificity of CT towards equatorial and axial p-nitrophenyl esters of 3-t-butylcyclohexane- carboxylic acids.Cohen and his co-workersgO* also favour the equatorial 91 hypothesis and like Hein and Niemann visualise the active site of CT as containing three binding sites (28). These authors consider the am interaction 86 W. B. Lawson J. Biol. Chem. 1967,242 3397. 87 B. Belleau and R Chevalier J. Amer. Chem SOC. 1968,90 6864. B. F.Erlanger Proc. Nat. Acad. Sci. U.S.A. 1967,s 703. a9 M.S.Silver and T. Sone J. Amer. Chem SOC. 1968,90 6193. S. G. Cohen L. H. Klee and S. Y. Weinstein J. Amer. Chem SOC. 1966,88,5302;S. G. Cohen Z Neuwirth and S. Y. Weinstein ibid. p. 5306;S. G.Cohen R M.Schultz and S. Y. Weinstein ibid. p. 5315. S. G. Cohen and R. M. Schultz Proc. Nat. Acad. Sci. U.S.A.,1967,57 243. 620 M. R. Hollaway unimportant in the CT-catalysed hydrolysis of (24) since ~-3,4-dihydroiso- coumarin-3-carboxylate the oxygen analogue of (24) is also a good substrate for CT.” Of especial significance is the observation by Cohen and Milovano- vic92 that the h site is unable to accommodate a methyl group. A different approach to stereospecificity in CT-reactions has been developed by Ingles and Kno~les.~~~ 94 The deacylation rates (k 3) of acyl-CT derivatives were found to decrease in the order N-acetyl-L-phenylalanyl > N-acetyl-L-tryptophanyl %-N-acetyl-L-leucy % N-acetylglycyl. However the k + values for the corresponding D-isomers were much smaller and the above order was reversed.The results were interpreted in terms of a three-site interaction as in (28). Ingles and Knowles emphasized the fact that specificity is expressed in the catalytic rather than in the binding steps. These authors have presented further evidenceg4 for the hydrogen bonding site (am) by comparing the ratio of deacylation rates for the L-and D-isomers of a series of substrates in which hydrogen-bonding capacity was systematically decreased. The free energy of formation of the hydrogen bond with the substrates employed was estimated at 4 kcal./mole and since this is not expressed as a difference in bonding of derivatives with and without hydrogen-bonding potential it was suggested that CT binds a high-energy conformation of the substrate (an aspect of the strain theory discussed by Jencks16).The X-ray crystallographic studies14*65 are rapidly approaching the refinement where it will be possible to discriminate between the various postulates for substrate binding. Thus studies on the CT-N-formyl-L-trypto- phan complex have revealedg5 that the tryptophan ring is located in a hydro- phobic pocket known as the ‘tosyl-hole’ (since this is the position occupied by the p-tolylsulphonyl group in tosyl-CT). This would correspond to the ar site in (28). Ser-189. at the bottom of the ‘tosyl hole’ is hydrogen bonded to the nitrogen atom of the tryptophan ring of the ligand and further interactions are with Tyr-146 which is C-terminal in the B chain Met-192 and the carbonyl group of Gly-193 which is hydrogen-bonded to the amido-nitrogen atom of the ligand and so may represent the am site.It seems likely that the non- specific hydrophobic site at the active centre of CT described by Canadyg6* 97 and his co-workers and the hydrophobic locus with which the p-nitrophenyl- sulphonyl group in the Ser-195 substituted enzyme interacts described by Kallos and A~atis,~~ are both identical to the ‘tosyl-hole’. McClure and Edelman” have observed that the non-competitive inhibitor 92 S. G. Cohen and A. Milovanovic J. Amer. Chem SOC. 1968,90,3495. 93 D. W. Ingles and J. R Knowles Biochem J. 1967,104,369. y4 D. W. Ingles and J. R Knowles Biochem J. 1968,108 568. 95 D. M. Blow Proc. Roy. SOC. December 1968 in preparation 96 R Wildnauer and W.J. Canady Biochemistry 1966,5,2885; A. J. Hymes D. A. Robinson and W. J. Canady J. Biol. Chem. 1965 240 134. 9’ G.Royer and W. J. Canady Arch. Biochem Biophys. 1968,124 530. 98 J. .Kallos and K. Avatis Biochemistry 1966,5 1979. 99 W. 0.McClure and G. E. Edelman Biochemistry 1967,6 559 567. Enzyme Mechanisms 62 1 of CT 2-p-toluidinylnaphthalene-6-sulphonate(TNS) fluoresces when bound to the enzyme. Since TNS is only fluorescent in non-polar solvents this finding represents evidence for a second hydrophobic regon in the active site of CT probably distinct from the tosyl hole. The fluorescent TNS peak was not observed with benzylsulphonyl-CT or with chymotrypsinogen. Another interesting study on the active site of CT using a chromophoric probe has been carried out by Hille and Koshland.'OO A catalytically active CT derivative was prepared in which Met-192 was substituted with a 2-aceta- mido-4-nitrophenol group.The spectrum of this derivative suggested that the chromophore was in a polar environment and the pH-dependence of the spectrum indicated that (i) the phenolic group of the chromophore was a stronger acid when com- bined with Met-192 (its pK shifts from 6.1 to 5-8); (ii) titration of a group in the enzyme of pK 7.6 alters the spectrum of the phenolate anion of the reporter group; and (iii) there is another group pK ca. 9 which also affects the spectrum. Also substrate binding or covalent substitution of Ser-195 abolishes the dependency of the spectrum on the group of pK 7.6.Hille and Koshland'" attributed the ionisation of pK 7.6 to the imidazolium group of a histidine residue presumably that of His-57 and suggested that the anomalously high value was due to the proximity of the phenolate anion of the reporter group. As indicated by these authors the finding of a polar region in the active site of CT is not incompatible with other studies which indicate non-polar areas but merely serves to illustrate the heterogeneous nature of the active site. The use of n.m.r. spectroscopy for the investigation of CT-substrate-analogue interactions has also proved informative. Thus Spotswood Evans and Richards' O' observed that the binding of N-acetyl-D-p-fluorophenylalanine to CT was attended by a downfield shift in the 19Fresonance consistent with the binding of this competitive inhibitor in a hydrophobic pocket.Gerig,lo2 in investigating the binding of tryptophan to CT by 'H n.m.r. observed that complex formation was accompanied by line-broadening in the alkyl and aromatic regions of the tryptophan spectrum. The extent of broadening suggested that binding was tight enough for tryptophan to assume the rota- tional characteristics of the enzyme. Line broadening of the tryptophan 'H n.m.r. spectrum was not apparent when di-isopropylphosphoryl-Ser-195or S-(N-3-trifluoromethylphenyl)carbamoylmethyl-Met-192 derivatives of CT were employed. Complementary to this study is the investigation of the binding of DL-N-trifluoroacetylphenylalanine to CT carried out by Zeffren and Rea~i1l.l'~ A downfield shift in the "F n.m.r.peak was considered compatible with the location of the trifluoroacetyl group in a polar environment when bound to the enzyme. loo M. B. Hille and D. E.Koshland,jun.,J. Amer. Chem SOC.,1967,89,5945. lo' T. McL. Spotswood J. M. Evans and J. H. Richards J. Amer. Chem SOC. 1967,89,5054. J. T.Gerig J. Amer. Chem SOC. 1968,90 2681. lo' E. Zeffren and R E. Reavill Biochem Biophys. Res. Comm. 1968,32 73. M. R. Hollaway An important contribution to the understanding of CT catalysis has recently been made by Bernhard and Rossi.lo4 These authors have discussed the im- portance of conformation differences between native and acyl-CT derivatives. Also some elegant experiments were described indicating that the rate-limiting step in the deacylation of indoleacryloyl-CT involves the formation of a tetrahedral intermediate.Direct observation of an acyl-enzyme intermediate in the CT hydrolysis of a specific substrate has been reported by Miller and Bender.'" Thus incuba- tion of N-(2-furyl)acryloyl-~-tryptophan (FAT) with CT at low pH gave a derivative considered to be acylated enzyme. The first-order breakdown of this derivative to give free enzyme and FAT was dependent on a group of pK 6.95 required in the base form. Furthermore k, values for the CT hy- drolysis of FAT methyl ester also exhibited a sigmoid pH dependence pK 6.95 and the absolute values of k, were only slightly less than those for the breakdown of the low pH CT-FAT intermediate.Miller and Bender concluded that the CT-catalysed hydrolysis of FAT methyl ester at neutral pH values involved a single kinetically important intermediate namely the acyl-enzyme. Elastase. Elastase another representative of the serine proteases exhibits the usual characteristics of these enzymes.56* 57 In particular the pig pancreatic enzyme contains a single reactive serine residue per molecule which is in identical amino-acid sequence to the reactive serines in chymotrypsin-A and trypsin.lo6Furthermore the remainder of the elastase sequence is so extensively homologous with the other pancreatic serine proteases that Hartley and his collaborator^^^^ have suggested that they have all evolved from a common ancestor. Elastase also shows some functional similarity to the other serine protease in that it catalyses a biphasic release of p-nitrophenol from p-nitrophenyl pivalate,'08 a finding consistent with the formation of an acyl-enzyme inter- mediate.As in the case of chymotrypsin-catalysed reactions both acylation and deacylation steps exhibit large deuterium isotope effects and are dependent on an ionising group of pK 6.7 required in the base form.'os However elastase is less specific in its action than chymotrypsin in that it readily catalyses the hydrolysis of bonds involving the carboxy-group of a side variety of neutral amino-acids including Leu Ile Val Ala Ser Gly Tyr and Gln.Io9 An extremely rapid X-ray crystallographic determination of the structure of pig pancreatic tosyl-elastase at 3.5 A resolution has recently been accom- plished by Watson and Shottonl'O (a concise account of this and other work lo4 S.A. Bernhard and G. L. Rossi in 'Structural Chemistry and Molecular Biology,' ed. A. Rich and N. Davidson W. H. Freeman and Co. San Francisco 1968,p. 98. lo5 C. G. Miller and M. L. Bender J. Amer. Chem SOC. 1968,90,6850. lo6 M.A.Naughton F. Sanger B. S. Hartley and D. C. Shaw Biochem. J. 1960,77 149. lo' L.R Smillie and B. S.Hartley Biochem. J. 1966,101,232;J. R Brown D. L. Kauffman and B. S. Hartley ibid. 1967,103 497. lo* M. L. Bender and T. H. Marshall J. Amer. Chem SOC. 1968,90,201. log M. k Naughton and F. Sanger Biochem J. 1961,78 156. 'lo H.C.Watson and D. M. Shotton Proc. Roy. SOC.,December 1968,in preparation.Enzyme Mechanisms presented at the British Biophysical Society meeting December 1968 on 'Structural Aspects of Enzymatic Activity' has been given by Johnson' '). This work shows that the high degree of homology between chymotrypsin-A and elastase is reflected in the close similarity of their crystallographic struc- tures. Especially significant is the fact that of the 72 residues in the interior of the elastase molecule 78 yi &is homologous with the interior chymotrypsin residues whereas the 169 external residues are oniy 200,; homologous. Also the arrangement of His-57 Asp-102 and Ser- 195 in tosyi-elastase although not identical to that in chymotrypsin could allow the operation of the type of 'charge-relay system' suggested for the functioning of the latter enzyme.110* In view of the many similarities in structure and function between chy- motrypsin and elastase it would be surprising were they to catalyse reactions by different mechanisms.Thus it becomes of interest to account for their differing specificities. This is almost certainly connected with the fact revealed by the X-ray structure"'* ' '' that the entrance to the 'tosyl hole' in elastase is blocked by the bulky side chain of Val-216. This does not occur in chymotryp- sin where residue 216 is a glycine so that the major difference between the enzymes is probably in the presence of an ar site in chymotrypsin (28) but not in elastase. Trypsin. The extensive homology between chymotrypsin and trypsin has been discussed by several authors.60* '12 'l3 In particular Smillie and his co-workers' l3 have pointed out that the amino-acid sequences of chymo- trypsinogen A chymotrypsinogen B and trypsin exhibit a common pattern of invariant non-polar residues which are almost certainly located in the interior of the molecules.This implies that these enzymes have very similar structures confirmed by the fact that it has proved possible to build a model of trypsin based on the crystallographic structure of a-chymotrypsin. l4 These considerations together with many functional ~imilarities,'~ strongly suggest that chymotrypsin and trypsin have a common catalytic mechanism. The chief difference between trypsin and chymotrypsin is in their speci- ficities chymotrypsin hydrolyses bonds involving the carboxy-groups of amino-acids with bulky non-polar side-chains whereas trypsin shows specificity for bonds involving amino-acids with positively charged side-chains (for a review see ref.59). This suggests the participation of a negatively charged group on the enzyme in substrate binding. Smillie et u!.''~ consider this group to be the carboxylate anion of Glu-188 (corresponding to Ser-186 in the chymotrypsinogen sequence) on the basis that this is the only substitution of a negatively charged group for a neutral one within a sequence of otherwise invariant non-polar residues. Furthermore the corresponding residue in L. N. Johnson F.E.B.S.Letters 1969,2 201. B. S. Hartley J. R Brown D. L. Kauffman and L. B. Smillie Nature 1965,207 1157.L. B. Smillie A. Furka N. Nagabhusan K. J. Stevenson and C. 0.Parkes Nature 1968,218 343. P. B. Sigler D. M. Blow B. W. Matthews and R Henderson J. Mol. Biol. 1968,35 143. M. R.Hollaway chymotrypsin lies close to the active centre. However in view of the uncer- tainties in discriminating between aspargine and aspartic acid residues,65 it would also seem that the replacement of Sex-189 in the chymotrypsinogen sequence with what has been considered to be an asparagine in trypsinogen could also represent the crucial amino-acid substitution. This replacement is particularly attractive since Ser-189 in chymotrypsin is located114 at the bottom of the 'tosyl-hole' which seem likely to be the region involved in the expression of chymotrypsin specificity.Some elegant experiments on the interaction of trypsin with the com- petitive inhibitor' l5 benzamidine (29) as measured by difference spectroscopy have been reported by East and Trowbridge.'16 The dependency of the dif- ference spectra on benzamidine concentration at different pH values was consistent with the competition between inhibitor and protons for a single site in the trypsin molecule. This site was characterised by a pK value of 4.6 (29) R' = H R2= H (30) R' = any1 or alkyl R2 = H (31) R' = H R2= phenoxyalkoxy consistent with that expected for the ionisation of a side-chain carboxy-group. East and Trowbridge' l6also made the interesting observation that benzamidine binds to di-isoprop ylphosphoryltrypsin albeit more weakly than to native enzyme but not to trypsinogen or to trypsin modified by reaction with p-tolylsulphonyl-L-lysyl chloromethyl ketone.Baker and Erickson' '' have studied the inhibition of trypsin action by a number of substituted benzami- dines (30) and (31) and shown that although substitution at the R2 position (31) did not greatly diminish binding substitution at R' (30) gave a large decrease in affinity. The latter result would be anticipated on steric grounds if the benzamidine derivatives were to bind in the modified 'tosyl-hole'. This would correspond to the hydrophobic slit suggested by Mares-Guia Shaw and C~hen.~~~*''~ 'ls M. Mares-Guia and E. Shaw J. Biol. Chem. 1965,240 1579. E. J. East and C. G. Trowbridge Arch.Biochem. Biophys. 1968 125 334. 11' B. R. Baker and E. H. Erickson J. Medicin Chem. 1967,10 1123. '18 M. Mares-Guia E. Shaw and W. Cohen J. Biol. Chem. 1967,242,5777; M. Mares-Guia and E. Shaw ibid. p. 5782. Enzyme Mechanisms 625 An important finding by Beeley and Neurath'lg is that the reaction of the active centre histidine residue of trypsin (His-46 in the trypsinogen sequence corresponding to His-57 in chymotrypsinogen) reacts with bromoacetone at a similar rate to that of the model compound a-N-benzoyl-L-histidine methyl ester. In the resulting substituted trypsin derivative Ser-183 had lost its unusual reactivity towards acylating reagents. However His46 was still reactive towards bromoacetone in di-isopropylphosphoryl-Ser-183-trypsin.These findings serve to emphasize that in the absence of substrate the catalytic residues of trypsin do not possess exceptional reactivity and that the type of 'charge-relay system' represented in (19) is unlikely to be present in the free enzyme. Elmore and Smyth12' have made the interesting observation that the rate of deacylation of a-N-p-tolysulphonyl-L-lysyl-trypsin is 260 times that of a-N-methyl-a-N-p-tolylsulphonyl-L-lysyl-trypsin. This finding illustrates the importance of the am site interaction (28) during catalytic steps in trypsin reactions and is thus reminiscent of chymotrypsin-catalysed reactions.93* 94 The elegant rapid-reaction studies by Bernhard and Gutfreund12' using chromophoric substrates and dye-displacement techniques have shown that the simple three-step mechanism (17) is probably insufficient to describe trypsin-catalysed reactions.Some of the evidence for this conclusion is given in the excellent book by Bernhard7 together with a discussion of possible mechanistic consequences. Subtilisin. The extracellular bacterial proteases subtilisin Carlsberg and subtilisin BPN' (probably identical to subtilisin Novo'~~) are similar to other serine proteases in their reaction with acylating agents'23-' 27 and involvement of a histidine residue in cataly~is.'~~*~~* In addition the specificity of the subtilisins resembles that of chymotrypsin in that they preferentially hydrolyse derivatives of N-acyl-L-amino-acids with aromatic side chains.' '27 Ho wever 39 although the amino-acid sequences of Carlsberg and BPN subtilisins are 70 % homologous neither exhibits any obvious homology with the chymo- trypsin or trypsin sequences.' 24 Wright Alden and Kraut'29 have recently determined the crystallographic structure of benzylsulphonyl-subtilisinBPN (PMS-subtilisin) at 23A resolu-tion.The roughly spherical molecule which has few structural features in li9 J. G. Beeley and H. Neurath Biochemistry 1968,7 1239. 120 D. T. Elmore and J. J. Smyth Biochern. J. 1968 107 97. 12' H. Gutfreund and S. A. Bernhard Proc. Roy. SOC.,December 1968 in preparation. lZ2 S. A. Olaitin R J. DeLange and E. L. Smith J. Biol. Chem. 1968,243,5296. 123 A. 0.Barel and A. N. Glazer J. BioZ. Chem 1968,243 1344. 124 E. L. Smith F. S. Markland C. B.Kasper R. J. DeLange M. Landon and W. H. Evans J. Bid. Chem. 1966,241 5974. lZ5 A. N. Glazer J. Biol. Chem. 1968,243 3639. 126 K. E. Neet and D. E. Koshland Proc. Nat. Acad. Sci. U.S.A.,1966,56 1606. 127 A. N. Glazer,J. Biol. Chem. 1967,242,433. 12* L. Polgar and M. L. Bender Biochemistry 1967,6,610. 129 C. S. Wright R. A. Alden and J. Kraut Nature 1969 221 235. 626 M. R. Hollaway common with chymotrypsin contains eight a-helical regions involving 3 1% of the residues and much of the hydrophobic interior is built up from the side-chains of residues arranged in the parallel pleated-sheet structure. The most striking feature however is that the reactive Ser-221 residue lies close to His-64 whose imidazole 3-nitrogen atom is within hydrogen-bonding dis- tance of Asp-35.This arrangement is reminiscent of the Ser-195 His-57 Asp-102 arrangement seen in the chymotrypsin structure (19) and its occurrence in the active sites of enzymes with otherwise widely different structures indicates that it plays a key role in the catalytic processes. Wright Alden and Kraut'29 have also calculated a difference electron density map between PMS-subtilisin BPN and the native enzyme. This map reveals that the imidazole ring of His-64 which is within 3.5 A of the oxygen atom of Ser-221 in the native enzyme is rotated through approximately 80"in the PMS-enzyme to a position 4 A away from its original location. It is also noteworthy that the side-chain of Met-222 moves by about 1 A on substitution of Ser-221 a similar movement to that of Met-192 in chymotrypsin on substitution of Ser-195 with a tosyl group.I4 The essential serine of subtilisin has been chemically converted into a cysteine residue by two groups of workers.12'*'27 The product thiolsubtilisin possessed less than 1 % of the esterolytic activity of the native enzyme when specific substrates were employed,'26 although exhibiting comparable activity if substrates with good leaving groups (such as p-nitrophenol) were used.These findings were contrary to those expected on the basis of the superior nucleo- philicity of the thiol group and so Koshland and Neet'26 attributed the loss of activity in the thiol-enzyme to the larger van der Waals radius and different bond angle of sulphur failing to conform to the stringent steric requirements of the subtilisin active site.A not unrelated interpretation has been offered by Wang,20 who has suggested that subtilisin functions uiaa catalytic mechanism involving facilitated proton transfer along rigid accurately-held hydrogen- bonds. The sulphur atom because of its larger size and bond angle is considered unable to take part in the correct steric arrangement. Thiol Proteas=.-The thiol proteases are low molecular weight proteolytic enzymes containing a single thiol group per molecule essential for catalytic activity. Reviews of earlier studies with these enzymes have been given by Cunningham" and by Kimmel and Smith.'30 The 2.8 A resolution X-ray crystallographic structure of papain the most extensively studied of the thiol proteases has recently been published13' and confirms many of the conclusions reached by chemical studies.Papain. An examination of the detailed X-ray structure of papain deter- mined by Drenth and his c~llaborators,'~~ has shown that it is necessary to 30 E. L. Smith and J. R Kimmel in 'The Enzymes,' vol. IV,ed. P. D. Boyer H. Lardy and K. Myrback Academic Press New York 1960 p. 133. 131 J. Drenth J. N. Jansonius R. Koekoek H. M. Swen and B. G. Wolthers Nature 1968,218 929; J. Drenth et al. Proc. Roy. Soc. December 1968 in preparation. Enzyme Mechanisms 627 amend the amino-acid sequence proposed by Light Frater Kimmel and Smith.132 In this report the revised numbering from the X-ray sequence is employed.' 31 Prior to the X-ray crystallographic studies a considerable body of evidence existed for the implication of a thiol group' 303 132*'33 and the imidazole group of a histidine residue' 34-' 36 in papain-catalysed reactions.Furthermore the active-centre thiol group has been identified as that of Cys-25 and convincing evidence has been presented that this group is acylated by substrate during the course of cataly~is.'~' The amino-acid sequence'32 in the vicinity of this group is given in (32) together with the strikingly similar sequences around the essential thiol groups of ficin'38 and bromelain.'39 25 Papain Lys-Asn-Gln-Gly-Ser-Cys-Gly-Ser-Cys* -Ficin Arg- Gly-Gln-Gly -Gln-Cys- Gly- Ser- Cy s*-Stem Bromelain -Am-Gln-Asp-Pro-Cys-Gly-Ala-Cys*-(32) It is noteworthy that Ser-24 in papain is replaced by alanine in bromelain and is thus unlikely to play a crucial role in catalysis emphasizing the value of comparative studies of enzymes which operate by similar catalytic mechanisms.The earlier somewhat indirect kinetic evidence' 34 for the involvement of a histidine residue in papain catalysis has recently been substantiated by the elegant chemical modification studies of Husain and LOW^.'^^^ 13' Th ese authors demonstrated that the reaction of papain with the bifunctional reagent 1,3-dibromoacetone led to the intramolecular cross-linking of the thiol group of a cysteine residue to N-1 of a histidine residue indicating that these groups are within 5 A of each other. Sequencing studies identified these residues as Cys-25 and His-158.A similar procedure was employed to determine the sequences around the active centre histidine and cysteine residues in brornelain.l3' The histidine sequences for the two enzymes are given in (33). -_--_ I----Papain -Val-Asp+ His*- Ala-Va$ Ala~Al~Val-IGly-Tyr-/ 1158 I I IIlll l I I I Ill I Stem Bromelain XiHis*-Ala-Val~-Thr-jAla+ L---IleJGly-Tyr-/ --- - - - -. __ (33) i32 A. Light R Frater J. R. Kimmel and E. L. Smith Proc. Nut. Acud. Sci. U.S.A. 1964,52 1276. 133 S. S. Husain and G. Lowe Chem Comm. 1965,345. lJ4G. Lowe and A Williams Biochem J. 1965,96,194; A W. Lake and G. Lowe ibid. 1966,101 402. 13' S. S. Husain and G. Lowe Chem Comm. 1968 310; S. S. Husain and G. Lowe Biochem. J. 1968,108,861. S.S. Husain and G. Lowe,Biochem J. 1968,108,855; G.Lowe,Proc. Roy. SOC. December 1968 in preparation. IJ7 G. Lowe and A. Williams Proc. Chem. SOC. 1964 140; G.Lowe and A. Williams Biochem. J. 1965,96 189; L. J. Brubacher and M. L. Bender J. Amer. Chem. SOC.,1966,88,5871. 13' R C. Wong and I. E. Liener Biochem. Biophys. Res. Comm. 1964,17,470. lJ9 S. S. Husain and G. Lowe Chem. Comm. 1968 1387; Li-Pen Chao and I. E. Liener Biochem. Biophys. Res. Comm. 1967,27 100. 628 M. R.Hollaway The group X is not an aspartic acid residue,'39 so that Lowe and Husain have concluded that Asp-157 is unlikely to play a crucial role in papain catalysis. Wallenfels and EiseleI4* have also suggested the presence of a histidine residue in the active site of papain largely on the basis of a bell-shaped pH profile for the rate of inhibition of the enzyme by (-)-L-a-iodopropionic acid The prototropic groups giving the bell-shaped curve had pK values of 4.0 and 7-82 and were designated as a carboxy-group and an imidazolium group respectively.It was proposed that electrostatic interaction between the latter group and the carboxylate anion of the inhibitor served to orientate the inhibitor favourably for reaction with the essential thiol group. Kinetic studies have also led Cohen and PetraI4' to suggest the involvement of a histidine residue in the deacylation of (a-N-benzoyl-L-citrulliny1)papain. The detailed X-ray structure of papain'3' confirms many of the findings from chemical investigations. The molecule is made up from a single polypep- tide chain of 21 1or 212 residues folded so as to form two 'wings' separated by a marked cleft.The active site lies at the surface of the cleft and contains the imidazole side-chain of His-158 at a distance of 4 A from the sulphur atom of Cys-25. Trp-176 is also in the active site close to His-158 an observation of particular interest since Shinitzky and G01dman'~~ have obtained evidence from fluorometric measurements for an indole-imidazolium charge-transfer interaction in papain. Examination of the structure of papain given in ref. 131 shows that His-81 the only other histidine residue in papain is not near to a tryptophan so that the charge-transfer interaction is almost certainly between His-158 and Trp-176. Shinitzky and G~ldman'~' have also calculated from the dependence of fluorescence on pH that the interacting imidazolium group has a pK value of 7.22.Other groups in the active site identified from the X-ray stru~ture'~' are Gln-19 Asp-64 and Asp-157. Furthermore His- 158 is hydrogen-bonded to residue 174 (122 in the sequence of Light et which has been identified as an a~paragine.'~~ However in view of uncer- tainties in distinguishing between aspartic acid and a~paragine.~' residue 174 could be an aspartic acid thereby providing an interesting parallel to the serine proteases (18). Drenth and his collaborators' 31 have also calculated a difference electron- density map between native papain and papain substituted at Cys-25 by reaction with tosyl-L-lysyl chloromethyl ketone (TLCK) an irreversible inhibitor with some substrate-like pr0~erties.I~~ In this derivative the lysyl side-chain is seen to point towards A~p-64,'~' suggesting a possible role for this group in orientating substrates with positively charged side-chains during catalysis.The presence of the large cleft in the papain in the active site region had been predicted by some earlier studies by Schechter and Berger.'44 140 K. Wallenfels and B. Eisele European J. Biochem. 1968,3,267. 14' W. Cohen and P. H. Petra Biochemistry 1967,6 1047. 142 M.Shinitzky and R Goldman European J. Biochem. 1967,3,139. 143 B. G. Wolthers F.E.B.S. Letters 1969,2,143. 144 I. Schechter and A. Berger Biochem Biophys. Res. Comm.,1967,27 157; 1968,32 898. Enzyme Mechanisms 629 From a comparison of the rates of papain-catalysed hydrolysis of a series of pairs of diastereoisomeric alanine peptides (up to six residues) it was concluded that the active site was able to accommodate seven amino-acid residues.'44 The binding subsites were designated S S3 S2 S, S; Si and Sl, where the bond cleaved is between S and S;.In the extension of these investigations to map the subsites Schechter and Berger'44 have demon- strated that the S2 site has a high affinity for a phenylalanyl side-chain. Different mechanisms for papain action have been proposed by Husain and L~we'~~ The mechanism proposed by Husain and and by S1~yterman.l~~ L~we,'~~ which is similar to that proposed by Wang2' for chymotrypsin action starts with enzyme in which the thiol group of Cys-25 is hydrogen bonded to the imidazole side-chain of His-158 (34).Acylation of the thiol group by substrate is then con~idered'~~ to take pkce by a process involving a four-membered transition state as represented schematically in (34). H I H i I His-158 I 'Cys -25 ' \ (34) \ (35) In the deacylation step the leaving group X is replaced by water or another nucleophile YH and the imidazole group functions as a general base. Evidence for His-158 acting as a general base in deacylation has been presented by L0~e.l~~ This is based on the observation that the pK value of a group required in the deprotonated form for deacylation to occur decreased from 4-65 in water to 4-15 in 20% dioxan a shift expected for a cationic rather than a neutral acid i.e.an imidazolium rather than a carboxylate ionisation. Sl~yterrnan'~~, however favours a mechanism (35) whereby the imidazolium form of His-158 functions as a general acid catalyst at the substrate carbonyl oxygen atom during the rate-limiting acylation of Cys-25. The key difference between this mechanism (35) and that of Lowe (36) is that His-158 is required in the acid form during acylation whereas in Lowe's mechanism it is required in the base form. Hence the discrimination between the mechanisms depends on 145 L. A. Ae. Sluyterman Proc. Roy. SOC.,December 1968 in preparation. M. R. Hollaway H2 Hh . the interpretation of pH profiles of rate constants for separate processes in catalysis.Such a discrimination is not possible at the moment because of the conflicting conclusions'46~ 147 about which step is rate-limiting in the papain- catalysed hydrolysis of benzoly-L-arginine ethyl ester probably the most in- tensively studied substrate for papain. As pointed out by Brocklehurst Crook and Whart~n,'~~~ non-productive binding of this substrate to papain could lead to the assignation of incorrect values of rate constants for acylation and deacylation. Further complications could also ensue from the suggestion by Henry and Kirsch14* that the three-step mechanism may be insufficient to describe some papain-catalysed reactions although this appears not to be the case149 with benzyloxycarbonylglycine aryl esters as substrates.Despite these considerations however it seems likely from the data of Whitaker and Bender that acylation of papain by substrate is dependent on two ionising groups of pKk values 4-3and 85 one required in the acid form and the other in the base form. It remains to identify the systems contributing these ionisations un- ambiguously and to establish their exact role in catalysis. Other thiol proteases. The other less extensively studied thiol proteases resemble papain in both structure and function.59* 13* The similarities of the amino-acid sequences in the vicinity of the active-site cysteine residues of fi~in,'~' br~melain,'~~ and papain'32* 1339 13* have already been indicated (32) and there is convincing evidence that the thiol groups of these residues are acylated during ~atalysis.'~' Furthermore a histidine residue is a component of the 146 J.R Whitaker and M. L. Bender. J. Amer. Chem SOC. 1965. 87.2728. 14' L. A. Ae. Sluyterman Biochim. Biophys. Acta 1968 151 178. 148 A. C. Henry and J. F. Kirsch Biochemistry 1967,6 3536. 148 'K. Brocklehurst E. M. Crook and C. M. Wharton F.E.B.S. Letters 1968 2 1969. C. D. Hubbard and J. F. Kirsch Biochemistry 1968,7 2569. 14' Enzyme Mechanisms 631 active site in all three enzymes'35. and the amino-acid sequences around these residues in papain and bromelain are similar.' 39 These considerations together with the overall similarities in amino-acid composition,' 50* 15' strongly suggest that these enzymes have a common catalytic mechanism.' 51 like papain,143- lS3are It is noteworthy that fi~in'~~.and br~melain,'~~ inhibited by the chloromethyl ketone derivatives of tosyl-L-phenylalanine (TPCK) and tosyl-L-lysine (TLCK) through reaction at the active centre thiol groups. Also each enzyme reacts faster with TLCK than with TPCK although these compounds react with cysteine at similar suggesting that TLCK binds to the enzymes so as to facilitate subsequent reaction with the essential thiol group. As with papain,'46 evidence has been presented'54 that the acylation of ficin is dependent on two ionising groups of pK values 4.4 and 8.5 at 25" and I = 01 one of which is required in the protonated form and the other in the deprotonated form. It is tempting to assign the ionisation of pK 8.5 to the active centre thiol residue which has been shown'55 to have a pK value of 8.55 (25" I = 0-1)but it is not possible to decide from available data whether this group is required in the acid or the base form for catalysis to occur.How- ever in view of the greater nucleophilicity of the ionised form of thiols it seems likely that the base form of the thiol group would be required in acylation which means that the group of pK 4.4 would be required as an acid. Another thiol protease known to contain a histidine residue in the active site,' 56 is the so-called streptococcal proteinase. Of particular interest was the finding by Gerwin157 that whereas the rate of inhibition of this enzyme by chloroacetamide gave a sigmoid pH-dependence a bell-shaped pH-profile was obtained with chloroacetate ion as inhibitor.Furthermore at pH 4.2 the rate with chloroacetate ion was at least one hundred times the rate with chloro- acetamide. These results which suggest' 57 the presence of an ionising cationic group in the active centre of the enzyme are essentially similar to those ob- tained in the reaction of papain with the (-)-L-antipodes of a-iodopropionate ion and a-iodopropionamide. 140 Carboxypeptidase A.-Reviews of earlier work on this enzyme are by ningham5' and Vallee.' 58 Carboxypeptidase A (CPA) specifically catalyses the hydrolysis of peptide bonds at the carboxy-terminus of polypeptide chains. It is a small extracellular enzyme of molecular weight 34,600 containing a single zine atom per molecule essential for catalysis.However the derivatives S. S. Husain and G. Lowe. Biochem. J.. 1968.110 53. P. T. Englund T. P. King L. C. Craig and k Walti Biochemistry 1968,7 163. 152 W. J. Stein and I. E. Liener Biochem Biophys. Res. Comm. 1967,26 376. 15' T. Murachi and K. Kato J. Biochem (Japan) 1967,62,627. M. R Holliaway European J. Biochem 1968,5 366. M. R Hollaway A P. Mathias and B. R Rabin Biochim. Biophys. Act4 1964,92,111. T-Y.Liu J. Biol. Chem 1967,242,4029. 15' B. I. Gerwin J. Biol. Chem. 1967,242,451. "* B. L. Vallee Fed. Proc. 1964,32 8. 632 M.R.Hollaway of CPA in which Zn2 + is replaced by Mn2 +,Co2+,or Ni2 + are still catalytic- ally active.158* lS9 The elegant X-ray crystallographic studies by Lipscomb and his co-workers'60 have resulted in the determination of the structure of bovine CPA at 2 A resolution.The molecule contains a single disulphide link and 30% of the 307 amino-acid residues are in an a-helical arrangement with a further 20% in a twisted P-pleated sheet conformation of either the parallel or the antiparallel type. As with some of the other hydrolytic enzymes there is a pronounced cleft in the molecule at the active site.'60 This observation confirms the prediction by Abramowitz Schechter and Berger16' from peptide 'mapping' studies that the active site of CPA would be able to accom- modate a substrate comprising 5 residues i.e. should be at least 18 A in length. The zinc atom is bound within the cleft through the side-chains of residues His-69 Glu-72 and Lys- 196 the fourth co-ordination position being occupied by a water molecule.Lipscomb and his collaborators' 6o have also calculated difference electron density maps for CPA and CPA containing bound glycyr-L-tyrosine or L-~YSY~-L-tyrosinamide and have made the following striking observations. (a) The tyrosine side-chain corresponding to the normal C-terminus of a substrate is located in a pocket sufficiently large to accommodate a tryptophan side-chain. This pocket is located at the end of the cleft. (b) The guanidinium side-chain of Arg-145 moves through about 2A to form an ionic bond with the carboxylate anion of the substrate. This movement is largely effected by rotation about the C(P)-C(y) bond of Arg-145. (c) The water molecule bound at the active centre zinc atom is displaced by the carbonyl oxygen of the substrate peptide bond.(d) In the glycl-L-tyrosine-CPA complex the carboxy-side-chain of Glu-270 is hydrogen-bonded to the amino-group of the glycyl residue. However model-building has shown that in substrates with more than two amino-acid residues this interaction may not occur and that Glu-270 may then approach the peptide bond which is bound at the zinc atom. The glycyl-Glu-270 inter- action could explain the stability of the CPA-glycyl-L-tyrosine complex. (e) On binding glycyl-L-tyrosine the phenolic oxygen atom of Tyr-248 moves through a distance of 12 A as a result of a shift in the peptide backbone and rotation about the C(ol)-C(P) bond. (f) The zinc atom moves through about 1 A.159 R. C. Davies J. F. Riordan D. S. Auld and B. L. Vallee Biochemistry 1968,7,1090. M. L. Ludwig J. A. Hartsuck T. A. Steitz H. Muirhead,J. C. Coppola G. N. Reeke and W. N. Lipscomb Proc. Nut. Acud. Sci. U.S.A. 1967,57,511;G.N. Reeke J. A. Hartsuck,M. L. Ludwig F. A. Quiocho T. A. Steitz and W. N. Lipscomb ibid. p. 2220; W. N. Lipscomb in 'Structural Chemistry and Molecular Biology,' ed. A. Rich and N. Davidson W. H. Freeman and Co. San Francisco 1968 p. 38; W. N. Lipscomb et al. Proc. Roy. SOC. December 1968 in preparation; W. N. Lipscomb J. A. Hartsuck G.N. Reeke F. A. Quiocho P. H. Bethge M. L. Ludwig T.A Steitz H. Muirhead and J. C. Coppola Brookhaven Symp. Biol. 1968,21 in the press 16' N. Abramowitz I. Schechter and A.Berger Biochem Biophys. Res. Comm. 1967,29,862 Enzyme Mechanisms (g) The conformation changes do not take place on the binding of poly- peptides in which the C-terminal carboxy-group is blocked e.g. with L-lysyl- L-tyrosinamide.'60 This suggests that the movements of CPA side-chains are triggered off by the substrate-carboxy-group-to-Arg-145ionic bond possibly by disruption of the hydrogen-bonding around this residue.' 6o A diagrammatic representation of the postulated interactions' 6o in a CPA-substrate complex is given in (36). The foregoing observations seem to provide one of the most clear-cut examplesof Koshland's 'induced-fit' theory 162 whereby the specific interaction between an enzyme and its substrate results in a conformation change in the enzyme protein so as to give a catalytically active structure.(37) R' \ 2+. Zn- / I Glu-270 On the basis of the structural work Lipscomb and his co-workers'60 have suggested two possible mechanisms. In the first (37) the carboxylate anion acts as a nucleophile and attacks the carbonyl carbon atom of the C-terminal peptide bond of the substrate. This is facilitated by the zinc atom acting as a Lewis acid at the carbonyl oxygen atom and removal of the leaving group by interaction with the phenolic hydroxy-group of Try-248 acting as an acid. This mechanism therefore involves an intermediate anhydride of Glu-270 and the acyl portion of the substrate. lci2 D. E. Koshland jun.,Fed. Proc. 1964,23 719. 634 M. 4.Holloway In the second mechanism involving a single displacement (38) the carboxy- late anion of Glu-270 functions as a general base in abstracting a proton from a water molecule as it adds to the substrate carbonyl carbon atom.The latter mechanism is favoured'60 on the grounds that CPA does not catalyse trans- peptidation reactions. Vallee and his co-workers163 have presented a schematic model based on multiple modes of substrate binding which attempts to explain the complex kinetic behaviour of CPA towards ester and peptide substrates. This model has correctly predicted a reciprocal effect of bound ligands on esterase and peptidase activities of CPA. Ribonuc1ease.-Bovine pancreatic ribonuclease A (RNAse A) molecular weight 13,600 comprising a single polypeptide chain of 124 residues catalyses the hydrolysis of internucleotide phosphodiester bonds in ribonucleic acids only if the phosphate of the susceptible P-0 bond is attached to the 3'-OH of a pyrimidine nucleotide.The reaction is known to proceed with the inter- mediate formation of 2'3'-cyclic phosphates which are themselves substrates for ribonuclease action. Reviews of earlier work are by Anfin~on'~~ and by Mathias Deavin and Rabin.' 65 Kinetic evidence has been presented'66 for the implication of two imidazole groups in RNAse-catalysed reactions and this has been substantiated by the results of chemical modification studies 67 which have identified the active- site imidazole residues as those of His-12 and His-119. A mechanism involving concerted acid-base catalysis by these groups has been suggested by Rabin and his c~llaborators.'~~~ 166 The detailed X-ray crystallographic structures of RNAse and RNAse Si70 have been published and confirm many of the chemical predictions.In the structure of RNAse A presented by Harker and his co-~orkers,~~~ the active site region which contains a bound phosphate ion is at the surface of a distinct cleft and contains the side-chains of residues His-12 His-1 19 Lys-7 Gln-11 Lys-41 and His-48. The 3.5 A resolution structure of RNAse S (a derivative of RNAse A with the Ala-20-Ser-21 peptide bond cleaved by treat- ment with subtilisin) published by Richards and his collaborators' 70 closely resembles the RNAse A structure apart from the region of Ala-20 and Ser-21 which are separated by 10-15 A in RNAse S.163 B. L. Vallee J. F. Riordan J. L. Bethune T.L. Coombs D. S. Auld and M. Sokolovsky Bio-chemistry 1968,7 3547. 164 C. B. Anfiison Brookhaven Symp. Biol. 196515 184. 165 A. P. Mathias A. Deavin and B. R. Rabin in 'Structure and Activity of Enzymes,' ed. T. W. Goodwin J. I. Harris and B. S. Hartley Academic Press New York 1964 p. 19. 166 D. Findlay D. G. Herries A. P. Mathias B. R Rabin and C. A. Ross Biochem J. 1962,85,152. 16' A. M. Crestfield W. H. Stein and S. Moore J. Biol Chem. 1963 US,2413. 16' A. Deavin A. P. Mathias and B. R Rabin Nature 1966,211,252; A. Deavin A. P. Mathias and B. R. Rabin Biochem J. 1966 101 14C. G. Kartha J. Bello and D. Harker Nature 1967,213,862; G. Kartha J.Bello and D. Harker in 'Structural Chemistry in Molecular Biology' ed A Rich and N. Davidson W. H. Freeman and Co. San Francisco 1968 p. 29. ''O H. Wyckoff K. D. Hardman N. M. Allewell T. Inagami L. N. Johnson and F. M. Richards J. Biol. Chem. 1967 242 3984 3749. 16' Enzyme Mechanisms 63 5 Electron-density maps for the complexes of RNAse S with the competitive inhibitors uridine 2'(3')-phosphate and 5-iodouridine 2' (3')-phosphate have also been calculated' 70 and the following locations of active-site residues established. (a)His-12 and His-119 are close to the nucleotide binding site. Furthermore His-119 is freely accessible to solvent whereas His-12 is partly buried with the N-1 atom inaccessible to solvent. This observation is consistent with the fact that iodoacetate ion reacts at the N-3 atom of this residue.'67 (b) The uracil portion of the inhibitors is located in a groove beneath the histidine residues bounded by residues Val-43 Phe-120 Thr-45 and Ser-123.Position 5 of the uracil ring points towards the liquid and is close to Ala-122. (c) Gln-11 and Asn-44 are above the histidine residues and Asp-121 within a slot to the left of His-119 (positions as depicted in ref. 170). (d) The E-amino-groups of Lys-7 and Lys-41 are in the active site region as predicted by chemical st~dies,'~' and could move close to the bound nucleo- tide. It is noteworthy that there are no marked conformation changes in the enzyme on binding the inhibitors."' More recent work by Richards Wyckoff and their collaborators has extended the crystallographic data for RNAse S and several RNAse S-sub- strate analogue complexes to the level of 2 A resolution (an account of this work was given by N.Allewell at the British Biophysical Society Meeting Dec. 1968 on 'Structural Aspects of Enzymatic Activity' which has been reported by Johnson'"). Thus the electron density map of the RNAse S-cytidine 3'-phosphate complex reveals that the 2-OH group of the ribose ring is hydrogen bonded to His-12 and the 2- and 6-positions of the cytosine ring are hydrogen-bonded to Thr-45. These interactions probably account for the specificity of the enzyme for pyrimidine nucleotides. His-119 appears to be displaced by the phosphate group of the inhibitor towards Asp-121.A schematic representation of these interactions is given in (39). An additional interaction shown in (39) is that of the E-NH group of Lys-7 with the N-7 atom of the guanine base in the guanosine nucleotide binding site. Such an interaction has been predicted on the basis of spectrophoto-metric investigation of the binding or guanosine nucleotides to RNAse.'" The crystallographic studies are thus consistent with the predicti~n'~~.'~~~ 168 that catalysis by RNAse involves concerted acid-base catalysis by the imidazole side chains of His-12 and His-119. A schematic representation of a possible mechanism,' ''which may involve the formation of a pentacovalent phosphate intermediate is given in (40). The formation of this intermediate (39) would be 17' C.H. W. Hirs Brookhaven Symp. Biol. 1962 15 154; P. S. Marfey M. Uziel and J. Little J. Biol. Chem 1965,240 3270. 17* A. Deavin R Fisher C. M. Kemp k P. Mathias and B. R Rabin European J. Biochem 1968,7 21. W M. R.Holloway facilitated by the interactions with His-12 and His-119 as indicated by the elegant studies of model systems by Westheimer and his collaborators.17 + Lys-7 -NH2 I H 0 I-r12 I His-12 173 F. H. Westheimer Accounts Chem. Rex 1968 1 70 Enzyme Mechanisms Some other interesting studies involving the histidine residues of RNAse are those by Jardetzky and his colleagues.'74 These authors have resolved the C(2r-H n.m.r. absorptions of the histidine residues and from the change in position of the peaks with pH determined the individual pK values of these groups.Furthermore the binding of cytidine 3'-phosphate to RNAse was shown to shift the pK value of His-119 from 5.8 to 7.4 and of His-12 from 6-2 to 8.0. These results have been discussed in terms of a model involving the binding of the dianion of the inhibitor to the active site of RNAse with both His-12 and His-1 19 in the protonated form. 174 D. H. Meadows and 0.Jardetzky Proc. Nut. Acad. Sci. U.S.A. 1968,61,406; D. H. Meadows J. L. Markley J. S. Cohen and 0.Jardetzky ibid. 1!?67,58,1307;D. H. Meadows 0.Jardetzky R M. Epand H. H. Ruterjans and H. A. Scheraga ibid. 1968.60 746.