3 Reaction Mechanisms Part (iii) Enzyme Mechanisms By M. AKHTAR and D. C. WILTON Dept. of Physiology and Biochemistry University of Southampton SO9 5NH 1 Introduction We have continued our practice of reviewing only those developments in Enzyme Mechanisms which fall within well-defined areas. In this chapter the previous reviews on ‘P,yridine Nucleotide Linked Reactions’ and ‘Active Site Directed Inhibitors’2 have been up-dated. In addition. we have introduced for the first time a section on mechanistic aspects of ‘Flavin-linked Enzymic Reactions’. Because enzyme mechanisms cover large areas of organic chemistry biochemistry and physiology it has not been possible for us to do justice to all the important developments in this field. We are particularly conscious of not having covered literature on the use of e.s.r.n.m.r. and fast reaction techniques as extensively as some may regard as necessary. We hope that this omission will be put right by the next reporter. 2 Flavin-linked Enzymic Reactions D-Amino-acid 0xidase.-Riboflavin (1) after being enzymically converted into either flavin mononucleotide (la) or flavin adenine dinucleotide (1 b) takes part in several biological redox reactions. The precise mechanism through which the flavin coenzymes participate in these reactions has been extensively investigated3 during 40 years since the first flavoprotein old yellow enzyme was dis~overed.~ The most popular view3 until 1970had been that flavin-linked biological reactions occur through the direct transfer of hydrogen atom(sj between the substrate and the coenzyme oxidized flavin + RH reduced flavin + R Then Hamilton made the interesting suggestion’ that the overall oxidation- reduction observed in flavin-linked enzymic reactions may not be due to direct ’ M.Akhtar and D. C. Wilton Ann. Reports (B) 1970 67 557. M. Akhtar and D. C. Wilton Ann. Reports (B) 1971 68 167. For review see (a) D. Wellner Ann. Rev. Biochem. 1967 36 669; (b) P. Strittmatter Ann. Rev. Biochem. 1966 35 125; (c) K. Yagi Ado. Enzymol. 1971 34 41 ;(4A. H. Neims and L. Hellerman Ann. Rev. Biochem. 1970,39,867. 0. Warburg and W. Christian Biochem. Z. 1932,254 438. L. E. Brown and G. A. Hamilton J. Amer. Chem. SOC. 1970,92,7225. 98 Reaction Mechanisms -Part (iii) Enzyme Mechanisms CH,-(CHOH)3-CH2-O-X I 0 (1) X = H; oxidized flavin 0 II (la) X = -P-O-;oxidized flavin mononucleotide I 0- 00 I1 II (lb) X = -P-0-P-0-CH adenine;oxidized flavin adenine dinucleotide t--f OH OH hydrogen transfer but the indirect consequence of a multi-step reaction sequence.The hypothesis is illustrated in Scheme 1 with reference to the enzyme D-amino-acid oxidase which catalyses the reaction RCH(NH2)COZH + 02 + RC(O)CO,H + NH + H202 (1) The first step in the mechanism postulated by Hamilton is the formation of a covalent adduct (4) between the substrate and flavin which then decomposes to give the reduced coenzyme (5) and the imino-compound (6). The latter (6) on subsequent hydrolysis gives the keto-acid.The intermediary role of an imino-acid in the D-amino-acid oxidase reaction was established by conducting the enzymic reaction in the presence of tritiated NaBH, when labelled alanine accumulated.6 During the past two years results obtained with substrate analogues have given further insight into the structure of the substrate-flavin adduct and the mechanistic sequence involved in transformations catalysed by D-amino-acid oxidase and lactic oxidase. It has been shown that D-amino-acid oxidase which catalyses the physiological reaction of equation (l) can also utilize b-chloroalanine (8)and the carbanion of nitroethane (10)as substrates. With B-chloroalanine u-amino-acid oxidase gives two types of product. In the presence of molecular oxygen the reaction is similar to that with the physiological substrate.and p-chloropyruvate is formed (8)-+(9). In the absence of oxygen however a novel reaction occurs7 which results in the elimination of HCl and the formation of pyruvate (8)-+(7). in an oxygendependent reaction the carbanion of nitroethane is converted into acetaldehyde and NO,-(10)+(1 1) by D-amino-acid oxidase.' Nitroethane itself is not a substrate for the enzyme. ' E. W. Hafner and D. Wellner. Proc. Nut. Acad. Sci. U.S.A. 1971 68 987. C. T. Walsh A. Schonbrunn and R. H. Abeles J. Biol.Chem. 1971 246 6855. * (a) D. J. T. Porter J. G. Voet. and H. J. Bright J. Biol. Chem. 1973 248 4400;see also (6) D. J. T. Porter J. G. Voet and H. J. Bright J. Bid. Chem. 1972 247 1951. M.Akhtar and D. C. Wilton Me CO,H \/ c -+ /\ H NH (3) (2) Throughout ;formula (2) represents 'C0,H oxidized flavin coenzyme bound Me (4) to the appropriate enzyme. Me Me \ (2) + H,O 0 + //C-CO,H -!% \C=O + NH, / HN HO,C (6) (7) (5) Scheme I 0 II NH + HCI + Me-C-CO,H I HI /aerobic (7) CH -C-C0,H AH \? 0 II (8) NH + H,O + H0,C-C-CH,CI 0 Na+[Me-CH-NO,] -% Me-Cc + NO,-+ Na' + H,O Several mechanisms for the action of D-amino-acid oxidase which may also permit rationalization of reaction pathways followed by the artificial substrates (8) and (10) have been ~onsidered.~ In the Reporters' view however the suggestion' of Porter et al.,which assumes that in the metabolism of the carbanion of nitroethane the adduct (12)is formed merits special attention.The outstanding feature of this mechanism (Scheme 2) is the proposal that an early step in the enzymic reaction is nucleophilic attack by the carbanion of nitroethane at N-5of the flavin moiety to give the adduct (12). The further conversion of the latter '1 Reaction Mechanisms -Part (iii) Enzyme Mechaiiisms adduct (12) into products then occurs by unorthodox reaction steps shown in Scheme 2. The extension of this mechanism to the physiological reaction using H’ -Me ‘-1 J i-(2) 0,(11) + (5) Me-C-H I CN (14) Scheme 2 D-alanine will involve the participation of an additional a-C-H bond cleavage step prior to the formation of the adduct (15)(Scheme 3). The further decompo- sition of the latter adduct (15)then furnishes the reduced flavin (5)and the imino- acid (6).The keto-acid is produced from the latter by hydrolysis and the original structure of the flavin moiety is regenerated through the involvement of 0,. When the mechanism is applied to the rationalization of the results obtained with P-chloroalanine the introduction of a modification is necessary. The Reporters’ opinion is that in the physiological-type reaction (Scheme 3) oxygen is not involved in the final step but participates more directly in the decomposition of the adduct (15) as shown in Scheme 4. With this modification the initial adduct (16) formed from P-chloroalanine decomposes with the involvement of O=O to give products as shown in the oxidative pathway of Scheme 5.It is now apparent why in the absence of oxygen the same adduct (16)may decompose by an alter- native pathway (anaerobic pathway Scheme 5)to give pyruvate HCl and NH,. It is interesting to note that in the anaerobic metabolism of P-chloroalanine labelled with tritium in the a-position the product pyruvate contains a substantial amount of tritium at the P-p~sition.~ This experiment suggests that the hydrogen atom removed in the first step of the conversion is transferred to the P-position of the product without exchange with the protons of the medium. a-Amino-P- chlorobutyrate is also accepted as a substrate by the D-amino-acid oxidase but M. Akhtar and D. C. Wilton + ’(5; + N NHi ‘Me (2) + H,02 * + H02C NH %(7) + NH, Me> (5) Scheme 3 + (6)+ (7) + H (15) Scheme 4 in this case the reaction follows predominantly an anaerobic pathway leading only to the formation of a-ketobutyrate.’ The support for the existence of a covalent complex involving the a-carbon atom of the substrate and N-5 of the coenzyme comes from two types of experi- ments.It has been shown that when the reaction of D-amino-acid oxidase with the carbanion of nitroethane (10)is performed in the presence ofCN- the enzyme is inactivated and 1 mole of CN-per mole of flavin is incorporated into the enzyme.8” The inactivation has been explained 8u by assuming that CN-reacts with the Schiff-base intermediate (13) to give a ‘dead-end’ complex (14). Using ’ C.T. Walsh E. Krodel V.Massey and R. H. Abeles J. Biol. Chem. 1973 246 1946. Reaction Mechanisms -Part (iii) Enzyme Mechanisms I z-u, + h X t? + h + c + a t X 0 u em E I ps'l M. Akhtar and D. C. Wilton spectroscopic techniques several laboratories have reported that the interaction of D-amino-acid oxidase with D-alanine results in the formation of a species having absorption properties characteristic of a 'dihydroflavin type' of chromo-phore.3c Qualitatively similar species are formed through the interaction of D-amino-acid oxidase with either P-chl~roalanine'~ or cr-amino-P-chlorob~tyrate.~ When substrates labelled with deuterium at the a-position are used the formation of the spectral species is attended by a three-fold kinetic isotope effe~t.~~.~-' The isotope effect data are consistent with the formulation of the initial complex as shown in structure (15),rather than with structures of the type (4)or (17) since the formation of the former complex requires the cleavage of the a-C-H in the crucial stage of the reaction.No C-H bond is labilized in the formation of either (4)or (17). It should be noted however that the isotope effect data are in accord with another formulationg for the substrate-flavin complex (18) therefore the mechanistic discussion should be considered prelimi- nary pending the reliable elucidation of the structure of the substrate-flavin adduct. Me-C-CO,H I H (17) An unpalatable aspect of the mechanism proposed in Scheme 3 is the cleavage of the relatively non-activated C-H bond of alanine in the initial stage of the reaction.A careful examination of the biological literature however reveals other precedents for the removal of a proton from the a-position of carboxylic acids. One may argue that such a deprotonation may be aided by enhancing the electron-withdrawing property of the carboxylic group by complexing the substrate with enzyme active-site groups which decrease the ionization of the 0-H bond and also co-ordinate with the carbonyl group as in Scheme 6. H 0-I/ -c-c I No Scheme 6 lo V. G. Voet D. J. T. Porter and H. J. Bright Z. Nuturforsch 1972,27b 1054 as quoted in ref. 8a. K. Yagi M. Nishikimi N. Ohishi and A. Takai F.E.B.S.Letters 1970,6 22. Reaction Mechanisms -Part (iii) Enzyme Mechanisms Flavoenzyme L-Lactate 0xidase.-Another flavin-linked enzyme L-lactate oxidase catalyses the oxygen-dependent conversion of L-lactic acid into acetic acid and CO MeCH(OH)CO,H + 0 + MeC0,H + CO + H,O (2) Unlike the D-amino-acid oxidase reaction shown in equation (l) this process does not form hydrogen peroxide. The type of mechanistic approach detailed above for the amino-acid oxidase reactions has been extended to the study of L-lactate 0~idase.l~ It has been shown that Q-chlorolactate is a substrate for L-lactate oxidase and may be converted into two different types of products. In an oxygen-dependent decarboxylation chloroacetate and CO are generated while molecular oxygen is reduced to the level of H,O (19)-+ (20);alternatively Q-chlorolactate can be converted in a reaction independent of oxygen into pyruvate and chloride ion (19) -+(7).In the latter conversion the a-proton of the substrate is transferred to the Q-position of the product without exchange with the protons of the medium. The conversion of the physiological substrate OH 0 P aI CO + H,O + CI-CH,-CO,H 2CH,-C-H -6H,-&-CO,H + HCI I1 I II C1 C0,H H O L-lactate labelled with deuterium at the a-position is attended by an isotope effect kH kD = 1.8 thus suggesting that an early step in the substrate oxidation by L-lactate oxidase is the abstraction of the a-hydrogen atom. The above reactions of Q-chlorolactate with the lactate oxidase are strikingly similar to the reaction of Q-chloroalanine with both D-and L-amino-acid oxidases.In the light of this fact a mechanism similar to that suggested for amino-acid oxidases is outlined for lactate oxidase reaction Scheme 7 except that in this case the for- mation of the decarboxylated product necessitates the assumptions that both H,O and pyruvate remain bound to lactate oxidase and that the enzyme contains lactate + -+ 01 -+ (2) + MeC0,H (2) BH Me-C-CO,H + H,O + H,O + CO (22) Scheme 7 W. B. Sutton J. Biol. Chem. 1957 226 395 and references therein. l3 C. Walsh 0. Lockridge V. Massey and R. H. Abeles J. Biol. Chem. 1973 248 7049. M. Akhtar and D. C. Wilton the additional activity for catalysing their conversion into acetate and CO,.This deduction is consistent with the earlier finding that one of the oxygen atoms of acetate orginates from molecular oxygen. l4 N-Methylglutamate Synthetase.-The involvement of an oxidized flavin moiety has recently been established for another enzyme catalysing the interconversion of glutamate (23) and N-methylglutamate (27). When viewed casually the reaction of equation (3)appears to involve a direct displacement of -NH by MeNH- ; however the knowledge of the compulsory requirement of oxidized flavin for the conversion necessitates the consideration of a more complex reaction pathway. L-glutamate +methylamine S N-methylglutamate+ammonia (3) The ability of the flavin moiety to form covalent intermediates in lactate and amino-acid oxidase reactions may explain this interconversion as shown in Scheme 8.The first step is the formation of a substrateflavin complex (24) which undergoes elimination to give NH and the Schiff base (25). The reversal of the B:) C R’ QHz {=-\+ N H I MeNH, -*(2) +R-C-CO,H -I B-H ‘C-CO,H NHMe / NHMe (27) R =HOzC(CHz)z-R Scheme 8 I4 0.Hayaishi and W. B. Sutton J. Amer. Chem. SOC.,1957 79 4809. Is R. J. Pollock and L. B. Hersh. J. Biol. Chem. 1973 248,6724. Reaction Mechanisms -Part (iii) Enzyme Mechanisms overall sequence but using MeNH instead of NH 3 provides a simple mechanistic basis for the interconversion involving the oxidized flavin moiety in a major catalytic role. In the enzymic reaction when glutamate labelled with tritium at the a-position is used all the radioactivity is retained in the product thus suggesting that the hydrogen atom removed in the conversion (23)+(24) is transferred to N-methylglutamate without exchange with the protons of the medium.3 Pyridine Nucleotide-linked Dehydrogenases The dehydrogenases were last reviewed in this series' at a tantalizing stage when the amino-acid sequences coupled with the high-resolution X-ray structures of a number of enzymes were nearing completion. However very few conclusions could then be drawn regarding the individual amino-acid residues at the active site that were involved in binding and catalysis. This situation is now resolved in the case of lactic dehydrogenase. Lactic Dehydrogenase (LDH).-This enzyme catalyses the reversible reaction MeCHOHC0,H + NAD+ MeCOC0,H + NADH + H+ and at neutral pH has an obligatory binding order of coenzyme followed by substrate.The 2.0 A resolution X-ray structure and the primary sequence have now been worked out for the dogfish muscle M enzyme.16 The enzyme consists of four identical sub-units with a cleft present in each sub-unit in which coenzyme and substrate bind. A peptide loop residues 98-114 folds down over this active centre pocket in the ternary complex. The major events occurring during binding and catalysis are summarized below (see Scheme 9). (a) The adenine part of the coenzyme binds first to a hydrophobic pocket with its NN group pointing outwards to the enzyme surface. NAD' linked to Sepharose through this residue will still bind to LDH.l7 (b) The pyrophosphate group of the coenzyme is positioned by binding to residues in the cleft and is finally locked in place by a salt bridge with Arg-101. This residue moves through 13 A to achieve this interaction in the ternary complex. The two ribose residues are hydrogen-bonded to various groups in the cleft as shown in Scheme 9. (c) In order to facilitate the binding of NAD' it is suggested that the side-chain carboxyl of Glu-140 is available to balance the positive charge of the quaternary nitrogen in the nicotinamide ring. This carboxyl approaches the A face of the pyridine ring while the B face resides in a hydrophobic environment. It is the A face of the coenzyme which is positioned towards the substrate.*'M. J. Adams M. Buehner K. Chandrasekhar G. C. Ford M. L. Hackert A. Liljas, M. G. Rossman I.E. Smiley W. S. Allison J. Everse N. 0.Kaplan. and S. S. Taylor Proc. Nat. Acad. Sci. U.S.A. 1973. 70 1968. " K. Mosbach H. Guilford R. Ohlsson and M. Scott Biochern. J.. 1972 127 625. 108 M. Akhtar and D. C. Wilton Hydrophobic His OH OH coo-Asp 53 I * NYN NH Arg 109 NH I Arg 171 Scheme 9 A diagram of the active site of lactic dehydrogenase (d) The substrate is bound by its carboxyl through a salt bridge to Arg-171 while the methyl group is orientated so that at least one of the hydrogens extends towards the surface of the protein. It is noteworthy that phenyl pyruvate is a substrate analogue with the same V,, as pyruvate.I6 (e) His-195 in the ternary complex is correctly positioned to act as either a proton donor for pyruvate reduction or a proton acceptor for lactate oxidation.(f) The essential thiol group Cys-165 of lactate dehydrogenase,' which is situated at the bottom of the active site cleft is not concerned with the binding of substrate or the coenzyme. (g) There is no tryptophan residue' situated near the active site. The precise mechanism by which the chemical transformation is brought about may now be looked at in more detail. The first point that must be clarified is that whereas a negatively charged carboxy-group from Glu- 140 would facilitate the binding of NAD the presence of such a charge would tend to reduce the electron- accepting ability of the pyridine ring and hence deactivate the coenzyme.Arg- 109 is a loop residue which has moved through 23 A in the ternary complex. The Reporters therefore would suggest that a possible role for this residue is to attract the negative charge of the glutamate in the ternary complex so that the positive charge on the pyridine nitrogen of NAD' may be neutralized by hydride transfer from the lactate. However in the oxidation of NADH the movement of this glutamate residue towards the pyridine ring would facilitate the formation of the positively charged nitrogen. There is much interest regarding the stage at which the 'essential' His-195 is protonated or deprotonated during the reaction. Using o-nitrophenyl pyruvate as a substrate the reaction has been studied Reaction Mechanisms -Part (iii) Enzyme Mechanisms by monitoring the proton concentration of the medium." It was observed that in the reduction of the pyruvate analogue protonation of the histidine occurs after the binding of the NADH and is intimately linked with the binding of the substrate.Moreover it has been established that during the conversion of lactate into pyruvate the proton is conserved in the ternary complex and is only lost after the release of the pyruvate. l9 These studies also indicate that a group of pK 6.8 the 'essential' histidine is the group involved in proton uptake and release. The central role of this essential histidine at the active site of the enzyme coupled with a compulsory requirement for protonation at the time of pyruvate binding prior to the actual chemical transformation highlights the importance of the proton transfer step in the overall chemical reaction.On the basis of work carried out on the pyridine nucleotide-linked reduction of carbon-carbon double bonds it was shown that the first step in the chemical transformation is the activation of the substrate by an enzyme mediated protonation giving an electron deficient carbon centre which is subsequently neutralized by hydride transfer from the pyridine nucleotide.20 It would appear that a similar sequence (Scheme 10) of bond-forming events may operate in the reduction of carbonyl compounds as was suggested previously.'*21 Scheme 10 Liver Alcohol Dehydrogenase.-Although the primary sequence of this enzyme has been known for some time22 the X-ray structure at high resolution has only recently been el~cidated.~~ The enzyme is a dimer of two identical sub-units each of which is made up of two lobes of unequal size separated by a wide deep cleft that contains the active site.The coenzyme binding site is very similar to that observed with LDH. The smaller lobe of the sub-unit binds the adenine in a hydrophobic pocket while the rest of the coenzyme in the wide open conformation points down into the cleft. It is estimated that the nicotinamide ring is located in close in J. J. Holbrook Biochem. J. 1973 133 847. 19 J. J. Holbrook and H. Gutfreund F.E.B.S. Letters 1973 31 157. 20 D. C. Wilton K. A. Munday S. J. M. Skinner and M. Akhtar Biochem.J. 1968 106 803; I. A. Watkinson D. C. Wilton K. A. Munday and M. Akhtar Biochem. J. 1971 121 131; I. A. Watkinson D. C. Wilton A. D. Rahimtula and M. Akhtar European J. Biochem. 1971 23 1. 21 M. Akhtar D. C. Wilton I. A. Watkinson and A. D. Rahimtula Proc. Roy. SOC.,1972 B180,167. 22 H. Jornvall European J. Biochem. 1970 16,25. 23 C. 1. Branden H. Eklund B. Nordstrom T. Boiwe G. Soderlund E. Zeppezauer I. Ohlsson and A. Akeson Proc. Nat. Acad. Sci. U.S.A. 1973 70 2439. I10 M. Akhtar and D. C. Wifton proximity to one of the two zinc atoms which is located at the bottom of the active site cleft. It is to this zinc atom that the chelating agent 1,lO-phenanthroline a competitive inhibitor binds. The substrate binding site is thought to be located at the bottom of the active-site cleft.The other zinc atom is found near the surface of the sub-unit far removed from the active-site cleft and is completely surrounded by protein. Its function is unknown but it may be to maintain the proper conformation of the sub-unit. Cytoplasmic Malate Dehydrogenase.-The X-ray structure at 2.9 A resolution has now been determined24 and a preliminary examination of the picture suggests a striking homology with LDH which includes a similar coenzyme binding area. Glyceraldehyde-%phosphate Dehydroge-e.-The 3.0 A electron-density map of the lobster enzyme has now been determined and preliminary observations have been made concerning its structure.25 Perhaps the most exciting observation concerning this enzyme is the fact that Lys-183 of one sub-unit binds to the pyrophosphate of the coenzyme in the active site of the adjacent sub-unit thus creating the unique situation in which the catalytic centre contains residues from two different sub-units.It should be remembered that a feature of the binding of ligands to this enzyme is that although a tetramer it shows two-fold symmetry and behaves as a pair of dimers as far as its reaction with coenzyme and substrate is concerned.26 Of particular relevance is the fact that the coenzyme binding site is very similar to that found in LDH. The Role of Pyridine Nucleotides in Carbohydrate Transformations.-In the 1971 Annual Report’ attention was drawn to the involvement of enzyme-bound pyridine nucleotides in several interesting transformations of carbohydrates.It was reported that the enzyme uridine diphosphate galactose Cepimerase which catalyses the reaction UDP-glucose (28) TUDP-galactose (30) contains tightly bound NAD and that the conversion may involve an oxidation-reduction sequence occurring through the intermediacy of the 4-keto-compound (29). Subsequently a different mechanistic sequence involving C-3 was considered for the conversion (28) (30).2’ The latest evidence however supports the original mechanistic proposal of Scheme 11. It has been shown28 that when UDP- galactose is incubated with substrate amounts of UDP-galactose 4-epimerase the intermediate keto-compound can be trapped by the addition of NaB3H4. The chemical degradation of the reduced hexoses showed that all the tritium was associated with C-4 thus providing evidence for the involvement of a 4-keto- compound in the enzyme reaction.28 Alternatively two groups have shown that when physiological and ‘artificial’ sugar substrates containing tritium at C-4 were 24 E.Hill D. Tsernoglou L. Webb and L. Banaszak J. Mol. Biol. 1972 72 577. 25 M. Buehner G. C. Ford D. Moras K. W. Olsen and M. G. Rossmann Proc. Nar. Acad. Sci. U.S.A. 1973 70 3052. 2b 0. P. Malhotra and S. A. Bernhard Proc. Nut. Akad. Sci. U.S.A. 1973 70 2077 and references therein. 27 L. Davis and L. Glaser Biochem. Biophys. Res. Comm. 1971,43 1429. U. S. Maitra and H. Ankel J. Biol. Chem. 1973 248 1477. Reaction Mechanisms -Part (iii) Enzyme Mechanisms incubated with UDP-galactose 4-epimerase the tritium was transferred to enzyme-bound NAD.leading to the formation of a complex consisting of enzyme- [3H]-NADH-4-ketohexose from which free E3H]-NADH was isolated and characterized by conventional method^.^'*^' Radioactive label was not trans- ferred to NADH when the corresponding sugars containing tritium at C-3 were used. CH,OH - NAD Scheme 11 Another example of the involvement of pyridine nucleotides in carbohydrate transformations is the conversion of CDP-D-glucose (3 1) into CDP4keto-3,6- dideoxy-D-glucose (37) shown in Scheme 12. The overall conversion occurs through stepwise removal of OH-6 and OH-3.31 The former is achieved by the sequence (31) +(32)--* (33)+(34) involving oxidation at C-4 and concomitant formation of enzyme-bound NADH ;j?-elimination then gives the unsaturated derivative (33) which upon reduction by NADH is converted into CDP-4-keto- 6-deoxy-~-glucose (34).In this process the hydrogen at C-4 of the substrate is transferred to C-6 of CDP4-keto-6-deoxy-~-glucose (34) via the coenzyme. The reduction of the latter compound (34) to the 3,6-dideoxy-derivative (37) requires the participation of NADH or NADPH. The reaction may therefore be con-sidered to involve a rather difficult displacement of the 3-hydroxy-group by an H-ion from the coenzyme. The recent demonstration however that 29 W. L. Adair. 0.Gabriel D. Ullrey and H. M.Kalckar J. Biol. Chem. 1973,248,4635; see also W. L. Adair 0. Gabriel D. Stathakos and H. M. Kalckar J.Biol. Chem. 1973,248,4640. 30 J. N. Ketley and K. A. Schellenberg Biochemistry 1973 12 315. 31 V. P. Gonzalez-Porque and J. L. Strominger Proc. Nut. Acad. Sci. U.S.A. 1972 69 1625. M. Akhtar and D. C. Wilton pyridoxamine-5'-phosphateis required as a cofactor for the conversion permits the consideration of an indirect reaction sequence. The first step in the conversion is the formation of the Schiff base intermediate (35). The presence of the pyridinium ion in the species (35) facilitates deprotonation at C* and elimination forming (36). The reduction of the conjugated double bond by NAD(P)H in (36) then occurs through a mechanism for which there are several precedents. OH OH (35) (34) (36) (37) + pyridoxamine-5-phosphate Scheme 12 Th central role of the coenzyme in bringing bout the movement of the loop residues in LDH (see above) highlights the importance of this ligand in the overall active-site conformation of pyridine nucleotide-linked enzymes.This point is illustrated by recent observations concerning the enzyme 6-phosphogluconate dehydrogena~e.~~ This enzyme catalyses the reaction (38) +(42) shown in 32 M. Rippa M. Signorini and F. Dallocchio J. Bioi. Chem. 1973 248 4920. 113 Reaction Mechanisms -Part (iii) Enzyme Mechanisms Scheme 13. Using a modified substrate (39)it was possible to isolate the predicted intermediate (41). The decarboxylation of (41)by the enzyme to give (43) required the presence of NADPH even though this coenzyme is not directly involved in the decarboxylation step.(38) R = OH (40) R = OH (42) R = OH (39) R = H (41) R = H (43) R = H Scheme 13 4 Active-site Directed Inhibitors The use of active-site directed alkylating inhibitors continues to be a major tool for studying enzyme mechanisms and identifying amino-acid residues at the active site. The purpose of this section is not to give a comprehensive list of compounds that have been used since this subject was last dealt with in these Reports2 but to highlight novel compounds and new methods of approach. The use of epoxides as alkylating agents has been further developed by Rose’s group as a means of investigating the mechanism of isomerases. Triose phosphate isomerase was inactivated by both the D-and L-isomers of glycidol phosphate (44);33 however inactivation by the D-isomer was ten times faster.Similarly the H2Y\ HC-0 I R (44) R = -CH,08 (45) R = -(CHOH)3CH20@ enzyme phosphexoisomerase was inactivated by 1,2-anhydro-~-mannitoI-6-phos-phate the R-isomer of the epoxide (45).34With both enzymes nucleophilic attack is by a glutamate residue at C-1 of the inhibitor. In the case of the triose phosphate isomerase the peptide sequence suggests that it is the same glutamate residue that is modified by both isomers of the epoxide. These results are consistent with a single base mechanism of proton transfer for both these enzymes (Scheme 14). Thus the carboxylate anion of the glutamate side-chain first removes a proton from the hydroxyl carbon to produce a cis-enediol intermediate.The reaction may 33 K. J. Schray E. L. O’Conneli and I. A. Rose J. Biol. Chem. 1973,248,2214. 34 E. L. O’Connell and I. A. Rose J. Biof. Chem. 1973 248 2225. 114 M. Akhtar and D. C. Wilton be completed by the carboxyl protonating the carbon atom originally present as the carbonyl. It is proposed that an electrophilic group which normally polarizes the carbonyl to allow proton transfer also acts as an acidic group to facilitate nucleophilic substitution at the epoxide (see Scheme 14). Another active-site H H I \ C-0-H C=& H-?-Em Enz -COO- _C I -+ Enz-COOH -110-fX-Enz W C-0-H I /c O R R CH,OH Enz-COO-I H -X -Em c=o I Scheme 14 directed epoxide that has been used successfully to alkylate a carboxy-group is the 2’,3‘-epoxypropyl P-glycoside of di-(N-acetyl-D-glucosamine) (46).This compound is an effective irreversible inhibitor of lysozyme and has been shown to modify the P-carboxy-group of A~p-52.~’ (N-acetyl glucosamine) -0 -CH -CH -CH, ‘d The use of a bifunctional and hence potentially cross-linking alkylating agent has always had the added attraction that it allows certain conclusions to be drawn concerning the spatial relationship of two amino-acid residues in the native protein. The covalent intermediate that is observed during the course of many enzymic reactions can often provide one of these points of attachment to the protein. A suitable electrophilic group on this intermediate may then react with amino-acid residues at the active site.The active-site serine of a-chymotrypin was acylated by the inhibitor 3,4-dihydro-3,4-dibromo-6-bromomethyl coumarin (47).36 The acetyl-enzyme (48) then readily underwent nucleophilic substitution at the bromomethyl group probably via a quinonemethide intermediate (48a). It would appear that His-57 of a-chymotrypsin was alkylated by this reagent. Another example of the use of a bifunctional alkylating agent is the inactivation of the enzyme 2-keto-3deoxy-6-phosphogluconicaldolase by brom~pyruvate.~’ 35 Y. Eshdat J. F. McKelvy and N. Sharon,J. Biol. Chem. 1973 248 5892. 36 J. J. Bechet A. Dupaix J. Yon,M. Wakselman J. C. Robert and M. Vilkas European J. Biochem. 1973 35 527. 37 H. P. Meloche J. Biol.Chem. 1973 248 6945. Reaction Mechanisms -Part (iii) Enzyme Mechanisms I15 Enz-X \ Br Br I c=o cH&Br 0 I 0 I (47) This inhibitor reacts with a carboxylate residue as well as with the &-amino-group of lysine which on subsequent treatment with borohydride gives a secondary amine. Thus the carboxylate and the lysine must be in close proximity at the active site of the enzyme as shown in Scheme 15. Br H I I I CH,-C-C0,-Br-YHz-t-coz-YHz-Y-CoZ-II 0 'NH BHd-0 0 I I &=o i" A Scheme 15 Isocyanates are now starting to be used as alkylating groups for affinity labelling of enzyme active sites. The mechanism of their reactivity is shown in Scheme 16. Butyl isocyanate (49) has been shown to alkylate specifically a cysteine residue at the active site of yeast alcohol dehydr~genase.~' Interestingly this is not the same 'reactive' cysteine that is alkylated by iodoacetamide.H+ Me(CH,),-N-C=O "1 -P Me(CH,),-NH-C //O Y \ (49) s S I I Enz Enz Scheme 16 Acetylenic compounds are now finding considerable use as alkylating agents and they have proved to be powerful inactivators of decanoyl dehydra~e,~~?~' thio1ase:l y-cystathionase,42 and plasma amine ~xidase.~~ In all cases it is J. Twu and F. Wold Biochemistry 1973 12 381. 39 G. M. Helmkamp R. R. Rando D. J. H. Brock and K. Bloch J. Biol. Chem. 1968 243 3229. 40 M. Morisaki and K. Bloch Biochemistry 1972 11 309. *I P. C. Holland M. G. Clark and D. P. Bloxham Biochemistry 1973 12 3309.42 R. H. Abeles and C. T. Walsh J. Amer. Chem. SOC. 1973.95 6124. 43 R. C. Hevey J. Babson A. L. Maycock and R. H. Abeles J. Amer. Chem. SOC., 1973 95 6125. M. Akhtar and D. C. Wilton proposed that the enzyme first isomerizes the acetylene to an allene which is subject to nucleophilic attack by a suitable group at the active site. This nucleo- phile has been identified as a histidine in the case ofdecanoyl dehydra~e,~’ while a cysteine residue is proposed for thi~lase.~~ Scheme 17 proposed for thiolase Scheme 17 illustrates the mechanism of inactivation of enzymes by acetylenic compounds. Enzyme-generated allene intermediates have also been implicated in the inactiva- tion of plasma amine oxidase by 2-chloroallylamine43 (50),as shown in Scheme 18.- C1 H I ICH,=C-CHI IX IX ICCl b H I c‘ -+ CH,=C-CH -+ CH,=C=CHI 1 NH +NH +NH II II (50) C C /\ /\ pyridoxal enzyme? Scheme 18 In the 1971 report2 evidence was reported which suggested that catalytic activity of several proteolytic enzymes is greatly enhanced by interaction between the enzyme active site and groups of the substrates remote from the amide bond to be cleaved. Studies of this nature have now been extended to ela~tase.~~ which is a serine proteinase elaborated by the pancreas and is involved in the hydrolysis of elastin the insoluble protein of connective tissue.45 A systematic study of the types of product formed when several synthetic peptides of varying sizes are hydrolysed by elastase has led to the deduction44 that the enzyme active site possesses at least six sub-sites for the binding of substrates.At least four of these 44 D. Atlas and A. Berger Biochemistry 1973 12 2573; D. Atlas S. Levit I. Schechter and A. Berger F.E.B.S. Letters 1970 11 281; R. C. Thompson and E. R. Blout Biochemistry 1973 12 51 57 66 and references therein. 45 For a review see B. S. Hartley and D. M. Shotton in ‘The Enzymes’ ed. P. D. Boyer Academic Press 1971 Vol. 3 p. 323. Reaction Mechanisms -Part (iii) Enzyme Mechanisms 117 sub-sites are occupied by the N-terminal and the remaining two by the C-terminal residues of the substrate and hydrolysis occurs at the amide bond residing between sub-sites S and S; as shown in Scheme 19. The knowledge that the occupancy Ala-Ala-Ala-AlaFLys-Phe s s s s S,’ S,‘ Scheme 19 of sub-sites S,,S ,S3,and preferably S4 is mandatory for the ‘perfect’ interaction of the substrate with the active-site region has led to the synthesis of effective active-site directed inhibitors.Thus single amino-acid derivatives containing electrophilic centres such as (51)do not interfere with the activity of the enzyme;46 however chloromethyl ketone derivatives of tri- and tetra-peptides are potent inhibitors of enzyme action.47 The tetrapeptide analogue (53)is ca. 12 times better as an alkylator of elastase than is the tripeptide derivative (52). This order of 0 0 II II Tosyl -NH -CH(R)-C -CH -C1 Ac -(Ala),-Ala -C -CH $1 (51) (52) n = 2 (53) n = 3 reactivity of the inhibitors is similar to the rates of hydrolysis of the corresponding peptide substrates by elastase.Preliminary evidence suggests that the inhibitors (52) and (53) act by modifying a single histidine residue of the enzyme. Another type of inhibitor48 which does not inactivate the enzyme but binds tightly to elastase is the aldehyde derivative (54). The tight binding of the inhibitor to the enzyme has been attributed48 to the reversible formation of a tetrahedral inter- mediate (53,which closely corresponds to an intermediary stage in the hydrolysis of amide substrates. 0 /-Ac-Pro -Ala-Pro-NH -CHMe-C \ H (54) Enz-CH -OH H I R-C-OH I Enz-CH,-0 (55) 46 H. Kaplan V. B. Symonds V. B. Dugas and D. R. Whitkar Canad. J. Biochem.1970,48 649; L. Visser D. S. Sigman and E. R. Blout Biochemistry 1971 10 735. 47 (a)R. C. Thompson and E. R. Blout Biochemistry 1973 12 44; (b) J. C. Powers and P. M. Tuhy ibid. p. 4767. 48 R. C. Thompson Biochemistry 1973 12 47. M. Akhtar and D. C. Wilton Somewhat similar results had previously been obtained with leucine amino- peptidase. This enzyme cleaves N-terminal amide bonds in peptides and proteins and also hydrolyses synthetic substrates such as L-leucine-p-nitroanilide and ~-1eucinamide.~~ The chloromethyl ketone derivative of leucine (56) does not inactivate the enzyme but is a potent reversible inhibitor of the enzyme action.” The strong binding ofthe inhibitor to the enzyme may be rationalized by assuming that the carbonyl group of the inhibitor activated by the a-chloro-substituent reversibly forms tetrahedral intermediateas shown in theconversion (57a) -+(58a).For comparison the first step of the mechanism postu1ated5l for the hydrolysis of amide bonds by leucine aminopeptidase is also shown (57b) -+ (58b). Me 0 \ II CH-CH2-CH(NH,)-C-CH -CI / Me (56) NH2 \ /“ H CH \ M2cs-E I1 2,- X (57) a; X = CH,CI (58) a; X = CH,CI b;X = NH b;X = NH *9 For a review see R. J. Delange and E. L. Smith in ‘The Enzymes’ ed. P. D. Boyer Academic Press 1971 Vol. 3 p. 81. P. L. Birch H. A. El-Obeid and M. Akhtar Arch. Biochem. Biophys. 1972 148,447. ’ G. F. Bryce and B. R. Rabin Biochem. J. 1964,90 513.