3 Reaction Mechanisms Part (iii) Enzyme Mechanisms ~~~ By M. AKHTAR and D. C. WILTON Dept. of Physiology and Biochemistry University of Southampton SO9 5NH In order to cover adequately large areas included in enzyme mechanism new topics have been surveyed in each of the last two Reports. This approach is continued this year except that literature on coenzyme-B ,-dependent biological reactions originally reviewed' in 1970 has now been brought up to date. Section 2 'Conformations of Substrates and Conformational Changes at the Active Sites' complements last year's Report,2 which was exclusively devoted to the identification and regulation of groups involved in enzyme catalysis. A section on the stereochemistry of enzymic reactions and another on non-RNA-dependent synthesis of peptides have also been included.1 Coenzyme-B A common feature of most coenzyme-B (1) dependent biological reactions involves the counter transfer of a hydrogen atom and a leaving group between two adjacent carbon atoms I IH-CB-'C-X I 1+ X-CB-"C-H I I I I {-NAN 5' 5' CO"'j-CH,-R E (CO"' +CH,-R fi (1) Partial structure of coenzyme-B -* Me Me 5' N A Of Co"'+CH -R ' M. Akhtar and D. C. Wilton Ann. Reports (B) 1970 67 557. * M. Akhtar and D. C. Wilton Ann. Reports (B),1971 68 167. 140 Reaction Mechanisms-Part (iii) Enzyme Mechanisms OH OH R= in all coenzyme structures 0 Adenine The most extensively studied examples from a mechanistic viewpoint are the dioldehydrase- and ethanolamine deaminase-catalysed transformations (Scheme l).l H HO > <O; Dtoldehydrase R H R OH R H (2) R = Me or H (3) (4) Scheme 1 The experimental evidence available at the time of writing the 1970 Report permitted the deduction,' which has received strong support from recent work,3 that the crucial event in the coenzyme-B ,-linked transformations may involve the transfer of the migrating hydrogen atom of the substrate to the C-5' of the coenzyme to give rise to an enzyme-bound 5'-deoxyadenosine [see structure (13)] as a transitory intermediate.One of the hydrogen atoms of the methyl group of 5'-deoxyadenosine is subsequently transferred to form the product and regenerate the coenzyme. The cleavage of the carbon-cobalt bond during catalysis could occur through one of three possible dissociation mechanisms involving car- banion @) radical (9) or carbonium ion (10)species.(C~III)CH,-R (C+)CH,-R (cot)~H,-R (8) (9) (10) Recently attention has been focused mainly on the identification of the primary species formed upon the cleavage of the C-Co bond. It was shown that the enzyme glyceroldehydrase which normally converts glycerol into P-hydroxy- propionaldehyde but can also catalyse the analogous conversion (2)-B (4; R = Me) when mixed with propane-1,2-diol and coenzyme-B, gave rise to the appear- ance of absorption bands at 61 1 and 655 nm attributed to the formation of CO" cobalamin specie^.^ The concentration of the latter species increased to a steady- state concentration of 80 % of the enzyme-bound coenzyme present initially.T. H. Finlay J. Valinsky K. Sato and R. H. Abeles J. Biol. Chern. 1972 247 4197. S. A. Cockle H. A. 0.Hill R. J. P. Williams S. P. Davies and M. A. Foster J. Amer. Chem. SOC.,1972,94 27.5. M. Akhtar and D. C. Wilton When the substrate had been consumed the concentration of Co" cobalamin decreased and the coenzyme was reformed. In another experiment the e.s.r. spectrum of the enzymic reaction mixture was also measured and showed the appearance of two signals with g values of 1.944and 2.035 which were assigned4 by the authors to the biradical species (9). Although e.s.r. signals during co- enzyme-B ,-dependent enzymic reactions have been observed in several other lab~ratories,~ the concentrations of the unpaired electrons in these studies were low and the possibility that the radical species may be formed in a side reaction could not be excluded.In recent reports the demonstration of the presence of high concentrations of the radical species4 (accounting for about 25 % of the coenzyme) coupled with the spectral evidence3v4 for the existence of the Co" cobalamin intermediate lends support to the view that the initial dissociation of the carbon- cobalt linkage may occur through a homolytic mechanism as was originally suggested by Eggerer et aL6 Another approach for identifying the species formed on the dissociation of the C-Co bond was designed by Law et aL7 These workers took advantage of the fact that the 5,6-dimethylbenzimidazolylmoiety of coenzyme-B, may be substi-tuted by 4-hydroxy-2,2,6,6-tetramethylpiperidine N-oxyl to give the spin-label coenzyme-B, analogue (la) which can replace the natural coenzyme in the ethanolamine deaminase reaction (5)-+(7).The rationale behind this approach is that if radical intermediates are involved during the enzymic reaction addi- tion of the substrate to the spin-label coenzyme-B ,,-enzyme complex will result in the disappearance of the nitroxide e.s.r. radical owing to its presence in the vicinity of another radical species (1 l) during catalysis. The signal will however Co"'+CH2-R + HO Co".)CH2-R reappear when the substrate has been utilized. Experimentally this was found to be the case.7 For the remaining two possibilities the authors argued as follows.If heterolytic cleavage occurs to give a carbanion and a Co"' cobalamin the nitroxide e.s.r. signal will remain unchanged. In the case of heterolytic cleavage to a carbonium ion and a Co' cobalamin species the nitroxide will be expected to be displayed' as a ligand for the cobalt and its radical character irreversibly (a) J. A. Hamilton and R. L. Blakley Biochim Biophys. Acta 1969 184 224; (6) B. Babior and D. C. Gould Biochem. Biophys. Res. Comm. 1969 34 441; (c) J. A. Hamilton R. L. Blakley F. D. Looney and M. E. Winfield Biochim. Biophys. Acta 1969 177 374; (d) J. A. Hamilton R. Yamada R. L. Blakley H. P. C. Hogenkamp, F. D. Looney and M. E. Winfield Biochemistry 1971 10 347; (e) R. Yamada Y. Tamao and R. L. Blakley ibid. p. 3959. H. Eggerer F.Overath F. Lynen and E. R. Stadtman J. Amer. Chem. SOC.,1960 82 2643. P. Y. Law D. G. Brown E. L. Lien B. M. Babior and J. M. Wood Biochemistry 1971 10 3428. J. Brodie and M. Poe Biochemistry 1971 10 914. Reaction Mechanisms-Part (iii) Enzyme Mechanisms lost owing to reduction by the Co' species. The validity of this last assumption however must be questioned in view of a recent report in which it was shown that nitroxide radicals may not be reduced' by such strong reducing agents as Li(OR),AlH or LiAlH,. A different approach was used by the Brandeis group3 who argued that if a heterolytic cleavage resulting in the formation of a carbonium ion and a Co' species operates during enzyme catalysis the latter species owing to its nucleo- philic character may react with a suitable electrophilic reagent.Dioldehydrase which is involved in the reaction (2)-+(4:R = H) also catalyses an exchange of the C-5' hydrogen atoms of the coenzyme with C-2 hydrogen atoms of the product acetaldehyde. Chloroacetaldehyde can replace acetaldehyde in the hydrogen-exchange reaction ; since during the course of the reaction involving chloroacetaldehyde the displacement of the type ]p3 //O (CO.) CH -C -H was not observed the author concluded that a Co' species is not involved in coenzyme-B ,-linked reactions. The conclusion may be correct ;the rationale is however questionable in view of a recent report highlighting the fact that the presence of an alkylating agent in the vicinity of a nucleophilic centre at the enzyme active site does not necessarily lead to alkylation in preference to other modes of reaction." In spite of some reservations the evidence presented points to (but does not prove) a homolytic cleavage of the C-Co bond of the coenzyme as an early step in catalysis.In the light of this information the mechanistic sequence of Scheme 2 H (Co"'+CH,-R H H H Htf".H + 11 a HO H-OH + [(colI-);+] HO H (coii.)CH -R (2) .Ib (13) (1) = (9) :HHOH c H+---((H H H (coq OH H H \ \ H-C-R (15) (14) (co*~~)H-c-R / / H' 1Id H' "49 ' O€ (CO~~-)H-C-R / H' Scheme 2 D. J. Kosman and L. H. Piette Chem. Comm. 1969 926. lo H. P. Meloche M. A. Luczak and J. M. Wurster J. Biol. Chem. 1972 247 4186. M. Akhtar and D.C.Wilton may be considered3v4 for the reaction catalysed by dioldehydrase. Reaction a of Scheme 2 resulting in the abstraction of a hydrogen atom from the substrate by the coenzyme requires a special mention. In several other coenzyme-B,,- dependent reactions the equivalent stage requires the transfer of a hydrogen atom from a non-activated C-H bond of the substrate to the C-5’ of the coenzyme. A carbon with radical character as in (9) is no doubt better suited for this role than an equivalent ionic species. However in order to furnish the product the intermediate (14) must rearrange to (15) with the migration of a OH group which requires unorthodox behaviour from a radical species. If this mechanistic principle is extended to ribonucleotide reductasel even greater difficulty arises since in this case the equivalent stage requires the expulsion of OH as a hydroxyl radical.One must therefore invoke an ionic rearrangement preferably involving a C’ for reaction c of Scheme 2. Thus the current mechanistic view on B ,-dependent enzymes involves attribut- ing a dual role to the C-Co bond a homolytic bond fission involved in the hydro- gen-transfer step and a heterolytic cleavage permitting rearrangement displace- ment and elimination reactions during catalysis. A detailed discourse on the mechanism of action of B ,,-dependent enzymes with particular emphasis on /I-methylaspartate-glutamate mutase is available.’ Although in general the mechanistic features of coenzyme-B,,-dependent mutases of type Pa P R-CH,-CH,-NH R-CH(NH,)-be are similar”12 to those of the enzymes discussed above there appears to be one difference.In these cases the migration of the amino-group may occur via Schiff-base formation with carbonyl compounds such as pyridoxal phosphate. Several other corrinoid-linked biological reactions such as the formation of acetate from methyltetrahydrofolate occur through the participation of methyl- B (19).l4 Although the precise mechanism of action for this reaction is not yet known the conversion occurs through the retention of all three hydrogen atoms of the methyl group of (18) or (19) in acetic acid.” Enzf Co) + Me-Tetrahydrofolate + Enzf Co+Me NADPH (17) + MeC0,H co + (17) (18) (19) R. G. Eager B. G. Baltimore M. M. Herbst H.A. Barker and J. H. Richards Bio-chemistry 1972 11 253. C. G. D. Morley and T. C. Stadtman Biochemistry 1971 10 2325. l3 C. G. D. Morley and T. C. Stadtman Biochemistry 1972 11 600 and references cited therein. l4 For a review on coenzyme-B,,-linked reactions see H. A. Barker Ann. Rev. Biochem. 1972 41 55; H. Wcissbach and R. T. Taylor Vitamins and Hormones 1970 28 415. l5 D. J. Parker H. G. Wood R. K. Ghambeer and L. G. Ljungdahl Biochemistry 1972 11. 3074. 145 Reaction Mechanisms-Part (iii) Enzyme Mechanisms 2 Conformations of Substrates and Conformational Changes at the Active Sites of Enzymes A great deal of emphasis has recently been laid on the determination of con- formations which substrates attain at the active sites of enzymes and also on monitoring the subtle conformational changes in protein structure occurring when allosteric effectors and substrates progressively bind to enzymes to give catalytically active complexes.Conformational and Mechanistic Studies with Glutamine Synthetase.-Sheep brain glutamine synthetase catalyses the formation of glutamine from glutamate NH, and ATP and has been extensively studied by Meister and co-workers. The overall reaction is believed to occur in two stages involving the intermediacy of enzyme-bound glutamyl phosphate (211 followed by its decomposition by NH .16-18 The precise configuration which an open-chain flexible molecule such as glutamate (20) may attain at the enzyme active site is usually difficult to 2-H02C-(CH2)2-CH(NH2)C02H 03-P-O-CO-(CH2)2-delineate.In the case of glutamine synthetase however Meister and co-workers took advantage of the relatively broad substrate specificity" of the enzyme and showed that cis-l-amino-1,3-dicarboxycyclohexane (23) was a good substrate for the enzyme and was converted into the corresponding 5-amido-deri~ative.'~ The alignment of the glutamyl moiety of(23) at the active site is thus limited to two arrangements allowed by the two favourable conformations of the cyclohexane ring as shown in (23a) and (23b). The subsequent demonstration that cis-l-amino-1,3-dicarboxycyclopentane(24)is also a good substrate2' for the enzyme suggests that the geometry of the atom of the glutamyl skeleton as present in (23a) is the most likely one since the great rigidity of the cyclopentane ring will permit only this arrangement at the enzyme active site.B3NH2 q c 0 2 H +02H qHCo2" 'COzH HOzC H02C 0 (234 (23b) (24) (25) l6 For a review see A. Meister Adt.. Enzymol. 1968 31 183; subsequent developments in the field are summarized in refs. 17 and 18. " J. D. Gass and A. Meister Biochemistry 1970 9 1380. S. S. Tate Fang-Yun Lew and A. Meister J. Biol. Chem. 1972 241 5312. l9 J. D. Gass and A. Meister Biochemistry 1970 9 842. *' R. A. Stephani W. B. Rowe J. D. Gass and A. Meister Biochemisrry 1972 11 4094. 146 M. Akhtar and D. C. Wilton An aspect of the glutamine synthetase reaction which has aroused much controversy is whether y-glutamyl phosphate (21) is a ‘true’ intermediate in the reaction.Using glutamate as the substrate attempts to isolate the intermediate (21) were unsuccessful and resulted in the isolation of pyrrolidonecarboxylic acid (25) presumably formed by a cyclization reaction. The use of the rigid derivative (23) for this purpose offered a distinct advantage since in this case the cyclization is unfavourable and the phosphate derivative was isolated from an enzymic incubation.21 In contrast to the involvement of a phosphoryl interme- diate with the sheep enzyme Boyer and co-workers on the basis of kinetic evidence and isotope-exchange studies have concluded that glutamine synthetase from E. coli catalyses the reaction via a concerted mechanism without the partici- pation of a phosphoryl intermediate.22 Mechanistic and Allosteric Studies with Cytidine Triphosphate Syntheta~e.~ ,-This is an allosteric enzyme ;it consists of four identical sub-units and is activated by the effector guanosine triphosphate (GTP) to catalyse the following reaction :24 UTP + Glu-NH + ATP Glu-OH + ADP + Pi + CTP (1) (32) (22) (20) (35) The information available to date allows the overall conversion in reaction (1) to be separated into two broad stages involving the liberation of enzyme-bound NH, reaction (2) followed by its utilization in CTP formation reaction (3).Glu-NH + H,O + Glu-OH + ‘NH,’ (2) UTP + ‘NH,’ + ATP -+CTP + ADP + Pi (3) The evidence for the operation of reaction (2) includes the demonstration that in the absence of the acceptor UTP the enzyme shows glutaminase activity and also the fact that added NH can replace glutamine in reaction (3).It has now been shown25 that the activator GTP acts almost exclusively on the glutaminase activity [reaction (2)]. The NH,-dependent CTP-forming activity of the enzyme is not altered by the allosteric effector. The reaction (2) resulting in ‘NH,’ formation has been shown to occur through the sequence (26) +(27)-+ (28)-+(Glu-OH) involving the participation of a covalent intermediate26 in which the glutamyl residue is linked to a SH group on the enzyme. The same SH group reacts27 with the diazoketone analogue [see (29)] through the reaction (29)- (31). This latter reaction is enhanced2* about eightfold by the effector GTP and requires only part of the catalytic 21 Y.Tsuda R. A. Stephani and A. Meister Biochemistry 1971 10 3186. 22 F. C. Wedler and P. D. Boyer J. Biol. Chem. 1972 247 984. 23 I. Liebrrman J. Biol. Chem. 1956 222 765. 24 K. P. Chakraborty and R. B. Hurlbert Biochim. Biophys. Acta 1961 47 607; C. W. Long and A. B. Pardee J. Biol. Chem. 1967 242 4715. 25 A. Levtizki and D. E. Koshland Biochemistry 1972 11 241 247. z6 A. Levitzki and D. E. Koshland Biochemistry 1971 10 3365. 27 C. W. Long A. Levitzki and D. E. Koshland J. Biol. Chem. 1970 245 80. 28 A. Levitzki W. B. Stallcup and D. E. Koshland Biochemistry 1971 10 3371. Reaction Mechanisms-Part (iii) Enzyme Mechanisms 0 e h 00t? P rn h3 " w w c:,"' h I-t? N w v,kNs 0 M.Akhtar and D.C. Wilton machinery needed for NH formation namely the glutamylation step.This particular feature permits the precise deductionz5 to be made that the binding of allosteric effector GTP to the enzyme enhances the activity of the groups partici- pating in glutamylation as represented by the conversion (26)+ (28). The mechanism of the second stage of the reaction (3) has also been studiedz6 and is elaborated in the sequence (32)- (35). Particular attention is drawn to the step (33) +(34) which requires ATP to shift the equilibrium towards product formation. 3 Subunit Interactions in Dimeric Enzymes A large number of oligomeric enzymes consist of two or four apparently identical subunits.29 Recently efforts have been directed towards the delineation of the mutual catalytic relationship between the two identical subunits of dimeric enzymes.In this connection three broad possibilities may be considered (a)only one of the subunits of the dimer contains a catalytically functional active site (b)both of the subunits contain active sites which are functional simultaneously or (c) both of the subunits contain active sites but these function sequentially. Recent studies suggest that the last mechanism may operate for several dimeric enzymes. One such study has been carried out with alkaline phosphatase (E.coli) which is a dimer3' of molecular weight 86 OOO and consists of two identical sub- unit~.~ ' At pH > 7.0 the alkaline phosphatase catalyses the hydrolysis of mono- phosphate esters R-0-PO -(in most studies p-nitrophenyl phosphate is used as a convenient substrate) via a phosphoryl+nzyme intermediate involving an OH groupofa serinere~idue.~',~~ Incubation oftheenzyme with an organic [3zP]- phosphate at pH 4-5 resulted in no net hydrolysis but in the formation of a monophosphated enzyme which was separated from the substrate by Sephadex chr~matography.~~ The liberation of 3zP from the purified monophosphorylated enzyme was greatly enhanced in the presence of a substrate analogue p-chloro- anilidophosphonate.The experiments were interpreted in terms of a 'Flip-Flop' mechanism34 in which binding of substrate on the site A stimulates dephosphory- lation on the site B(reaction d Scheme 3). In the next cycle site A is phosphorylated 29 For a review see I. M. Klotz N. R.Langerman and D. W. Darnall Ann. Rev. Biochem. 1970 39 25. 30 M. J. Schlesinger Brookhaven Symp. Biol. 1964,17 66. 3t F. Rothman and R. Byrne J. Mol. Biol.,1963 6,330. 32 J. H. Schwartz Proc. Nar. Acad. Sci. U.S.A. 1963 49 871. 33 L. Engstrom Arkiu Kemi 1962 19 129. 34 M. Lazdunski C. Petitclerc D. Chappelet and C. Lazdunski European J. Biochem. 1971 20 124. Reaction Mechanisms-Part (iii) Enzyme Mechanisms B Enz-OH hz-OH(R-O-PO:-) RoH R-0 -PO -1. a (A ' (A Enz-OH Enz-OH Enz-O-PO -Enz-o-PO -p. R-0-PO -"(A Enz-OH Enz-OH(R-O-PO;-) Scheme 3 and its dephosphorylation is stimulated by the binding of the substrate to B (Scheme 3). A related phenomenon was observed with horse liver alcohol dehydrogenase which catalyses the reaction RCHO + NADH + H+ S R-CH,OH + NAD The amino-acid sequence of the ethanol-active isozyme indicates that the two polypeptide chains making the dimer are identical.35 The mechanism of the enzyme has been studied by the stop-flow technique under conditions such that product formation may be studied during a single turnover of the enzyme catalysis.Such experiments using a variety of aromatic aldehydes gave biphasic kinetics in which half the product corresponding to one active site per mole of the enzyme was released at a considerably greater rate than the other half.36-3 * Two broad conclusions may be drawn from such studies :36-39 first that both subunits contain catalytically functional active sites but that the sites are kinetically non-equivalent and secondly that the state of liganding at one subunit in the dimer regulates the activity of the other.4 N.M.R. and E.S.R. Currently n.m.r. and e.s.r. techniques are being increasingly applied to the eluci- dation of conformational interactions and the determination of distances between 35 H. Jornvall European J. Biochem. 1970,16,25,41. 36 S. A. Bernhard M. F. Dunn P. L. Luisi and P. Shack Biochemistry 1970,9 185. 37 J. T. Mcfarland and S. A. Bernhard Biochemistry 1972,11 1486. 38 P. L. Luisi and R. Favilla Biochemistry 1972 11 2303. 39 For similar studies on malic dehydrogenase see K. Harada and R. G. Wolfe J. Biof. Chem. 1968 243,4123 4131. M. AkhtarandD. C. Wilton a paramagnetic centre and sensitive nuclei at the active site.40 Examples of this approach include work on the enzyme creatine kina~e,~'.~' which in the presence of Mn" catalyses the reaction Me ATP + \ YNH -+ ADP + Me \ YNH /N-c\ /N-c\ HOzCCHz NHZ HOzC*CHz HN-PO;-Using Mn" as a paramagnetic probe it was shown that the relaxation rate of water protons (PRR) expressed by an empirical enhancement factor E was increased (E = 8.1) only when ADP Mn" and enzyme were present together.Since the PRR of water for Enz-Mn-ADP was the same as for Mn-ADP it was concluded that in the ternary complex (Enz-Mn-ADP) Mn" was bound only to ADP and not to the enzyme.42 The addition of creatine to the ternary 42343 complex caused a decrease in the enhancement value (E~,for the ternary complex Enz-Mn-ADO = 8.1 and E~, for the quaternary complex Enz-Mn-ADP- creatine = 5.3).That the decrease accompanying the conversion ternary + quaternary was not due to the binding of creatine to Mn resulting in the displace- ment of water ligand from the enzyme-bound Mn was shown by determining E and cq at different temperatures. The decrease was therefore attributed to the creatine-induced structural rearrangement at the enzyme active site. The addi- tion of NO; to the quaternary complex resulted in a further decrease in PRR (E for Enz-Mn-ADP*reatine-NO = 2). These results coupled with parallel using e.s.r. indicated a further modification of the structure at the active site as had been previously suggested from SH reactivity and kinetic A schematic arrangement of Mn ADP NO; and creatine on the active site is shown in (36) and is based on distances derived from relaxation rates of the ADP creatine ~//////////,,////,/// // / /,,,,,,,/,-,,,/,,,/,//, (36) Nitrate occupies the site of the y-phosphate of ATP substrate nuclei.42 Another example is the study of relaxation rates of various protons of toluene-p-sulphonamide when bound to Mn-carbonic anhydrase 40 For reviews and comments see (a)A.S. Mildvan and M. Cohn Adc.. Enzymol. 1970 33 1 (6) 0.Jardetzky and N. G. Wade-Jardetzky Ann. Rev. Biochem. 1971,40 605; (c)P. Knowles 'Essays in Biochemistry' ed. P. N. Campbell and F. Dickens Academic Press London and New York 1972 Vol. 8 p. 79. 41 G. H. Reed H. Diefenbach and M. Cohn J. Biof. Chem. 1972 247 3066. 42 G. H. Reed and M.Cohn J. Biol. Chem. 1972.247 3073. 43 Similar studies have been carried out using analogues of creatine ;see A. C. McLaughlin M. Cohn and G. L. Kenyon J. Biol. Chem. 1972,247,4382. 44 E. J. Milner-White and D. C. Watts Biochem. J. 1971 122 727. Reaction Mechanisms-Part (iii) Enzyme Mechanisms 151 which permitted the determination of the distances between Mn" and various protons as shown in (37).45 In another related approach a paramagnetic nitrox- ide radical was covalently attached to histidine-1 5 of lys~zyme.~~ The resulting spin-labelled enzyme (38) broadened the nuclear resonance spectra of N-acetyl- -a-D-glucosamine bound at the active site and these broadenings were used to estimate the distance from histidine-15 to the acetamido-group of the sugar.The paper describes several other related observation^.^^ 3 0-N H q NH-CO-CH,-His-15 (37) 5 The Stereochemistry of Enzyme Reactions The synthesis of stereospecifically labelled substrates especially those which undergo extensive further metabolism has allowed many diverse enzymic reactions to be studied and important stereochemical information to be obtained concerning the mechanisms of individual reactions. Perhaps the most notable example of this approach was the synthesis of the various stereoisomers of mevalonic acid in which hydrogens at the pro-chiral centres C-2 C-4 and C-5 were stereospecifically deuteriated or tritiated.47 More recently considerable attention has been focused on the synthesis of acetate in which the hydrogens of the methyl group were chirally labelled with protium deuterium and tritium.48 The synthesis of these chiral acetates and their use to establish the stereochemistry of the malate synthase system has been reported previ~usly.~~ These acetates have now been used to investigate other biochemical problems involving the intraconversion of methyl and methylene groups.The conversion of acetyl CoA into citrate is catalysed by two enzymes the better known si-citrate synthase which is found as part of the tricarboxylic acid cycle and the other re-citrate synthase found in certain bacteria (the terms re and si refer to the side of a trigonal carbon atom or double bond as determined by the convention proposed by Hanson5*). The cleavage of citrate may be achieved by citrate lyase or ATP-citrate lyase.In order to study the stereo- chemical course of the citrate lyase reaction it is necessary to label the methylene 45 A. Lanir and G. Navon Biochemistry 1972 11 3536. 4h R. H. Wein J. D. Morrisett and H. M. McConnell Biochemistry 1972 11 3707. " J. W. Cornforth R. W. Cornforth C. Donniqger G. Popjak G. Ryback and G. J. Schroepfer Proc. Roy. Soc. 1966 B163 436. '* J. W. Cornforth J. W. Redmond H. Eggerer W. Buckel and C. Gutschow Nature 1969 221 1212; J. Luthy J. Retey and D. Arigoni Nature 1969 221 1213. 49 J. Staunton Ann. Reports (B) 1969 66 555. '' K. R. Hanson .I.Amer. Chem. Soc. 1966,88 2731. 152 M. Akhtar and D. C. Wilton group of citrate that will give rise to the methyl group of acetate.This was achieved enzymically by Cornforth and his associates using stereospecifically tritiated oxaloacetate unlabelled acetyl CoA and re-citrate synthase” (Scheme 4). The Acetyl CoA + HO,C\ T H1 ,;‘c-c* ~ \ CO,H re-Cit rate synthase ’ Citrate lyase (D,O) ’ H D ‘*. /C* T‘ ‘C02H ( 3R)-[3-3 H Oxaloacetate S-Citra te Oxaloacetate TD f H HO,C.CH D**. / H si-Citrate %* *,, Citrate lyase D.**c*/ 0 \2 ,c; ’ C synthase ,; j H02d OH T‘ \!-SSCoA S-Acetyl CoA CO,H T’ \CO,H S-Citrate S-Acetate Scheme 4 citrate has the labelled methylene in the acetate-producing part of the molecule so that subsequent cleavage with citrate lyase in the presence of deuterium oxide gave chiral acetate.The chirality of this acetate was determined using the malate synthase system (Scheme 5) in which owing to kinetic isotope effects the protium D H02C D 7 HO,C H-.. / Malqte /T T‘%* \ c-c*1 ,; >c=c* syn thase’ \ \CO,H HO H \CO,H D,O H CO,H R-Acetate H. ‘-. / T H0,C \ T \ D Fumy HO,C \/c=c*/D Malate synthase’ D‘ ‘CO,H C* HO..‘ c-c* H \CO,H T20 H \C02H S-Acetate Scheme 5 #,, H. Eggerer W. Buckel H. Lenz P. Wunderwald G. Gottschalk J. W. Cornforth C. Donninger R. Mallaby and J. W. Redmond Nature 1970 226 517; W. Buckel H. Lenz P. Wunderwald V. Buschmeier H. Eggerer and G. Gottschalk European J. Biochem. 1971 24 201; H. Lenz W. Buckel P. Wunderwald G. Biedermann V. Buschmeier H. Eggerer J. W. Cornforth J. W. Redmond and R.Mallaby ibid. p. 207; P. Wunderwald W. Buckel H. Lenz V. Buschmeier H. Eggerer G. Gottschalk J. W. Cornforth J. W. Redmond and R. Mallaby ibid. p. 216. Reaction Mechanisms-Part (iii)Enzyme Mechanisms 153 is preferentially removed. The percentage loss of tritium obtained when the malate is dehydrated using fumarase is a measure of the chirality of the methylene in the malate. The results established that citrate lyase catalysed an inversion of configuration between citrate and acetate as shown in Scheme 4. The above results were then used to study the stereochemistry of the si-citrate synthase.’ Conversion of (R)-or (S)-acetate into citrate followed by reconver-sion back into acetate using citrate lyase resulted in an overall retention of con-figuration in the actate and since it has already been shown that the lyase reaction proceeded with inversion the citrate synthase reaction must also have occurred with inversion as shown in Scheme 4.The reactions catalysed by ATP-citrate lyase and re-citrate synthase also showed inversion of configuration.’ In an alternative approach5 Arigoni’s group studied the stereochemistry of the si-citrate synthase system using (R)-and (S)-chiral acetates and analysed the resulting citrate by conversion through isocitrate to succinate. The absolute configuration of the methylene of the succinate was determined using succinic dehydrogenase. Rose’s group has confirmed the results on the si-citrate synthase and the lyases by making use of the enzyme aconitate isomerase which was shown to catalyse the labilization of the 4-pro-S-hydrogen of both cis-and trans-aconitate5 (Scheme 6).This observation allowed [4S-’H]citrate to be prepared from trans-aconitate in which the labelled methylene was in the acetate-producing part of HO,C H HO,C \ ,,‘,‘ \ H \ /H H H+ ,,c 7 H-.\-Y5Hi \&c/ H / c-c /\ \ HO,C C0,H HO,C CO H trans-Aconitate cis-Aconitate Scheme 6 the citrate molecule and hence cleavage of this citrate in deuterium oxide gives a species of chiral acetate. The chiral acetates required for this study of the si-citrate synthase’ were prepared biosynthetically from phosphoenol pyruvate (see below). The enzyme isopentenyl pyrophosphate isomerase catalyses the reaction shown in Scheme 7T.The methylene at C-4 may be specifically Iabelled with tritium by using [2R-or 2S-3H]mevalonate and the chiral methyl group that should result from isomerization in deuterium oxide was analysed by conversion 52 J. Retey J. Luthy and D. Arigoni Nature 1970 226 519. ’’ J. P. Klinman and I. A. Rose Biochemistry 1971 10 2259. 54 J. P. Klinman and I. A. Rose Biochemistry 1971 10 2267. t The numbers a b and c refer to decreasing order of priority and make a clockwise (R) rearrangement. This makes the side of C-3 viewed by the reader re. In designating C-4 of isopentenyl pyrophosphate where identical ligands normally prevent the application of this rule the sides at C-4 are named according to those of the adjoining trigonal carbon C-3 in this case.M. Akhtar and D.C. Wilton Enz-X I D Me H DT H Dimethylallyl pyrophosphate Enz-X-Isopentenyl pyrophosphate Scheme 7 of the dimethylallyl pyrophosphate into farnesol followed by ozonolysis and oxidation to give acetate. Analysis of the acetate (Scheme 5)showed that protona- tion during the isomerization had occurred to the re side of the double bond and a concerted addition-abstraction process has been proposed5 (Scheme 7). Apart from the stereospecifically labelled mevalonates and chirally labelled acetates a third specifically labelled compound that has found considerable use in solving a number of stereochemical problems is phosphoenol pyruvate (PEP) in which the two vinyl hydrogens are labelled specifically with either deuterium or tritium (or both).The method used for the specific labelling of the vinyl hydrogens of PEP with deuterium involved biosynthesis from (1R)-[1-2H,]-fructose 6-phosphate. Assignment of the absolute configuration of the product was achieved by n.m.r.56 That this assignment could be made meant that the enolase reaction was stereospecific and that removal of the OH group occurred anti to the C-2 proton whereas in the reverse direction the OH is added to the re side of the double bonds6 as shown in Scheme 8. The synthesis of PEP labelled with both deuterium and tritium of known absolute configuration allowed an investigation of the stereochemical course of the conversion of PEP into pyruvate catalysed by pyruvate kina~e.'~ This conver- sion should result in the methyl group of the pyruvate attaining chirality and this chirality may be determined by conversion of the pyruvate into acetyl CoA and analysis of the methyl group by the malate synthase system (Scheme 5).The approach established that the proton donated by the pyruvate kinase adds to the si side of the PEP. The chirally labelled pyruvate was used to establish the stereochemical course of the biotin-dependent carboxylation to give o~aloacetate.~~ This was made possible because only 8 % of the tritium was lost during the carboxylation instead 55 K. Clifford J. W. Cornforth R. Mallaby and G. T. Phillips Chem. Comm. 1971 1599; J. W. Cornforth K. Clifford R. Mallaby and G. T. Phillips Proc. Roy. Soc. 1972 B182 277.56 M. Cohn J. E. Pearson E. L. O'Connell and I. A. Rose J. Amer. Chem. Soc. 1970 92 4095. 57 I. A. Rose J. Biol. Chem. 1970 245 6052. Reaction Mechanisms-Part (iii) Enzyme Mechanisms CO2H T-C-D I Oxaloacetate c=o P TD t! \/ Tcc-oH Enolase C Pyruvate) T-C-D I ___* II kinase I H-C--08 C c=o /\ I AO,H C0,H 08 C0,H 2-Phosphoglyceric Phosphoenol Pyruvate acid pyruvate T-C-D Malate I HO-C-H ~ I CO,H Scheme 8 of a theoretical 33 % and it therefore follows that there is a considerable isotope discrimination in favour of the removal of the protium. The determination of the absolute configuration at C-3 of the resulting oxaloacetate by conversion through malate into fumarate established that addition of the carbon dioxide to the C-3 of pyruvate occurred with retention of configuration as outlined in Scheme 8.Similarly the conversion of the deuteriated and tritiated malate back into pyruvate using malate enzyme was also shown to occur by retention of configuration5 (Scheme 8) a result confirmed by Cornforth's group using an alternative approach.58 There are three enzymes which catalyse the carboxylation of PEP namely PEP carboxylase PEP carboxytransphosphorylase and PEP carboxykinase. Incubation of specifically tritiated PEP with these enzymes and .analysis of the resulting oxaloacetate established that in all three cases addition of CO occurred to the si side of the double The condensation of PEP and erythrose 4-phosphate to give 3-deoxy-~- aribinoheptulosonate 7-phosphate (DAHP)provided another system for utilizing the specifically labelled PEP.The reaction which is outlined in Scheme 9 was shown to proceed through si attack on C-3 of PEP.60 It is significant that a K. H. Clifford J. W. Cornforth C. Donninger and R. Mallaby European J. Biochem. 1972 26 40 1. 59 I. A. Rose E. L. O'Connell P. Noce M. F. Utter H. G. Wood J. M. Willard T. G. Cooper and M. Benziman J. Biol. Chem. 1969 244 61 30. '' H. G. Floss D. K. Onderka and M. Carroll J. Biol. Chem. 1972 247 736; D. K. Onderka and H. G. Floss J. Amer. Chem. SOC.,1969,91 5894. M. Akhtur and D.C. Wilton CO,H I O=C T CHZO 8 Lf I ,C-OH H\ + ,C-CHO HI T OH Phosphoenol pyruvate Erythrose 4-phosphate H OH 3-Deoxy-~-Aribinoheptulo-sonate 7-phosphate Scheme 9 combined chemical and enzymic degradation showed at least an 80 % stereo-specificity at C* in the overall reaction because it had previously been proposed that the mechanism for DAHP synthetase proceeded uia the intermediacy of a free methyl group at C-3 of PEP.61 Such a proposal is not consistent with the above retention of stereospecificity unless this methyl group is sufficiently hindered to prevent rotation before the subsequent deprotonation step occurs.On the other hand DAHP synthetase does catalyse an exchange of protons of PEP with the medium and this could be explained by a side reaction in which PEP forms a carbanion in the presence of erythrose 4-phosphate that is protonated and de-protonated during the course of the reaction thus resulting in a certain amount of scrambling at C-3.60 The chiral centre produced at C* of DAHP was utilized to establish the stereo- chemistry of the 1,4-conjugated elimination of phosphoric acid occurring in the synthesis of chorismic acid uia shikimic acid.The conversion of doubly labelled [7-I4C,6R or 6S3H]shikimic acid to chorismic acid and measurement of the resulting tritium :carbon ratio established that the 6-pro-R-hydrogen is lost and hence that the elimination was anti.60 In an alternative pathway the dehydra- tion of dehydroshikimate to give protocatechuate also lost the 6-pro-R-hydrogen in a syn elimination of water.62 Apart from the information obtained by using chirally labelled acetate and specifically labelled PEP stereochemical aspects of a number of other enzyme systems have been investigated.It has been established that in the NAD-linked reaction catalysed by the flavoprotein enzyme dihydro-orotic acid dehydro- genase there is an anti elimination of the C-4-and C-SS-hydr~gens.~~ In the presence of catalytic amounts of NAD both hydrogens exchange with the medium; however the 5s-hydrogen exchanges twice as fast as that at C-4 (Scheme 10). The mechanism for the conversion of urocanate (39) into 3-(imidazol-4’-on- 5’-y1)propionate(40) which is catalysed by the enzyme urocanase has been described in a previous Report.’ The stereochemistry of proton addition from A. B. DeLeo and D. B. Sprinson Biochem. Biophys. Res. Comm. 1968 32,873. 62 K.H. Scharf M. H. Zenk D. K. Onderka M. Carroll and H. G. Floss Chem. Comm. 197 1,765. 63 P. Blattmann and J. Retey European J. Biochem. 1972 30 130. 157 Reaction Mechanisms-Part (iii) Enzyme Mechanisms Dihydro-orota te _____+ dehydrogenase H I H Dihydro-orotic acid Orotic acid Scheme 10 the medium has now been investigated by carrying out the reaction in deuterium oxide. Chemical degradation of the product established that both the protons in the side-chain were added to the re side of the double bond64 (Scheme 11). CO,H / Urocanase D*O ' Scheme 11 An interesting example of an inhibitor which is also a partial substrate is found in the case of bromopyruvate and the enzyme 2-keto-3-deoxy-6-phospho-gluconate aldolase.The 3R-hydrogen of the bromopyruvate is exchanged at about 50 times the rate at which the inhibitor alkylates the enzyme. Using [3S-3H]bromopyruvate it was established that alkylation of the enzyme involved inversion of configuration at C-3 of the pyruvate. lo One reason for establishing the stereochemistry of enzymic reactions is to see if there are correlations between reactions which might suggest either similar mechanisms or common evolutionary origins for the enzymes. A particularly striking example of the latter possibility is found for mammalian NADPH-linked steroid reductases and dehydrogenases where all known examples that transfer the hydride ion to the a side of the steroid use the 4B-hydrogen of the cofactor. On the other hand hydride transfer to the p face of steroids originates from the 4A position of the c~factor.~~ 6 The Biosynthesis of Peptides by Non-ribosomal Multi-enzyme Complexes The biosynthesis of the cyclic polypeptides Gramacidin S (41) and Tyrocidine (42) has received considerable attention over the past few years.66 It is now well 64 F.Kaeppeli and J. Retey European J. Biochem. 1971 23 198. 65 M. Akhtar D. C. Wilton I. A. Watkinson and A. D Rahimtula Proc. Roy. SOC. 1972 B180 167. 66 Y. Saito S. Otani and S. Otani Adu. Em.vtnoI. 1970 33 337; F. Lipmann Science 197 1 173 875 ;S. G. Laland 0. Froyshov C. Gilhuus-Moe and T. L. Zimmer Nature New Biol. 1972 239 43. 158 M. Akhtar and D.C. Wilton established that these compounds are synthesized on a multi-enzyme complex which itself codes for the sequence of amino-acids in the peptide by having multiple binding sites one for each amino-acid residue.Activation is achieved by linking amino-acids to their respective sites as thioesters after which the growing peptide chain is transferred from one site to the next as each amino- acid is incorporated into the sequence. In the case of the biosynthesis of Gramacidin S the synthetase consists of two units. The smaller unit (Enzyme 11) is responsible for the racemization as well as activation of phenylalanine. The larger unit (Enzyme I) first primes itself with the four remaining amino-acids in sequence as thioesters on four separate sites while the fifth site binds the completed pentapeptide until the second penta- peptide is synthesized when the two are linked in head-to-tail condensations.It is suggested that the transfer of the growing peptide chain from one amino-acid residue to the next is achieved by an arm consisting of a pantetheine residue in a manner analogous to that in fatty-acid synthesis. The sequence of events for synthesis is shown in Scheme 12. /NHz SH SH SH Phe SH I Phe ‘,A I Pantetheine arm 5 NH D-Phe-+Pro-+Val-Om-’Leu Leu+-Orn+Val+ Pro+ D-Phe t I - SH HS‘.G . (41) D-Phe-+Pro+PheQ-Phe+AspNH2t Leu+Orn+-Val+Tyr+GluNH (42) Scheme 12 There are also three linear Gramacidins A B and C which are pentadeca- peptides and these are also biosynthesized by a non-rib~somal~’ system which ’’ 0.Froyshov T. L. Zimmer and S.G. Laland F.E.B.S. Letters 1972 20 249. Reaction Mechanisms-Part (iii)Enzyme Mechanisms has been purified.68 It is interesting to note that the biosynthesis of the cyclic trimer 2,3-dihydroxy-N-benzoyl-~-serine also involves enzyme-bound thioester intermediate^.^' '' K. Bauer R. Roskoski jun. H. Kleinkauf and F. Lipmann Biochemistry 1972 11 3266. 69 G. F. Bryce and N. Brot Biochemistry 1972 11 1708.