14 Biological Chemistry Part (ii)Enzyme Chemistry By C. A. ROSS Department of Biochemistry University College Cork Lee Makings Prospect Row Cork Ireland 1 Introduction In stark contrast to recent times the past year has seen very little activity by writers of texts on enzymology. There has been the appearance of one book,’ which is cast in much the same mould as Foster’s ‘The Nature of Enzymology’. It succeeds well at the level of introductory university courses but poor diagrams and typesetting detract from its appearance. From Academic Press come two new series. One is entitled ‘Benchmark Papers in Biochemistry’ the first volume of which is devoted to enzymes.’ Here is a chance to read in English original papers by many of the pioneer workers in the field some of whom have tended to be overlooked by modern authors.Oddly enough that most quoted of papers by Lineweaver and Burk3 is not included. The other new series is entitled ‘Selected Methods in Enzymology’. Such has been the rapid advance of the subject matter and the technology of biochemistry since 1955 when the first of the four original volumes of ‘Methods in Enzymology’ appeared as a compendium of available techniques that the original series has run to over sixty volumes with a bewildering sequence of titles. It has now been decided to produce single volumes in selected areas that will contain basic methodology. The first of these volumes4 is significantly devoted to RNA and protein synthesis. It has of course been apparent for some years past that the main growth area in biochemistry has been in the area of molecular biology.Following the discovery of the double helical structure of DNA and subsequently the nature of the genetic code in the 1960’s a lull ensued which has ended with the ability to sequence DNA and to manipulate genetic material. The importance that this work holds is highlighted by the Nobel Awards in chemistry and in medicine or physiology in recent years. F. Sanger achieved a most notable success in being awarded in 1980 a second Nobel Prize in chemistry and his Lecture has now appeared in print’” as have those of P. Berg” and W. Gilbert.” This report begins with a short review of the chemistry of the enzymes that are the very basis of this new area. For biochemistry the 1980’s will see enormous strides in the field of T.Palmer ‘Understanding Enzymes’ Ellis Horwood Ltd. Chichester 1981. ’Benchmark Papers in Biochemistry Vol. 1 Enzymes’ ed. H. C. Friedman Academic Press New York 1981. H. Lineweaver and D. Burk J. Am. Chem. Soc. 1934,56,658. ‘RNA and Protein Synthesis’ ed. K. Moldave Academic Press New York 1981. ’ (a)F. Sanger Biosc. Rep. 1981.1 3; (b)P. Berg ibid. p. 269; (c) W.Gilbert ibid.,p. 353. 381 382 C. A. Ross molecular genetics with all its attendant moral and ethical problems. With the passing of the era of metabolic biochemistry it is poignant to record too the passing of Sir Hans Krebs at the age of 81. He had made that era his own. 2 Restriction Endonucleases Bacteria are capable of destroying foreign DNA by splitting the double-stranded polynucleotide at a limited number of sites.This ability resides in strain-specific restriction endonucleases. At the same time the DNA of the host cell is protected against its own enzymes by methylation of adenine and cytosine residues at the specific sites recognized by the restriction enzymes. Originally it was thought that the restriction site and the methylation site were one and the same for a particular restriction and modifying (methylating) enzyme but this is now known to be not necessarily true. Sufficient numbers of restriction endonucleases have been detected and characterized to enable at least three classes to be identified and the basis of the classification can best be summarized in table form.From Table 1 it can be Table 1 Characteristics of restriction endonucleases (modified from ref. 6) Type I Type 11 Type 111 1. Restriction and modification activities Single multifunctional enzyme Separate endonuclease and methylase Single multifunctional enzyme 2. Protein structure 3 different subunits Simple 2 different subunits 3. Requirements for restriction ATP.Mg2+ SAM Mg" ATP.Mg2' (SAM) 4. Sequence of host specificity sites' sB:T-G-A-Ng-T-sK:A-A-C-N6-G-C-T Two-fold symmetry SP1 A-G-A-C-C ~1.5 C-A-G-C- A-G G-T-G-C 5. Cleavage sites Possibly random at least 1000 bp from host specificity site Generally at host specificity site 24-26 bp to 3' of host specificity site 6. Requirements for me thylation (ATP.Mg2+) SAM SAM (ATP.Mg2+) SAM 7. Restriction us.methylation Mutually exclusive Separate reactions Simultaneous 8. Site of methylation Host specificity site Host specificity site Host specificity site The host specificity sites sB sK sP1 and s15 refer to E. coli B E. coli K phage PI and plasmid P1S. Type I enzymes have a hyphenated site with two constant domains of 3 and 4 bases separated by a nonspecified spacer of 6 or 8 bases (Reproduced with permission from Annu. Rev. Biochem. 1981 50 285) seen that type I and type I11 endonucleases are multimeric multifunctional enzymes that require ATP S-adenosyl methionine (SAM) and Mg2+ as factors or activators of the cleavage and modifying activities. Furthermore type I enzymes are powerful ATPases. Both types have recently been reviewed.6 Whereas types I and I11 enzymes cleave the polynucleotide chain at a point distant from the specificity site (in itself ' R.Yuan Annu. Rev. Biochem. 1981,50 285. Biological Chemistry -Part (ii) Enzyme Chemistry a remarkable phenomenon) type I1 endonucleases usually cleave and methylate at the same site. They have the simplest protein structure and minimal requirements for both restriction (nucleolytic cleavage) and methylation. The detailed study of the type I1 has focused primarily on the Eco RI enzyme.'* At this point a short explanation is required of the notation used for the naming of restriction enzymes due to Smith and Nathans.' An italicized three-letter abbreviation is used for the host organism followed by where necessary a fourth letter for strain and a roman numeral to indicate each RM system in the organism.Thus Hind I1 is the name of an RM system in Haemophilus influenrue strain Rd. Similarly Eco RI describes an enzyme found in Escherichia coli R that has the recognition sequence shown in Scheme 1,where an arrow indicates the point of cleavage and the asterisk 5'-G?A-i-T-T-C-3' 3'-C-T-T-I-A7G-5' indicates centre of two-fold symmetry Scheme 1 the group which is methylated. The availability of enzymes with highly specific and limited sites for cleaving DNA molecules has been immensely valuable in the task of determining nucleotide sequences and they have been described as 'the molecular scalpels of the contemporary biologist'.'" In fact DNA sequencing is faster and more accurate than protein sequencing largely because of the techniques introduced by Sanger.sa.9b This is remarkable since it was considered that DNA sequencing would be extremely difficult if not impossible owing to the nature of the primary structure.Degradation by chemical means or by the nucleases then known led to the produc- tion of large numbers of small oligonucleotides. The discovery of endonucleases with specificities involving sequences of bases that occur relatively rarely in the polynucleotide has served to revolutionize sequence analysis. At a higher level of complexity using techniques employing the same principles it has been possible to construct physical maps of chromosomes in which restriction enzyme cleavage sites serve as defined reference points.Referring to Scheme 1 the two strands of duplex DNA are seen to be cleaved at points that are not opposite each other and so give rise to cohesive ends that hold the strands together by hydrogen bonding until steps are taken to separate them. Originally cohesive ends composed of several bases of one type (e.g. oligo-dA) were added to a polynucleotide chain which would then anneal with a chain to which a short length of complementary base (oligo-dT in this case) had been attached by a polymerase enzyme. However new restriction enzymes are con- tinually being reported and a list compiled annually by Roberts" now contains ' (a)R. D. Wells R. D. Klein and C. K. Singleton in 'The Enzymes Third Ed. Vol. 14 Nucleic Acids Part A' ed. P. D. Boyer Academic Press New York 1981; (b) P.Modrich Quart. Rev. Biophys. 1979 12 315. * H. 0.Smith and D. Nathans J. Mol. Bid 1973,81 419. (a)M. Zabeau and R. J. Roberts in 'Molecular Genetics Part 111' ed. J. H. Taylor Academic Press New York 1979; (b)P.A. Biro and S. Weissman ibid. in (a)R. J. Roberts Nuckic Acid Res. 1981,9 75; (b)Gene 1980 8 329. 384 C.A. Ross over 200 entries. Each enzyme has a different specificity site and/or cleavage site so that it is possible to cleave one DNA molecule to produce a ‘sticky end’ that will anneal with another DNA strand produced by a different restriction enzyme. Covalent joining is then effected by means of a ligase enzyme. The isolation of particular DNA segments by molecular cloning5b involves the insertion of DNA fragments at particular sites in suitable receptor DNA molecules called ‘vectors’ which are capable of replicating autonomously.A large number of these are now available for cloning DNA segments in E. coli. Thus the synthesis of human interferon all and interferon PI,’’ has recently been achieved in E. coli infected by a A phage recombinant containing a human genomic fragment. Such genetic engineering feats hold out the best hope for producing scarce biological materials on a commercial basis. In the field of clinical diagnosis the new technology is proving to be a powerful tool. For example,” DNA from a normal human digested with MbO I1 gives three bands of 1600 580 and 400 base pairs that hybridize with cloned human insulin cDNA probes.DNA from a diabetic patient with a mutant insulin gave an additional band of 980 base pairs on digestion and hybridization. A MbO I1 cleavage site has thus been lost in the mutant. The effect of this on insulin is expressed as a leucine residue being incorporated at either position B-24 or B-25 instead of phenylalanine. 3 Kinetic Analysis The proceedings of a satellite symposium on Design and Analysis of Enzyme and Pharmacokinetic Experiments held in conjunction with the XI International Con- gress of Biochemistry have now been p~blished.’~ There are sections on nonlinear regression robust parameter estimation and the design of experiments kinetic model identification and kinetic data analysis. Further contributions on the statis- tical treatment of kinetic data have been made by Cornish-Bowden and Endrenyi” and by Oppenheimer et ~1.’~ Waley” had produced an easy method for determining initial rates of reactions where the Michaelis-Menten equation is obeyed.Instead of attempting to draw tangents to the progress curve at zero time a chord is drawn between two points on the curve and its slope gives the rate at an intermediate substrate concentration. The value of this intermediate concentration is calculated from relationships derived from the integrated form of the Michaelis-Menten equation. Thus when the concentration of product is being monitored the intermediate concentration of product formed P3 is given by equation (1). (1) S. Nagata N. Mantei and C. Weissmann Name (London) 1980 287,401.Y.Mory er al. Eur. J. Biochem. 1981 120 197. l3 S. C. M. Kwok S. J. Chan A. H. Rubenstein R. Poncher and D. F. Steiner Biochem. Biophys. Res. Commun. 1981.98 844. l4 ‘Kinetic Data Analysis’ ed. L. Endrenyi Plenum Press New York 1981. *5 A. Cornish-Bowden and L. Endrenyi Biochem. J. 1981,193,1005. l6 L.Oppenheimer T. P. Capizzi and G. T. Miwa Biochem. J. 1981,197,721. S. G. Waley Biochem. J. 1981 193 1009. Biological Chemistry -Part (ii) Enzyme Chemistry 385 Only one datum point need be used when the calculation is of course simplified since P1 =0 at zero time. In order to obtain the correct values for Vmaxand K, two chords are required (e.g.one from 0-20°/0 reaction of slope v and the other from 040% reaction of slope v’) in order to calculate the rate vo when the concentration of product is zero.Then So6 -S’) where S and S’are the substrate concentrations corresponding to the rates v and v’.The use of this equation is recommended if the product is a powerful inhibitor. Otherwise a plot of v against p will give an intercept at p =0 equal to the initial rate vo. The ratio of the two slopes v/v’ already mentioned will lie between 1 and 1.145 when there is no product inhibition nor enzyme inactivation. However if there is product inhibition the upper bound of this quotient is approximately 2.2 so that it is possible to detect product inhibition from the results of a single experiment ! A new method has been described for plotting kinetic results for inhibited enzyme reactions18 and consists of plotting data in a normalized fashion.Thus vo/vi the ratio of the initial velocities for the non-inhibited and inhibited reactions is plotted versus the specific velocity r/(l+CT) where CT is the [S]/K ratio. The method provides a simple way of determining the inhibition constants Ki and KI the dissociation constants of the EI and ESI complexes respectively by replots of intercepts of the primary specific velocity plot. The method is applied to two actual examples elastase inhibition by a glycosaminoglycan polysulphate and glycogen phosphorylase inhibition by glucose 6-phosphate. The kinetics of suicide substrates based on the scheme due to Walsh et a/.” (see Scheme 3 in last year’s Report) have been reworked and results more consistent with exact solutions obtained.The most important factor was found to be the term (1 +r)p instead of rp as proposed by Waley where r is the ratio of the rate constant of product formation to that of enzyme inactivation and p is the ratio of initial concentration of enzyme to that of suicide substrate. Continuing his thesis that enzymes do not usually obey the Michaelis-Menten equation Bardsley has published further papers” on the computation of the probabilities of obtaining complex curves from simple kinetic schemes and on the probability that complex kinetic curves can be caused by activators or inhibitors. Considering crude enzyme samples may contain unknown amounts of endogenous substrate reaction product inhibitor and even contaminating enzymes and yet are used to evaluate the kinetic parameters K and V,an interesting novel analytical method has been described.” In the presence of endogenous substrate (concentra- tion x) and contaminating enzyme (activity u),the Michaelis-Menten equation can A.Baici Eur. J. Bioehem. 1981 119 9. 19 S.Tatsunami N Yago and M. Hosoe Biochim. Biophys. Am 1981,662,226. *’ (a)F. Solano-Mufioz P. B. McGuilay R. Woolfson and W. G. Bardsley Eiochem. J. 1981,193,339; (b)F.Solano-Mufioz W. G. Bardsley and K. J. Indge ibid. 1981 195 589. 21 M. G. Kato and N. Inoue Biochim. Biophys. Aefa,1981,661 1. 386 C.A. Ross be rewritten as where vi is an observed reaction rate at a concentration of exogenously added substrate Sl. When substrate is not added (S; = 0) DO = (VX/K + X) + u (4) These equations can be rearranged to give which is of the general form y/b = x/a = 1 representing a straight line in xy space with intercept a on the x axis and intercept b on the y axis.Thus the equation represents a straight line when -Sl is plotted on the x axis as a and (ui -vo)/vo is plotted on the y axis as b (cf. the direct plot of Eisenthal and Cornish-Bowden) as shown in Figure 1.In the case where no contaminating enzyme is present (u = 0))A = (K,+ x) and B = K,/x so that K = AB/(B+ 1)andx = A(B + 1). (6) Where a contaminating enzyme is present (u # O) A = (K,+ x) and B = K,/x(l -u/vo).Two concentration levels of endogenous substrate are used to give two pairs of co-ordinates for the evaluation of the parameters.X a -Si I Figure 1 Linear plot An increasingly important aspect of enzyme kinetics is that in which the reaction occurs between two phases as with insoluble substrates or immobilized enzymes. In such situations the reaction rates are directly related to the area concentration Biological Chemistry -Part (ii)Enzyme Chemistry 387 of bulk solution which is defined as the two thirds power of volume concentration.” Thus for an immobilized enzyme the Michaelis-Menten equation is modified to v[s]2’3 v= KL + [S]’” where KL is the modified Michaelis-Menten constant. In similar fashion the Langmuir adsorption isotherm may also be modified. The modified equations have been tested experimentally and do appear to be obeyed over a greater concentration range than the unmodified expressions.On the subject of immobilized enzyme systems Mosbach the immobilization of the four urea cycle enzymes to the same Sepharose particles resulting in greater catalytic efficiency than the soluble system and operating in a cyclic manner. The same worker also reportsz4 an easy method for activating hydroxy-group-carrying supports with 2,2,2-trifluoroethanesulphonyl chloride resulting in much improved coupling yields (up to 80%) and retained specific activities (up to 50%).An interesting use of immobil- ized enzymes has resulted in the demon~tration’~ of physical interactions between different enzymes of the citric acid cycle and the aspartate-malate shuttle. It is shown that maximally four molecules of malate dehydrogenase bind to one fumarase molecule and that this complex is then able to bind to either citrate synthase or aspartate aminotransferase.The suggestion is that the enzymes bind alternatively in order to allow the cell to perform either cycle or shuttle reactions according to its needs. An intriguing concept that of energy cost of catalytic turnover owing to enzyme degradation in viuo has been explored.26 On the basis that four ATPs are required in the formation of a single peptide bond it is possible to calculate the energy required to make an enzyme molecule. Given the half-life of the enzyme and its turnover number it is then possible to calculate the number of ATPs amortized per catalytic turnover. Data on some twenty rat liver enzymes has been examined in this way and an eight-order-of-magnitude range results.4 pHEffects The variation in enzyme-catalysed reaction rates with pH has always been an obvious line to follow in investigating mechanisms of action. However the interpre- tation of the experimental data is fraught with difficulties. One basic source of error in the pH measurements themselves has been rep~rted.~’ In glass-electrode assem- blies in which the reference half-cell contains a porous ceramic plug errors arise from substantial liquid-junction potentials. The size of the error is proportional to the ratio between the salt concentration in the standard buffers and in the unknown solutions. Although the error varies from one electrode specimen to another the average was found to be 0.2pH unit per 10-fold salt concentration difference between standard and test solutions.The author claims that errors of this order must be widespread in the recent literature since the introduction of this design of electrode. 22 M.-G. Lee and W.-C. Chiang Biochem. Int. 1981,2,1. 23 N.Siegbahn and K. Mosbach FEBSLett. 1982 137 6. 24 K.Nilsson and K. Mosbach Biochem. Biophys. Res. Commun. 1981,102,449. 25 S.Beeckmans and L. Kanarek Eur. J. Biochem. 1981,117 527. 26 M.N. Kazarinoff Arch. Biochem. Biophys. 1981,208 131. 27 J. A.Illingworth Biochem. J. 1981,195,259. 388 C. A. Ross A standard procedure for the analysis of kinetic data for regulatory enzymes as a function of pH has been presented2* using the limiting hyperbolae predicted by the exponential model for a regulatory enzyme VAa0exp kPa v= 1 + Aaoexp kPa where a.is the association-binding constant characterizing the original state of the enzyme k is the interaction free energy and P. is the fractional saturation. Such an expression predicts two limiting hyperbolae as Paapproaches its limiting values of 0 and 1.The relationship between the exponential model parameters k and ao the Hill coefficient h and the substrate concentration at half-saturation So.5,is shown as k = 4(h -l)/h and In a. = -(In So.5+ k/2). (9) Data from the literaturezg on the pH dependence of the reaction catalysed by aspartate transcarbamylase (aspartate carbamoyl-transferase) has been analysed and shows a significant shift in the pK values for the free enzyme as the conforma- tional state of the enzyme varies.These shifts reflect changes in the environment of ionizable groups within the active site. From Cleland's laboratory comes the results of a pH study to elucidate the mechanism of creatine kinase which catalyses the reversible phosphorylation of creatine by MgATPe3' The pH profiles from initial velocity measurements are interpreted as suggesting the presence of a single group with a pK near 7 that acts as an acid-base catalyst and must be unprotonated in the direction of creatine phosphorylation and protonated for Mg-ADP phosphorylation. In addition another group with a pK near 6 must be unprotonated for activity in either direction. Solvent perturbation of apparent pK values allows one to determine whether the groups involved are neutral or cationic acids.Cationic acids such as the nitrogen bases exhibit little change in pK when organic solvent is added since a net positive charge is present before and after dissociation of the acid. Neutral acids such as carboxy-groups on the other hand show an increase in pK since a new positive and negative charge are present after dissociation which is impeded by the presence of the solvent. Since the pK is near 7 the decrease in the apparent pK of the acid-base catalytic group in these experiments is interpreted as indicating that the group is probably histidine but that it is a lysine with a low pK cannot be ruled out. Similarly the group that must be ionized and that has a pK near 6 has been shown to be a carboxylic acid group.The chemical reaction catalysed by creatine kinase may be written in the form -+R-NH2 + MgATP2-$ -+R-NH-PO:-+ MgADP-+ H' (10) where -+R-NH2 represents creatine. Hence an acid-base catalyst such as histidine will accept a proton from the guanidinium group during phosphorylation of creatine and donate a proton during the reverse reaction. It is interesting to note that in other kinases such as hexokinase and fructokinase the acid-base catalyst has been identified as a carboxy-group (as an aspartate in the case of hexokinase3*). Cleland 28 R. B. Gregory and S. Ainsworth Biochim. kiophys. Acta 1981,659,249. 29 S. C. Pastra-Landis D. R. Evans and W. N. Lipscomb J. Eiol. Chem. 1978 253,4624. '' P.F.Cook G.L. Kenyon and W. W. Cleland Eiochemisfry 1981 20 1204. 31 C. M.Anderson R. E. Stenkamp R. C. McDonald and T.A. Steitz J. Mol. Eiol. 1978 123 207. Biological Chemistry -Part (ii) Enzyme Chemistry argues that as creatine kinase catalyses a reaction which must be freely reversible in muscle the acid-base catalyst must have a pK closely matching the pH of the cell i.e. that for a histidine rather than of a carboxy-group. Furthermore the phosphorylation of a positively charged guanidinium group would present difficulties if the NH2 group formed a hydrogen bond with an ionized carboxy-group. The negative charge of the carboxy-group would be likely to attract much of the positive charge of the guanidinium group to the nitrogen and make it a much poorer nucleophile A neutral histidine would not have the same effect.A possible structure of the transition state for phosphoryl transfer is shown in Scheme 2 in which the (Ad = adenosine; E = enzyme) Scheme 2 ionized carboxy-group is depicted interacting with one of the guanidinium nitrogens while the other is shown speculatively as interacting with the carboxy-group of creatine in order to localize the positive charge away from the nitrogen being phosphorylated. The effect of ionic strength on the pK of an essential acidic group on glucose oxidase has been rep~rted.~’ The kinetics of the reaction are easily analysed according to Scheme 3. I dissociation Of an essential Eo’H202 acidic group (G = glucose; P = product) Scheme 3 Analysis of the ionic strength dependence of pK4.0br assuming that the enzyme is an homogeneously charged impenetrable sphere predicts that the intrinsic pK of the acidic group is 6.7.A van’t Hoff plot of the temperature dependence of K4,0br gave values for AHo and AS0 consistent with the group responsible being a histidine residue. J. G.Voet J. Coe 1. Epstein V. Matossian and T.Shipley Biochemistry 1981,20 7182. 390 C.A. Ross 5 Dehydrogenases In a series of papers,33 Cook and Cleland have presented some detailed mechanistic deductions from isotope effects. First from the magnitude of the isotope effects on V and V/K for various substrates using both labelled and unlabelled substrates information can be obtained on the order of addition of reactants as well as on the location and degree of rate limitation of the isotope-sensitive steps.Secondly by making use of the equilibrium perturbation method to determine the effect of reactants other than those involved in the perturbation on the rate of release of the perturbant molecules from the enzyme it has been possible to distinguish between ordered and random binding. Thus for liver alcohol dehydrogenase NAD is not released at an appreciable rate from the E-NAD-cyclohexanol complex whereas cyclohexanol is released much more readily than it reacts to give products so that the mechanism appears ordered. With NADH and cyclohexanone however since NADH is released at a finite rate from the complex the reaction is random. Conversely for the yeast enzyme NADH release from the enzyme is prevented by acetone so giving the appearance of an ordered mechanism whereas the reverse reaction is random since propan-2-01 and NAD are released at equal rates.In order to enhance observed isotope effects and obtain more useful information on the kinetic mechanism the variation of pH on isotope effects has been studied. In the first case where the bond-breaking (i.e. isotope-dependent) step is pH depen- dent the model predicts that; at the optimum pH an increase in the V/K isotope effect shows that the substrate is sticky (i.e. reacts to give products as fast as or faster than it dissociates) whereas an increase in the V isotope effect shows that some other pH-independent step is at least partly limiting.The theory has been verified in the case of three dehydrogenases. In the second case where the isotope- dependent step is not pH-dependent the isotope effects fall to 1.0 in the forward direction and this model accurately describes the kinetics of yeast and liver alcohol dehydrogenases where hydride transfer to an aldehyde or ketone from NADH to give an alkoxide is isotope sensitive but largely pH independent. The subsequent protonation of the alkoxide requires that an enzyme group (possibly His-51 in the liver enzyme) be protonated. In the final paper secondary isotope effects have been employed to investigate the chemical mechanism of liver alcohol 7toHis-51 H R R E-NAD-alcohol E-NAD-alkoxide E-NADH-ketone E-NADH Scheme 4 33 P.F.Cook and W.W. Cleland Biochemistry 1981,20 1790;ibid. p. 1797,ibid. p. 1805,ibid. p. 1817. Biological Chemistry -Part (ii)Enzyme Chemistry 391 dehydrogenase. Secondary isotope effects result from isotopic substitution in atoms that do not undergo cleavage but change the stiffness of their bonding during the reaction. Thus it is claimed that NAD develops carbonium ion character at C-4 as the result of geometric distortion before hydride transfer. In Scheme 4 the E-NAD-alkoxide complex is shown with the sugar-base bond bent out of the plane of the nicotinamide ring so that N-1 is no longer trigonal and carbonium ion character is developed at C-4. As a result the proton on the Zn-co-ordinated alcohol is transferred to His-5 1. Hydride transfer from alkoxide to C-4 then permits the fast conversion of bent NADH to planar NADH.NPO p-nitrophenol octanoate I 7 t E EA $ EPO EQ E (NPO = p-nitrophenyloctanoate) Scheme 5 0 R-C-0 S- II R-C-OH 0 O \ v I1 R-'2-OH yRJn I H2O Scheme 6 " c.s. Tsai Arch. Biochem. Biophys. 1982 213 635. 392 C. A. Ross Alcohol dehydrogenase is known to have a broad substrate specificity catalysing the reversible oxidation of various primary and secondary alcohols. In addition it has other activities such as mediating the oxidation of aldehydes and hydrolysis of esters. It is this latter activity that has received further attention.34 Kinetic experi- ments were performed on the enzyme using p-nitrophenyl octanoate as substrate and the effect of chemical modification of arginyl sulphydryl and lysyl groups tested.The inhibition of the esterase activity by p-nitrophenol (non-competitively) and octanoic acid (competitively) support the Uni-Bi kinetic scheme shown in Scheme 5. The implication of cysteinyl and lysyl residues in the catalytic site is suggested by the results from chemical modification of these groups and thus Scheme 6 is proposed which resembles that of the sulphydryl esterases without however histidine activation of the cysteinyl group. Kinetic studies on two other oxido-reductases are worthy of mention here. It was recently rep~rted’~ that aldehyde reductase from pig kidney obeyed a strict compulsory-order mechanism coenzyme reacting first. On further e~amination,~~ the reverse reaction in the direction of alcohol oxidation exhibits random-order addition of substrates.Succinic-semialdehyde dehydrogenase from rat brain has also been re-investigated3’ and evidence produced for a compulsory-order mechan- ism coenzyme binding first though it was not possible to distinguish between a ternary complex mechanism and a Theorell-Chance mechanism. 6 Co-operativity Linking this section with the previous one is a report of pressure relaxation experiments3’ showing that the rate of NAD dissociation from liver alcohol dehy- drogenase is controlled by the rate of isomerization of the enzyme-NAD complex and that it is this isomerization which is the rate-determining step in the reduction of acetaldehyde. If this is the case it represents one of the first demonstrations of a rate-limiting conformational change in a dehydrogenase-catalysed reaction.Cornish-Bowden has published further observations on the mechanism of glucokinase from rat liver.39 Glucokinase is a monomeric enzyme with a single binding site for glucose on each molecule and yet which exhibits strong positive co-operativity. Thus the Monod and Koshland models do not apply and a kinetic model must be invoked. Rabin’s original idea of ‘enzyme memory’40 has been developed by Ricard et aL4’ into the so-called mnemonical mechanism. The model requires that the free enzyme exists in two forms with differing affinities for the substrate (glucose in this case) which give rise to the same complex. If the complex reacts fast enough at high ATP concentrations so that equilibrium between the free forms of the enzyme is prevented then apparent co-operativity of glucose binding is observed as the relative proportions of the two forms of the free enzyme change with increasing substrate concentration.At low ATP concentrations no Is W. S. Davidson and T. G. Flynn Biochem. J.,1979,177,595. I6 F. F. Morpeth and F. M. Dickinson Biochem. J. 1981 193,485. A. J. Rivett and K. F. Tipton Eur. J. Biochem. 1981,117 187. M. J. Hardman Biochem. J. 1981 195 773. I9 M. Gregoriou I. P. Trayer and A. Cornish-Bowden Biochemistry 1981,20 499. 40 B. R. Rabin Biochem. I. 1967,102,22C. 41 J. Ricard J.-C. Meunier and J. Buc Eur. J.Biochem. 1974 49 195. Biological Chemistry -Part (ii) Enzyme Chemistry 393 co-operativity will be observed since equilibration of free enzyme forms occurs.The model for glucokinase based on steady-state experiments follows an ordered binding sequence so that ATP does not bind to free enzyme. The isotope-exchange data now presented supports this model rather than one which includes random- order binding.42 In contrast to the previous example citrate synthase which catalyses the condensation of acetyl-CoA and oxaloacetate consists of two identical subunits each independently binding one molecule of acetyl CoA and oxaloacetate. The enzyme exhibits both ligase and hydrolase activity and it might well be thought that these two diverse reactions would occur at different sites. Preliminary evidence has now been to show that when the conformational change occurs in the enzyme in the presence of oxaloacetate the free enzyme representing the hydrolase is converted into a ligase.In the presence of both substrates the enzyme is reconverted to the hydrolase form upon the formation of the intermediate (3S)-citryl-CoA. Co-operativity is ruled out and it is suggested that on binding oxaloacetate a ‘hinged movement’ takes place unmasking catalytic groups and bringing the two binding sites into proximity (ligase form). Hexokinase is an important enzyme and is regarded as the pacemaker of glycolysis in brain tissue and in erythrocytes. It is subject to regulation by the product glucose 6-phosphate for the action of which two different mechanisms have been proposed. One mechanism has glucose 6-phosphate at an allosteric site remote from the active site whereas the other envisages glucose 6-phosphate acting as a simple competitive inhibitor by binding to the y-phosphate subsite of the ATP-binding region.An investigation of the reverse reaction44 shows that glucose 6-phosphate exhibits classical Michaelis-Menten kinetics with a dissociation constant in close agreement with the inhibition constant of glucose 6-phosphate in the forward reaction. From’m therefore maintains his earlier contention that regulation of hexokinase is a manifestation of product inhibition glucose 6-phosphate binding to a high-affinity site that is also the active site. Malate thiokinase catalyses the reversible formation of malyl-CoA or succinyl-CoA from ATP CoA and malate or succinate respec- tively.It is a tetramer each subunit being composed of two non-identical units. Incubation of the enzyme with succinyl-CoA leads to the formation of a tight complex in which four succinyl-CoA molecules are bound per tetramer.45 The ligand is released on denaturation of the enzyme by SDS or on the addition of ADP + inorganic phosphate. However addition of ATP results in the loss of two molecules of bound succinyl-CoA with concomitant incorporation of 2 molecules of phosphate into the enzyme. This enzyme form will combine with succinate plus CoA but will only release phosphate. Hence it is claimed for this enzyme that it displays half-of-the-sites activity. Finally in this Section is a report46 on that most investigated of allosteric enzymes aspartate transcarbamylase (ATCase).The kinetics of this enzyme are complicated by the fact that the holoenzyme is a dodecamer and contains separate 42 M. L. Cardenas E. Rabajille and H. Niemeyer Arch. Biol. Med. Exp. 1979 12 571 43 E. Bayer B. Bauer and H. Eggerer Eur. J. Biochem. 1981 120 155. 44 L. P. Solheim and H. J. Fromm Arch. Biochem. Biophyr. 1981 211 92. 45 L. B. Hersh and M. Peet J. Bid. Chem. 1981 256 1732. 46 J. Foote and W. N. Lipscomb J. Bid. Chem. 1981 256 11 428. 394 C. A. Ross catalytic and regulatory sub-units so that it does not easily conform to either the Monod or the Koshland models. The reverse reaction involving the formation of carbamyl phosphate and aspartate has now been studied using a new assay system involving the removal of products in order to overcome the unfavourable equili- brium position which strongly favours the forward reaction (Scheme 7).Under these conditions co-operativity is not observed. In connection with this novel assay system it may be noted that a method has appeared4' for optimizing the amount(s) of enzyme(s) used in coupling assays. carbamyl aspartate xpLse carbamyl phosphate aspartate a-ketoglutarate x transaminase xaloacetate dehydrogenase'"""XNAD+ malate glutamate Scheme 7 '' F. Garcia-Carmona F. Garcia-Cbnovas and J. A. Lozano Anal. Biochem. 1981,113 286.