15 Biological Chemistry Part (iv) Enzyme Chemistry By C. A. ROSS Department of Biochemistry University College Cork Lee Makings Prospect Row Cork Ireland 1 Introduction The main event to occur in the publishing field in this area of interest during the past year must be the appearance of the third edition of the now classical text ‘Enzymes’ by Dixon and Webb.’ Some twenty-two years have elapsed since the first appearance of this book and eighteen since the second edition so obviously an overhaul was long overdue. Two new authors have been added Thorne and Tipton both from the Cambridge school so that the unique character of the book remains. This includes a classified list of known enzymes with references which takes up a quarter of the book and one must question the value of such a list especially as the I.U.B.3‘Enzyme Nomenclature’ ’also arrived on the bookshelves during the year.Over 2000 enzymes are now classified but the latest edition of this essential reference work unfortunately omits the useful chapter on Symbols of Kinetic Analysis. The I.U.B. was also responsible for another notable event in 1980 the appearance of a new monthly journal Biochemistry International which commenced in July. The new journal is aimed at the rapid (6-8 weeks) publication of papers in the whole field of biochemistry and the manuscript supplied is reproduced directly. It is claimed that the editors will ensure the scientific merit and originality of material selected for publication. At the other end of the scale in 1980 the Journal of Biological Chemistry achieved its 75th anniversary and still does not require its contributors to give in their reports the EC reference numbers for enzymes.Reports from the 1.U.B Meeting in Toronto in July 1979 are now becoming available. The Ayerst Award Lecture given by James on the X-ray crystallographic approach to enzyme structure and function has now been p~blished,~ as have Gutfreund’s views on future problems in en~ymology.~ Fluorescent probes of active sites of proteinases have been reviewed by Fruton’ and of the flavoproteins by ‘ M. Dixon E. C. Webb C. J. R. Thorne and K. F. Tipton ‘Enzymes’ Longman Group Ltd. London 3rd edn. 1979 [Reviewed by E. M. Crook,Biochem. SOC.Trans. 1980,8,401]. * ‘Enzyme Nomenclature’ Recommendations (1978) of the Nomenclature Committee of the International Union of Biochemistry Academic Press London 1979 [Reviewed by P.Bohley FEBS Lett. 1980 116 1361. M. N. G. James Can. J. Biochem. 1980,58,251. H. Gutfreund Can. J. Biochem. 1980,58 1. ’ J. S. Fruton Mol. Cell. Biochem. 1980,32 105. 347 348 C. A. Ross Massey and Hemmerich.6 Among the structures reported during the year was the crystal structure of cytochrome c peroxidase7 and the primary structures of triosephosphate isomerase8 and of glyceraldehyde 3-phosphate dehydrogenase.' Two reviews and a number of books" have appeared on the topic of immobilized enzymes and there is an interesting paper on the immobilization of NADH in a soluble and enzymically active form.'' Intriguing reports were noted on the bio- chronological applications of amino-acid racemization l2 on sarcosine oxidase which is claimed to be a unique enzyme with both covalently and non-covalently bound flavins,13 and on the preparation of a cytidine-specific RNase from chicken 1i~er.I~ Finally mention must be made of a review of Lowry's method for protein estimation which is widely used in enzyme preparative work." 2 Calmodulin Brief mention of calmodulin was made in last year's Report and since then several important reviews have appeared.16 Recognition of the importance of calmodulin in regulating the metabolism and motility of living cells has now reached such a point that it is almost more pertinent to enquire which cellular mechanisms are not dependent upon calmodulin than to enumerate those which are.Calmodulin exists as a monomer of 148 amino-acids of molecular weight 17 000 with remarkable stability in the presence of Ca2'. The complete sequence of the protein from bovine brain has been publi~hed'~ (see Figure l) and the evidence to date is that the structure has been highly conserved throughout the plant and animal kingdoms. While many similarities exist between calmodulin and other Ca2'-binding proteins such as troponin-C it has as a distinctive feature a post-transcriptionally trimethy- lated E-amino-group of lysine-115 and is also some seven to eight amino-acid residues shorter at the amino-terminal region. Because of the high degrees of homology displayed at the Ca2'-binding sites (especially between sites I and I11 and between I1 and IV in calmodulin) it would seem that the Ca2'-binding proteins arose by gene duplication from a smaller ancestral precursor which bound a single Ca2+ ion.V. Massey and P. Hemmerich Biochem. SOC.Trans. 1980,8 246. T. L. Poulos S. T. Freer R. A. Alden S. L. Edwards U. Skogland K. Takio B. Eriksson N-H. Xuong T. Yonetani and J. Kraut J. Biol. Chem. 1980 255 575. * S. Artavarris-Tskonas and J. I. Harris Eur. J. Biochem. 1980 108 599. J. D. Hocking and J. I. Harris Eur. J. Biochem. 1980,108 567. lo (a)G. G. Bickerstaff Int. J. Biochem. 1980 11 201; (b)T. T. Ngo ibid. p. 459; (c) 'Enzymic and Non-Enzymic Catalysis' ed. P. Dunnill A. Wiseman and N. Blakebrough Ellis Horwood Chichester 1979; (d)'Enzyme Technology Applied Biochemistry and Bioengineering' Vol.2 ed. L. B. Wingard E. Katehalski-Katzir and L. Goodstein Academic Press New York 1979; (e)'Enzyme Engineering Future Directions' ed. L. B. Wingard I. V. Berezin and A. A. Klyasov Plenum Press New York 1980; (f)M. D. Trevan 'Immobilized Enzymes' John Wiley & Sons Chichester 1980. l1 C. W. Fuller J. R. Rubin and H. J. Bright Eur. J. Biochem. 1980,103,421. '* J. L. Bada and S. E. Brown Trends Biochem. Sci. 1980 5 Sept. p. 111. l3 S. Hayashi S. Nakamura and M. Suzuki Biochem. Biophys. Res. Commun. 1980,96 924. l4 (a)C. C. Levy and T. P. Karpetsky J. Biol. Chem. 1980,255 2153; (6) M. S. Boguski P. A. Hieter and C. C. Levy ibid.,p. 2160. l5 G. L. Peterson Anal. Biochem. 1979 100 201.l6 (a)A. R. Means and J. R. Dedman Nature 1980 285 73; (6) C. B. Klee T. H. Crouch and P. G. Richman Annu. Rev. Biochem. 1980 49 489; (c) J. H. Wang and D. M. Waisman Curr. Top. Cell. Regul. 1979,15,47;(d)W. Y. Cheung Science 1980,207,19. D. M. Watterson F. Sharief and T. C. Vanaman J. Biol. Chem. 1980 255 962. Biological Chemistry -Part (iu) Enzyme Chemistry Figure 1 The sequence of bovine brain calmodulin showing the four proposed Ca2+-binding domains with the stretches ofa-helix in darker circles (Reproduced by permission from Annu. Rev. Biochem. 1980,49,489) The tertiary structure of calmodulin is markedly altered in the presence of Ca2' the CY -helical content that is detected by far-u.v. circular dichroism being variously reported as increasing by 5-10%.Conflicting observations on the affinities of the four Ca2+-binding sites and on the co-operativity that exists between them have been published. In a recent paper,'* Crouch and Klee have determined the dissocia- tion constants for the four sites as being between 3 x lop6and 2 x lo-' moll-' and have observed positive co-operativity with Hill coefficients of 1.33 and 1.22 in the absence and in the presence of 3mM-Mg2+ respectively. The binding of metal ion appears to be by a sequential mechanism generating at least four different conformers of the protein (CaM) in its free and liganded state CaM + nCa2+ $ CaM.Ca2+ $ CaM*-Ca2+ (1) Subsequently the active conformer of calmodulin (indicated by an asterisk) interacts with an enzyme which may be inactive or only partially active CaM*.Ca2+ + E $ (CaM*.Ca2+,).E $ (CaM*Ca2+,).E* (2) The stoicheiometry of both equations can clearly vary so giving rise to a number of possible combinations.These will have to be determined for each calmodulin- dependent protein which may itself be capable of binding Ca2' directly. In con- sequence the calmodulin activation system has the capacity to exist in multiple T.H. Crouch and C. B. Klee Biochemistry 1980,19.3692. 350 C. A. Ross forms for the differential regulation of a large number of cellular mechanisms. Thus cyclic nucleotide phosphodiesterase appears to require the calmodulin Ca2+34 complex which it binds with a Kd =1-3 x mol 1-'.l8 On the other hand phosphorylase kinase has the structure where the y-subunit is identical with calmodulin and which can in the presence of calcium ions interact with a further molecule of calmodulin (termed the $-subunit) or with troponin-C of skeletal muscle so that the enzymic activity is increased." This enzyme which exists in a high-activity phosphorylated form (phosphorylase a) and a low-activity dephosphorylated form (phosphorylase b) has varying affinities for calmodulin troponin-C and Ca2'.The interconversion of.forms a and b is under hormonal control and so a most elaborate control system exists linking the regulation by cyclic nucleotides and by Ca2' mediated by at least two Ca*'-binding proteins. In general terms this can be depicted as in Scheme 1. hormonal nervous stimulation stimulation + intracellular Ca" extracellular Ca2+ aden ylate phosphodiesterase I cyclase .\ I \ \ \ . \ \ \ 7... .\ kina< -.\ -21E.CaM.Ca2' dephosphoprotein phosphoprotein / bioc6emical response (E.CaM.Ca2+ may stand for those enzymes indicated by the dashed lines or others listed in refs 16a and 164 Scheme 1 To the lists of calmodulin-mediated processes in refs. 16a and 16d can now be added DNA synthesis,20 which can be initiated in rat liver cells in a low-calcium medium (0.02 mmol 1-') by raising the concentration of metal ion to 1.25 mmol 1-'. The initiation could be blocked by the putative Ca-calmodulin blockers chlor- promazine and trifluoperagine and the block could be removed by adding purified calmodulin to lo-' mol I-').That stimulation of calmodulin does not necessarily require Ca2' has been demonstrated in the case of cyclic-GMP-depen- dent protein kinase." Calmodulin affinity chromatography has been employed in l9 P. Cohen Eur. J. Biochem. 1980 111 563. *' A. L. Boynton J. F. Whitfield and J. P. MacManus Biochem. Biophys. Res. Commun. 1980,95 745. T. Yamaki and H. Hidaka Biochem. Biophys. Res. Commun. 1980,94727. Biological Chemistry -Part (iv) Enzyme Chemistry 35 1 the purification of (Ca” + Mg2+)-dependent ATPase from human erythrocyte membranes.** It is evident that much chemical and physical work is waiting to be done before all the reactions involving calmodulin as a regulator of cellular metabol- ism can be fully delineated. 3 Superoxide Dismutases The univalent reduction of molecular oxygen gives rise to the superoxide radical 02-, which is the conjugate base of the weak acid H02’ (pK = 4.8).Two superoxide anions can interact with the formation of hydrogen peroxide and oxygen 02-+ 02-+ 2H’ -+ H202 + 02 (3) All oxygen-metabolizing cells apparently contain metallo-enzymes that are capable of catalysing reaction (3),and these have been named the superoxide dismutases (EC 1.15.1.1).Just as the catalases and peroxidases have evolved to protect cells from the harmful effects of H202 so it is thought that the dismutases act as a protection against the much more reactive oxygen free-radicals. However that this is not the whole story is now clear since phagocytosis (the ingestion of particles and their destruction by certain cells) results in the abundant production of super- oxide.Such phagocytosis by leucocytes plays a major role in the combating of infections in man and animals.23 It is generally agreed that superoxide itself is not the main antimicrobial or anti-tumour agent but that its importance lies in its ability to react with H202to produce singlet oxygen and hydroxyl radical as shown in reaction (4). metal 02-+ H202 +lo2+ OH-+ OH’ (4) The development of our knowledge of the superoxide dismutase enzymes (SOD) has recently been reviewed.24 There are three distinctly different classes of super-oxide dismutases depending upon the metal content. Copper-zinc SOD was originally found in ox blood and has now been isolated from the cytoplasm of a wide range of animal and plant cells.The enzymes are all of molecular weights near 32 000 consist of two identical subunits and contain one atom of copper and one of zinc per subunit. The SOD isolated from prokaryotic cells was found to be a larger molecule with a molecular weight of 40000 composed of two identical subunits and containing one atom of manganese per molecule. The enzyme originally isolated from mitochondria is also a mangano-enzyme but it has a molecular weight of 80000 and is tetrameric. Subsequently an enzyme has been located in the cytoplasm which contains four manganese atoms per molecule. However it is dissimilar to the mitochondria1 enzyme which closely resembles the prokaryote enzyme and which has been cited as evidence for the symbiosis theory of the origin of mitochondria.The mangano-enzymes appear to be the most widely distributed being found in both eukaryotes and prokaryotes. Finally an iron- containing dimeric SOD has been isolated from bacterial sources. 22 K. Gietzen M. TejEka and H. U. Wolf Biochem. J. 1980 189 81. 23 J. A. Badwey and M. L. Karnovsky Annu. Rev. Biochem. 1980,49,695. 24 (a) I. Fridovich Ado. Enzymol. 1974 41 35; (b)I. Fridovich Annu. Rev. Biochem. 1975 44 147; (c) ‘Superoxide and Superoxide Disrnutases’ ed. A. M.Nicholson J. M. McCord and I. Fridovich Academic Press London 1977; (d)J. V. Bannister and W. H. Bannister Biochem. Educ. 1981,9,42. 352 C. A. Ross The assay of SOD activity presents formidable difficulties since the substrate the superoxide radical has a transitory existence and has to be generated in situ.Pulse radiolysis of pure water generates e-aqand the H' and OH' radicals. In the presence of oxygen 02-and HO,' will be produced ePaq+ o2+ 02- (k= 2 x 10'Ol mol-' s-') (5) H' + 02 -+ H02 (k= 2 x 10" 1 mol-'s-') (6) Formate doubles the yield of 02-and eliminates OH' as shown HCOO-+ OH' -+ COO-+ H20 (7) coo-+ 02 -* COZ + 02- (8) It is thus possible to introduce 02-into aqueous solutions to a concentration as high as 2 x moll-' and the rate constant for the reaction catalysed by bovine erythrocyte SOD was found to be 1.9 x lo91mol-' s-' by direct spectrophotometric monitoring. However more convenient indirect methods of assay have been developed whereby 0,-is generated enzymically by say the action of xanthine oxidase on xanthine in the presence of oxygen.The superoxide is scavenged by a suitable indicator molecule such as cytochrome c or nitroblue tetrazolium and SOD is detected by its ability to inhibit the modification of the detector. One unit of SOD activity can be defined as that which causes 50% inhibition of the modification under specified conditions. Differentiation between the copper-zinc and the manganese enzymes may be made on the basis of the sensitivity of the former to cyanide. To overcome these inherently difficult techniques radioim- munoassay has been de~eloped.'~ In the past year a number of primary structures has been reported. The Cu/Zn superoxide dismutase from human erythrocytes has been investigated by two groups,26 and the published amino-acid sequences are virtually identical with only two amino-acid differences (at positions 17 and 98).The resulting structure has been compared with sequence data for the enzyme from other sources and it shows a high degree of homology. The mammalian enzymes so far examined all have the N-terminal amino-acid blocked by an N-acetyl group. The main structural feature of the protein is a barrel formed by eight antiparallel 0-strands. The region from residues 17 to 30 which shows the most variability in sequence is located on the surface of the molecule and it may be related to the immunochemical properties of this class of enzymes. The amino-acid sequence of Cu/Zn SOD from bakers' yeast has also been as has that for the manganous enzyme from Bacillus stear~thermophilus.~~" In the case of the latter two identical subunits comprise 203 amino-acids each and they show 60% homology with the enzyme of Escherichia coli B.The predicted secondary structure indicates that it is unlike that of the Cu/Zn enzymes. 25 A. Baret P. Schiavi P. Michel A. M. Michelson and K. Puget FEBS Lett. 1980,112 25. 26 (a)D. Barra F. Martini J. V. Bannister M. E. Schinina G. Rotilio W. H. Bannister and F. Bossa FEBS Lett. 1980 120,53;(b) J. R. Jabusch D. L. Farb D. A. Kerschensteiner and H. F. Dentsch Biochemistry 1980,19,2310. '' (a)J. T. Johansen C. Overballe-Petersen B. Martin V. Hasemann and I. Svendsen Carlsberg Res. Commun. 1979,44,201;(6) H. M.Steinman J. Biol. Chem. 1980 255 6758; (c) C.J. Brock and J. E. Walker Biochemistry 1980,19 2873. Biological Chemistry -Part (iu) Enzyme Chemistry The metal-binding sites in the subunit of Cu/Zn SOD consist of clusters of four residues (His-61 His-69 His-78 and Asp-9 1) in an approximate tetrahedral arrangement around the zinc atom and His-44 His-46 His-61 and His-118 in a plane with the copper. N.m.r. studies on the bovine erythrocyte enzyme and on two isoenzymes from wheat germ28 have revealed substantial structural homology. It has been shown that zinc plays a structural role while copper is catalytically essential being alternately reduced and re-oxidized. The two subunits are capable of exhibiting the same activity when paired with a native partner or with a modified inactive partner and Fridovich could find no evidence for half-of -the-sites reac- ti~ity.,~ Other investigations into the nature of the active site have implicated Arg-143 in the catalytic process because the enzyme is rendered inactive when that residue is modified with phenylgly~xal.~' The presence of a guanidinium group in close (6 A) proximity to the Cu" is seen either as providing electrostatic guidance to the incoming 0,-or as proton conduction.Transfer of an electron from Cu' and of a proton from the arginine in the second half of the catalytic cycle would yield a leaving group H20- which would be protonated to H202 in solution. Alternatively modification of Arg- 143 may simply perturb the active-site region. A somewhat surprising result from perturbed angular correlation (PAC) of y-rays spectroscopy on a yeast Cu/Cd SOD derivative has shown that His-63 (equivalent to His-61 in the mammalian enzyme) is not simultaneously co-ordinated to both the Cu and Zn.31 Two 'cautionary tales' have come from the Fridovich laboratory at Duke Univer- sity Medical Center during the past year.32 The first concerns the possible discovery of an inhibitor of SOD activity.Since mammalian cells do not contain a ferro- superoxide dismutase but many prokaryotes do a very extensive search has been made to find an inhibitor of this enzyme which might well then act as a bactericidal agent. Pamoic acid (1)was proposed as one such compound and 2,5-dimethylphenol as another. However it has been shown that such compounds react with 0,-to yield radicals which can reduce the scavenging indicators that are used in a number of assay methods for SOD.When the compounds were tested on the inhibition of the 02-dependent generation of ethylene from methional by SOD they had no effect. The other re-examination concerns the putative superoxide dismutase activity CH I (1) 28 A. R. Burger S. J. Lippard M. W. Pantoliano and J. S. Valentine Biochemistry 1980 19 4139. 29 D. P. Malinowski and I. Fridovich Biochemistry 1979,18,237. 30 (a)D. P. Malinowski and I. Fridovich Biochemistry 1979 18 5909; (b)C. L. Borders Jr. and J. T. Johansen Biochem. Biophys. Res. Commun. 1980,96 1071. 31 R. Baner I. Derneter V. Hasemann and J. T. Johansen Biochem.Biophys. Res. Commun. 1980,94 1296. 32 (a)H. M. Hassan H. Dougherty and I. Fridovich Arch. Biochem. Biophys. 1980 199 349; (b)J. Diguiseppi and I. Fridovich ibid. 1980 203 145. 354 C.A. Ross of Fe-EDTA which has variously been reported as exhibiting 0.01 to 0.1% of the efficiency of the enzyme. Again it has been shown that the Fe-EDTA complex interfered with the usual assay methods of SOD activity i.e. those utilizing cyto- chrome c or nitroblue tetrazolium by inhibiting the xanthine oxidase activity. The complex does not catalytically scavenge 02-,as does SOD and it had no effect upon the method of assay involving the photo-oxidation of dianisidine. The lesson appears to be that a wide range of assay systems must be employed when evaluating inhibitors and mimics of superoxide dismutase activity.4 Enzyme Kinetics The eightieth birthday of Malcolm Dixon (18 April 1979) was honoured by the Molecular Enzymology Group of the Biochemical Society in the form of a Col- loquium the proceedings of which have now been published. In this publication Tipt~n~~ discusses the concept of kinetic mechanisms being advantageous in terms of specific metabolic functions of enzymes. New books devoted to the subject of enzyme kinetics (which have been so plentiful in recent years) have been virtually absent in 1980. A revised edition of Cornish-Bowden’s textbook (under a slightly different title34) arrived as did two volumes in the ‘Methods in Enzymology’ series devoted to enzyme kinetics and mechani~m.~~ The first of these deals with initial-rate methods and with inhibitor and substrate effects and it consists of twenty contribu- tions from many of the leaders in the field.The second volume comprises six articles on isotope probes and eight on complex enzyme systems including hysteretic and immobilized enzymes and co-operativity in enzyme function. We are promised more volumes edited by Purich in this series. With regard to kinetic data and analysis the debate continues on the best way to proceed. In a contribution to the ‘Textbook Errors’ series in Trends Biochem. Sci. Cha~lin~~ observes that one of the reservations made when deriving the Michaelis-Menten equation namely that substrate concentration always exceeds enzyme concentration to such a degree that ([So] -[ES]) [So] is unnecessarily restrictive.At high enzyme concentrations a modified Michaelis-Menten equation can be simply derived vo = vmax[s0l/(~m + [Sol + LEO]) (9) Chaplin has a point for in this day and age enzymes are frequently available in such purity and of known molecular weight that [E,] can be calculated. In a review article Atkins and Nimmo3’ have examined common approaches to determining the usual kinetic parameters and they favour the determination of initial velocities and analysing the data by least-squares treatment rather than attempting the analysis of progress curves. In contrast Fukagawa3* has published a computer program for 33 K. F. Tipton Biochem. SOC. Trans. 1980,8,242. 34 A. Cornish-Bowden ‘Fundamentals of Enzyme Kinetics’ Butterworths Boston 1979 [Reviewed by H.Gutfreund FEBS Lett. 1980,116 1251. 35 ‘Enzyme Kinetics and Mechanism Part A’ and ‘Enzyme Kinetics and Mechanism Part B’ ed. D. L. Purich ‘Methods in Enzymology’ Vol. 63 and Vol. 64 ed. S. P. Colowick and N. 0.Kaplan Academic Press New York 1979 and 1980. 36 M. F. Chaplin Trends Biochem. Sci. 1981,6 Jan. p. IV. 37 G. L. Atkins and I. A. Nimmo Anal. Biochem. 1980,104 1. Y.Fukagawa Biochem. J. 1980,185 186. Biological Chemistry -Part (iv)Enzyme Chemistry the analysis of the progress curve of a single reaction (obtained by spec- trophotometry) and has applied it to p-lactamase. Chou has described a new graphical method for simplifying the calculation of reaction rates of steady-state enzyme-catalysed The Lee and Wilson modification of the double-reciprocal plot has been re-e~amined.~’ It was claimed for this modified treatment in which the so-called log mean substrate concentration {([So]-[S,])/ln([S,]/[S,])} was replaced by the arithmetic mean substrate concentra- tion {([So] + [S,])/2} in the integrated Michaelis-Menten equation that the para- meters V,, and K could be determined with little error even when half of the substrate had been consumed in a reaction.This now appears not to be the case. The validity of the use of the ‘jack-knife’ statistical technique in analysing kinetic data for enzymes has been defended41 against criticism. Kohberger4’ has evaluated the Direct Linear Plot and has produced statistical evidence to show that whereas harmonic spacing of substrate concentrations is in general a more efficient experi- mental procedure it should be used with a weighted least-squares analysis.The Direct Linear Plot performs better with substrate values chosen in arithmetic spacing and has been used to evaluate individual rate and association constants.43 In the serine-protease-catalysed hydrolysis of substrate S added nucleophile (N) provides an alternative pathway as shown in Scheme 2. Additions of nucleophile K (PI P2 P3 are products) k EA+yL+p, E+S $ ES+ P1 Scheme 2 lead to decreases in k,, (which eqFals k2kz/kz+ k3)and to increases in K,. The intersection points (co-ordinates K and V) lie on a straight line with intercept K on the K axis and k3[Eo]on the V axis (see Figure 2).The rate constant of acylation k2,may be determined from the ratio a/b = k2/k3. So-called suicide substrates (mechanism-based inactivators) are useful as probes of active sites or in the search for effective drugs. It is characteristic of their action that formation of products and inactivation of enzymes proceed concurrently and in a constant ratio (the partition ratio) throughout the course of the reaction. Scheme 3 due to Walsh et al. (1978) where X and Y are enzyme complexes and Ei is inactive enzyme has now been analysed by wale^,^^ and rate equations for inactivation have been derived. E+S $ X+Y-*E+P L Ei Scheme 3 39 (a)K. C. Chou and S. Forsin Biochem. J. 1980 187 829; (b)K. C. Chou Eur. J. Biochem. 1980 113 195.40 N. G. Karanth and A. K. Srivastava Biochim. Biophys. Actu 1980,615 279. 41 A. Cornish-Bowden and J. T.-F. Wong Biochem. J. 1980,185 535. 42 R. C. Kohberger Anal. Biochem. 1980,101 1. 43 V.Dorovska-Taran and D. Raykova Biochim. Biophys. Actu 1980,615 509. 44 S.G. Waley Biochem.J. 1980 185 771. 356 C.A. Ross V A 1 I bl I >I I I I I I a1 I I I I 1 I I Figure 2 Determination of individual rate and association constants of hydrolysis in Eisenthal and Cornish-Bowden co-ordinates (1) without added nucleophile ;(2)and (3),with added nucleophile (Modified from Biochim. Biophys. Acta 1980 615 509) Non-hyperbolic behaviour of enzymes that bind ligands has received attention from a number of authors.Knack and R~hm~~ have developed a least-squares fit of the generalized Hill equation s = s,,,[L]"'L'/K + [LpL' (10) where the Hill coefficient is allowed to vary with [L] and is no longer treated as a constant. As a result curves for incomplete saturation may be analysed i.e. where S,, is unknown and the concentration dependence of Hill slopes that is arrived at may indicate more elaborate models for testing in specific cases. A review46 has been made of the usual graphical plots used in ligand-binding studies i.e. [bound]/[free] us [bound] [bound]/[free] us [free] and [bound]/[total] us [total] and the conclusion has been drawn that these can only provide a first approximation and that weighted non-linear least-squares treatment is required.In a novel approach to the estimation of the Hill parameters V and n) substrate was continuously added to a single reaction mixture and absorbance was analysed by a tangent-slope pr~cedure.~' A Gilford spectrophotometer has been suitably adapted to allow for the continuous addition of substrate to a stirred assay mixture and the absorbance changes were recorded encoded on paper tape. A number of enzymes has been investigated including immobilized lactate dehy- drogenase and the results obtained are in reasonable agreement with those already 45 1. Knack and K.-H. Rohm Biochim. Biophys. Actu 1980,614,613. 46 A. K. Thakur M. L. Jaffe andD. Rodbard Anal. Biochem. 1980,107,279. 47 D. J. LeBlond C. L. Ashendel and W. A. Wood Anal. Biochem..1980,104,355,370. Biological Chemistry -Part (iv) Enzyme Chemistry published. Finally in an extensive review48 of twelve enzymes out of some 800 which are said to exhibit deviations from Michaelis-Menten kinetics the conclusion is reached that at present without simplifying assumptions theoretical studies of involved mechanisms create equations of such complexity that it is virtually imposs- ible to interpret them in terms of K,, and V. 5 Enzyme Mechanisms Molybdenum and molybdenum-containing enzymes of which xanthine oxidase and dehydrogenase are examples have been the subject of recent reviews,49 and the kinetic mechanism of the xanthine-oxidizing enzymes has been re-examined.” These enzymes are dimeric and contain a molybdenum atom a flavin moiety and two iron-sulphur clusters in each subunit.The reducing substrate (xanthine) binds to the molybdenum and the oxidizing substrate (NAD’ or 0,) to the flavin; the iron-sulphur centres act as electron sinks. The generally accepted view was that these enzymes functioned in a classical ‘ping-pong’ manner (see Scheme 4) and indeed double-reciprocal plots did yield sets of parallel lines when one substrate was held at constant concentration while the other was varied and vice versa. This mechanism requires that the reducing substrate binds to the oxidized form of the enzyme and leaves as an oxidized product before the addition of the second substrate (oxidizing). However it has now been shown (by e.p.r.) that xanthine can bind to the reduced form of xanthine oxidase and (by spectrophotometry) that NAD’ can bind to the oxidized form of xanthine dehydrogenase.In consequence Coughlan and Rajagopalan’’ have proposed a ‘rapid-equilibrium random (two site) hybrid ping-pong’ mechanism such as has been proposed for the biotin-containing transcar- boxylase and they have applied Cha’s treatment to their scheme consisting of three rapid-equilibrium random segments (Scheme 5). The letter E represents the X U NAD’ NADH (X= xanthine; U =uric acid) Scheme 4 Scheme 5 48 W. G. Bardsley P. Leff J. Kavanagh and R. D. Waight Biochem. J. 1980,189 739. 49 (a)‘Molybdenum and Molybdenum-containing Enzymes’ ed. M. P. Coughlan Pergamon Press Oxford 1980; (6)‘Molybdenum Chemistry of Biological Significance’ ed.W. E. Newton and S. Otsuka Plenum Press New York and London 1980; (c) R. C. Bray Adv. Enzymol. 1981,51 107. 50 M. P. Coughlan and K. V. Rajagopalan Eur. J. Biochem. 1980,105,81. 358 C.A. Ross enzyme with the top and bottom bars denoting the molybdenum and flavin sites respectively. When all substrates and products are present there are eight possible complexes (and one form of the free enzyme) at each of the segments (see Scheme 6). From published data on a number of other enzymes the authors state their belief that many multi-component redox enzymes will be found to operate via such a two-site mechanism. Jp NAD + E" L_. NAD + u+E+x NAD + G -Ex + + + NADH NADH NADH Jp Jp Scheme 6 The control of the pentose phosphate pathway (hexose monophosphate shunt) is not well understood but it is generally agreed that the first committed step i.e.that catalysed by glucose 6-phosphate dehydrogenase (EC 1.1.1.49),is critical to such control. The mammalian enzymes are NADP-preferring and are inhibited by reduced coenzyme. In view of the high NADPH/NADP' ratios prevailing in many tissues it is of interest to ascertain how the enzymes can function at all let alone be controlled. Following reports that some glucose 6-phosphate dehydrogenases exhibit dual nucleotide specificity and that the kinetic mechanism differs for the NAD'-and NADP'-linked reactions the enzyme from lactating rat mammary gland has been in~estigated.~~ The authors are unable to state unequivocally that the mechanism is 'partial rapid-equilibrium random' and their explanation for the two classes of binding site for glucose-6-P appears to be at variance with their own evidence of competitive inhibition by NADPH with respect to both NADP' and NAD' indicating that there is only one catalytically active form of the enzyme.The kinetic mechanism of glutamate dehydrogenase (EC 1.4.1.3),previously considered to be random in both directions has been re-examined by Rife and Cleland.52 The forward reaction does indeed appear to be random with the nucleotide dissociating more rapidly than the amino-acid from the ternary complex. D. S. Shrere and H. R. Levy J. Biol. Chem. 1980,255 2670. 52 J. E. Rife and W. W. Cleland Biochemistry 1980 19 2321 2328. Biological Chemistry -Part (iv) Enzyme Chemistry L B A Scheme 7 In the reverse direction however the addition of keto-acid and ammonia to the enzyme-NADPH complex appears to be ordered and with the combination of nucleotide and keto-acid largely ordered as shown in Scheme 7 which correctly predicts the non-competitive or uncompetitive nature of substrate inhibition pat- terns.However ketoglutarate will combine with E-NADP' and so gives rise to the inhibitory E-NADP'-KG complex as well as combining with E-NADPH to give E-NADPH-KG to which ammonia binds. The two complexes occur in constant ratio as the ketoglutarate concentration is altered. The Scheme does not take into account the fact that NADP' can dissociate from the E-NADP'-KG complex and so allow the reaction to proceed as evidenced by the partial inhibition that is seen when the concentration of ammonia is varied.In a second paper the authors report the pH profiles for the binding of keto-acids and analogues to E-NADPH and show that a group with a pK of 5 (possibly carboxyl) has to be protonated for binding to occur and that a second group (possibly lysine) with a pK of 7.8 has to be protonated to bind the 5-carboxyl of dicarboxylic acids. The V/K profiles for ammonia indicate that the substrate is the neutral molecule and those for glutamate (or norvaline) show that the amino-group of the amino-acid needs to be protonated while a group (possibly lysine) with a pK of 7.6-8 in the enzyme must be unprotonated for activity. Based on these observations a mechan- ism is proposed in which there is direct attack of ammonia on ketoglutarate to give a carbinolamine with a lysine residue supplying a proton.The transfer of a proton from N to 0 of the carbinolamine (possibly catalysed by a carboxyl group) is followed by elimination of water and the resulting iminoglutarate is reduced by NADPH to glutamate. The reaction is completed by the protonation of the amino- group of glutamate by the catalytic carboxyl group (Scheme 8). From amino-acid modification studies it is known that the lysine residues 27 and 126 are unusually reactive and have low pK values (8.2 and 7.8 respectively). It is not yet possible to assign the catalytic role or the binding role to a particular lysine residue. Also from Cleland's laboratory have come reports5 of an elegant kinetic method for measuring dissociation constants of metal complexes with ATP and ADP.The '' (a)J. F. Morrison and W. W. Cleland Biochemistry 1980 19 3127; (b)R. E. Viola J. F. Morrison and W. W. Cleland ibid. p. 3 13 1. 360 C.A. Ross F+.+ HH I I I 00 0 o\c40 \/ 0\c/ I I I .wc IM. H HH I I 0\&0 0 -o\cl.p ac/0 I uI* I I Scheme 8 determination requires that the metal-ATP complex acts as an inhibitor of an enzyme which requires MgATP and it is the ratio of the dissociation constants for MgATP and inhibitory metal-ATP complex that is measured. Since the values of dissociation constants for MgATP under a wide variety of conditions are known the dissociation constant of metal-ATP can be readily calculated.Using the method with hexokinase the dissociation constants for EuATP and GdATP have been calculated. In the course of the work it has been confirmed that the apparent activation of hexokinase by citrate is due to the chelation of aluminium in AlATP which is a contaminant of commercial ATP to form aluminium citrate. 6 Phosphofructokinase Phosphofructokinase (EC 2.7.1.1 1) catalyses the phosphorylation of fructose 6- phosphate to fructose 1,6-bisphosphate [equation (1l)],and being the first com- mitted step in glycolysis is not unexpectedly subject to a high degree of regulation. fructose-6-P2-+ MgATP4-$ fr~ctose-1,6-P~~-+ MgADP3-+ H+ (11) The enzyme has naturally excited the interest of researchers since it was first reported by Dische in 1935 and continues to do so as evidenced by two recent reviews54 and a 54 (a) K.Uyeda Adv. Enzymol. Relat. Areas Mol. Biol.,1979 48 193; (b) A. R. Goldhammer and H. H. Paradies Curr. Top. Cell. Regul. 1980 15 109. Biological Chemistry -Part (iu)Enzyme Chemistry 361 number of important papers which have appeared during the past couple of years. Phosphofructokinases (PFK) from various mammalian tissues and from yeast and prokaryotes have been shown to differ in their physical and chemical properties. The mammalian enzymes are more complex; the smallest enzymically active form is a tetramer of molecular weight ca. 3.2-3.8 x lo5and of S20.w= 13 but aggregations occur by self-association of subunits dependent upon the concentration of protein.The latest available electron micrograph^'^" show the porcine liver enzyme to occur in tetramers under suitable dissociating conditions; a few octamers occur dimers are rare and hexamers are not seen. For this reason the authors argue that Figure 3 represents the most likely orientation of identical subunits since it alone features distinct isologous dimerdimer and tetramer-tetramer bonds. Lardy at Wisconsin has continued his long association with research on PFK by publishing observations on the aggregation of the enzyme as observed by fluorescence p~larization.~~ He has shown that MgATP or fructose 6-phosphate (Fru6P) are the dominant influencing ligands and that binding Fru6P shifts the equilibrium in favour of association.In a steady-state analysis of rat liver PFK Lardy has shown the enzyme to have an extremely low affinity and a high degree of positive co-operativity towards Fru6P (K,= 6 mmol I-*; Hill coefficient >4). Figure 3 Subunit interactions within phosphofructokinase. Arbitrarily assigned ‘top’ subunits are marked T;‘bottom’ faces are blank. Subunit-binding sets are marked with one two or three lines to show the orientation of each subunit ilt the plane of the page The existence of five isomers of human PFK has been demonstrated.” The liver and muscle enzymes are homotetramers i.e. L4 and M4 respectively but the enzyme isolated from erythrocytes is a heterogenous mixture of five isoenzymes arising from the association of L and M subunits. Another human isoenzyme F4 the fibroblast type has recently been reported and chara~terized.~~ It exhibits allosteric behaviour to a much lesser degree than the other enzymes in keeping with its occurrence in cells where the control of glycolysis is not of paramount importance.The phosphofructokinases exhibit regulatory behaviour towards both substrates and both products and they are activated by AMP ADP CAMP NH4+ and inorganic phosphate (and also sulphatessb) and are inhibited by citrate phosphoglyc- erates and creatine phosphate. In addition the activity of the enzyme appears to be controlled by aggregation (referred to above) and by phosphorylation. The ’’ L. G. Foe and J. L. Trujillo (a)I. Biol. Chem. 1980,255 10 537; (b)Arch. Biochern. Biophys. 1980 199,l.56 G.D. Reinhart and H. A. Lardy Biochemistry 1980,19,1484,1491,1477. 57 S.Vora C.Seaman S. Durham and S. Piomelli Proc. Narl. Acad. Sci. USA 1980,77,62. ’* M. C.Meienhofer D. Cottreau J. C. Dreyfus and A. Kahn FEBSLert. 1980,110 219. 362 C.A. Ross catalytic subunit of CAMP-dependent protein kinase is capable of phosphorylating a serine residue in a terminal sequence that has been isolated by limited proteolysis; the reaction is profoundly affected by the presence of allosteric ligand~.~' The binding of ATP to the enzyme has recently been extensively investigated by Kellett in York.60 Using techniques similar to those developed by Trentham for probing the kinases Kellett has used a fluorescent analogue of ATP namely 1,N6-ethano-ATP for use in stopped-flow fluorimetry.The binding of the probe (L) to the catalytic site is consistent with a two-step mechanism E + E* E* + L $ E*L in which there is interconversion of the enzyme from one form E to another E". The extent of the allosteric transition from the R to the T conformation as induced by the ATP analogue was determined from the amplitude of the slow phase of its fluorescence enhancement. The activators cyclic AMP and fructose 1,6-bisphos- phate decreased the amplitude while an inhibitor citrate increased it. The confor- mation of the enzyme itself was monitored by intrinsic protein fluorescence the R conformation having diminished fluorescence compared with the T conformation. MgATP exerted a complicated effect enhancement at low concentrations and quenching at high concentrations resulting from binding to the inhibitory site followed by allosteric transition.Enhancement reflects the extent of the transition and it involves tyrosine and tryptophan probably in the vicinity of the active centre. Quenching reflects occupancy of the inhibitory site and involves a tyrosine. The binding site for cyclic AMP has been investigated by Hammes61 and its distance from a reactive sulphydryl group has been measured by fluorescence resonance energy transfer. The nucleotide-binding site was labelled with FSBas A {Sf-[ p-(fluorosulphonyl)benzoyl]-2-aza-l,N6-ethenoadenosine} and the sulphydryl group with NBD-Cl (7-chloro-4-nitrobenzo-2,1,3-oxadiazole)tand DDPM {N-[4-(dimethylamino)-3,5-dinitrophenyl]maleimide}.The distance was found to be 28 A; Hammes had previously shown the distance between the sulphydryl group and the citrate-binding site to be 40 A.By covalent modification with p-fluorosulphonyl ['4C]benzoyl-5'-adenosine,a peptide has been isolated that contains a lysine residue that has been identified as being in the binding site for the allosteric activators CAMP AMP and ADP.62 The complete amino-acid sequence of the subunit of the PFK from Bacillus stearothermophilus has been reported from the laboratory of the late J.I. Harris (see also refs. 8 and 9).63 t [4-Chloro-7-nitrobenzofurazan. Senior Reporter] 59 P. T. Riquelme and R. G. Kemp J. Biol. Chem. 1980,255,4367. 6o D.Roberts and G. L. Kellett Biochem. J. 1979,183 349; 1980,189 561,569.61 D.W.Craig and G. G. Hammes Biochemistry 1980,19 330. 62 L.Weng R. L. Heinrickson and T. E. Mansour J. Biol. Chem. 1980,255 1492. 63 E.Kolb P. J. Hudson and J. I. Harris Eur. J. Biochem. 1980,108 587.