The Nature of the Proton Transfer from an Acid Group at the Active Site of an Enzyme to Solvent Water The Extent of 2Hand 3HTransfer in the Reaction catalysed by Triose Phosphate Isomerase BY L. MARKFISHER,* w. JOHNALBERY? AND JEREMYR.KNOWLES" Department of Chemistry Harvard University Cambridge Mass.( *) and Physical Chemistry Laboratory University of Oxford(?) Received 2nd May 1975 The extent of transfer of a 2H and of a 3H label in 1-R-[*H or 3H]-dihydroxyacetone phosphate to the 2-position of D-glyceraldehyde 3-phosphate in the reaction catalysed by triose phosphate isomerase has been determined. Since the enzymic base that abstracts the substrate's isotopic label is a carboxylate group this enzyme-substrate system effectively provides -COOZH and -CO03H in a solution of 'H2O,and allows an investigation of the transfer of ZH and 3H from the carboxyl group to unlabelled water.From the value of the fractionation factor for this proton transfer it is evident that the mechanism of exchange of isotope on -COOL with the protons of the solvent does not involve a transition state in which L is in flight. The glycolytic enzyme triose phosphate isomerase catalyses the interconversion of dihydroxyacetone phosphate (I) and D-glyceraldehyde 3-phosphate (IIj through abstraction of the 1 pro-R hydrogen of the dihydroxyacetone phosphate or of the hydrogen at C-2 of the D-glyceraldehyde 3-phosphate by a catalytic base at the active site of the enzyme.l Experiments with specific active-site-directed inhibitors and recent crystallo- graphic work on the enzyme strongly suggest that this base is the y-carboxyl group of a glutamate residue.This assignment is consistent with the pH-dependence of the enzyme activity and attractively accommodates a number of other features of the catalysed reaction. When 1-R-[3H]-dihydroxyacetonephosphate is incubated with isomerase under conditions where the initial product D-glyceraldehyde 3-1 54 L. M. FISHER W.J. ALBERY AND J. R. KNOWLES phosphate is oxidised to 3-phosphoglycerate as fast as it is formed only some 6% of the tritium label originally at C-1 is transferred (after complete reaction) to C-2 of the final product 3-phosphoglycerate. This demonstrates that the catalytic base having abstracted the carbon-bound proton can exchange this proton with the solvent water during the course of the enzyme-catalysed reaction.A plausible scheme for catalysis by the isomerase involving an enediol intermediate has been proposed (see fig. I) and a generalised kinetic treatment has been derived that allows the rate constants for the individual steps of the reaction to be evaluated from the results of a number of isotopic experimenk6 This paper is concerned with the isotope exchange reaction between the conjugate acid of the catalytic base B-L (fig. l) and solvent FIG.1.-Transfer of hydrogen isotope in the isomerisation of dihydroxyacetone phosphate by triose phosphate isomerase. B-is the active-site carboxyl group L is 2Hor 3H,@ represents a phosphoryl group and the heavy arrows indicate the fate of the isotopic label.All the species represented are enzyme-bound. water. The extent of deuterium transfer from 1-R-[2H]-dihydroxyacetonephosphate to the final product 3-phosphoglycerate has been measured under conditions essentially identical to those of the tritium transfer experiment mentioned above.5 The results from these two experiments are used to derive the isotope effect for the reaction involving proton exchange between solvent water and the carboxyl group at the active site of the isomerase. The enzyme-substrate system effectively provides -C002H and -CO03H in a solution of 'H20 and allows an investigation of the transfer of 2Hand of 3H from the carboxyl group to unlabelled water. From the full kinetic treatment of the reaction scheme laid out in fig.1 the fraction of isotope transferred (p") from 2H- or 3H-labelled I -R-dihydroxyacetone phosphate to the 2-position of the product 3-phosphoglycerate under conditions where the isomerase reaction is rate-limiting approximates to where k refers to the rate of the irreversible isotope exchange reaction with solvent ks B-L + H2O + B-H + HOL and kt is the composite rate constant for the collapse of the enediol to D-glyceraldehyde phosphate and its departure from the enzyme's active site. From the knowledge of the partitioning ratio of the labelled enzyme-bound enediol intermediate (i.e. the extent of transfer of 2H and of 3H) the isotope effect for the k exchange reaction of eqn (2) is obtained from the Swain-Schaad equation.' PROTON TRANSFER TO WATER FROM AN ENZYME ACID EXPERIMENTAL Dihydroxyacetone phosphate stereospecifically labelled with deuterium in the 1-R position was prepared by equilibration of dihydroxyacetone phosphate in 2Hz0 (99.8 %) using chicken muscle triose phosphate isomerase.8 Ion exchange chromatography yielded l-R-[2H]-dihydroxyacetonephosphate essentially free of ~-2-[~H]-glyceraldehyde 3-phosphate (at equilibrium the mixture contains 96% of dihydroxyacetone phosphate and 4% of D-glyceraldehyde phosphate) and the extent of deuterium incorporation was determined by mass spectrometric analysis of the volatile derivative 1-[2H]-tetrakis(trimethyIsilyl)-a-glycerolphosphate using an AEI MS9 instrument.For the transfer experiments 1-R- [2H]-dihydroxyacetone phosphate NAD+ EDTA and sodium arsenate were dissolved in 100 mM triethanolamine-HC1 buffer pH 7.6 and equilibrated in an optical cuvette at 30°C.The amount of any contaminating ~-2-[~H]-glyceraldehyde 3-phosphate was determined by enzymic assay using isomerase-free glyceraldehyde 3-phosphate dehydrogenase. The isomerase reaction was then initiated in the same cuvette by the addition of a small (rate- limiting) quantity of triose phosphate isomerase. The progress of the reaction was foilowed to completion by monitoring the increase in absorbance at 366 nm due to the formation of NADH. The product of the reaction 3-phosphoglycerate was isolated and purified by ion exchange chromatography. This material was methylated using diazomethane.Mass spectra of the methylated product were obtained at a probe temperature of 45" and 70 eV by direct insertion using an AEI MS9 instrument. The peaks at mle 169 and mle 170 were scanned slowly and repeatedly and the extent of deuterium transfer in the isomerase- catalysed reaction was determined from the averaged mle 170:169 intensity ratios of the labelled methylated 3-phosphoglycerate and of the unZabeZZed material obtained similarly. As a check of both the extent of labelling of the original dihydroxyacetone phosphate and of the position of the isotopic label in the product 3-phosphoglycerate a very little tritiated water was added to the 2H20 used for the labelling of the dihydroxyacetone phosphate. This radioactive tracer was also used (see below) as a check on the fate of the deuterium label.RESULTS AND DISCUSSION The intense (M+-59) peak at m/e 169 in the mass spectrum of methylated 3- phosphoglycerate is formed from (111) by facile loss of -COOCH,. Incorporation of deuterium at the C-2 position of 3-phosphoglycerate increased the intensity at m/e 170 relative to that at m/e 169. Data from the mass spectra of four samples of L \ OCH 3-phosphoglycerate obtained from separate deuterium-transfer experiments gave the following per cent deuterium contents 6.5 6.3 5.0 and 5.9 %. Each of these values was obtained from between 5 and 14 scans of the mass spectrum. The starting 1-R-[2Hj-dihydroxyacetone phosphate was shown by mass spectrometry to be essentially completely deuterated at C-1 and the percentage deuterium transfer was L.M. FISHER W. J. ALBERY AND J. R. KNOWLES corrected in each case for any small contribution from contaminating D-2-L2 HI-glyceraldehyde 3-phosphate. The extent of deuterium transfer from 1-R-[2H]-dihydroxyacetone phosphate to 3-phosphoglycerate catalysed by the isomerase is thus about 5.9%. To confirm the location of the isotopic label in the product 3-phosphoglycerate use was made of the tritium tracer initially incorporated into the starting 1-R-[2H]-dihydroxyacetone phosphate. A portion of the 3-phospho- glycerate product was treated with the enzymes phosphoglycerate mutase and enolase and the tritium specifically labilised from C-2 of the 3-phosphoglycerate (see fig. 2) was recovered as tritiated water by distillation.coo-COO-coo-\ L-C-OH L-C-O-@ \ I yLI I/ + CH2 CH2 CH LOH 0-@ OH FIG.2.-Check on the position of the isotopic label in 3-phosphoglycerate. a phosphoglycerate mutasei-2 3-diphosphoglycerate as cofactor ; b enolase. The isotopically-labelled conjugate acid of the base at the active site of the isomerase may suffer one of two fates (see fig. I). The label may be transferred to the enediol intermediate forming labelled D-glyceraldehyde 3-phosphate which is then lost from the enzyme and converted to 3-phosphoglycerate labelled at C-2 (the k route). Alternatively the label may be irreversibly exchanged for a proton from solvent water resulting ultimately in the formation of unlabelled 3-phosphoglycerate (the k route). Clearly the extent of transfer of label to C-2 of 3-phosphoglycerate is a measure of the relative fluxes of material along these two pathways.From the work of Herlihy et a/.,5we know that for tritium transfer pT = 0.058 f. 0.004. That is after complete conversion of 1-R-[ 3H]-dihydroxyacetone phosphate to product 5.8f0.4% of the tritium label is transferred intramolecularly. In the present work we have found that for deuterium transfer pD = 0.059+0.005. From eqn (1) we have where P is derived from the experimental values of pT and pD. From the Swain Schaad relationship for any reaction kD/kH= (kT/kD)2*3, so In terms of the deuterium fractionation factors & and 4t for the transition states for the two routes 4 = 4 P-2*3. (5) Now by performing the enzyme-catalysed reaction in tritiated water lo and comparing the tritium content of the product with that of the solvent we know the tritium fractionation factor on the k route Qt = 0.83f0.01 from which the Swain- Schaad relationship gives q5t = 0.88 f.O.01.This fractionation factor is slightly less than unity since the rate of transfer is controlled mainly by the loss of product from the enzyme (4 = 1.0) but partly by the proton transfer step as the bound enediol collapses to bound D-glyceraldehyde 3-phosphate (4-0.2 to 0.3). From the above PROTON TRANSFER TO WATER FROM AN ENZYME ACID equations we find the values of +s from our four experiments to be 1.1 1.1 0.61 and 0.92. +s is therefore 0.94f0.17. This value is close enough to unity for it to be clear that the mechanism of exchange of the isotope on -COOL with the protons of the solvent does not involve a transition state in which L is in flight.We may envisage three kinds of pathway for this isotopic exchange. First the pathway may involve ionisation solvent exchange and reprotonation (fig. 3). At pH 7 the solvent exchange step will be much faster than the diffusive encounter of H30+and B- and provided that the diffusive step is rate-limiting our observed value for the fractionation factor is consistent with this mechanism. diffusion -1.0 Jt solvent B-..*H-O + H,O+ reorgoniration far? B-... L -0 + H,O+ I 4 -1.0 H H FIG.3.-Stepwise proton exchange process between a labelled acid B-L and HzO. The exchange of a carboxyl proton -COOL however may occur in a cyclic manner within a hydrogen-bonded complex such as has been proposed to account for the H n.m.r.of acetic acid-acetate buffers (see fig. 4). In such a mechanism H /” L transfer H -0. c ‘H -R-C Lo-c ...0 ’ 4-0.2 ‘H H transfer L -4 -1.0 fast solvent I reorganisation FIG.4.-Cyclic proton exchange for a carboxylic acid the second proton transfer must be rate-determining since H20 is a weaker base than R-COO- and solvent reorganisation will be faster (at 108-109 s-l) than the cyclic proton transfer (106-107 s-l). The effect of L on the rate-determining transition state is now secondary and since the transition state is symmetrical the fractionation factor for this step will be ,/l = 40.69 = 0.83 (see ref.(12)) [where I is the fraction- ation factor for the process L2HO++3D20 +L2DO+++H20(ignoring statistical factors)]. This is also consistent with our experimental value. Since the cyclic mechanism is some 102-fold faster than the normal dissociation path we may expect L. M. FISHER W. J. ALBERY AND J. R. KNOWLES this route to be more probable provided that the steric constraints of the active site allow it. Thirdly it is conceivable that during the lifetime of the enediol intermediate (i.e. of B-L) there is complete and rapid exchange off with a limited pool of about 3 water molecules that are isolated from the bulk water during the reaction. In this case as soon as the substrate or product leaves the active site these water molecules would be free to exchange with bulk solvent and a fresh set of unlabelled water molecules would be trapped by the next substrate to bind.This situation could explain the absence of a measurable difference between the extent of transfer of 2H and 3H. We do not favour this interpretation since it does not agree with the characteristics of enzyme active sites in general nor with the active site of triose phosphate isomerase in partic~lar.~ From the crystal structure of the isomerase at high res~lution,~ it seems unlikely that the substrate could entrap 3 water molecules at the active site though this point will be completely clarified when the crystal structure of the isomerase-dihydroxyacetonephosphate complex has been solved at high resolution.In summary it is clear that a process not involving proton transfer from -COOL is the rate determining step for the overall stepwise proton exchange reaction. Analogous situations have been proposed for amines l3 and for a number of carbon acids,14 but this is the first analysis of the nature of the ionisation of a carboxylic acid in aqueous solution. S. V. Rieder and I. A. Rose J. Biol. Chem. 1959 234 1007. * S. de la Mare A. F. W. Coulson J. R. Knowles J. D. Priddle and R. E. Offord Biochem. J. 1972 129 321 ; S. G. Waley J. C. Miller I. A. Rose and E. L. O’Connell Nature 1970,227 181; F. C. Hartman Biochem. 1971 10 146. D. W. Banner A. C. Bloomer G. A. Petsko D. C. Phillips C. I. Pogson and I. A. Wilson Nature 1975 255 609. B. Plaut and J.R. Knowles Biochem. J. 1972 129 31 1. J. M. Herlihy W. J. Albery and J. R. Knowles to be published. W. J. Albery and J. R. Knowles to be published. For a preliminary description of some of the isotope experiments see J. R. Knowles P. F. Leadlay and S.G. Maister CoZd Spring Harbor Symposia on Quantitatice Biology 1971 36 157. ’C. G. Swain E. C. Strivers J. F. Reuwer Jr. and L. J. Schaad J. Amer. Chem. SOC.,1958 80 5885. * S. J. Putman A. F. W. Coulson I. R. T. Farley B. Riddleston and J. R. Knowles Biochem. J. 1972 129 301. T. Curstedt Eur. J. Biochem. 1974 49 355. lo S. G. Maister C. P. Pett W. J. Albery and J. R. Knowles to be published. Z. Luz and S. Meiboom J. Amer. Chem. SOC.,1963,85,3923. V. Gold Trans. Faraday Soc. 1968 64,2770; A. J. Kresge Pure Appl. Chem. 1964 8 243. l3 E. Grunwald and E. K. Ralph Acc. Chem. Res. 1971 4 107. l4 e.g. D. J. Cram D. A. Scott and W. D. Nielsen J. Amer. Chem. SOC.,1961,83 3696.