Aiochem. J. (1974) 139, 797-800 797 Printed in Great Britain Histidine Residues and the Enzyme Activity of Pig Heart Supernatant Malate Dehydrogenase By J. JOHN HOLBROOK, A. LODOLA and NICHOLAS P. ILLSLEY Department ofBiochemistry, University ofBristol, Bristol BS8 1TD, U.K. (Received 25 February 1974) 1. Supernatant pig heart malate dehydrogenase is completely inhibited by reaction with diethyl pyrocarbonate at pH6.5, when 0.58+0.1 residue of ethoxycarbonylhistidine is formed per NADH-binding site. 2. Oxaloacetate and hydroxymalonate protect the en- zyme from inhibition in the absence of coenzyme. 3. Limited ethoxycarbonylation does not alter the binding of NADH to the enzyme but prevents the enzyme-NADH complex from interacting with hydroxymalonate in a ternary complex. Supematant malate dehydrogenase and lactate de- and the crystals were dissolved in 20mM-H3P04 hydrogenase have very similar tertiary structures adjusted to pH6.5 with 5M-NaOH and then filtered (Hill et al., 1972; Adams et al., 1972). Histidine-195 through a column of Sephadex G-50 equilibrated in plays a central role in the activity of lactate dehydro- the same buffer. If (NH4)2SO4 was not completely genase as the acceptor and donor of the proton in- removed it competed effectively for the diethyl pyro- volved in the reaction (Holbrook & Gutfreund, 1973). carbonate. This latter reagent was obtained from A proton is also a reactant for malate dehydrogenase, Sigma (London) Chemical Co., Kingston-on- and the work reported here is an attempt to discover Thames, U.K., and the batch used in these experi- if this enzyme also has a histidine residue with the ments was 2.6M when assayed as described by same function as histidine-195 of lactate dehydro- Holbrook & Ingram (1973). genase. Anderton & Rabin (1970) found that a histi- Preliminary experiments established that enzyme dine residue was involved in the activity of the mito- was most rapidly inhibited in a buffer prepared by chondrial malate dehydrogenase, although it is not adjusting a solution of 20mM-H3PO4 with 5M-NaOH known ifthis form ofmalate dehydrogenase is a struc- to pH6.5. This buffer was used for all theexperiments tural homologue of lactate dehydrogenase. reported in this paper. The rate of inhibition was half The method ofGerding & Wolfe (1969) was used to itsmaximumvalue at pH6 and 7.5. The enzyme could prepare the supernatant enzyme from pig heart. The be protected from inhibition by diethyl pyrocarbon- negative absorptions (steps 4-7) were not necessary. ate if substrates and coenzymes were present in the In our hands the hydroxyapatite column chromato- incubation mixture. For example, the rate of loss of graphy did not reproducibly give a purification and enzyme activity (kobl. = 2min-'; samples assayed was omitted. Enzyme activity was expressed as units every i5s) when 7p/M-enzyme (calculated from a rela- of umol of NADH produced/min in an incubation tive subunit mass of 35000) was incubated with mixture containing at 25°C 5mM-NAD+, O.1M-L- 0.78mM-inhibitor was decreased 10-fold by the inclu- malate and 0.1M-glycine adjusted to pH 10 with SM- sion of either 3mM-oxaloacetate or 25,UM-NADH in NaOH. Protein was determined from absorbance the incubation mixture. In another experiment with measurements at 280nm by using E1',1 = 9.3 given by 284uM-enzyme and 0.5mM-inhibitor the rate of inhibi- Gerding & Wolfe (1969). The specific enzyme activity tion was decreased 10-fold by 3.3mM-hydroxy- of our preparations was 77 units/mg; this is only 70% malonate or 0.4mM-NADH and no inhibition was of the 108 units/mg reported by Gerding & Wolfe detected over 20min if the hydroxymalonate and (1969) for the pure enzyme. The presence of impurity NADH were both present. Oxaloacetate did not re- could also be deduced from a determination of the act with the inhibitor. There is as yet no definitive NADH-binding capacity of the preparation by using steady-state kinetic investigation of the mechanism of the supernatant enzyme. The form of the initial the continuous micro-fluorimetric titration as re- velocity equations is consistent with a sequential Bi Bi ported by Holbrook & Wolfe (1972). Only 1.4mol of mechanism (A. Lodola, unpublished work), but it is NADH was bound per mol ofprotein, assuming that not known whether there is a compulsory, preferred the protein molecule consists of two subunits of or random order of addition of substrate and co- mol.wt. 35000. Before the experiments reported here enzyme. The above results suggest that a preferred or the crystalline enzyme suspension was centrifuged random order should be considered, although the Vol. 139 798 J. J. HOLBROOK, A. LODOLA AND N. P. ILLSLEY protection by oxaloacetate in the absence of NADH measured against that ofan uninhibited control. With does not necessarily mean that this substrate binds at greater degrees of inhibition a negative absorbance the active site. peak at 280am appeared (ethoxycarbonyltyrosine; Inhibition after reaction with diethyl pyrocarbon- Muhlrad et al., 1967). The rate of appearance of ate is due to the modification of histidine residues. ethoxycarbonylhistidine was approx. 40M1-s-1 at Although the inhibitor will react with other side pH6.5, and this is smilar to the rate of reaction of chains it was possible to reactivate the inhibited free imidazoles at the same pH (Holbrook & Ingram, enzyme with hydroxylamine. For example 4uM- 1973). The rate of reaction was not quite fast enough enzyme was 98% inhibited after incubation with the for it to be possible to add 1 equivalent of inhibitor inhibitor (2.6mM) for 10min and then diluted into to 1 equivalent of enzyme and observe 100% 0.7M-hydroxylamine hydrochloride, which had been inhibition with the formation of 1 equivalent of adjusted to pH6.5 with SM-NaOH. When assayed ethoxycarbonylhistidine. The closest that this ideal after 10min the enzyme activity was 100±5% of that has been approached is in experiments similar to that of an untreated control. Hydroxylamine regenerates shown in Fig. 1. In that case 86gM-inhibitor reacted histidine fromethoxycarbonylhistidine, but iswithout with 72,gM-active enzyme to yield 36gM-ethoxycarb- effect on the ethoxycarbonyl derivatives of tyrosine onylhistidine and 68% inhibition after 18min. Thus and lysine (Melchior & Fahrney, 1970). Cysteine there is sufficient inhibitor unaccounted for to allow residues are not essential for the activity of this the reaction of up to 0.5mol of another amino acid enzyme (Devenyi etal., 1966). When inhibited enzyme side chain/mol of NADH-binding sites. Such a was freed from excess ofinhibitor byfiltration through calculation is the worst case and does not take into a column of Sephadex G-50 the enzyme activity of account the loss of inhibitor by hydrolysis during the the protein peak slowly reappeared. The initial rate of 18min incubation. regain of activity was approx. 2%/h. This is much The results described in detail in the legend to Fig. 1 slower than with lactate dehydrogenase or glutamate show that a sample of enzyme (3.6mg/ml) containing dehydrogenase (Holbrook & Ingram, 1973; Wallis & 72gM-NADHrbinding sites is completely inhibited Holbrook, 1973). by the reaction of0.58+±0.1 histidine residue/NADH- The fonnation of ethoxycarbonylhistidine in an binding site. It seems unlikely that this unusual ratio enzyme can be monitored from the absorbance peak arises from the use of only 70 %-active enzyme, since at 240nm. In samples of enzyme that had not been the results are referred to NADH-binding sites. The more than 90% inhibited this was the only absorbance reactivation of the enzyme in hydroxylamine indi- peak visible in a spectrum of the inhibited enzyme cates that it is the histidine modification that causes I100 0 Time (min) [Eth.oxycarbonylhistidinelI(#M) Fig. 1. Inkibition ofsupernatant malate dehydrogenase by ethoxycarbonylation ofhitidine residues in the enzyme Fig. 1(a) is a record ofthe increase in absorbance at 250nm in a solution containing enzyme (3.6mg/ml, 70% active, 72pM in sites that can bind NADH) monitored on a double-bearn spectrophotometer with a recorder output sensitivity of 0.15 absorbance unit full scale during reaction with 86pM-diethyl pyrocarbonatp at pH6.5. After 19min the inhibitor concentra- tion was increased to 133 gM. During this and similar reactions samples of the incubation mixture were assayed for enzyme activity, and Fig. l(b) shows the relation between the residual enzyme activity and the concentration of ethoxycarbonyl- histidine formed on the enzyme: 1 mm-ethoxycarbonylhistidine has an absorbance of 1.9 in a 1cm light-path at 250nm(Hol- brook & Ingran, 1973). The two broken lines enclose the results and indicate that t-he enzyme is inhibited by formation of 0.58±O.Imol -of ethoxycarbonylhistidine/mol of N-AD-binding sites. 1974 SHORT COMMUNICATIONS 799 (a) (b) - 6 5o48 4U4c a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0~~~~~~~~~~~~~~~ 0~~~~~~~~~~~~~~~~~ C) ~ [AH A)Ihbtonb ity yoabnt (%) a)~~~~~~~~~~~~~~~~ 0 a) AD- (m Ihbtonb ithlprcabnt Fig. 2. Nucleotide bindingbyethoxycarbonylatedmalatedehydrogenase Samples of enzyme (l65,UM in NADH-binding sites) were incubated with concentrations of diethyl pyrocarbonate between 130,M and 2.6mM for 5min. At exactly the same moment one portion ofthe incubation mixture was diluted 1:100 (v/v) and immediately assayed, while another portion was diluted 35-fold into lOmM-hydroxymalonate in 20mM-H3PO4 (which had been adjusted to pH6.5 with 5M-NaOH) and was titrated micro-fluorimetrically with NADH. The traces in Fig. 2(a) show the increases in NADH fluorescence (excited at 320nm, emission at about 430nm via a Kodak Wratten no. 98 filter) after subtraction of the fluorescence of NADH added to a solution without enzyme. The apparatus was described by Holbrook (1972) and the application to malate dehydrogenase by Holbrook & Wolfe (1972).