The CuZn superoxide dismutases (SODs) from ox, sheep, pig and yeast were investigated by pulse radiolysis in order to evaluate the role of electrostatic interactions between O2.- and SOD proteins in the mechanism of action of the SOD enzymes. The protein net charge in this series varies, as evaluated by the protein pI values spanning over a large range of pH: 8.0 (sheep), 6.5 (pig), 5.2 (ox) and 4.6 (yeast). The amino acid sequences are largely conserved, with the three mammalian proteins being highly homologous and the yeast protein having some distinct variations in the region surrounding the active site. At pH 8.0 the activities of the SODs from various sources are similar, though the minor differences observed suggest that in the highly homologous mammalian series the most acidic protein is the most enzymically efficient one. The pH-dependences of the various activities in the pH range 7-12 are similar, and the related curves are best fitted by two pK values, which are approx. 9.2 and 11.0 for the mammalian enzymes and 9.1 and 11.4 for the yeast enzyme. The activities of the proteins at I 0.1 are decreased by approx. 20% when compared with the activity at I 0.02 at pH 8.5, whereas at pH above 10 the pH-dependence of the activity approaches that determined at I 0.02 and at pH 11.9 the activity is essentially independent of ionic strength. The dependence upon ionic strength also depends on the salt used, with perchlorate being more effective than phosphate or borate or Mops and still effective at pH above 10.5, where the effect of other salts becomes negligible. The dual and concerted dependence of the activities of different SODs on pH and salt concentration is explained with the encounter of O2.- with the active-site copper being governed by the protonation of two positively charged groups in the vicinity of the active site. The gradient between these localized charges and the rest of the protein may explain the different activities of the mammalian proteins at lower pH. On the basis of the sequence variation of the SODs examined it is not possible to definitely identify these groups. Likely candidates are conserved basic amino acid side chains in the vicinity (less than or equal to 1.2 nm) of the active site, i.e. Lys-134 and Arg-141, but co-ordination of OH- in the first copper co-ordination sphere may be an additional factor accounting for the higher pK.(ABSTRACT TRUNCATED AT 400 WORDS)

The interaction of one-electron reduced metronidazole (ArNO2.-) with native and Type-2-copper-depleted ascorbate oxidase were studied in buffered aqueous solution at pH 6.0 and 7.4 by using the technique of pulse radiolysis. With ArNO2.-, reduction of Type 1 copper of the native enzyme and of the Type-2-copper-depleted ascorbate oxidase occurs via a bimolecular step and at the same rate. Whereas the native protein accepts, in the absence of O2, 6-7 reducing equivalents, Type-2-copper-depleted ascorbate oxidase accepts only 3 reducing equivalents with stoichiometric reduction of Type 1 copper. On reaction of O2.- with ascorbate oxidase under conditions of [O2.-] much greater than [ascorbate oxidase], removal of Type 2 copper results in reduction of all the Type 1 copper atoms, in contrast with reduction of the equivalent of only one Type 1 copper atom in the holoprotein. From observations at 610 nm, the rate of reduction of ascorbate oxidase by O2.- is not dependent on the presence of Type 2 copper. For the holoprotein, no significant optical-absorption changes were observed at 330 nm. It is proposed that electrons enter the protein via Type 1 copper in a rate-determining step followed by a fast intramolecular transfer of electrons within the protein. For the Type-2-copper-depleted protein, intramolecular transfer within the protein, however, is slow or does not occur. In the presence of O2, it is also suggested that re-oxidation of the partially reduced holoprotein occurs at steady state, as inferred from the observations at 330 nm and 610 nm. The role of Type 2 copper in ascorbate oxidase is discussed in terms of its involvement in redistribution of electrons within the protein or structural considerations.

The interactions of one-electron reduced metronidazole (ArNO2.-) and O2.- with native and Type-2-copper-depleted Vietnamese- and Japanese-lacquer-tree laccases were studied in aqueous solution at pH 6.0 and 7.4 by using the technique of pulse radiolysis. On reaction with ArNO2.-, in the absence of O2, the holo- and the Type-2-copper-depleted proteins accept, with reduction of Type 1 copper, 2 and 1 reducing equivalents respectively. On reaction with O2.- of both holo- and Type-2-copper-depleted Vietnamese-lacquer-tree laccase, almost complete reduction of Type 1 copper was observed and, after completion of the reaction, some (less than 20%) reoxidation of Type 1 copper occurs. Reduction of Type 1 copper of the laccases by these one-electron donors occurs via a bimolecular step; however, the rate of reduction of Vietnamese-lacquer-tree laccase is over 10 times that of Japanese-lacquer-tree laccase. It is inferred that electrons enter the protein via Type 1 copper with, in the case of the holoprotein, subsequent rapid intramolecular transfer of 1 reducing equivalent within the protein. Furthermore it is suggested that intra-molecular electron transfer to Type 3 copper atoms is slow and, in the case of Type-2-copper-depleted protein, may not occur. This slow process may partially account for the variation of the catalytic activities of ‘blue’ oxidases.

The interaction of e-aq., CO2-. and one-electron reduced nitroaromatics (RNO2-.) with ascorbate oxidase (AAO) was studied in aqueous solution at pH 6.0 and 7.5 by using the technique of pulse radiolysis. From observations at 330, 410 and 610 nm, interaction of e-aq. and CO2-. with AAO results in non-specific reduction of the protein followed by reduction of Type 1 Cu in a rate-determining intramolecular step. Only a few per cent of the reducing equivalents ultimately results in reduction of Type 1 Cu. With large excesses of reducing equivalents (e-aq. and CO2-.) with respect to the copper concentration, the amount of Type 1 copper reduced never exceeds 50% of the total amount of Type 1 copper after a single radiation pulse. With less-powerful reducing agents, e.g. RNO2-. reduction of Type 1 Cu occurs via a bimolecular step, and there is no evidence for formation of radicals on protein residues. From observations at 330 nm it is evident that Type 2 and/or Type 3 Cu may also be reduced along with Type 1 Cu. Almost stoichiometric reduction of AAO by RNO2-. was observed, e.g. the protein accepts 6-7 reducing equivalents. It is inferred that the various types of redox couples Cu2+/Cu+ are in equilibrium and that intramolecular electron transfer between the different types of Cu is not rate-determining when using RNO2-. as reducing agent.

By using the technique of pulse radiolysis to generate O2-., it is demonstrated that Co(II) derivatives of bovine superoxide dismutase in which the copper alone and both the copper and zinc of the enzyme have been substituted by Co(II), resulting in (Co, Zn)- and (Co, Co)-proteins, are capable of catalytically dismutating O2-. with ‘turnover’ rate constants of 4.8×10(6) dm3.s-1.mol-1 and 3.1×10(6) dm3.s-1.mol-1 respectively. The activities of the proteins are independent of the pH (7.4-9.4) and are about three orders of magnitude less than that of the native (Cu, Zn)-protein. The rate constants for the initial interaction of O2-. with the Co-proteins were determined to be (1.5-1.6) X 10(9) dm3.s-1.mol-1; however, in the presence of phosphate, partial inhibition is apparent [k approximately (1.9-2.3) X 10(8) dm3.s-1.mol-1]. To account for the experimental observations, two reaction schemes are presented, involving initially either complex-formation or redox reactions between O2-. and Co(II). This is the first demonstration that substitution of a metal into the vacant copper site of (Cu, Zn)-protein results in proteins that retain superoxide dismutase activity.

In the preceding paper the mechanism of catalysis of the manganese-containing superoxide dismutase from Bacillus stearothermophilus was shown to involve a ‘fast cycle’ and a ‘slow cycle’ [McAdam, Fox, Lavelle & Fielden, 1977 (Biochem. J. 165, 71-79)]. Further properties of the enzyme was considered in the present paper. Pulse-radiolysis studies, under conditions of low substrate concentration to (i.e. when the fast cycle predominates), showed that enzyme activity decreases as pH increases (6.5-10.2). Activity was unaffected by the addition of H2O2 or NaN3 but slightly decreased by KCN. Both H2O2 and the reducing radical anion CO2– caused a decrease in A480 of the native enzyme. The rate of the fast catalytic cycle was independent of temperature (5-55 degrees C), and as temperature increases the slow cycle becomes relatively more important. Arrhenius parameters of the rate contants were estimated. The possible identity of the various forms of the enzyme is considered.

The enzymic reaction mechanism of a manganese-containing superoxide dismutase from Bacillus stearothermophilus was studied by using pulse radiolysis. During catalysis (pH 8.9; 25 degrees C), changes occurring in the kinetics of substrate disappearance and in the visible absorption of the enzyme at 480 nm established that the simple two-step mechanism found for copper- and iron-containing superoxide dismutases is not involved. At a low ratio (less than 15) of substrate concentration to enzyme concentration the decay of O2–is close to exponetial, whereas at much higher ratios (greater than 100) the observed decay is predominantly zero-order. The simplest interpretation of the results invokes a rapid one-electron oxidation-reduction cycle (‘the fast cycle’) and, concurrently, a slower reaction giving a form of the enzyme that is essentially unreactive towards O2– but which undergoes a first-order decay to yield fully active native enzyme (‘the slow cycle’). The fast cycle involves the native enzyme EA and a form of the enzyme EB which can be obtained also by treating the form EA with H2O2. Computer calculations made with such a simple model predict behaviour in excellent agreement with the observed results.

The mechanism of the enzymic reaction of an iron-containing superoxide dismutase purified from the marine bacterium Photobacterium leiognathi was studied by using pulse radiolysis. Measurements of activity were done with two different preparations of enzyme containing either 1.6 or 1.15 g-atom of iron/mol. In both cases, identical values of the second-order rate constant for reaction between superoxide dismutase and the superoxide ion in the pH range 6.2-9.0 (k=5.5 X 10(8) M-1-S-1 at pH 8.0) were found. As with the bovine erythrocuprein, there was no evidence for substrate saturation. The effects of reducing agents (H2O2, sodium ascorbate or CO2 radicals) on the visible and the electron-paramagnetic-resonance spectra of the superoxide dismutase containing 1.6 g-atom of ferric iron/mol indicate that this enzyme contains two different types of iron. Turnover experiments demonstrate that only that fraction of the ferric iron that is reduced by H2O2 is involved in the catalysis, being alternately oxidized and reduced by O2; both the oxidation and the reduction steps have a rate constant equal to that measured under turnover conditions. These results are interpreted by assuming that the superoxide dismutase isolated from the organism contains 1 g-atom of catalytic iron/mol and a variable amount of non-catalytic iron. This interpretation is discused in relation to the stoicheiometry reported for iron-containing superoxide dismutases prepared from several other organisms.