Human xanthine oxidase was purified from breast milk. The dehydrogenase form of the enzyme, which predominates in most mammalian tissues, catalyses the oxidation of NADH by oxygen, generating superoxide anion significantly faster than does the oxidase form. The corresponding forms of bovine enzyme behave very similarly. The steady-state kinetics of NADH oxidation and superoxide production, including inhibition by NAD, by the dehydrogenase forms of both enzymes, are analysed in terms of a model involving two-stage recycling of oxidised enzyme.Established inhibitors of xanthine oxidoreductases (allopurinol oxypurinol, amflutizole and BOF 4272), which block all other reducing substrates, were ineffective in the case of NADH. Diphenyleneiodonium, on the other hand, was a powerful inhibitor of NADH oxidation.The potential involvement of reactive oxygen species arising from NADH oxidation by xanthine oxidoreductase in ischaemia-reperfusion injury and other disease states, as well as in normal signal transduction, is discusssed.

Resonance Raman spectra are reported for recombinant horseradish peroxidase C (HRP-C*) and three protein variants prepared by in vitro refolding after Escherichia coli expression. The spectra of their FeII and FeIII forms and of their complexes with benzohydroxamic acid (BHA) were recorded at neutral pH. The residues mutated were on the distal [Phe41-->Trp or Val (F41W, F41V) and Arg38-->Lys (R38K)] side of the heme. The spectra give information on the spin and ligation states via the frequencies of the core size marker bands. No detectable modification in the enzyme structure or in the heme group has been observed in the wild-type recombinant HRP-C*. The FeIII forms of both the recombinant and the plant proteins show the coexistence of a 5-(5-cHS) and a 6-coordinate high-spin (6-cHS) heme, characterized by the anomalous frequencies of certain bands, namely, v3 and v10, which we attribute to a different degree of distortion of the heme planarity with respect to other heme proteins and model compounds, resulting from external forces such as steric contacts within the protein. This effect is partially relieved upon complexation with BHA or as a result of mutation. F41W and F41V are characterized by an increase in a 6-cHS form at the expense of the 5-cHS species, and the R38K by an increase in both the 6-c high-(HS) and low-spin (LS) hemes. The 6-cHS and -LS species are characterized by normal core size marker band frequencies. The FeII-His RR band is at 243 cm-1 in HRP-C*, the high frequency value being due to hydrogen-bonding interactions between the proximal His170 N delta and the carboxylate acceptor group on Asp247. Mutation at position 38 causes a downshift of 3 cm-1 in the v(Fe-Im) stretching mode, suggesting a weakening of the Fe-Im bond strength. By comparing the results obtained with HRP-C* mutants with those previously reported for the corresponding cytochrome c peroxidase (CCP) mutants, it appears that the distal heme pocket architecture is significantly different in the two peroxidases, although the hydrogen-bonding network coupling the distal and the proximal sides of the heme appears to be conserved. Mutations on the distal side dramatically affect the capability of the protein to bind BHA. F41W and R38K mutants do not bind the substrate, whereas the F41V variant shows a 2-fold increase in affinity.(ABSTRACT TRUNCATED AT 400 WORDS)

A horseradish peroxidase variant ([F41 V] HRP-C*), in which Val replaces the conserved Phe at position 41 adjacent to the distal His, has been constructed. Its composition and spectroscopic, catalytic and substrate-binding properties were compared with those of the wild-type recombinant (HRP-C*) and plant (HRP-C) enzymes. Presteady-state kinetic measurements of the rate constant for compound 1 formation ( k , ) revealed an eightfold decrease in the reactivity of the Phe41+Val variant towards H202, in comparison with HRP-C or HRP-C*. Measurement of the remaining rate constants, k2 and k3, for the two single-electron reduction reactions of [F41V] HRP-C with paraaminobenzoic acid as reducing substrate, showed that they were 2.5-fold and 1.3-fold faster, respectively. In contrast, analysis of data from steady-state assays with 2,2'-azinobis(3-ethylbenzthiazoline-6-sulphonate) as reducing substrate, showed decreased reactivity of the mutant enzyme to this compound, indicating a change in substrate specificity. Over the substrate range studied, the data for HRP-C* and for [F41V] HRP-C conformed to a simple modification of the accepted peroxidase mechanism in which a first-order step (k,), assumed to be product dissociation, becomes rate-limiting under our standard assay conditions. Calculations of rate constants from steady-state data yielded values of kl for both enzyme forms in adequate agreement with those from pre-steady state measurements. They showed, furthermore, that both k3 for 2,2'-azinobis(3-ethyIbenzthiazoline-6-sulphonate) and k , were substantially decreased, fivefold and tenfold, respectively, in the mutant. Analogous to the decrease ink,, we observed a twofold increase in the affinity of the mutant variant for the inhibitor benzhydroxamic acid. The coordination-state equilibrium of the haem iron also appeared shifted towards the hexacoordinate high-spin form. These observations indicate that in addition to affecting reactivity to H20z, mutations in the distal region and close to the haem iron also affect reactivity towards different reducing substrates, inducing perturbations in the neighbourhood of the aromaticsubstrate-binding site, known to be 0.8 -1.2 nm from the haem iron.

In all sequenced genomes, a large fraction of predicted genes encodes proteins of unknown biochemical function and up to 15% of the genes with "known" function are mis-annotated. Several global approaches are routinely employed to predict function, including sophisticated sequence analysis, gene expression, protein interaction, and protein structure. In the first coupling of genomics and enzymology, Phizicky and colleagues undertook a screen for specific enzymes using large pools of partially purified proteins and specific enzymatic assays. Here we present an overview of the further developments of this approach, which involve the use of general enzymatic assays to screen individually purified proteins for enzymatic activity. The assays have relaxed substrate specificity and are designed to identify the subclass or sub-subclasses of enzymes (phosphatase, phosphodiesterase/nuclease, protease, esterase, dehydrogenase, and oxidase) to which the unknown protein belongs. Further biochemical characterization of proteins can be facilitated by the application of secondary screens with natural substrates (substrate profiling). We demonstrate here the feasibility and merits of this approach for hydrolases and oxidoreductases, two very broad and important classes of enzymes. Application of general enzymatic screens and substrate profiling can greatly speed up the identification of biochemical function of unknown proteins and the experimental verification of functional predictions produced by other functional genomics approaches.