The composition of milk xanthine oxidase has been reinvestigated. When the enzyme is prepared by methods that include a selective denaturation step in the presence of sodium salicylate the product is obtained very conveniently and in high yield, and is homogeneous in the ultracentrifuge and in recycling gel filtration. It has specific activity higher than previously reported preparations of the enzyme and its composition approximates closely to 2mol of FAD, 2g-atoms of Mo and 8g-atoms of Fe/mol of protein (molecular weight about 275000). In contrast, when purely conventional preparative methods are used the product is also homogeneous by the above criteria but has a lower specific activity and is generally comparable to the crystallized enzyme described previously. Such samples also contain 2mol of FAD/mol of protein but they have lower contents of Mo (e.g. 1.2g-atom/mol). Amino acid compositions for the two types of preparation are indistinguishable. These results confirm the previous conclusion that conventional methods give mixtures of xanthine oxidase with an inactive modification of the enzyme now termed ;de-molybdo-xanthine oxidase', and show that salicylate can selectively denature the latter. The origin of de-molybdo-xanthine oxidase was investigated. FAD/Mo ratios show that it is present not only in enzyme purified by conventional methods but also in ;milk microsomes' (Bailie & Morton, 1958) and in enzyme samples prepared without proteolytic digestion. We conclude that it is secreted by cows together with the active enzyme and we discuss its occurrence in the preparations of other workers. Studies on the milks of individual cows show that nutritional rather than genetic factors determine the relative amounts of xanthine oxidase and de-molybdo-xanthine oxidase. A second inactive modification of the enzyme, now termed ;inactivated xanthine oxidase', causes variability in activity relative to E(450) or to Mo content and formation of it decreases these ratios during storage of enzyme samples including samples free from demolybdo-xanthine oxidase. We conclude that even the best purified xanthine oxidase samples described here and by other workers are contaminated by significant amounts of the inactivated form. This may complicate the interpretation of changes in the enzyme taking place during the slow phase of reduction by substrates. Attempts to remove iron from the enzyme by published methods were not successful.

Earlier studies on inactivation of xanthine oxidase induced by methanol and accompanied by development of a specific electron paramagnetic resonance signal (the Inhibited signal) have been extended. The same reaction takes place when enzyme is treated with formaldehyde. With this, time-courses for signal development and disappearance of enzymic activity cannot be distinguished. Under a variety of conditions, disappearance of the signal accompanied by partial restoration of activity could be achieved. Side reactions also take place in the presence of formaldehyde. Electron paramagnetic resonance parameters of the Inhibited signal are reported. I n the signal-giving species a single non-exchangeable proton interacts with Mo(V) in the enzyme active site. Experiments with C2H30H and with 2HC2H0 show that this proton is derived from the inactivating agent. It is proposed that a formyl residue, -CHO, becomes attached to a group in the active centre and that interaction of molybdenum with this residue stabilizes the five-valent state of the metal, thus preventing enzyme turnover. Methanol is presumed to be oxidized to formaldehyde a t the active centre before taking part in the reaction.Studies on the interaction of enzymes with specific irreversible inhibitors may yield information important in understanding their catalytic mechanisms. Specific inhibition of xanthine oxidase by methanol seems first to have been noted by Polonovski [l]. More detailed studies were carried out by Handler and co-workers (e.g. [2,3]), who found that inhibition develops progressively and only while the enzyme is turning over in the presence of substrates plus methanol. They proposed [4] that this somewhat unusual inhibition involves a molecule of methanol substituting for one of water in the coordination sphere of molybdenum in the active site of the enzyme. They further reported [4] tion [6]) that the doublets were replaced by single lines, establishing the interacting proton t o be derived from the methyl group of methanol. We now report further studies carried out with the aim of understanding the reaction between milk xanthine oxidase and methanol.

No abstract

Active milk xanthine oxidase has 2 moles of FAD, 2g.atoms of Mo and 8g.atoms of Fe/mole of protein of mol.wt. 275000 (Hart, McGartoll, Chapman & Bray, 1969). There has been a tendency to assume (cf. Massey, Brumby, Komai & Palmer, 1969) that all of these constituents would not be likely to function as a single catalytic unit in the molecule and that there must be two presumably equivalent active centres (Bray, 1963). We now present evidence based on FAD loss (Bray, Chisholm, Hart, Meriwether & Watts, 1966) during inactivation of the reduced enzyme by iodoacetamide (Bray & Watts, 1965) indicating that xanthine oxidase operates, during aerobic xanthine oxidation at least, with a single active centre.The interpretation of earlier quantitative studies on reaction ofxanthine oxidase with iodoacetamide, by measuring iodide liberation (Bray & Watts, 1965) and following FAD loss (A. J. Chisholm, A. M. Scott & R. C. Bray, unpublished work), was complicated by the use as starting material of preparations with non-stoicheiometric FAD/Mo ratios. These problems were overcome for the present work by the use of the salicylate method for preparing the enzyme, this giving samples free from demolybdoxanthine oxidase (Hart et al. 1969). The enzyme used was also free from detectable impurities in the ultracentrifuge and in recycling gel filtration on Sephadex G-200. Techniques and analytical methods were as described by Hart et al. (1969). Reaction of enzyme, reduced with xanthine, with iodo[14C]acetamide was carried out at pH5-5 to minimize side reactions (Bray & Watts, 1965) and was terminated by the addition of cysteine. The product from the reaction was dialysed, then purified by recycling gel filtration on Sephadex G-200. After three cycles it showed a single symmetrical peak, radioactivity (detected with a flow scintillation counter from Packard Instruments Ltd., Wembley, Middx.), residual xanthine oxidase