Studies by electron paramagnetic resonance spectroscopy of xanthine oxidase enriched with molybdenum-95 and with molybdenum-97

Investigations have been carried out on the nature of the species from the enzyme xanthine oxidase that give rise to two molybdenum (V) electron paramagnetic resonance (EPR) signals. Isotopic enrichment with 95Mo, 97Mo, 33S, and 17O was employed. Computer simulations of the EPR spectra recorded at 9- and 35-GHz microwave frequencies were used to evaluate the various hyperfine couplings and angular relations between the principal axes of g and A, as well as the nuclear electric quadrupole interaction for 97Mo. The results support the presence of an oxo ligand in the Rapid and of both an oxo and a sulfido ligand in the Very Rapid signal-giving species.     

Studies by electron-paramagnetic-resonance spectroscopy on the mechanism of action of xanthine dehydrogenase from Veillonella alcalescens.

E.p.r- (electron-paramagnetic-resonance) spectroscopy was used to compare chemical environment and reactivity of molybdenum, flavin and iron-sulphur centres in the enzyme xanthine dehydrogenase from Veillonella alcalescens (Micrococcus lactilyticus) with those of the corresponding centres in milk xanthine oxidase. The dehydrogenase is frequently contaminated with small but variable amounts of a species resistant to oxidation and giving a new molybdenum (V) e.p.r. signal, "Resting I". There is also a "desulpho" form of the enzyme giving a Slow Mo(V) signal, indistinguishable from that of the milk enzyme. Molybdenum of the active enzyme behaves in a manner analogous to that of the milk enzyme, giving a Rapid Mo(V) signal on partial reduction with substrates or dithionite. Detailed comparison shows that molybdenum in each enzyme must have the same ligand atoms arranged in the same manner. As with the milk enzyme, complex-formation between reduced dehydrogenase and purine substrate molecules, presumably interacting at the normal substrate-binding site, modifies the Rapid signal, confirming that such substrates interact near molybdenum. The dehydrogenase-flavin semiquinone signal is identical with that of the oxidase but, in contrast, there is only one iron-sulphur signal. The latter gives an e.p.r. spectrum similar to that of aldehyde oxidase.     

Coupled electron/proton transfer in complex flavoproteins: solvent kinetic isotope effect studies of electron transfer in xanthine oxidase and trimethylamine dehydrogenase

A solvent kinetic isotope effect study of electron transfer in two complex flavoproteins, xanthine oxidase and trimethylamine dehydrogenase, has been undertaken. With xanthine oxidase, electron transfer from the molybdenum center to the proximal iron-sulfur center of the enzyme occurs with a modest solvent kinetic isotope effect of 2.2, indicating that electron transfer out of the molybdenum center is at least partially coupled to deprotonation of the Mo(V) donor. A Marcus-type analysis yields a decay factor, beta, of 1.4 A(-1), indicating that, although the pyranopterin cofactor of the molybdenum center forms a nearly contiguous covalent bridge from the molybdenum atom to the proximal iron-sulfur center of the enzyme, it affords no exceptionally effective mode of electron transfer between the two centers. For trimethylamine dehydrogenase, rates of electron equilibration between the flavin and iron-sulfur center of the one-electron reduced enzyme have been determined, complementing previous studies of electron transfer in the two-electron reduced form. The results indicate a substantial solvent kinetic isotope effect of 10 +/- 4, consistent with a model for electron transfer that involves discrete protonation/deprotonation and electron transfer steps. This contrasts to the behavior seen with xanthine oxidase, and the basis for this difference is discussed in the context of the structures for the two proteins and the ionization properties of their flavin sites. With xanthine oxidase, a rationale is presented as to why it is desirable in certain cases that the physical layout of redox-active sites not be uniformly increasing in reduction potential in the direction of physiological electron transfer.     

The molybdenum centre of native xanthine oxidase. Evidence for proton transfer from substrates to the centre and for existence of an anion-binding site.

The observation by Bray & Knowles [Proc. R. Soc. London Ser. A (1968) 302, 351--353] of direct transfer, during the catalytic reaction, of hydrogen atoms from substrate molecules to the enzyme xanthine oxidase was reinvestigated. The experimental phenomenon and its basic interpretation were confirmed and extended. In the reduced functional enzyme, molybdenum(V) interacts with two enzyme-bound protons, which are exchangeable with solvent protons. One of these is coupled to the metal with AHav. 1.4mT and the other with AHav. 0.3mT. The molecule also contains a site for the binding of anions, presumably as ligands of molybdenum. This is shown by effects of nitrate ions on the e.p.r. spectra. The spectra of the nitrate and 1-methylxanthine complexes of the reduced enzyme are very similar to one another, and are designated Rapid type-1 spectra. It is concluded that, in the Michaelis complex, the substrate molecule occupies the anion site, probably being bound to molybdenum via the nitrogen in its 9-position. During the turnover process, hydrogen from the substrate C-8 position, after transfer to the enzyme, appears as the proton more strongly coupled to molybdenum. This proton then exchanges with solvent deuterium with a rate constant of 27s-1, at pH 8.2 and 12 degrees C. It has been confirmed that substrate molecules occupying the anion site do not interfere with observation of the transfer and exchange processes.     

Formamide as a substrate of xanthine oxidase.

Formamide is a substrate of xanthine oxidase. At pH 8.2 and 1.14 mM-O2, Vmax.(app.) is 3.1 s-1 and Km (app.) is 0.7 M. Mo(V) e.p.r. signals obtained by treating the enzyme with formamide were studied, and these provide new information about the ligation of molybdenum in the enzyme and about the enzymic mechanism. The substrate is the first compound that is not a nitrogen-containing heterocycle to give a Very Rapid signal. This supports the hypothesis that the Very Rapid signal, though it is not detectable with all substrates, represents an essential intermediate in turnover. Formamide also gives the Inhibited signal and is the first non-aldehyde substrate to do so. The Rapid type 1 signal obtained in the presence of formamide was examined in H2O enriched with 2H or with 17O. The single oxygen atom detectable in the signal is shown to be strongly and anisotropically coupled. This indicates that this atom remains as an oxo ligand of molybdenum in this signal-giving species. Other structural features of this species are discussed.     

Hyperfine coupling of 17O to Mo(V) in the rapid epr signal of xanthine oxidase: Evidence for an oxygen ligand of the metal

The spectrum of the Rapid Mo(V) electron paramagnetic resonance signal from xanthine oxidase dissolved in ¹⁷O-enriched water is presented. Difference technqiues have been used to eliminate the ¹⁶O contribution. Clearly observed structure in the spectrum is attributed to moderately strong hyperfine coupling of one oxygen atom to molybdenum. Though complete interpretation of the spectrum has not been attempted, one component of A(¹⁷O) is about 1.6 mT. The possibility that the oxygen is present in a MoOH group, whose proton is the strongly-coupled proton of the Rapid signal, is discussed.     

X-ray absorption spectroscopy of xanthine oxidase. The molybdenum centres of the functional and the desulpho forms.

X-ray absorption spectra have been recorded for the molybdenum K-edge region of xanthine oxidase. Both the absorption edge and the extended fine structure (e.x.a.f.s.) regions were investigated. Spectra were obtained for samples of the desulpho enzyme as well as for mixtures of this with the active enzyme. The spectrum of the pure active form was then obtained by difference. The desulpho enzyme shows a pronounced step in the absorption edge, of a type previously associated terminal oxygen ligands. In the active enzyme this step has decreased markedly. Satisfactory simulations of the e.x.a.f.s. spectrum of the desulpho enzyme could be obtained by assuming the molybdenum to be bonded to two terminal oxygen atoms (Mo = O about .175 nm), two sulphur atoms (presumably from cysteine residues, Mo-S about .0250 nm) and one sulphur atom (presumably from a methionine residue, Mo-S about 0.290 nm). E.x.a.f.s. of the active enzyme differed appreciably from this. In keeping with earlier proposals [Gutteridge, Tanner & Bray (1978) Biochem. J. 175, 887-897], the spectrum of the active enzyme could be simulated if a sulphur atom at about 0.225 nm (i.e. presumably a terminal sulphur atom) replaced one of the terminal oxygen atoms of the desulpho from, with small changes in the other bond distances. Validity of the interpretative procedures, which involved phase shift and amplitude calculations ab initio, was demonstrated by using low molecular weight compounds of known structure.     

Structure of the molybdenum site of Rhodobacter sphaeroides biotin sulfoxide reductase

Conditions for heterologous expression of Rhodobacter sphaeroides biotin sulfoxide reductase in Escherichia coli were modified, resulting in a significant improvement in the yield of recombinant enzyme and enabling structural studies of the molybdenum center. Quantitation of the guanine and the molybdenum as compared to that found in R. sphaeroides DMSO reductase demonstrated the presence of the bis(MGD)molybdenum cofactor. UV-visible absorption spectra were obtained for the oxidized, NADPH-reduced, and dithionite-reduced enzyme. EPR spectra were obtained for the Mo(V) state of the enzyme. X-ray absorption spectroscopy at the molybdenum K-edge has been used to probe the molybdenum coordination of the enzyme. The molybdenum site of the oxidized protein possesses a Mo(VI) mono-oxo site (Mo=O at 1.70 A) with additional coordination by approximately four thiolate ligands at 2.41 A and probably one oxygen or nitrogen at 1.95 A. The NADPH- and dithionite-reduced Mo(IV) forms of the enzyme are des-oxo molybdenum sites with approximately four thiolates at 2.33 A and two different Mo-O/N ligands at 2.19 and 1.94 A.     

Comparison of the molybdenum centres of native and desulpho xanthine oxidase. The nature of the cyanide-labile sulphur atom and the nature of the proton-accepting group.

The non-functional form of xanthine oxidase known as the desulpho enzyme was compared with the functional enzyme in various ways, to obtain information on the structure of the molybdenum centre and the mechanism of the catalytic reaction. The desulpho enzyme, like the functional one, possesses a site for the binding of anions, presumably as ligands of molybdenum. Evidence is presented that in the Mo(V) e.p.r. signal from the desulpho-enzyme, as in that from the functional enzyme, a weakly coupled proton, in addition to a strongly coupled proton, interacts with the metal. Measurements were carried out by e.p.r. on the rate at which the proton strongly coupled to molybdenum exchanged, on diluting enzyme samples with 2H2O. For the desulpho enzyme the exchange rate constant was 0.40s-1, at pH 8.2 and 12 degrees C, and for the functional enzyme it was 85 s-1. It is shown that the great majority of reported differences between the enzyme forms are consistent with functional enzyme containing an (Enzyme)-Mo=S grouping, replaced in the desulpho form by (Enzyme)-Mo=O. Protonation of these groups, with pK values of about 8 and 10 respectively, would give (Enzyme)-Mo-SH and (Enzyme)-Mo-OH, these being the forms observed by e.p.r. The accepting group in the functional enzyme, for the proton transferred from the substrate while molybdenum is reduced in the catalytic reaction [Gutteridge, Tanner & Bray (1978) Biochem J. 175 869-878], is thus taken to be Mo=S.     

Hyperfine coupling of O to Mo(V) in the rapid epr signal of xanthine oxidase: Evidence for an oxygen ligand of the metal

The spectrum of the Rapid Mo(V) electron paramagnetic resonance signal from xanthine oxidase dissolved in ¹⁷O-enriched water is presented. Difference technqiues have been used to eliminate the ¹⁶O contribution. Clearly observed structure in the spectrum is attributed to moderately strong hyperfine coupling of one oxygen atom to molybdenum. Though complete interpretation of the spectrum has not been attempted, one component of A(¹⁷O) is about 1.6 mT. The possibility that the oxygen is present in a Mo---OH group, whose proton is the strongly-coupled proton of the Rapid signal, is discussed.     

Proton electron-nuclear double-resonance spectra of molybdenum(V) in different reduced forms of xanthine oxidase

Electron-nuclear double-resonance (ENDOR) spectra of protons coupled to molybdenum(V) in reduced xanthine oxidase samples have been recorded. Under appropriate conditions these protons may be studied without interference from protons coupled to reduced iron-sulfur centers. Spectra have been obtained for the molybdenum(V) species known as Rapid, Slow, Inhibited, and Desulfo Inhibited. Resonances corresponding to at least nine protons or sets of protons are observed for all four species, with coupling constants in the range 0.08-4 MHz. Most of these protons do not exchange when 2H2O is used as solvent. Additional protons giving couplings up to 40 MHz are also detected. These correspond to EPR-detectable protons studied in earlier work. The strongly coupled protons may be replaced by 2H, through appropriate use of 2H2O or of 2H-substituted substrates, with consequent disappearance of the 1H resonances. In most cases the corresponding 2H ENDOR features have also been observed. The nature of the various coupled protons is briefly discussed. Results permit specific conclusions to be drawn about the structures of the Inhibited and Desulfo Inhibited species. In particular, the data indicate that the aldehyde residue of the Inhibited species has been oxidized and that the four protons derived from the ethylene glycol molecule in the Desulfo Inhibited species are not all equivalent. Recent assignments [Edmondson, D.E., & D'Ardenne, S.C. (1989) Biochemistry 28, 5924-5930] of the weakly coupled protons in the latter species appear not to be soundly based. The possibility of obtaining more detailed structural information from the spectra is briefly considered.(ABSTRACT TRUNCATED AT 250 WORDS)     

Studies by electron-paramagnetic-resonance spectroscopy of the molybdenum centre of aldehyde oxidase.

Molybdenum(V) e.p.r. spectra from reduced forms of aldehyde oxidase were obtained and compared with those from xanthine oxidase. Inhibited and Desulpho Inhibited signals from aldehyde oxidase were fully characterized, and parameters were obtained with the help of computer simulations. These differ slightly but significantly from the corresponding parameters for the xanthine oxidase signals. Rapid type 1 and type 2 and Slow signals were obtained from aldehyde oxidase, but were not fully characterized. From the general similarities of the signals from the two enzymes, it is concluded that the ligands of molybdenum must be identical and that the overall co-ordination geometries must be closely similar in the enzymes. The striking differences in substrate specificity must relate primarily to structural differences in a part of the active centre concerned with substrate binding and not involving the catalytically important molybdenum site.     

`Rapidly Appearing'' molybdenum electron-paramagnetic-resonance signals from reduced xanthine oxidase

Further electron-paramagnetic-resonance studies relating to the role of molybdenum in the enzymic mechanisms of xanthine oxidase were carried out. The classification of the various molybdenum signals obtained on reducing the enzyme is briefly discussed. The group of `Rapidly appearing'' signals, which are obtained with all substrates within the turnover time and which show interaction with exchangeable protons, were studied in detail. Signals with salicylaldehyde, purine and xanthine in H₂O and in 95% D₂O were examined at 9 and 35GHz and interpreted with the help of computer simulation. Molybdenum atoms in a number of different chemical environments are involved, each substrate giving rise to two superimposed spectra with slightly different parameters; g values and proton splittings were determined. The spectrum with salicylaldehyde is believed to represent the reduced enzyme alone not in the form of a complex with substrate and its two constituents are believed to represent the two molybdenum atoms bonded slightly differently within the enzyme molecule. With purine and xanthine the spectra are thought to represent complexes of reduced enzyme with substrate molecules. With xanthine one signal-giving species shows coupling to two equivalent protons, whereas in all the other species observed two non-equivalent protons are involved. The origin of the protons is discussed in the light of the direct hydrogen-transfer mechanism implicated earlier for the enzyme. It is concluded that the proton derived from the substrate is located at least 3å from the molybdenum atom with which it interacts.     

Rapid type 2 molybdenum(V) electron-paramagnetic resonance signals from xanthine oxidase and the structure of the active centre of the enzyme.

Rapid type 2 molybdenum(V) e.p.r. signals from reduced functional xanthine oxidase have been further investigated. These signals, which show strong coupling of two protons to molybdenum, have been obtained under a variety of new conditions: specifically either at pH 8.2 in the presence of borate ions, or at pH 10.1--10.7 with or without various other additions. Parameters of the signals were obtained with the help of computer simulations. In at least some of these signals, the coupled protons must be located on the enzyme rather than on bound species. The relationship between type 1 and type 2 Rapid signals is discussed. They may represent geometrical isomers, or alternatively, hydroxyl uptake as a ligand of molybdenum may be involved in formation of type 2 species.     

Transfer of exogenous cholesterol to microsomes of hepatocytes investigated with [3H]desmosterol tracer.

The molybdenum centre of spinach (Spinacia oleracea) nitrate reductase has been investigated by e.p.r. spectroscopy of molybdenum(V) in reduced forms of the enzyme. The resting enzyme gives no signals attributable to Mo(V). However, on reduction with NADH, Mo(V) signals appeared at relatively short reaction times but decreased again on prolonged exposure to excess of the substrate as the enzyme was further reduced. On brief treatment of such samples with nitrate, Mo(V) signals reappeared but disappeared again on longer exposure to excess nitrate as the enzyme became fully reoxidized. Detailed investigation of the signals carried out in both 1H2O and 2H2O revealed the presence of two signal-giving species, referred to as 'signal A' and 'signal B', analogous to corresponding signals from nitrate reductase from Escherichia coli and from liver sulphite oxidase. Signal A has gav. 1.9767 and shows coupling to a single proton, exchangeable with the solvent, with A(1H)av. 1.3mT, whereas signal B shows no more than weak coupling to protons. Investigation of interconversion between the two species indicated that decreasing the pH from 8.0 to 6.7 had little effect, but that signal A was favoured by the presence of Cl-. This suggests, by analogy with recent work on sulphite oxidase by Bray, Gutteridge, Lamy & Wilkinson [Biochem. J. (1983) 211, 227-236] that Cl- is a ligand of molybdenum in the species giving signal A.     

Oxidation--reduction potentials of turkey liver xanthine dehydrogenase and the origins of oxidase and dehydrogenase behaviour in molybdenum-containing hydroxylases.

Redox potentials for the various centres in the enzyme xanthine dehydrogenase (EC 1.2.1.37) from turkey liver determined by potentiometric titration in the presence of mediator dyes, with low-temperature electron-paramagnetic-resonance spectroscopy. Values at 25 degrees C in pyrophosphate buffer, pH 8.2, are: Mo(VI)/Mo(V)(Rapid),-350 +/- 20mV; Mo(V) (Rapid)/Mo(IV), -362 +/- 20mV; Fe-S Iox./Fe-S Ired., -295 +/- 15mV; Fe-S IIox./Fe-S IIred., -292 +/- 15mV; FAD/FADH,-359+-20mV; FADH/FADH2, -366 +/- 20mV. This value of the FADH/FADH2 potential, which is 130mV lower than the corresponding one for milk xanthine oxidase [Cammack, Barber & Bray (1976) Biochem. J. 157, 469-478], accounts for many of the differences between the two enzymes. When allowance is made for some interference by desulpho enzyme, then differences in the enzymes' behaviour in titration with xanthine [Barber, Bray, Lowe & Coughlan (1976) Biochem. J. 153, 297-307] are accounted for by the potentials. Increases in the molybdenum potentials of the enzymes caused by the binding of uric acid are discussed. Though the potential of uric acid/xanthine (-440mV) is favourable for full reduction of the dehydrogenase, nevertheless, during turnover, for kinetic reasons, only FADH and very little FADH2 is produced from it. Since only FADH2 is expected to react with O2, lack of oxidase activity by the dehydrogenase is explained. Reactivity of the two enzymes with NAD+ as electron acceptor is discussed in relation to the potentials.     

Differences between tissues in response of S-adenosylmethionine decarboxylase to administration of polyamines.

Different reduced forms of xanthine oxidase, labelled specifically in the cyanide-labile site with 33S, were prepared and examined by electron paramagnetic resonance. Coupling of this isotope to molybdenum(V) was quantified with the help of computer simulations and found to differ markedly from one reduced form to another. The xanthine Very Rapid signal shows strong, highly anisotropic, coupling with A(33S)av. 1.27 mT. For this signal, axes of the g- and A(33S)-tensors are rotated relative to one another. One axis of the A-tensor is in the plane of gxx ang gyy, but rotated by 40 degrees relative to the gxx axis, whereas the direction of weakest coupling to sulphur deviates by 10 degrees from the gzz axis. In contrast with this signal, only rather weaker coupling was observed in different types of Rapid signal [A(33S)av. 0.3--0.4 mT], and in the Inhibited signal coupling was weaker still [A(33S)av. 0.1--0.2 mT]. Clearly, there must be substantial differences in the structures of the molybdenum centre in the different signal-giving species, with the sulphur atom perhaps in an equatorial type of ligand position in the Very Rapid species but in a more axial one in the other species. Structures are discussed in relation to the mechanism of action of the enzyme and the nature of the proton-accepting group that participates in turnover.     

Oxygen-17 splitting of the very rapid molybdenum(V) e.p.r. signal from xanthine oxidase. Rate of exchange with water of the coupled oxygen atom.

Studies have been carried out of effects of 17O substitution on a Mo(V) e.p.r. signal from xanthine oxidase, known as Very Rapid. This transient signal is believed to represent an intermediate in enzymic turnover. When Very Rapid was developed from enzyme equilibrate with 17O-enriched water, strong coupling of Mo(V) to a single oxygen atom was observed, with A(17O)1,2,3 1.34, 1.40, 1.36 mT. The isotropic character of the splittings is interpreted as favouring a structure of the type Mo--O--C. The rate of exchange with water of the oxygen atom detected in the signal was studied. In oxidized enzyme, which contains a terminal oxygen ligand, the exchange rate constant was 2--4 h-1 (pH 5.9--7.8 and about 20 degrees C). However, if the exchange was allowed to take place whilst the enzyme was turning over a substrate, then the process occurred within a few seconds. The present and previous results are interpreted as favouring an enzymic mechanism in which a terminal oxygen ligand reacts, as a nucleophile, with a substrate carbonium ion. To complete the reaction, product liberation, by hydrolysis of the enzyme-bound species, occurs in such a way as to cleave the Mo--O bond, thus explaining the fast oxygen exchange in the presence of the substrate.     

On the mechanism of action of xanthine oxidase. Evidence in support of an oxo transfer mechanism in the molybdenum-containing hydroxylases

The mechanism of action of xanthine oxidase has been investigated using single-turnover experiments in an effort to determine the primary source of the oxygen atom incorporated into product in the course of catalysis. It is found from mass spectroscopic analysis of the uric acid generated in these experiments that when 16O-labeled enzyme in [18O]H2O is reacted with substoichiometric amounts of xanthine (under conditions where no enzyme molecule is likely to react with more than one substrate molecule), the uric acid isolated from the reaction mixture contains 16O at position 8 of the purine ring. Conversely, when 18O-labeled enzyme in [16O]H2O is exposed to substoichiometric xanthine, 18O is incorporated into the product uric acid. These results strongly support a variety of chemical studies with model molybdenum complexes suggesting that the oxygen atom of the Mo = O group known to be present at the active site of xanthine oxidase is transferred to product in the course of catalysis. The mechanistic implications of the present work are discussed.     

The heterogeneity of arginases in rat tissues.

1. The mid-point reduction potentials of the various groups in xanthine oxidase from bovine milk were determined by potentiometric titration with dithionite in the presence of dye mediators, removing samples for quantification of the reduced species by e.p.r. (electron-paramagnetic-resonance) spectroscopy. The values obtained for the functional enzyme in pyrophosphate buffer, pH8.2, are: Fe/S centre I, -343 +/- 15mV; Fe/S II, -303 +/- 15mV; FAD/FADH-; -351 +/- 20mV; FADH/FADH2, -236 +/-mV; Mo(VI)/Mo(V) (Rapid), -355 +/- 20mV; Mo(V) (Rapid)/Mo(IV), -355 +/- 20mV. 2. Behaviour of the functional enzyme is essentially ideal in Tris but less so in pyrophosphate. In Tris, the potential for Mo(VI)/Mo(V) (Rapid) is lowered relative to that in pyrophosphate, but the potential for Fe/S II is raised. The influence of buffer on the potentials was investigated by partial-reduction experiments with six other buffers. 3. Conversion of the enzyme with cyanide into the non-functional form, which gives the Slow molybdenum signal, or alkylation of FAD, has little effect on the mid-point potentials of the other centres. The potentials associated with the Slow signal are: Mo(VI)/Mo(V) (Slow), -440 +/- 25mV; Mo(V) (Slow)/Mo(IV), -480 +/- 25 mV. This signal exhibits very sluggish equilibration with the mediator system. 4. The deviations from ideal behaviour are discussed in terms of possible binding of buffer ions or anti-co-operative interactions amongst the redox centres.