Xanthine dehydrogenase from Pseudomonas putida 86: specificity, oxidation-reduction potentials of its redox-active centers, and first EPR characterization.

Xanthine dehydrogenase (XDH) from Pseudomonas putida 86, which was induced 65-fold by growth on hypoxanthine, was purified to homogeneity. It catalyzes the oxidation of hypoxanthine, xanthine, purine, and some aromatic aldehydes, using NAD+ as the preferred electron acceptor. In the hypoxanthine:NAD+ assay, the specific activity of purified XDH was 26.7 U (mg protein)(-1). Its activity with ferricyanide and dioxygen was 58% and 4%, respectively, relative to the activity observed with NAD+. XDH from P. putida 86 consists of 91.0 kDa and 46.2 kDa subunits presumably forming an alpha4beta4 structure and contains the same set of redox-active centers as eukaryotic XDHs. After reduction of the enzyme with xanthine, electron paramagnetic resonance (EPR) signals of the neutral FAD semiquinone radical and the Mo(V) rapid signal were observed at 77 K. Resonances from FeSI and FeSII were detected at 15 K. Whereas the observable g factors for FeSII resemble those of other molybdenum hydroxylases, the FeSI center in contrast to most other known FeSI centers has nearly axial symmetry. The EPR features of the redox-active centers of P. putida XDH are very similar to those of eukaryotic XDHs/xanthine oxidases, suggesting that the environment of each center and their functionality are analogous in these enzymes. The midpoint potentials determined for the molybdenum, FeSI and FAD redox couples are close to each other and resemble those of the corresponding centers in eukaryotic XDHs.

Probing magnetic properties of the reduced [2Fe-2S] cluster of the ferredoxin from Arthrospira platensis by 1H ENDOR spectroscopy.

The 1H electron nuclear double resonance (ENDOR) spectra in frozen solutions of the reduced [2Fe-2S] cluster in ferredoxin from Arthrospira (Spirulina) platensis have been measured at low temperatures (5-20 K) and simulated using orientational selection methods. The analysis confirmed the existence of a single paramagnetic species with iron valence states II and III connected uniquely to the cluster irons. The experimental ENDOR spectra were fitted to a model including the spin distribution on the centre, the orientation of the g-matrix, and the isotropic and anisotropic hyperfine couplings of the nearest protons in the crystallographically determined structure. In order to partially simulate ENDOR line shapes, a statistical distribution of the corresponding torsion angles between the Fe(III) centre and one of the beta-CH2 protons was introduced. From the analysis, four of the larger hyperfine couplings found were assigned to the cysteine beta-protons near the Fe(III) ion of the cluster, with isotropic hyperfine couplings ranging from 1.6 to 4.1 MHz. The spin distribution on the two iron ions was estimated to be +1.85 for the Fe(III) ion and -0.9 for the Fe(II) ion. The Fe(III) ion was identified as being coordinated to the cysteine ligands Cys49 and Cys79, confirming previous NMR results. The direction of the g-tensor with respect to the cluster was deduced. The g1-g2 plane is parallel to the planes through each iron and its adjacent cysteine sulfurs; the g2-g3 plane is nearly perpendicular to the latter planes and deviates by 25 degrees from the FeSSFe plane. The g1 direction is dominated by the bonding geometry of Fe(II) and does not align with the Fe(II)-Fe(III) vector.

Kinetics and interactions of molybdenum and iron-sulfur centers in bacterial enzymes of the xanthine oxidase family: mechanistic implications.

For isoquinoline 1-oxidoreductase (IsoOr), the reaction mechanism under turnover conditions was studied by EPR spectroscopy using rapid-freeze methods. IsoOr displays several EPR-active Mo(V) species including the "very rapid" component found also in xanthine oxidase (XanOx). For IsoOr, unlike XanOx or quinoline 2-oxidoreductase (QuinOr), this species is stable for about 1 h in the absence of an oxidizing substrate [Canne, C., Stephan, I., Finsterbusch, J., Lingens, F., Kappl, R., Fetzner, S., and Hüttermann, J. (1997) Biochemistry 36, 9780-9790]. Under rapid-freeze conditions in the presence of ferricyanide the very rapid species behaves as a kinetically competent intermediate present only during steady-state turnover. To explain the persistence of the very rapid species in IsoOr in the absence of an added oxidant, extremely slow product dissociation is required. This new finding that oxidative conditions facilitate decay of the very rapid signal for IsoOr supports the mechanism of substrate turnover proposed by Lowe, Richards, and Bray [Lowe, D. J., Richards, R. L., and Bray, R. C. (1997) Biochem. Soc. Trans. 25, 774-778]. Additional stopped-flow data reveal that alternative catalytic cycles occur in IsoOr and show that the product dissociates after transfer of a single oxidizing equivalent from ferricyanide. In rapid-freeze measurements magnetic interactions of the very rapid Mo(V) species and the iron-sulfur center FeSI of IsoOr and QuinOr were observed, proving that FeSI is located close to the molybdopterin cofactor in the two proteins. This finding is used to relate the two different iron-sulfur centers of the aldehyde oxidoreductase structure with the EPR-detectable FeS species of the enzymes.

Comparative EPR and redox studies of three prokaryotic enzymes of the xanthine oxidase family: quinoline 2-oxidoreductase, quinaldine 4-oxidase, and isoquinoline 1-oxidoreductase.

For three prokaryotic enzymes of the xanthine oxidase family, namely quinoline 2-oxidoreductase, quinaldine 4-oxidase, and isoquinoline 1-oxidoreductase, the electron transfer centers were investigated by electron paramagnetic resonance. The enzymes are containing a molybdenum-molybdopterin cytosine dinucleotide cofactor, two distinct [2Fe-2S] clusters and, apart from isoquinoline 1-oxidoreductase, a flavin adenine dinucleotide. The latter cofactor yields two different organic radical signals in quinoline 2-oxidoreductase and quinaldine 4-oxidase, typical for the neutral and anionic form, respectively. A "rapid" Mo(V) species is present in all enzymes with small differences in magnetic parameters. From spectra simulation of 95Mo-substituted quinoline 2-oxidoreductase, a deviation of 25 degrees between the maximal g and 95Mo-hyperfine tensor component was derived. The very rapid Mo(V) species was detected in small amounts upon reduction with substrates in quinoline 2-oxidoreductase and quinaldine 4-oxidase, but showed a different kinetic behavior with considerable EPR intensities in isoquinoline 1-oxidoreductase. The FeSI and FeSII centers produced different signals in all three enzymes and, in case of isoquinoline 1-oxidoreductase, revealed a dipolar interaction, from which a maximum distance of 15 A between FeSI and FeSII was estimated. The midpoint potentials of the FeS centers were surprisingly different and determined for FeSI/FeSII with -155/-195 mV in quinoline 2-oxidoreductase, -250/-70 mV in quinaldine 4-oxidase, and +65/+10 mV in isoquinoline 1-oxidoreductase. The slopes of the fitting curves for the Nernst equation are indicative for nonideal behavior. Only in quinoline 2-oxidoreductase, an averaged midpoint potential of the molybdenum redox pairs of about -390 mV could be determined. Both of the other enzymes did not produce Mo(V) signals in redox titration experiments, probably because of direct reduction of Mo(VI) to Mo(IV) in the presence of dithionite.