Structure of the oxidized active site of galactose oxidase from realistic in silico models

A systematic in silico approach is employed to generate an accurate model for the catalytically important oxidized state of galactose oxidase (GO) using spectroscopically calibrated hybrid density-functional theory. GO displays three distinct oxidation states: oxidized [Cu(II)-Y*], semireduced [Cu(II)-Y], and fully reduced [Cu(I)-Y], but only the [Cu(II)-Y*] and the [Cu(I)-Y] states are assumed to be involved in catalysis. We have developed multiple models for the oxidized [Cu(II)-Y*] state, whose structure has not yet been fully characterized. These models were evaluated by comparison of calculated and experimental structural data, singlet-triplet energy gaps, and electronic transitions for the antiferromagnetically coupled oxidized [Cu(II)-Y*] state. An extended model system that includes explicit solvent molecules and second coordination sphere residues (R330, Y405, and W290) is essential to obtain the correct electronic structure of the active site. The model with all the residues that have been shown to affect the radical stability and catalysis resulted in a singlet ground state with the radical centered on the Y272-C228 cofactor. The optimized structure of the oxidized GO [Cu(II)-Y*] reveals a five-coordinated square pyramidal coordination geometry very similar to [Cu(II)-Y] with considerably different Cu-ligand distances. The hydrogen-bonding interactions involving Y495 modulates the spin density distribution and the singlet-triplet energy gaps. The final model as the most reasonable structure of the oxidized [Cu(II)-Y*] state in GO reproduces the spectroscopic signature of oxidized GO.     

Indirect evidence of direct electron communication between the active site of galactose oxidase and a graphite electrode

Bi-enzymatic biosensor based on galactose oxidase (GalOD) and horseradish peroxidase (HRP) using ferrocene as an efficient mediator was constructed. When a dependence of a working potential on the sensor performance was examined, an unusual behaviour was observed. With increasing of an applied working potential a lower concentration of substrate to attain full linear range was needed. A fully linear dependence from the first substrate addition was observed at and above the working potential of 150 mV. This activation of the biosensor response by an applied working potential very well corresponds with a formal potential of GalOD (156 mV). When a membrane prevented GalOD access to the electrode surface was applied, no activation effect of a working potential on the sensor performance was observed. Thus, it can be assumed that direct electron communication between GalOD and the electrode occurred.     

Dissecting an enzyme—model compounds for the galactose oxidase radical site

The active site of the enzyme galactose oxidase (GOase) contains square‐pyramidal monocopper site, one of whose ligands is a tyrosinate side‐chain that is oxidized to an unusually stable radical in the active enzyme. The structure of this non‐innocent tyrosinate is unique in two ways. First, the tyrosine ring is crosslinked to a neighboring cysteine residue, affording an orthoalkylsulfanyl‐substituted phenoxide ligand. Second, this assembly is protected by a π–π interaction to a tryptophan indole group. We describe here a series of compounds designed to model various aspects of the structure of this unusual cofactor. Our studies have shown that the thermodynamic stability of the GOase radical can be attributed almost exclusively to its thioether substituent, that the π–π interaction contributes little to this stability, and that the assignment of the optical spectrum of the GOase radical is more complex than had been previously suggested. © 2002 Wiley Periodicals, Inc. Heteroatom Chem 13:494–500, 2002; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/hc.10091     

Models for the active site in galactose oxidase: Structure, spectra and redox of copper(II) complexes of certain phenolate ligands

Galactose oxidase (GOase) is a fungal enzyme which is unusual among metalloenzymes in appearing to catalyse the two electron oxidation of primary alcohols to aldehydes and H₂O₂. The crystal structure of the enzyme reveals that the coordination geometry of mononuclear copper(II) ion is square pyramidal, with two histidine imidazoles, a tyrosinate, and either H₂O (pH 7.0) or acetate (from buffer,pH 4-5) in the equatorial sites and a tyrosinate ligand weakly bound in the axial position. This paper summarizes the results of our studies on the structure, spectral and redox properties of certain novel models for the active site of the inactive form of GOase. The monophenolato Cu(II) complexes of the type [Cu(L1)X][H(L1) = 2-(bis(pyrid-2-ylmethyl)aminomethyl)-4-nitrophenol and X⁻ = Cl⁻ 1, NCS⁻ 2, CH₃COO⁻ 3, ClO₄ ⁻ 4] reveal a distorted square pyramidal geometry around Cu(II) with an unusual axial coordination of phenolate moiety. The coordination geometry of 3 is reminiscent of the active site of GOase with an axial phenolate and equatorial CH₃COO⁻ ligands. All the present complexes exhibit several electronic and EPR spectral features which are also similar to the enzyme. Further, to establish the structural and spectroscopic consequences of the coordination of two tyrosinates in GOase enzyme, we studied the monomeric copper(II) complexes containing two phenolates and imidazole/pyridine donors as closer structural models for GOase. N,N-dimethylethylenediamine and N,N’-dimethylethylenediamine have been used as starting materials to obtain a variety of 2,4-disubstituted phenolate ligands. The X-ray crystal structures of the complexes [Cu(L5)(py)], (8) [H₂(L5) = N,N-dimethyl-N’,N’-bis(2-hydroxy-4-nitrobenzyl) ethylenediamine, py = pyridine] and [Cu(L8)(H₂O)] (11), [H₂(L8) = N,N’-dimethyl-N,N’-bis(2-hydroxy-4-nitrobenzyl)ethylenediamine] reveal distorted square pyramidal geometries around Cu(II) with the axial tertiary amine nitrogen and water coordination respectively. Interestingly, for the latter complex there are two different molecules present in the same unit cell containing the methyl groups of the ethylenediamine fragmentcis to each other in one molecule andtrans to each other in the other. The ligand field and EPR spectra of the model complexes reveal square-based geometries even in solution. The electrochemical and chemical means of generating novel radical species of the model complexes, analogous to the active form of the enzyme is presently under investigation.     

The active site of arsenite oxidase from Alcaligenes faecalis

Arsenite oxidase, a member of the DMSO reductase family of molybdenum enzymes, has two molecules of guanosine dinucleotide molybdenum cofactor coordinating the molybdenum at the active site. X-ray absorption spectroscopy indicates that the Mo-S bonds shorten from 2.47 to 2.37 A upon reduction with the physiological substrate. It also indicates the presence of an oxo ligand at 1.70 A in both oxidized and reduced forms of the enzyme, together with a short, 1.83 A, Mo-O bond in the oxidized form that is lost upon reduction. Resonance Raman spectroscopy indicates that the two pterin dithiolene moieties have different aromaticities, with one, the Q-pterin, having a more discrete dithiolate structure while the other, the P-pterin, has considerable pi-delocalization. Our results indicate that the structure of arsenite oxidase is intermediate between that seen in other molybdenum enzymes, in which one ligand to the metal is provided by the polypeptide (serine, cysteine, or selenocysteine), and tungsten enzymes that lack a peptide ligand.     

Role of the electronic properties of azurin active site in the electron‐transfer process

Electron transfer proteins, such as azurin (a blue copper protein), are promising candidates for the implementation of biomolecular nanoelectronic devices. To understand the details of electron transfer in redox active azurin molecules, we performed plane‐wave pseudo‐potential density functional theory (DFT) calculations of the protein active site in the two possible oxidation states Cu(I) and Cu(II). The ab initio results are used to discuss how the electronic spectrum and wavefunctions may mediate the shuttling of electrons through the copper ion. We find that the Cu‐ligand hybridization is very similar in the two charge states of the metal center, but the energy spectrum changes substantially. This result might indicate important effects of electronic correlations in the redox activity and consequent electron transfer through the Cu site. © 2004 Wiley Periodicals, Inc. Int J Quantum Chem, 2005     

Bioelectrochemical response of the polyaniline galactose oxidase electrode

Galactose oxidase has been immobilized in a polyaniline film. The response current of the galactose oxidase electrode is a function of the applied potential and increases as the pH increases from 5.61 to 7.25. The optimum pH of the immobilized galactose oxidase is 7.25. The activation energy of the enzyme-catalysed reaction is 41.8 kJ mol⁻¹. The response current of the enzyme electrode shows good reproducibility at temperatures below the optimum temperature of 30.4°C and increases as the galactose concentration increases from 0.2 to 6 mmol dm⁻³. Thus the polyaniline galactose oxidase electrode can be used to determine galactose concentration.     

Nature of the catalytically labile oxygen at the active site of xanthine oxidase

In this paper we report the results of molybdenum K-edge X-ray absorption studies performed on the oxidized active site of xanthine oxidase at pH 6 and 10. These results indicate that the active site possesses one terminal oxygen ligand (Mo=O), two thiolate ligands (Mo-S), one terminal sulfido ligand (Mo=S), and one Mo-OH moiety. EXAFS analysis demonstrates that the Mo-OH bond shortens from 1.97 A at pH 6 to 1.75 A at pH 10, which is consistent with the generation of a Mo-O- moiety. This study provides convincing structural evidence that the catalytic oxygen donor at the oxidized active site of xanthine oxidase is Mo-OH rather than the Mo-OH2 ligation previously suggested by X-ray crystallography. These results support a mechanism initiated by base-assisted nucleophilic attack of the substrate by Mo-OH.     

Calculated properties of the active site of oxidized rubredoxins

For a molecular model of the Fe-S active site complex in oxidized rubredoxin, we have calculated the spin-orbit coupling between the ground sextet state and excited quartet and doublet states which gives rise to the observed zero field splitting of the sextet ground state into three spin-mixed Kramers doublets. Additionally, we have used the six spin-mixed sextet state components to calculate effective magnetic moments, magnetic field energies and nine g values corresponding to transitions between the three pairs of Kramers doublets in applied magnetic fields along three perpendicular axes. We have calculated these properties for eight conformational variations of the ligands around the Fe at the active site. The results of these calculations clearly show the origin of the observed g=4.3 signal previously described only in terms of the phenomenological spin-Hamiltonian formalism. For the eight conformations considered, five have this characteristic signal. Zero field splitting comparable to the observed values could be obtained for all symmetries studied. In addition, the calculated values of magnetic moment in all symmetries correspond to that of high spin ferric ion and do not vary appreciably with temperature above 77° K, in agreement with experimental results. From comparison of all our calculated results with experiment, it appears that the active site in oxidized rubredoxins could have small conformational variations in different rubredoxins and under the various experimental conditions used.     

Pi-interaction tuning of the active site properties of metalloproteins

The influence of pi-interactions with a His ligand have been investigated in a family of copper-containing redox metalloproteins. The Met16Phe and Met16Trp pseudoazurin, and Leu12Phe spinach and Leu14Phe Phormidium laminosum plastocyanin variants possess active-site pi-contacts between the introduced residue and His81 and His87/92 respectively. The striking overlap of the side chain of Phe16 in the Met16Phe variant and that of Met16 in wild type pseudoazurin identifies that this position provides an important second coordination sphere interaction in both cases. His-ligand protonation and dissociation from Cu(I) occurs in the wild type proteins resulting in diminished redox activity, providing a [H(+)]-driven switch for regulating electron transfer. The introduced pi-interaction has opposing effects on the pKa for the His ligand in pseudoazurin and plastocyanin due to subtle differences in the pi-contact, stabilizing the coordinated form of pseudoazurin whereas in plastocyanin protonation and dissociation is favored. Replacement of Pro36, a residue that has been suggested to facilitate structural changes upon His ligand protonation, with a Gly, has little effect on the pKa of His87 in spinach plastocyanin. The mutations at Met16 have a significant influence on the reduction potential of pseudoazurin. Electron self-exchange is enhanced, whereas association with the physiological partner, nitrite reductase, is only affected by the Met16Phe mutation, but kcat is halved in both the Met16Phe and Met16Trp variants. Protonation of the His ligand is the feature most affected by the introduction of a pi-interaction.     

Shape-shifting tetranuclear oxo-bridged manganese cluster: relevance to photosystem II water oxidase active site

The redox properties of the "dimer-of-dimers" complex, [{Mn2(mu-O)2(tphpn)}2]4+ (1) (where Htphpn = N,N,N',N'-tetra(2-methylpyridyl)-2-hydroxypropane-diamine) were investigated. The structure changes dramatically to an adamantane-shaped core upon one-electron oxidation. On the other hand, the one-electron reduced product of 1, [Mn4O4(tphpn)2]3+, exhibits a hyperfine-structured multiline EPR signal very similar to the so-called S0 state of the tetramanganese cluster, which resides at the Photosystem II water oxidase active site.     

Modified active site coordination in a clinical mutant of sulfite oxidase

The molybdenum site of the Arginine 160 --> Glutamine clinical mutant of the physiologically vital enzyme sulfite oxidase has been investigated by a combination of X-ray absorption spectroscopy and density functional theory calculations. We conclude that the mutant enzyme has a six-coordinate pseudo-octahedral active site with coordination of Glutamine Oepsilon to molybdenum. This contrasts with the wild-type enzyme which is five-coordinate with approximately square-based pyramidal geometry. This difference in the structure of the molybdenum site explains many of the properties of the mutant enzyme which have previously been reported.     

Calculated properties of the active site complex of oxidized rubredoxins

The active site of rubredoxins, the simplest class of iron-sulfur, electron-transfer proteins consists of a single Fe atom surrounded by a distorted tetrahedral array of four cysteine sulfur atoms. With the aid of a newly formulated computer program, we have calculated the electric field gradient at the Fe⁵⁷ nucleus, the resultant quadrupole splitting (ΔE _Q) and the isotropic and anisotropic parts of the hyperfine interaction between the Fe nuclear spin and the net electron spin in the sextet ground state for ten conformational variations of the active site complex. For each conformer we used an electron distribution obtained from an Iterated Extended Huckel Theory (IEHT) molecular orbital calculation. Comparison of calculated and experimental results of (ΔE _Q) obtained from Mossbauer resonance spectra indicated that a number of conformers have field gradient values in excellent agreement with experiment. Good agreement with experiment was also found for the calculated hyperfine coupling constant. In this interaction, the isotropic contribution is dominant while the anisotropic contribution is more symmetry dependent. Since a single average experimental value is observed, hyperfine interaction in this system is not a very sensitive probe of active site conformation.     

Functional biomimetic models for the active site in the respiratory enzyme cytochrome c oxidase

A functional analog of the active site in the respiratory enzyme, cytochrome c oxidase (CcO) reproduces every feature in CcO's active site: a myoglobin-like heme (heme a3), a distal tridentate imidazole copper complex (Cu(B)), a phenol (Tyr244), and a proximal imidazole. When covalently attached to a liquid-crystalline SAM film on an Au electrode, this functional model continuously catalyzes the selective four-electron reduction of dioxygen at physiological potential and pH, under rate-limiting electron flux (as occurs in CcO).     

Galactosyl-biomimetic dye-ligands for the purification of Dactylium dendroides galactose oxidase

Two anthraquinone galactosyl-biomimetic dye-ligands comprising, as terminal biomimetic moiety, galactose analogues (1-amino-1-deoxy-beta-D-galactose and D(+)-galactosamine) were designed for the enzyme galactose oxidase (GAO), using molecular modelling, synthesized and characterized. The biomimetic ligands were immobilized on agarose beads and the affinity adsorbents, together with a non-biomimetic adsorbent bearing Cibacron Blue 3GA, were studied for their ability to purify GAO from Dactylium dendroides. Both biomimetic adsorbents showed higher purifying ability for GAO compared to the non-biomimetic adsorbent, thus demonstrating their superior effectiveness as affinity chromatography materials. In particular, the affinity adsorbent comprising, as terminal biomimetic moiety, 1-amino-1-deoxy-beta-D-galactose (BM1) exhibited the highest purifying ability for GAO. This affinity adsorbent did not bind galactose dehydrogenase, glucose dehydrogenase, alcohol dehydrogenase, or glucose oxidase. The dissociation constant (K(D)) of the immobilized BM1 ligand with GAO was found to be equal to 45.8 microM, whereas the binding capacity was equal to 709 U per ml adsorbent. Therefore, the BMI adsorbent was integrated in a facile two-step purification procedure for GAO. The purified enzyme showed a specific activity equal to 2038 U/mg, the highest reported so far, approximately 74% overall recovery and a single band after sodium dodecylsulfate-polyacrylamide gel electrophoresis analysis.     

Catalytic effects of galactose oxidase on micelle-forming galactolipids

Catalytic effects of galactose oxidase on the oxidation of beta-D-galactose-carrying lipids with an oligo-ethylene glycol spacer (number of ethylene glycol units (n)=1, 2, 3, 6, 9, 13, and 20) were examined. The affinity of galactose oxidase for the galactose residue in the amphiphile (estimated by the inverse of the Michaelis constant, K(m)) was much higher than those for free D-galactose and small beta-D-galactopyranosides, and dependent on the length of the ethylene glycol spacer. That is, both below and above the critical micellar concentration, the 1/K(m) values decreased with an increase in the n value. The effectiveness of the enzyme, which can be estimated by the k(cat)/K(m) value, showed the same tendency as the 1/K(m) value. These results could be attributed to the role of the nonpolar environment around the galactose residue in the binding by the enzyme. A significant enhancement of the enzymatic oxidation of galactose residue on the liposome surface was also observed.     

Heavily fluorinated carbohydrates as enzyme substrates: oxidation of tetrafluorinated galactose by galactose oxidase

Galactose oxidase (GOase) was shown to oxidise several C2/C3 fluorinated galactose analogues. Interestingly, the enzyme was able to distinguish between the 2,3-tetrafluorinated galactose and its epimeric glucose analogue, and this represents the first reported biotransformation of a heavily fluorinated sugar.