JBIC Journal of Biological Inorganic Chemistry (v.11, #3)

The results of studies performed in the author’s laboratory are surveyed, with particular emphasis on demonstrating the value of a multidisciplinary synthetic modeling approach for discovering new and unusual chemistry helpful for understanding the properties of the active sites of copper proteins or assessing the feasibility of mechanistic pathways they might follow during catalysis. The discussion focuses on the progress made to date toward comprehending the nitrite reductase catalytic site and mechanism, the electronic structures of copper thiolate electron transfer centers, the sulfido-bridged “CuZ” site in nitrous oxide reductase, and the processes of dioxygen binding and activation by mono- and dicopper centers in oxidases and oxygenases.
Keywords: Copper proteins; Thiolate; Sulfide; Nitric oxide; Dioxygen

Oxoiron(IV) complexes of the tris(2-pyridylmethyl)amine ligand family: effect of pyridine α-substituents by Tapan K. Paine; Miquel Costas; József Kaizer; Lawrence Que Jr (272-276).
The oxoiron(IV) complexes of two 6-substituted tris(2-pyridylmethyl)amine ligand derivatives have been generated and characterized with respect to their spectroscopic and reactivity properties. The introduction of an α-substituent maintains the low-spin nature of the oxoiron(IV) unit but weakens the ligand field, as evidenced by red shifts in its characteristic near-IR chromophore. While its hydrogen-atom abstraction ability is only slightly affected, the oxo-transfer reactivity of the oxoiron(IV) center is significantly enhanced relative to that of the parent complex. These results demonstrate that the ligand environment plays a key role in modulating the reactivity of this important biological oxidant.
Keywords: Oxoiron(IV) complexes; Tris(2-pyridylmethyl)amine; Nonheme iron intermediates; Taurine/α-ketoglutarate dioxygenase

Reinvestigation of the method used to map the electronic structure of blue copper proteins by NMR relaxation by D. Flemming Hansen; Serge I. Gorelsky; Ritimukta Sarangi; Keith O. Hodgson; Britt Hedman; Hans E. M. Christensen; Edward I. Solomon; Jens J. Led (277-285).
A previous method for mapping the electron spin distribution in blue copper proteins by paramagnetic nuclear magnetic resonance (NMR) relaxation (Hansen DF, Led JJ, 2004, J Am Chem Soc 126:1247–1253) suggested that the blue copper site of plastocyanin from Anabaena variabilis (A.v.) is less covalent than those found for other plastocyanins by other experimental methods, such as X-ray absorption spectroscopy. Here, a detailed spectroscopic study revealed that the electronic structure of A.v. plastocyanin is similar to those of other plastocyanins. Therefore, the NMR approach was reinvestigated using a more accurate geometric structure as the basis for the mapping, in contrast to the previous approach, as well as a more complete spin distribution model including Gaussian-type natural atomic orbitals instead of Slater-type hydrogen-like atomic orbitals. The refinement results in a good agreement between the electron spin density derived from paramagnetic NMR and the electronic structure description obtained by the other experimental methods. The refined approach was evaluated against density functional theory (DFT) calculations on a model complex of the metal site of plastocyanin in the crystal phase. In general, the agreement between the experimental paramagnetic relaxation rates and the corresponding rates obtained by the DFT calculations is good. Small deviations are attributed to minor differences between the solution structure and the crystal structure outside the first coordination sphere. Overall, the refined approach provides a complementary experimental method for determining the electronic structure of paramagnetic metalloproteins, provided that an accurate geometric structure is available.
Keywords: Electronic structure; Blue copper proteins; Plastocyanin; NMR; Paramagnetic nuclear relaxation; DFT

Density functional study of the catalytic cycle of nickel–iron [NiFe] hydrogenases and the involvement of high-spin nickel(II) by Alejandro Pardo; Antonio L. De Lacey; Víctor M. Fernández; Hua-Jun Fan; Yubo Fan; Michael B. Hall (286-306).
In light of recent experiments suggesting high-spin (HS) Ni(II) species in the catalytic cycle of [NiFe] hydrogenase, a series of models of the Ni(II) forms Ni-SI(I,II), SI-CO and Ni-R(I,II,III) were examined in their high-spin states via density functional calculations. Because of its importance in the catalytic cycle, the Ni–C form was also included in this study. Unlike the Ni(II) forms in previous studies, in which a low-spin (LS) state was assumed and a square–planar structure found, the optimized geometries of these HS Ni(II) forms resemble those observed in the crystal structures: a distorted tetrahedral to distorted pyramidal coordination for the NiS4. This resemblance is particularly significant because the LS state is 20–30 kcal/mol less stable than the HS state for the geometry of the crystal structure. If these Ni(II) forms in the enzyme are not high spin, a large change in geometry at the active site is required during the catalytic cycle. Furthermore, only the HS state for the CO-inhibited form SI-CO has CO stretching frequencies that match the experimental results. As in the previous work, these new results show that the heterolytic cleavage reaction of dihydrogen (where H2 is cleaved with the metal acting as a hydride acceptor and a cysteine as the proton acceptor) has a lower energy barrier and is more exothermic when the active site is oxidized to Ni(III). The enzyme models described here are supported by a calibrated correlation of the calculated and measured CO stretching frequencies of the forms of the enzyme. The correlation coefficient for the final set of models of the forms of [NiFe] hydrogenase is 0.8.
Keywords: [NiFe] hydrogenases; Density functional theory; High-spin Ni(II); Catalytic cycle; Quantum mechanics

Desulfovibrio gigas ferredoxin II: redox structural modulation of the [3Fe–4S] cluster by Pedro M. Rodrigues; Anjos L. Macedo; Brian J. Goodfellow; Isabel Moura; José J. G. Moura (307-315).
Desulfovibrio gigas ferredoxin II (DgFdII) is a small protein with a polypeptide chain composed of 58 amino acids, containing one Fe3S4 cluster per monomer. Upon studying the redox cycle of this protein, we detected a stable intermediate (FdIIint) with four 1H resonances at 24.1, 20.5, 20.8 and 13.7 ppm. The differences between FdIIox and FdIIint were attributed to conformational changes resulting from the breaking/formation of an internal disulfide bridge. The same 1H NMR methodology used to fully assign the three cysteinyl ligands of the [3Fe–4S] core in the oxidized state (DgFdIIox) was used here for the assignment of the same three ligands in the intermediate state (DgFdIIint). The spin-coupling model used for the oxidized form of DgFdII where magnetic exchange coupling constants of around 300 cm−1 and hyperfine coupling constants equal to 1 MHz for all the three iron centres were found, does not explain the isotropic shift temperature dependence for the three cysteinyl cluster ligands in DgFdIIint. This study, together with the spin delocalization mechanism proposed here for DgFdIIint, allows the detection of structural modifications at the [3Fe-4S] cluster in DgFdIIox and DgFdIIint.
Keywords: Fe3S4 cluster; Ferredoxin; Disulfide bridge; Paramagnetic protein; Desulfovibrio gigas

In order to understand the detailed mechanism of the stereoselective photoinduced electron-transfer (ET) reactions of zinc-substituted myoglobin (ZnMb) with optically active molecules by flash photolysis, we designed and prepared new optically active agents, such as N,N′-dimethylcinchoninium diiodide ([MCN]I2) and N,N′-dimethylcinchonidinium diiodide ([MCD]I2). The photoexcited triplet state of ZnMb, 3(ZnMb)*, was successfully quenched by [MCN]2+ and [MCD]2+ ions to form the radical pair of ZnMb cation (ZnMb·+) and reduced [MCN]·+ and [MCD]·+, followed by a thermal back ET reaction to the ground state. The rate constants (k q) for the ET quenching at 25 °C were obtained as k q(MCN)=(1.9±0.1)×106 M−1 s−1 and k q(MCD)=(3.0±0.2)×106 M−1 s−1, respectively. The ratio of k q(MCD)/k q(MCN)=1.6 indicates that the [MCD]2+ preferentially quenches 3(ZnMb)*. The second-order rate constants (k b) for the thermal back ET reaction from [MCN]·+ and [MCD]·+ to ZnMb·+ at 25 °C were k b(MCN)=(0.79±0.04)×108 M−1 s−1 and k b(MCD)=(1.0±0.1)×108 M−1 s−1, respectively, and the selectivity was k q(MCD)/k q(MCN)=1.3. Both quenching and thermal back ET reactions are controlled by the ET step. In the quenching reaction, the energy differences of ΔΔH (MCD–MCN) and ΔΔS (MCD–MCN) at 25 °C were obtained as −1.1 and 0 kJ mol−1, respectively. On the other hand, ΔΔH (MCD–MCN)=11±2 kJ mol−1 and TΔΔS (MCD–MCN)=−10±2 kJ mol−1 were given in the thermal back ET reaction. The highest stereoselectivity of 1.7 for [MCD]·+ found at low temperature (10 °C) was due to the ΔΔS value obtained in the thermal back ET reaction.
Keywords: Stereoselective photoinduced electron transfer; Zinc myoglobin; Cinchonine; Cinchonidine; Kinetics

Crosstalk between metal ions in Bacillus subtilis ferrochelatase by Mattias D. Hansson; Mats Lindstam; Mats Hansson (325-333).
Ferrochelatase (EC 4.99.1.1), the terminal enzyme in the heme biosynthetic pathway, catalyzes the insertion of Fe2+ into protoporphyrin IX, generating heme. In vitro assays have shown that all characterized ferrochelatases can also incorporate Zn2+ into protoporphyrin IX. Previously Zn2+ has been observed at an inner metal binding site close to the porphyrin binding site. Mg2+, which stimulates Zn2+ insertion by Bacillus subtilis ferrochelatase, has been observed at an outer metal binding site. Exchange of Glu272 to a serine eliminated the stimulative effect of Mg2+. We found that Zn2+ quenched the fluorescence of B. subtilis ferrochelatase and this quenching was used to estimate the metal affinity. Trp230 was identified as the intrinsic fluorophore responsible for the observed quenching pattern. The affinity for Zn2+ could be increased by incubating the ferrochelatase with the transition state analogue N-methyl mesoporphyrin IX, which reflected a close collaborative arrangement between the two substrates in the active site. We also showed that the affinity for Zn2+ was lowered in the presence of Mg2+ and that bound Zn2+ was released upon binding of Mg2+. In the ferrochelatase with a Glu272Ser modification, the interaction between Zn2+ and Mg2+ was abolished. It could thereby be demonstrated that the presence of a metal at one metal binding site affected the metal affinity of another, providing the enzyme with a site that regulates the enzymatic activity.
Keywords: Ferrochelatase; hemH ; Metal interaction; N-Methyl mesoporphyrin IX; Porphyrin metallation

Hydrogen-bonding conformations of tyrosine B10 tailor the hemeprotein reactivity of ferryl species by Walleska De Jesús-Bonilla; Anthony Cruz; Ariel Lewis; José Cerda; Daniel E. Bacelo; Carmen L. Cadilla; Juan López-Garriga (334-342).
Ferryl compounds [Fe(IV)=O] in living organisms play an essential role in the radical catalytic cycle and degradation processes of hemeproteins. We studied the reactions between H2O2 and hemoglobin II (HbII) (GlnE7, TyrB10, PheCD1, PheE11), recombinant hemoglobin I (HbI) (GlnE7, PheB10, PheCD1, PheE11), and the HbI PheB10Tyr mutant of L. pectinata. We found that the tyrosine residue in the B10 position tailors, in two very distinct ways, the reactivity of the ferryl species, compounds I and II. First, increasing the reaction pH from 4.86 to 7.50, and then to 11.2, caused the the second-order rate constant for HbII to decrease from 141.60 to 77.78 M−1 s−1, and to 2.96 M−1 s−1, respectively. This pH dependence is associated with the disruption of the heme–tyrosine (603 nm) protein moiety, which controls the access of the H2O2 to the hemeprotein active center, thus regulating the formation of the ferryl species. Second, the presence of compound I was evident in the UV–vis spectra (648-nm band) in the reactions of HbI and recombinant HbI with H2O2, This band, however, is completely absent in the analogous reaction with HbII and the HbI PheB10Tyr mutant. Therefore, the existence of a hydrogen-bonding network between the heme pocket amino acids (i.e., TyrB10) and the ferryl compound I created a path much faster than 3.0×10−2 s−1 for the decay of compound I to compound II. Furthermore, the decay of the heme ferryl compound I to compound II was independent of the proximal HisF8 trans-ligand strength. Thus, the pH dependence of the heme–tyrosine moiety complex determined the overall reaction rate of the oxidative reaction limiting the interaction with H2O2 at neutral pH. The hydrogen-bonding strength between the TyrB10 and the heme ferryl species suggests the presence of a cycle where the ferryl consumption by the ferric heme increases significantly the pseudoperoxidase activity of these hemeproteins.
Keywords: Tyrosine B10; Ferryl species; Hydrogen peroxide; Lucina pectinata hemoglobins; Proximal HisF8 trans effect

Cluster N1 of complex I from Yarrowia lipolytica studied by pulsed EPR spectroscopy by T. Maly; L. Grgic; K. Zwicker; V. Zickermann; U. Brandt; T. Prisner (343-350).
After reduction with nicotinamide adenine dinucleotide (NADH), NADH:ubiquinone oxidoreductase (complex I) of the strictly aerobic yeast Yarrowia lipolytica shows clear signals from five different paramagnetic iron–sulfur (FeS) clusters (N1–N5) which can be detected using electron paramagnetic resonance (EPR) spectroscopy. The ligand environment and the assignment of several FeS clusters to specific binding motifs found in several subunits of the complex are still under debate. In order to characterize the hyperfine interaction of the surrounding nuclei with FeS cluster N1, one- and two-dimensional electron spin echo envelope modulation experiments were performed at a temperature of 30 K. At this temperature only cluster N1 contributes to the overall signal in a pulsed EPR experiment. The hyperfine and quadrupole tensors of a nitrogen nucleus and the isotropic and dipolar hyperfine couplings of two sets of protons could be determined by numerical simulation of the one- and two-dimensional spectra. The values obtained are in perfect agreement with a ferredoxin-like binding structure by four cysteine amino acid residues and allow the assignment of the nitrogen couplings to a backbone nitrogen nucleus and the proton couplings to the β-protons of the bound cysteine residues.
Keywords: Complex I; Iron–sulfur clusters; Ferredoxins; Electron spin echo envelope modulation; Hyperfine sublevel correlation

Site-selective binding of Zn(II) to metallo-β-lactamase L1 from Stenotrophomonas maltophilia by Alison Costello; Gopalraj Periyannan; Ke-Wu Yang; Michael W. Crowder; David L. Tierney (351-358).
Extended X-ray absorption fine structure studies of the metallo-β-lactamase L1 from Stenotrophomonas maltophilia containing 1 and 2 equiv of Zn(II) and containing 2 equiv of Zn(II) plus hydrolyzed nitrocefin are presented. The data indicate that the first, catalytically dominant metal ion is bound by L1 at the consensus Zn1 site. The data further suggest that binding of the first metal helps preorganize the ligands for binding of the second metal ion. The di-Zn enzyme displays a well-defined metal–metal interaction at 3.42 Å. Reaction with the β-lactam antibiotic nitrocefin results in a product-bound species, in which the ring-opened lactam rotates in the active site to present the S1 sulfur atom of nitrocefin to one of the metal ions for coordination. The product bridges the two metal ions, with a concomitant lengthening of the Zn–Zn interaction to 3.62 Å.
Keywords: Extended X-ray absorption fine structure; Metallo-β-lactamase; L1

Iron nitrosyl complexes as models for biological nitric oxide transfer reagents by Chao-Yi Chiang; Marcetta Y. Darensbourg (359-370).
Owing to the indiscriminate reactivity of the free NO radical, intricate control mechanisms are required for storage, transport and transfer of NO to its various biological targets. Among the proposed storage components are protein-bound thionitrosyls (Rprotein–SNO) and protein-bound dinitrosyl iron complexes. Current knowledge suggests the latter are derived from iron–sulfur cluster degradation in the presence of excess NO. Mobilization of protein-bound NO could involve NO or Fe(NO)2 unit transfer to small serum molecules such as glutathione, free cysteine, or iron-porphyrins. The study reported is of a reaction model which addresses the key steps in NO transfer from a prototypal iron dinitrosyl complex. While the N,N′-bis(2-mercaptoethyl)-N,N′-diazacyclooctane (bme-daco) ligand typically binds in square-planar N2S2 coordination, it also serves as a bidentate dithiolate donor for tetrahedral structures in the preparation of the (H+bme-daco)Fe(NO)2 derivative (Chiang et al., J. Am. Chem. Soc. 126:10867–10874, 2004). The removal of one NO produces the mononitrosyl complex, (bme-daco)Fe(NO), and simplifies studies of NO release mechanisms. We have used heme-type model complexes, Fe or Co porphyrins as NO acceptors, yielding (porphyrin)M(NO), where M is  Fe or Co, and monitored reactions by ν(NO) Fourier transform IR spectroscopy. Reaction products were verified by electrospray ionization mass spectrometry. Rudimentary mechanistic studies suggest a role for HNO in the NO release from the dinitrosyl; the mononitrosyl benefits as well from acid catalysis. Other NO uptake complexes such as [(N2S2)Fe]2 [N2S2 is bme-daco or N,N’-bis(2-mercapto-2-methylpropyl)-daco] are shown to form Fe(NO) mononitrosyls with stability and spectroscopic signatures similar to those of the porphyrins.
Keywords: Iron dinitrosyl; Dinitrosyl iron complex; NO transfer; N2S2 ligand; Biomimetic

Function of the tunnel in acetylcoenzyme A synthase/carbon monoxide dehydrogenase by Xiangshi Tan; Anne Volbeda; Juan C. Fontecilla-Camps; Paul A. Lindahl (371-378).
Acetylcoenzyme A synthase/carbon monoxide dehydrogenase (ACS/CODH) contains two Ni–Fe–S active-site clusters (called A and C) connected by a tunnel through which CO and CO2 migrate. Site-directed mutants A578C, L215F, and A219F were designed to block the tunnel at different points along the region between the two C-clusters. Two other mutant proteins F70W and N101Q were designed to block the region that connects the tunnel at the ββ interface with a water channel also located at that interface. Purified mutant proteins were assayed for Ni/Fe content and examined by electron paramagnetic resonance spectroscopy. Analyses indicate that same metal clusters found in wild-type (WT) ACS/CODH (i.e., the A-, B-, C-, and probably D-clusters) are properly assembled in the mutant enzymes. Stopped-flow kinetics revealed that these centers in the mutants are rapidly reducible by dithionite but are only slowly reducible by CO, suggesting an impaired ability of CO to migrate through the tunnel to the C-cluster. Relative to the WT enzyme, mutant proteins exhibited little CODH or ACS activity (using CO2 as a substrate). Some ACS activity was observed when CO was a substrate, but not the cooperative CO inhibition effect characteristic of WT ACS/CODH. These results suggest that CO and CO2 enter and exit the enzyme at the water channel along the ββ subunit interface. They also suggest two pathways for CO during synthesis of acetylcoenzyme A, including one in which CO enters the enzyme and migrates through the tunnel before binding at the A-cluster, and another in which CO binds the A-cluster directly from the solvent.
Keywords: Nickel; Iron–sulfur clusters; Metabolic channeling

Inhibition of cyclic AMP dependent protein kinase by vanadyl sulfate by Kioumars A. Jelveh; Rachel Zhande; Roger W. Brownsey (379-388).
Vanadium salts influence the activities of a number of mammalian enzymes in vitro but the mechanisms by which low concentrations of vanadium ameliorate the effects of diabetes in vivo remain poorly understood. The hypothesis that vanadium compounds act by inhibiting protein tyrosine phosphatases has attracted most support. The studies described here further evaluate the possibility that vanadyl sulfate trihydrate (VS) can also inhibit 3′,5′-cyclic adenosine monophosphate (cAMP) dependent protein kinase (PKA). Using conventional assay conditions, VS inhibited PKA only at high concentrations (IC50>400 μM); however, PKA inhibition was seen at dramatically lower concentrations of VS (IC50<10 μM) when sequestration of vanadyl ions was minimized. Vanadyl appears to be the effective PKA inhibitor because sodium orthovanadate did not inhibit PKA and inhibition by vanadyl was abolished by potential chelators such as ethylenediaminetetraacetic acid or glycyl peptides. PKA inhibition by vanadyl appears to be mixed rather than strictly competitive or uncompetitive and may replicate the inhibitory effects of high concentrations of Mg2+. The effect of vanadyl on PKA provides a possible explanation for the effects of vanadium salts on fat tissue lipolysis and perhaps on other aspects of energy metabolism that are controlled by cAMP-dependent mechanisms. Considering the high degree of conservation of the active sites of protein kinases, vanadyl may also influence other members of this large protein family.
Keywords: Biomimesis; Enzyme kinetics; Protein phosphorylation; Protein kinases; Metabolism