Desulfovibrio gigas formate dehydrogenase is the first representative of a tungsten-containing enzyme from a mesophile that has been structurally characterized. It is a heterodimer of 110 and 24 kDa subunits. The large subunit, homologous to E. coli FDH-H and to D. desulfuricans nitrate reductase, harbors the W site and one [4Fe-4S] center. No small subunit ortholog containing three [4Fe-4S] clusters has been reported. The structural homology with E. coli FDH-H shows that the essential residues (SeCys158, His159, and Arg407) at the active site are conserved. The active site is accessible via a positively charged tunnel, while product release may be facilitated, for H(+) by buried waters and protonable amino acids and for CO(2) through a hydrophobic channel.

Desulfovibrio gigas desulforedoxin (Dx) consists of two identical peptides, each containing one [Fe-4S] center per monomer. Variants with different iron and zinc metal compositions arise when desulforedoxin is produced recombinantly from Escherichia coli. The three forms of the protein, the two homodimers [Fe(III)/Fe(III)]Dx and [Zn(II)/Zn(II)]Dx, and the heterodimer [Fe(III)/Zn(II)]Dx, can be separated by ion exchange chromatography on the basis of their charge differences. Once separated, the desulforedoxins containing iron can be reduced with added dithionite. For NMR studies, different protein samples were prepared labeled with (15)N or (15)N + (13)C. Spectral assignments were determined for [Fe(II)/Fe(II)]Dx and [Fe(II)/Zn(II)]Dx from 3D (15)N TOCSY-HSQC and NOESY-HSQC data, and compared with those reported previously for [Zn(II)/Zn(II)]Dx. Assignments for the (13)C(alpha) shifts were obtained from an HNCA experiment. Comparison of (1)H-(15)N HSQC spectra of [Zn(II)/Zn(II)]Dx, [Fe(II)/Fe(II)]Dx and [Fe(II)/Zn(II)]Dx revealed that the pseudocontact shifts in [Fe(II)/Zn(II)]Dx can be decomposed into inter- and intramonomer components, which, when summed, accurately predict the observed pseudocontact shifts observed for [Fe(II)/Fe(II)]Dx. The degree of linearity observed in the pseudocontact shifts for residues >/=8.5 A from the metal center indicates that the replacement of Fe(II) by Zn(II) produces little or no change in the structure of Dx. The results suggest a general strategy for the analysis of NMR spectra of homo-oligomeric proteins in which a paramagnetic center introduced into a single subunit is used to break the magnetic symmetry and make it possible to obtain distance constraints (both pseudocontact and NOE) between subunits.

Using a myosin preparation containing endogenous myosin light-chain (LC2) kinase and phosphatase and calmodulin, i.e. near physiological ones, the interaction of vanadate oligomers with phosphorylated myosin was evaluated. Decavanadate or metavanadate solutions (2-15 mM total vanadate) did not prevent the phosphorylation state of the regulatory myosin lightchain, as observed by urea-polyacrylamide gel electrophoresis. The relative order of line broadening upon protein addition, reflecting the interaction of the vanadate oligomers with phosphorylated myosin, was V10 > V-4 > V-1 = 1 whereas, no changes were observed for monomeric vanadate. In the presence of ATP, V-1 signal was shifted upfield 2 ppm and became broadened, while V4 signal became narrowed. Moreover, a significant increase in myosin ATPase inhibition (60%) was observed when decameric vanadate species were present (1.4 mM). It is concluded that, under conditions near physiological ones, decameric vanadate differs from vanadate oligomers present in metavanadate solutions due to its strong interaction with the phosphorylated enzyme and myosin ATPase inhibition. Besides, ATP decreases the affinity of myosin for tetravanadate, induces the interaction with monomeric vanadate, whereas it does not affect decameric vanadate interaction. (C) 2002 Elsevier Science B.V. All rights reserved.

The accumulation of toxic bile salts within the hepatocyte plays a key role in organ injury during liver disease. Deoxycholate (DC) and glycochenodeoxycholate (GCDC) induce apoptosis in vitro and in vivo, perhaps through direct perturbation of mitochondrial membrane structure and function. In contrast, ursodeoxycholate (UDC) and its taurine-conjugated form (TUDC) appear to be protective. We show here that hydrophobic bile salts induced apoptosis in cultured rat hepatocytes, without modulating the expression of pro-apoptotic Bax protein, and caused cytochrome c release in isolated mitochondria. Co-incubation with UDC and TUDC prevented cell death and efflux of mitochondrial factors. Using spin-labelling techniques and EPR spectroscopy analysis of isolated rat liver mitochondria, we found significant structural changes at the membrane-water surface in mitochondria exposed to hydrophobic bile salts, including modified lipid polarity and fluidity, altered protein order and increased oxidative injury. UDC, TUDC and cyclosporin A almost completely abrogated DC- and GCDC-induced membrane perturbations. We conclude that the toxicity of hydrophobic bile salts to hepatocytes is mediated by cytochrome c release, through a mechanism associated with marked direct effects on mitochondrial membrane lipid polarity and fluidity, protein order and redox status, without modulation of pro-apoptotic Bax expression. UDC and TUDC can directly suppress disruption of mitochondrial membrane structure, which may represent an important mechanism of hepatoprotection by these bile salts.

BACKGROUND/AIMS: Unconjugated bilirubin (UCB) impairs crucial aspects of cell function and induces apoptosis in primary cultured neurones. While mechanisms of cytotoxicity begin to unfold, mitochondria appear as potential primary targets. METHODS: We used electron paramagnetic resonance spectroscopy analysis of isolated rat mitochondria to test the hypothesis that UCB physically interacts with mitochondria to induce structural membrane perturbation, leading to increased permeability, and subsequent release of apoptotic factors. RESULTS: Our data demonstrate profound changes on mitochondrial membrane properties during incubation with UCB, including modified membrane lipid polarity and fluidity (P<0.01), as well as disrupted protein mobility (P<0.001). Consistent with increased permeability, cytochrome c was released from the intermembrane space (P<0.01), perhaps uncoupling the respiratory chain and further increasing oxidative stress (P<0.01). Both ursodeoxycholate, a mitochondrial-membrane stabilising agent, and cyclosporine A, an inhibitor of the permeability transition, almost completely abrogated UCB-induced perturbation. CONCLUSIONS: UCB directly interacts with mitochondria influencing membrane lipid and protein properties, redox status, and cytochrome c content. Thus, apoptosis induced by UCB may be mediated, at least in part, by physical perturbation of the mitochondrial membrane. These novel findings should ultimately prove useful to our evolving understanding of UCB cytotoxicity.

Increased levels of unconjugated bilirubin, the end product of heme catabolism, impair crucial aspects of nerve cell function. In previous studies, we demonstrated that bilirubin toxicity may be due to cell death by apoptosis. To characterize the sequence of events leading to neurotoxicity, we exposed developing rat brain astrocytes and neurons to unconjugated bilirubin and investigated whether changes in membrane dynamic properties can mediate apoptosis. Bilirubin induced a rapid, dose-dependent increase in apoptosis, which was nevertheless preceded by impaired mitochondrial metabolism. Using spin labels and electron paramagnetic resonance spectroscopy analysis of whole cell and isolated mitochondrial membranes exposed to bilirubin, we detected major membrane perturbation. By physically interacting with cell membranes, bilirubin induced an almost immediate increase in lipid polarity sensed at a superficial level. The enhanced membrane permeability coincided with an increase in lipid fluidity and protein mobility and was associated with significant oxidative injury to membrane lipids. In conclusion, apoptosis of nerve cells induced by bilirubin is mediated by its primary effect at physically perturbing the cell membrane. Bilirubin directly interacts with membranes influencing lipid polarity and fluidity, protein order, and redox status. These data suggest that nerve cell membranes are primary targets of bilirubin toxicity.

Resonance Raman (RR) spectroscopy is used to examine porphyrin substrate, product, and inhibitor interactions with the active site of murine ferrochelatase (EC 4.99.1.1), the terminal enzyme in the biosynthesis of heme. The enzyme catalyzes in vivo Fe2+ chelation into protoporphyrin IX to give heme. The RR spectra of native ferrochelatase show that the protein, as isolated, contains varying amounts of endogenously bound high- or low-spin ferric heme, always at much less than 1 equiv. RR data on the binding of free-base protoporphyrin IX and its metalated complexes (Fe(III), Fe(II), and Ni(II)) to active wild-type protein were obtained at varying ratios of porphyrin to protein. The binding of ferric heme, a known inhibitor of the enzyme, leads to the formation of a low-spin six-coordinate adduct. Ferrous heme, the enzyme's natural product, binds in the ferrous high-spin five-coordinate state. Ni(II) protoporphyrin, a metalloporphyrin that has a low tendency toward axial ligation, becomes distorted when bound to ferrochelatase. Similarly for free-base protoporphyrin, the natural substrate of ferrochelatase, the RR spectra of porphyrin-protein complexes reveal a saddling distortion of the porphyrin. These results corroborate and extend our previous findings that porphyrin distortion, a crucial step of the catalytic mechanism, occurs even in the absence of bound metal substrate. Moreover, RR data reveal the presence of an amino acid residue in the active site of ferrochelatase which is capable of specific axial ligation to metals.

Crystallographic studies of the hydrogenases (Hases) from Desulfovibrio gigas (Dg) and Desulfovibrio vulgaris Miyazaki (DvM) have revealed heterodinuclear nickel-iron active centers in both enzymes. The structures, which represent the as-isolated (unready) Ni-A (S = (1)/(2)) enzyme state, disclose a nonprotein ligand (labeled as X) bridging the two metals. The bridging atom was suggested to be an oxygenic (O(2)(-) or OH(-)) species in Dg Hase and an inorganic sulfide in DvM Hase. To determine the nature and chemical characteristics of the Ni-X-Fe bridging ligand in Dg Hase, we have performed 35 GHz CW (17)O ENDOR measurements on the Ni-A form of the enzyme, exchanged into H(2)(17)O, on the active Ni-C (S = (1)/(2)) form prepared by H(2)-reduction of Ni-A in H(2)(17)O, and also on Ni-A formed by reoxidation of Ni-C in H(2)(17)O. In the native state of the protein (Ni-A), the bridging ligand does not exchange with the H(2)(17)O solvent. However, after a reduction/reoxidation cycle (Ni-A --> Ni-C --> Ni-A), an (17)O label is introduced at the active site, as seen by ENDOR. Detailed analysis of a 2-D field-frequency plot of ENDOR spectra taken across the EPR envelope of Ni-A((17)O) shows that the incorporated (17)O has a roughly axial hyperfine tensor, A((17)O) approximately [5, 7, 20] MHz, discloses its orientation relative to the g tensor, and also yields an estimate of the quadrupole tensor. The substantial isotropic component (a(iso)((17)O) approximately 11 MHz) of the hyperfine interaction indicates that a solvent-derived (17)O is indeed a ligand to Ni and thus that the bridging ligand X in the Ni-A state of Dg Hase is indeed an oxygenic (O(2)(-) or OH(-)) species; comparison with earlier EPR results by others indicates that the same holds for Ni-B. The small (57)Fe hyperfine coupling seen previously for Ni-A (A((57)Fe) approximately 0.9 MHz) is now shown to persist in Ni-C, A((57)Fe) approximately 0.8 MHz. However, the (17)O signal is lost upon reductive activation to the Ni-C state; reoxidation to Ni-A leads to the reappearance of the signal. Consideration of the electronic structure of the EPR-active states of the dinuclear center leads us to suggest that the oxygenic bridge in Ni-A(B) is lost in Ni-C and is re-formed from solvent upon reoxidation to Ni-A. This implies that the reductive activation to Ni-C opens Ni/Fe coordination sites which may play a central role in the enzyme's activity.

Spectroscopy coupled with density functional calculations has been used to define the spin state, oxidation states, spin distribution, and ground state wave function of the mu4-sulfide bridged tetranuclear CuZ cluster of nitrous oxide reductase. Initial insight into the electronic contribution to N2O reduction is developed, which involves a sigma superexchange pathway through the bridging sulfide.

This article aims to study hydrogen production/consumption in Desulfovibrio (D.) desulfuricans strain New Jersey, a sulfate reducer isolated from a medium undergoing active biocorrosion and to compare its hydrogen metabolism with two other Desulfovibrio species, D. gigas and D. vulgaris Hildenborough. Hydrogen production was followed during the growth of these three bacterial species under different growth conditions: no limitation of sulfate and lactate, sulfate limitation, lactate limitation, pyruvate/sulfate medium and in the presence of molybdate. Hydrogen production/consumption by D. desulfuricans shows a behavior similar to that of D. gigas but a different one from that of D. vulgaris, which produces higher quantities of hydrogen on lactate/sulfate medium. The three species are able to increase the hydrogen production when the sulfate became limiting. Moreover, in a pyruvate/sulfate medium hydrogen production was lower than on lactate/sulfate medium. Hydrogen production by D. desulfuricans in presence of molybdate is extremely high. Hydrogenases are key enzymes on production/consumption of hydrogen in sulfate reducing organisms. The specific activity, number and cellular localization of hydrogenases vary within the three Desulfovibrio species used in this work, which could explain the differences observed on hydrogen utilization.

Superoxide reductases have now been well characterized from several organisms. Unique biochemical features include the ability of the reduced enzyme to react with O2- but not dioxygen (reduced SORs are stable in an aerobic atmosphere for hours). Future biochemical assays that measure the reaction of SOR with O2- should take into account the difficulties of assaying O2- directly and the myriad of redox reactions that can take place between components in the assay, for example, direct electron transfer between cytochrome c and Dfx. Future prospects include further delineation of the reaction mechanisms, characterization of the putative (hydro)peroxo intermediate, and studies that uncover the components between reduced pyridine nucleotides and SOR in the metabolic pathway responsible for O2- detoxification.

The sulfate-reducing bacterium aldehyde oxidoreductase from Desulfovibrio gigas (MOP) is a member of the xanthine oxidase family of enzymes. It has 907 residues on a single polypeptide chain, a molybdopterin cytosine dinucleotide (MCD) cofactor and two [2Fe-2S] iron-sulfur clusters. Synchrotron data to almost atomic resolution were collected for improved cryo-cooled crystals of this enzyme in the oxidized form. The cell constants of a=b=141.78 A and c=160.87 A are about 2% shorter than those of room temperature data, yielding 233,755 unique reflections in space group P6(1)22, at 1.28 A resolution. Throughout the entire refinement the full gradient least-squares method was used, leading to a final R factor of 14.5 and Rfree factor of 19.3 (4sigma cut-off) with "riding" H-atoms at their calculated positions. The model contains 8146 non-hydrogen atoms described by anisotropic displacement parameters with an observations/parameters ratio of 4.4. It includes alternate conformations for 17 amino acid residues. At 1.28 A resolution, three Cl- and two Mg2+ ions from the crystallization solution were clearly identified. With the exception of one Cl- which is buried and 8 A distant from the Mo atom, the other ions are close to the molecular surface and may contribute to crystal packing. The overall structure has not changed in comparison to the lower resolution model apart from local corrections that included some loop adjustments and alternate side-chain conformations. Based on the estimated errors of bond distances obtained by blocked least-squares matrix inversion, a more detailed analysis of the three redox centres was possible. For the MCD cofactor, the resulting geometric parameters confirmed its reduction state as a tetrahydropterin. At the Mo centre, estimated corrections calculated for the Fourier ripples artefact are very small when compared to the experimental associated errors, supporting the suggestion that the fifth ligand is a water molecule rather than a hydroxide. Concerning the two iron-sulfur centres, asymmetry in the Fe-S distances as well as differences in the pattern of NH.S hydrogen-bonding interactions was observed, which influences the electron distribution upon reduction and causes non-equivalence of the individual Fe atoms in each cluster.

Dg ferredoxin gene was cloned using the polymerase chain reaction (PCR), inserted into vector pT7-7, and overexpressed in Escherichia coli (E. coli) grown in aerobic media. The recombinant protein is a dimer and contains a [3Fe-4S] cluster per monomer. EPR and (1)H NMR data of recombinant and wild-type protein are compared.

Ferrochelatase (EC 4.99.1.1) is the terminal enzyme of the haem biosynthetic pathway and catalyses iron chelation into the protoporphyrin IX ring. Glutamate-287 (E287) of murine mature ferrochelatase is a conserved residue in all known sequences of ferrochelatase, is present at the active site of the enzyme, as inferred from the Bacillus subtilis ferrochelatase three-dimensional structure, and is critical for enzyme activity. Substitution of E287 with either glutamine (Q) or alanine (A) yielded variants with lower enzymic activity than that of the wild-type ferrochelatase and with different absorption spectra from the wild-type enzyme. In contrast to the wild-type enzyme, the absorption spectra of the variants indicate that these enzymes, as purified, contain protoporphyrin IX. Identification and quantification of the porphyrin bound to the E287-directed variants indicate that approx. 80% of the total porphyrin corresponds to protoporphyrin IX. Significantly, rapid stopped-flow experiments of the E287A and E287Q Variants demonstrate that reaction with Zn2+ results in the formation of bound Zn-protoporphyrin IX, indicating that the endogenously bound protoporphyrin IX can be used as a substrate. Taken together, these findings suggest that the structural strain imposed by ferrochelatase on the porphyrin substrate as a critical step in the enzyme catalytic mechanism is also accomplished by the E287A and E287Q variants, but without the release of the product. Thus E287 in murine ferrochelatase appears to be critical For the catalytic process by controlling the release of the product.

The periplasmic hydrogenase of Desulfovibrio vulgaris (Hildenbourough) is an all Fe-containing hydrogenase. It contains two ferredoxin type [4Fe-4S] clusters, termed the F clusters, and a catalytic H cluster. Recent X-ray crystallographic studies on two Fe hydrogenases revealed that the H cluster is composed of two sub-clusters, a [4Fe-4S] cluster ([4Fe-4S](H)) and a binuclear Fe cluster ([2Fe](H)), bridged by a cysteine sulfur. The aerobically purified D. vulgaris hydrogenase is stable in air. It is inactive and requires reductive activation. Upon reduction, the enzyme becomes sensitive to O(2), indicating that the reductive activation process is irreversible. Previous EPR investigations showed that upon reoxidation (under argon) the H cluster exhibits a rhombic EPR signal that is not seen in the as-purified enzyme, suggesting a conformational change in association with the reductive activation. For the purpose of gaining more information on the electronic properties of this unique H cluster and to understand further the reductive activation process, variable-temperature and variable-field Mossbauer spectroscopy has been used to characterize the Fe-S clusters in D. vulgaris hydrogenase poised at different redox states generated during a reductive titration, and in the CO-reacted enzyme. The data were successfully decomposed into spectral components corresponding to the F and H clusters, and characteristic parameters describing the electronic and magnetic properties of the F and H clusters were obtained. Consistent with the X-ray crystallographic results, the spectra of the H cluster can be understood as originating from an exchange coupled [4Fe-4S]-[2Fe] system. In particular, detailed analysis of the data reveals that the reductive activation begins with reduction of the [4Fe-4S](H) cluster from the 2+ to the 1+ state, followed by transfer of the reducing equivalent from the [4Fe-4S](H) subcluster to the binuclear [2Fe](H) subcluster. The results also reveal that binding of exogenous CO to the H cluster affects significantly the exchange coupling between the [4Fe-4S](H) and the [2Fe](H) subclusters. Implication of such a CO binding effect is discussed.

Unconjugated bilirubin is a neurotoxic pigment that interacts with membrane lipids. In this study we used electron paramagnetic resonance and the spin labels 5-, 7-, 12-, and 16-doxyl-stearic acid (DSA) to evaluate the depth of the hydrocarbon chain at which interaction of bilirubin preferentially occurs. In addition, we used different pH values to determine the molecular species involved. Resealed right-side-out ghosts were incubated (1-60 min) with bilirubin (3.4-42.8 microM) at pH 7.0, 7.4, and 8.0. Alterations of membrane dynamic properties were maximum after 15 min of incubation with 8.6 microM bilirubin at pH 7.4 and were accompanied by a significant release of phospholipids. Interestingly, concentrations of bilirubin up to 42.8 microM and longer incubations resulted in the elution of cholesterol and further increased that of phospholipids while inducing less structural alterations. Variation of the pH values from 8.0 to 7.4 and 7.0, under conditions of maximum perturbation, led to a change from an increased to a diminished polarity sensed by 5-DSA. Conversely, a progressive enhancement in fluidity was reported by 7-DSA, followed by 12- and 16-DSA. These results indicate that bilirubin while enhancing membrane lipid order at C-5 simultaneously has disordering effects at C-7. Furthermore, recovery of membrane dynamics after 15 min of bilirubin exposure along with the release of lipids is compatible with a membrane adaptive response to the insult. In addition, our data provide evidence that uncharged diacid is the species primarily interacting with the membrane as perturbation is favored by acidosis, a condition frequently associated with hyperbilirubinemia in premature and severely ill infants.

The structural changes in the heme macrocycle and substituents caused by binding of Ca2+ to the diheme cytochrome c peroxidase from Paracoccus pantotrophus were clarified by resonance Raman spectroscopy of the inactive fully oxidized form of the enzyme. The changes in the macrocycle vibrational modes are consistent with a Ca2+-dependent increase in the out-of-plane distortion of the low-potential heme, the proposed peroxidatic heme. Most of the increase in out-of-plane distortion occurs when the high-affinity site I is occupied, but a small further increase in distortion occurs when site II is also occupied by Ca2+ or Mg2+. This increase in the heme distortion explains the red shift in the Soret absorption band that occurs upon Ca2+ binding. Changes also occur in the low-frequency substituent modes of the heme, indicating that a structural change in the covalently attached fingerprint pentapeptide of the LP heme occurs upon Ca2+ binding to site I. These structural changes may lead to loss of the sixth ligand at the peroxidatic heme in the semireduced form of the enzyme and activation.

The intermolecular electron transfer kinetics between nitrite reductase (NiR, cytochrome cd1) isolated from Pseudomonas nautica and three cytochromes c isolated from the same strain, as well as the intramolecular electron transfer between NiR heme c and NiR heme d1, were investigated by cyclic voltammetry. All cytochromes (cytochrome c552, cytochrome c553 and cytochrome C553(548)) exhibited well-behaved electrochemistry. The individual diffusion coefficients and mid-point redox potentials were determined. Under the experimental conditions, only cytochrome c552 established a rapid electron transfer with NiR. At acidic pH, the intermolecular electron transfer (cytochrome c(552red)-->NiR heme cox) is a second-order reaction with a rate constant (k2) of 4.1+/-0.1x10(5) M(-1) s(-1) (pH=6.3 and 100 mM NaCl). Under these conditions, the intermolecular reaction represents the rate-limiting step. A minimum estimate of 33 s(-1) could be determined for the first-order rate constant (k1) of the intramolecular electron transfer reaction NiR heme c(red)-->NiR heme d1ox. The pH dependence of k2 values was investigated at pH values ranging from 5.8 to 8.0. When the pH is progressively shifted towards basic values, the rate constant of the intramolecular electron transfer reaction NiR heme c(red)-->NiR heme d1ox decreases gradually to a point where it becomes rate limiting. At pH 8.0 we determined a value of 1.4+/-0.7 s(-1), corresponding to a k2 value of 2.2+/-1.1x10(4) M(-1) s(-1) for the intermolecular step. The physiological relevance of these results is discussed with a particular emphasis on the proposed mechanism of "dead-end product" formation.

Mitochondria have been implicated in the cytotoxicity of amyloid beta-peptide (A beta), which accumulates as senile plaques in the brain of Alzheimer's disease patients. Tauroursodeoxycholate (TUDC) modulates cell death, in part, by preventing mitochondrial membrane perturbation. Using electron paramagnetic resonance spectroscopy analysis of isolated mitochondria, we tested the hypothesis that A beta acts locally in mitochondrial membranes to induce oxidative injury, leading to increased membrane permeability and subsequent release of caspase-activating factors. Further, we intended to determine the role of TUDC at preventing A beta-induced mitochondrial membrane dysfunction. The results demonstrate oxidative injury of mitochondrial membranes during exposure to A beta and reveal profound structural changes, including modified membrane lipid polarity and disrupted protein mobility. Cytochrome c is released from the intermembrane space of mitochondria as a consequence of increased membrane permeability. TUDC, but not cyclosporine A, almost completely abrogated A beta-induced perturbation of mitochondrial membrane structure. We conclude that A beta directly induces cytochrome c release from mitochondria through a mechanism that is accompanied by profound effects on mitochondrial membrane redox status, lipid polarity, and protein order. TUDC can directly suppress A beta-induced disruption of the mitochondrial membrane structure, suggesting a neuroprotective role for this bile salt.

The reduction of nitrate to nitrogen gas via nitrite, nitric oxide and nitrous oxide is the metabolic pathway usually known as denitrification, a key step in the nitrogen cycle. As observed for other elemental cycles, a battery of enzymes are utilized, namely the reductases for nitrate, nitrite, nitric oxide and nitrous oxide, as well as multiple electron donors that interact with these enzymes, in order to carry out the stepwise reactions that involve key intermediates. Because of the importance of this pathway (of parallel importance to the nitrogen-fixation pathway), efforts are underway to understand the structures of the participating enzymes and to uncover mechanistic aspects. Three-dimensional structures have been solved for the majority of these enzymes in the past few years, revealing the architecture of the active metal sites as well as global structural aspects, and possible mechanistic aspects. In addition, the recognition of specific electron-transfer partners raises important questions regarding specific electron-transfer pathways, partner recognition and control of metabolism.

The tungsten-containing formate dehydrogenase (W-FDH) isolated from Desulfovibrio gigas has been crystallized in space group P2(1), with cell parameters a = 73.8 A, b = 111.3 A, c = 156.6 A and beta = 93.7 degrees. These crystals diffract to beyond 2.0 A on a synchrotron radiation source. W-FDH is a heterodimer (92 kDa and 29 kDa subunits) and two W-FDH molecules are present in the asymmetric unit. Although a molecular replacement solution was found using the periplasmic nitrate reductase as a search model, additional phasing information was needed. A multiple-wavelength anomalous dispersion (MAD) dataset was collected at the W- and Fe-edges, at four different wavelengths. Anomalous and dispersive difference data allowed us to unambiguously identify the metal atoms bound to W-FDH as one W atom with a Se-cysteine ligand as well as one [4Fe-4S] cluster in the 92 kDa subunit, and three additional [4Fe-4S] centers in the smaller 29 kDa subunit. The D. gigas W-FDH was previously characterized based on metal analysis and spectroscopic data. One W atom was predicted to be bound to two molybdopterin guanine dinucleotide (MGD) pterin cofactors and two [4Fe-4S] centers were proposed to be present. The crystallographic data now reported reveal a selenium atom (as a Se-cysteine) coordinating to the W site, as well as two extra [4Fe-4S] clusters not anticipated before. The EPR data were re-evaluated in the light of these new results.