Site-selective X-ray spectroscopy discriminated the cubane and diiron units in the H-cluster of [FeFe]-hydrogenase revealing its electronic and structural configurations.

[FeFe]-hydrogenase (H 2 ase) 4 proteins are the most active biological catalysts for the production of molecular hydrogen (H 2 ) from proton reduction, with reported turnover rates of up to 10 4 s Ϫ1 (1, 2). Therefore, these enzymes are of high interest for biotechnology, aiming at the generation of H 2 as a renewable fuel (1, 3-5). However, a severe limitation for such applications is the rapid inactivation of [FeFe]-H 2 ases by dioxygen (O 2 ) (6, 7). Understanding the mechanism of O 2 -induced inactivation may allow the improvement of enzyme features to yield increased O 2 tolerance (e.g. by genetic engineering, which has already been demonstrated for NiFe hydrogenases) (8 -10).[FeFe]-H 2 ases are found in certain bacteria and green algae (11, 12). All of these enzymes contain an active site that consists of an inorganic iron complex, which is denoted as the H-cluster (13-15). The [FeFe]-H 2 ase enzymes from anaerobic bacteria in addition bind several iron-sulfur (FeS) clusters, serving as a relay for electron transfer to and from the active site (13-15).[FeFe]-H 2 ases from green algae represent the minimal unit for biological H 2 production because they contain only the H-cluster, whereas accessory FeS clusters are absent (16). This feature renders these enzymes most suitable for spectroscopic investigations on O 2 -induced inactivation (17), focusing on the active site reactions. The general structure of the H-cluster has been unraveled by crystallography, x-ray absorption spectroscopy (XAS), EPR, and FTIR spectroscopy on H 2 ase protein from various organisms (18). The H-cluster structure in both bacteria and green algae appears to be similar overall (16,19,20). It features a [4Fe4S] cubane cluster, which is bound by four cysteine residues to the protein and is linked by one of them to a binuclear iron unit (2Fe H ) (Fig. 1). The latter carries two cyanide (CN Ϫ ) ligands and three carbon monoxide (CO) ligands (21, 22) and presumably an azadithiolate group (adt; (SCH 2 ) 2 NH) in the metal-bridging position (21,23,24). Both types of diatomic ligands are probably derived from a biosynthetic pathway starting with tyrosine (25-27). The nitrogen atom of the adt has been proposed to be involved in proton transfer at the active site (21, 28). The H-cluster structure is assembled in a complex reaction involving three maturation proteins (29 -34). H 2 formation has been proposed to involve the binding and reduction of a single proton, which transiently creates a hydride ligand, either located in a bridging position between the two iron ions or terminally bound at the distal iron ion of the 2Fe H moiety (Fig. 1). After a second protonation step, H 2 is released from the H-cluster (35)(36)(37) 4 The abbreviations used are: H 2 ase, hydrogenase; adt, azadithiolate; EXAFS, extended x-ray absorption fine structure; NaDT, sodium dithionite; ROS, reactive oxygen species; XANES, x-ray absorption near edge structure; XAS, x-ray absorption spectroscopy; FT, Fourier transform.

[FeFe]-hydrogenase from green algae (HydA1) is the most efficient hydrogen (H2) producing enzyme in nature and of prime interest for (bio)technology. Its active site is a unique six-iron center (H-cluster) composed of a cubane cluster, [4Fe4S]H, cysteine-linked to a diiron unit, [2Fe]H, which carries unusual carbon monoxide (CO) and cyanide ligands and a bridging azadithiolate group. We have probed the molecular and electronic configurations of the H-cluster in functional oxidized, reduced, and super-reduced or CO-inhibited HydA1 protein, in particular searching for intermediates with iron-hydride bonds. Site-selective X-ray absorption and emission spectroscopy were used to distinguish between low- and high-spin iron sites in the two subcomplexes of the H-cluster. The experimental methods and spectral simulations were calibrated using synthetic model complexes with ligand variations and bound hydride species. Distinct X-ray spectroscopic signatures of electronic excitation or decay transitions in [4Fe4S]H and [2Fe]H were obtained, which were quantitatively reproduced by density functional theory calculations, thereby leading to specific H-cluster model structures. We show that iron-hydride bonds are absent in the reduced state, whereas only in the super-reduced state, ligand rotation facilitates hydride binding presumably to the Fe-Fe bridging position at [2Fe]H. These results are in agreement with a catalytic cycle involving three main intermediates and at least two protonation and electron transfer steps prior to the H2 formation chemistry in [FeFe]-hydrogenases.

No abstract

The active site for hydrogen production in [FeFe] hydrogenase comprises a diiron unit. Bioinorganic chemistry has modeled important features of this center, aiming at mechanistic understanding and the development of novel catalysts. However, new assays are required for analyzing the effects of ligand variations at the metal ions. By high-resolution X-ray absorption spectroscopy with narrow-band X-ray emission detection (XAS/XES = XAES) and density functional theory (DFT), we studied an asymmetrically coordinated [FeFe] model complex, [(CO)(3)Fe(I)1-(bdtCl(2))-Fe(I)2(CO)(Ph(2)P-CH(2)-NCH(3)-CH(2)-PPh(2))] (1, bdt = benzene-1,2-dithiolate), in comparison to iron-carbonyl references. Kβ emission spectra (Kβ(1,3), Kβ') revealed the absence of unpaired spins and the low-spin character for both Fe ions in 1. In a series of low-spin iron compounds, the Kβ(1,3) energy did not reflect the formal iron oxidation state, but it decreases with increasing ligand field strength due to shorter iron-ligand bonds, following the spectrochemical series. The intensity of the valence-to-core transitions (Kβ(2,5)) decreases for increasing Fe-ligand bond length, certain emission peaks allow counting of Fe-CO bonds, and even molecular orbitals (MOs) located on the metal-bridging bdt group of 1 contribute to the spectra. As deduced from 3d → 1s emission and 1s → 3d absorption spectra and supported by DFT, the HOMO-LUMO gap of 1 is about 2.8 eV. Kβ-detected XANES spectra in agreement with DFT revealed considerable electronic asymmetry in 1; the energies and occupancies of Fe-d dominated MOs resemble a square-pyramidal Fe(0) for Fe1 and an octahedral Fe(II) for Fe2. EXAFS spectra for various Kβ emission energies showed considerable site-selectivity; approximate structural parameters similar to the crystal structure could be determined for the two individual iron atoms of 1 in powder samples. These results suggest that metal site- and spin-selective XAES on [FeFe] hydrogenase protein and active site models may provide a powerful tool to study intermediates under reaction conditions.

High-resolution X-ray absorption spectroscopy with narrow-band X-ray emission detection, supported by density functional theory calculations (XAES-DFT), was used to study a model complex, ([Fe(2)(μ-adt)(CO)(4)(PMe(3))(2)] (1, adt = S-CH(2)-(NCH(2)Ph)-CH(2)-S), of the [FeFe] hydrogenase active site. For 1 in powder material (1(powder)), in MeCN solution (1'), and in its three protonated states (1H, 1Hy, 1HHy; H denotes protonation at the adt-N and Hy protonation of the Fe-Fe bond to form a bridging metal hydride), relations between the molecular structures and the electronic configurations were determined. EXAFS analysis and DFT geometry optimization suggested prevailing rotational isomers in MeCN, which were similar to the crystal structure or exhibited rotation of the (CO) ligands at Fe1 (1(CO), 1Hy(CO)) and in addition of the phenyl ring (1H(CO,Ph), 1HHy(CO,Ph)), leading to an elongated solvent-exposed Fe-Fe bond. Isomer formation, adt-N protonation, and hydride binding caused spectral changes of core-to-valence (pre-edge of the Fe K-shell absorption) and of valence-to-core (Kß(2,5) emission) electronic transitions, and of Kα RIXS data, which were quantitatively reproduced by DFT. The study reveals (1) the composition of molecular orbitals, for example, with dominant Fe-d character, showing variations in symmetry and apparent oxidation state at the two Fe ions and a drop in MO energies by ~1 eV upon each protonation step, (2) the HOMO-LUMO energy gaps, of ~2.3 eV for 1(powder) and ~2.0 eV for 1', and (3) the splitting between iron d(z(2)) and d(x(2)-y(2)) levels of ~0.5 eV for the nonhydride and ~0.9 eV for the hydride states. Good correlations of reduction potentials to LUMO energies and oxidation potentials to HOMO energies were obtained. Two routes of facilitated bridging hydride binding thereby are suggested, involving ligand rotation at Fe1 for 1Hy(CO) or adt-N protonation for 1HHy(CO,Ph). XAES-DFT thus enables verification of the effects of ligand substitutions in solution for guided improvement of [FeFe] catalysts.

The geometry and electronic structure of cis-[Ru(II)(bpy)(2)(H(2)O)(2)](2+) and its higher oxidation state species up formally to Ru(VI) have been studied by means of UV-vis, EPR, XAS, and DFT and CASSCF/CASPT2 calculations. DFT calculations of the molecular structures of these species show that, as the oxidation state increases, the Ru-O bond distance decreases, indicating increased degrees of Ru-O multiple bonding. In addition, the O-Ru-O valence bond angle increases as the oxidation state increases. EPR spectroscopy and quantum chemical calculations indicate that low-spin configurations are favored for all oxidation states. Thus, cis-[Ru(IV)(bpy)(2)(OH)(2)](2+) (d(4)) has a singlet ground state and is EPR-silent at low temperatures, while cis-[Ru(V)(bpy)(2)(O)(OH)](2+) (d(3)) has a doublet ground state. XAS spectroscopy of higher oxidation state species and DFT calculations further illuminate the electronic structures of these complexes, particularly with respect to the covalent character of the O-Ru-O fragment. In addition, the photochemical isomerization of cis-[Ru(II)(bpy)(2)(H(2)O)(2)](2+) to its trans-[Ru(II)(bpy)(2)(H(2)O)(2)](2+) isomer has been fully characterized through quantum chemical calculations. The excited-state process is predicted to involve decoordination of one aqua ligand, which leads to a coordinatively unsaturated complex that undergoes structural rearrangement followed by recoordination of water to yield the trans isomer.

Background: Some molybdoenzymes in prokaryotes contain the bis-molybdopterin guanine dinucleotide cofactor. Results: The bis-Mo-MPT cofactor is a novel intermediate in Moco biosynthesis in E. coli. Conclusion: Bis-MGD formed by MobA is fully functional and restores the catalytic activity in apoTorA. Significance: Bis-Mo-MPT assembles spontaneously on MobA prior to forming bis-MGD.