Structure-function relationships in [FeFe]-hydrogenase active site maturation

Since the discovery that, despite the active site complexity, only three gene products suffice to obtain active recombinant [FeFe]-hydrogenase, significant light has been shed on this process. Both the source of the CO and CN(-) ligands to iron and the assembly site of the catalytic subcluster are known, and an apo structure of HydF has been published recently. However, the nature of the substrate(s) for the synthesis of the bridging dithiolate ligand to the subcluster remains to be established. From both spectroscopy and model chemistry, it is predicted that an amine function in this ligand plays a central role in catalysis, acting as a base in the heterolytic cleavage of hydrogen.     

Pathways of H2 toward the active site of [NiFe]-hydrogenase

Hydrogenases catalyze the reversible oxidation of molecular hydrogen (H(2)), but little is known about the diffusion of H(2) toward the active site. Here we analyze pathways for H(2) permeation using molecular dynamics (MD) simulations in explicit solvent. Various MD simulation replicates were done, to improve the sampling of the system states. H(2) easily permeates hydrogenase in every simulation and it moves preferentially in channels. All H(2) molecules that reach the active site made their approach from the side of the Ni ion. H(2) is able to reach distances of <4 A from the active site, although after 6 A permeation is difficult. In this region we mutated Val-67 into alanine and perform new MD simulations. These simulations show an increase of H(2) inside the protein and at lower distances from the active site. This valine can be a control point in the H(2) access to the active center.     

The role of the maturase HydG in [FeFe]-hydrogenase active site synthesis and assembly

[FeFe]-hydrogenases catalyze the protons/hydrogen interconversion through a unique di-iron active site consisting of three CO and two CN ligands, and a non-protein SCH₂XCH₂S (X = N or O) dithiolate bridge. Site assembly requires two “Radical-S-adenosylmethionine (SAM or AdoMet)” iron-sulfur enzymes, HydE and HydG, and one GTPase, HydF. The sequence homology between HydG and ThiH, a Radical-SAM enzyme which cleaves tyrosine into p-cresol and dehydroglycine, and the finding of a similar cleavage reaction catalyzed by HydG suggests a mechanism for hydrogenase maturation. Here we propose that HydG is specifically involved in the synthesis of the dithiolate ligand, with two tyrosine-derived dehydroglycines as precursors along with an [FeS] cluster of HydG functioning both as electron shuttle and source of the sulfur atoms.     

O2 reactions at the six-iron active site (H-cluster) in [FeFe]-hydrogenase

Irreversible inhibition by molecular oxygen (O(2)) complicates the use of [FeFe]-hydrogenases (HydA) for biotechnological hydrogen (H(2)) production. Modification by O(2) of the active site six-iron complex denoted as the H-cluster ([4Fe4S]-2Fe(H)) of HydA1 from the green alga Chlamydomonas reinhardtii was characterized by x-ray absorption spectroscopy at the iron K-edge. In a time-resolved approach, HydA1 protein samples were prepared after increasing O(2) exposure periods at 0 °C. A kinetic analysis of changes in their x-ray absorption near edge structure and extended X-ray absorption fine structure spectra revealed three phases of O(2) reactions. The first phase (τ(1) ≤ 4 s) is characterized by the formation of an increased number of Fe-O,C bonds, elongation of the Fe-Fe distance in the binuclear unit (2Fe(H)), and oxidation of one iron ion. The second phase (τ(2) ≈ 15 s) causes a ～50% decrease of the number of ～2.7-? Fe-Fe distances in the [4Fe4S] subcluster and the oxidation of one more iron ion. The final phase (τ(3) ≤ 1000 s) leads to the disappearance of most Fe-Fe and Fe-S interactions and further iron oxidation. These results favor a reaction sequence, which involves 1) oxygenation at 2Fe(H(+)) leading to the formation of a reactive oxygen species-like superoxide (O(2)(-)), followed by 2) H-cluster inactivation and destabilization due to ROS attack on the [4Fe4S] cluster to convert it into an apparent [3Fe4S](+) unit, leading to 3) complete O(2)-induced degradation of the remainders of the H-cluster. This mechanism suggests that blocking of ROS diffusion paths and/or altering the redox potential of the [4Fe4S] cubane by genetic engineering may yield improved O(2) tolerance in [FeFe]-hydrogenase.     

Mechanism of proton transfer in [FeFe]-hydrogenase from Clostridium pasteurianum

[FeFe]-Hydrogenases are complex metalloproteins that catalyze the reversible reduction of protons to molecular hydrogen utilizing a unique diiron subcluster bridged to a [4Fe4S] subcluster. Extensive studies have concentrated on the nature and catalytic activity of the active site, yet relatively little information is available concerning the mechanism of proton transport that is required for this activity. Previously, structural characterization of [FeFe]-hydrogenase from Clostridium pasteurianum indicated a potential proton transport pathway involving four residues (Cys-299, Glu-279, Ser-319, and Glu-282) that connect the active site to the enzyme surface. Here, we demonstrate that substitution of any of these residues resulted in a drastic reduction in hydrogenase activity relative to the native enzyme, supporting the importance of these residues in catalysis. Inhibition studies of native and amino acid-substituted enzymes revealed that Zn(2+) specifically blocked proton transfer by binding to Glu-282, confirming the role of this residue in the identified pathway. In addition, all four of these residues are strictly conserved, suggesting that they may form a proton transport pathway that is common to all [FeFe]-hydrogenases.     

Optimized expression and purification for high-activity preparations of algal [FeFe]-hydrogenase

Background

Recombinant expression and purification of metallo-enzymes, including hydrogenases, at high-yields is challenging due to complex, and enzyme specific, post-translational maturation processes. Low fidelities of maturation result in preparations containing a significant fraction of inactive, apo-protein that are not suitable for biophysical or crystallographic studies.

Principal Findings

We describe the construction, overexpression and high-yield purification of a fusion protein consisting of the algal [2Fe2S]-ferredoxin PetF (Fd) and [FeFe]-hydrogenase HydA1. The maturation of Fd-HydA1 was optimized through improvements in culture conditions and media components used for expression. We also demonstrated that fusion of Fd to the N-terminus of HydA1, in comparison to the C-terminus, led to increased expression levels that were 4-fold higher. Together, these improvements led to enhanced HydA1 activity and improved yield after purification. The strong binding-affinity of Fd for DEAE allowed for two-step purification by ion exchange and StrepTactin affinity chromatography. In addition, the incorporation of a TEV protease site in the Fd-HydA1 linker allowed for the proteolytic removal of Fd after DEAE step, and purification of HydA1 alone by StrepTactin. In combination, this process resulted in HydA1 purification yields of 5 mg L⁻¹ of culture from E. coli with specific activities of 1000 U (U = 1 µmol hydrogen evolved mg⁻¹ min⁻¹).

Significance

The [FeFe]-hydrogenases are highly efficient enzymes and their catalytic sites provide model structures for synthetic efforts to develop robust hydrogen activation catalysts. In order to characterize their structure-function properties in greater detail, and to use hydrogenases for biotechnological applications, reliable methods for rapid, high-yield expression and purification are required.     

Molecular dynamics study of the proposed proton transport pathways in [FeFe]-hydrogenase

Possible proton transport pathways in Clostridium pasteurianum (CpI) [FeFe]-hydrogenase were investigated with molecular dynamics simulations. This study was undertaken to evaluate the functional pathway and provide insight into the hydrogen bonding features defining an active proton transport pathway. Three pathways were evaluated, two of which consist of water wires and one of predominantly amino acid residues. Our simulations suggest that protons are not transported through water wires. Instead, the five-residue motif (Glu282, Ser319, Glu279, H₂O, Cys299) was found to be the likely pathway, consistent with previously made experimental observations. The pathway was found to have a persistent hydrogen bonded core (residues Cys299 to Ser319), with less persistent hydrogen bonds at the ends of the pathway for both H₂ release and H₂ uptake. Single site mutations of the four residues have been shown experimentally to deactivate the enzyme. The theoretical evaluation of these mutations demonstrates redistribution of the hydrogen bonds in the pathway, resulting in enzyme deactivation. Finally, coupling between the protein dynamics near the proton transport pathway and the redox partner binding regions was also found as a function of H₂ uptake and H₂ release states, which may be indicative of a correlation between proton and electron movement within the enzyme.     

Purification and characterization of the [NiFe]-hydrogenase of Shewanella oneidensis MR-1

Shewanella oneidensis MR-1 possesses a periplasmic [NiFe]-hydrogenase (MR-1 [NiFe]-H(2)ase) that has been implicated in H(2) production and oxidation as well as technetium [Tc(VII)] reduction. To characterize the roles of MR-1 [NiFe]-H(2)ase in these proposed reactions, the genes encoding both subunits of MR-1 [NiFe]-H(2)ase were cloned and then expressed in an MR-1 mutant without hyaB and hydA genes. Expression of recombinant MR-1 [NiFe]-H(2)ase in trans restored the mutant's ability to produce H(2) at 37% of that for the wild type. Following purification, MR-1 [NiFe]-H(2)ase coupled H(2) oxidation to reduction of Tc(VII)O(4)(-) and methyl viologen. Change of the buffers used affected MR-1 [NiFe]-H(2)ase-mediated reduction of Tc(VII)O(4)(-) but not methyl viologen. Under the conditions tested, all Tc(VII)O(4)(-) used was reduced in Tris buffer, while in HEPES buffer, only 20% of Tc(VII)O(4)(-) was reduced. The reduced products were soluble in Tris buffer but insoluble in HEPES buffer. Transmission electron microscopy analysis revealed that Tc precipitates reduced in HEPES buffer were aggregates of crystallites with diameters of ～5 nm. Measurements with X-ray absorption near-edge spectroscopy revealed that the reduction products were a mixture of Tc(IV) and Tc(V) in Tris buffer but only Tc(IV) in HEPES buffer. Measurements with extended X-ray adsorption fine structure showed that while the Tc bonding environment in Tris buffer could not be determined, the Tc(IV) product in HEPES buffer was very similar to Tc(IV)O(2)·nH(2)O, which was also the product of Tc(VII)O(4)(-) reduction by MR-1 cells. These results shows for the first time that MR-1 [NiFe]-H(2)ase catalyzes Tc(VII)O(4)(-) reduction directly by coupling to H(2) oxidation.     

The CO and CN(-) ligands to the active site Fe in [NiFe]-hydrogenase of Escherichia coli have different metabolic origins

The Fe atom in the bimetallic active site of [NiFe]-hydrogenases has one CO and two cyanide ligands. To determine their metabolic origin, [NiFe]-hydrogenase-2 was isolated from Escherichia coli grown in the presence of L-[ureido-(13)C]citrulline, purified and analyzed by infrared spectroscopy. The spectra indicate incorporation of (13)C only into the cyanide ligands and not into the CO, showing that cyanide and CO have different metabolic origins. After growth of E. coli in the presence of (13)CO only the CO ligand was labelled with (13)C. Labelling did not result from an exchange of the intrinsic CO ligand with the exogenous CO.     

A Threonine Stabilizes the NiC and NiR Catalytic Intermediates of [NiFe]-hydrogenase

The heterodimeric [NiFe] hydrogenase from Desulfovibrio fructosovorans catalyzes the reversible oxidation of H₂ into protons and electrons. The catalytic intermediates have been attributed to forms of the active site (NiSI, NiR, and NiC) detected using spectroscopic methods under potentiometric but non-catalytic conditions. Here, we produced variants by replacing the conserved Thr-18 residue in the small subunit with Ser, Val, Gln, Gly, or Asp, and we analyzed the effects of these mutations on the kinetic (H₂ oxidation, H₂ production, and H/D exchange), spectroscopic (IR, EPR), and structural properties of the enzyme. The mutations disrupt the H-bond network in the crystals and have a strong effect on H₂ oxidation and H₂ production turnover rates. However, the absence of correlation between activity and rate of H/D exchange in the series of variants suggests that the alcoholic group of Thr-18 is not necessarily a proton relay. Instead, the correlation between H₂ oxidation and production activity and the detection of the NiC species in reduced samples confirms that NiC is a catalytic intermediate and suggests that Thr-18 is important to stabilize the local protein structure of the active site ensuring fast NiSI-NiC-NiR interconversions during H₂ oxidation/production.     

Production of biohydrogen by heterologous expression of oxygen-tolerant Hydrogenovibrio marinus [NiFe]-hydrogenase in Escherichia coli

Oxygen sensitivity of hydrogenase is a critical issue in efficient biological hydrogen production. In the present study, oxygen-tolerant [NiFe]-hydrogenase from the marine bacterium, Hydrogenovibrio marinus, was heterologously expressed in Escherichia coli, for the first time. Recombinant E. coli BL21 expressing H. marinus [NiFe]-hydrogenase actively produced hydrogen, but the parent strain did not. Recombinant H. marinus hydrogenase required both nickel and iron for biological activity. Compared to the recombinant E. coli [NiFe]-hydrogenase 1 described in our previous report, recombinant H. marinus [NiFe]-hydrogenase displayed 1.6- to 1.7-fold higher hydrogen production activity in vitro. Importantly, H. marinus [NiFe]-hydrogenase exhibited relatively good oxygen tolerance in analyses involving changes of surface aeration and oxygen proportion within a gas mixture. Specifically, recombinant H. marinus [NiFe]-hydrogenase produced ∼7- to 9-fold more hydrogen than did E. coli [NiFe]-hydrogenase 1 in a gaseous environment containing 5-10% (v/v) oxygen. In addition, purified H. marinus [NiFe]-hydrogenase displayed a hydrogen evolution activity of ∼28.8 nmol H₂/(min mg protein) under normal aerobic purification conditions. Based on these results, we suggest that oxygen-tolerant H. marinus [NiFe]-hydrogenase can be employed for in vivo and in vitro biohydrogen production without requirement for strictly anaerobic facilities.     

Integrated thermodynamic analysis of electron bifurcating [FeFe]-hydrogenase to inform anaerobic metabolism and H2 production

Electron bifurcating, [FeFe]-hydrogenases are recently described members of the hydrogenase family and catalyze a combination of exergonic and endergonic electron exchanges between three carriers (2 ferredoxin_red⁻ + NAD(P)H + 3 H⁺ = 2 ferredoxin_ox + NAD(P)⁺ + 2 H₂). A thermodynamic analysis of the bifurcating, [FeFe]-hydrogenase reaction, using electron path-independent variables, quantified potential biological roles of the reaction without requiring enzyme details. The bifurcating [FeFe]-hydrogenase reaction, like all bifurcating reactions, can be written as a sum of two non-bifurcating reactions. Therefore, the thermodynamic properties of the bifurcating reaction can never exceed the properties of the individual, non-bifurcating, reactions. The bifurcating [FeFe]-hydrogenase reaction has three competitive properties: 1) enabling NAD(P)H-driven proton reduction at pH₂ higher than the concurrent operation of the two, non-bifurcating reactions, 2) oxidation of NAD(P)H and ferredoxin simultaneously in a 1:1 ratio, both are produced during typical glucose fermentations, and 3) enhanced energy conservation (~10 kJ mol⁻¹ H₂) relative to concurrent operation of the two, non-bifurcating reactions. Our analysis demonstrated ferredoxin E°′ largely determines the sensitivity of the bifurcating reaction to pH₂, modulation of the reduced/oxidized electron carrier ratios contributed less to equilibria shifts. Hydrogenase thermodynamics data were integrated with typical and non-typical glycolysis pathways to evaluate achieving the ‘Thauer limit’ (4 H₂ per glucose) as a function of temperature and pH₂. For instance, the bifurcating [FeFe]-hydrogenase reaction permits the Thauer limit at 60 °C if pH ₂ ≤ ~10 mbar. The results also predict Archaea, expressing a non-typical glycolysis pathway, would not benefit from a bifurcating [FeFe]-hydrogenase reaction; interestingly, no Archaea have been observed experimentally with a [FeFe]-hydrogenase enzyme.     

Distinct physiological roles of the three [NiFe]-hydrogenase orthologs in the hyperthermophilic archaeon Thermococcus kodakarensis

Hydrogenases catalyze the reversible oxidation of molecular hydrogen (H₂) and play a key role in the energy metabolism of microorganisms in anaerobic environments. The hyperthermophilic archaeon Thermococcus kodakarensis KOD1, which assimilates organic carbon coupled with the reduction of elemental sulfur (S⁰) or H₂ generation, harbors three gene operons encoding [NiFe]-hydrogenase orthologs, namely, Hyh, Mbh, and Mbx. In order to elucidate their functions in vivo, a gene disruption mutant for each [NiFe]-hydrogenase ortholog was constructed. The Hyh-deficient mutant (PHY1) grew well under both H₂S- and H₂-evolving conditions. H₂S generation in PHY1 was equivalent to that of the host strain, and H₂ generation was higher in PHY1, suggesting that Hyh functions in the direction of H₂ uptake in T. kodakarensis under these conditions. Analyses of culture metabolites suggested that significant amounts of NADPH produced by Hyh are used for alanine production through glutamate dehydrogenase and alanine aminotransferase. On the other hand, the Mbh-deficient mutant (MHD1) showed no growth under H₂-evolving conditions. This fact, as well as the impaired H₂ generation activity in MHD1, indicated that Mbh is mainly responsible for H₂ evolution. The copresence of Hyh and Mbh raised the possibility of intraspecies H₂ transfer (i.e., H₂ evolved by Mbh is reoxidized by Hyh) in this archaeon. In contrast, the Mbx-deficient mutant (MXD1) showed a decreased growth rate only under H₂S-evolving conditions and exhibited a lower H₂S generation activity, indicating the involvement of Mbx in the S⁰ reduction process. This study provides important genetic evidence for understanding the physiological roles of hydrogenase orthologs in the Thermococcales.     

Genetic analyses of the functions of [NiFe]-hydrogenase maturation endopeptidases in the hyperthermophilic archaeon Thermococcus kodakarensis

The maturation of [NiFe]-hydrogenases requires a number of accessory proteins, which include hydrogenase-specific endopeptidases. The endopeptidases carry out the final cleavage reaction of the C-terminal regions of [NiFe]-hydrogenase large subunit precursors. The hyperthermophilic archaeon Thermococcus kodakarensis harbors two [NiFe]-hydrogenases, a cytoplasmic Hyh and a membrane-bound Mbh, along with two putative hydrogenase-specific endopeptidase genes. In this study, we carried out a genetic examination on the two endopeptidase genes, TK2004 and TK2066. Disruption of TK2004 resulted in a strain that could not grow under conditions requiring hydrogen evolution. The Mbh large subunit precursor (pre-MbhL) in this strain was not processed at all whereas Hyh cleavage was not affected. On the other hand, disruption of TK2066 did not affect the growth of T. kodakarensis under the conditions examined. Cleavage of the Hyh large subunit precursor (pre-HyhL) was impaired, but could be observed to some extent. In a strain lacking both TK2004 and TK2066, cleavage of pre-HyhL could not be observed. Our results indicate that pre-MbhL cleavage is carried out solely by the endopeptidase encoded by TK2004. Pre-HyhL cleavage is mainly carried out by TK2066, but TK2004 can also play a minor role in this cleavage.

     

Reactions of H2, CO, and O2 with active [NiFe]-hydrogenase from Allochromatium vinosum. A stopped-flow infrared study

The Ni-Fe site in the active membrane-bound [NiFe]-hydrogenase from Allochromatium vinosum can exist in three different redox states. In the most oxidized state (Ni(a)-S) the nickel is divalent. The most reduced state (Ni(a)-SR) likewise has Ni(2+), while the intermediate state (Ni(a)-C) has Ni(3+). The transitions between these states have been studied by stopped-flow Fourier transform infrared spectroscopy. It is inferred from the data that the Ni(a)-S --> Ni(a)-C* and Ni(a)-C* --> Ni(a)-SR transitions induced by dihydrogen require one of the [4Fe-4S] clusters to be oxidized. Enzyme in the Ni(a)-S* state with all of the iron-sulfur clusters reduced reacts with dihydrogen to form the Ni(a)-SR state in milliseconds. By contrast, when one of the cubane clusters is oxidized, the Ni(a)-S state reacts with dihydrogen to form the Ni(a)-C state with all of the iron-sulfur clusters reduced. The competition between dihydrogen and carbon monoxide for binding to the active site was dependent on the redox state of the nickel ion. Formation of the Ni(a)-S.CO state (Ni(2+)) by reacting CO with enzyme in the Ni(a)-SR and Ni(a)-S states (Ni(2+)) is considerably faster than its formation from enzyme in the Ni(a)-C* (Ni(3+)) state. Excess oxygen converted hydrogen-reduced enzyme to the inactive Ni(r)* state within 158 ms, suggesting a direct reaction at the Ni-Fe site. With lower O(2) concentrations the formation of intermediate states was observed. The results are discussed in the light of the present knowledge of the structure and mechanism of action of the A. vinosum enzyme.     

Patterns of [FeFe] hydrogenase diversity in the gut microbial communities of lignocellulose-feeding higher termites

Hydrogen is the central free intermediate in the degradation of wood by termite gut microbes and can reach concentrations exceeding those measured for any other biological system. Degenerate primers targeting the largest family of [FeFe] hydrogenases observed in a termite gut metagenome have been used to explore the evolution and representation of these enzymes in termites. Sequences were cloned from the guts of the higher termites Amitermes sp. strain Cost010, Amitermes sp. strain JT2, Gnathamitermes sp. strain JT5, Microcerotermes sp. strain Cost008, Nasutitermes sp. strain Cost003, and Rhyncotermes sp. strain Cost004. Each gut sample harbored a more rich and evenly distributed population of hydrogenase sequences than observed previously in the guts of lower termites and Cryptocercus punctulatus. This accentuates the physiological importance of hydrogen for higher termite gut ecosystems and may reflect an increased metabolic burden, or metabolic opportunity, created by a lack of gut protozoa. The sequences were phylogenetically distinct from previously sequenced [FeFe] hydrogenases. Phylogenetic and UniFrac comparisons revealed congruence between host phylogeny and hydrogenase sequence library clustering patterns. This may reflect the combined influences of the stable intimate relationship of gut microbes with their host and environmental alterations in the gut that have occurred over the course of termite evolution. These results accentuate the physiological importance of hydrogen to termite gut ecosystems.     

Carbamoylphosphate serves as the source of CN(-), but not of the intrinsic CO in the active site of the regulatory [NiFe]-hydrogenase from Ralstonia eutropha

Within the catalytic centre of [NiFe]-hydrogenases one carbonyl and two cyanide ligands are covalently attached to the iron. To identify the metabolic origins of these ligands, the regulatory [NiFe] hydrogenase in conjunction with the indigenous Hyp maturation proteins of Ralstonia eutropha H16 were heterologously overproduced in E. coli grown in the presence of L-[ureido-(13)C] citrulline and NaH(13)CO(3). Infrared spectroscopy of purified hydrogenase provided direct evidence that only the cyanide ligands, but not the CO ligand, originate from CO(2) and carbamoylphosphate. Incorporation of label from (13)CO exclusively into the carbonyl ligand indicates that free CO is a possible precursor in carbonyl ligand biosynthesis.     

Pre-steady-state kinetics of the reactions of [NiFe]-hydrogenase from Chromatium vinosum with H2 and CO.

Results are presented of the first rapid-mixing/rapid-freezing studies with a [NiFe]-hydrogenase. The enzyme from Chromatium vinosum was used. In particular the reactions of active enzyme with H2 and CO were monitored. The conversion from fully reduced, active hydrogenase (Nia-SR state) to the Nia-C* state was completed in less than 8 ms, a rate consistent with the H2-evolution activity of the enzyme. The reaction of CO with fully reduced enzyme was followed from 8 to 200 ms. The Nia-SR state did not react with CO. It was discovered, contrary to expectations, that the Nia-C* state did not react with CO when reactions were performed in the dark. When H2 was replaced by CO, a Nia-C* EPR signal appeared within 11 ms; this was also the case when H2 was replaced by Ar. With CO, however, the Nia-C* state decayed within 40 ms, due to the generation of the Nia-S.CO state (the EPR-silent state of the enzyme with bound CO). The Nia-C* state, induced after 11 ms by replacing H2 by CO in the dark, could be converted, in the frozen enzyme, into the EPR-detectable state with CO bound to nickel (Nia*.CO) by illumination at 30 K (evoking the Nia-L* state), followed by dark adaptation at 200 K. This can be explained by assuming that the Nia-C* state represents a formally trivalent state of nickel, which is unable to bind CO, whereas nickel in the Nia-L* and the Nia*.CO states is formally monovalent.     

The active site of the [FeFe]-hydrogenase from Desulfovibrio desulfuricans. I. Light sensitivity and magnetic hyperfine interactions as observed by electron paramagnetic resonance

The hydrogen-activating cluster (H cluster) in [FeFe]-hydrogenases consists of two moieties. The [2Fe]_H subcluster is a (L)(CO)(CN)Fe(μ-RS₂)(μ-CO)Fe(CysS)(CO)(CN) centre. The Cys-bound Fe is called Fe1, the other iron Fe2. The Cys-thiol forms a bridge to a [4Fe–4S] cluster, the [4Fe–4S]_H subcluster. We report that electron paramagnetic resonance (EPR) spectra of the ⁵⁷Fe-enriched enzyme from Desulfovibrio desulfuricans in the H_ox–CO state are consistent with a magnetic hyperfine interaction of the unpaired spin with all six Fe atoms of the H cluster. In contrast to the inactive aerobic enzyme, the active enzyme is easily destroyed by light. The [2Fe]_H subcluster in some enzyme molecules loses CO by photolysis, whereupon other molecules firmly bind the released CO to form the H_ox–CO state giving rise to the so-called axial 2.06 EPR signal. Though not destroyed by light, the H_ox–CO state is affected by it. As demonstrated in the accompanying paper [49] two of the intrinsic COs, both bound to Fe2, can be exchanged by extrinsic ¹³CO during illumination at 2 °C. We found that only one of the three ¹³COs, the one at the extrinsic position, gives an EPR-detectable isotropic superhyperfine interaction of 0.6 mT. At 30 K both the inhibiting extrinsic CO bound to Fe2 and one more CO can be photolysed. EPR spectra of the photolysed products are consistent with a 3d ⁷ system of Fe with the formal oxidation state +1. The damaged enzyme shows a light-sensitive g=5 signal which is ascribed to an S=3/2 form of the [2Fe]_H subcluster. The light sensitivity of the enzyme explains the occurrence of the g=5 signal and the axial 2.06 signal in published EPR spectra of nearly all preparations studied thus far.