Superclusters : a search for novel structures and functions of biological iron-sulfur clusters

Iron sulfur (Fe-S) proteins are found in a variety of organisms. They usually function in electron transport, but they may also be involved in other functions like gene regulation and Lewis acid catalysis. The structure and spectroscopic properties of Fe-S clusters holding one, two, three, or four iron atoms is known to a great extent. These 'common' clusters share some basic properties. Firstly, they contain not more than four iron atoms. Secondly, despite the fact that they may contain more than one iron atom, they exist in two (physiological) redox states only. Thirdly, they are characterized by a low electron spin, i.e. they are usually S = 1/2. However, there is a strong indication that Fe-S clusters exist which do not obey these general rules. These clusters may hold more than four iron atoms, they are usually high-spin (S ≥3/2), and they may exist in more than two redox states. Because of these properties these clusters are referred to as superclusters. When I started this research project, six potential systems were proposed to contain larger (≥4Fe) or uncommon (WFe3S4) Fe-S clusters. These enzymes are involved in (multi-electron) redox catalysis:1. Nitrogenase2. Fe-only hydrogenase3. Dissimilatory sulfite reductase4. carbonmonoxide dehydrogenase5. Prismane protein6. Pyrococcus furiosus aldehyde oxidoreductaseThe aim of my thesis is to study physical, chemical, and biological properties of multi- electron transferring enzymes, in a quest for possible new structures and functions of biological Fe-S clusters.Chapter 2 describes the purification and characterization of a dissimilatory sulfite reductase from Desulfosarcina variabilis. The enzyme belongs to the class of desulforubidins, as was deduced from its UV/vis absorption spectrum. It is a a 2β2γ2 hexamer of ≈208 kDa, and it was found to contain ≈15 Fe and ≈19 S 2-. The oxidized enzyme exhibited S = 9/2 Fe-S EPR signals (g = 16 ). Similar signals have previously been found in Desulfovibrio vulgaris (Hildenborough) desulfoviridin by Pierik and Hagen, who suggested the presence of a larger Fe-S cluster. With the finding of similar very high spin signals also in D. variabilis desulforubidin, it appears that the presence of a S = 9/2 Fe-S (super) cluster is common in all dissimilatory sulfite reductases. The sirohemes in D. variabilis desulforubidin were found to be fully metalated, and none of the Fe-S EPR signals gave indication for dipolar and/or exchange coupling with siroheme. These observations are interpreted as supportive evidence against the previously proposed model of a bridged cubane/siroheme as the active site for dissimilatory sulfite reductases.The extreme hyperthermophile Pyrococcus furiosus contains a NiFe hydrogenase which not only reversibly oxidizes hydrogen but also reduces elemental sulfur (S 0) to H 2 S. The Archaeon was succesfully grown in a 200 l fermentor at 90°C on potato starch, and the hydrogenase could be purified aerobically without loss of activity. In contrast to previous reported data the enzyme was found to contain 17 Fe, 17 S 2-, and 0.74 Ni. Three EPR signals were found; a near-axial (g = 2.02, 1.95, 1.92) S = 1/2 signal ( Em,7.5 = - 303 mV) indicative of a [2Fe-2S] (2+;1+)cluster, a broad spectrum of unknown origin (g = 2.25, 1.89; Em,7.5 = -3 10 mV), and a novel rhombic S = 1/2 EPR signal (g = 2.07, 1.93, 1.89) reminiscent of a [4Fe-4S] (2+;1+)cluster. This rhombic signal appears with a reduction potential of Em,7.5 = -90 mV, and disappears at Em,7.5 = -328 mV. The latter observation suggested that this cluster is capable of taking up two electrons, and, therefore, that it is a supercluster. However, it is hypothesized that the disappearance of the signals at low potential is caused by magnetic interaction of the rhombic g = 2.07 signal with a third paramagnet, resulting in a broad interaction signal. Hence, there is no indication for the presence of a supercluster in P.furiosus NiFe hydrogenase (chapter 3).In the NiFe hydrogenases of several organisms as well as in the Fe-only hydrogenases of Megasphaera eldenii and Desulfovibrio vulgaris (Hildenborough) novel Fourier transform infrared (FTIR) detectable groups were found. The bands occur in the region of 2100-1800 cm -1, which corresponds to stretching vibrations of polar triple bonds, metal hydrides, or asymetrically coupled vibrations of two adjacent double bonds. The position of these bands shifted upon oxidation and reduction of the enzymes. FTIR bands in this region were not detected in a large control group of Fe-S and/or nickel containing proteins including a metal-free hydrogenase. The FTIR groups in NiFe hydrogenases are assigned to the three unidentified small non-protein ligands that coordinate the Fe as observed in the X-ray structure of Desulfovibrio gigas NiFe hydrogenase. Thus far, the structural difference between NiFe- and Feonly hydrogenases had been thought to reside in the absence or presence, respectively, of a novel Fe-S cluster (H-cluster) which is proposed to be the site of hydrogen activation. The finding of similar FTIR groups in both Fe-only and NiFe-hydrogenases might suggest that the hydrogen-activating site of both classes of hydrogenases encompasses of a bimetallic center involving a low spin Fe ion with FTIR- detectable groups.During the purification of several proteins from Desulfovibrio vulgaris strain Hildenborough a yellowish fraction eluted from the first anion exchange column at high NaCl concentration. It was observed that this fraction absorbed strongly at 260 nm. Attempts were made to purify the protein. The protein turned out to be a ferredoxin, as concluded from its size (a homodimer of subunits, each of 7.5 kDa), its pi (3.9) and its EPR spectrum in the reduced state, indicating the presence of a [4Fe-4S] (2+;1+)cluster. The protein was associated with RNA having a typical size of 9-17 nucleotides. Hybridization experiments with extracted, radiolabeled RNA and digested D. vulgaris genomic DNA indicated that the ferredoxin binds either to total RNA or specifically to rRNA. The suggestion is made that D. vulgaris ferredoxin may not be a redox protein, but that it may have a regulatory function. This suggestion is supported by the unusually high standard reaction entropy of reduction of -230 J . K -1.mol -1. This would be the first prokaryotic Fe-S protein known to function in translation regulation.In chapter 6 an EPR/redox study is presented on the tungsten-containing aldehyde oxidoreductase (AOR) from the hyperthermophile Pyrococcus furiosus. The enzyme had previously been suggested to hold a [WFe3S4] cluster. Highly active AOR could be obtained by rapid, anaerobic purification (i.e. within two days). Only active enzyme was used for this study. The fully reduced enzyme exhibited a mixture of S = 1/2 and S = 3/2 Fe-S EPR signals. Oxidized AOR afforded signals typical for W 5+(g = 1.982, 1.953, 1.885). Shortly after this research project started the X-ray structure of P.furiosus AOR was elucidated by Chan et al., who showed that the enzyme contains one [4Fe-4S] cluster and one tungsten cofactor per subunit. Our data are in agreement with the crystal structure, which excluded the possibility of a [WFe3S4] cluster. Such a cluster would have been a completely novel Fe-S cluster. The W 5+spectrum could be simulated using 183W hyperfine splitting constants A xyz of 7.7, 4.5, and 4.2 mT. A reduction potential Em,7.5 = +180 mV was determined for the couple W 4+/W 5+. Given the low reduction potential of the substrate, it is suggested that the biologically relevant redox chemistry may not not be located on the tungsten, but rather on the pterin cofactor.The prismane proteins of Desulfovibrio vulgaris (Hildenborough) and Desulfovibrio desulfuricans ATCC 27774 are proposed to contain a [6Fe-6S] cluster. A similar cluster has been proposed to be present in the active site of Fe-only hydrogenases. Unfortunately, crystallographic evidence for the presence of [6Fe-6S] clusters is still lacking. Several attempts were made to crystallize the D. vulgaris (H) protein in our lab, but these were all unsuccessful. Eventually, high quality crystals were obtained in Daresbury, U.K. in collaboration with prof Lindley. Crystals grew within four days, and grew to a maximum size of 0.7 mm within ten days. The resolution of the diffraction pattern extends to 1.7 Å. The unit cell is orthorhombic, with spacegroup P2 1 2 1 2 1 . The unit cell will readily hold four molecules of molecular mass of 60 kDa, with a solvent content of approximately 48%.Generally, I looked for superclusters in the following multi-electron transferring enzymes: dissimilatory sulfite reductase, hydrogenase, and P.furiosus aldehyde oxidoreductase. For Desulfosarcina variabilis dissimilatory sulfite reductase I found indication for the presence of a supercluster, whereas for hydrogenases the finding of novel FTIR resonances suggest a unique metal structure to be part of the active site. No indication was found for the presence of a supercluster in Pyrococcus furiosus aldehyde oxidoreductase, but the enzyme showed to be an interesting case for the study of biological tungsten. An apparent supercluster in Pyrococcus furiosus NiFe hydrogenase turned out to be most likely a [4Fe-4S] cluster. An RNA-binding ferredoxin from Desulfovibrio vulgaris (H) may be the first example of a prokaryotic, gene regulating Fe-S protein. Finally, elucidation of the crystal structure of the prismane protein from Desulfovibrio vulgaris (Hildenborough) will be a major step towards the development of the concept of larger Fe-S clusters.