Molecular studies on iron-sulfur proteins in Desulfovibrio

<p><strong><em>Desulfovibrio vulgaris</em> (Hildenborough)</strong> . The organism described in this thesis, is an anaerobic gram-negative sulfate reducing bacterium (SRB). Its natural environments are the anaerobic sediments in lower levels of lakes and pools. This habitat is rich in sulfate that is used as terminal electron-acceptor by the organism and by performing this, <em>D. vulgaris</em> contributes to the important sulfur-cycle in nature. <em>D. vulgaris</em> can both utilize lactate (by anaerobic oxidation) and molecular hydrogen as energy source. The oxidation of lactate to acetate and C0 <sub>2</sub> occurs in the cytoplasm or at the cytoplasmic membrane and results in the production of ATP and the release of protons and electrons. When <em>D. vulgaris</em> uses molecular hydrogen as substrate, the oxidation of the hydrogen occurs at the periplasmic side of the inner membrane. This creates a proton motive force that drives ATP synthesis.<p><strong>Periplasmic Fe-hydrogenase</strong> . Molecular hydrogen has been shown to play an important role in both energy-evolving systems described above. So far, three hydrogenases, catalyzing the reversible H <sub>2</sub> oxidation and reduction of protons have been identified in <em>D. vulgaris</em> (Hildenborough). The precise physiological function of each of these hydrogenases remains unclear. Two of these enzymes are localized in the cytoplasmic membrane and contain nickel in addition to iron-sulfur clusters as cofactor. The third enzyme contains only iron-sulfur clusters as cofactors and resides in the periplasmic space of the bacterium. This enzyme exhibits one of the highest catalytic activities ever described for hydrogenases. It occurs as a heterodimer that is composed of a large cc subunit (46 kDa) and a small 0 subunit (10 kDa). Only the small subunit is translated as a precursor (13 kDa) with a cleavable signal sequence for export that is probably involved in the export of both hydrogenase subunits across the cytoplasmic membrane. The catalytically active enzyme contains three iron-sulfur clusters as cofactors. Two of them are typical ferredoxin-like [4Fe-4S] clusters (F-clusters) involved in electron transport. The third (H-cluster) cluster contains six iron and sulfide ions coordinated in an unknown structure and is part of the catalytic center of the enzyme.<p>Ile gene encoding the Fe-hydrogenase was the first hydrogenase gene that was isolated and expressed in <em>E. coli</em> . From these expression studies it became apparent that only very small amounts of αand βsubunits were assembled into an αβdimer and transported across the membrane. Also the iron-sulfur cluster incorporation was incomplete in the recombinant enzyme. The enzyme contained sub-stoichiometric amounts of F-clusters, while the H-cluster was not incorporated at all. These results indicated that the assembly and export of hydrogenase generating a catalytically active enzyme, are not spontaneously occurring processes, but involve specific helper components, as has been shown for other enzymes with redox-active metal clusters (reviewed in <strong>Chapter 1</strong> ).<p><strong>Studies regarding the biosynthesis of Fe-hydrogenase: the <em>hydC</em> gene</strong> . As genes serving a single pathway are often clustered in the genome, the identification of genes encoding these additional activating components, was started by the isolation of large DNA fragments surrounding the structural hydrogenase genes. Surprisingly, one of the large isolated DNA fragments contained a gene, <em>hydC</em> ( <strong>Chapter 2</strong> ) with homology in primary structure to the αand βsubunits of the Fe-hydrogenase.<br/>HydC has a high degree of similarity with both the αsubunit of the Fe-hydrogenase (in its central part) and with the βsubunit, minus the leader peptide (in its C-terminal part). Analogous to the FeMo co-factor insertion in nitrogenase component 1 which involves genes ( <em>nifEN</em> ) <em></em> with high similarities to the structural subunits, it was speculated that the <em>hydC</em> gene might code for a helper protein that is involved in the processing of the hydrogenase. 'Me primary structure of <em>hydC</em> contains a N-terminus with no homology with one of the hydrogenase subunits. Subsequently, it was found that this N-terminal segment has homology with mitochondrial NADH-ubiquinone reductase, with subunits of a NAD+-reducing NiFe-hydrogenase from <em>Alcaligenes eutrophus</em> and with the Fe-hydrogenase I from <em>Clostridium pasteurianum</em> ( <strong>Chapter 3</strong> ). On the basis of what is known about iron-sulfur cluster contents of these three enzymes and the conservation of cysteine motifs in these proteins, it was suggested that these motifs coordinate [2Fe-2S] clusters. Unfortunately, the HydC protein could not be purified from <em>D. vulgaris,</em> because no growth conditions were found resulting in a sufficient production of HydC protein. This hampered a further<br/>biochemical and spectroscopical characterization of the protein. On the other hand, the high degree of homology with the <em>C. pasteurianum</em> Fe-hydrogenase, strongly suggests that HydC is a second alternative Fe-hydrogenase and not a helper protein involved in the processing of Fe-hydrogenase.<p>Numerous attempts have been made to exchange the genes for the subunits of the Fe- hydrogenase and the <em>hydC</em> gene in the <em>D. vulgaris</em> genome with inactivated, interrupted copies of the genes. This type of marker exchange experiments would also be very useful for the identification of genes involved in biosynthesis of hydrogenase. One of the requirements for marker exchange is a system for the introduction of plasmids into <em>Desulfovibrio.</em> Such a plasmid transfer system has been developed, but subsequent experiments to apply it for marker exchange have been unsuccessful.<p><strong>The prismane protein</strong> . The inability to design a system for marker-exchange mutagenesis in <em>Desulfovibrio</em> blocked further study of the biosynthesis of the Fe-hydrogenase Therefore, investigations on another protein from <em>D. vulgaris,</em> the prismane protein, were started that are described in the second part of this thesis. As mentioned earlier, some indications were obtained that the H-cluster of Fe-hydrogenase is a [6Fe-6S] cluster.<p>Stronger indications for the existence of such "supercluster" were obtained by Hagen and Pierik in our department for another iron-sulfur containing protein from <em>D. vulgaris,</em> the prismane protein. They isolated a protein containing six irons and sulfide ions coordinated in only one [6Fe-6S] cluster. The putative [6Fe-6S] prismane cluster occurs in four different redox- states: the three-electron reduced state [6Fe-6S] <sup>3+</SUP>(S=1/2), [6Fe-6S] <sup>4+</SUP>(S = even), [6Fe-6S] <sup>5+</SUP>(S = 1/2 and S = 9/2) and the fully oxidized [6Fe-6S] <sup>6+</SUP>(S=0) that shows no EPR spectrum. <strong>Chapter 4</strong> and <strong>5</strong> describe the isolation of the genes for the prismane proteins from <em>D. vulgaris</em> (Hildenborough) and <em>D. desulfuricans</em> (ATCC 27774) and the determination of the amino acid sequence. Both proteins are highly conserved (66% identical residues), except for a 100 residues segment (residue 50-150). Besides this, both proteins contain typical cysteine motifs at the N-terminus. These motifs have also been found in the sequence of the a subunit of CO dehydrogenase from <em>Methanothrix soehngenii</em> and, in a slightly modified form, in that of CO dehydrogenase from <em>Clostridium thermoaceticum.</em> Also <em></em> for the CO dehydrogenase from <em>M. soehngenii</em> a supercluster has been proposed. Therefore, it is tempting to speculate about the involvement of this motif in the ligation of the [6Fe-6S] prismane cluster.<p>Prismane protein is produced only in small amounts in <em>D. vulgaris.</em> Since large amounts of purified prismane protein are required for X-ray crystallography and Mössbauer studies, efforts were made for overproduction of the protein ( <strong>Chapter 6</strong> ). In a first attempt, the protein was overproduced in <em>E. coli.</em> In this host, a high production of prismane protein was obtained, but no iron-sulfur cluster was incorporated into the protein. The overproduced protein occurred as large insoluble protein-complexes. A second attempt for the overproduction of prismane protein was performed in <em>D. vulgaris</em> by using the aforementioned cloning system. <em>A</em> 25-fold overproduction of prismane protein was obtained by the introduction of extra copies of the gene encoding the prismane protein on a stable plasmid. Biochemical and spectroscopic properties of the protein overproduced in <em>D. vulgaris</em> were shown to be identical to wild-type prismane with one exception: in the as-isolated, oneelectron-reduced state the protein shows EPR signals belonging to a second (S=1/2), spin system that was not observed in the wild-type protein. These additional signals were also described for the wild-type prismane protein purified from <em>D. desulfuricans</em> by Moura and co-workers in Portugal. EPR signals belonging to this second (S=1/2), spin system disappear upon reduction/re-oxidation of the overproduced prismane protein, indicating that this spin system represents a different magnetic form of the [6Fe-6S] cluster. There are no indications for a second cluster as proposed by Moura et al. Determination of the three dimensional structure by X-ray crystallography and further Mössbauer spectroscopy of the overproduced prismane protein are subject for further study in our department and will ultimately lead to insight into the structure of this novel iron-sulfur cofactor