The increasing sequence information on oxygen reductases of the haem-copper superfamily, together with the available three-dimensional structures, allows a clear identification of their common, functionally important features. Taking into consideration both the overall amino acid sequences of the core subunits and key residues involved in proton transfer, a novel hypothesis for the molecular evolution of these enzymes is proposed. Three main families of oxygen reductases are identified on the basis of common features of the core subunits, constituting three lines of evolution: (i) type A (mitochondrial-like oxidases), (ii) type B (ba3-like oxidases) and (iii) type C (cbb3-type oxidases). The first group can be further divided into two subfamilies, according to the helix VI residues at the hydrophobic end of one of the proton pathways (the so-called D-channel): (i) type A1, comprising the enzymes with a glutamate residue in the motif -XGHPEV-, and (ii) type A2, enzymes having instead a tyrosine and a serine in the alternative motif -YSHPXV-. This second subfamily of oxidases is shown to be ancestor to the one containing the glutamate residue, which in the Bacteria domain is only present in oxidases from Gram-positive or purple bacteria. It is further proposed that the Archaea domain acquired terminal oxidases by gene transfer from the Gram-positive bacteria, implying that these enzymes were not present in the last common ancestor before the divergence between Archaea and Bacteria. In fact, most oxidases from archaea have a higher amino acid sequence identity and similarity with those from bacteria, mainly from the Gram-positive group, than with oxidases from other archaea. Finally, a possible relation between the dihaemic subunit (FixP) of the cbb3 oxidases and subunit II of caa3 oxidases is discussed. As the families of haem-copper oxidases can also be identified by their subunit II, a parallel evolution of subunits I and II is suggested.

The Bacillus subtilis endospore coat protein CotA shows laccase activity. By using comparative modeling techniques, we were able to derive a model for CotA based on the known x-ray structures of zucchini ascorbate oxidase and Cuprinus cereneus laccase. This model of CotA contains all the structural features of a laccase, including the reactive surface-exposed copper center (T1) and two buried copper centers (T2 and T3). Single amino acid substitutions in the CotA T1 copper center (H497A, or M502L) did not prevent assembly of the mutant proteins into the coat and did not alter the pattern of extractable coat polypeptides. However, in contrast to a wild type strain, both mutants produced unpigmented colonies and spores unable to oxidize syringaldazine (SGZ) and 22-azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS). The CotA protein was purified to homogeneity from an overproducing Escherichia coli strain. The purified CotA shows an absorbance and a EPR spectra typical of blue multicopper oxidases. Optimal enzymatic activity was found at

Heme-copper oxygen reductases (HCO) reduce O(2) to water being the last enzymatic complexes of most aerobic respiratory chains. These enzymes promote energy conservation coupling the catalytic reaction to charge separation and charge translocation across the prokaryotic cytoplasmatic or mitochondrial membrane. In this way they contribute to the establishment and maintenance of the transmembrane difference of electrochemical potential, which is vital for solute/nutrient cell import, synthesis of ATP and motility. The HCO enzymes most probably share with the nitric oxide reductases, NORs, a common ancestor. We have proposed the classification of HCOs into three different types, A, B and C; based on the constituents of their proton channels (Pereira, Santana and Teixeira (2001) Biochim Biophys Acta, 1505, 185-208). This classification was recently challenged by the suggestion of other different types of HCOs. Using an enlarged sampling we performed an exhaustive bioinformatic reanalysis of HCOs family. Our results strengthened our previously proposed classification and showed no need for the existence of more divisions. Now, we analyze the taxonomic distribution of HCOs and NORs and the congruence of their sequence trees with the 16S rRNA tree. We observed that HCOs are widely distributed in the two prokaryotic domains and that the different types of enzymes are not confined to a specific taxonomic group or environmental niche.

The copper content of recombinant CotA laccase from Bacillus subtilis produced by Escherichia coli cells is shown to be strongly dependent on the presence of copper and oxygen in the culture media. In copper-supplemented media, a switch from aerobic to microaerobic conditions leads to the synthesis of a recombinant holoenzyme, while the maintenance of aerobic conditions results in the synthesis of a copper-depleted population of proteins. Strikingly, cells grown under microaerobic conditions accumulate up to 80-fold more copper than aerobically grown cells. In vitro copper incorporation into apoenzymes was monitored by optical and electron paramagnetic resonance (EPR) spectroscopy. This analysis reveals that copper incorporation into CotA laccase is a sequential process, with the type 1 copper center being the first to be reconstituted, followed by the type 2 and the type 3 copper centers. The copper reconstitution of holoCotA derivatives depleted in vitro with EDTA results in the complete recovery of the native conformation as monitored by spectroscopic, kinetic and thermal stability analysis. However, the reconstitution of copper to apo forms produced in cultures under aerobic and copper-deficient conditions resulted in incomplete recovery of biochemical properties of the holoenzyme. EPR and resonance Raman data indicate that, presumably, folding in the presence of copper is indispensable for the correct structure of the trinuclear copper-containing site.

A sulfide:quinone oxidoreductase (SQR) was isolated from the membranes of the hyperthermoacidophilic archaeon Acidianus ambivalens, and its X-ray structure, the first reported for an SQR, was determined to 2.6 A resolution. This enzyme was functionally and structurally characterized and was shown to have two redox active sites: a covalently bound FAD and an adjacent pair of cysteine residues. Most interestingly, the X-ray structure revealed the presence of a chain of three sulfur atoms bridging those two cysteine residues. The possible implications of this observation in the catalytic mechanism for sulfide oxidation are discussed, and the role of SQR in the sulfur dependent bioenergetics of A. ambivalens, linked to oxygen reduction, is addressed.

Complex I of respiratory chains is an energy transducing enzyme present in most bacteria, mitochondria and chloroplasts. It catalyzes the oxidation of NADH and the reduction of quinones, coupled to cation translocation across the membrane. The complex has a modular structure composed of several proteins most of which are identified in other complexes. Close relations between complex I and group 4 membrane-bound [NiFe] hydrogenases and some subunits of multiple resistance to pH (Mrp) Na(+)/H(+) antiporters have been observed before and the suggestion that complex I arose from the association of a soluble nicotinamide adenine dinucleotide (NAD(+)) reducing hydrogenase with a Mrp-like antiporter has been put forward. In this article we performed a thorough taxonomic profile of prokaryotic group 4 membrane-bound [NiFe] hydrogenases, complexes I and complex I-like enzymes. In addition we have investigated the different gene clustering organizations of such complexes. Our data show the presence of complexes related to hydrogenases but which do not contain the binding site of the catalytic centre. These complexes, named before as Ehr (energy-converting hydrogenases related complexes) are a missing link between complex I and group 4 membrane-bound [NiFe] hydrogenases. Based on our observations we put forward a different perspective for the relation between complex I and related complexes. In addition we discuss the evolutionary, functional and mechanistic implications of this new perspective. This article is part of a Special Issue entitled: The evolutionary aspects of bioenergetic systems.

A comprehensive phylogenetic analysis of the core subunits of succinate:quinone oxidoreductases and quinol:fumarate oxidoreductases is performed, showing that the classification of the enzymes as type A to E based on the type of the membrane anchor fully correlates with the specific characteristics of the two core subunits. A special emphasis is given to the type E enzymes, which have an atypical association to the membrane, possibly involving anchor subunits with amphipathic helices. Furthermore, the redox properties of the SQR/QFR proteins are also reviewed, stressing out the recent observation of redox-Bohr effect upon haem reduction, observed for the Desulfovibrio gigas and Rhodothermus marinus enzymes, which indicates a direct protonation event at the haems or at a nearby residue. Finally, the possible contribution of these enzymes to the formation/dissipation of a transmembrane proton gradient is discussed, considering recent experimental and structural data.

SummaryA prerequisite for any rational drug design strategy is understanding the mode of protein-ligand interaction. This motivated us to explore protein-substrate interaction in Type-II NADH:quinone oxidoreductase (NDH-2) from Staphylococcus aureus, a worldwide problem in clinical medicine due to its multiple drug resistant forms. NDHs-2 are involved in respiratory chains and recognized as suitable targets for novel antimicrobial therapies, as these are the only enzymes with NADH:quinone oxidoreductase activity expressed in many pathogenic organisms.We obtained crystal and solution structures of NDH-2 from S. aureus, showing that it is a dimer in solution. We report fast kinetic analyses of the protein and detected a charge-transfer complex formed between NAD + and the reduced flavin, which is dissociated by the quinone. We observed that the quinone reduction is the rate limiting step and also the only half-reaction affected by the presence of HQNO, an inhibitor. We analyzed protein-substrate interactions by fluorescence and STD-NMR spectroscopies, which indicate that NADH and the quinone bind to different sites. In summary, our combined results show the presence of distinct binding sites for the two substrates, identified quinone reduction as the rate limiting step and indicate the establishment of a NAD + -protein complex, which is released by the quinone.