Structure and function of the radical enzyme ribonucleotide reductase

Ribonucleotide reductases (RNRs) catalyze all new production in nature of deoxyribonucleotides for DNA synthesis by reducing the corresponding ribonucleotides. The reaction involves the action of a radical that is produced differently for different classes of the enzyme. Class I enzymes, which are present in eukaryotes and microorganisms, use an iron center to produce a stable tyrosyl radical that is stored in one of the subunits of the enzyme. The other classes are only present in microorganisms. Class II enzymes use cobalamin for radical generation and class III enzymes, which are found only in anaerobic organisms, use a glycyl radical. The reductase activity is in all three classes contained in enzyme subunits that have similar structures containing active site cysteines. The initiation of the reaction by removal of the 3′-hydrogen of the ribose by a transient cysteinyl radical is a common feature of the different classes of RNR. This cysteine is in all RNRs located on the tip of a finger loop inserted into the center of a special barrel structure. A wealth of structural and functional information on the class I and class III enzymes can now give detailed views on how these enzymes perform their task. The class I enzymes demonstrate a sophisticated pattern as to how the free radical is used in the reaction, in that it is only delivered to the active site at exactly the right moment. RNRs are also allosterically regulated, for which the structural molecular background is now starting to be revealed.     

Structural basis of different substrate preferences of two old yellow enzymes from yeasts in the asymmetric reduction of enone compounds

Old yellow enzymes (OYEs) are potential targets of protein engineering for useful biocatalysts because of their excellent asymmetric reductions of enone compounds. Two OYEs from different yeast strains, Candida macedoniensis AKU4588 OYE (CmOYE) and Pichia sp. AKU4542 OYE (PsOYE), have a sequence identity of 46%, but show different substrate preferences; PsOYE shows 3.4-fold and 39-fold higher catalytic activities than CmOYE toward ketoisophorone and (4S)-phorenol, respectively. To gain insights into structural basis of their different substrate preferences, we have solved a crystal structure of PsOYE, and compared its catalytic site structure with that of CmOYE, revealing the catalytic pocket of PsOYE is wider than that of CmOYE due to different positions of Phe²⁴⁶ (PsOYE)/Phe²⁵⁰ (CmOYE) in static Loop 5. This study shows a significance of 3D structural information to explain the different substrate preferences of yeast OYEs which cannot be understood from their amino acid sequences.

Abbreviations: OYE: Old yellow enzymes, CmOYE: Candida macedoniensis AKU4588 OYE, PsOYE: Pichia sp. AKU4542 OYE     

A homolog of old yellow enzyme in tomato. Spectral properties and substrate specificity of the recombinant protein

A cDNA was isolated and characterized from a tomato shoot cDNA library, the deduced amino acid sequence of which exhibited similarity with yeast Old Yellow Enzymes (OYEs) and related enzymes of bacterial and plant origin. Sequence identity was particularly high with 12-oxophytodienoate 10,11-reductase (OPR) from Arabidopsis thaliana. The cDNA-encoded protein was expressed as a glutathione S-transferase fusion protein in Escherichia coli and was purified from bacterial extracts. The protein was found to be a flavoprotein catalyzing the NADPH-dependent reduction of the olefinic bond of alpha,beta-unsaturated carbonyl compounds, including 12-oxophytodienoic acid. Thus, the tomato enzyme was termed LeOPR. The catalytic efficiency of LeOPR was highest with N-ethylmaleimide followed by 12-oxophytodienoic acid and maleic acid as substrates. Photoreduction of the LeOPR-bound FMN resulted in the formation of a red, anionic semiquinone prior to the formation of the fully reduced flavin dihydroquinone. Spectroscopic characterization of LeOPR revealed the formation of charge transfer complexes upon titration with para-substituted phenolic compounds, a distinctive feature of the enzymes of the OYE family. The ligand binding properties were compared between LeOPR and OYE, and the findings are discussed with respect to structural differences between the active sites of OYE and LeOPR.     

The Evolution of Acetyl-CoA Synthase

Acetyl-coenzyme A synthases (ACS) are Ni-Fe-S containing enzymes found in archaea and bacteria. They are divisible into 4 classes. Class I ACS's catalyze the synthesis of acetyl-CoA from CO2 + 2e-, CoA, and a methyl group, and contain 5 types of subunits (alpha, beta, gamma, delta, and epsilon). Class II enzymes catalyze essentially the reverse reaction and have similar subunit composition. Class III ACS's catalyze the same reaction as Class I enzymes, but use pyruvate as a source of CO2 and 2e-, and are composed of 2 autonomous proteins, an alpha 2 beta 2 tetramer and a gamma delta heterodimer. Class IV enzymes catabolize CO to CO2 and are alpha-subunit monomers. Phylogenetic analyses were performed on all five subunits. ACS alpha sequences divided into 2 major groups, including Class I/II sequences and Class III/IV-like sequences. Conserved residues that may function as ligands to the B- and C-clusters were identified. Other residues exclusively conserved in Class I/II sequences may be ligands to additional metal centers in Class I and II enzymes. ACS beta sequences also separated into two groups, but they were less divergent than the alpha's, and the separation was not as distinct. Class III-like beta sequences contained approximately 300 residues at their N-termini absent in Class I/II sequences. Conserved residues identified in beta sequences may function as ligands to active site residues used for acetyl-CoA synthesis. ACS gamma-sequences separated into 3 groups (Classes I, II, and III), while delta-sequences separated into 2 groups (Class I/II and III). These groups are less divergent than those of alpha sequences. ACS epsilon-sequence topology showed greater divergence and less consistency vis-à-vis the other subunits, possibly reflecting reduced evolutionary constraints due to the absence of metal centers. The alpha subunit phylogeny may best reflect the functional diversity of ACS enzymes. Scenarios of how ACS and ACS-containing organisms may have evolved are discussed.     

Ligand-induced conformational changes in the capping subdomain of a bacterial old yellow enzyme homologue and conserved sequence fingerprints provide new insights into substrate binding

We have recently reported that Shewanella oneidensis, a Gram-negative gamma-proteobacterium with a rich arsenal of redox proteins, possesses four old yellow enzyme (OYE) homologues. Here, we report a series of high resolution crystal structures for one of these OYEs, Shewanella yellow enzyme 1 (SYE1), in its oxidized form at 1.4A resolution, which binds a molecule of PEG 400 in the active site, and in its NADH-reduced and p-hydroxybenzaldehyde- and p-hydroxyacetophenone-bound forms at 1.7A resolution. Although the overall structure of SYE1 reveals a monomeric enzyme based on the alpha(8)beta(8) barrel scaffold observed for other OYEs, the active site exhibits a unique combination of features: a strongly butterfly-bent FMN cofactor both in the oxidized and NADH-reduced forms, a collapsed and narrow active site tunnel, and a novel combination of conserved residues involved in the binding of phenolic ligands. Furthermore, we identify a second p-hydroxybenzaldehyde-binding site in a hydrophobic cleft next to the entry of the active site tunnel in the capping subdomain, formed by a restructuring of Loop 3 to an "open" conformation. This constitutes the first evidence to date for the entire family of OYEs that Loop 3 may indeed play a dynamic role in ligand binding and thus provides insights into the elusive NADH complex and into substrate binding in general. Structure-based sequence alignments indicate that the novelties we observe in SYE1 are supported by conserved residues in a number of structurally uncharacterized OYEs from the beta- and gamma-proteobacteria, suggesting that SYE1 represents a new subfamily of bacterial OYEs.     

Allosteric control of three B12-dependent (class II) ribonucleotide reductases. Implications for the evolution of ribonucleotide reduction

Three separate classes of ribonucleotide reductases are known, each with a distinct protein structure. One common feature of all enzymes is that a single protein generates each of the four deoxyribonucleotides. Class I and III enzymes contain an allosteric substrate specificity site capable of binding effectors (ATP or various deoxyribonucleoside triphosphates) that direct enzyme specificity. Some (but not all) enzymes contain a second allosteric site that binds only ATP or dATP. Binding of dATP to this site inhibits the activity of these enzymes. X-ray crystallography has localized the two sites within the structure of the Escherichia coli class I enzyme and identified effector-binding amino acids. Here, we have studied the regulation of three class II enzymes, one from the archaebacterium Thermoplasma acidophilum and two from eubacteria (Lactobacillus leichmannii and Thermotoga maritima). Each enzyme has an allosteric site that binds ATP or various deoxyribonucleoside triphosphates and that regulates its substrate specificity according to the same rules as for class I and III enzymes. dATP does not inhibit enzyme activity, suggesting the absence of a second active allosteric site. For the L. leichmannii and T. maritima enzymes, binding experiments also indicate the presence of only one allosteric site. Their primary sequences suggest that these enzymes lack the structural requirements for a second site. In contrast, the T. acidophilum enzyme binds dATP at two separate sites, and its sequence contains putative effector-binding amino acids for a second site. The presence of a second site without apparent physiological function leads to the hypothesis that a functional site was present early during the evolution of ribonucleotide reductases, but that its function was lost from the T. acidophilum enzyme. The other two B12 enzymes lost not only the function, but also the structural basis for the site. Also a large subgroup (Ib) of class I enzymes, but none of the investigated class III enzymes, has lost this site. This is further indirect evidence that class II and I enzymes may have arisen by divergent evolution from class III enzymes.     

Pathotype-specific genetic factors in chickpea (Cicer arietinum L.) for quantitative resistance to ascochyta blight

Ascochyta blight in chickpea (Cicer arietinum L.) is a devastating fungal disease caused by the necrotrophic pathogen, Ascochyta rabiei (Pass.) Lab. To elucidate the genetic mechanism of pathotype-dependent blight resistance in chickpea, F₇-derived recombinant inbred lines (RILs) from the intraspecific cross of PI 359075(1) (blight susceptible) × FLIP84-92C(2) (blight resistant) were inoculated with pathotypes I and II of A. rabiei. The pattern of blight resistance in the RIL population varied depending on the pathotype of A. rabiei. Using the same RIL population, an intraspecific genetic linkage map comprising 53 sequence-tagged microsatellite site markers was constructed. A quantitative trait locus (QTL) for resistance to pathotype II of A. rabiei and two QTLs for resistance to pathotype I were identified on linkage group (LG)4A and LG2+6, respectively. A putative single gene designated as Ar19 (or Ar21d) could explain the majority of quantitative resistance to pathotype I. Ar19 (or Ar21d) appeared to be required for resistance to both pathotypes of A. rabiei, and the additional QTL on LG4A conferred resistance to pathotype II of A. rabiei. Further molecular genetic approach is needed to identify individual qualitative blight resistance genes and their interaction for pathotype-dependent blight resistance in chickpea.     

Cyclin-dependent kinase 5 phosphorylates mammalian HMGB1 protein only if acetylated

Old yellow enzyme (OYE) is an NADPH oxidoreductase capable of reducing a variety of compounds. It contains flavin mononucleotide (FMN) as a prosthetic group. A ternary complex structure of OYE from Trypanosoma cruzi (TcOYE) with FMN and one of the substrates, p-hydroxybenzaldehyde, shows a striking movement around the active site upon binding of the substrate. From a structural comparison of other OYE complexed with 12-oxophytodienoate, we have constructed a complex structure with another substrate, prostaglandin H(2) (PGH(2)), to provide a proposed stereoselective reaction mechanism for the reduction of PGH(2) to prostaglandin F(2α) by TcOYE.     

Ca-asp bound X-ray structure and inhibition of Bacillus anthracis dihydroorotase (DHOase)

Dihydroorotase (DHOase) is the third enzyme in the de novo pyrimidine synthesis pathway and is responsible for the reversible cyclization of carbamyl-aspartate (Ca-asp) to dihydroorotate (DHO). DHOase is further divided into two classes based on several structural characteristics, one of which is the length of the flexible catalytic loop that interacts with the substrate, Ca-asp, regulating the enzyme activity. Here, we present the crystal structure of Class I Bacillus anthracis DHOase with Ca-asp in the active site, which shows the peptide backbone of glycine in the shorter loop forming the necessary hydrogen bonds with the substrate, in place of the two threonines found in Class II DHOases. Despite the differences in the catalytic loop, the structure confirms that the key interactions between the substrate and active site residues are similar between Class I and Class II DHOase enzymes, which we further validated by mutagenesis studies. B. anthracis DHOase is also a potential antibacterial drug target. In order to identify prospective inhibitors, we performed high-throughput screening against several libraries using a colorimetric enzymatic assay and an orthogonal fluorescence thermal binding assay. Surface plasmon resonance was used for determining binding affinity (K_D) and competition analysis with Ca-asp. Our results highlight that the primary difference between Class I and Class II DHOase is the catalytic loop. We also identify several compounds that can potentially be further optimized as potential B. anthracis inhibitors.     

Characteristics of mammalian class III alcohol dehydrogenases, an enzyme less variable than the traditional liver enzyme of class I

Class III alcohol dehydrogenase, whose activity toward ethanol is negligible, has defined, specific properties and is not just a "variant" of the class I protein, the traditional liver enzyme. The primary structure of the horse class III protein has now been determined, and this allows the comparison of alcohol dehydrogenases from human, horse, and rat for both classes III and I, providing identical triads for both these enzyme types. Many consistent differences between the classes separate the two forms as distinct enzymes with characteristic properties. The mammalian class III enzymes are much less variable in structure than the corresponding typical liver enzymes of class I: there are 35 versus 84 positional differences in these identical three-species sets. The class III and class I subunits contain four versus two tryptophan residues, respectively. This makes the differences in absorbance at 280 nm a characteristic property. There are also 4-6 fewer positive charges in the class III enzymes accounting for their electrophoretic differences. The substrate binding site of class III differs from that of class I by replacements at positions that form the hydrophobic barrel typical for this site. In class III, two to four of these positions contain residues with polar or even charged side chains (positions 57 and 93 in all species, plus positions 116 in the horse and 140 in the human and the horse), while corresponding intraclass variation is small. All these structural features correlate with functional characteristics and suggest that the enzyme classes serve different roles. In addition, the replacements between these triad sets illustrate further general properties of the two mammalian alcohol dehydrogenase classes.(ABSTRACT TRUNCATED AT 250 WORDS)     

Discovery and Characterization of FMN-Binding β-Glucuronidases in the Human Gut Microbiome

The human gut microbiota encodes β-glucuronidases (GUSs) that play key roles in health and disease via the metabolism of glucuronate-containing carbohydrates and drugs. Hundreds of putative bacterial GUS enzymes have been identified by metagenomic analysis of the human gut microbiome, but less than 10% have characterized structures and functions. Here we describe a set of unique gut microbial GUS enzymes that bind flavin mononucleotide (FMN). First, we show using mass spectrometry, isothermal titration calorimetry, and x-ray crystallography that a purified GUS from the gut commensal microbe Faecalibacterium prausnitzii binds to FMN on a surface groove located 30 Å away from the active site. Second, utilizing structural and functional data from this FMN-binding GUS, we analyzed the 279 unique GUS sequences from the Human Microbiome Project database and identified 14 putative FMN-binding GUSs. We characterized four of these hits and solved the structure of two, the GUSs from Ruminococcus gnavus and Roseburia hominis, which confirmed that these are FMN binders. Third, binding and kinetic analysis of the FMN-binding site mutants of these five GUSs show that they utilize a conserved site to bind FMN that is not essential for GUS activity, but can affect K_M. Lastly, a comprehensive structural review of the PDB reveals that the FMN-binding site employed by these enzymes is unlike any structurally characterized FMN binders to date. These findings reveal the first instance of an FMN-binding glycoside hydrolase and suggest a potential link between FMN and carbohydrate metabolism in the human gut microbiota.     

Characterization of glycerol trinitrate reductase (NerA) and the catalytic role of active-site residues

Glycerol trinitrate reductase (NerA) from Agrobacterium radiobacter, a member of the old yellow enzyme (OYE) family of oxidoreductases, was expressed in and purified from Escherichia coli. Denaturation of pure enzyme liberated flavin mononucleotide (FMN), and spectra of NerA during reduction and reoxidation confirmed its catalytic involvement. Binding of FMN to apoenzyme to form the holoenzyme occurred with a dissociation constant of ca. 10(-7) M and with restoration of activity. The NerA-dependent reduction of glycerol trinitrate (GTN; nitroglycerin) by NADH followed ping-pong kinetics. A structural model of NerA based on the known coordinates of OYE showed that His-178, Asn-181, and Tyr-183 were close to FMN in the active site. The NerA mutation H178A produced mutant protein with bound FMN but no activity toward GTN. The N181A mutation produced protein that did not bind FMN and was isolated in partly degraded form. The mutation Y183F produced active protein with the same k(cat) as that of wild-type enzyme but with altered K(m) values for GTN and NADH, indicating a role for this residue in substrate binding. Correlation of the ratio of K(m)(GTN) to K(m)(NAD(P)H), with sequence differences for NerA and several other members of the OYE family of oxidoreductases that reduce GTN, indicated that Asn-181 and a second Asn-238 that lies close to Tyr-183 in the NerA model structure may influence substrate specificity.     

Phosphate group binding "cup" of PLP-dependent and non-PLP-dependent enzymes: leitmotif and variations

Pyridoxal-5'-phosphate (PLP) is widely used by many enzymes in reactions where amino acids are interconverted. Whereas the role of the pyridoxal ring in catalysis is well understood, the functional role of the single phosphate group in PLP has been less studied. Here we construct unambiguous connection diagrams that describe the interactions among the three non-ester phosphate oxygen atoms of PLP and surrounding atoms from the protein binding site and from water molecules, the so-called phosphate group binding "cup". These diagrams provide a simple means to identify common recognition motifs for the phosphate group in both similar and different protein folds. Diagrams were constructed and compared in the cases of five newly determined structures of PLP-dependent transferases (fold type I enzymes) and, additionally, two non-PLP protein complexes (indole-3-glycerol phosphate synthase (IGPS) with bound indole-3-glycerol phosphate (IGP) and old yellow enzyme (OYE) with bound flavin mononucleotide (FMN)). A detailed comparison of the diagrams shows that three positions out of ten in the structure of the phosphate group binding "cup" contain invariant atoms, while seven others are occupied by conserved atom types. This level of similarity was also observed in the fold type III (TIM beta/alpha-barrel) enzymes that bind three different ligands: PLP, IGP and FMN.     

New spectroscopic and electrochemical insights on a class I superoxide reductase: evidence for an intramolecular electron-transfer pathway

SORs (superoxide reductases) are enzymes involved in bacterial resistance to reactive oxygen species, catalysing the reduction of superoxide anions to hydrogen peroxide. So far three structural classes have been identified. Class I enzymes have two iron-centre-containing domains. Most studies have focused on the catalytic iron site (centre II), yet the role of centre I is poorly understood. The possible roles of this iron site were approached by an integrated study using both classical and fast kinetic measurements, as well as direct electrochemistry. A new heterometallic form of the protein with a zinc-substituted centre I, maintaining the iron active-site centre II, was obtained, resulting in a stable derivative useful for comparison with the native all-iron from. Second-order rate constants for the electron transfer between reduced rubredoxin and the different SOR forms were determined to be 2.8 × 10? M?1 · s?1 and 1.3 × 10? M?1 · s?1 for SORFe(IIII)-Fe(II) and for SORFe(IIII)-Fe(III) forms respectively, and 3.2 × 10? M?1 · s?1 for the SORZn(II)-Fe(III) form. The results obtained seem to indicate that centre I transfers electrons from the putative physiological donor rubredoxin to the catalytic active iron site (intramolecular process). In addition, electrochemical results show that conformational changes are associated with the redox state of centre I, which may enable a faster catalytic response towards superoxide anion. The apparent rate constants calculated for the SOR-mediated electron transfer also support this observation.     

Role of active site residues and solvent in proton transfer and the modulation of flavin reduction potential in bacterial morphinone reductase

The reactions of several active site mutant forms of bacterial morphinone reductase (MR) with NADH and 2-cyclohexen-1-one as substrates have been studied by stopped-flow and steady-state kinetic methods and redox potentiometry. The enzymes were designed to (i) probe a role for potential proton donors (Tyr-72 and Tyr-356) in the oxidative half-reaction of MR; (ii) assess the function of a highly conserved tryptophan residue (Trp-106) in catalysis; (iii) investigate the role of Thr-32 in modulating the FMN reduction potential and catalysis. The Y72F and Y356F enzymes retained activity in both steady-state and stopped-flow kinetic studies, indicating they do not serve as key proton donors in the oxidative reaction of MR. Taken together with our recently published data (Messiha, H. L., Munro, A. W., Bruce, N. C., Barsukov, I., and Scrutton, N. S. (2005) J. Biol. Chem. 280, 4627-4631) that rule out roles for Cys-191 (corresponding with the proton donor, Tyr-196, in the structurally related OYE1 enzyme) and His-186 as proton donors, we infer solvent is the source of the proton in the oxidative half-reaction of MR. We demonstrate a key role for Thr-32 in modulating the reduction potential of the FMN, which is decreased approximately 50 mV in the T32A mutant MR. This effects a change in rate-limiting step in the catalytic cycle of the T32A enzyme with the oxidizing substrate 2-cyclohexenone. Despite the conservation of Trp-106 throughout the OYE family, we show this residue does not play a major role in catalysis, although affects on substrate and coenzyme binding are observed in a W106F enzyme. Our studies show some similarities, but also major differences, in the catalytic mechanism of MR and OYE1, and emphasize the need for caution in inferring mechanism by structural comparison of highly related enzymes in the absence of solution studies.     

Fractionation of RNA polymerase subspecies from soybean hypocotyls by isoelectric focusing in Sephadex

RNA polymerase enzymes isolated from chromatin of 6-day-old soybean hypocotyls are resolved into two major and one minor species of activity by DEAE-Sephadex chromatography. A soluble form of the enzyme, isolated from the postchromatin supernatant fraction, shows a broad peak of activity when fractionated by this method. The elution characteristics and α-amanitin sensitivity data indicate the two major chromatin-bound activities to be Class I and III enzymes, while the minor chromatin-bound activity and the soluble enzyme are representative of the Class II enzymes. In contrast to these profiles, fractionation of these enzyme preparations by the new method of isoelectric focusing in Sephadex G-15 yields five distinct chromatin-bound and four soluble subspecies. The relationships of these observed activities to their parent DEAE classes are investigated, showing two subspecies within the Class I and III RNA polymerase enzymes, respectively, and four subspecies within the Class II enzymes.     

Efficacy ofTrichoderma harzianum (Rifaii) on inhibition of ascochyta blight disease of chickpea

Trichoderma harzianum is a widely distributed soil fungus that antagonies numerous fungal phytopathogens. In this study, interactions between theT. harzianum isolates andAscochyta rabiei in experiments on agar growth medium were studied. All testedT. harzianum isolates produced metabolite that inhibited growth ofA. rabiei the agent of ascochyta blight disease of chickpea in culture. Isolates ofT. harzianum produced chitinase and β-1,3-glucanase when grown in liquid cultures containingA. rabiel cell wall, laminarin and chitin as sole carbon sources. Levels higher of these enzymes were induced inT. harzianum T15 isolate.     

Crystal structure of bacterial morphinone reductase and properties of the C191A mutant enzyme

The crystal structure of the NADH-dependent bacterial flavoenzyme morphinone reductase (MR) has been determined at 2.2-A resolution in complex with the oxidizing substrate codeinone. The structure reveals a dimeric enzyme comprising two 8-fold beta/alpha barrel domains, each bound to FMN, and a subunit folding topology and mode of flavin-binding similar to that found in Old Yellow Enzyme (OYE) and pentaerythritol tetranitrate (PETN) reductase. The subunit interface of MR is formed by interactions from an N-terminal beta strand and helices 2 and 8 of the barrel domain and is different to that seen in OYE. The active site structures of MR, OYE, and PETN reductase are highly conserved reflecting the ability of these enzymes to catalyze "generic" reactions such as the reduction of 2-cyclohexenone. A region of polypeptide presumed to define the reducing coenzyme specificity is identified by comparison of the MR structure (NADH-dependent) with that of PETN reductase (NADPH-dependent). The active site acid identified in OYE (Tyr-196) and conserved in PETN reductase (Tyr-186) is replaced by Cys-191 in MR. Mutagenesis studies have established that Cys-191 does not act as a crucial acid in the mechanism of reduction of the olefinic bond found in 2-cyclohexenone and codeinone.     

Ribonucleotide reduction in Pseudomonas species: simultaneous presence of active enzymes from different classes.

Three separate classes of ribonucleotide reductases exist in nature. They differ widely in protein structure. Class I enzymes are found in aerobic bacteria and eukaryotes; class II enzymes are found in aerobic and anaerobic bacteria; class III enzymes are found in strict and facultative anaerobic bacteria. Usually, but not always, one organism contains only one or two (in facultative anaerobes) classes. Surprisingly, the genomic sequence of Pseudomonas aeruginosa contains sequences for each of the three classes. Here, we show by DNA hybridization that other species of Pseudomonas also contain the genes for three classes. Extracts from P. aeruginosa and P. stutzeri grown aerobically or microaerobically contain active class I and II enzymes, whereas we could not demonstrate class III activity. Unexpectedly, class I activity increased greatly during microaerobic conditions. The enzymes were separated, and the large proteins of the class I enzymes were obtained in close to homogeneous form. The catalytic properties of all enzymes are similar to those of other bacterial reductases. However, the Pseudomonas class I reductases required the continuous presence of oxygen during catalysis, unlike the corresponding Escherichia coli enzyme but similar to the mouse enzyme. In similarity searches, the amino acid sequence of the class I enzyme of P. aeruginosa was more related to that of eukaryotes than to that of E. coli or other proteobacteria, with the large protein showing 42% identity to that of the mouse, suggesting the possibility of a horizontal transfer of the gene. The results raise many questions concerning the physiological function and evolution of the three classes in Pseudomonas species.