Spinach glycolate oxidase and yeast flavocytochrome b2 are structurally homologous and evolutionarily related enzymes with distinctly different function and flavin mononucleotide binding

Crystal structure of Tritrichomonas foetus inosine-5'-monophosphate dehydrogenase and the enzyme-product complex

Inosine-5'-monophosphate dehydrogenase (IMPDH) is an attractive drug target for the control of parasitic infections. The enzyme catalyzes the oxidation of inosine monophosphate (IMP) to xanthosine monophosphate (XMP), the committed step in de novo guanosine monophosphate (GMP) biosynthesis. We have determined the crystal structures of IMPDH from the protozoan parasite Tritrichomonas foetus in the apo form at 2.3 A resolution and the enzyme-XMP complex at 2.6 A resolution. Each monomer of this tetrameric enzyme is comprised of two domains, the largest of which includes an eight-stranded parallel beta/alpha-barrel that contains the enzyme active site at the C termini of the barrel beta-strands. A second domain, comprised of residues 102-220, is disordered in the crystal. IMPDH is expected to be active as a tetramer, since the active site cavity is formed by strands from adjacent subunits. An intrasubunit disulfide bond, seen in the crystal structure, may stabilize the protein in a less active form, as high concentrations of reducing agent have been shown to increase enzyme activity. Disorder at the active site suggests that a high degree of flexibility may be inherent in the catalytic function of IMPDH. Unlike IMPDH from other species, the T. foetus enzyme has a single arginine that is largely responsible for coordinating the substrate phosphate in the active site. This structural uniqueness may facilitate structure-based identification and design of compounds that specifically inhibit the parasite enzyme.

Resonance Raman study on the oxidized and anionic semiquinone forms of flavocytochrome b2 and L-lactate monooxygenase. Influence of the structure and environment of the isoalloxazine ring on the flavin function

The oxidized and semiquinone anion radical forms of flavin mononucleotide carried by flavocytochrome b2 and L-lactate monooxygenase have been studied by resonance Raman (RR) spectroscopy. The RR spectra of their oxidized forms are compared with previously published RR data on various flavins and flavoproteins. Taking as a support available X-ray crystallographic data on flavoproteins, we have found correlations between the frequencies of RR bands II (1575-1588 cm-1), III (1534-1557 cm-1), and X (1244-1266 cm-1) and the H-bonding environment and/or the structure of the flavin ring. The present RR data provide strong evidence that the electron density, the conformation, and the H-bonding environment of the oxidized flavin mononucleotide of flavocytochrome b2 and L-lactate monooxygenase are different. As far as the anionic semiquinone form of flavoproteins is concerned, the behavior of two bands observed at 1280-1300 and 1320-1350 cm-1 suggests that they have vibrational origins similar to those of RR bands II and III of oxidized compounds. On this basis, the differences in conformation and H-bonding environment of the isoalloxazine ring, observed for the oxidized form of flavocytochrome b2 and L-lactate monooxygenase, appear to be preserved upon one-electron reduction of the flavin. For both flavoproteins, the RR spectra of the semiquinone form are affected by pyruvate binding. The data are interpreted in the frame of a change in H-bonding interaction of the C4&dbd;O carbonyl group of the flavin without significant alteration of the isoalloxazine conformation. This modification in electrostatic interaction quantitatively accounts for the pyruvate-induced changes of the oxidized/semiquinone and semiquinone/reduced redox potentials of the flavoproteins. Considering the high homology in the flavin catalytic sites of flavocytochrome b2 and L-lactate monooxygenase, the observed differences in H-bonding environment and conformation of the FMN ring are related to the different biological functions of the two flavoproteins.

Molecular interpretation of inhibition by excess substrate in flavocytochrome b2: a study with wild-type and Y143F mutant enzymes

The crystal structure of flavocytochrome b2 (L-lactate dehydrogenase) from Saccharomyces cerevisiae suggests that Tyr143 plays a dual role at the active site: it contributes to substrate binding and, most importantly, makes a hydrogen bond to a heme propionate, which could facilitate communication between the domains. Previous work on the Y143F mutant enzyme provided support for these hypotheses [Miles, C. S., Rouvière-Fourmy, N., Lederer, F., Mathews, F. S., Reid, G. A., Black, M. T., & Chapman, S. K. (1992) Biochem. J. 285, 187-192; Rouvière-Fourmy, N., Capeillère-Blandin, C., & Lederer, F. (1994) Biochemistry 33, 798-806]. In the course of kinetic comparisons between the wild-type (WT) enzyme and the Y143F mutant protein, we observed for the latter signs of inhibition by excess substrate at much lower concentrations than observed for the former. A detailed investigation of the phenomenon has shown that, for the wild-type and Y143F forms, lactate at high concentrations inhibits both cytochrome c and ferricyanide reduction. In these cases, inhibition appears to be a specific effect, since acetate at identical concentrations exerts an inhibitory effect that is markedly weaker than that of lactate. In the pre-steady-state, in the absence of acceptor, flavin and heme reduction are unaffected by high substrate concentrations in the WT enzyme case. For the Y143F mutant, flavin reduction is similarly unaffected, but heme reduction is inhibited to nearly the same extent by high lactate and acetate concentrations. In this case, inhibition can probably be ascribed to ionic strength effects. The combination of stopped-flow and steady-state results suggests that lactate binds with weak affinity at the active site when the flavin is in the semiquinone state, preventing electron transfer to heme b2 and hence to acceptors. This phenomenon is analogous to the inhibition exerted by pyruvate when bound to the enzyme at the semiquinone stage [Tegoni, M., Janot, J. M., & Labeyrie, F. (1990) Eur. J. Biochem. 190, 329-342]. We suggest that the substrate carboxylate and the heme propionate of the mobile heme-binding domain compete for the Tyr143 hydroxyl group, hence for approach to the flavin. In the Y143F mutant enzyme, in which the interdomain interaction is impaired, competition would play in favor of the substrate, resulting in the inhibition at lower lactate concentrations than observed for the wild-type enzyme.

Three-dimensional structures of glycolate oxidase with bound active-site inhibitors

A key step in plant photorespiration, the oxidation of glycolate to glyoxylate, is carried out by the peroxisomal flavoprotein glycolate oxidase (EC 1.1.3.15). The three-dimensional structure of this alpha/beta barrel protein has been refined to 2 A resolution (Lindqvist Y. 1989. J Mol Biol 209:151-166). FMN dependent glycolate oxidase is a member of the family of alpha-hydroxy acid oxidases. Here we describe the crystallization and structure determination of two inhibitor complexes of the enzyme, TKP (3-Decyl-2,5-dioxo-4-hydroxy-3-pyrroline) and TACA (4-Carboxy-5-(1-pentyl)hexylsulfanyl-1,2,3-triazole). The structure of the TACA complex has been refined to 2.6 A resolution and the TKP complex, solved with molecular replacement, to 2.2 A resolution. The Rfree for the TACA and TKP complexes are 24.2 and 25.1%, respectively. The overall structures are very similar to the unliganded holoenzyme, but a closer examination of the active site reveals differences in the positioning of the flavin isoalloxazine ring and a displaced flexible loop in the TKP complex. The two inhibitors differ in binding mode and hydrophobic interactions, and these differences are reflected by the very different Ki values for the inhibitors, 16 nM for TACA and 4.8 microM for TKP. Implications of the structures of these enzyme-inhibitor complexes for the model for substrate binding and catalysis proposed from the holo-enzyme structure are discussed.

The crystal structure of the flavin containing enzyme dihydroorotate dehydrogenase A from Lactococcus lactis

BACKGROUND

. Dihydroorotate dehydrogenase (DHOD) is a flavin mononucleotide containing enzyme, which catalyzes the oxidation of (S)-dihydroorotate to orotate, the fourth step in the de novo biosynthesis of pyrimidine nucleotides. Lactococcus lactis contains two genes encoding different functional DHODs whose sequences are only 30% identical. One of these enzymes, DHODA, is a highly efficient dimer, while the other, DHODB, shows optimal activity only in the presence of an iron-sulphur cluster containing protein with which it forms a complex tetramer. Sequence alignments have identified three different families among the DHODs: the two L. lactis enzymes belong to two of the families, whereas the enzyme from E. coli is a representative of the third. As no three-dimensional structures of DHODs are currently available, we set out to determine the crystal structure of DHODA from L. lactis. The differences between the two L. lactis enzymes make them particularly interesting for studying flavoprotein redox reactions and for identifying the differences between the enzyme families.

RESULTS

. The crystal structure of DHODA has been determined to 2.0 resolution. The enzyme is a dimer of two crystallographically independent molecules related by a non-crystallographic twofold axis. The protein folds into and alpha/beta barrel with the flavin molecule sitting between the top of the barrel and a subdomain formed by several barrel inserts. Above the flavin isoalloxazine ring there is a small water filled cavity, completely buried beneath the protein surface and surrounded by many conserved residues. This cavity is proposed as the substrate-binding site.

CONCLUSIONS

. The crystal structure has allowed the function of many of the conserved residues in DHODs to be identified: many of these are associated with binding the flavin group. Important differences were identified in some of the active-site residues which vary across the distinct DHOD families, implying significant mechanistic differences. The substrate cavity, although buried, is located beneath a highly conserved loop which is much less ordered than the rest of the protein and may be important in giving access to the cavity. The location of the conserved residues surrounding this cavity suggests the potential orientation of the substrate.

Functional properties of the histidine-aspartate ion pair of flavocytochrome b2 (L-lactate dehydrogenase): substitution of Asp282 with asparagine

The FMN prosthetic group of flavocytochrome b2 or L-lactate dehydrogenase oxidizes lactate to pyruvate. The reducing equivalents are then transferred one by one, intramolecularly, to heme b2 and then to external acceptors. Substrate oxidation is thought to begin with abstraction of the substrate alpha-hydrogen as a proton by an enzyme base. It has been proposed that this role is played by His373, which lies close to the flavin in the crystal structure and interacts with Asp282. It has also been shown before, using hydrogen exchange measurements, that the pKa of His373 is substantially increased in the wild-type reduced enzyme compared to that in the oxidized state. We report here the enzymatic properties of the D282N mutant flavocytochrome b2. Steady-state rate measurements with [2-1H]lactate and [2-2H]-lactate indicate that, as predicted, the Michaelis complex stability is hardly affected, whereas the transition state for proton abstraction increases in energy by 2.8 kcal/mol. Steady-state inhibition studies were conducted with a number of active-site ligands: sulfite, D-lactate, pyruvate, and oxalate. Binding was found to be most affected for oxalate, but kinetic patterns indicated oxalate and pyruvate were still capable of binding to the enzyme both at the oxidized and semiquinone stages, whereas inhibition by excess substrate, due to lactate binding at the semiquinone stage, was lost. Finally, analysis of the intermolecular hydrogen transfer catalyzed by the enzyme between [2-3H]lactate and fluoropyruvate indicated that the substitution with asparagine facilitates exchange of the histidine-bound proton and hence induces a decrease in the pKa value of H373 in the reduced enzyme of about 1.4 pH units. Nevertheless, the rate constant value for exchange with the solvent of the enzyme-bound substrate alpha-proton indicates that H373 is still protonated in the reduced mutant enzyme at neutral pH. Thus, the D282N mutation destabilizes the transition state for proton abstraction and decreases the pKa of H373 in the reduced enzyme but is insufficient to bring it back to a normal value.

The role of a beta barrel loop 4 extension in modulating the physical and functional properties of long-chain 2-hydroxy-acid oxidase isozymes

Peroxisomal long-chain 2-hydroxy-acid oxidase, an FMN-dependent enzyme, catalyzes the oxidation of a variety of L-2-hydroxy acids into keto acids at the expense of oxygen. We recently reported the cloning and sequencing of its CDNA and the existence of a weakly expressed isozyme [Belmouden, A., Le, K. H. D., Lederer, F. & Garchon, H. J. (1993) Eur. J. Biochem. 214, 17-251. This isozyme, beta 2 differs from the major one in having a three-residue insertion, -VRK-, in loop 4 of the beta 8 alpha 8 barrel. In the crystal structures of homologous flavocytochrome beta 2, and glycolate oxidase, the corresponding region of loop 4 is disordered. We now report on the constitutive high-level expression of isozymes beta 1, and beta 2 in Escherichia coli under control of the lambda pL promoter, and on the influence of the E. coli genetic background and the growth medium on the expression level. We describe the properties of isozyme beta 2 and compare them with those of pure isoform beta 1. The visible spectra of the purified enzymes differ in the position of the near-ultraviolet band of the prosthetic group. pH titration studies indicate that the FMN ionizes at N3 at a lower pH than free flavin and that there is a small pKa difference between the isozymes. To our knowledge, the only other known case of a lowered pKa for the protein-bound flavin is that of glycolate oxidase. In the CD spectra of the FMN region, a marked difference between isozymes in the 270-300-nm region appears to be related to the pKa difference for the N3-H bond. Kinetic parameters for a number of substrates and inhibitors are indistinguishable within the limits of experimental error, with the exception of values for kcat for mandelate (the most active substrate), Km for hydroxyhippurate (a new substrate), Ki for cinnamate and oxalate, and Kd for sulfite. The differences are no larger than twofold. The foregoing comparison between isozymes beta 1 and beta 2 shows that the naturally engineered insertion in loop 4 exerts some influence on the flavin spectral properties and the active-site reactivity. Since the corresponding loop 4 regions in the three-dimensional structures of flavocytochrome 2 and glycolate oxidase are 1.5-2.0 nm removed from the flavin, it would appear either that loop 4 has a very different conformation in hydroxy-acid oxidase, or that it may interact with the active site due to mobility.

Invariant glycines and prolines flanking in loops the strand beta 2 of various (alpha/beta)8-barrel enzymes: a hidden homology?

The question of parallel (alpha/beta)8-barrel fold evolution remains unclear, owing mainly to the lack of sequence homology throughout the amino acid sequences of (alpha/beta)8-barrel enzymes. The "classical" approaches used in the search for homologies among (alpha/beta)8-barrels (e.g., production of structurally based alignments) have yielded alignments perfect from the structural point of view, but the approaches have been unable to reveal the homologies. These are proposed to be "hidden" in (alpha/beta)8-barrel enzymes. The term "hidden homology" means that the alignment of sequence stretches proposed to be homologous need not be structurally fully satisfactory. This is due to the very long evolutionary history of all (alpha/beta)8-barrels. This work identifies so-called hidden homology around the strand beta 2 that is flanked by loops containing invariant glycines and prolines in 17 different (alpha/beta)8-barrel enzymes, i.e., roughly in half of all currently known (alpha/beta)8-barrel proteins. The search was based on the idea that a conserved sequence region of an (alpha/beta)8-barrel enzyme should be more or less conserved also in the equivalent part of the structure of the other enzymes with this folding motif, given their mutual evolutionary relatedness. For this purpose, the sequence region around the well-conserved second beta-strand of alpha-amylase flanked by the invariant glycine and proline (56_GFTAIWITP, Aspergillus oryzae alpha-amylase numbering), was used as the sequence-structural template. The proposal that the second beta-strand of (alpha/beta)8-barrel fold is important from the evolutionary point of view is strongly supported by the increasing trend of the observed beta 2-strand structural similarity for the pairs of (alpha/beta)8-barrel enzymes: alpha-amylase and the alpha-subunit of tryptophan synthase, alpha-amylase and mandelate racemase, and alpha-amylase and cyclodextrin glycosyltransferase. This trend is also in agreement with the existing evolutionary division of the entire family of (alpha/beta)8-barrel proteins.

Three-dimensional model of the alpha-subunit of bacterial luciferase

The predicted secondary structure of both subunits of bacterial luciferase is in accordance with a regular 8-fold alpha/beta-barrel structure. The 3D profile confirmed that luciferase subunits are compatible with the alpha/beta-barrel despite the absence of sequence similarity with any alpha/beta-barrel protein. The three-dimensional structure of 260 residues of the alpha-chain of luciferase was modeled from coordinates of glycolate oxidase and then energy minimized. The model obtained satisfies the criteria for the structure of a globular protein and is in accordance with known experimental data. From the model it is possible to predict active site residues involved in binding and catalysis. These predictions, and thus also the model, can be tested by protein engineering experiments.

The primary structure of Hyphomicrobium X dimethylamine dehydrogenase. Relationship to trimethylamine dehydrogenase and implications for substrate recognition

The gene encoding dimethylamine dehydrogenase from Hyphomicrobium X has been cloned and over-expressed in Escherichia coli. Using the chemically determined protein sequence, primers were designed to amplify DNA fragments encoding the proximal and distal parts of the gene. These fragments were used to synthesise two probes and the dmd gene was cloned as part of two BamHI fragments isolated from digested genomic DNA. The sequence of the complete open reading frame was determined on both strands and contained 2211 bp coding for a protein of 736 amino acids, including the N-terminal methionine residue that is removed when expressed in the native host. The molecular mass of the processed apoprotein predicted from the DNA sequence is 82,523 Da. Dimethylamine dehydrogenase is closely related to the trimethylamine dehydrogenase of Methylophilus methylotrophus W3A1 (63.5% identical) and other class I FMN-binding beta 8 alpha 8 barrel flavoproteins. Residues in the active site of trimethylamine dehydrogenase that are known, or implicated, to be important in catalysis are conserved in dimethylamine dehydrogenase. Sequence alignment of dimethylamine and trimethylamine dehydrogenases suggests that the specificity for secondary and tertiary amines resides in a single amino acid substitution in a substrate-binding aromatic bowl located in the active site of the enzymes.

An enzyme-substrate complex involved in bacterial cell wall biosynthesis

The crystal structure of UDP-N-acetylenolpyruvylglucosamine reductase in the presence of its substrate, enolpyruvyl-UDP-N-acetylglucosamine, has been solved to 2.7 A resolution. This enzyme is responsible for the synthesis of UDP-N-acetylmuramic acid in bacterial cell wall biosynthesis and consequently provides an attractive target for the design of antibacterial agents. The structure reveals a novel flavin binding motif, shows a striking alignment of the flavin with the substrate, and suggests a catalytic mechanism for the reduction of this unusual enol ether.

Similarity of different beta-strands flanked in loops by glycines and prolines from distinct (alpha/beta)8-barrel enzymes: chance or a homology?

Many (alpha/beta)8-barrel enzymes contain their conserved sequence regions at or around the beta-strand segments that are often preceded and succeeded by glycines and prolines, respectively. alpha-Amylase is one of these enzymes. Its sequences exhibit a very low degree of similarity, but strong conservation is seen around its beta-strands. These conserved regions were used in the search for similarities with beta-strands of other (alpha/beta)8-barrel enzymes. The analysis revealed an interesting similarity between the segment around the beta 2-strand of alpha-amylase and the one around the beta 4-strand of glycolate oxidase that are flanked in loops by glycines and prolines. The similarity can be further extended on other members of the alpha-amylase and glycolate oxidase subfamilies, i.e., cyclodextrin glycosyltransferase and oligo-1,6-glucosidase, and flavocytochrome b2, respectively. Moreover, the alpha-subunit of tryptophan synthase, the (alpha/beta)8-barrel enzyme belonging to the other subfamily of (alpha/beta)8-barrels, has both investigated strands, beta 2 and beta 4, similar to beta 2 of alpha-amylase and beta 4 of glycolate oxidase. The possibilities of whether this similarity exists only by chance or is a consequence of some processes during the evolution of (alpha/beta)8-barrel proteins are briefly discussed.

On the lack of coordination between protein folding and flavin insertion in Escherichia coli for flavocytochrome b2 mutant forms Y254L and D282N

Wild-type flavocytochrome b2 (L-lactate dehydrogenase) from Saccharomyces cerevisiae, as well as a number of its point mutants, can be expressed to a reasonable level as recombinant proteins in Escherichia coli (20-25 mg per liter culture) with a full complement of prosthetic groups. At the same expression level, active-site mutants Y254L and D282N, on the other hand, were obtained with an FMN/heme ratio significantly less than unity, which could not be raised by addition of free FMN. Evidence is provided that the flavin deficit is due to incomplete prosthetic group incorporation during biosynthesis. Flavin-free and holo-forms for both mutants could be separated on a Blue-Trisacryl M column. The far-UV CD spectra of the two forms of each mutant protein were very similar to one another and to that of the wild-type enzyme, suggesting the existence of only local conformational differences between the active holo-enzymes and the nonreconstitutable flavin-free forms. Selective proteolysis with chymotrypsin attacked the same bond for the two mutant holo-enzymes as in the wild-type one, in the protease-sensitive loop. In contrast, for the flavin-free forms of both mutants, cleavage occurred at more than a single bond. Identification of the cleaved bonds suggested that the structural differences between the mutant flavin-free and holo-forms are located mostly at the C-terminal end of the barrel, which carries the prosthetic group and the active site. Altogether, these findings suggest that the two mutations induce an alteration of the protein-folding process during biosynthesis in E. coli; as a result, the synchrony between folding and flavin insertion is lost. Finally, a preliminary kinetic characterization of the mutant holo-forms showed the Km value for lactate to be little affected; kcat values fell by a factor of about 70 for the D282N mutant and of more than 500 for the Y254L mutant, compared to the wild-type enzyme.

Kinetic limitation and cellular amount of pyridoxine (pyridoxamine) 5'-phosphate oxidase of Escherichia coli K-12

We report the purification and enzymological characterization of Escherichia coli K-12 pyridoxine (pyridoxamine) 5'-phosphate (PNP/PMP) oxidase, which is a key committed enzyme in the biosynthesis of the essential coenzyme pyridoxal 5'-phosphate (PLP). The enzyme encoded by pdxH was overexpressed and purified to electrophoretic homogeneity by four steps of column chromatography. The purified PdxH enzyme is a thermally stable 51-kDa homodimer containing one molecule of flavin mononucleotide (FMN). In the presence of molecular oxygen, the PdxH enzyme uses PNP or PMP as a substrate (Km = 2 and 105 microM and kcat = 0.76 and 1.72 s-1 for PNP and PMP, respectively) and produces hydrogen peroxide. Thus, under aerobic conditions, the PdxH enzyme acts as a classical monofunctional flavoprotein oxidase with an extremely low kcat turnover number. Comparison of kcat/Km values suggests that PNP rather than PMP is the in vivo substrate of E. coli PdxH oxidase. In contrast, the eukaryotic enzyme has similar kcat/Km values for PNP and PMP and seems to act as a scavenger. E. coli PNP/PMP oxidase activities were competitively inhibited by the pathway end product, PLP, and by the analog, 4-deoxy-PNP, with Ki values of 8 and 105 microM, respectively. Immunoinhibition studies suggested that the catalytic domain of the enzyme may be composed of discontinuous residues on the polypeptide sequence. Two independent quantitation methods showed that PNP/PMP oxidase was present in about 700 to 1,200 dimer enzyme molecules per cell in E. coli growing exponentially in minimal medium plus glucose at 37 degrees C. Thus, E. coli PNP/PMP oxidase is an example of a relatively abundant, but catalytically sluggish, enzyme committed to PLP coenzyme biosynthesis.

Divergent evolution of a beta/alpha-barrel subclass: detection of numerous phosphate-binding sites by motif search

Study of the most conserved region in many beta/alpha-barrels, the phosphate-binding site, revealed a sequence motif in a few beta/alpha-barrels with known tertiary structure, namely glycolate oxidase (GOX), cytochrome b2 (Cyb2), tryptophan synthase alpha subunit (TrpA), and the indoleglycerolphosphate synthase (TrpC). Database searches identified this motif in numerous other enzyme families: (1) IMP dehydrogenase (IMPDH) and GMP reductase (GuaC); (2) phosphoribosylformimino-5-aminoimidazol carboxamide ribotide isomerase (HisA) and the cyclase-producing D-erythro-imidazole-glycerolphosphate (HisF) of the histidine biosynthetic pathway; (3) dihydroorotate dehydrogenase (PyrD); (4) glutamate synthase (GltB); (5) ThiE and ThiG involved in the biosynthesis of thiamine as well as related proteins; (6) an uncharacterized open reading frame from Erwinia herbicola; and (7) a glycerol uptake operon antiterminator regulatory protein (GlpP). Secondary structure predictions of the different families mentioned above revealed an alternating order of beta-strands and alpha-helices in agreement with a beta/alpha-barrel-like topology. The putative phosphate-binding site is always found near the C-terminus of the enzymes, which are all at least about 200 amino acids long. This is compatible with its assumed location between strand 7 and helix 8. The identification of a significant motif in functionally diverse enzymes suggests a divergent evolution of at least a considerable fraction of beta/alpha-barrels. In addition to the known accumulation of beta/alpha-barrels in the tryptophan biosynthetic pathway, we observe clusters of these enzymes in histidine biosynthesis, purine metabolism, and apparently also in thiamine biosynthesis. The substrates are mostly heterocyclic compounds.(ABSTRACT TRUNCATED AT 250 WORDS)

Flavocytochrome b2-cytochrome c interactions: the electron transfer reaction revisited

This review is concerned with the kinetics and mechanism of electron transfer processes which occur intermolecularly between reduced flavocytochrome b2 and cytochrome c molecules within an encounter complex. Analyses are given of previous reports which aimed at describing the formation of stable complexes obtained at low ionic strength in solution and in the crystalline state with a binding stoichiometry of 1 to 1 heme ratio. Relevant data allow to define the respective role of flavin and heme b2 in the electron transfer towards cytochrome c and give a description of the recognition areas on the two redox partners. The paper also refers to a recent computer model of their postulated interactions as based on the three-dimensional structure of the Saccharomyces cerevisiae single molecules. Special emphasis is given to rapid kinetic investigations of the electron transfer reaction between Hansenula anomala flavocytochrome b2 and cytochrome c studied as a function of concentration, ionic strength and temperature. Data showed that reaction rates were modulated by ionic strength, reaching a saturation behaviour at low ionic strength. In the present paper the temperature effects on Kd and kET have been re-examined. Thermodynamic analysis of the dissociation constant points out the importance of hydrophobic interactions in the complex formation. Analysis of the variations of rate constants in terms of semiclassical theory of electron-transfer reaction yields values of 1.12 eV for the reorganization energy and 0.05 cm-1 for the electronic coupling factor. Interpretation of the electronic coupling in terms of through-bond and/or through-space pathways takes into account the hypothetical model proposed for the binary complex. The functional implications of this model in the electron transfer reaction are discussed. Finally the existence of a conformational equilibrium between the initial binding complex and the complex from which electron transfer occurs is considered.

L-lactate oxidase and L-lactate monooxygenase: mechanistic variations on a common structural theme

Properties of L-lactate oxidase from Aerococcus viridans are described. The gene encoding the enzyme has been isolated. From its cDNA sequence the amino acid sequence has been derived and shown to have high similarity with those of other enzymes catalyzing oxidation of L-alpha-hydroxy acids, including flavocytochrome b2, lactate monooxygenase, glycolate oxidase, mandelate dehydrogenases and a long chain alpha-hydroxy acid oxidase. The enzyme is expressed in Escherichia coli, and is a flavoprotein containing FMN as prosthetic group. It shares many properties of other alpha-hydroxy acid oxidizing enzymes, eg stabilization of the anionic semiquinone form of the flavin, facile formation of flavin-N(5)-sulfite adducts and a set of conserved amino acid residues around the bound flavin. Steady-state and rapid reaction kinetics of the enzyme have been studied and found to share many characteristics with those of L-lactate monooxygenase, but to differ from the latter in quantitative aspects. It is these quantitative differences between the two enzymes which account for the differences in the overall reactions catalyzed. These differences arise from different stabilities of a common intermediate of reduced flavin enzyme and pyruvate. In the case of the monooxygenase this complex is very stable and is the form that reacts with O2 to give a complex in which the oxidative decarboxylation occurs, yielding the products, acetate, CO2, and H2O (Lockridge O, Massey V, Sullivan PA (1972) J Biol Chem 247, 8097-8106). With lactate oxidase, the complex dissociates rapidly, with the result that it is the free reduced flavin form of the enzyme that reacts with O2, to give the observed products, pyruvate and H2O2.

On the evolution of alternate core packing in eightfold beta/alpha-barrels

Two sequence-related subfamilies of flavin-binding beta/alpha-barrels have been identified (the type I and type II proteins) that differ in the nature of residue packing in the core of the barrel domain. Similar observed differences in the packing of internal amino acid side chains in beta/alpha-barrels have previously been used to argue that these domains have evolved convergently toward a stable structural framework. Using structural alignments of flavin-binding barrel proteins, we demonstrate that simple genetic alterations may be responsible for switching the nature of side-chain packing observed in beta/alpha-barrels. The implication is that the 2 structural classes of beta/alpha-barrel cores can arise divergently from an ancestral barrel framework and that convergent evolution to a stable fold need not be invoked to account for the emergence of 2 classes of beta/alpha-barrel core.

The 2.6-A refined structure of the Escherichia coli recombinant Saccharomyces cerevisiae flavocytochrome b2-sulfite complex

Flavocytochrome b2 from Saccharomyces cerevisiae catalyzes the oxidation of L-lactate to pyruvate and the electron transfer to cytochrome c in the mitochondrial intermembrane space. It is a homotetramer with a molecular weight of 4 x 58 kDa, each monomer of which is composed of 2 distinct domains, the one carrying FMN and the other, a "b5-like" heme. The native structure has been described at a resolution of 2.4 A (Xia ZX, Mathews FS, 1990, J Mol Biol 212:837-863). The heme domains protrude from the central body of the tetramer consisting of the 4 FMN binding domains. Because only 2 heme domains are visible in the electron density map, the other 2 are probably disordered. We crystallized the Escherichia coli recombinant flavocytochrome b2 from S. cerevisiae inhibited by sulfite. Although the crystals were obtained under very different conditions from those of the pyruvate-containing native enzyme, they were found to be isostructural (P 3(2) 2 1, a = b = 164.5 A, c = 114.0 A). The 2.6-A X-ray structure was extensively refined with X-PLOR (R = 17.3%), which made it possible to describe in detail the recombinant flavocytochrome b2 molecular structure. There exist few differences between the native and recombinant structures, in line with the fact that they show similar kinetic behavior, and they further confirm the intrinsic mobility of the heme domain (Labeyrie F, Beloil JC, Thomas MA, 1988, Biochim Biophys Acta 953:134-141). This structure will be used as a starting model in the structural resolution of flavocytochrome b2 point mutants.

Structural studies on recombinant and point mutants of flavocytochrome b2

Flavocytochrome b2 from S cerevisiae is a homotetramer with a molecular mass of 4 x 58 kDa. It catalyses the oxidation of L-lactate into pyruvate and the electron transfer to cytochrome c in the mitochondrial intermembrane space. Each monomer is composed of a flavinmononucleotide (FMN) carrying domain and a 'b5-like' heme domain. The wild type structure has been described at a resolution of 2.4 A. We report here on the refined structure of the E. coli native recombinant flavocytochrome b2 from S cerevisiae inhibited by sulphite and that of two point mutants, Y143F and Y254F, in which pyruvate is bound to the active site. The crystals, obtained under very different conditions from those of the native enzyme, are isostructural (P 3(2) 2 1, a=b=164.5 A, c=114.0 A). In line with the similarities found to exist in the kinetic behaviour of the native and recombinant protein, few structural differences were observed here, and the crystallographic data further confirm the intrinsic mobility of the heme domain. The superimposable position of the aromatic rings of Phe 143 in the mutant Y143F and Tyr 143 in the native protein makes it seem unlikely that the aromatic ring may be directly involved in the intramolecular electron transfer. The fact that a very restricted number of domain interactions was observed in Y143F shows that Tyr 143 is one of the amino acids essential to the formation of the productive complex. In the Y143F mutant, the number of catalytically efficient complexes is probably drastically decreased, which will severely limit the rate of intramolecular election transfer. The structure of Y254F shows a reorientation of the substrate at the active site. Together with the kinetic results, this finding definitely excludes the possibility that Tyr 254 may act as general base and that the substrate may interact directly with Phe 254 in the mutant. The model between flavocytochrome b2 and cytochrome c will serve as a basis for designing suitable mutants of the amino acids involved either in the interaction or the electron transfer.

On the rate of proton exchange with solvent of the catalytic histidine in flavocytochrome b2 (yeast L-lactate dehydrogenase)

The family of FMN-dependent, alpha-hydroxy acid-oxidizing enzymes catalyzes substrate dehydrogenation by a mechanism the first step of which is abstraction of the substrate alpha-proton (so-called carbanion mechanism). For flavocytochrome b2 and lactate oxidase, it was shown that once on the enzyme this proton is lost only slowly to the solvent (Lederer F, 1984, In: Bray RC, Engel PC, Mayhew SG, eds, Flavins & flavoproteins, Berlin: Walter de Gruyter & Co., pp 513-526; Urban P, Lederer F, 1985, J Biol Chem 260:11115-11122). This suggested the occurrence of a pKa increase of the catalytic histidine upon enzyme reduction by substrate. For flavocytochrome b2, the crystal structure indicated 2 possible origins for the stabilization of the imidazolium form of His 373: either a network of hydrogen bonds involving His 373, Tyr 254, flavin N5 and O4, a heme propionate, and solvent molecules, and/or electrostatic interactions with Asp 282 and with the reduced cofactor N1 anion. In this work, we probe the effect of the hydrogen bond network at the active site by studying proton exchange with solvent for 2 mutants: Y254F and the recombinant flavodehydrogenase domain, in which this network should be disrupted. The rate of proton exchange, as determined by intermolecular hydrogen transfer experiments, appears identical in the flavodehydrogenase domain and the wild-type enzyme, whereas it is about 3-fold faster in the Y254F mutant. It thus appears that specific hydrogen bonds to the solvent do not play a major role in stabilizing the acid form of His 373 in reduced flavocytochrome b2. Removal of the Y254 phenol group induces a pKa drop of about half a pH unit for His 373 in the reduced enzyme. Even then, the rate of exchange of the imidazolium proton with solvent is still lower by several orders of magnitude than that of a normally ionizing histidine. Other factors must then also contribute to the pKa increase, such as the electrostatic interactions with D282 and the anionic reduced cofactor, as suggested by the crystal structure.

On the origin of enzymatic species

The diversity of enzyme catalytic function is remarkable, particularly when one considers that ancestral life forms must have started with a much smaller ensemble of proteins. In this article, we discuss the evolution of the mandelate pathway in pseudomonads as an example of how catalytic diversity may have evolved. We suggest that existing enzymes that catalyse the chemistry needed to accomplish a transformation were recruited, followed by the evolution of specific binding.

Evolution of parallel beta/alpha-barrel enzyme family lightened by structural data on starch-processing enzymes

The parallel beta/alpha-barrel domain consisting of eight parallel beta-sheets surrounded by eight alpha-helices has been currently identified in crystal structures of more than 20 enzymes. This type of protein folding motif makes it possible to catalyze various biochemical reactions on a variety of substrates (i.e., it seems to be robust enough so that different enzymatic functionalities could be designed on it). In spite of many efforts aimed at elucidation of evolutionary history of the present-day beta/alpha-barrels, a challenging question remains unanswered: How has the parallel beta/alpha-barrel fold arisen? Although the complete sequence comparison of all beta/alpha-barrel amino acid sequences is not yet available, several sequence similarities have been revealed by using the highly conserved regions of alpha-amylase as structural templates. Since many starch-processing enzymes adopt the parallel beta/alpha-barrel structure these enzymes might be useful in the search for evolutionary relationships of the whole parallel eight-folded beta/alpha-barrel enzyme family.

Crystal structure of apo-glycolate oxidase

The crystal structure of the apoform of the alpha/beta-barrel enzyme glycolate oxidase has been determined to 2.6 A resolution. Removal of the tightly bound cofactor FMN has a very strong influence on the protein structure; it is converted into a very flexible state, verging on a molten globule type of structure. The asymmetric unit contains two subunits with different conformations to each other and to the holo-enzyme. The secondary structures are preserved, but their mutual arrangement has changed to some extent introducing cavities into the protein. The largest structural shifts are, however, found in the loops.

A hypothetical complex between crystalline flavocytochrome b2 and cytochrome c

Flavocytochrome b2 and cytochrome c are physiological electron transfer partners in yeast mitochondria. The formation of a stable complex between them has been demonstrated both in solution and in the crystalline state. On the basis of the three-dimensional structures, using molecular modeling and energy minimization, we have generated a hypothetical model for the interaction of these redox partners in the crystal lattice. General criteria such as good charge and surface complementarity, plausible orientation, and separation distance of the prosthetic groups, as well as more specific criteria such as the stoichiometry determined in the crystal, and the involvement of both domains and of more than one subunit of flavocytochrome b2 led us to discriminate between several possible interaction sites. In the hypothetical model we present, four cytochrome c molecules interact with a tetramer of flavocytochrome b2. The b2 and c hemes are coplanar, with an edge-to-edge distance of 14 A. The contact surface area is ca. 800 A2. Several electrostatic interactions involving the flavin and the heme domains of flavocytochrome b2 stabilize the binding of cytochrome c.

Molecular cloning and nucleotide sequence of cDNA encoding rat kidney long-chain L-2-hydroxy acid oxidase. Expression of the catalytically active recombinant protein as a chimaera

Long-chain L-alpha-hydroxy acid oxidase from rat kidney is a member of the family of FMN-dependent alpha-hydroxy-acid-oxidizing enzymes. With the knowledge of the recently determined amino acid sequence, the cDNA encoding the enzyme has now been cloned using the polymerase chain reaction. The 1648-bp cDNA contains an open reading frame coding for the 352 residues of the previously determined sequence, preceded by a methionine codon. In addition, several clones were found to present a nine-base insertion, predicting the existence of an isoform with a tripeptide VRK inserted between residues 188 and 189 of the mature protein. The presence of about 10% of this isoform in the oxidase purified from rat kidney was indeed identified by amino acid sequencing. A recombinant active enzyme was obtained as a protein fused to glutathione S-transferase using the bacterial expression plasmid pGEX-3X. Physico-chemical characterization indicated, for the fused enzyme, properties similar to those of the rat kidney protein. When the chimaera was submitted to factor Xa, proteolysis at the engineered cleavage point was poor. Separation of hydroxy acid oxidase from glutathione S-transferase could not be achieved with trypsin either. With both proteases, the initial cleavage point appeared to be in a peptide loop internal to the hydroxy acid oxidase sequence, close to or in the tripeptide insertion locus and not at the engineered factor-Xa-cleavage point. Comparative tryptic proteolysis of the rat kidney enzyme yielded a form cleaved in the same loop.

Role of tyrosine 129 in the active site of spinach glycolate oxidase

The enzymatic properties and the three-dimensional structure of spinach glycolate oxidase which has the active-site Tyr129 replaced by Phe (Y129F glycolate oxidase) has been studied. The structure of the mutant is unperturbed which facilitates interpretation of the biochemical data. Y129F glycolate oxidase has an absorbance spectrum with maxima at 364 and 450 nm (epsilon max = 11400 M-1 cm-1). The spectrum indicates that the flavin is in its normal protonated form, i.e. the Y129F mutant does not lower the pKa of the N(3) of oxidized flavin as does the wild-type enzyme [Macheroux, P., Massey, V., Thiele, D. J., and Volokita, M. (1991) Biochemistry 30, 4612-4619]. This was confirmed by a pH titration of Y129F glycolate oxidase which showed that the pKa is above pH 9. In contrast to wild-type glycolate oxidase, oxalate does not perturb the absorbance spectrum of Y129F glycolate oxidase. Moreover oxalate does not inhibit the enzymatic activity of the mutant enzyme. Typical features of wild-type glycolate oxidase that are related to a positively charged lysine side chain near the flavin N(1)-C(2 = O), such as stabilization of the anionic flavin semiquinone and formation of tight N(5)-sulfite adducts, are all conserved in the Y129F mutant protein. Y129F glycolate oxidase exhibited about 3.5% of the wild-type activity. The lower turnover number for the mutant of 0.74 s-1 versus 20 s-1 for the wild-type enzyme amounts to an increase of the energy of the transition state of about 7.8 kJ/mol. Steady-state analysis gave Km values of 1.5 mM and 7 microM for glycolate and oxygen, respectively. The Km for glycolate is slightly higher than that found for wild-type glycolate oxidase (1 mM) whereas the Km for oxygen is much lower. As was the case for wild-type glycolate oxidase, reduction was found to be the rate-limiting step in catalysis, with a rate of 0.63 s-1. The kinetic properties of Y129F glycolate oxidase provide evidence that the main function of the hydroxyl group of Tyr129 is the stabilization of the transition state.

Sequence similarities in (alpha/beta)8-barrel enzymes revealed by conserved regions of alpha-amylase

The parallel (alpha/beta)8-barrel is a frequently occurring protein-folding motif. Although the arrangement of secondary structural elements along the barrel is very similar in different (alpha/beta)8-barrel enzymes, there is a very low mutual amino acid sequence homology among the enzymes, contributing in part to the hazy view of their evolution. Here an approach to identifying at least the rough of evolutionarily conserved (alpha/beta)8-barrel sequence is presented. Based on the idea that highly conserved sequence regions of a particular enzyme should be more or less conserved in the sequences of the other evolutionary related enzymes, five sequence similarities of ten different (alpha/beta)8-barrel enzymes were revealed, using the five conserved regions of the amino acid sequence of the alpha-amylase (alpha/beta)8-barrel as the templates.

The refined crystal structure of Pseudomonas putida lipoamide dehydrogenase complexed with NAD+ at 2.45 A resolution

The three-dimensional structure of one of the three lipoamide dehydrogenases occurring in Pseudomonas putida, LipDH Val, has been determined at 2.45 A resolution. The orthorhombic crystals, grown in the presence of 20 mM NAD+, contain 458 residues per asymmetric unit. A crystallographic 2-fold axis generates the dimer which is observed in solution. The final crystallographic R-factor is 21.8% for 18,216 unique reflections and a model consisting of 3,452 protein atoms, 189 solvent molecules and 44 NAD+ atoms, while the overall B-factor is unusually high: 47 A2. The structure of LipDH Val reveals the conformation of the C-terminal residues which fold "back" into the putative lipoamide binding region. The C-terminus has been proven to be important for activity by site-directed mutagenesis. However, the distance of the C-terminus to the catalytically essential residues is surprisingly large, over 6 A, and the precise role of the C-terminus still needs to be elucidated. In this crystal form LipDH Val contains one NAD+ molecule per subunit. Its adenine-ribose moiety occupies an analogous position as in the structure of glutathione reductase. However, the nicotinamide-ribose moiety is far removed from its expected position near the isoalloxazine ring and points into solution. Comparison of LipDH Val with Azotobacter vinelandii lipoamide dehydrogenase yields an rms difference of 1.6 A for 440 well defined C alpha atoms per subunit. Comparing LipDH Val with glutathione reductase shows large differences in the tertiary and quaternary structure of the two enzymes. For instance, the two subunits in the dimer are shifted by 6 A with respect to each other. So, LipDH Val confirms the surprising differences in molecular architecture between glutathione reductase and lipoamide dehydrogenase, which were already observed in Azotobacter vinelandii LipDH. This is the more remarkable since the active sites are located at the subunit interface and are virtually identical in all three enzymes.

Extreme pKa displacements at the active sites of FMN-dependent alpha-hydroxy acid-oxidizing enzymes

Flavocytochrome b2 (or L-lactate dehydrogenase) from baker's yeast is thought to operate by the initial formation of a carbanion, as do the evolutionarily related alpha-hydroxy acid-oxidizing FMN-dependent oxidases. Previous work has shown that, in the active site of the unligated reduced flavocytochrome b2, the group that has captured the substrate alpha-proton has a high pKapp, calculated to lie around 15 through the use of Eigen's equation. A detailed inspection of the now known three-dimensional structure of the enzyme leads to the conclusion that the high pKa belongs to His 373, an active site group that plays the role of general base in the forward reaction and of general acid in the reverse direction. Moreover, consideration of the kinetics of proton transfer during the catalytic cycle suggests that the pKa of the reduced FMN N5 position should be lowered by several pH units compared to its pKa of 20 or more when free. The features of the three-dimensional structure possibly responsible for these pK shifts are analyzed; they are proposed to consist of a network of hydrogen bonds with the solvent and of a mutual electrostatic stabilization of anionic reduced flavin and the imidazolium ion. Finally, it is suggested that similar pK shifts affect the active sites of the alpha-hydroxy acid-oxidizing flavooxidases, which are homologous to flavocytochrome b2. The functional significance of these pK shifts in terms of catalysis and semiquinone stabilization is discussed.

Novel NADPH-binding domain revealed by the crystal structure of aldose reductase

Aldose reductase is the first enzyme in the polyol pathway and catalyses the NADPH-dependent reduction of D-glucose to D-sorbitol. Under normal physiological conditions aldose reductase participates in osmoregulation, but under hyperglycaemic conditions it contributes to the onset and development of severe complications in diabetes. Here we present the crystal structure of pig lens aldose reductase refined to an R-factor of 0.232 at 2.5-A resolution. It exhibits a single domain folded in an eight-stranded parallel alpha/beta barrel, similar to that in triose phosphate isomerase and a score of other enzymes. Hence, aldose reductase does not possess the expected canonical dinucleotide-binding domain. Crystallographic analysis of the binding of 2'-monophospho-adenosine-5'-diphosphoribose, which competitively inhibits NADPH binding reveals that it binds into a cleft located at the C-terminal end of the strands of the alpha/beta barrel. This represents a new type of binding for nicotinamide adenine dinucleotide coenzymes.