Structure of L-aspartate oxidase: implications for the succinate dehydrogenase/fumarate reductase oxidoreductase family.

BACKGROUND: Given the vital role of NAD+ in cell metabolism, the enzymes involved in bacterial de novo NAD+ biosynthesis are possible targets for drug design against pathogenic bacteria. The first reaction in the pathway is catalysed by L-aspartate oxidase (LASPO), a flavoenzyme that converts aspartate to iminoaspartate using either molecular oxygen or fumarate as electron acceptors. LASPO has considerable sequence homology with the flavoprotein subunits of succinate dehydrogenase (SDH) and fumarate reductase (FRD). RESULTS: The crystal structure of the apoform of LASPO from Escherichia coli has been determined to 2.2 A resolution. The enzyme shows a novel fold for an FAD-dependent protein, comprising a three-domain structure: an FAD-binding domain with the dinucleotide-binding fold, a C-terminal three-helical bundle domain, and an alpha + beta capping domain, which is topologically similar to the small subunit of spinach ribulose-1,5-bisphosphate carboxylase/oxygenase. The interface between the FAD-binding and capping domains defines a cleft in which the active site is located. CONCLUSIONS: A number of strictly conserved residues present in all three domains indicate that LASPO, SDH and FRD share the same overall folding topology. Many of these conserved residues are in the FAD-binding site and active centre, suggesting a similar catalytic mechanism. Thus, LASPO, SDH and FRD form a class of functionally and structurally related oxidoreductases that are all able to reduce fumarate and to oxidise a dicarboxylate substrate.     

Fumaraldehydic acid-induced inactivation of aspartase

Fumaraldehydic acid (FAA) induced a time-dependent inactivation of aspartase (L-aspartate ammonia-lyase, EC 4.3.1.1) from Escherichia coli at 30 degrees C and pH 7.4 following pseudo-first order kinetics. The rate of inactivation increased in proportion to the FAA concentration. In addition, the rate of inactivation increased, as the pH was increased. Determination of sulfhydryl groups showed that approximately one among 11 sulfhydryl groups was modified by FAA concomitant with the inactivation. L-Aspartate and fumarate protected the enzyme against FAA-inactivation, when Mg2+ ions were present. Unlike E. coli aspartase, P. fluorescens aspartase was not inactivated by FAA.     

Two biochemically distinct classes of fumarase in Escherichia coli

Biochemical studies with strains of Escherichia coli that are amplified for the products of the three fumarase genes, fumA (FUMA), fumB (FUMB) and fumC (FUMC), have shown that there are two distinct classes of fumarase. The Class I enzymes include FUMA, FUMB, and the immunologically related fumarase of Euglena gracilis. These are characteristically thermolabile dimeric enzymes containing identical subunits of Mr 60,000. FUMA and FUMB are differentially regulated enzymes that function in the citric acid cycle (FUMA) or to provide fumarate as an anaerobic electron acceptor (FUMB), and their affinities for fumarate and L-malate are consistent with these roles. The Class II enzymes include FUMC, and the fumarases of Bacillus subtilis, Saccharomyces cerevisiae and mammalian sources. They are thermostable tetrameric enzymes containing identical subunits Mr 48,000-50,000. The Class II fumarases share a high degree of sequence identity with each other (approx. 60%) and with aspartase (approx. 38%) and argininosuccinase (approx. 15%), and it would appear that these are all members of a family of structurally related enzymes. It is also suggested that the Class I enzymes may belong to a wider family of iron-dependent carboxylic acid hydro-lyases that includes maleate dehydratase and aconitase. Apart from one region containing a Gly-Ser-X-X-Met-X-X-Lys-X-Asn consensus sequence, no significant homology was detected between the Class I and Class II fumarases.     

Enhancement of recombinant protein production in Escherichia coli by coproduction of aspartase

As commonly recognized, the excretion of acetate by the aerobic growth of Escherichia coli on glucose is a manifestation of imbalanced flux between glycolysis and the tricarboxylic acid (TCA) cycle. Accordingly, this may restrict the production of recombinant proteins in E. coli, due to the limited amounts of precursor metabolites produced in TCA cycle. To approach this issue, an extra supply of intermediate metabolites in TCA cycle was made by conversion of aspartate to fumarate, a reaction mediated by the activity of L-aspartate ammonia-lyase (aspartase). As a result, in the glucose minimal medium containing aspartate, the production of two recombinant proteins, beta-galactosidase and green fluorescent protein, in the aspartase-producing strain was substantially increased by 5-fold in association with 30-40% more biomass production. This preliminary study illustrates the great promise of this approach used to enhance the production of these two recombinant proteins.     

Expression of the E. coli nadB gene and characterization of the gene product L-aspartate oxidase

Quinolinic acid is synthesized in E. coli by the enzymes L-aspartate oxidase and quinolinate synthase A, the genes of which are named nadB and nadA. In our previous work we cloned and characterized the two genes (Flachmann, R., Kunz, N., Seifert, J., Gütlich, M., Wientjes, F.J., L?ufer, A. & Gassen, H.G. (1988) Eur. J. Biochem. 175, 221-228). Here we report on the expression of the nadB gene under control of the inducible left promoter of the bacteriophage lambda. The yield of the active gene product L-aspartate oxidase was enhanced up to 20% of the soluble cell protein. The enzyme was purified to homogeneity in a three-step procedure and the reading frame of the L-aspartate oxidase gene was confirmed by Edman degradation of five cyanogen bromide peptides. L-Aspartate oxidase shows no classical Michaelis-Menten behaviour but is subject to a substrate inactivation. The apparent Km values were different for substrate concentrations below and above 1mM and were determined to 0.5 mM and 4.1mM, respectively. The active form of the enzyme is a monomer of 60,284 Da and contains one molecule of FAD and nine cysteine residues, four of which built up two disulfide bonds. The isoelectric point of the protein was determined to be at pH 5.6. Chemical modifications of the enzyme showed that at least one tyrosine and one histidine residue are essential for enzyme activity. The coenzyme-binding domain is located in the amino-terminal part of the polypeptide chain as revealed by a sequence comparison to other dinucleotide binding enzymes. Furthermore, there is evidence for a relationship to fumarate reductase and succinate dehydrogenase of E. coli.     

Identification of

The succinate dehydrogenases (SDH: soluble, membrane-extrinsic subunits of succinate:quinone oxidoreductases) from Escherichia coli and beef heart mitochondria each adsorb at a pyrolytic graphite 'edge' electrode and catalyse the interconversion of succinate and fumarate according to the electrochemical potential that is applied. E. coli and beef heart mitochondrial SDH share only ca. 50% homology, yet the steady-state catalytic activities, when measured over a continuous potential range, display very similar catalytic operating potentials and energetic biases (the relative ability to catalyse succinate oxidation vs. fumarate reduction). Importantly, E. coli SDH also exhibits the interesting 'tunnel-diode' behaviour previously reported for the mitochondrial enzyme. Thus as the potential is lowered below ca. -60 mV (pH 7, 38 degrees C) the rate of catalytic fumarate reduction decreases abruptly despite an increase in driving force. Since the homology relates primarily to residues associated with active site regions, the marked similarity in the voltammetry reaffirms our previous conclusions that the tunnel-diode behaviour is a characteristic property of the enzyme active site. Thus, succinate dehydrogenase is an excellent fumarate reductase, but its activity in this direction is limited to a very specific range of potential.     

Acetylornithine deacetylase, succinyldiaminopimelate desuccinylase and carboxypeptidase G2 are evolutionarily related.

The nucleotide (nt) sequence of the Escherichia coli argE gene, encoding the acetylornithine deacetylase (AO) subunit, has been established and corresponds to a 43-kDa (M(r) 42,320) polypeptide. The enzyme has been purified to near homogeneity and it appears to be a dimer consisting of two 43-kDa subunits. The amino acid sequence deduced from the nt sequence was compared to that of the subunit of E. coli succinyldiaminopimelate desuccinylase (the dapE gene product involved in the diaminopimelate pathway for lysine biosynthesis), since both enzymes share functional and biochemical features. Significant similarity covering the entire sequence allows us to infer a common origin for both deacylases. This homology extends to the Pseudomonas sp. G2 carboxypeptidase (G2CP); this or a functionally related enzyme may be responsible for the minor AO activity found in organisms relying on ornithine acetyltransferase for ornithine biosynthesis.     

Structural studies on aspartate aminotransferase from Escherichia coli. Covalent structure

The amino acid sequence of aspartate aminotransferase from Escherichia coli was established by sequence analysis and alignment of 39 tryptic peptides and 7 cyanogen bromide peptides. The total number of amino acid residues of the subunit was 396, and the molecular weight was calculated to be 43,573. A comparison of the primary structure of the E. coli enzyme with all known sequences of the two types of isoenzyme (mitochondrial and cytosolic enzymes) in vertebrates revealed that approximately 25% of all residues are invariant. The amino acid residues which were proposed from crystallographic studies on the vertebrate enzymes to be essential for the enzymic action are well conserved in the E. coli enzyme. The E. coli enzyme shows a similar degree of sequence homology to both the mitochondrial and cytosolic isoenzymes (close to 40%). The finding that the positions of deletions introduced into the sequence of E. coli enzyme to give the maximum homology agree well with those of the mitochondrial enzymes supports the endosymbiotic hypothesis of mitochondrial origin.     

The primary structure of rabbit liver cytosolic serine hydroxymethyltransferase

The complete amino acid sequence of cytosolic serine hydroxymethyltransferase from rabbit liver was determined. The sequence was determined from analysis of peptides isolated from tryptic and cyanogen bromide cleavages of the enzyme. Special procedures were used to isolate and sequence the C-terminal and blocked N-terminal peptides. Each of the four identical subunits of the enzyme consists of 483 residues. The sequence could be easily aligned with the sequence of Escherichia coli serine hydroxymethyltransferase. The primary structural homology between the rabbit and E. coli enzymes is about 42%. The importance of the primary and predicted secondary structural homology between the two enzymes is discussed.     

Cloning, sequencing, and expression of the genes encoding the adenosylcobalamin-dependent ethanolamine ammonia-lyase of Salmonella typhimurium

Ethanolamine ammonia-lyase is a bacterial enzyme that catalyzes the adenosylcobalamin-dependent conversion of certain vicinal amino alcohols to oxo compounds and ammonia. Studies of ethanolamine ammonia-lyase from Clostridium sp. and Escherichia coli have suggested that the enzyme is a heterodimer composed of subunits of Mr approximately 55,000 and 35,000. Using a partial Sau3A Salmonella typhimurium library ligated into pBR328 and selecting by complementation of a mutant lacking ethanolamine ammonia-lyase activity, we have cloned the genes for the 2 subunits of the S. typhimurium enzyme. The genes were localized to a 6.5-kilobase fragment of S. typhimurium DNA, from which they could be expressed in E. coli under noninducing conditions. Sequencing of a 2526-base pair portion of this 6.5-kilobase DNA fragment revealed two open reading frames separated by 21 base pairs. The open reading frames encoded proteins of 452 and 286 residues whose derived N-terminal sequences were identical to the N-terminal sequences of the 2 subunits of the E. coli ethanolamine ammonia-lyase, except that residue 16 of the large subunit was asparagine in the E. coli sequence and aspartic acid in the S. typhimurium sequence.     

Fumarate reductase and succinate oxidase activity of Escherichia coli complex II homologs are perturbed differently by mutation of the flavin binding domain

The Escherichia coli complex II homologues succinate:ubiquinone oxidoreductase (SQR, SdhCDAB) and menaquinol:fumarate oxidoreductase (QFR, FrdABCD) have remarkable structural homology at their dicarboxylate binding sites. Although both SQR and QFR can catalyze the interconversion of fumarate and succinate, QFR is a much better fumarate reductase, and SQR is a better succinate oxidase. An exception to the conservation of amino acids near the dicarboxylate binding sites of the two enzymes is that there is a Glu (FrdA Glu-49) near the covalently bound FAD cofactor in most QFRs, which is replaced with a Gln (SdhA Gln-50) in SQRs. The role of the amino acid side chain in enzymes with Glu/Gln/Ala substitutions at FrdA Glu-49 and SdhA Gln-50 has been investigated in this study. The data demonstrate that the mutant enzymes with Ala substitutions in either QFR or SQR remain functionally similar to their wild type counterparts. There were, however, dramatic changes in the catalytic properties when Glu and Gln were exchanged for each other in QFR and SQR. The data show that QFR and SQR enzymes are more efficient succinate oxidases when Gln is in the target position and a better fumarate reductase when Glu is present. Overall, structural and catalytic analyses of the FrdA E49Q and SdhA Q50E mutants suggest that coulombic effects and the electronic state of the FAD are critical in dictating the preferred directionality of the succinate/fumarate interconversions catalyzed by the complex II superfamily.     

L-Aspartate-induced activation of aspartase

During the catalysis of the fumarate amination reaction, aspartase was markedly activated by the product, L-aspartate, as shown by a steep increase in the reaction rate. When NH4+ was replaced by NH2OH, the hydroxylamination reaction proceeded without any acceleration, and was activated upon addition of L-aspartate. The activation required the Mg2+ ion and the alkaline pH, and the half-saturation concentration of L-aspartate for activation was as low as 0.07 mM, which was far lower than the Km value for catalysis. Fumarate showed no activating effect in contrast to L-aspartate, and L-aspartate lowered the Km value for fumarate instead of acting as a competitive inhibitor. Besides L-aspartate, alpha-methyl-DL-aspartate exhibited an activating effect without serving as a substrate. These results suggest that the activation is mediated by an indirect action of L-aspartate which is bound to a site distinct from the catalytic site.     

Evolution of amino acid biosynthesis and enzymes with broad substrate specificity.

I selected 82 proteins that were related to amino acid biosynthesis in the genome of Escherichia coli. I then searched the extensive sequence homology for each of the selected proteins from among the proteins of E.coli. The result showed that 30 proteins of the selected proteins had extensive sequence homology within the selected proteins, and 21 proteins had extensive sequence homology to proteins outside the selected proteins. In addition, the enzymes with broad substrate specificity play an important role in the amino acid biosynthesis. I demonstrate here that some substrate-specific enzymes evolved from an ancestor enzyme with broad substrate specificity. CONTACT: hnishida@iam.u-tokyo.ac.jp     

Mechanism of action of Escherichia coli phosphoribosylaminoimidazolesuccinocarboxamide synthetase

The conversion of ATP, L-aspartate, and 5-aminoimidazole-4-carboxyribonucleotide (CAIR) to 5-aminoimidazole-4-(N-succinylcarboxamide) ribonucleotide (SAICAR), ADP, and phosphate by phosphoribosylaminoimidazolesuccinocarboxamide synthetase (SAICAR synthetase) represents the eighth step of de novo purine nucleotide biosynthesis. SAICAR synthetase and other enzymes of purine biosynthesis are targets of natural products that impair cell growth. Prior to this study, no kinetic mechanism was known for any SAICAR synthetase. Here, a rapid equilibrium random ter-ter kinetic mechanism is established for the synthetase from Escherichia coli by initial velocity kinetics and patterns of linear inhibition by IMP, adenosine 5'-(beta,gamma-imido)triphosphate (AMP-PNP), and maleate. Substrates exhibit mutual binding antagonism, with the strongest antagonism between CAIR and either ATP or L-aspartate. CAIR binds to the free enzyme up to 200-fold more tightly than to the ternary enzyme-ATP-aspartate complex, but the latter complex may be the dominant form of SAICAR synthetase in vivo. IMP is a competitive inhibitor with respect to CAIR, suggesting the possibility of a hydrogen bond interaction between the 4-carboxyl and 5-amino groups of enzyme-bound CAIR. Of several aspartate analogues tested (hadacidin, l-malate, succinate, fumarate, and maleate), maleate was by far the best inhibitor, competitive with respect to L-aspartate. Inhibition by IMP and maleate is consistent with a chemical mechanism for SAICAR synthetase that parallels that of adenylosuccinate synthetase.     

Crystal structure of a putative aspartate aminotransferase belonging to subgroup IV

Protein TT0402 from Thermus thermophilus HB8 exhibits about 30-35% sequence identity with proteins belonging to subgroup IV in the aminotransferase family of the fold-type I pyridoxal 5'-phosphate (PLP)-dependent enzymes. In this study, we determined the crystal structure of TT0402 at 2.3 A resolution (R(factor) = 19.9%, R(free) = 23.6%). The overall structure of TT0402 exhibits the fold conserved in aminotransferases, and is most similar to that of the Escherichia coli phosphoserine aminotransferase, which belongs to subgroup IV but shares as little as 13% sequence identity with TT0402. Kinetic assays confirmed that TT0402 has higher transamination activities with the amino group donor, L-glutamate, and somewhat lower activities with L-aspartate. These results indicate that TT0402 is a subgroup IV aminotransferase for the synthesis/degradation of either L-aspartate or a similar compound.     

Cloning, sequence, and expression of a chitinase gene from a marine bacterium, Altermonas sp. strain O-7.

The gene encoding an extracellular chitinase from marine Alteromonas sp. strain O-7 was cloned in Escherichia coli JM109 by using pUC18. The chitinase produced was not secreted into the growth medium but accumulated in the periplasmic space. A chitinase-positive clone of E. coli produced two chitinases with different molecular weights from a single chitinase gene. These proteins showed almost the same enzymatic properties as the native chitinase of Alteromonas sp. strain O-7. The N-terminal sequences of the two enzymes were identical. The nucleotide sequence of the 3,394-bp SphI-HindIII fragment that included the chitinase gene was determined. A single open reading frame was found to encode a protein consisting of 820 amino acids with a molecular weight of 87,341. A putative ribosome-binding site, promoter, and signal sequence were identified. The deduced amino acid sequence of the cloned chitinase showed sequence homology with chitinases A (33.4%) and B (15.3%) from Serratia marcescens. Regardless of origin, the enzymes of the two bacteria isolated from marine and terrestrial environments had high homology, suggesting that these organisms evolved from a common ancestor.     

Homology of Escherichia coli B glutathione synthetase with dihydrofolate reductase in amino acid sequence and substrate binding site

Glutathione synthetase from Escherichia coli B showed amino acid sequence homology with mammalian and bacterial dihydrofolate reductases over 40 residues, although these two enzymes are different in their reaction mechanisms and ligand requirements. The effects of ligands of dihydrofolate reductase on the reaction of E. coli B glutathione synthetase were examined to find resemblances in catalytic function to dihydrofolate reductase. The E. coli B enzyme was potently inhibited by 7,8-dihydrofolate, methotrexate, and trimethoprim. Methotrexate was studied in detail and proved to bind to an ATP binding site of the E. coli B enzyme with K1 value of 0.1 mM. The homologous portion of the amino acid sequence in dihydrofolate reductases, which corresponds to the portion coded by exon 3 of mammalian dihydrofolate reductase genes, provided a binding site of the adenosine diphosphate moiety of NADPH in the crystal structure of dihydrofolate reductase. These analyses would indicate that the homologous portion of the amino acid sequence of the E. coli B enzyme provides the ATP binding site. This report gives experimental evidence that amino acid sequences related by sequence homology conserve functional similarity even in enzymes which differ in their catalytic mechanisms.     

Purification and characterization of thermostable aspartate aminotransferase from a thermophilic Bacillus species.

Aspartate aminotransferase (EC 2.6.1.1) was purified to homogeneity from cell extracts of a newly isolated thermophilic bacterium, Bacillus sp. strain YM-2. The enzyme consisted of two subunits identical in molecular weight (Mr, 42,000) and showed microheterogeneity, giving two bands with pIs of 4.1 and 4.5 upon isoelectric focusing. The enzyme contained 1 mol of pyridoxal 5'-phosphate per mol of subunit and exhibited maxima at about 360 and 415 nm in absorption and circular dichroism spectra. The intensities of the two bands were dependent on the buffer pH; at neutral or slightly alkaline pH, where the enzyme showed its maximum activity, the absorption peak at 360 nm was prominent. The enzyme was specific for L-aspartate and L-cysteine sulfinate as amino donors and alpha-ketoglutarate as an amino acceptor; the KmS were determined to be 3.0 mM for L-aspartate and 2.6 mM for alpha-ketoglutarate. The enzyme was most active at 70 degrees C and had a higher thermostability than the enzyme from Escherichia coli. The N-terminal amino acid sequence (24 residues) did not show any similarity with the sequences of mammalian and E. coli enzymes, but several residues were identical with those of the thermoacidophilic archaebacterial enzyme recently reported.     

Computer analysis of the sequence relationships among 4.5S RNA molecular species from various sources

Nucleotide sequence homology among 4.5S RNAs from various organisms was examined by computer analysis to evaluate their sequence relationships. Chloroplast 4.5S rRNAs of wheat and tobacco were not significantly related to Escherichia coli 4.5S RNA, but were closely related to the 3'-terminus of bacterial 23S rRNA. Significant sequence homology was found between rat Novikoff hepatoma 4.5S RNAI and mouse and hamster 4.5S RNAs, suggesting that these RNAs are products of a family of genes with diverged sequences. E. coli 4.5S RNA had no significant sequence homology with any rodent 4.5S RNAs as a whole sequence. The E. coli, mouse and hamster 4.5S RNAs, however, were found to share a homologous 14-nucleotide sequence at the center of the molecules, which is known to exist as a conserved sequence in both Alu and Alu-equivalent sequences of mammalian DNAs.     

Fatty acid synthetase, malic enzyme and other NADP+ binding dehydrogenases have similar antigenic determinant(s) at the NADPH binding domain

The sequence of tryptic and chymotryptic peptides from cytosolic and mitochondrial rabbit liver serine hydroxymethyltransferase are compared to the proposed sequence of a protein coded for by the glyA gene of Escherichia coli. The E. coli glyA gene is believed to code for serine hydroxymethyltransferase. Extensive sequence homology between these peptides were found for the proposed E. coli enzyme in the aminoterminal two-thirds of the molecule. All three proteins have identical sequences from residue 222-231. This sequence is known to contain the lysyl residue which forms a Schiff's base with pyridoxal-P in the two rabbit liver enzymes. These results support the interpretation that the proposed sequence of E. coli serine hydroxymethyltransferase is correct. The data also show that cytosolic and mitochondrial serine hydroxymethyltransferase are homologous proteins.