Binding modes of DL-2-haloacid dehalogenase revealed by crystallography, modeling and isotope effects studies.

Several pathways of biotic dechlorination can be found in enzymes, each characterized by different chlorine isotopic fractionation, which can thus serve as a signature of a particular mechanism. Unlike other dehalogenases, DL-2-haloacid dehalogenase, DL-DEX, converts both enantiomers of the substrate. Chlorine isotope effects for this enzyme are larger than in the case of other dehalogenases. Recently, the 3D structure of this enzyme became available and enabled us to model these isotope effects and seek their origin. We show that the elevated values of the chlorine kinetic isotope effects originate in part in the processes of binding and migration within the enzyme active site that precede the dehalogenation step.

Substrate specificity of fluoroacetate dehalogenase: an insight from crystallographic analysis, fluorescence spectroscopy, and theoretical computations.

The high substrate specificity of fluoroacetate dehalogenase was explored by using crystallographic analysis, fluorescence spectroscopy, and theoretical computations. A crystal structure for the Asp104Ala mutant of the enzyme from Burkholderia sp. FA1 complexed with fluoroacetate was determined at 1.2 Å resolution. The orientation and conformation of bound fluoroacetate is different from those in the crystal structure of the corresponding Asp110Asn mutant of the enzyme from Rhodopseudomonas palustris CGA009 reported recently (J. Am. Chem. Soc. 2011, 133, 7461). The fluorescence of the tryptophan residues of the wild-type and Trp150Phe mutant enzymes from Burkholderia sp. FA1 incubated with fluoroacetate and chloroacetate was measured to gain information on the environment of the tryptophan residues. The environments of the tryptophan residues were found to be different between the fluoroacetate- and chloroacetate-bound enzymes; this would come from different binding modes of these two substrates in the active site. Docking simulations and QM/MM optimizations were performed to predict favorable conformations and orientations of the substrates. The F atom of the substrate is oriented toward Arg108 in the most stable enzyme-fluoroacetate complex. This is a stable but unreactive conformation, in which the small O-C-F angle is not suitable for the S(N)2 displacement of the F(-) ion. The cleavage of the C-F bond is initiated by the conformational change of the substrate to a near attack conformation (NAC) in the active site. The second lowest energy conformation is appropriate for NAC; the C-O distance and the O-C-F angle are reasonable for the S(N) 2 reaction. The activation energy is greatly reduced in this conformation because of three hydrogen bonds between the leaving F atom and surrounding amino acid residues. Chloroacetate cannot reach the reactive conformation, due to the longer C-Cl bond; this results in an increase of the activation energy despite the weaker C-Cl bond.

Reaction mechanism and molecular basis for selenium/sulfur discrimination of selenocysteine lyase.

Selenocysteine lyase (SCL) catalyzes the pyridoxal 5'-phosphate-dependent removal of selenium from l-selenocysteine to yield l-alanine. The enzyme is proposed to function in the recycling of the micronutrient selenium from degraded selenoproteins containing selenocysteine residue as an essential component. The enzyme exhibits strict substrate specificity toward l-selenocysteine and no activity to its cognate l-cysteine. However, it remains unclear how the enzyme distinguishes between selenocysteine and cysteine. Here, we present mechanistic studies of selenocysteine lyase from rat. ESI-MS analysis of wild-type and C375A mutant SCL revealed that the catalytic reaction proceeds via the formation of an enzyme-bound selenopersulfide intermediate on the catalytically essential Cys-375 residue. UV-visible spectrum analysis and the crystal structure of SCL complexed with l-cysteine demonstrated that the enzyme reversibly forms a nonproductive adduct with l-cysteine. Cys-375 on the flexible loop directed l-selenocysteine, but not l-cysteine, to the correct position and orientation in the active site to initiate the catalytic reaction. These findings provide, for the first time, the basis for understanding how trace amounts of a selenium-containing substrate is distinguished from excessive amounts of its cognate sulfur-containing compound in a biological system.

Crystal structure of a homolog of mammalian serine racemase from Schizosaccharomyces pombe.

D-serine is an endogenous coagonist for the N-methyl-D-aspartate receptor and is involved in excitatory neurotransmission in the brain. Mammalian pyridoxal 5'-phosphate-dependent serine racemase, which is localized in the mammalian brain, catalyzes the racemization of L-serine to yield D-serine and vice versa. The enzyme also catalyzes the dehydration of D- and L-serine. Both reactions are enhanced by Mg.ATP in vivo. We have determined the structures of the following three forms of the mammalian enzyme homolog from Schizosaccharomyces pombe: the wild-type enzyme, the wild-type enzyme in the complex with an ATP analog, and the modified enzyme in the complex with serine at 1.7, 1.9, and 2.2 A resolution, respectively. On binding of the substrate, the small domain rotates toward the large domain to close the active site. The ATP binding site was identified at the domain and the subunit interface. Computer graphics models of the wild-type enzyme complexed with L-serine and D-serine provided an insight into the catalytic mechanisms of both reactions. Lys-57 and Ser-82 located on the protein and solvent sides, respectively, with respect to the cofactor plane, are acid-base catalysts that shuttle protons to the substrate. The modified enzyme, which has a unique "lysino-D-alanyl" residue at the active site, also exhibits catalytic activities. The crystal-soaking experiment showed that the substrate serine was actually trapped in the active site of the modified enzyme, suggesting that the lysino-D-alanyl residue acts as a catalytic base in the same manner as inherent Lys-57 of the wild-type enzyme.

Structures of hypoxanthine-guanine phosphoribosyltransferase (TTHA0220) from Thermus thermophilus HB8.

Hypoxanthine-guanine phosphoribosyltransferase (HGPRTase), which is a key enzyme in the purine-salvage pathway, catalyzes the synthesis of IMP or GMP from alpha-D-phosphoribosyl-1-pyrophosphate and hypoxanthine or guanine, respectively. Structures of HGPRTase from Thermus thermophilus HB8 in the unliganded form, in complex with IMP and in complex with GMP have been determined at 2.1, 1.9 and 2.2 A resolution, respectively. The overall fold of the IMP complex was similar to that of the unliganded form, but the main-chain and side-chain atoms of the active site moved to accommodate IMP. The overall folds of the IMP and GMP complexes were almost identical to each other. Structural comparison of the T. thermophilus HB8 enzyme with 6-oxopurine PRTases for which structures have been determined showed that these enzymes can be tentatively divided into groups I and II and that the T. thermophilus HB8 enzyme belongs to group I. The group II enzymes are characterized by an N-terminal extension with additional secondary elements and a long loop connecting the second alpha-helix and beta-strand compared with the group I enzymes.

X-Ray crystallographic and mutational studies of fluoroacetate dehalogenase from Burkholderia sp. strain FA1.

Fluoroacetate dehalogenase catalyzes the hydrolytic defluorination of fluoroacetate to produce glycolate. The enzyme is unique in that it catalyzes the cleavage of a carbon-fluorine bond of an aliphatic compound: the bond energy of the carbon-fluorine bond is among the highest found in natural products. The enzyme also acts on chloroacetate, although much less efficiently. We here determined the X-ray crystal structure of the enzyme from Burkholderia sp. strain FA1 as the first experimentally determined three-dimensional structure of fluoroacetate dehalogenase. The enzyme belongs to the alpha/beta hydrolase superfamily and exists as a homodimer. Each subunit consists of core and cap domains. The catalytic triad, Asp104-His271-Asp128, of which Asp104 serves as the catalytic nucleophile, was found in the core domain at the domain interface. The active site was composed of Phe34, Asp104, Arg105, Arg108, Asp128, His271, and Phe272 of the core domain and Tyr147, His149, Trp150, and Tyr212 of the cap domain. An electron density peak corresponding to a chloride ion was found in the vicinity of the N(epsilon1) atom of Trp150 and the N(epsilon2) atom of His149, suggesting that these are the halide ion acceptors. Site-directed replacement of each of the active-site residues, except for Trp150, by Ala caused the total loss of the activity toward fluoroacetate and chloroacetate, whereas the replacement of Trp150 caused the loss of the activity only toward fluoroacetate. An interaction between Trp150 and the fluorine atom is probably an absolute requirement for the reduction of the activation energy for the cleavage of the carbon-fluorine bond.

Structure of D-alanine-D-alanine ligase from Thermus thermophilus HB8: cumulative conformational change and enzyme-ligand interactions.

D-Alanine-D-alanine ligase (Ddl) is one of the key enzymes in peptidoglycan biosynthesis and is an important target for drug discovery. The enzyme catalyzes the condensation of two D-Ala molecules using ATP to produce D-Ala-D-Ala, which is the terminal peptide of a peptidoglycan monomer. The structures of five forms of the enzyme from Thermus thermophilus HB8 (TtDdl) were determined: unliganded TtDdl (2.3 A resolution), TtDdl-adenylyl imidodiphosphate (2.6 A), TtDdl-ADP (2.2 A), TtDdl-ADP-D-Ala (1.9 A) and TtDdl-ATP-D-Ala-D-Ala (2.3 A). The central domain rotates as a rigid body towards the active site in a cumulative manner in concert with the local conformational change of three flexible loops depending upon substrate or product binding, resulting in an overall structural change from the open to the closed form through semi-open and semi-closed forms. Reaction-intermediate models were simulated using TtDdl-complex structures and other Ddl structures previously determined by X-ray methods. The catalytic process accompanied by the cumulative conformational change has been elucidated based on the intermediate models in order to provide new insights regarding the details of the catalytic mechanism.

Crystallization screening test for the whole-cell project on Thermus thermophilus HB8.

It was essential for the structural genomics of Thermus thermophilus HB8 to efficiently crystallize a number of proteins. To this end, three conventional robots, an HTS-80 (sitting-drop vapour diffusion), a Crystal Finder (hanging-drop vapour diffusion) and a TERA (modified microbatch) robot, were subjected to a crystallization condition screening test involving 18 proteins from T. thermophilus HB8. In addition, a TOPAZ (microfluidic free-interface diffusion) designed specifically for initial screening was also briefly examined. The number of diffraction-quality crystals and the time of appearance of crystals increased in the order HTS-80, Crystal Finder, TERA. With the HTS-80 and Crystal Finder, the time of appearance was short and the rate of salt crystallization was low. With the TERA, the number of diffraction-quality crystals was high, while the time of appearance was long and the rate of salt crystallization was relatively high. For the protein samples exhibiting low crystallization success rates, there were few crystallization conditions that were common to the robots used. In some cases, the success rate depended greatly on the robot used. The TOPAZ showed the shortest time of appearance and the highest success rate, although the crystals obtained were too small for diffraction studies. These results showed that the combined use of different robots significantly increases the chance of obtaining crystals, especially for proteins exhibiting low crystallization success rates. The structures of 360 of 944 purified proteins have been successfully determined through the combined use of an HTS-80 and a TERA.

Expression, purification and preliminary X-ray characterization of DL-2-haloacid dehalogenase from Methylobacterium sp. CPA1.

DL-2-Haloacid dehalogenase from Methylobacterium sp. CPA1 (DL-DEX Mb) is a unique enzyme that catalyzes the dehalogenation reaction without the formation of an ester intermediate. A recombinant form of DL-DEX Mb has been expressed in Escherichia coli, purified and crystallized using the hanging-drop vapour-diffusion method. The crystal belongs to the hexagonal space group P6(3), with unit-cell parameters a = b = 186.2, c = 114.4 A. The crystals are likely to contain between four and eight monomers in the asymmetric unit, with a V(M) value of 4.20-2.10 A3 Da(-1). A self-rotation function revealed peaks on the chi = 180 degrees section. X-ray data have been collected to 1.75 A resolution.

Engineering the substrate specificity of porcine kidney D-amino acid oxidase by mutagenesis of the "active-site lid".

Comparison of the primary structures of pig kidney D-amino acid oxidase (DAO) and human brain D-aspartate oxidase (DDO) revealed a notable difference at I215-N225 of DAO and the corresponding region, R216-G220, of DDO. A DAO mutant, in which I215-N225 is substituted by R216-G220 of DDO, showed D-aspartate-oxidizing activity that wild-type DAO does not exhibit, together with a considerable decrease in activity toward D-alanine. These findings indicate that I215-N225 of DAO contributes profoundly to its substrate specificity. Based on these results and the crystal structure of DAO, we systematically mutated the E220-Y224 region within the short stretch in question and obtained five mutants (220D224G, 221D224G, 222D224G, 223D224G, and 224D), in each of which an aspartate residue is mutated to E220-Y224. All of the mutants exhibited decreased apparent K(m) values toward D-arginine, i.e., to one-seventh to one-half that of wild type DAO. The specificity constant, k(cat app)/K(m app), for D-arginine increased by one order of magnitude for the 221D224G or 222D224G mutant, whereas that for D-alanine or D-serine decreased to marginal or nil.

Three-dimensional structure of rat-liver acyl-CoA oxidase in complex with a fatty acid: insights into substrate-recognition and reactivity toward molecular oxygen.

The three-dimensional structure of rat-liver acyl-CoA oxidase-II (ACO-II) in a complex with a C12-fatty acid was solved by the molecular replacement method based on the uncomplexed ACO-II structure. The crystalline form of the complex was obtained by cocrystallization of ACO-II with dodecanoyl-CoA. The crystalline complex possessed, in the active-site crevice, only the fatty acid moiety that had been formed through hydrolysis of the thioester bond. The overall dimeric structure and the folding pattern of each subunit are essentially superimposable on those of uncomplexed ACO-II. The active site including the flavin ring of FAD, the crevice embracing the fatty acyl moiety, and adjacent amino acid side chains are superimposably conserved with the exception of Glu421, whose carboxylate group is tilted away to accommodate the fatty acid. One of the carboxyl oxygens of the bound fatty acid is hydrogen-bonded to the amide hydrogen of Glu421, the presumed catalytic base, and to the ribityl 2'-hydroxyl group of FAD. This hydrogen-bonding network correlates well with the substrate recognition/activation in acyl-CoA dehydrogenase. The binding mode of C12-fatty acid suggests that the active site does not close upon substrate binding, but remains spacious during the entire catalytic process, the oxygen accessibility in the oxidative half-reaction thereby being maintained.

Crystal structures of CTP synthetase reveal ATP, UTP, and glutamine binding sites.

CTP synthetase (CTPs) catalyzes the last step in CTP biosynthesis, in which ammonia generated at the glutaminase domain reacts with the ATP-phosphorylated UTP at the synthetase domain to give CTP. Glutamine hydrolysis is active in the presence of ATP and UTP and is stimulated by the addition of GTP. We report the crystal structures of Thermus thermophilus HB8 CTPs alone, CTPs with 3SO4(2-), and CTPs with glutamine. The enzyme is folded into a homotetramer with a cross-shaped structure. Based on the binding mode of sulfate anions to the synthetase site, ATP and UTP are computer modeled into CTPs with a geometry favorable for the reaction. Glutamine bound to the glutaminase domain is situated next to the triad of Glu-His-Cys as a catalyst and a water molecule. Structural information provides an insight into the conformational changes associated with the binding of ATP and UTP and the formation of the GTP binding site.

The active site structure of quinohemoprotein amine dehydrogenase inhibited by p-nitrophenylhydrazine.

Quinohemoprotein amine dehydrogenase (QH-AmDH) catalyzes the oxidative deamination of aliphatic and aromatic amines. The enzyme from Pseudomonas putida has an alpha beta gamma heterotrimeric structure with two heme c groups in the largest alpha subunit, and a novel quinone cofactor [cysteine tryptophylquinone (CTQ)] and hitherto unknown internal cross-bridges in the smallest gamma subunit. The crystal structure of the enzyme in the complex with the inhibitor [p-nitrophenylhydrazine (pNPH)] has been determined at a 2.0 A resolution.(1) The hydrazone of the cofactor with the inhibitor was nicely modeled into the omit electron density map, identifying the C6 carbonyl group as the reactive site of the cofactor. The Asp33 gamma is unambiguously determined as the catalytic base to abstract the alpha-proton from a substrate, because N beta atom of the inhibitor corresponding to the C alpha atom of the substrate amine is neighbored to Asp33 gamma. The bound inhibitor is completely enclosed in the active site pocket formed by the residues from the beta- and gamma-subunits. The cofactor-inhibitor adduct may be predominantly in the hydrazone with the azo form as a minor component. The binding of the inhibitor causes minor but important conformational changes in the residues surrounding the active site. The inhibitor may have access to the active site pocket through the water-filled crevice between the beta- and gamma-subunits.

Characterization of histidinol phosphate aminotransferase from Escherichia coli.

Histidinol phosphate aminotransferase (HPAT) is a pyridoxal 5'-phosphate (PLP)-dependent aminotransferase classified into Subgroup I aminotransferase, in which aspartate aminotransferase (AspAT) is the prototype. In order to expand our knowledge on the reaction mechanism of Subgroup I aminotransferases, HPAT is an enzyme suitable for detailed mechanistic studies because of having low sequence identity with AspAT and a unique substrate recognition mode. Here we investigated the spectroscopic properties of HPAT and the effect of the C4-C4' strain of the PLP-Lys(214) Schiff base on regulating the Schiff base pK(a) in HPAT. Similar to AspAT, the PLP-form HPAT showed pH-dependent absorption spectral change with maxima at 340 nm at high pH and 420 nm at low pH, having a low pK(a) of 6.6. The pK(a) value of the methylamine-reconstituted K214A mutant enzyme was increased from 6.6 to 10.6. Mutation of Asn(157) to Ala increased the pK(a) to 9.2. Replacement of Arg(335) by Leu increased the pK(a) to 8.6. On the other hand, the pK(a) value of the N157A/R335L double mutant enzyme was 10.6. These data indicate that the strain of the Schiff base is the principal factor to decrease the pK(a) in HPAT and is crucial for the subsequent increase in the Schiff base pK(a) during catalysis, although the electrostatic effect of the arginine residue that binds the negatively charged group of the substrate is larger in HPAT than that in AspAT. Our findings also support the idea that the strain mechanism is common to Subgroup I aminotransferases.

Strain and catalysis in aspartate aminotransferase.

The notion of "ground-state destabilization" has been well documented in enzymology. It is the unfavourable interaction (strain) in the enzyme-substrate complex, and increases the k(cat) value without changing the k(cat)/K(m) value. During the course of the investigation on the reaction mechanism of aspartate aminotransferase (AAT), we found another type of strain that is crucial for catalysis: the strain of the distorted internal aldimine in the unliganded enzyme. This strain raises the energy level of the starting state (E+S), thereby reducing the energy gap between E+S and ES(++) and increasing the k(cat)/K(m) value. Further analysis on the reaction intermediates showed that the Michaelis complex of AAT with aspartate contains strain energy due to an unfavourable interaction between the main chain carbonyl oxygen and the Tyr225-aldimine hydrogen-bonding network. This belongs to the classical type of strain. In each case, the strain is reflected in the pK(a) value of the internal aldimine. In the historical explanation of the reaction mechanism of AAT, the shifts in the aldimine pK(a) have been considered to be the driving forces for the proton transfer during catalysis. However, the above findings indicate that the true driving forces are the strain energy inherent to the respective intermediates. We describe here how these strain energies are generated and are used for catalysis, and show that variations in the aldimine pK(a) during catalysis are no more than phenomenological results of adjusting the energy levels of the reaction intermediates for efficient catalysis.

A new family of NAD(P)H-dependent oxidoreductases distinct from conventional Rossmann-fold proteins.

A new family of NAD(P)H-dependent oxidoreductases is now recognized as a protein family distinct from conventional Rossmann-fold proteins. Numerous putative proteins belonging to the family have been annotated as malate dehydrogenase (MDH) or lactate dehydrogenase (LDH) according to the previous classification as type-2 malate/L-lactate dehydrogenases. However, recent biochemical and genetic studies have revealed that the protein family consists of a wide variety of enzymes with unique catalytic activities other than MDH or LDH activity. Based on their sequence homologies and plausible functions, the family proteins can be grouped into eight clades. This classification would be useful for reliable functional annotation of the new family of NAD(P)H-dependent oxidoreductases.

Structure of imidazole glycerol phosphate synthase from Thermus thermophilus HB8: open-closed conformational change and ammonia tunneling.

Imidazole glycerol phosphate synthase (IGPs) catalyzes the fifth step in the histidine biosynthetic pathway located at the branch point to de novo purine biosynthesis. IGPs is a multienzyme comprising glutaminase and synthase subunits. The glutaminase activity, which hydrolyzes glutamine to give ammonia, is coupled with substrate binding to the synthase subunit. The three-dimensional structure of the IGPs from Thermus thermophilus HB8 has been determined at 2.3 A resolution, and compared with the previously determined structures for the yeast and Thermotoga maritima enzymes. The structure of each subunit is similar to that of the corresponding domain in the yeast enzyme or subunit in the T. maritima enzyme. However, the overall structure is significantly different from the yeast and T. maritima enzymes, indicating that IGPs may change the relative orientation between the two subunits and close the glutaminase site upon glutamine binding. The putative ammonia tunnel, which carries nascent ammonia from glutaminase to the synthase site, has a closed gate comprising a cyclic salt bridge formed by four charged residues of the synthase subunit. The side chain of Lys100 in the cyclic salt bridge might change its side chain direction to form new interactions with the main chain carbonyl group of glutamine from the synthase subunit and the hydoxyl group of tyrosine from the glutaminase subunit, resulting in the opening of the gate for ammonia transfer.

Effects of hydrogen bonds in association with flavin and substrate in flavoenzyme d-amino acid oxidase. The catalytic and structural roles of Gly313 and Thr317.

According to the three-dimensional structure of a porcine kidney D-amino acid oxidase-substrate (D-leucine) complex model, the G313 backbone carbonyl recognizes the substrate amino group by hydrogen bonding and the side-chain hydroxyl of T317 forms a hydrogen bond with C(2)=O of the flavin moiety of FAD [Miura et al. (1997) J. Biochem. 122, 825-833]. We have designed and expressed the G313A and T317A mutants and compared their enzymatic and spectroscopic properties with those of the wild type. The G313A mutant shows decreased activities to various D-amino acids, but the pattern of substrate specificity is different from that of the wild type. The results imply that the hydrogen bond between the G313 backbone carbonyl and the substrate amino group plays important roles in substrate recognition and in defining the substrate specificity of D-amino acid oxidase. The T317A mutant shows a decreased affinity for FAD. The steady-state kinetic measurements indicate diminished activities of T317A to substrate D-amino acids. The transient kinetic parameters measured by stopped-flow spectroscopy revealed that T317 plays key roles in stabilizing the purple intermediate, a requisite intermediate in the oxidative half-reaction, and in enhancing the release of the product from the active site, thereby optimizing the overall catalytic process of D-amino acid oxidase.

Three-dimensional structure of the flavoenzyme acyl-CoA oxidase-II from rat liver, the peroxisomal counterpart of mitochondrial acyl-CoA dehydrogenase.

Acyl-CoA oxidase (ACO) catalyzes the first and rate-determining step of the peroxisomal beta-oxidation of fatty acids. The crystal structure of ACO-II, which is one of two forms of rat liver ACO (ACO-I and ACO-II), has been solved and refined to an R-factor of 20.6% at 2.2-A resolution. The enzyme is a homodimer, and the polypeptide chain of the subunit is folded into the N-terminal alpha-domain, beta-domain, and C-terminal alpha-domain. The X-ray analysis showed that the overall folding of ACO-II less C-terminal 221 residues is similar to that of medium-chain acyl-CoA dehydrogenase (MCAD). However, the N-terminal alpha- and beta-domains rotate by 13 with respect to the C-terminal alpha-domain compared with those in MCAD to give a long and large crevice that accommodates the cofactor FAD and the substrate acyl-CoA. FAD is bound to the crevice between the beta- and C-terminal domains with its adenosine diphosphate portion interacting extensively with the other subunit of the molecule. The flavin ring of FAD resides at the active site with its si-face attached to the beta-domain, and is surrounded by active-site residues in a mode similar to that found in MCAD. However, the residues have weak interactions with the flavin ring due to the loss of some of the important hydrogen bonds with the flavin ring found in MCAD. The catalytic residue Glu421 in the C-terminal alpha-domain seems to be too far away from the flavin ring to abstract the alpha-proton of the substrate acyl-CoA, suggesting that the C-terminal domain moves to close the active site upon substrate binding. The pyrimidine moiety of flavin is exposed to the solvent and can readily be attacked by molecular oxygen, while that in MCAD is protected from the solvent. The crevice for binding the fatty acyl chain is 28 A long and 6 A wide, large enough to accommodate the C23 acyl chain.

Glutamine:phenylpyruvate aminotransferase from an extremely thermophilic bacterium, Thermus thermophilus HB8.

A subfamily I aminotransferase gene homologue containing an open reading frame encoding 381 amino acid residues (Mr=42,271) has been identified in the process of the genome project of an extremely thermophilic bacterium, Thermus thermophilus HB8. Alignment of the predicted amino acid sequence using FASTA shows that this protein is a member of aminotransferase subfamily Igamma. The protein shows around 40% identity with both T. thermophilus aspartate aminotransferase [EC 2.6.1.1] and mammalian glutamine:phenylpyruvate aminotransferase [EC 2.6.1.64]. The recombinant protein expressed in Escherichia coli is a homodimer with a subunit molecular weight of 42,000, has one pyridoxal 5'-phosphate per subunit, and is highly active toward glutamine, methionine, aromatic amino acids, and corresponding keto acids, but has no preference for alanine and dicarboxylic amino acids. These substrate specificities are similar to those described for mammalian glutamine: phenylpyruvate aminotransferase. This is the first enzyme reported so far that has the glutamine aminotransferase activity in non-eukaryotic cells. As the presence of aromatic amino acid:2-oxoglutarate aminotransferase [EC 2.6.1.57] has not been reported in T. thermophilus, this enzyme is expected to catalyze the last transamination step of phenylalanine and tyrosine biosynthesis. It may also be involved in the methionine regeneration pathway associated with polyamine biosynthesis. The enzyme shows a strikingly high pKa value (9.3) of the coenzyme Schiff base in comparison with other subfamily I aminotransferases. The origin of this unique pKa value and the substrate specificity is discussed based on the previous crystallographic data of T. thermophilus and E. coli aspartate aminotransferases.