Specificity and kinetics of Rhodotorula gracilis D-amino acid oxidase.

D-Amino acid oxidase purified from the yeast Rhodotorula gracilis is a flavoenzyme which does not require exogenous FAD for maximum activity. The enzyme showed temperature and pH activity optima centred between 40 and 45 degrees C and between 8.0 and 8.5, respectively; a broad pH and ionic strength range of stability and a more limited range of thermostability was determined. The enzyme stability was markedly influenced by the presence of 2-mercaptoethanol. Apparent kinetic parameters for a number of substrates were determined: nonpolar and aromatic D-amino acids appeared to be the best substrates. Steady state measurements carried out at different oxygen concentrations indicated that for D-alanine the kinetic pattern is consistent with a Ping Pong Bi Bi mechanism; kcat values on D-alanine and D-valine are 43,250 min-1 and 31,370 min-1, respectively. L-Amino acids did not inhibit enzyme activity; several aromatic and aliphatic carboxylic acids proved to be competitive inhibitors of the enzyme and their ki values were determined. The reported properties of R. gracilis D-amino acid oxidase markedly distinguish it from other characterized D-amino acid oxidases.

Induction of cytotoxic oxidative stress by D-alanine in brain tumor cells expressing Rhodotorula gracilis D-amino acid oxidase: a cancer gene therapy strategy.

Hydrogen peroxide (H2O2) is a reactive oxygen species (ROS) generated in the stereoselective deamination of D-amino acids catalyzed by D-amino acid oxidase (DAAO). H2O2 readily crosses cellular membranes and damages DNA, proteins, and lipids. The scarcity of DAAO substrates in mammalian organisms and its co-localization with catalase in the peroxisomal matrix suggested that the cytotoxicity of ROS could be harnessed by administration of D-amino acids to tumor cells ectopically expressing DAAO in the cytoplasm. To evaluate this hypothesis, the cDNA encoding the highly active DAAO from the red yeast Rhodotorula gracilis was mutated to remove the carboxy-terminal peroxisomal targeting sequence. A clonal line of 9L glioma cells stably transfected with this construct (9Ldaao17) was found to synthesize active R. gracilis DAAO. Exposure of 9Ldaao17 cells to D-alanine resulted in cytotoxicity at concentrations that were nontoxic to parental 9L cells. Depletion of cellular glutathione further sensitized 9Ldaao17 cells to D-alanine (D-Ala). This result, combined with stimulation of pentose phosphate pathway activity and the production of extracellular H2O2 by 9Ldaao17 cells incubated with D-alanine implicates oxidative stress as the mediator of cytotoxicity. These results demonstrate that expression of R. gracilis DAAO in tumor cells confers chemosensitivity to D-alanine that could be exploited as a novel cancer gene therapy paradigm.

Expression of D-amino acid oxidase in Rhodotorula gracilis under induction conditions: a biochemical and cytochemical study.

D-amino acid oxidase is expressed to a high level in the yeast Rhodotorula gracilis (0.3% of total cell protein) through induction by D-alanine in a defined growth medium. Monospecific polyclonal antibodies against pure enzyme were obtained. Western blot analysis showed that the enzyme is synthesized as the mature polypeptide. The localization of the enzyme was investigated by immunoelectron microscopy using the postembedding immunogold technique and by submicroscopic enzyme cytochemistry. D-Amino acid oxidase was detected in peroxisomes, and quantitation of immunoelectron microscopic data indicated that the enzyme is exclusively confined to these organelles. Immunoelectron microscopic observations are in complete agreement with biochemical data showing that the enzyme is not expressed in the absence of D-alanine. Morphometric analysis demonstrated that induction of D-amino acid oxidase synthesis is associated with a 241% increase of peroxisome volume density and with a 31% increase of peroxisome size as compared to cells grown on non-inducing medium.

Reactivity of histidyl residues in D-amino acid oxidase from Rhodotorula gracilis.

Incubation of D-amino acid oxidase from the yeast Rhodotorula gracilis with excess dansyl chloride at pH 6.6 and 18 degrees C caused an irreversible inactivation of D-amino acid oxidase. Benzoate, a competitive inhibitor of the enzyme, completely protected the enzyme from inactivation. The dansylated-enzyme, isolated by gel-filtration, was in part still active while the substrate specificity was altered substantially. It was completely reduced by D-alanine in anaerobiosic conditions and did stabilize the red anion semiquinone upon photochemical reduction with EDTA. The results provide evidence for the presence of essential histidyl residue(s) in the active center of the yeast enzyme.

Identification and role of ionizing functional groups at the active center of Rhodotorula gracilis D-amino acid oxidase.

D-Amino acid oxidase (DAAO) is a flavoprotein oxidase that catalyzes the oxidation of amino acids and produces ketoacids and H(2)O(2). The rate of product release from reduced DAAO from Rhodotorula gracilis is pH dependent and reflects a pK(a) of approximately 9.3. Binding of benzoate and 3,3,3-trifluoro-D-alanine to wild-type and Y238F-DAAO is also pH dependent (pK(a)=9.8+/-0.1 and 9.05+/-0.1, respectively for benzoate binding). However, binding of benzoate to Y223F-DAAO is pH independent, indicating the pK(a) is due to Y223-OH. This latter residue is thus involved in substrate binding, and probably is the group that governs product release. In contrast to this, the second active site tyrosine, Y238, has little influence on ligand binding.

Cloning, sequencing and expression in E. coli of a D-amino acid oxidase cDNA from Rhodotorula gracilis active on cephalosporin C.

We have cloned the cDNA coding for the Rhodotorula gracilis D-amino acid oxidase (DAAO), an enzyme that performs with high catalytic efficiency biotechnologically relevant bioconversions, by PCR amplification. The first strand cDNA was synthesised from the total mRNA fraction isolated from R. gracilis cells grown under DAAO-inducing conditions. The R. gracilis DAAO cDNA consists of 1104 bp encoding a protein of 368 amino acids. The insertion of the cDNA into the pKK223-3 plasmid allowed the expression of recombinant DAAO in Escherichia coli as a wholly soluble and catalytically active holoenzyme (approximately 0.5 U mg-1 protein) with a fermentation yield, in terms of DAAO units, of 800 U l-1. This level of expression allowed the purification, in homogeneous form and high yield (50%), of the recombinant enzyme which showed a high catalytic activity on cephalosporin C as substrate. The nucleotide sequence reported in this paper will appear in the nucleotide sequence databases under accession number.

Physiological functions of D-amino acid oxidases: from yeast to humans.

D-Amino acid oxidase (DAAO) is a FAD-containing flavoenzyme that catalyzes the oxidative deamination of D-isomers of neutral and polar amino acids. This enzymatic activity has been identified in most eukaryotic organisms, the only exception being plants. In the various organisms in which it does occur, DAAO fulfills distinct physiological functions: from a catabolic role in yeast cells, which allows them to grow on D-amino acids as carbon and energy sources, to a regulatory role in the human brain, where it controls the levels of the neuromodulator D-serine. Since 1935, DAAO has been the object of an astonishing number of investigations and has become a model for the dehydrogenase-oxidase class of flavoproteins. Structural and functional studies have suggested that specific physiological functions are implemented through the use of different structural elements that control access to the active site and substrate/product exchange. Current research is attempting to delineate the regulation of DAAO functions in the contest of complex biochemical and physiological networks.

D-Amino acid oxidase: new findings.

The most recent research on D-amino acid oxidases and D-amino acid metabolism has revealed new, intriguing properties of the flavoenzyme and enlighted novel biotechnological uses of this catalyst. Concerning the in vivo function of the enzyme, new findings on the physiological role of D-amino acid oxidase point to a detoxifying function of the enzyme in metabolizing exogenous D-amino acids in animals. A novel role in modulating the level of D-serine in brain has also been proposed for the enzyme. At the molecular level, site-directed mutagenesis studies on the pig kidney D-amino acid oxidase and, more recently, on the enzyme from the yeast Rhodotorula gracilis indicated that the few conserved residues of the active site do not play a role in acid-base catalysis but rather are involved in substrate interactions. The three-dimensional structure of the enzyme was recently determined from two different sources: at 2.5-3.0 A resolution for DAAO from pig kidney and at 1.2-1.8 A resolution for R. gracilis. The active site can be clearly depicted: the striking absence of essential residues acting in acid-base catalysis and the mode of substrate orientation into the active site, taken together with the results of free-energy correlation studies, clearly support a hydrid transfer type of mechanism in which the orbital steering between the substrate and the isoalloxazine atoms plays a crucial role during catalysis.

On the holoenzyme reconstitution process in native and truncated Rhodotorula gracilis D-amino acid oxidase.

After developing a rapid gel filtration method to prepare pure and stable apoenzyme forms of D-amino acid oxidase from the yeast Rhodotorula gracilis, we carried out comparative kinetic studies on the reconstitution to holoenzyme (with FAD) of the intact (40 kDa) and proteolyzed (38.3 kDa) apoenzyme forms of this oxidase. Changes in catalytic activity and flavin and protein fluorescence revealed that in both cases reconstitution was biphasic. The proteolyzed enzyme was catalytically competent, but unlike the intact form was unable to dimerize following formation of the apoprotein-FAD complex. We present evidence that reconstitution of holoenzyme from apoenzyme plus FAD does not involve dimerization, and that dimerization is not necessary for expression of DAAO activity. We propose that both apoenzyme forms share a common reconstitution mechanism, which includes a step of conformational interconversion of an enzymatically active intermediate to the final holoenzyme.

Purification of Rhodotorula gracilis D-amino acid oxidase.

A protocol is presented for preparing Rhodotorula gracilis D-amino acid oxidase in homogeneous form and in high yield in 3 to 4 days. The method takes advantage of (a) cell rupture by alternate freeze-thawing, (b) use of DEAE-Sepharose to bind contaminants, and (c) enzyme binding to a Mono S column. The D-amino acid oxidase isolated by this means has the same spectral and catalytic properties as the enzyme previously obtained, and possesses improved long-term stability.

Reaction of phenylglyoxal with arginine groups in D-amino-acid oxidase from Rhodotorula gracilis.

D-Amino-acid oxidase from Rhodotorula gracilis was irreversibly inactivated by phenylglyoxal in a biphasic process. The fast phase was completed in less than 1 min. Its extent was linearly dependent on phenylglyoxal concentration and was not influenced by the presence of FAD or benzoate, a pseudo-substrate. The second phase of inactivation was due to a simple second-order reaction. The presence of FAD exerted only partial protection; the second-order rate constants of inactivation were 8.3 M-1 min-1 for holoprotein and 18.0 M-1 min-1 for apoprotein. The addition of benzoate completely protected against this second phase of inactivation. Efforts to isolate the enzyme modified at a single arginine residue at the end of the fast phase were unsuccessful, but analysis of the enzyme isolated at the end of the slow phase identified an arginine residue, protected by benzoate, that is highly conserved in all D-amino-acid oxidases and corresponds to Arg283 in the pig kidney enzyme. Modification of this residue is directly involved in the inactivation process during the slow phase. This arginine may represent the basic residue ion pairing with the carboxylate group of the substrate or the residue interacting with the flavin N1-C2 = O locus.

Chemical modification of lysyl residues of Rhodotorula gracilis D-amino acid oxidase.

D-amino acid oxidase from the yeast Rhodotorula gracilis is irreversibly inactivated by reaction with TNBS with complete inactivation accompanied by covalent modification of lysine residues of the protein. The inactivation was biphasic, the fast phase being dependent on TNBS concentration and completed in less than 1 minute. The competitive inhibitor benzoate afforded partial protection against inactivation during the fast phase of the process, with no effect on the slow phase. The pH curve of inactivation (slow phase) indicates the involvement of a residue(s) with a pK of 8.2. Amino acid analyses showed that in the fast phase of inactivation 1.6 lysine residues were modified, whereas up to 13 lysine residues were modified in the slow phase of inactivation. Our data show that TNBS behaves as an active site-directed reagent in yeast D-amino acid oxidase and they suggest the presence of at least one essential lysyl residue at or near the active site.

Kinetic mechanisms of cholesterol oxidase from Streptomyces hygroscopicus and Brevibacterium sterolicum.

The kinetic properties of two cholesterol oxidases, one from Brevibacterium sterolicum (BCO) the other from Streptomyces hygroscopicus (SCO) were investigated. BCO works via a ping-pong mechanism, whereas the catalytic pathway of SCO is sequential. The turnover numbers at infinite cholesterol and oxygen concentrations are 202 s-1 and 105 s-1 for SCO and BCO, respectively. The rates of flavin reduction extrapolated to saturating substrate concentration, under anaerobic conditions, are 235 s-1 for BCO and 232 s-1 for SCO (in the presence of 1% Thesit and 10% 2-propanol). With reduced SCO the rate of Delta5-6-->Delta4-5 isomerization of the intermediate 5-cholesten-3-one to final product is slow (0.3 s-1). With oxidized SCO and BCO the rate of isomerization is much faster ( approximately 300 s-1), thus it is not rate-limiting for catalysis. The kinetic behaviour of both reduced COs towards oxygen is unusual in that they exhibit apparent saturation with increasing oxygen concentrations (extrapolated rates approximately 250 s-1 and 1.3 s-1, for BCO and SCO, respectively): too slow to account for catalysis. For BCO the kinetic data are compatible with a step preceding the reaction with oxygen, involving interconversion of reactive and nonreactive forms of the enzyme. We suggest that the presence of micelles in the reaction medium, due to the necessary presence of detergents to solubilize the substrate, influence the availability or reactivity of oxygen towards the enzyme. The rate of re-oxidation of SCO in the presence of product is also too slow to account for catalysis, probably due to the impossibility of producing quantitatively the reduced enzyme-product complexes.

Studies on the active centre of Rhodotorula gracilis D-amino acid oxidase and comparison with pig kidney enzyme.

D-Amino acid oxidase (EC 1.4.3.3) from Rhodotorula gracilis has been reconstituted with 8-chloro-, 8-mercapto-, 6-hydroxy-, 2-thio-, 5-deaza- and 1-deaza-FAD, and the properties of the resulting complexes have been studied and compared with those of the correspondingly modified pig kidney D-amino acid oxidases. Binding appears to be tight for most analogues, at least as tight as for native FAD (approximately 10(-8) M). 8-Mercapto- and 6-hydroxy-FAD bind in their para- and ortho-quinoid forms respectively to yeast D-amino acid oxidase, inferring the presence of a positive charge near the flavin N(1) position, as in the case of the mammalian enzyme. On the other hand, important differences in active-site microenvironment emerge: solvent accessibility to flavin position 8 is drastically restricted in yeast D-amino acid oxidase as indicated by the unreactivity of 8-chloro- and 8-mercapto-FAD enzyme with thiolates and alkylating agents. Significantly different microenvironments are also likely to occur around the flavin positions N(1)-C(2) = 0, N(3)-H and N(5). This is deduced from the differences in interaction of the two proteins with 1-deaza-FAD, 5-deaza-FAD and 2-thio-FAD and from the properties of the respective complexes. The same re-side flavin stereospecificity as shown by the mammalian enzyme was determined for the yeast enzyme using 8-hydroxy-5-deaza-FAD. Thus we can deduce the presence of a similar pattern of functional groups at the active centres of the two enzymes, while the fine tuning of specificity and regulation correlate with environmental differences at specific flavin loci.

Overexpression in Escherichia coli of a recombinant chimeric Rhodotorula gracilis d-amino acid oxidase.

This paper reports a novel expression system constructed to maximize the production in Escherichia coli of d-amino acid oxidase from the yeast Rhodotorula gracilis (RgDAAO). We produced a recombinant plasmid by the insertion of the cDNA encoding for the RgDAAO into the multiple cloning site of the expression vector pT7.7 (pT7-DAAO), downstream of the T7 RNA polymerase binding site. The pT7-DAAO, which encodes a fully active fusion protein with six additional residues at the N-terminus of DAAO, was used to transform the BL21(DE3) and BL21(DE3)pLysS E. coli cells. In the latter host and under optimal IPTG induction conditions, soluble and active chimeric DAAO was expressed in these cells up to 930 U/g of cell (and a fermentation yield of 2300 U/liter of fermentation broth), with a specific activity of 8.8 U/mg protein. RgDAAO represents approximately 8% of the total soluble protein content of the cell.

Redox potentials and their pH dependence of D-amino-acid oxidase of Rhodotorula gracilis and Trigonopsis variabilis.

The redox potentials and pH characteristics of D-amino-acid oxidase (EC 1.4.3.3; DAAO) from the yeast Rhodotorula gracilis and Trigonopsis variabilis were measured in the pH range 6.5-8.5 at 15 degrees C. In the free enzyme form, the anionic red semiquinone is quantitatively formed in both DAAOs, indicating that a two single-electron transfer mechanism is active. The semiquinone species is also thermodynamically stable, as indicated by the large separation of the single-electron transfer potentials. The first electron potential is pH-independent, while the second electron transfer is pH-dependent exhibiting a approximately -60 mV/pH unit slope, consistent with a one-electron/one-proton transfer. In the presence of the substrate analogue benzoate, the two-electron transfer is the thermodynamically favoured process for both DAAOs, with only a quantitative difference in the stabilization of the anionic semiquinone. Clearly binding of the substrate (or substrate analogue) modulates the redox properties of the two enzymes. In both cases, in the presence and absence of benzoate, the slope of Em vs. pH (-30 mV/pH unit) corresponds to an overall two-electron/one-proton transfer in the reduction to yield the anionic reduced flavin. This behaviour is similar to that reported for DAAO from pig kidney. The differences in potentials and the stability of the semiquinone intermediate measured for the three DAAOs probably stem from different isoalloxazine environments. In the case of R. gracilis DAAO, the low stability of the semiquinone form in the DAAO-benzoate complex can be explained by the shift in position of the side chain of Arg285 following substrate analogue binding.

Kinetic mechanism of D-amino acid oxidases from Rhodotorula gracilis and Trigonopsis variabilis.

The reaction of two D-amino acid oxidases from the yeasts Rhodotorula gracilis and Trigonopsis variabilis with the substrates alanine and valine in their 2-1H and 2-2H forms was studied employing the stopped-flow spectrophotometric technique. The turnover numbers at infinite substrate and oxygen concentrations were: 20,700/4,250 and 1,730/360 ([2-1H]/[2-2H]alanine and valine, respectively) for the Rhodotorula and 3,150/440 and 2,500/520 ([2-1H]/[2-2H]alanine and valine, respectively) for the Trigonopsis enzymes. The rates of anaerobic enzyme flavin reduction were 20,100/4,000 and 1,820/350 ([2-1H]/[2-2H]alanine and valine, respectively) for the Rhodotorula and 3,470/350 and 2,460/480 ([2-1H]/[2-2H]alanine and valine, respectively) for the Trigonopsis enzymes. The isotope effects on enzyme reduction were 5.0 and 5.2 for Rhodotorula and 9.9 and 5.1 for Trigonopsis D-amino acid oxidases with alanine and valine, respectively. This suggests that the intrinsic isotope effect on rupture of the substrate alpha-C-H bond can be as high as 10. The rate-determining step corresponds to the enzyme reductive half-reaction in contrast to the mammalian kidney enzyme where it is the product release from oxidized enzyme (Massey, V., and Gibson, Q.H. (1964) Fed. Proc. 23, 18-29). Upon anaerobic reaction with substrate, the yeast enzymes do not form the transient long wavelength absorbing species which are characteristic of the mammalian protein. This is due only in part to rapid dissociation of iminoacid product and is ascribed to intrinsic differences between the charge-transfer complexes of reduced enzyme flavin and product of the yeast as compared to the mammalian enzyme. With the Trigonopsis enzyme the flavin radical anion appears to be strongly stabilized and can be produced quantitatively.

Studies on the structural and functional aspects of Rhodotorula gracilis D-amino acid oxidase by limited trypsinolysis.

The structure-function relationships of purified Rhodotorula gracilis D-amino acid oxidase (in its holo-, apo- and holo-enzyme-benzoate complex forms) was analysed by digestion with trypsin. In all cases trypsin cleaves this 80 kDa dimeric enzyme at the C-terminal region, since the peptide bonds sensitive to proteinase attack are clustered in this region. Digestion of native enzyme with trypsin produced a nicked and truncated form of 38.3 kDa containing two polypeptides of 34 and 5 kDa starting from Met1 and Ala319 respectively, and with detachment of the Thr306-Arg318 and Glu365-Leu368 peptides. Our results show that this 'core', folded into a compact structure, is catalytically competent. The acquisition of this nicked form was marked by a shift from a dimeric to a monomeric active enzyme, a result never previously obtained. The deleted sequences, Thr306-Arg318 and Glu365-Leu368, are essential for the monomer-monomer interaction, and, in particular, the region encompassing Thr306-Arg318 should play an essential role in the dimerization process. interestingly, the Ser308-Lys321 sequence present in the lost peptide corresponds to a sequence not present in other known D-amino acid oxidases [Faotto, Pollegioni, Ceciliani, Ronchi and Pilone (1995) Biotechnol. Lett. 17, 193-198]. A role of the cleaved-off region for the thermostabilization of the enzyme is also discussed.

Evaluation of D-amino acid oxidase from Rhodotorula gracilis for the production of alpha-keto acids: a reactor system.

The study reports on the development of a bioreactor for the production of alpha-keto acids from D,L- or D-amino acids using Rhodotorula gracilis D-amino acid oxidase. D-amino acid oxidase was co-immobilized with catalase on Affi-Gel 10 matrix, and the reactor was operated as a continuous-stirred tank reactor (CSTR) or stirred tank with medium recycling conditions. The optimum substrate concentration and quantity of biocatalyst were determined (5 mM and 1.2 mg/L, respectively). Under optimum operating conditions, product formation was linearly related to both substrate and enzyme concentration, showing the system to be highly flexible. Under these conditions, in a stirred tank, over 90% conversion was achieved in 30 min with a maximum production of 0.23 g of pyruvic acid/day/enzyme units. Product was recovered by ion exchange chromatography. The operational stability of the reactor was high (up to 9.5 h of operation without loss of activity) and the inactivation half-life was not reached even after 18 h or 36 bioconversion cycles. This represents the first case of a reactor developed successfully with a D-amino acid oxidase.

Cholesterol oxidase from Streptomyces hygroscopicus and Brevibacterium sterolicum: effect of surfactants and organic solvents on activity.

We have studied systematically the effect of the non-ionic surfactants Thesit and Triton X-100, and of propan-2-ol (used as a substrate solubilizer) on the activity of the cholesterol oxidases from Streptomyces hygroscopicus (SCO) and Brevibacterium sterolicum (BCO). Low concentrations of Thesit lead to an activity increase with both enzymes; at higher surfactant concentrations the opposite effect occurs. Triton X-100 inactivates both enzymes at all concentrations. It is deduced that these surfactants exert their effects by interaction with the enzymes and not by affecting micellar phenomena. The effect of propan-2-ol on SCO, in contrast with that on BCO, depends on the buffer concentration (potassium phosphate). Other organic solvents induce results similar to those obtained with SCO and propan-2-ol. A significant difference between the two cholesterol oxidases emerges when stability is tested at 25 degrees C and in the presence of different concentrations of propan-2-ol: BCO activity is rapidly inactivated, whereas SCO still has 70% of the initial activity after 5 h in the presence of 30% propan-2-ol. From our results, SCO seems to be the catalyst of choice in comparison with BCO for the exploitation of cholesterol oxidases in biotechnology and applied biochemistry.