Monitoring of phenylketonuria: a colorimetric method for the determination of plasma phenylalanine using L-phenylalanine dehydrogenase

A simple, rapid, accurate, and precise colorimetric assay for the determination of L-phenylalanine in plasma samples using L-phenylalanine dehydrogenase [L-phenylalanine:NAD+-oxidoreductase (deaminating)] from Rhodococcus sp. M 4 is described. The enzyme catalyzes the NAD-dependent oxidative deamination of L-phenylalanine. However, the equilibrium of reaction favors L-phenylalanine formation. By stoichiometric coupling of this reaction with diaphorase/iodonitro tetrazolium chloride (INT) the formed NADH converts INT to a formazan whereby the reaction is displaced in favor of phenylpyruvate. Using a kinetic approach the increase in absorbance at 492 nm shows linearity over more than 30 min. Deproteinized standard solutions of L-phenylalanine in the range from 30 to 1200 mumol/liter show a linearity between the dAformazan/30 min and the substrate concentration. In phenylketonuria (PKU) plasma samples no interferences caused by L-tyrosine or phenylpyruvic acid are seen. Applicability is demonstrated by comparative determination of plasma L-phenylalanine of treated PKU patients by the colorimetric method and automated amino acid analysis.     

Enzymatic cycling assay for phenylpyruvate

Enzymatic cycling assays for the determination of L-phenylalanine and phenylpyruvate in deproteinized tissue extracts are described. Assay 1 couples glutamine transaminase K with L-phenylalanine dehydrogenase. Assay 2 combines phenylalanine dehydrogenase, L-amino acid oxidase, and catalase. In both assays, tyrosine and some other amino acids (or their alpha-keto acid analogs) can replace phenylalanine (or phenylpyruvate) to a small extent. Thus, if phenylalanine is to be measured a correction must be made for the nonspecificity of the reaction. By removing phenylalanine on a cation-exchange column it was possible to measure phenylpyruvate in tissue extracts. Concentrations of phenylpyruvate (mumol/kg) in normal rat liver, kidney, and brain were 2.1 +/- 1.1 (n = 8), 1.8 +/- 0.4 (n = 4), and 3.3 +/- 0.6 (n = 4), respectively.     

Rhodococcus L-phenylalanine dehydrogenase: kinetics, mechanism, and structural basis for catalytic specificity

Phenylalanine dehydrogenase catalyzes the reversible, pyridine nucleotide-dependent oxidative deamination of L-phenylalanine to form phenylpyruvate and ammonia. We have characterized the steady-state kinetic behavior of the enzyme from Rhodococcus sp. M4 and determined the X-ray crystal structures of the recombinant enzyme in the complexes, E.NADH.L-phenylalanine and E.NAD(+). L-3-phenyllactate, to 1.25 and 1.4 A resolution, respectively. Initial velocity, product inhibition, and dead-end inhibition studies indicate the kinetic mechanism is ordered, with NAD(+) binding prior to phenylalanine and the products' being released in the order of ammonia, phenylpyruvate, and NADH. The enzyme shows no activity with NADPH or other 2'-phosphorylated pyridine nucleotides but has broad activity with NADH analogues. Our initial structural analyses of the E.NAD(+).phenylpyruvate and E.NAD(+). 3-phenylpropionate complexes established that Lys78 and Asp118 function as the catalytic residues in the active site [Vanhooke et al. (1999) Biochemistry 38, 2326-2339]. We have studied the ionization behavior of these residues in steady-state turnover and use these findings in conjunction with the structural data described both here and in our first report to modify our previously proposed mechanism for the enzymatic reaction. The structural characterizations also illuminate the mechanism of the redox specificity that precludes alpha-amino acid dehydrogenases from functioning as alpha-hydroxy acid dehydrogenases.     

Novel phenylalanine dehydrogenases from Sporosarcina ureae and Bacillus sphaericus. Purification and characterization

NAD+-dependent phenylalanine dehydrogenases were purified 1,500- and 1,600-fold, and crystallized from Sporosarcina ureae SCRC-R04 and Bacillus sphaericus SCRC-R79a, respectively. The purified enzymes were homogeneous as judged by disc gel electrophoresis. The enzyme from S. ureae has a molecular weight of 305,000, while that of B. sphaericus has a molecular weight of 340,000. Each is probably composed of eight subunits identical in molecular weight. The S. ureae enzyme showed a high substrate specificity in the oxidative deamination reaction acting on L-phenylalanine, while that of B. sphaericus acted on L-phenylalanine and L-tyrosine. The enzymes had lower substrate specificities in the reductive amination reaction acting on alpha-keto acids. The Sporosarcina enzyme acted on phenylpyruvate, alpha-ketocaproate, alpha-keto-gamma-methylthiobutyrate and rho-hydroxyphenylpyruvate. The Bacillus enzyme acted on rho-hydroxyphenylpyruvate, phenylpyruvate, and alpha-keto-gamma-methylthiobutyrate. The enzyme from B. sphaericus catalyzes The enzyme from B. sphaericus catalyzes the transfer of pro-S (B) hydrogen from NADH.     

The biochemical basis for growth inhibition by L-phenylalanine in Neisseria gonorrhoeae

Clinical isolates of Neisseria gonorrhoeae are commonly subject to growth inhibition by phenylpyruvate or by L-phenylalanine. A blockade of tyrosine biosynthesis is indicated since inhibition is reversed by either L-tyrosine or 4-hydroxyphenylpyruvate. Phenylalanine-resistant (PheR) and phenylalanine-sensitive (PheS) isolates both have a single 3-deoxy-D-arabino-heptulosonate 7-phosphate (DAHP) synthase that is partially inhibited by L-phenylalanine (80%). However, PheS and PheR isolates differ in that the ratio of phenylpyruvate aminotransferase to 4-hydroxyphenylpyruvate aminotransferase is distinctly greater in PheS isolates than in PheR isolates. A mechanism for growth inhibition is proposed in which phenylalanine exerts two interactive effects. (i) Phenylalanine decreases precursor flow to 4-hydroxyphenylpyruvate through its controlling effect upon DAHP synthase; and (ii) phenylalanine is largely transaminated to phenylpyruvate, which saturates both aminotransferases, preventing transamination of an already limited supply of 4-hydroxyphenylpyruvate to L-tyrosine.     

Spectral and kinetic studies on Pseudomonas L-phenylalanine oxidase (deaminating and decarboxylating)

Pseudomonas L-phenylalanine oxidase (deaminating and decarboxylating) contains two FAD molecules in one molecule of the enzyme (Koyama, H. (1983) J. Biochem. 93, 1313-1319). When the enzyme was mixed anaerobically with L-phenylalanine, beta-2-thienylalanine, L-tyrosine, or L-methionine, a spectral species (purple intermediate) with a broad absorption band around 540 nm was observed with each substrate, and decayed slowly. From the data on the overall reaction kinetics, the rate of the L-phenylalanine oxidase reaction was expressed as follows. e/v = e/Vm + A/[S] + B/[O2] where e represents the concentration of enzyme unit, v the rate of the overall reaction, Vm the maximum velocity, and A and B are constants. Furthermore, the reactions of the enzyme with beta-2-thienylalanine (mostly an oxygenase substrate) and L-methionine (an oxidase substrate) were analyzed by the "stopped flow" method. The following scheme for the mechanism of L-phenylalanine oxidase reaction with both substrates is proposed, based on the data obtained. (formula; see text) Where Eox represents the oxidized form of the enzyme unit, EoxS the enzyme unit (oxidized form)-substrate compound, X the purple intermediate with a characteristic broad absorption band around 540 nm, S the substrate and P the product.     

Purification, characterization and induction of L-phenylalanine ammonia-lyase in Phaseolus vulgaris

The enzyme L-phenylalanine ammonia-lyase was purified from leaves of Phaseolus vulgaris by Sephacryl S-200 gel filtration and Sepharose-4-B--succinyl-aminoethyl-L-phenylalanine affinity chromatography. L-Phenylalanine ammonia-lyase was specifically eluted from the affinity matrix with its substrate L-phenylalanine at 20-25 degrees C. The purified enzyme was shown to be homogeneous by gel electrophoresis both in presence and absence of SDS. Its Mr, determined by gel filtration and non-denaturing gel electrophoresis, was 320,000 +/- 9000 and 330,000 +/- 4000 respectively. After SDS electrophoresis only one band of Mr 83,000 +/- 4000 was detected, indicating that the enzyme is an oligomer containing four subunits. The pH optimum of enzyme activity was 8.8-9.2. Ampholyte isoelectrofocusing in polyacrylamide demonstrated the presence of a single charged species at pH 4.2. The homogeneous enzyme catalyzed the deamination of L-phenylalanine to trans-cinnamate but did not catalyze the transamination of L-phenylalanine to L-phenylpyruvate. The enzyme showed Km 1.25 mM for L-phenylalanine. Antibodies to homogeneous L-phenylalanine ammonia-lyase recognised specific epitopes on L-phenylalanine aminotransferase as demonstrated by immunoaffinity purification and immunoblotting. The induction of L-phenylalanine ammonia-lyase activity during phaseollin biosynthesis in the Phaseolus vulgaris--Colletotrichum lindemuthianum interaction was regulated by an increase in enzyme concentration resulting from an increase in de novo synthesis of L-phenylalanine ammonia-lyase protein.     

Purification and characterization of a dimeric phenylalanine dehydrogenase from Rhodococcus maris K-18.

NAD+-dependent phenylalanine dehydrogenase (EC 1.4.1.) was purified to homogeneity from a crude extract of Rhodococcus maris K-18 isolated from soil. The enzyme had a molecular mass of about 70,000 daltons and consisted of two identical subunits. The enzyme catalyzed the oxidative deamination of L-phenylalanine and several other L-amino acids and the reductive amination of phenylpyruvate and p-hydroxyphenylpyruvate. The enzyme required NAD+ as a natural coenzyme. The NAD+ analog 3-acetylpyridine-NAD+ showed much greater coenzyme activity than did NAD+. D-Phenylalanine, D-tyrosine, and phenylethylamine inhibited the oxidative deamination of L-phenylalanine. The enzyme reaction was inhibited by p-chloromercuribenzoate and HgCl2. Initial-velocity and product inhibition studies showed that the reductive amination proceeded through a sequential ordered ternary-binary mechanism. NADH bound first to the enzyme, followed by phenylpyruvate and then ammonia, and the products were released in the order L-phenylalanine and NAD+. The Michaelis constants were as follows: L-phenylalanine, 3.8 mM; NAD+, 0.25 mM; NADH, 43 microM; phenylpyruvate, 0.50 mM; and ammonia, 70 mM.     

Oxidation and oxygenation of L-amino acids catalyzed by a L-phenylalanine oxidase (deaminating and decarboxylating) from Pseudomonas sp. P-501

A number of L-amino acids and derivatives were tested as substrates for the purified Pseudomonas L-phenylalanine oxidase. The reaction products of these amino acids were analyzed by high performance liquid chromatography and the kinetic properties of the reactions were partially characterized. In addition to L-phenylalanine, L-tyrosine, DL-o-tyrosine, DL-m-tyrosine, p-fluoro-DL-phenylalanine and beta-2-thienyl-DL-alanine served as substrates for both oxidation and oxygenation catalyzed by the enzyme. On the other hand, L-methionine and L-norleucine were enzymically converted to the corresponding alpha-keto acids with the consumption of oxygen and with the formation of ammonia and hydrogen peroxide in stoichiometric amounts. Kinetic studies showed that the Km values for oxidation and oxygenation of L-phenylalanine by the enzyme were 2.04 mM and 1.96 mM for oxygen, and 13.3 microM and 11.1 microM for L-phenylalanine, respectively. omega-Phenyl fatty acids such as phenylacetic acid, 3-phenylpropionic acid and 4-phenylbutyric acid were competitive inhibitors of the enzyme towards L-phenylalanine. Both oxidation and oxygenation of L-phenylalanine by the enzyme were also inhibited by phenylacetic acid competitively.     

Distribution, purification, and characterization of thermostable phenylalanine dehydrogenase from thermophilic actinomycetes.

Phenylalanine dehydrogenase (L-phenylalanine:NAD oxidoreductase, deaminating; EC 1.4.1.-) was found in various thermophilic actinomycetes. We purified the enzyme to homogeneity from Thermoactinomyces intermedius IFO 14230 by heat treatment and by Red Sepharose 4B, DEAE-Toyopearl, Sepharose CL-4B, and Sephadex G-100 chromatographies with a 13% yield. The relative molecular weight of the native enzyme was estimated to be about 270,000 by gel filtration. The enzyme consists of six subunits identical in molecular weight (41,000) and is highly thermostable: it is not inactivated by incubation at pH 7.2 and 70 degrees C for at least 60 min or in the range of pH 5 to 10.8 at 50 degrees C for 10 min. The enzyme preferably acts on L-phenylalanine and its 2-oxo analog, phenylpyruvate, in the presence of NAD and NADH, respectively. Initial velocity and product inhibition studies showed that the oxidative deamination proceeds through a sequential ordered binary-ternary mechanism. The Km values for L-phenylalanine, NAD, phenylpyruvate, NADH, and ammonia were 0.22, 0.078, 0.045, 0.025, and 106 mM, respectively. The pro-S hydrogen at C-4 of the dihydronicotinamide ring of NADH was exclusively transferred to the substrate.     

Kinetic isotope effect of the L-phenylalanine oxidase from Pseudomonas sp. P-501

Pseudomonas L-phenylalanine oxidase (deaminating and decarboxylating) mainly catalyzes oxygenation when L-phenylalanine is used as the substrate, but oxidation when L-methionine is used as the substrate. Using [C(alpha)-H]-DL-methionine and [C(alpha)-D]-DL-methionine as substrate, the reductive half reaction of FAD cofactor of enzyme has been studied by stopped-flow spectrophotometry. The rate of reduction of FAD cofactor has a kinetic isotope effect (KIE) of 5.4 and 4.1 in the absence and presence of 30% glycerol, respectively. The KIE is independent of temperature, but the rates of the reductive half reaction are dependent on temperature, indicating that thermally induced motion at the active site drives the H-transfer reaction by H-tunneling.     

The Activity of Rat Pineal and Brain Tyrosine Hydroxylase During the Daily Cycle of Light and Darkness as Determined by the Modified 14CO2 Assay Method

A previous published assay method for tyrosine hydroxylase by the evolution of 14CO2 was modified to a two-step procedure to allow reliable measurement of large numbers of samples containing low tyrosine hydroxylase activity. The reliability of the method was examined in detail. Properties of rat brain and pineal tyrosine hydroxylase solubilized with 0.2% Triton X-100 were as follows. The apparent Km values of the brain enzyme for L-tyrosine with 1 mM-(6-DL)-5,6,7,8-tetrahydro-L-erythro-biopterin (BPH4) as cofactor and for BPH4 with 62 microM-L-tyrosine as substrate were approximately 25 microM and 85 microM, respectively. The Km's for L-tyrosine with 1 mM-(6-DL)-5,6,7,8-tetrahydro-6-methylpterin (6MPH4) as cofactor and for 6MPH4 with 210 microM-L-tyrosine as substrate were 68 microM and 270 microM, respectively. The marked substrate inhibition by high concentrations of L-tyrosine was observed only when BPH4 was used as cofactor. High concentrations of BPH4 inhibited the reaction slightly. The kinetic properties of tyrosine hydroxylase in the pineal extract were similar to those of the brain enzyme, except that a Lineweaver-Burk plot of reciprocal velocity versus the reciprocal concentration of BPH4 with 62 microM-L-tyrosine as substrate deviated downward at a BPH4 concentration of about 100 microM. Analyses of the plot indicated that the peculiar kinetic property may represent either the reaction occurring at two independent sites or with two forms (6L- and 6D-isomers) of the tetrahydrobiopterin cofactor, with apparent Km for BPH4 of 23 microM and 1025 microM, respectively, or the negatively cooperative ligand binding with a Hill coefficient of 0.72. Based on the results obtained as reported above the standard assay conditions of tyrosine hydroxylase in tissue extracts were established. Using the assay method and conditions, the absence of the daily rhythmicity of tyrosine hydroxylase in rat pineal glands and three discrete brain areas was demonstrated. The findings, especially on pineal tyrosine hydroxylase, are discussed in relation to the daily change of noradrenaline turnover.     

Production of L-phenylalanine from phenylpyruvate using resting cells of Escherichia coli

Summary We describe the production of L-phenylalanine from phenylpyruvate using resting cells of a genetically modified strain of Escherichia coli. Fermentations were carried out by continuously raising the feeding rate of D-glucose. We reached a biomass of 10 g/l dry weight and an aminotransferase activity of 14000 U/l The maximum phenylalanine concentration achieved was 173 mmol/l with a phenylpyruvate molar conversion yield of 95%.Dedicated to Prof. Dr. Heinz Harnisch on the occasion of his 60th birthday     

Stereospecificity of the effect of flupenthixol isomers on the substrate inhibition of brain tyrosine hydroxylase

Stereospecificity of the effect of neuroleptics on substrate inhibition of isolated brain tyroxine hydroxylase is shown. Flupentixole cis-isomer eliminates substrate inhibition of the enzyme. The effect is concentration-dependent and is well marked within the tyrosine concentration range 10-6-10-4 M. Flupentixole trans-isomer in the same concentrations has no effect on substrate inhibition of tyrosine hydroxylase. In the presence of cis-flupentixole, the reaction rate plotted against tyrosine concentration is a hyperbole with a plateau at 160-360 microM tyrosine. In the presence of trans-isomer, as in the control sample, the relationships between the reaction rate and tyrosine concentration are depicted by a curve with a maximum (at 110-140 microM tyrosine). Like ftorphenazine, flupentixole isomer fails to eliminate the inhibitory action of alpha-methyl paratyrosine, which indicates the interaction of neuroleptics with the tyrosine-binding site of the enzyme molecules in the noncatalytic centrer. It is suggested that the interaction of the neuroleptics with the regulatory area of tyrosine hydroxylase might by important in the molecular mechanism of their action.     

A monofunctional and thermostable prephenate dehydratase from the archaeon Methanocaldococcus jannaschii

Prephenate dehydratase (PDT) is an important but poorly characterized enzyme that is involved in the production of L-phenylalanine. Multiple-sequence alignments and a phylogenetic tree suggest that the PDT family has a common structural fold. On the basis of its sequence, the PDT from the extreme thermophile Methanocaldococcus jannaschii (MjPDT) was chosen as a promising representative of this family for pursuing structural and functional studies. The corresponding pheA gene was cloned and expressed in Escherichia coli. It encodes a monofunctional and thermostable enzyme with an N-terminal catalytic domain and a C-terminal regulatory ACT domain. Biophysical characterization suggests a dimeric (62 kDa) protein with mixed alpha/beta secondary structure elements. MjPDT unfolds in a two-state manner (Tm = 94 degrees C), and its free energy of unfolding [DeltaGU(H2O)] is 32.0 kcal/mol. The purified enzyme catalyzes the conversion of prephenate to phenylpyruvate according to Michaelis-Menten kinetics (kcat = 12.3 s-1 and Km = 22 microM at 30 degrees C), and its activity is pH-independent over the range of pH 5-10. It is feedback-inhibited by L-phenylalanine (Ki = 0.5 microM), but not by L-tyrosine or L-tryptophan. Comparison of its activation parameters (DeltaH(++)= 15 kcal/mol and DeltaS(++)= -3 cal mol-1 K-1) with those for the spontaneous reaction (DeltaH(++) = 17 kcal/mol and DeltaS(++)= -28 cal mol-1 K-1) suggests that MjPDT functions largely as an entropy trap. By providing a highly preorganized microenvironment for the dehydration-decarboxylation sequence, the enzyme may avoid the extensive solvent reorganization that accompanies formation of the carbocationic intermediate in the uncatalyzed reaction.     

Microbial L-phenylalanine ammonia-lyase. Purification, subunit structure and kinetic properties of the enzyme from Rhizoctonia solani.

1. Phenylalanine ammonia-lyase (EC 4.3.1.5) was purified to homogeneity from the acetone-dried powders of the mycelial felts of the plant pathogenic fungus Rhizoctonia solani. 2. A useful modification in protamine sulphate treatment to get substantial purification of the enzyme in a single-step is described. 3. The purified enzyme shows bisubstrate activity towards L-phenylalanine and L-tyrosine. 4. It is sensitive to carbonyl reagents and the inhibition is not reversed by gel filtration. 5. The molecular weight of the enzyme as determined by Sephadex G-200 chromatography and sucrose-density-gradient centrifugation is around 330000. 6. The enzyme is made up of two pairs of unidentical subunits, with a molecular weight of 70000 (alpha) and 90000 (beta) respectively. 7. Studies on initial velocity versus substrate concentration have shown significant deviations from Michaelis-Menten kinetics. 8. The double-reciprocal plots are biphasic (concave downwards) and Hofstee plots show a curvilinear pattern. 9. The apparent Km value increases from 0.18 mM to as high as 5.0 mM with the increase in the concentration of the substrate and during this process the Vmax, increases by 2-2.5-fold. 10. The value of Hill coefficient is 0.5. 11. Steady-state rates of phenylalanine ammonia-lyase reaction in the presence of inhibitors like D-phenylalanine, cinnamic, p-coumaric, caffeic, dihydrocaffeic and phenylpyruvic acid have shown that only one molecule of each type of inhibitor binds to a molecule of the enzyme. These observations suggest the involvement of negative homotropic interactions in phenylalanine ammonia-lyase. 12. The enzyme could not be desensitized by treatment with HgCl2, p-chloromercuribenzoic acid or by repeated freezing and thawing.     

Selective production of L-aspartic acid and L-phenylalanine by coupling reactions of aspartase and aminotransferase in Escherichia coli

With L-aspartate (L-Asp) as the amino donor, L-phenylalanine (L-Phe) can be prepared from phenylpyruvate (PPA) via an amination reaction mediated by aminotransferase (encoded by aspC). On the other hand, L-Asp can be produced by an aspartase (encoded by aspA) -catalyzed reaction using fumaric acid as substrate. To overproduce aspartase in Escherichia coli, the aspA gene was cloned and overexpressed 180 times over the wild-type level. The use of AspA-overproducing E. coli strain for L-Asp production exhibited an 83% conversion, approaching to the theoretical yield, whereas the wild-type strain obtained scarcely L-Asp. Furthermore, the recombinant strain overproducing both AspA and AspC was able to produce L-Asp and L-Phe simultaneously by using fumaric acid and PPA as substrates. As a result, the conversion yields obtained for L-Asp and L-Phe were 78% and 85%, respectively. In sharp contrast, the wild-type strain attained a conversion of L-Phe less than 15% and an undetectable level of L-Asp. This result illustrates a potential and attractive process to yield both L-Asp and L-Phe by coupling AspA and AspC. A further study on the repeated use of the recombinant strain immobilized with calcium alginate showed that after eight batch runs L-Asp conversion maintained roughly constant (around 75%), whereas L-Phe conversion dropped to 65% from 81%. This result indicates the stability of AspA being superior to AspC.     

1-Amino-2-phenylethylphosphonic acid: an inhibitor of L-phenylalanine ammonia-lyase in vitro

L(-)-, and D(+)-enantiomers of 1-amino-2-phenylethylphosphonic acid (PheP), a phosphonic analogue of phenylalanine, inhibit the activity of L-phenylalanine ammonia-lyase (EC 4.3.1.5) of potato tuber tissue in vitro. The apparent type of inhibition depends on concentration of PheP; as the concentration of D-PheP is raised from 10(-5) M to 2.5 X 10(-3) M, the type of inhibition shifts from competitive through mixed and non-competitive to uncompetitive. L-PheP exerts either a competitive or mixed-type inhibition at low (10(-6)-10(-5) M) or moderate (5 X 10(-5)-2 X 10(-4) M) concentration. Ki for the concentration range of competitive inhibition were 6.5 X 10(-6) M, 5.3 X 10(-5)M and 1.6 X 10(-5) M for L-, D-, and D,L-PheP, respectively. These Ki values are valid for a relatively narrow range of L-Phe concentration (0.2-4 mM) as L-phenylalanine ammonia-lyase does not follow the Michaelis-Menten kinetics of the reaction.     

Metabolic engineering of the E. coli L-phenylalanine pathway for the production of D-phenylglycine (D-Phg)

D-phenylglycine (D-Phg) is an important side chain building block for semi-synthetic penicillins and cephalosporins such as ampicillin and cephalexin. To produce d-Phg ultimately from glucose, metabolic engineering was applied. Starting from phenylpyruvate, which is the direct precursor of L-phenylalanine, an artificial D-Phg biosynthesis pathway was created. This three-step route is composed of the enzymes hydroxymandelate synthase (HmaS), hydroxymandelate oxidase (Hmo), and the stereoinverting hydroxyphenylglycine aminotransferase (HpgAT). Together they catalyse the conversion of phenylpyruvate via mandelate and phenylglyoxylate to D-Phg. The corresponding genes were obtained from Amycolatopsis orientalis, Streptomyces coelicolor, and Pseudomonas putida. Combined expression of these activities in E. coli strains optimized for the production of L-phenylalanine resulted in the first completely fermentative production of D-Phg.