Functional and Structural Characterization of d-Aspartate Oxidase from Porcine Kidney: Non-Michaelis Kinetics due to Substrate Activation

d-Aspartate exists abundantly in mammalian pineal, pituitary, and adrenal glands (1–7), which are members of the neuroendocrine system. Recent studies indicate that d-aspartate is involved in the synthesis of hormones and their secretion in the neuroendocrine system (2, 5, 8–13). Importantly, the localization of d-aspartate in the system is inversely correlated to that of d-aspartate oxidase (DDO; EC 1.4.3.1), an FAD-dependent peroxisomal enzyme (2). DDO is the only enzyme known to be responsible for metabolizing d-aspartate. DDO catalyzes the oxidative deamination of dicarboxylic d-amino acids using O₂ to yield the corresponding imino acids and hydrogen peroxide; the imino acids are rapidly hydrolyzed to 2-oxo acids and ammonia. DDO-deficient mice show elevated levels of d-aspartate in the pituitary intermediate lobe, leading to diminished synthesis of proopiomelanocortin, and then melanotropin-dependent influences decrease to elevate body mass (13–15). To further elucidate the roles of both d-aspartate and DDO in the neuroendocrine system, it is important to examine the biochemical properties of DDO.

Difficulty in the purification of DDO from mammalian tissues has hampered the biochemical characterization of this oxidase. Purification of DDO from bovine kidney has succeeded (16), whereas DDO has been only partially purified from porcine kidney (17) and thyroid glands (18). On the other hand, the recombinant mammalian DDO has been obtained for human (19), bovine (20, 21), and mouse (22). Among these mammalian DDOs, the most detailed kinetic analysis has been carried out for bovine DDO (23–25). Bovine DDO exhibits significant substrate activation at d-aspartate concentrations above 1.0 mM (24, 25): its Lineweaver-Burk plot shows an apparent downward curvature (25). As the physiological concentrations of d-aspartate are in the range of 0–3 mM (1, 3, 5), it is important to elucidate the mechanism of this substrate activation. However, prior to this study, all the apparent k_cat and K_m values reported for DDO have been determined using only data obtained at a limited range of substrate concentrations where a Michaelian behavior is apparently observed.

To explain the non-Michaelian kinetics of DDO described above, Hamilton has proposed a reaction mechanism where not only the oxidized form of DDO (E_o) but also the reduced form of DDO (E_r) accepts substrate and catalyzes the reaction (see Fig. 1) (25). This model has never been applied to systematically analyze the actual kinetic data of DDO.

In the present study, for the first time, we purified DDO from porcine kidney, cloned the cDNA for the DDO, and expressed the enzyme in Escherichia coli. Both the native and recombinant porcine DDOs showed significant substrate activation at higher concentrations of d-aspartate and d-glutamate. We have proposed a reaction mechanism which combines the Hamilton model for DDO (25) and the reaction mechanism proposed for d-amino acid oxidase (26, 25 and the references cited therein). On the basis of our model, the steady-state kinetic properties of DDO were successfully described with four kinetic parameters. This study indicates that under physiological concentrations of d-amino acid substrate and O₂, both the oxidized and reduced forms of DDO catalyze the reaction with different catalytic competence.

EXPERIMENTAL PROCEDURES

Materials

Porcine kidney from female Sus scrofa was purchased from a local slaughter house and kept at about −35°C until use. d-Aspartic acid, d-glutamic acid, N-methyl-d-aspartic acid (NMDA), sodium pyruvate, 2-oxoglutaric acid, 2,4-dinitrophenylhydrazine, FAD, sodium potassium l-tartrate, l-tartaric acid, d-tartaric acid, meso-tartaric acid, and malonic acid were all purchased from Wako Pure Chemical Industries (Osaka, Japan). 3-Methyl-2-benzothiazolinone hydrazone hydrochloride (MBTH) was obtained from Aldrich and catalase was from Boehringer-Mannheim. All other chemicals were of analytical grade.

Enzyme Assays

During the purification, the 2,4-dinitrophenylhydrazine method of Yamada et al. (28) was used to measure oxalacetate generation from d-aspartate. The standard assay mixture (0.1 ml) contained 50 mM sodium pyrophosphate, pH 8.3, 50 mM d-aspartate, 50 μM FAD, and 10 μg catalase. The reaction was initiated by adding enzyme solutions. One unit of DDO activity was defined as the amount of enzyme producing 1 μmol of oxalacetate per min at 37°C under the standard assay conditions.

Kinetic data was obtained in a wide concentration range of d-amino acid by two methods. In the first method, polarographic measurement using a Model 5331 O₂ electrode (Yellow Springs, OH, USA) was used to measure O₂ consumption. We used a glass reaction vessel maintained at 37°C and put the reaction vessel in a box that was easy to purge with argon, as described previously (29). The reaction mixture (1.78 ml) contained 50 mM sodium pyrophosphate, pH 8.3, 50 μM FAD, 180 μg catalase, and varying amounts of d-amino acid substrate and O₂. The O₂ concentration was varied over the range of 1–1,000 μM using the method described previously (29). The reaction was started by the addition of an enzyme solution. The second method was the HPLC assay developed recently by us (30). This method quantified the accumulation of oxalacetate during incubation at 37°C for 0–30 min. Reactions were carried out in 200 μl of 50 mM sodium pyrophosphate, pH 8.3, containing 50 μM FAD, 20 μg catalase, and varying amounts of d-amino acid substrate. The reaction was started by the addition of the enzyme (1 µl) and stopped by the addition of 12.5% trichloroacetic acid (40 µl).

The enzyme stock solution for kinetics was prepared by overnight dialysis of 200 μl of the purified DDO stored at −20°C against 2 liters of 50 mM potassium phosphate, pH 6.75, containing 1 μM FAD and 10 mM Na,K l-tartrate. The concentration of the holo-subunit of the DDO stock solution was determined spectrophotometrically using a Shimadzu UV-2550 spectrometer and assuming a molar absorption coefficient of 11.3 m M⁻¹ cm⁻¹ at 455 nm (31).

The effects of pH on the activity of native and recombinant DDO for d-aspartate were examined at 37°C and at the ionic strength of 0.3 M (adjusted with NaCl) using the following buffer solutions: 0.1 M Mes-NaOH (pH 5.5–6.5), 0.1 M HEPES-NaOH (pH 6.5–7.5), 0.1 M Tris-HCl (7.5–8.5), and 0.1 M CHES-NaOH (8.5–9.5). The reaction mixture contained 50 mM d-aspartate in air-saturated buffers. The oxalacetate generation was measured using the 2,4-dinitrophenylhydrazine method, as described above.

Purification of Native DDO

Unless otherwise stated, all procedures were conducted at about 4°C and all buffers contained 10 mM Na,K l-tartrate. HPLC was performed at room temperature. The buffers used for the homogenization of porcine kidney contained no Na,K l-tartrate. We purified DDO from 90 g of porcine kidney cortex. Extremely gentle homogenization of the cortex was essential to avoid the contamination of certain proteins that were difficult to remove by later chromatographic steps. Thus, we treated 10 g of the kidney cortex at one time by the following method, and repeated the same procedure 9 times. We diced the cortex and homogenized partially in a glass homogenizer with only three gentle strokes of the piston using 10 ml of 5 mM MOPS, pH 7.4, containing 250 mM sucrose, 1 mM EDTA, and 0.1% (v/v) ethanol. The partial homogenates were centrifuged at 200 × g for 5 min. The supernatant obtained was centrifuged at 5,500 × g for 10 min. The precipitates were then completely homogenized with 10 mL of 20 mM potassium phosphate, pH 5.4, containing 10 μM FAD and 0.3 mM EDTA using the glass homogenizer. The obtained homogenates were combined (crude extract). The pH of the crude extract was adjusted to 5.4 at 4°C using 1.67 M acetic acid. The crude extract was then incubated at 55°C for 15 min using a water bath. After cooling down to 10°C, the heat-treated extract was centrifuged at 10,000 × g for 10 min. The supernatant (heat-treatment) was applied to an SP-Toyopearl column (5 × 2 cm, Tosoh) preequilibrated with 20 mM potassium phosphate, pH 5.4, containing 10 μM FAD and 0.3 mM EDTA. The column was washed with 5 volumes of the equilibration buffer, and the enzyme was then eluted with 20 mM potassium phosphate, pH 8.4, containing 10 μM FAD and 0.3 mM EDTA. The active fractions were pooled (SP-Toyopearl) and loaded onto a Q-Sepharose column (2.5 × 5 cm, Pharmacia) preequilibrated with the buffer used above for the elution. The column was developed with the buffer and the active fractions in the flow-through were pooled (Q-Sepharose). The Q-Sepharose fraction was diluted 2-fold with 2 M potassium phosphate, pH 6.75, containing 10 μM FAD, and 0.3 mM EDTA. The diluted Q-Sepharose fraction was applied to a PPG-Toyopearl column (2.5 × 2 cm, Tosoh) preequilibrated with 1 M potassium phosphate, pH 6.75, containing 10 μM FAD, and 0.3 mM EDTA. The column was washed with five volumes of the buffer and then the enzyme was eluted with 200 mM potassium phosphate, pH 6.75, containing 10 μM FAD, 0.3 mM EDTA, and 200 mM KCl. The pooled active fractions (PPG-Toyopearl) were concentrated to about 0.2 ml using an Ultra PL-30 membrane filter (Amicon) and a Microcon YM-30 membrane filter (Amicon) in that order. An aliquot (20 μl) of the concentrate was repeatedly injected into a TSKgel SuperSW3000 column (4.6 × 300 mm, Tosoh) preequilibrated with 200 mM potassium phosphate, pH 6.75, containing 10 μM FAD, 0.3 mM EDTA, and 200 mM KCl. Elution was performed with the buffer at a flow rate of 0.2 ml/min. The active fractions (SuperSW3000) were pooled and concentrated to about 0.2 ml with an Amicon Ultra PL-30 membrane filter. The concentrated SuperSW3000 fraction was dialyzed against 1 liter of 1 mM potassium phosphate, pH 6.75, containing, 0.3 mM EDTA overnight. The dialyzed sample was applied to a hydroxyapatite column (1 × 1.5 cm, Nacalai tesque) preequilibrated with 1 mM potassium phosphate, pH 6.75, containing 10 μM FAD, 1 mM Na,K l-tartrate, and 0.3 mM EDTA. The column was washed with five volumes of the equilibration buffer and then the enzyme was eluted with 20 mM potassium phosphate, pH 5.4, containing 10 μM FAD and 0.3 mM EDTA. The pooled active fractions (hydroxyapatite) were applied to a Bioassist-5S column (4.6 × 50 mm, Tosoh) preequilibrated with 20 mM potassium phosphate, pH 5.4, containing 10 μM FAD and 0.3 mM EDTA. The column was developed with a linear gradient of KCl concentration (16.7 mM/min) in the equilibration buffer at a flow rate of 0.5 ml/min. Each of the active fractions was subjected to SDS-PAGE, and the fractions showing a single protein band were pooled. The pooled fraction (Bio-5S) was stored at −20°C. No loss of activity was observed after at least one month storage.

Protein content was determined using a BCA protein assay kit (Wako Pure Chemical Industries) with bovine serum albumin as a standard.

Cloning of Porcine DDO cDNA

Total RNA was extracted from porcine kidney cortex using an RNeasy Midi kit (Qiagen). First strand cDNA was prepared from the total RNA using Super Script III First Strand Synthesis System (Invitrogen). The cDNA strand of porcine DDO was amplified by PCR using the first strand cDNA as a template and the following two sets of primers: forward primers, 5′-CCCATGGATACAGTACGGATTG-3′ and 5′-GCTTTTTCAGAGACAGGCCCATG-3′; reverse primers, 5′-GAGCTTAACGCCCTATGTCATAGC-3′ and 5′-GCTAATTTCTGCATCTGGGGAC-3′. These primers were designed on the basis of the nucleotide sequence of bovine kidney DDO (20). PCR was performed using Platinum pfx DNA polymerase kit (Invitrogen) using the following conditions: 94°C for 15 s, 55°C for 30 s, and 68°C for 1 min, for 35 cycles. The main PCR product obtained was cloned into pCR4-TOPO vector (Invitrogen) and sequenced using an ABI Prism 310 DNA Sequencer (Applied Biosystems) with a DYEnamic ET terminator kit (Amersham). The nucleotide sequence of the PCR product was similar to the corresponding sequences of bovine and human DDOs. To obtain the complete cDNA by rapid amplification of cDNA ends (RACE), we designed primers specific to porcine DDO on the basis of this nucleotide sequence. RACE was performed using a Gene Racer kit (Invitrogen) and the following primer sets: for the primary PCR, Gene Racer 5′ Primer and 5′-TTCTGGGGCCATCTGCTGTTGT-3′ for 5′-RACE, Gene Racer 3′ Primer and 5′-CCCATGGATACAGTACGGATTG-3′ for 3′-RACE; for the secondary nested PCR, Gene Racer 5′ nested Primer and 5′-CTACAGCTTTGATTTAGGAGCAGGGG-3′ for nested 5′-RACE, Gene Racer 3′ nested Primer and 5′-ACTTTGAGCACTTGGCCCCTCA-3′ for nested 3′-RACE. The nucleotide sequences of the PCR products were determined as described above.

Expression of Recombinant Porcine DDO

The open reading frame for porcine DDO cDNA was amplified by PCR using the first strand cDNA and the following primers, 5′-CACCATGGATACAGTACGGATTG-3′ and 5′-CTACAGCTTTGATTTAGGAGCAGGG-3′ (the sequence of the forward primer indicated in bold was added for directional TOPO cloning). The PCR product was cloned into pET100/D-TOPO vector (Invitrogen). The obtained expression plasmid (pETDDO) was used to transform BL 21 Star (DE 3) E. coli cells (Invitrogen). The cells were grown at 37°C to a turbidity of 0.5–0.8 at 600 nm in LB medium containing 100 μg/ml ampicillin. After the addition of isopropyl-1-thio-β-d-galactopyranoside to a final concentration of 1 mM, the culture was grown for a further 4–6 h to induce the expression of His₆-tagged DDO. The cells were harvested by centrifugation, and stored at −20°C until use.

Purification of Recombinant DDO

The E. coli cells (10 g) were resuspended in 90 ml of 20 mM sodium phosphate, pH 7.4, containing 10 μM FAD and 10 mM Na,K l-tartrate. The cells were disrupted on ice by sonication while maintaining the temperature below 10°C. The homogenates were centrifuged at 8,000 × g for 10 min. Solid ammonium sulfate was added to the supernatant to 30% saturation. After incubation for 30 min with continuous stirring, the solution was centrifuged at 10,000 × g for 10 min. Solid ammonium sulfate was further added to the resultant supernatant to 50% saturation and the mixture was incubated for 30 min, and then the precipitate was collected by centrifugation. The precipitate was dissolved in 3.0 ml of 20 mM sodium phosphate, pH 7.4, containing 10 μM FAD, 10 mM Na,K l-tartrate, 0.5 M NaCl, and 20 mM imidazole, and dialyzed against 3 liters of the phosphate buffer for 4 h. After removing insoluble materials by centrifugation, the dialysate was applied to a His Trap HP column (1 ml bed volume, Amersham) preequilibrated with the same buffer as used for the dialysis. The column was washed with five volumes of the buffer and then the recombinant enzyme was eluted with 20 mM sodium phosphate, pH 7.4, containing 10 μM FAD, 10 mM Na,K l-tartrate, 0.5 M NaCl, and 0.5 M imidazole. The pooled active fractions were concentrated to about 2.0 ml using an Ultra PL-30 membrane filter and a Microcon YM-30 membrane filter in that order. An aliquot of the concentrate (200 μl) was applied to a TSKgel G3000SW_XL column (7.8 × 300 mm) preequilibrated with 200 mM potassium phosphate, pH 6.75, containing 10 μM FAD and 0.3 mM EDTA. Elution was performed with the phosphate buffer at a flow rate of 0.8 ml/min and the active fractions were collected. To remove the N-terminal His₆-tag, the pooled active fractions were digested with 10 U of enterokinase using an EK Max kit (Invitrogen) for 24 h at 37°C. The recombinant DDO was recovered from the digests by gel filtration on a TSKgel SuperSW3000 column (4.6 × 300 mm) using the same phosphate buffer at a flow rate of 0.2 ml/min.

SDS-Polyacrylamide Gel Electrophoresis

SDS-PAGE was performed as described by Laemmli (32) using a 12.5% gel. Proteins in the gel were stained using a silver staining kit (Wako Pure Chemical Industries). N-Terminal sequencing was performed on polyvinylidene difluoride electrotransferred samples using an automated protein sequencer (Model 477A, Applied Biosysytem).

Molecular Weight Determination

The molecular weight of native DDO was determined by low-angle laser light scattering measurement combined with gel chromatography as described previously (33) using a TSKgel SuperSW3000 column (4.6 × 300 mm). The molecular weight standards used were ovalbumin monomer (M_r = 45,000), bovine albumin monomer (M_r = 66,300) and catechol 2,3-dioxygenase (Pseudomonas putida, homotetramer of M_r = 140,624); the former two proteins were from Sigma and the third was purified by the reported method (34). The subunit molecular weight of DDO was determined by matrix-assisted laser desorption mass spectrometry using a Voyager-DE RP mass spectrometer (PerSeptive Biosystems) and α-cyano-4-hydroxycinnamic acid as a matrix-forming material. Catechol 2,3-dioxygenase (Pseudomonas putida, subunit molecular weight of 35,156) was used for mass calibration.

Thin-layer Chromatography

To remove the exogenous FAD added in the buffer used for purification, an aliquot of the purified DDO (50 μg) was passed through a Sephadex G-25 column (5 × 30 mm) equilibrated with 50 mM potassium phosphate, pH 6.75. The enzyme eluted at the void volume was incubated at 80°C for 20 min in the dark, and then centrifuged to pellet the denatured protein. An aliquot of the supernatant was analyzed by thin-layer chromatography on a Linear-K Preadsorbent TLC Plate (Whatman, USA) with 5% Na₂HPO₄ · 12H₂O as the eluent. As a control, authentic FAD, FMN, and riboflavin were run separately or as a mixture under the same conditions.

Titration of DDO with Dicarboxylic Acids

Analysis of Kinetic Data

RESULTS

Purification of Porcine DDO

The DDO purified from porcine kidney migrated as a single band with molecular weight of 38,000 during SDS-PAGE (Fig. 2A), indicating that the purified enzyme was homogeneous. The results of a typical purification procedure are shown in Table 1. The purified DDO had a specific activity of 62 μmol/min/mg protein. Only a single peak of DDO activity appeared for all chromatographic steps performed.

The subunit molecular weight was estimated to be 37,000 by mass spectrometry (Fig. 2B). The molecular weight of native enzyme was determined to be of 146,000 by low-angle laser light scattering photometry (Fig. 2C), predicting a homotetrameric structure for native DDO. Determination of the N-terminal amino acid sequence of the purified enzyme failed probably due to the modification of the N-terminal residue.

The absorption spectrum of the purified DDO (Fig. 3) showed maxima at 273, 374, and 452 nm and a shoulder around 480 nm (A₂₇₃/A₄₅₅ ratio of 7.9). Unlike the DDO from bovine kidney (16), the purified enzyme showed no absorption above 600 nm, indicating that the enzyme did not contain 6-hydroxy-FAD. The yellow chromophore contained in the enzyme was completely released from the protein moiety by heat-denaturation. The isolated chromophore migrated as a single spot on the thin-layer chromatograph with the same R_f value of 0.83 as FAD migrated; R_f values for FMN and riboflavin were 0.64, and 0.54, respectively (data not shown). These results strongly suggest that the purified enzyme contains noncovalently bound FAD as the sole coenzyme. The recombinant enzyme (see later section) also contained only FAD as judged by the absorption spectrum and the thin-layer chromatographic analysis.

Cloning and Sequencing of the DDO Gene

The nucleotide sequence of porcine DDO cDNA and the deduced amino acid sequence are shown in Fig. 4. The cDNA contains 1984 base-pairs coding for a 341-amino-acid protein with a molecular weight of 37,315. This is in agreement with the molecular weight estimated by SDS-PAGE and mass spectrometry (Fig. 2). The cDNA exhibits a 5′ untranslated region of 73 base-pairs and a 3′ untranslated region of 885 base-pairs. A polyadenylation signal (AATAAA) is located 21 base-pairs upstream from the beginning of the poly A tail.

Porcine DDO consisted of exactly the same number of amino acids (341) as other mammalian DDOs, and the identity in amino-acid sequence between porcine DDO and other mammalian DDOs was 88%, 83%, and 75% for bovine, human, and mouse DDOs, respectively (Fig. 5). The GAGVMG sequence in the N-terminal sequence of porcine DDO (indicated by solid underline in Fig. 4) has a feature identical to the dinucleotide binding motif of GXGXXG, suggesting that these residues are responsible for FAD binding. The C-terminal SKL sequence (indicated by broken underline in Fig. 4) is identical to one of the minimal peroxisome-targeting signals (35), indicating that porcine DDO is a peroxisomal enzyme. As indicated in Fig. 5, the three arginine residues (R216, R237, and R278) and one tyrosine residue (Y223) are conserved; these are proposed as catalytically important residues (19, 22, 36).

Overexpression of Recombinant Porcine DDO

Recombinant DDO expressed in E. coli cells carrying pETDDO was purified by a four-step procedure as described under “experimental procedures”. About 1 mg of homogeneous DDO was obtained from 3 liters of isopropyl-1-thio-β-d-galactopyranoside-induced culture. The recombinant DDO had a specific activity of 58 μmol/min/mg protein, and showed a single protein band corresponding to a molecular weight of 38,000 on an SDS-PAGE gel (data not shown). Unlike the native DDO, the N-terminal amino acid sequence of the recombinant DDO could be determined to be identical to that predicted from the nucleotide sequence up to residue 41 (data not shown).

Comparison between Recombinant and Native DDO Activities

Both native and recombinant DDO showed similar specific activities of 62 and 58 μmol/min/mg protein under standard assay conditions. First, we compared the pH-dependence of the initial velocity between the native and recombinant DDO using 50 mM d-aspartate and air-saturated buffers (Fig. 6A). Both enzymes showed an essentially identical pH-profile of their activity: the activity increased gradually as pH increased from pH 5.5 to 7.5, exhibited an almost constant value over pH 7.5 to 9.0, and decreased as the pH increased over pH 9.0. At pH 8.3, the activity (40–45 μmol/min/mg protein, Fig. 6A) was less than the specific activities determined under standard assay conditions (60 μmol/min/mg protein). This difference is probably due to the effect of the buffer species pyrophosphate or Tris on the enzyme, as observed for d-amino acid oxidase. Next, we examined at pH 8.3 the dependence of the initial velocity on the concentration of d-aspartate. No significant difference was found in the dependence of the reaction catalyzed by the native and recombinant DDO on the concentration of d-aspartate (Fig. 6B). Therefore, in the following, we used the recombinant enzyme for the study of the steady-state kinetics.

Kinetic Properties of Porcine DDO

First, we obtained initial velocity of DDO over a wide concentration range of d-aspartate (0.01–80 mM) using air-saturated buffer (pH 8.3) (Fig. 7A–D). As is apparent in Fig. 7C and D, the [S]/v versus [S] plot curved downward over the concentration range of 0.1 to 1 mM, indicating that substrate activation occurred as the d-aspartate concentration increased above 0.2 mM. Second, we examined the dependence of the activity on the concentration of d-glutamate (0.01–100 mM) (Fig. 7E–H). The [S]/v versus [S] plot curved downward over the concentration range of 2 to 40 mM (Fig. 7G and H), although the curvature was less prominent than that observed for d-aspartate. Significant substrate activation occurred at concentrations of d-glutamate above 4 mM. Lastly, we obtained the initial velocity as a function of the concentration of NMDA (0.01–100 mM) (Fig. 7I–L). In this case, the [S]/v versus [S] plot was almost linear (Fig.7K and L), and the initial velocity showed a weak tendency to decrease as the NMDA concentration increased over 20 mM (Fig. 7I). This result suggested that a weak but significant level of substrate inhibition occurred when the NMDA concentration was higher than 20 mM.

All of the kinetic data shown in Fig. 7 fitted to Eq. 3 with good quality; the best fit curves are drawn in Fig. 7. The best-fit parameters are summarized in Table 2. Relationships between the elementary rate constants in Fig. 1 and the kinetic parameters of A₁, A₃, A₄, and A₅ are listed in Table 3. As explained in the  Appendix, the initial part of the [S]/v versus [S] plot is approximately linear and the slope of this linear portion is A₄/A₁. At higher values of [S], the plot becomes approximately linear again, but the slope of the liner portion is A₅. Using the parameter values listed in Table 2, the value of the initial slope (A₄/A₁) was calculated to be 0.380 and 0.313 s for d-aspartate and d-glutamate, respectively, about 14- and 3-fold larger than the corresponding value of the slope at higher [S] (A₅), whereas the A₄/A₁ value (0.0215 s) was slightly smaller than the A₅ value of for NMDA. Hence, these calculations confirm the downward curvature of the [S]/v versus [S] plot (substrate activation) for d-aspartate and d-glutamate, and the upward curvature of the plot (substrate inhibition) for NMDA.

Figure 8 shows the dependence of the initial velocity of DDO on a range of the O₂ concentration (10–10,00 μM) in the presence of 25 mM d-aspartate, 50 mM d-glutamate, or 25 mM NMDA. We could not determine the initial velocity at the O₂ concentration above 1 mM due to the low solubility of O₂. All of the data obtained apparently obeyed the ordinary Michaelis–Menten equation. Therefore, we fitted the data to Eq. 5 as described in “experimental procedures”. The best-fit values of apparent k_cat and K_mO2 are listed in Table 4, and the theoretical curves are drawn in Fig. 8 using these values. The apparent k_cat/K_mO2 value of porcine DDO for d-aspartate and NMDA was 2.81 × 10⁵ and 3.16 × 10⁵ M⁻¹s⁻¹, respectively, and that for d-glutamate was 5.48 × 10⁴ M⁻¹s⁻¹, about one-order smaller compared to the two former d-amino-acids.

Titration of DDO by Dicarboxylic Acids

To purify native and recombinant DDOs with higher levels of specific activity, the addition of Na,K l-tartaric acid to the buffer used was indispensable. We examined the interaction of DDO with malonate and three stereo-isomers of tartrate by measuring the change induced in the visible absorption of DDO upon the addition of these dicarboxylates. Figure 9 shows typical titration data for l-tartrate. The visible spectrum of the l-tartrate-DDO complex (Fig. 9A) showed a considerable increase of the shoulder around 480 nm and a red shift of the absorption maximum from 452 to 456 nm. All of the data obtained fitted well to Eq. 1 (see the theoretical curve drawn in Fig. 9B). The K_d values are shown in Table 5. Among the three stereo-isomers of tartrate, DDO showed the highest affinity to meso-tartrate (K_d value of 118 μM).

DISCUSSION

In spite of the physiological importance of DDO especially in the mammalian neuroendocrine system, biochemical characterization of DDO has been very limited compared to d-amino acid oxidase, which catalyzes similar oxidative deamination of d-amino acids with non-anionic side-chain groups and plays important roles in the metabolism of d-serine, a gliotransmitter, in the mammalian brain (e.g.30, 30). Purification of mammalian DDO has succeeded only from bovine kidney prior to this study. The purified porcine DDO is a homotetramer, whereas bovine DDO is a monomeric enzyme (16). The porcine DDO contained noncovalently bound FAD as a coenzyme. Unlike DDO purified from bovine kidney cortex (16), the porcine enzyme contained no 6-hydroxy-FAD. The N-terminal amino acid of the native porcine DDO was somewhat modified, whereas that of the recombinant DDO was not modified. However, the presence and absence of the modification of the N-terminal residue was not important for activity. Multiple forms of DDO mRNA have been isolated from human brain (19) and mouse tissues (14). However, a single active peak of DDO always appeared during all of the chromatography stages used for the purification of DDO from porcine kidney, suggesting that only one active form of DDO protein is expressed in porcine kidney.

We isolated and sequenced the complete cDNA coding for the porcine DDO for the first time, and overexpressed the recombinant DDO in E. coli cells. We did not find any significant difference in catalytic properties between the native and recombinant DDOs.

DDO from bovine kidney showed substrate activation at high d-aspartate concentrations (23–25), resulting in a downward curved Lineweaver-Burk plot (the 1/v versus 1/[S] plot). Porcine DDO showed similar substrate activation for d-aspartate. In addition to d-aspartate, substrate activation of porcine DDO was also found for d-glutamate. Interestingly, porcine DDO was subject to substrate inhibition at high NMDA concentrations.

Both activation and inhibition by high concentrations of d-amino acid substrates can be explained by the reaction model shown in Fig. 1. There are two important points in this model, as Hamilton has pointed out (25). The first point is that the complex between the reduced form of enzyme and the imino-acid product (E_r–P) releases the product (P) more rapidly than the oxidation of the E_r–P complex by O₂. Due to this rapid release of product, substantial levels of free reduced form DDO (E_r) are produced under steady-state conditions. In the case of the d-amino acid oxidase reaction, it is supposed that the E_r–P complex is rapidly oxidized to the oxidized form enzyme–product complex (E_o–P) without the release of the product from the E_r–P complex (26, 37); the bound product facilitates the oxidation of the reduced FAD coenzyme by O₂ (26). In fact, probably due to the lack of the production of free reduced enzyme, neither substrate activation nor substrate inhibition has been observed for d-amino acid oxidase. The second point is that the free reduced form of DDO (E_r) catalyzes the reaction as follows: (1) binding of d-amino acid substrate (S) to form the E_r–S complex, (2) oxidation of the E_r–S complex by O₂ to produce the E_o–S complex and hydrogen peroxide, (3) dehydrogenation of the substrate in the E_o–S complex to produce the E_r–P complex, and finally (4) the release of the product from the E_r–P complex, and regeneration of the catalyst E_r.

If the oxidation rate of the free reduced form of the enzyme (E_r) by O₂ is higher than that of the E_r–S complex by O₂, then substrate inhibition will appear at higher concentrations of substrate, as observed for NMDA (Fig. 7I). If the oxidation rate of the E_r–S complex is higher than that of free E_r, then substrate activation will appear at higher substrate concentrations, as observed for d-aspartate and d-glutamate (Fig. 7C, D, G, and H). If there is no partition between the substrate binding to E_r and the oxidation of E_r (either k₆ = 0 or k₈ = 0 in Fig. 1), then Eq. 2 is transformed to a Michaelian type equation and no substrate activation/inhibition occurs.

Recently, engineering the substrate specificity of d-amino acid oxidase to accept dicarboxylic d-amino acids as substrates has been reported (38, 38). However, substrate activation/inhibition has never been reported for these engineered d-amino acid oxidases. Although models of the active site structures of mouse and bovine DDOs have been built using the three-dimensional structure of d-amino acid oxidase and the homology in the amino acid sequences (22, 36), the three-dimensional structure of DDO has never been solved hitherto. Interestingly, bacterial L-aspartate oxidase, which has no sequence homology with DDO, also exhibits prominent substrate activation at L-aspartate concentrations above 1.0 mM (39).

Apparent k_cat and K_m values for DDO from various sources have been determined using a linear portion of the Lineweaver-Burk plots obtained at higher concentrations of d-amino acid substrates. In the present model of DDO reaction, the corresponding values of k_cat and K_m can be calculated according to Eqs E20 and E21 (see  Appendix) using the kinetic parameters listed in Table 2. The results are summarized in Table 6 together with the apparent k_cat and K_m values reported for bovine and human DDOs. The apparent k_cat/K_m values of porcine and bovine DDOs for d-aspartate are smaller than those for NMDA by 3.8- and 3.4-fold, respectively, whereas the apparent k_cat/K_m value of human DDO for d-aspartate is 3.5-fold larger than that for NMDA. All of the apparent k_cat/K_m values of porcine, bovine, and human DDOs for d-aspartate are in the order of 1 × 10⁴ M⁻¹ s⁻¹. These results suggest that the catalytic efficiency and the substrate preference of porcine DDO at higher substrate concentrations are similar to those of bovine and human DDO.

As the concentration of d-aspartate at physiological conditions is in the range of 0–3 mM (1, 3, 5), the kinetic properties of DDO at lower concentrations of d-aspartate are important. As explained in the  Appendix, the initial portion of the v versus [S] curve is characterized by the apparent k_cat and K_m values defined by Eqs E16 and E17. These values are summarized in Table 7. At lower concentrations of d-amino acid substrates, the apparent k_cat/K_m value of porcine DDO for d-aspartate is 3.7-fold larger than that for NMDA, indicating that the relative catalytic efficiency is reversed by low or high concentrations of substrate.

In conclusion, we have revealed for the first time the kinetic and structural properties of porcine DDO. Substrate activation/inhibition of porcine DDO was successfully analyzed by a reaction model where the oxidase shows different catalytic properties depending on the redox states of FAD. To further elucidate the molecular mechanism of substrate activation/inhibition of the oxidase, crystallographic and transient-state kinetic studies are necessary. In addition, the crystal structure of FAD-bound l-aspartate oxidase from E. coli has been solved (40). Considering the information accumulated for DDO and the related enzymes together, the redox-dependent interaction of DDO with substrate and imino-acid product seems to play key roles in the substrate activation/inhibition of DDO.

Supplementary data are available at JB online.

The present study was supported by a Grant-in Aid for Scientific Research (C) (NO. 18590528) (to T. I., and H. T.) from the Ministry of Education, Culture, Sports, Sciences, and Technology of Japan, and Grant-in Aid (Heisei era 18) from Shiga University of Medical Science. The nucleotide sequence data reported in this paper will appear in the DDBJ, EMBL and GenBank nucleotide sequence databases with accession number AB271762.

REFERENCES

APPENDIX

If k₃ = 0, then the model becomes the same as Hamilton's model (25). If k₆ = 0, then the model becomes the same one proposed for d-amino acid oxidase catalysis (26, 27), and Eq. E1 becomes an ordinary equation of Michaelis–Menten type.

Approximate Expressions

Dependence of the Initial Velocity on O₂ Concentration

Abbreviations:

Abbreviations:
 
• DDO

  d-aspartate oxidase

 
• NMDA

  N-methyl-d-aspartic acid

 
• MBTH

  3-methyl-2-benzothiazolinone hydrazone hydrochloride

 
• RACE

  rapid amplification of cDNA ends