YGL196W of Saccharomyces cerevisiae encodes a putative protein that is unidentified but is predicted to have a motif similar to that of the N-terminal domain of the bacterial alanine racemase. In the present study we found that YGL196W encodes a novel D-serine dehydratase, which belongs to a different protein family from that of the known bacterial enzyme. The yeast D-serine dehydratase purified from recombinant Escherichia coli cells depends on pyridoxal 5′-phosphate and zinc, and catalyses the conversion of D-serine into pyruvate and ammonia with the Km and kcat values of 0.39 mM and 13.1 s−1 respectively. D-Threonine and β-Cl-D-alanine also serve as substrates with catalytic efficiencies which are approx. 3 and 2% of D-serine respectively. L-Serine, L-threonine and β-Cl-L-alanine are inert as substrates. Atomic absorption analysis revealed that the enzyme contains one zinc atom per enzyme monomer. The enzyme activities toward D-serine and D-threonine were decreased by EDTA treatment and recovered by the addition of Zn2+. Little recovery was observed with Mg2+, Mn2+, Ca2+, Ni2+, Cu2+, K+ or Na+. In contrast, the activity towards β-Cl-D-alanine was retained after EDTA treatment. These results suggest that zinc is involved in the elimination of the hydroxy group of D-serine and D-threonine. D-Serine dehydratase of S. cerevisiae is probably the first example of a eukaryotic D-serine dehydratase and that of a specifically zinc-dependent pyridoxal enzyme as well.
Several D-amino acids have been discovered in eukaryotes, including mammals, and they are reported to have various physiological functions. For example, D-aspartate has been found in the mammalian central nervous system and endocrine tissues and has been suggested to be involved in the regulation of hormone release [1,2] and melatonin  and testosterone syntheses . D-Serine occurs primarily in the mammalian brain, with the highest concentrations in the regions that are rich in NMDA (N-methyl-D-aspartate) receptors, and modulates brain function as a coagonist of the NMDA receptor [5,6]. D-Alanine has been found in the anterior pituitary gland and pancreas , but its physiological role is currently unclear. To understand the function of D-amino acids in eukaryotes, we have been studying the roles of D-amino acids and their metabolizing enzymes in two yeasts, Saccharomyces cerevisiae and Schizosaccharomyces pombe, which are model organisms of eukaryotic cells. S. pombe, which utilizes several D-amino acids as a nitrogen source, contains enzymes acting on D-amino acids such as alanine racemase , serine racemase (PDB accession number 1V71) and D-amino acid oxidase . On the other hand, S. cerevisiae cannot utilize D-amino acids as a nitrogen source , although D-amino acids are incorporated into the cells via a general amino acid permease (Gap1p) . Moreover D-amino acids show toxicity and inhibit S. cerevisiae cell growth [10,11]. In Escherichia coli cells, several D-amino acids have been reported to serve as a substrate of aminoacyl tRNA synthases. The formation of D-aminoacyl tRNA is suggested to lower the amount of normal L-aminoacyl tRNAs and cause a delay of cell growth . In S. cerevisiae cells, D-tyrosine was found to serve as a substrate of tyrosine tRNA synthase . It is possible that the formation of D-aminoacyl tRNAs is one of the reasons for the toxicity of D-amino acids to S. cerevisiae cells. S. cerevisiae possesses several protective systems against D-amino acid toxicity, such as D-Tyr-tRNATyr deacylase catalysing the deacylation of D-Tyr-tRNATyr  and D-amino acid N-acetyltransferase . D-Amino acid N-acetyltransferase catalyses the acetyl group transfer from acetyl CoA to an amino group of various D-amino acids . The resultant N-acetylated D-amino acids are excluded from the S. cerevisiae cells more efficiently than free D-amino acids .
Heterologous expression of the D-amino acid-metabolizing enzyme also protects the S. cerevisiae cells from D-amino acid toxicity. For example, S. cerevisiae cell growth was inhibited by D-alanine but was rescued by the expression of S. pombe alanine racemase . Alanine racemase depends on PLP (pyridoxal 5′-phosphate) and catalyses the interconversion between L- and D-alanine. The enzyme occurs ubiquitously in eubacteria, and is responsible for the formation of D-alanine, an essential component of the peptidoglycan layer of bacterial cell walls. Eukaryotes, except S. pombe, have no homologous genes with those of the bacterial alanine racemase. S. pombe probably obtained the alanine racemase gene through horizontal transfer from proteobacteria . S. cerevisiae does not contain alanine racemase but has two genes, YBL036C and YGL196W, encoding proteins with a motif common to alanine racemase. X-ray crystallography of Ybl036cp revealed that the protein is bound with PLP and the whole structure closely resembles that of the N-terminal domain of alanine racemase . An unidentified hypothetical protein, Ygl196wp, shows weak sequence similarity to the bacterial alanine racemase and, on the Pfam database (http://www.sanger.ac.uk/Software/Pfam/), is predicted to have a common motif with the N-terminal domain of alanine racemase.
In the present study, we cloned and expressed the YGL196W gene into E. coli cells and purified the gene products. We have demonstrated that YGL196W of S. cerevisiae encodes a PLP-dependent D-serine dehydratase. The enzyme, which we name DsdSC (D-serine dehydratase of S. cerevisiae), is probably the first example of a eukaryotic D-serine dehydratase and belongs to a different protein family from that of the bacterial D-serine dehydratase . DsdSC is also unique in its zinc dependency.
Amino acids and lactate dehydrogenase from rabbit muscle were purchased from Sigma–Aldrich. Glutamate dehydrogenase from bovine liver was from Wako Pure Chemicals. 2,4-DNP (2,4-dinitrophenylhydrazine) was from Kanto Chemicals. Restriction enzymes were from Takara. KOD-plus DNA polymerase was from TOYOBO. Synthetic oligonucleotides were from Fasmac. Histidine-binding resin was from Novagen (EMD Bioscience). DEAE-TOYOPEARL 650M was obtained from Tosoh. All other chemicals were of analytical grade.
Bacterial strain and culture conditions
S. cerevisiae BY4742 (MAT α, his3Δ1, leu2Δ0, lys2Δ0, ura3Δ0) and BY4742Δygl196w strains were obtained from Invitrogen. S. cerevisiae BY4742 cells were grown at 30 °C in YPD [1% (w/v) yeast extract/2% (w/v) peptone/2% (w/v) glucose] or SC medium . E. coli XL1-Blue and BL21(DE3) cells were used for the plasmid construction and gene expression respectively. Recombinant E. coli cells were cultivated at 37 °C in LB (Luria–Bertani) medium containing 50 μg/ml ampicillin.
Cloning of the YGL196W gene
The chromosomal DNA of S. cerevisiae BY4742 was isolated with an ISOPLANT Kit (Wako Pure Chemicals), and digested with HindIII. The resultant chromosomal DNA fragment was used as the template for the amplification of YGL196W by PCR with KOD-plus DNA polymerase, and the following oligonucleotides as primers: DsdSCup, 5′-ATGCTCGAGGTTCTATCTCAATATAAAGGGTGCTCAG-3′ (forward primer, the XhoI site is underlined) and DsdSCrv, 5′-ACCAAGCTTACCATTTCTGAAAAGGTAACCAAACATCG-3′ (reverse primer, the HindIII site is underlined). The amplified DNA fragment was digested with XhoI and HindIII, separated by agarose gel electrophoresis, and purified with a GeneClean II kit (Q-BIO gene). The amplified DNA was then ligated into pET15b (Novagen) digested with XhoI and HindIII. The resultant plasmid, pDsdSC, encodes the Ygl196wp with an N-terminal tag, which consists of six histidine residues and ten other amino acids residues providing a thrombin cleavage site. Construction of the plasmid was verified by DNA sequencing.
Expression and purification of the recombinant D-serine dehydratase
The E. coli BL21(DE3) cells harbouring pDsdSC were grown at 37 °C in LB medium containing 50 μg/ml ampicillin. When D610 reached 0.5, YGL196W was expressed by the addition of IPTG (isopropyl β-D-thiogalactoside) to a final concentration of 0.5 mM, and the cultivation was continued at 30 °C for a further 3 h. All subsequent steps were carried out at 0–4 °C. The E. coli cells (wet weight of 12 g) were harvested by centrifugation (4000 g for 5 min at 4 °C), resuspended in binding buffer consisting of 20 mM Tris/HCl (pH 7.9), 500 mM NaCl and 5 mM imidazole, and sonicated with a UP2005 Ultraschallprozessor (Hielscher Ultrasonics). The lysate was centrifuged at 20000 g for 30 min, and the supernatant was applied to a Ni2+-chelating column (5 ml) pre-equilibrated with the binding buffer. The column was washed with 50 ml of washing buffer, consisting of 20 mM Tris/HCl (pH 7.9), 500 mM NaCl and 80 mM imidazole. Ygl196wp was eluted with elution buffer consisting of 20 mM Tris/HCl (pH 7.9), 500 mM NaCl and 1 M imidazole. The enzyme solution was dialysed against PBS0 buffer (pH 7.5) consisting of 10 mM Na2HPO4, 2.7 mM KCl, 1.8 mM KH2PO4, 20 μM ZnCl2 and 20 μM PLP, and then applied to the DEAE-TOYOPEARL column (7 ml) pre-equilibrated with PBS0 buffer. The column was thoroughly washed with the PBS0 buffer containing 20 mM NaCl, and the enzyme was eluted with a linear gradient of 20–80 mM NaCl in PBS0 buffer. The fractions showing the enzymatic activity were pooled and concentrated by ultrafiltration. Purity of the enzyme was confirmed by SDS/PAGE.
The absorption spectrum of Ygl196wp was recorded in PBS buffer (pH 7.5) consisting of 140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4 and 20 μM PLP, with a Shimadzu UV-2450 spectrophotometer.
SDS/PAGE was performed using a 10% gel by the method of Laemmli [17a]. After electrophoresis, the gel was stained with 0.1% Coomassie Brilliant Blue R-250 and destained with 7% acetic acid containing 5% ethanol.
HPLC analyses were carried out with a Shimadzu SCL-10A system (Shimadzu) equipped with a Cosmosil 5C18 column (4.6 mm×150 mm; Nacalai Tesque) unless otherwise stated.
Enantioselective determination of amino acids with HPLC
We examined whether Ygl196wp catalysed the reactions of pyridoxal enzymes, such as racemization, decarboxylation and dehydratation, with various amino acids as substrates. Decreased amounts and formation of the amino acids during the Ygl196wp reaction were analysed by the enantioselective measurement of amino acids with HPLC after the amino acids were derivatized to fluorescent diastereomers . The Ygl196wp reaction was carried out in a 200 μl mixture containing 50 mM Hepes/NaOH buffer (pH 8.0), 20 μM PLP, 10 mM amino acid and an appropriate amount of enzyme at 30 °C for 1h. The reaction was stopped by the addition of 200 μl of 10% TCA (trichloroacetic acid) and incubated at 4 °C for 15 min. After centrifugation of the mixture at 19000 g for 10 min, the supernatant was extracted three times with water-saturated diethyl ether for the removal of TCA. Then, amino acids were derivatized with Boc-L-Cys-OPA (where Boc is t-butoxycarbonyl and OPA is o-phthaladehyde) as described previously . Separation of the derivatized amino acids with HPLC was accomplished by a linear gradient of 0–60% mobile phase B (47% acetonitrile in a 0.1 M acetate buffer, pH 6.0) in mobile phase A (7% acetonitrile in a 0.1 M acetate buffer, pH 6.0) in 60 min at a flow rate of 0.8 ml/min at room temperature (25 °C). Elution was monitored with an RF-10A fluorescence detector (Shimadzu) with excitation and emission wavelengths of 344 and 443 nm respectively. The method of the enantioselective determination of amino acids was also used for the analyses of D-serine contamination in D-threonine and β-Cl-D-alanine.
Identification of oxo acid produced from D-serine
The oxo acid formed from D-serine through the Ygl196wp reaction was identified with HPLC after the product was derivatized with MBTH (3-methyl-2-benzothiazolinone hydrazone) as described previously . The reaction mixture (200 μl) containing 50 mM Hepes/NaOH buffer (pH 8.0), 20 μM PLP, 1 mM D-serine and 2 μg of Ygl196wp was incubated at 30 °C for 1 h. A control experiment was carried out with a similar reaction mixture except that D-serine was replaced by pyruvate or hydroxypyruvate. The reaction was stopped by the addition of 200 μl of 10% TCA. After the reaction mixture was diluted 10-fold, a 240 μl portion was withdrawn and mixed with 10 μl of 1 mM 2-oxoglutarate as an internal standard. After centrifugation of the mixture at 19000 g for 15 min, 100 μl of the supernatant was mixed with 200 μl of 1 M sodium acetate buffer (pH 5.0) and 80 μl of 0.1% MBTH. The mixture was incubated at 50 °C for 30 min, and a 20 μl portion was subjected to HPLC. The derivatized oxo acids were separated by a linear gradient of 0–100% buffer B [90% acetonitrile, 0.1% TFA (trifluoroacetic acid)] in buffer A (20% acetonitrile, 0.1% TFA) over 15 min at a flow rate of 1.2 ml/min at room temperature. Elution was detected with a SDP-10A UV-visible detector (Shimadzu) at a wavelength of 350 nm.
The amount of ammonia formed from D-serine was assayed with an indophenol reagent as described previously . A reaction mixture (200 μl) consisting of 50 mM potassium phosphate (pH 8.0), 20 μM PLP, various concentrations of D-serine and 1 μg of DsdSC was incubated at 30 °C for 60 min. An aliquot of the reaction mixture (112 μl) was withdrawn, and mixed with 40 μl of a phenol/nitroprusside solution and 48 μl of alkaline hypochlorite solution. The mixture was incubated for 45 min at 30 °C. The intensity of the developed blue colour was measured at 635 nm.
Ygl196wp (DsdSC) was routinely assayed by measuring the amount of pyruvate formed with 2,4-DNP. The reaction mixture (50 μl) containing 50 mM Hepes/NaOH buffer (pH 8.0), 20 μM PLP, 10 mM D-serine and an appropriate amount of the enzyme was incubated at 30 °C for 15 min. The reaction was stopped by the addition of 50 μl of 0.05% 2,4-DNP in 2 M HCl. The mixture was incubated for 5 min at 30 °C, followed by the addition of 100 μl of ethanol and 125 μl of 10 M NaOH. After incubation of the mixture for 10 min at room temperature, the absorbance at 515 nm was measured. For kinetic analyses with D-serine and β-Cl-D-alanine, the pyruvate formed was assayed spectrophotometrically by following the oxidation of NADH in the coupling system with rabbit lactate dehydrogenase. The reaction mixture contained 50 mM Hepes/NaOH buffer (pH 8.0), 20 μM PLP, various concentrations of each substrate, 0.3 mM NADH, 10 units of lactate dehydrogenase and 1.5 μg of DsdSC in a final volume of 1 ml. The reaction was started by the addition of DsdSC, and the decrease in absorbance at 340 nm was monitored at 30 °C.
For kinetic analyses with D-threonine, the ammonia formed was assayed spectrophotometrically by following the oxidation of NADH in the coupling system with bovine glutamate dehydrogenase. The reaction mixture contained 50 mM Hepes/NaOH buffer (pH 8.0), 20 μM PLP, various concentrations of D-threonine, 0.3 mM NADH, 5 mM 2-oxoglutarate, 10 units of glutamate dehydrogenase and an appropriate amount of DsdSC in a final volume of 1 ml. The reaction was started by the addition of DsdSC, and the decrease in absorbance at 340 nm was monitored at 30 °C.
Determination of proteins
The concentration of DsdSC was assayed with its molar absorption coefficient at 280 nm, 47850 M−1·cm−1, which was estimated from the tyrosine and tryptophan contents .
EDTA treatment of DsdSC
To obtain the metal-free enzyme, DsdSC was dialysed against PBS buffer containing 5 mM EDTA (pH 7.3) at 4 °C for 16 h, followed by dialysis against PBS buffer (pH 7.3) without EDTA under the same conditions.
Atomic absorption measurement
The amount of zinc atom in DsdSC was determined by atomic absorption spectroscopy on a Varian SpectrAA-50 spectrometer.
To obtain the molecular mass of the enzyme, the purified DsdSC was subjected to gel-permeation chromatography with a TSK G3000 SW column (7.8 mm×300 mm) (Tosoh) equipped on a Shimadzu SCL-10A HPLC system. DsdSC was eluted with 50 mM potassium phosphate buffer (pH 7.0) containing 300 mM KCl at a flow rate of 1 ml/min at room temperature. Elution of the protein was monitored by measuring the absorbance at 280 nm. The column was calibrated with β-galactosidase (molecular mass, 464 kDa), glutamate dehydrogenase (246 kDa), BSA (69.2 kDa), ovalbumin (42.8 kDa) and ribonuclease A (13.7 kDa).
To evaluate the secondary structure, CD spectra from 200 to 260 nm were measured with 5 μM enzyme at room temperature with a Jasco J-720WI spectropolarimeter with a 0.1 cm light-path cell. To assess the thermostability of DsdSC, the CD at 220 nm was measured with increasing temperature from 5 °C to 75 °C at a rate of 1 °C/min.
Expression and purification of the recombinant Ygl196wp in E. coli cells
To study the properties of Ygl196wp, we cloned and expressed the YGL196W gene in E. coli cells. The recombinant cells produced Ygl196wp with six N-terminal histidine residues and a thrombin cleavage site when IPTG was added to the culture. Ygl196wp was purified from the recombinant E. coli cells by Ni2+-chelating and DEAE-TOYOPEARL column chromatography. The purified enzyme showed a single protein band upon SDS/PAGE with a molecular mass of 50 kDa (Figure 1A), which is compatible with that calculated from its amino acid sequence (50059.2 Da). The absorption spectrum of the purified Ygl196wp exhibited absorption maxima at 278 and 417 nm, suggesting that Ygl196wp contained PLP (Figure 1B). The 417 nm peak was decreased by the addition of NaCNBH3 with a concomitant increase in the absorption maximum at 320 nm. These are the typical phenomena occurring upon the reduction of an aldimine linkage between PLP and the ω-amino group of the lysine residue of protein.
SDS/PAGE (A) and absorption spectra (B) of DsdSC
Identification of Ygl196wp as a D-serine dehydratase
We examined whether Ygl196wp catalysed the reactions of pyridoxal enzymes, such as racemization, decarboxylation and dehydratation, with various D- and L-amino acids as substrates. After the reaction, D- and L-amino acids, oxo acids and ammonia in the reaction mixture were assayed by enantioselective HPLC measurement, the 2,4-DNP method, and indophenol methods respectively. We found that Ygl196wp converted D-serine into oxo acid and ammonia. To identify the oxo acid produced, which was expected to be pyruvate or hydroxypyruvate, the reaction product was modified with MBTH and subjected to HPLC analysis . As shown in Figure 2, the oxo acid formed from D-serine was identified as pyruvate, not hydroxypyruvate. Stoichiometric analysis revealed that 1 mol of D-serine was converted into 1 mol of pyruvate and ammonia (see Supplementary Table S1 at http://www.biochemj.org/bj/409/bj4090399add.htm). During the reaction, no hydrogen peroxide was produced, which was confirmed by the horseradish peroxidase reaction with TOOS [N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3-methylaniline sodium salt dihydrate] and 4-aminoantipyrine . The results indicated that Ygl196wp was not a D-amino acid oxidase. Neither NAD+ nor NADP+ was required in the Ygl196wp reaction, which suggested that Ygl196wp was not a dehydrogenase. Eukaryotic serine racemase is known to catalyse the dehydration of D- and L-serine , but Ygl196wp showed no reactivity toward L-serine and had no racemase activity. On the basis of these results, we concluded that Ygl196wp is a D-serine dehydratase. We thus named Ygl196Wp DsdSC (D-serine dehydratase of S. cerevisiae). DsdSC is probably the first example of a eukaryotic D-serine dehydratase.
Identification of the oxo acid formed from D-serine by the DsdSC reaction
A Blast search on the protein databases with DsdSC as a query sequence gave several unknown proteins. Most of them were putative proteins of yeasts and fungi. Supplementary Figure S1 (at http://www.biochemj.org/bj/409/bj4090399add.htm) shows the alignment of the amino acid sequences of some of these proteins. DsdSC showed little overall sequence similarity with any identified pyridoxal enzymes, but had some similar motifs to D-threonine aldolase of Arthrobacter sp. strain DK-38 [24,25] as shown in Supplementary Figure S1. The primary structure of DsdSC was completely different from those of bacterial D-serine dehydratase  and eukaryotic serine racemase  possessing D-serine dehydratase activity. Both enzymes showed similarity to L-threonine dehydratase in their primary structures [16,23] and had two characteristic common motifs. One is the Ser-Xaa-Lys-Ile-Arg-Gly sequence, of which lysine is the PLP-binding site (see  for the bacterial D-serine dehydratase and PDB accession number 1V7l for the S. pombe serine racemases). Another conserved motif is the glycine-rich sequence, which is considered to interact with the phosphate group of PLP . DsdSC has neither overall sequence homology with these enzymes nor the two conserved sequences mentioned above. These results suggest that DsdSC belongs to a protein family that differs from that of the bacterial D-serine dehydratase and the eukaryotic serine racemase. This is compatible with the prediction that DsdSC has a motif such as that in the N-terminal domain of the alanine racemase, which belongs to the fold-type III groups of pyridoxal enzymes . In contrast, bacterial D-serine dehydratase belongs to the fold-type II group .
Cofactors essential for enzyme activity
Upon reduction with NaCNBH3, DsdSC lost its activity, which was not recovered by the addition of PLP (Table 1). Hydroxylamine and phenylhydrazine (both 5 mM) lowered the DsdSC activity to 23.7 and 4.2% respectively, when the activity was measured in the absence of PLP, and to 64.6 and 97.1% when it was measured in the presence of PLP (Table 1). These results suggest that PLP serves as a cofactor of DsdSC.
|Compound||With PLP||Without PLP|
|Compound||With PLP||Without PLP|
EDTA treatment was carried out as described in the text.
Effect of pH on enzyme activity
The effect of pH on the DsdSC activity was examined (see Supplementary Figure S2 at http://www.biochemj.org/bj/409/bj4090399add.htm). The maximum activity was obtained at around pH 8.0. Tris/HCl buffer showed an inhibitory effect.
Substrate specificity and kinetic properties of DsdSC
The substrate specificity of the enzyme was examined. In addition to D-serine, DsdSC showed little activity towards D-threonine and β-Cl-D-alanine with 3.2 and 1.7% efficiency (kcat) of that towards D-serine respectively (Table 2). We confirmed that there was no contamination of D-serine in the D-threonine and β-Cl-D-alanine used in the present study. No reactions were observed with L-serine, L-threonine or β-Cl-L-alanine. Neither the D- nor the L-form of alanine, cysteine, tyrosine or tryptophan served as a substrate of DsdSC. The kinetic parameters of DsdSC in the D-serine dehydratase reaction were obtained by the coupling method with lactate dehydrogenase. The apparent Km and kcat values for D-serine were 0.39 mM and 13.1 s−1 respectively (Table 2).
|Untreated||EDTA-treated||EDTA-treated+2.5 μM ZnCl2|
|Substrates||Km (mM)||kcat (s−1)||Km (mM)||kcat (s−1)||Km (mM)||kcat (s−1)|
|Untreated||EDTA-treated||EDTA-treated+2.5 μM ZnCl2|
|Substrates||Km (mM)||kcat (s−1)||Km (mM)||kcat (s−1)||Km (mM)||kcat (s−1)|
Metal requirement for the dehydratase activity
DsdSC was inhibited by EDTA treatment (Table 1). To examine the metal-dependency of the enzyme, we attempted to obtain the metal-free enzyme by the dialysis of DsdSC against 5 mM EDTA for 16 h, followed by dialysis against the buffer without EDTA. The activity towards D-serine of the resultant enzyme was about 5–15% of that of the EDTA-untreated enzyme, and was fully recovered by the addition of 2.5–5.0 μM Zn2+ (Figure 3). Over 10 μM, zinc showed inhibitory effects (Figure 3). No activation of the EDTA-treated enzyme was obtained with monovalent cations, such as K+ or Na+, and divalent cations, such as Mg2+, Mn2+, Ca2+, Ni2+ or Cu2+ (results not shown). We carried out atomic absorption measurements, and found that the purified DsdSC contained 1.02 mol of zinc per mol of the DsdSC monomer (the zinc content is the average value from the duplicated measurements). In contrast, the EDTA-treated enzyme with 5.6% of the kcat value of the untreated enzyme contained 0.07 mol of zinc per mol of the monomer (7% of the zinc content of the untreated enzyme).
Effect of zinc on the D-serine dehydratase activity of EDTA-treated DsdSC
The activity toward D-threonine was also decreased below the detection limit by EDTA-treatment and recovered by the addition of Zn2+ (Table 2). On the other hand, α,β-elimination of β-Cl-D-alanine was slightly increased rather than decreased by EDTA treatment (Table 2). These results suggest that zinc is involved in the elimination of the hydroxy group of D-serine and D-threonine. CD spectra were nearly the same with the EDTA-treated and untreated enzymes (see Supplementary Figure S3 at http://www.biochemj.org/bj/409/bj4090399add.htm), suggesting that the gross conformation was not affected by zinc-binding.
Subunit structure of DsdSC
Gel-permeation chromatography of the purified DsdSC gave one peak at approx. 7.9 min corresponding to the molecular mass of 118 kDa (Figure 4). Because the calculated molecular mass of the DsdSC monomer is 50059.2 Da, DsdSC probably exists as a dimer. The EDTA-treated enzyme and the enzyme from which the N-terminal histidine-tag was removed by thrombin digestion, also gave a similar elution pattern upon gel filtration (results not shown). Zinc or a histidine-tag did not affect the subunit structure of DsdSC.
Gel-permeation chromatography of the purified DsdSC
Effect of EDTA treatment on the thermostability of DsdSC
Thermostability of DsdSC with or without EDTA treatment was compared by measuring the temperature-dependent CD change at 220 nm. As shown in Supplementary Figure S4 (at http://www.biochemj.org/bj/409/bj4090399add.htm), the CD at 220 nm of the EDTA-treated and untreated enzymes started to decrease at approx. 50 and 60 °C respectively. The EDTA treatment decreased the thermostability of DsdSC.
Effect of exogenous D-serine on the growth of the S. cerevisiae Δygl196w strain
Exogenous D-serine has been reported to delay S. cerevisiae cell growth [10,11]. We studied the possibility that DsdSC participates in D-serine detoxification in yeast cells. The growth of S. cerevisiae BY4742 cells and that of BY4742Δygl196w cells were compared in a medium containing D-serine. In the presence of 1 mM D-serine, the Δygl196w strain showed slight growth retardation; in contrast the mother strain showed a similar growth rate to that without D-serine (Figure 5). These results suggest that YGL196W contributes to D-serine detoxification. In the presence of 5 mM D-serine, the mother strain also suffered from growth retardation. The detoxification effect of DsdSC is limited, effectively below 5 mM D-serine.
Effect of D-serine on the cell growth between the mother strain and the Δygl196w mutant of S. cerevisiae BY4742
In the present study we demonstrated that YGL196W of S. cerevisiae encodes a PLP- and zinc-dependent D-serine dehydratase and named it DsdSC. DsdSC is probably the first eukaryotic D-serine dehydratase, and its primary structure has no similarity to that of the bacterial enzyme. PLP-dependent enzymes whose structures have been solved to date belong to one of the five distinct fold types [27,28]. Each enzyme belonging to a different fold-type has completely different folding and has probably evolved from a different ancestor. DsdSC is predicted to have a motif similar to that of the N-terminal domain of the bacterial alanine racemase. If this is the case, DsdSC is classified into the fold-type III group of pyridoxal enzymes. In contrast, the bacterial D-serine dehydratase belongs to the fold-type II group . DsdSC and the bacterial D-serine dehydratase are examples of convergent evolution, which is also exemplified by aspartate aminotransferase and D-amino acid aminotransferase [27,28].
DsdSC shows no overall sequence similarities with the known pyridoxal enzymes; however, we found that DsdSC had some similarity in segments with D-threonine aldolase of Arthrobacter sp. strain DK-38 [24,25] (see Supplementary Figure S1). D-Threonine aldolase is a pyridoxal enzyme and depends on divalent cations. Although the overall sequence similarity between D-threonine aldolase and DsdSC is approx. 10%, both enzymes possess several common amino acid sequence segments. One of them, RPHAK59AHKC of D-threonine aldolase, is conserved in DsdSC as RAHVK57THKT (see Supplementary Figure S1). The Lys59 of D-threonine aldolase is the active-site lysine bound with PLP . The Lys57 of DsdSC is also considered to be the PLP-binding site because DsdSC lost its activity and ability to form a Schiff base with PLP by the mutation of Lys57 to an alanine residue (results not shown). ArthrobacterD-threonine aldolase is predicted to have similar folding to that of the bacterial alanine racemase , suggesting that D-threonine aldolase and DsdSC belong to the same superfamily of PLP enzymes (fold-type III group). However, both enzymes differ from each other in their metal dependency. D-Threonine aldolase completely lost its activity as a consequence of EDTA treatment and recovered it by the addition of divalent cations, such as Mn2+, Mg2+, Co2+, Ni2+, Fe2+ and Ca2+. No activation was observed with Zn2+ [24,25]. In contrast with the D-threonine aldolase, DsdSC was dominantly activated with Zn2+ (Figure 3). To the best of our knowledge, DsdSC is the only PLP enzyme that is activated specifically with Zn2+.
Besides DsdSC and the ArthrobacterD-threonine aldolase, several PLP enzymes have been reported to be activated by metals. Tyrosine phenol-lyase , tryptophanase , dialkylglycine decarboxylase  and E. coliD-serine dehydratase  are activated by monovalent cations, and the eukaryotic serine racemase is activated by divalent cations . Whoel and Dunn  proposed two roles for the monovalent cations in the enzyme catalysis, structural and dynamic roles. The structural role indicates that the metal binding activates the enzyme by stabilizing the catalytically active conformation of the enzyme. The dynamic role means that metals are bound with the protein at the transition state and lower the activation energy. The EDTA-treated DsdSC lost the catalytic activity towards D-serine and D-threonine but retained that towards β-Cl-D-alanine (Table 2). The dehydration of D-serine and D-threonine, and the α,β-elimination of β-Cl-D-alanine are expected to proceed through common steps, α-proton abstraction and elimination of the leaving group on the β-carbon . As a leaving group, chloride is more efficient than a hydroxy group. Chloride can probably be eliminated without the help of zinc from the deprotonated complex of β-Cl-D-alanine and PLP. In contrast, zinc is required for lowering the activation energy for the hydroxy group elimination from the D-serine and D-threonine moiety of the intermediate. The D-serine dehydratase activity of DsdSC was correlated with the zinc content of the enzyme, suggesting that zinc was indispensable for the reaction. However, we cannot rule out the possibility that the absolutely zinc-free enzyme still has a little activity. We are currently investigating the zinc-binding site to construct the absolutely zinc-free enzyme by site-directed mutagenesis. The results in the present study also demonstrated that the EDTA treatment lowered the thermostability of the enzyme. Zinc also plays a role in stabilizing the enzyme structure as well as activating catalysis. X-ray crystallography of the enzyme currently in progress will provide a clue to elucidate the detailed roles of zinc in DsdSC, a unique zinc-dependent pyridoxal enzyme.
DsdSC gave a little contribution to the detoxification of D-serine (Figure 5). D-Amino acid toxicity has been explained by the formation of the D-aminoacyl tRNA and the resultant decrease in the size of the tRNA pool . It is possible that DsdSC prevents the cells from forming D-serine-tRNA by lowering the D-serine concentration in the cells. Analysis of the genetic interaction network of the S. cerevisiae genes ( andhttp://www.yeastgenome.org/) suggests that DsdSC interacts with 1,3-β-D-glucan synthase (Fks1p) and β-glucan synthesis-associated protein (Kre6p), which are related to cell-wall synthesis. In the vegetative cell wall of S. cerevisiae, the serine content reaches approx. 17% of the total amino acids . If the serine residue is in the L-form, it is possible that DsdSC prevents the contamination of D-serine to the S. cerevisiae cell walls. On the other hand, Vorachek-Warren and McCusker  reported that the heterologous expression of the E. coliD-serine dehydratase gene (dsdA) conferred resistance to 19 mM D-serine and the ability to use D-serine as a nitrogen source on S. cerevisiae strain S019 cells. The authors have developed a new method of yeast gene disruption with dsdA as a selection marker. These experiments, as well as the results in the present study (Figure 5), suggest that intrinsic DsdSC is not sufficient to reduce the toxicity of the high concentration of D-serine. Studies of the YGL196W gene expression should be important for understanding the physiological roles of DsdSC and D-serine.
This work was supported in part by the Grant-in-Aid for Scientific Research 19380059 (to T.Y.) from the Ministry of Education, Culture, Sports, Science, and Technology.