Identification, Cloning, and Characterization of l-Phenylserine Dehydrogenase from Pseudomonas syringae NK-15

The gene encoding d-phenylserine dehydrogenase from Pseudomonas syringae NK-15 was identified, and a 9,246-bp nucleotide sequence containing the gene was sequenced. Six ORFs were confirmed in the sequenced region, four of which were predicted to form an operon. A homology search of each ORF predicted that orf3 encoded l-phenylserine dehydrogenase. Hence, orf3 was cloned and overexpressed in Escherichia coli cells and recombinant ORF3 was purified to homogeneity and characterized. The purified ORF3 enzyme showed l-phenylserine dehydrogenase activity. The enzymological properties and primary structure of l-phenylserine dehydrogenase (ORF3) were quite different from those of d-phenylserine dehydrogenase previously reported. l-Phenylserine dehydrogenase catalyzed the NAD+-dependent oxidation of the β-hydroxyl group of l-β-phenylserine. l-Phenylserine and l-threo-(2-thienyl)serine were good substrates for l-phenylserine dehydrogenase. The genes encoding l-phenylserine dehydrogenase and d-phenylserine dehydrogenase, which is induced by phenylserine, are located in a single operon. The reaction products of both enzymatic reactions were 2-aminoacetophenone and CO2.


Identification of the Gene Encoding d-Phenylserine
Dehydrogenase and Gene Organization. Based on the N-terminal amino acid sequence of d-phenylserine dehydrogenase, determined as described previously [16], and the internal amino acid sequence of the enzyme determined in this work, inverse PCR was performed to identify the gene encoding d-phenylserine dehydrogenase. PCR products were sequenced with an Applied Biosystems 373A DNA sequencer and a DNA sequencing kit (ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit). Inverse PCR was also used to determine the nucleotide sequence of the regions upstream and downstream of the d-phenylserine dehydrogenase gene.

Cloning and Expression of the Gene Encoding d-
Phenylserine Dehydrogenase and the Orf3 Gene in Escherichia coli. Chromosomal DNA was prepared from P. syringae NK-15 by the method of Saito and Miura [17]. A DNA fragment containing the gene encoding d-phenylserine dehydrogenase was amplified by PCR with Ex Taq DNA polymerase (Takara Shuzo, Kyoto, Japan) using a sense primer containing an EcoRI site (5 -GCGGAATTCGAA-TCCGCCAACCCACGCCAAGGAATAACGCA -3 ) and an antisense primer containing a PstI site (5 -GCGAAGCTT-CTGCAGCAAGCAGCGCTCACGTCGAAGCGCACA-3 ).
The amplified DNA fragment was ligated into the EcoRI-PstI site of pUC18. The resultant plasmid, pUPsDH, was introduced into E. coli JM109 to provide recombinant dphenylserine dehydrogenase. E. coli JM109 carrying pUPsDH was cultivated in LB medium containing 50 μg/ml ampicillin and 0.1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at 37 • C for 20 hours. A DNA fragment containing the orf3 gene was amplified using a sense primer containing an EcoRI site and the ATG start codon (5 -GGGAATTCA-GGAAACAGACCATGAGTTTTCCGGTTTGTCTCGTCA-3 ) and an antisense primer containing a HindIII site (5 -GGAAGCTTATGTGTTGAGCAGCAGCCCGxTTCTCGA-TCG 3 ). The amplified DNA fragment was ligated into the EcoRI-HindIII site of pSE420D (Daicel Chemical Industries, Osaka, Japan) [18]. The resultant plasmid, pSORF3, was deposited in the International Patent Organism Depositary, National Institute of Advanced Industrial Science and Technology (Ibaraki, Japan) under accession number FERM-P-20287. To obtain recombinant ORF3, E. coli JM109 carrying pSORF3 was cultivated in LB medium containing 50 μg/ml ampicillin and 0.1 mM IPTG at 37 • C for 16 hours.
2.6. Purification of the orf3 Gene Product. The standard buffer used throughout purification was 10 mM potassium phosphate buffer (pH 8.0), and all operations were done at 4 • C. Cultured E. coli cells expressing ORF3 were harvested by centrifugation, resuspended in 0.1 M potassium phosphate buffer (pH 8.0) containing 0.02% 2-mercaptoethanol (2-ME) and 2 mM phenylmethylsulfonyl fluoride (PMSF), and disrupted using a Micro Smash MS-100 (TOMY, Tokyo, Japan). After centrifugation, the supernatant was fractionated by ammonium sulfate precipitation (0-50% saturation). The enzyme-containing fraction was resuspended in 0.1 M potassium phosphate buffer (pH 8.0) containing 0.02% 2-ME and 2 mM PMSF, and dialyzed against the same buffer. The enzyme fraction was applied to a Q-Sepharose FF column (Pharmacia, Columbus, OH) equilibrated with the standard buffer containing 0.01% 2-ME. The enzyme was eluted with a linear gradient of 0-0.5 M NaCl in the same buffer. The enzyme fractions were collected, concentrated, dialyzed against the standard buffer containing 0.01% 2-ME and 20% saturated ammonium sulfate, and centrifuged. The supernatant was applied to a Phenyl superose HP 26/10 column (Pharmacia, Columbus, OH) equilibrated with the standard buffer containing 0.01% 2-ME and 30% saturated ammonium sulfate. The enzyme was eluted with a linear gradient of 20-0% saturated ammonium sulfate in the buffer. The enzyme fractions were collected, concentrated and dialyzed against the standard buffer containing 0.01% 2-ME. The final preparation of the enzyme was stored at −80 • C until use.

Enzyme
Assay. l-Phenylserine dehydrogenase activity was assayed by monitoring the increase in absorbance at 340 nm due to the production of NADH at 30 • C in a 1-ml reaction mixture containing 20 mM dl-threo-βphenylserine and 2.5 mM NAD + in 0.2 M Glycine-KCl-KOH buffer (pH 10.5). d-Phenylserine dehydrogenase activity was determined as previously described [16].
2.9. Analytical Methods for Enzyme. Protein concentration was determined using a Protein assay kit (Bio-Rad, Hercules, CA) with bovine serum albumin as standard. The molecular mass of the subunit of l-phenylserine dehydrogenase was examined by SDS-PAGE using Protein Markers for SDS-PAGE (Nacalai Tesque, Kyoto, Japan). The molecular mass of native l-phenylserine dehydrogenase was estimated by HPLC on a TSK-GEL G3000SW column (0.75×60 cm) operating at room temperature. The column was eluted with 0.1 M potassium phosphate buffer (pH 7.0) containing 0.2 M NaCl at a flow rate of 0.7 ml/min. Amino acid sequences were obtained from PubMed at NCBI (http://www.ncbi.nlm.nih.gov/). A homology search was performed using the BLAST program [19] at GenomeNet (http://www.genome.ad.jp/). Multiple alignments were obtained with the ClustalW program [20] at GenomeNet (http://www.genome.ad.jp/).

Nucleotide Sequence Accession Number.
The nucleotide sequence data have been deposited in the DDBJ/EMBL/ GenBank nucleotide sequence databases under accession number AB499092.

Identification of a Gene Encoding d-Phenylserine
Dehydrogenase. Purified d-phenylserine dehydrogenase was obtained as previously described [16]. The enzyme was digested with lysyl endopeptidase, and the peptide products were purified by reversed-phase HPLC. The amino acid sequences of only two internal peptides could be determined ( Figure 1). Based on the N-terminal amino-acid sequence and the internal amino acid sequences determined, an 897bp nucleotide sequence was identified as the gene encoding d-phenylserine dehydrogenase ( Figure 1). A crude extract of E. coli JM109 transformed with the pUPsDH expression vector containing the gene showed d-phenylserine dehydrogenase activity (33 U/mg), while that of wild-type E. coli JM109 was inactive.  of the gene encoding d-phenylserine dehydrogenase, inverse PCR was carried out. As a result, a 9,246-bp nucleotide sequence containing at least six open reading frames (ORFs) was determined (Figure 2). The transcriptional directions of orf1 and orf6 are opposite to those of the four other ORFs. Postulated promoter and terminator sequences are located immediately upstream of orf2 and downstream of d-phenylserine dehydrogenase encoding orf5, respectively. These observations suggest that orf2, orf3, orf4, and orf5 may form an operon. orf1 encodes a protein of 320 amino acids that is similar to amino acid sequences of putative LysR-type transcriptional regulators. Thus, orf1 probably plays a role in the regulation of transcription of the operon. orf2 encodes a protein of 436 amino acids that shows sequence similarity to putative major facilitator superfamily (MFS) transporters. orf4 encodes a protein of 579 amino acids that is similar to amino acid sequences of putative dihydroxy acid dehydratases (ilvD). The ilvD gene has previously been identified in the ilv operon involved in branched-chain amino acids biosynthesis [21][22][23]; however, the operon containing the gene for d-phenylserine dehydrogenase did not contain other genes related to branched-chain amino acids metabolism. orf5 encodes d-phenylserine dehydrogenase, which was previously characterized [16]. orf6 encoded a protein of 520 amino acids that showed high similarity with amino acid sequences of putative ABC peptide transporters. orf3 encodes a protein of 259 amino acids that shares 37% identity with ketoreductase (RED2) from Streptomyces violaceoruber Tü22 [24] and 28% identity with 1,3,8trihydroxynaphthalene reductase (3HNR) from Magnaporthe grisea (Figure 4) [25,26]. The amino acid sequence of ORF3 also shows high similarity to that of putative short chain dehydrogenases and putative 3-oxoacyl-(acylcarrier protein) reductases and 24% identity with serine dehydrogenase from Agrobacterium tumefaciens ICR 1600 [27]. A common GXXXGXG sequence, which is characteristic of an NAD(P)-binding site conserved in serine dehydrogenase and its homologs [27], was found in the Nterminal region of ORF3. For these reasons, we assumed that ORF3 has dehydrogenase activity, and considered that 3-hydroxy amino acids were likely to serve as a substrate for the enzyme, so cloning of orf3 was done.

Purification of l-Phenylserine Dehydrogenase (ORF3).
ORF3 was purified to homogeneity from the recombinant E. coli JM109 cell carrying pSORF3. ORF3 has a calculated molecular mass of 27498.3 Da. The purified protein gave a single band with a molecular mass of 27 kDa on SDS-PAGE. The molecular mass of the native protein was determined to be 98 kDa by gel filtration. Because the elution of ORF3 was likely slightly slowed by nonspecific hydrophobic and ionic interactions between ORF3 and the gel filtration resin, the apparent molecular mass of the protein was most likely an underestimate. Therefore, ORF3 probably consists of four identical subunits. A summary of the specific activity and recovery of ORF3 during purification is shown in Table 1.

Properties of l-Phenylserine Dehydrogenase (ORF3).
The molecular characteristics of the enzyme are shown in Tables 2, 3, and 4. The enzyme was significantly inhibited by 0.05 mM p-chloromercuribenzoate and 0.01 mM HgCl 2 . However, thiol reagents, such as N-ethylmaleimide and iodoacetamide, the chelating agent EDTA, and bivalent metal cations did not affect the enzyme ( Table 2). The enzyme acted in an NAD + -dependent way on dlthreo-β-phenylserine but not on d-threo-β-phenylserine. Because we could not obtain pure l-threo-β-phenylserine, we were unable to perform enzyme assays with l-threo-βphenylserine as a substrate. However, the data we obtained indicate that the enzyme showed activity towards only the lform. The enzyme also acted on dl-erythro-β-phenylserine and dl-threo-(2-thienyl)serine. Pure l-forms of these compounds are also unavailable, but the enzyme likely acted on only the l-forms of erythro-β-phenylserine and threo-(2-thienyl)serine. Other amino acids tested did not serve as a substrate. The enzyme showed weak activity toward (S)-phenylethanol (Table 3). TLC analysis revealed that the enzyme converted l-β-phenylserine (R f = 0.52) into 2aminoacetophenone (R f = 0.63). Therefore, we considered that the enzyme catalyzed the oxidation of the β-hydroxyl group of l-β-phenylserine and that the reaction product, lα-amino-β-keto-γ-phenylpropionate, spontaneously decarboxylated to form 2-aminoacetophenone (Scheme 2). The enzyme preferred NAD + to NADP + as a coenzyme. The enzyme showed maximal activity at pH 11.2 and was stable between pH 6.1 and 11.2 at 30 • C. The enzyme was stable at temperatures lower than 55 • C for at least 10 minutes and showed the highest activity at 40 • C ( Table 4). The apparent K m values for dl-threo-β-phenylserine and NAD + were 59 and 2.1 mM, respectively.

Discussion
The enzymological properties of d-phenylserine dehydrogenase have already been reported [16], but the nucleotide sequence of the gene encoding d-phenylserine dehydrogenase was determined in this work. The amino acid sequence of d-phenylserine dehydrogenase shares 24% identity with 3hydroxyisobutyrate dehydrogenase (TTHA0237) from Thermus thermophilus HB8 [28] and 24% identity with a possible 3-hydroxyisobutyrate dehydrogenase (PA0743) from Pseudomonas aeruginosa PAO1. An alignment of the amino acid sequences of d-phenylserine dehydrogenase, TTHA0237, and PA0743 is shown in Figure 3. Many NAD/NADPdependent dehydrogenases contain the Rossmann fold for nucleotide binding; the pyrophosphate group interacts with the GXGXX(G/A) motif found in the Rossmann fold [29]. This characteristic glycine-rich fingerprint motif was highly conserved in the N-termini of d-phenylserine dehydrogenase, TTHA0237, and PA0743. Similarly, alignment of the amino acid sequence of d-phenylserine dehydrogenase with the sequences of 6-phosphogluconate dehydrogenase from Ovis aries [30], Saccharomyces cerevisiae [31], Lactococcus    lactis [32], and Trypanosoma brucei [33] showed that the GX(G/A)XXG motif and residues interacting with 2phosphate group of NADP + (Gln36 and Arg37) were highly conserved among these enzymes. d-Phenylserine dehydrogenase and these 6-phosphogluconate dehydrogenases prefer NADP + to NAD + as a coenzyme. Moreover, a catalytic residue, Lys177, was also conserved in d-phenylserine dehydrogenase, TTHA0237, and PA0743. The molecular characteristics of l-phenylserine dehydrogenase and d-phenylserine dehydrogenase are summarized in Table 4. The amino acid sequences of these enzymes showed no homology to each other and each enzyme belongs to a different protein family. The amino acid sequence of lphenylserine dehydrogenase was similar to those of ketoreductase (RED2) from Streptomyces violaceoruber Tü22 [24] and 1,3,8-trihydroxynaphthalene reductase (3HNR) from Magnaporthe grisea [25,26]. The amino acid sequences of l-phenylserine dehydrogenase and two homologs belonging to the short chain dehydrogenase/reductase (SDR) family Table 3: Substrate specificity of l-phenylserine dehydrogenase. aligned well (Figure 4). Members of the SDR family contain a similar structural fold, which shows a common nucleotidebinding site characterized by a GXXXGXG fingerprint motif [25]. Moreover, Arg or Asp residues located 18-20 residues downstream from the motif are responsible for nucleotide specificity [25,34]. The characteristic glycinerich fingerprint motif was conserved in the N-terminus of l-phenylserine dehydrogenase. Acidic residues, Asp36 or Asp37, which are 20 and 21 residues downstream, respectively, from the motif probably recognize the 2 -hydroxy group of NAD + . Our kinetic analysis also indicated that lphenylserine dehydrogenase prefers NAD + to NADP + as the coenzyme. An X-ray structure of 3HNR complexed with NADPH and tricyclazole revealed that Ser164, Tyr178, and Lys182 compose the catalytic triad [25]. These residues were highly conserved in l-phenylserine dehydrogenase, RED2, and 3HNR (corresponding to Ser148, Tyr162, and Lys166 in l-phenylserine dehydrogenase) (Figure 4). Although threonine [14,[35][36][37], serine [38,39], and phenylalanine [40] serve as substrates for many enzymes acting on phenylserine, these amino acids were not accepted as substrates by l-phenylserine dehydrogenase. Among the 8 Enzyme Research  amino acids tested, l-phenylserine and l-threo-(2-thienyl)serine were good substrates for l-phenylserine dehydrogenase. The genes encoding l-phenylserine dehydrogenase (ORF3) and d-phenylserine dehydrogenase (ORF5) were located within a single operon, and the reaction product of both enzymes is 2-aminoacetophenone. Moreover, dphenylserine dehydrogenase is induced by addition of dlthreo-β-phenylserine to a culture medium as a sole source of carbon and nitrogen [16]. Therefore, we consider that d-phenylserine dehydrogenase acts physiologically on dthreo-β-phenylserine. For these reasons, we assume that the physiological function of l-phenylserine dehydrogenase is an NAD + -dependent conversion of l-phenylserine into 2aminoacetophenone and carbon dioxide.