Characterization of thermostable aminoacylase from hyperthermophilic archaeon Pyrococcus horikoshii Koichi Tanimoto 1 , Noriko Higashi 1 , Motomu Nishioka 1 , Kazuhiko Ishikawa 2 and Masahito Taya 1 1 Division of Chemical Engineering, Graduate School of Engineering Science, Osaka University, Japan 2 Research Institute for Cell Engineering, National Institute of Advanced Industrial Science and Technology (AIST), Osaka, Japan Today, many kinds of enzyme are used to manufac- ture commercial products in various fields, such as the food, chemical and pharmaceutical industries. Compared with their chemically conducted counter- parts, enzymatic reactions have several advantages in selectivity and specificity, as well as environmental load and operational safety [1]. However, enzymatic reactions need to be practiced under mild conditions of temperature and pH, because of instability of the enzymes. In the past two decades, thermostable enzymes found in many hyper- or extreme thermo- philes have been studied intensively in an attempt to overcome the limitation in application of enzymatic reactions. Previous studies have demonstrated that most hyperthermophile-originating enzymes show high stability to extreme pH conditions and organic solvents, in addition to durability against high temperature [2]. In general, reactions performed at relatively high temperature have certain advantages, such as extended availability of less water-soluble substrates, decreased viscosity of reaction solutions and increased diffusibility of substrates or products. Keywords aminoacylase; hyperthermophilic archaeon; metal ligand; Pyrococcus horikoshii; substrate specificity Correspondence M. Taya, Division of Chemical Engineering, Graduate School of Engineering Science, Osaka University, 1–3 Machikaneyama-cho, Toyonaka, Osaka 560-8531, Japan Fax: +81 6 6850 6254 Tel: +81 6 6850 6251 E-mail: taya@cheng.es.osaka-u.ac.jp (Received 6 November 2007, revised 5 January 2008, accepted 7 January 2008) doi:10.1111/j.1742-4658.2008.06274.x The gene encoding putative aminoacylase (ORF: PH0722) in the genome sequence of a hyperthermophilic archaeon, Pyrococcus horikoshii, was cloned and overexpressed in Escherichia coli. The recombinant enzyme was determined to be thermostable aminoacylase (PhoACY), forming a homotetramer. Purified PhoACY showed the ability to release amino acid molecules from the substrates N-acetyl-l-Met, N-acetyl-l-Gln and N-acetyl-l-Leu, but had a lower hydrolytic activity towards N-acetyl-l-Phe. The kinetic parameters K m and k cat were determined to be 24.6 mm and 370 s )1 , respectively, for N-acetyl-l-Met at 90 °C. Purified PhoACY con- tained one zinc atom per subunit. EDTA treatment resulted in the loss of PhoACY activity. Enzyme activity was fully recovered by the addition of divalent metal ions (Zn 2+ ,Mn 2+ and Ni 2+ ), and Mn 2+ addition caused an alteration in substrate specificity. Site-directed mutagenesis analysis and structural modeling of PhoACY, based on Arabidopsis thaliana indole-3- acetic acid amino acid hydrolase as a template, revealed that, amongst the amino acid residues conserved in PhoACY, His106, Glu139, Glu140 and His164 were related to the metal-binding sites critical for the expression of enzyme activity. Other residues, His198 and Arg260, were also found to be involved in the catalytic reaction, suggesting that PhoACY obeys a similar reaction mechanism to that proposed for mammalian aminoacylases. Abbreviations AcMet, N-acetyl- L-methionine; AcPhe, N-acetyl-L-phenylalanine; AfaACY, Alcaligenes faecalis strain DA1 D-aminoacylase; AthIAAH, Arabidopsis thaliana indole-3-acetic acid amino acid hydrolase; BsbAH, Bacillus subtilis putative amidohydrolase; BstACY, Bacillus stearothermophilus aminoacylase; hACY1, human aminoacylase-1; ICP-OES, inductively coupled plasma-optical emission spectrometry; pACY1, porcine aminoacylase-1; PfuACY, Pyrococcus furiosus aminoacylase; PhoACY, Pyrococcus horikoshii aminoacylase; PhoCP, Pyrococcus horikoshii carboxypeptidase; PseCP, Pseudomonas sp. strain RS-16 carboxypeptidase; PSSM, position-specific scoring matrix; SsoCP, Sulfolobus solfataricus carboxypeptidase; TkoACY, Thermococcus kodakaraensis aminoacylase; TliACY, Thermococcus litralis aminoacylase. 1140 FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS Aminoacylase (EC 3.5.1.14) is one of the most important enzymes for industrial applications and lib- erates enantioselectively l-amino acid from a corre- sponding N-acyl-amino acid racemate [3]. Methionine, alanine and phenylalanine are produced using amino- acylase on a commercial scale [4–7]. Aminoacylase is an essential enzyme found in almost all organisms, and many of these enzymes are classified into the M20 family of metallopeptidases [4,8–11]. In hyperthermo- philes, several enzymes showing aminoacylase activity have been identified [5,12,13], but, to our knowledge, few reports have dealt with the amino acid residues involved in metal binding. Information on the struc- ture and catalytic mechanism of thermostable amino- acylase will help in the development of valuable enzymes and their modified variants for industrial utili- zation. The work reported here describes the successful expression of active aminoacylase from a hyperthermo- philic archaeon, Pyrococcus horikoshii [14], in Escheri- chia coli to characterize the enzymatic properties in terms of substrate specificity, stability and kinetics. In addition, the amino acid residues for metal binding were identified by site-directed mutagenesis, and the metal-binding domain and catalytic mechanism of ami- noacylase are discussed on the basis of a constructed structural model. Results Identification of recombinant aminoacylase from P. horikoshii Amongst thermophilic microorganisms, several ther- mostable enzymes exhibiting aminoacylase activity have been identified and characterized [5,7,12,13]. A hyperthermophilic archaeon, P. horikoshii, has been reported to produce a bifunctional carboxypepti- dase ⁄ aminoacylase (PhoCP ⁄ ACY), whereas Thermo- coccus litralis and Pyrococcus furiosus have been demonstrated to produce an aminoacylase [5,12]. In the genome database of P. horikoshii (DDBJ, http:// www.ddbj.nig.ac.jp/, as of June 2007), an alternative candidate for the aminoacylase gene (PH0722, Uni- ProtKB O58453) has been annotated as a putative amino acid amidohydrolase, as suggested in a previous report [12]. Figure 1 shows the alignment of the amino acid sequence of the gene (PH0722) from P. horikoshii, together with those of P. furiosus aminoacylase (PfuACY, 82% homology) [12], Thermococcus kodaka- raensis aminoacylase ( TkoACY, 72% homology) (UniProtKB Q5JD73), P. horikoshii carboxypepti- dase ⁄ aminoacylase (PhoCP ⁄ ACY, 58% homology) [13], Thermococcus litralis aminoacylase (TliACY, 55% homology) [5], Sulfolobus solfataricus carboxypeptidase (SsoCP, 40% homology) [15], Bacillus stearothermophi- lus aminoacylase (BstACY, 34% homology) [7], Bacil- lus subtilis putative amidohydrolase (BsbAH, 39% homology) (UniProtKB P54955) and Arabidopsis thaliana indole-3-acetic acid amino acid hydrolase (AthIAAH, 45% homology) (UniProtKB P54970). These enzymes are classified into the M20 family of metallopeptidases, in which zinc is included as an essential metal for the catalytic activity [16,17]. The alignments clearly demonstrate that the sequences of these enzymes share highly conserved regions, espe- cially in the vicinity of Cys104, His106, Glu139, Glu140, His164 and His361 (indicated by the positions in the PhoACY sequence), which have been reported to be the amino acid residues related to metal binding from mutational analysis [18] and structural informa- tion from BsbAH [Protein Data Bank (PDB) ID: 1YSJ]. With regard to functionally related enzymes with available structural information, the carboxypep- tidase from Pseudomonas sp. strain RS-16 (PseCP, PDB ID: 1CG2) and the d-aminoacylase from Alcali- genes faecalis strain DA1 ( AfaACY, PDB ID: 1V4Y) exhibit 19% and 15% homology, respectively, to PhoACY. In this study, the gene (PH0722) was overexpressed in E. coli cells using the pET-11a expression system to confirm whether it encodes a protein with enzymatic function. The recombinant protein was successfully produced in a soluble form and the purified enzyme showed a molecular mass of 42 kDa as a single poly- peptide on SDS-PAGE, in accordance with the molec- ular mass estimated from the amino acid sequence. Amongst the aminoacylases and carboxypeptidases that belong to the M20 family of metallopeptidases, most enzymes are known to exist in a dimeric form [9,19–22]. HPLC gel filtration demonstrated that the molecular mass of PhoACY under native conditions was approximately 165 kDa, indicating that PhoACY forms a homotetrameric structure. Catalytic properties of PhoACY The enzyme encoded by the cloned gene was expected to be a thermostable aminoacylase (PhoACY), and the activity of recombinant PhoACY at various pH values and temperatures was measured using N-acetyl- l-methionine (AcMet) as a substrate. The pH depen- dence of the activity was examined in the range pH 6.0–8.5, and the maximum activity was observed at pH 7.5 (Fig. 2A). The effect of temperature on the K. Tanimoto et al. Characterization of thermostable aminoacylase FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 1141 activity was also determined in the range 60–100 °Cat pH 7.5, and the optimum activity was observed in the vicinity of 90 °C. The activity at 60 °C fell to less than one-half of the activity at the optimum condition (Fig. 2B). Figure 3 shows the thermostability of PhoACY at 75, 85 and 90 °C for the indicated time periods at pH 7.5. From the slopes of the lines, the half-lives of the enzyme were estimated to be 217, 119 and 72 min at 75, 85 and 90 °C, respectively. It should be noted that no substantial loss in activity was observed during the examined period when zinc ions (0.1 mm) were present in the enzyme solution. The substrate specificity of PhoACY was examined with respect to various N-acetyl-l-amino acid com- pounds. As shown in Table 1, the order of favorite substrates was Met > Leu > Gln > Glu > Arg > Ala > Asn > Phe = Gly, indicating that PhoACY has a tendency to exhibit a high activity for amino acids with long side chains. PfuACY and PhoCP ⁄ ACY also show hydrolase activity against N-acetyl-l-amino acid compounds in the order of preferred l-amino acid moieties of Met > Ala = Asn > Glu > Leu for Pfu- ACY [12] and Met > Phe > Ala > Trp > Gly for PhoCP ⁄ ACY [13], which is rather different from that of PhoACY. The dependence of the reaction rate on substrate concentration was examined for AcMet in the range 1–100 mm, and the kinetic parameters were estimated according to Michaelis–Menten kinetics (Fig. 4). The value of the Michaelis constant for Fig. 1. Alignment of amino acid sequences of P. horikoshii aminoacylase with aminoacylases from various sources and carboxypeptidases from Sulfolobus solfataricus and P . horikoshii, amidohydrolase from Bacillus subtilis and indole-3-acetic acid hydrolase from A. thaliana. Iden- tical residues are indicated by asterisks, and the residues with strong and weak similarity are shown by colons and full points, respectively. The residues mutated in this work are indicated by filled circles. The abbreviations are as follows and the entry names of the protein data- base are given in parentheses: PhoACY, P. horikoshii aminoacylase (PH0722) (this study, UniProtKB O58453); PfuACY, P. furiosus amino- acylase (UniProtKB Q8U375); TkoACY, Thermococcus kodakaraensis aminoacylase (UniProtKB Q5JD73); PhoCP ⁄ ACY, P. horikoshii carboxypeptidase ⁄ aminoacylase (UniProtKB O58754); TliACY, Thermococcus litoralis aminoacylase [5]; SsoCP, S. solfataricus carboxypepti- dase (UniProtKB P80092); BstACY, Bacillus stearothermophilus aminoacylase (UniProtKB P37112); BsbAH, B. subtilis putative amidohydro- lase (UniProtKB P54955); AthIAAH, A. thaliana indole-3-acetic acid amino acid hydrolase (UniProtKB P54970). Characterization of thermostable aminoacylase K. Tanimoto et al. 1142 FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS AcMet was high (K m = 24.6 mm) compared with that of PhoCP ⁄ ACY, and the turnover number of PhoACY (k cat = 370 s )1 ) was 10 times larger than that of PhoCP ⁄ ACY, indicating that these two enzymes have different kinetic characteristics as well as substrate specificity. Substrate inhibition for PhoACY was observed at an AcMet concentration above 40 mm,as often reported for other aminoacylases [11,13]. Some aminoacylases are capable of cleaving dipep- tide or tripeptide. TliACY can cleave the dipeptide N-benzyloxycarbonyl-Phe-Gly, and PhoCP ⁄ ACY can release the carboxyl-terminus amino acid residue from di-, tri- and tetrapeptides, whereas PfuACY is not able to hydrolyze N-formyl-Met-Phe, N-acetyl-Met-Ala and N-acetyl-Met-Leu-Phe [12]. PhoACY can cleave the examined dipeptides Ala-Ala and Phe-Ala, but cannot release any amino acid monomer from the tripeptides Ala-Ala-Ala and N-acetyl-Ala-Ala-Ala (data not shown). Effect of metal ions on PhoACY activity and amino acid residues related to metal binding and catalysis Inductively coupled plasma-optical emission spectro- metry (ICP-OES) demonstrated that PhoACY contained 6 6.5 7 7.5 8 8.5 0 20 40 60 80 100 A B 60 70 80 90 100 20 40 60 80 100 pH Relative enzyme activityRelative enzyme activity pH: 7.5 Temperature: 90 °C Temperature (°C) Fig. 2. Effects of pH (A) and temperature (B) on the activity of PhoACY. Incubation time (min) Residual activity pH = 7.5 0 30 60 90 120 150 180 0.2 0.4 0.6 0.8 1 Fig. 3. Thermostability of PhoACY. The enzyme was incubated at 75 ( ), 85 ( ) and 90 °C(d) in the absence of Zn 2+ and 90 °Cin the presence of 0.1 m M Zn 2+ (s). Table 1. Substrate specificity of PhoACY. The substrate concentra- tion was 30 m M. Mn-PhoACY was prepared by EDTA treatment and subsequent addition of 1 m M Mn 2+ . Substrate Specific activity (UÆmg )1 ) Untreated PhoACY Mn-PhoACY N-acetyl- L-Met 256 239 N-acetyl- L-Leu 133 134 N-acetyl- L-Gln 131 208 N-acetyl- L-Glu 92 56 N-acetyl- L-Arg 61 116 N-acetyl- L-Ala 38 47 N-acetyl- L-Asn 18 68 N-acetyl- L-Phe 13 69 N-acetyl- L-Gly 13 19 10 0 0.1 0.2 0.3 0.01 0.02 0.03 0.09 0.1 1/v (min·nmol –1 ) 1/s (mM –1 ) Fig. 4. Double reciprocal plots of enzyme reaction rate (v) versus AcMet concentration (s) for untreated PhoACY (d) and Mn-PhoACY ( ). Mn-PhoACY was prepared by EDTA treatment and subsequent addition of 1 m M Mn 2+ . K. Tanimoto et al. Characterization of thermostable aminoacylase FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 1143 one zinc atom per subunit. The effect of metal chela- tion on PhoACY activity was examined. The dialyzing treatment of PhoACY with EDTA ⁄ Tris ⁄ HCl buffer resulted in the loss of activity, as shown in Table 2. The activity was fully restored by the addition of 10 mm of Zn 2+ ,Mn 2+ or Ni 2+ to the reaction mixture. The presence of Cu 2+ or Co 2+ was less effective on the recovery of the activity, and Mg 2+ exerted no effect on the recovery. On addition of Mn 2+ , the substrate specificity of EDTA-treated Pho- ACY was altered, the order of favorite substrates being Met > Gln > Leu > Arg > Phe = Asn > Glu > Ala > Gly (Table 1). Moreover, under this condition, the values of the Michaelis constant and turnover number for AcMet were K m = 15.0 mm and k cat = 381 s )1 , respectively (Fig. 4), indicating that Mn 2+ addition causes a change in affinity of PhoACY to the substrate. In zinc-containing SsoCP, with similarity to Pho- ACY (40% homology), site-directed mutagenesis revealed that residues His108, Asp109 and His168 were zinc-binding ligands, and Glu142 interacted with a water molecule as an acidic residue [18]. The corre- sponding amino acid residues in PhoACY are His106, Asp107, Glu139, Glu140 and His164. For Glu142 in SsoCP, there are two possible corresponding amino acid residues, Glu139 and Glu140, in PhoACY by comparison of their amino acid sequences. As shown in Table 3, the aminoacylase activities of mutants H106A, E139Q, E140Q and H164A were decreased below the limit of detection, and the zinc atom content was approximately 1.5–2.0 times that of wild-type Pho- ACY. By contrast, mutant D107N retained approxi- mately 20% of the activity of wild-type PhoACY, and the zinc content was equivalent to that in the wild-type enzyme. Some aminoacylases share other highly conserved amino acid residues in addition to those for the zinc ligands. Of these, we examined the contribution of His198 and Arg260 to the enzyme activity of PhoACY. The residues His206 in human aminoacylase-1 (hACY1) and His205 in porcine aminoacylase-1 (pACY1) are involved in enzyme activity through interaction with an active site of the counter-subunit in dimer formation [16,17,23]. When His198, a corre- sponding residue in PhoACY, was substituted for ala- nine, the activity of mutant H198A was completely lost (Table 3). Arg260 in PhoACY is also a conserved residue that is expected to be associated with substrate binding to an active site through conformational change, as proposed in hACY1 [23]. Mutant R260A of PhoACY lost its enzyme activity (Table 3), as did the mutant of hACY1. Discussion This work reports a novel thermostable aminoacylase, PhoACY, identified in the hyperthermophilic archaeon P. horikoshii. Recombinant PhoACY was obtained as a soluble and active form. Substrate specificity is an interesting property of aminoacylases, and has been characterized extensively in several aminoacylases, including those from thermo- philic microorganisms. Aminoacylases from different sources show a wide variety of substrate specificity for amino acid moieties and acyl groups [5,12,13,24]. With regard to the aminoacylases from hyperthermophiles, the substrate specificities of PfuACY, TliACY and PhoCP ⁄ ACY have been examined in detail. As in the case of PhoACY, these three enzymes tend to prefer AcMet, but the preference for other N-acetyl-l-amino acids is different in spite of the high homology of their amino acid sequences. For instance, N-acetyl-l-Phe (AcPhe) is the most favorable substrate with an acetyl group for TliACY, but PhoCP ⁄ ACY, which shows Table 2. Effect of metal ions on the activity of PhoACY treated with EDTA. The concentration of metal ion added was 1 m M. ND, not detectable. Metal ion Relative enzyme activity (%) Without treatment 100 (103 a ) With treatment No addition ND Mg 2+ 9 Cu 2+ 9 Co 2+ 34 Zn 2+ 91 Mn 2+ 154 Ni 2+ 221 a The activity was measured in the reaction mixture with the addi- tion of 1 m M Zn 2+ . Table 3. Relative enzyme activities and zinc contents of various mutants. ND, not detectable. Enzyme Relative enzyme activity (%) Zn 2+ content Predicted role in PhoACY Wild-type 100 1.0 Zn 2+ ligand H106A ND 1.3 Zn 2+ ligand D107N 19 1.0 Zn 2+ ligand E139Q ND 2.3 Zn 2+ ligand E140Q ND 1.5 Zn 2+ ligand H164A ND 1.9 Zn 2+ ligand H198A ND – Catalysis R260A ND – Catalysis Characterization of thermostable aminoacylase K. Tanimoto et al. 1144 FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 82% homology to TliACY, hydrolyzes AcPhe with half of the activity obtained for AcMet. As shown in Table 1, PhoACY shows a much lower activity for AcPhe, which cannot be hydrolyzed by PfuACY. In addition, some aminoacylases show an ability to cleave dipeptides. PhoCP ⁄ ACY is a dipeptidase with a higher specificity for the C-terminal carboxyl group and a lower specificity for the acyl group. PhoACY seems to be primarily an aminoacylase, but exhibits peptidase activity towards a dipeptide substrate, whereas TliACY and PfuACY show no peptidase activity [5,12]. These observations suggest that amino acid residues close to substrate binding and catalytic sites make a great contribution to substrate specific- ity, and that modification of amino acid residues in these local sites may be useful to improve enzymatic properties. PhoACY belongs to the M20 family of metallopep- tidases which contain one or two divalent metal ions, usually Zn 2+ , in their subunits. The role of zinc ions is still uncertain, although treatment with metal chelating reagents results in the complete loss of activity of most aminoacylases. It is known that TliACY activity is inhibited by approximately 50% by treatment with EDTA, but that metal-free TliACY still shows some activity [5]. PhoACY treated with EDTA shows no activity, but the activity can be entirely recovered in the presence of Zn 2+ ,Mn 2+ ,Ni 2+ and, to a lesser extent, Co 2+ . Although the activity of PfuACY treated with EDTA can be restored by the addition of Zn 2+ or Co 2+ ,Mn 2+ and Ni 2+ are inert additives, suggest- ing that the coordination of the metal-binding sites in these two enzymes may be different in spite of their great similarity. Although zinc involvement in the PhoACY molecule is essential for its expression of aminoacylase activity, the external addition of Zn 2+ to the reaction mixture (1 mm) does not lead to elevated activity (Table 2). In the presence of excess Zn 2+ , however, the thermosta- bility of PhoACY is strongly enhanced (Fig. 3), sug- gesting that additional Zn 2+ binding to PhoACY, which may occur in a second metal-binding site, induces a possible alteration in enzyme conformation to improve the thermostability. Future research will focus on the effect of external metal ions on the ther- mal durability of PhoACY from an enzymological and biotechnological aspect. PhoACY whose Zn 2+ was exchanged for Mn 2+ (Mn-PhoACY) showed an altered substrate specificity. A significant increase in activity for N-acetyl-l-Asn and AcPhe substrates was observed for Mn-PhoACY, whereas the activity for N-acetyl-l-Glu was reduced by half, compared with untreated PhoACY, indicating that the catalytic metal ions associated with the enzyme have an effect on the affinity between enzyme and substrate. This is also supported by the increase in the K m value for AcMet in Mn-PhoACY despite a small change in the k cat value. These findings suggest that the substrate recognition of l-aminoacylases, including PhoACY, is largely influenced by local dis- tortion in the structures of the enzymes. Further detailed examinations are needed to understand the mechanisms which can dominate substrate specificities in l-aminoacylases. With regard to the amino acid residues related to metal binding in SsoCP, His108, Asp109, Glu142 and His168 have been identified as Zn 2+ ligands on the basis of site-directed mutagenesis and computational modeling [18]. In the case of PhoACY with mutation in the corresponding residues, mutants H106A, E139Q, E140Q and H164A lost their activities and mutant D107N retained 20% of the activity of wild- type PhoACY, whereas the zinc content of all mutants did not decrease. In order to determine whether these amino acid residues are close to a putative metal-binding site, a molecular model of PhoACY was predicted using the 3d-jigsaw program. The model showed that the overall structure (Fig. 5A) was very similar to the structure of BsbAH (PDB ID: 1YSJ), which has 39% homology to PhoACY and two atoms of nickel in one subunit, and that the spa- tial configuration of the amino acid residues related to zinc binding (Fig. 5B) also resembled that of BsbAH, pACY1 [17] and the zinc-binding domain of the T347G mutant from hACY1 (PDB ID: 1Q7L) [23], although His164 was replaced by a glutamic acid residue in pACY1 and hACY1. In this model, the residues His106, Glu139, Glu140 and His164 were located close to each other and formed a metal-bind- ing pocket, as found in the other aminoacylases, lead- ing to the conclusion that these four residues are zinc ligands. It is expected that, in PhoACY, the mutation of one zinc ligand allows the local loosening of Zn 2+ and the intervention of another Zn 2+ in the metal- binding pocket, resulting in the incorrect coordination of Zn 2+ and a lack of aminoacylase activity. With regard to the mutant D107N, the model showed that Asp107 was not located in the metal-binding site. Therefore, it is probable that Asp107 is not a zinc ligand in PhoACY. The molecular model of PhoACY and the three- dimensional structural information of BsbAH also predict a role of the cysteine residue at position 104, which is one of the most conserved residues in vari- ous aminoacylases. It has been reported that this cys- teine residue is one of the key amino acids for K. Tanimoto et al. Characterization of thermostable aminoacylase FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 1145 aminoacylase activity by site-directed mutagenesis and chemical modification analyses, although its role in activity is still unclear [13]. In the structure of BsbAH, the distance between the cysteine residue and the metal ions is approximately 2.4–2.6 A ˚ , which per- mits a direct interaction. Likewise, in the structural model of PhoACY, the cysteine residue occupies a similar coordination position, suggesting that cysteine at position 104 is an alternative Zn 2+ -associating residue. When PseCP, which is included in the M20 family of metallopeptidases, was compared with PhoACY, the structure of PseCP resembled that of PhoACY in the overall shape and active site conformation in spite of the low homology between the amino acid sequences. The active site of PseCP contains two zinc atoms bridged by a water molecule (which acts as an attack- ing hydroxyl ion nucleophile), together with a glutamic acid residue (Glu175 of PseCP) (which plays a role as a general base in hydrolytic catalysis) [25]. The hydro- lytic catalysis of l-aminoacylases, including PhoACY, seems to obey a mechanism similar to that proposed in pACY1, where the water molecule binding the zinc metal and glutamic acid residue act as a nucleophile and general base, respectively [17]. By contrast, in the case of d-aminoacylase, AfaACY from Al. faecalis strain DA1 is quite different from PhoACY in its overall structure. AfaACY includes one catalytically essential Zn 2+ in its active site, but an additional metal ion can be kept at a second metal- binding site. According to the proposed catalytic mechanism for d-aminoacylase, Asp366 of AfaACY abstracts a proton from the water molecule; then, the catalytically essential Zn 2+ polarizes the carbonyl–oxy- gen bond to facilitate nucleophilic attack on the amide carbon atom, leading to the formation of a tetrahedral intermediate [26]. In the predicted mechanism for l-aminoacylase, Glu146 accepts a proton from the zinc-binding water and a hydroxide group attacks the carbonyl–oxygen bond in a substrate polarized by Arg348 to form the tetrahedral intermediate in pACY1 [17]. In this context, the hydrolysis of the amide bond of the substrate occurs in a similar manner, but the contribution of catalytically essential Zn 2+ to the reac- tion is different between l- and d-aminoacylases. In addition, the metal ion at the second position also seems to have a different role in the two types of aminoacylase. The presence of excess Zn 2+ causes no inhibition of the activity in the case of PhoACY (Table 2), but additional Zn 2+ can inhibit the enzyme activity of AfaACY [27]. With regard to hACY1, Lindner et al. [23] have proposed that Arg276 interacts with the substrate accompanying a conformational change of the enzyme. Arg276 corresponds to Arg260 in PhoACY, and this residue is located in a similar manner to Arg276 in hACY1, being apart from the active site in PhoACY. The mutant R260A loses its enzyme activity, suggesting AB Fig. 5. Molecular model presentation of PhoACY and BsbAH. (A) Superimposition of PhoACY (blue) on BsbAH (PDB ID: 1YSJ) (yellow). The model of PhoACY was based on the structure of AthIAAH (PDB ID: 1XMB). The nickel ions binding to BsbAH are shown by a pink sphere. (B) Speculated zinc-binding domain of PhoACY. Zinc ion is shown by a pink sphere. The residues related to zinc binding are shown by the balls and sticks. Characterization of thermostable aminoacylase K. Tanimoto et al. 1146 FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS that the same reaction mechanism exists in PhoACY and hACY1. The mutant H198A also loses its enzyme activity. The corresponding residues His206 in hACY1 and His205 in pACY1 are presumed to interact with an active site of the counter-subunit of a dimeric struc- ture [16,17,23]. The molecular model showed that His198 is located in the vicinity of Arg260 in the coun- ter-subunit (data not shown), suggesting that His198 can interact with an active site of the counter-subunit with conformational change of the enzyme. The molec- ular model prediction and mutagenesis experiments support the view that PhoACY catalyzes the hydrolysis of a carbon–nitrogen bond, according to a mechanism similar to that proposed for hACY1 and pACY1, although further work is needed to clarify the detailed catalytic mechanism. In conclusion, we have identified and characterized a new thermostable aminoacylase from the hyperthermo- philic archaeon P. horikoshii (PhoACY). Differences in the substrate specificity between thermostable amino- acylases were found in spite of the high similarity in their amino acid sequences. The substrate specificity could be altered by the replacement of Zn 2+ with Mn 2+ . The site-directed mutagenesis and molecular model predicted that four amino acid residues were critical as zinc-binding ligands of PhoACY, and also suggested that a common catalytic mechanism may occur in archaeal and mammalian aminoacylases. Experimental procedures Bacteria, plasmids, medium and chemicals E. coli JM109 and E. coli Rosetta (DE3) pLysS cells (Nov- agen, Madison, WI, USA) were used as hosts for DNA manipulation and overexpression of the cloned genes, respectively. The plasmid carrying the DNA fragment con- taining the PhoACY gene, PH0722 (NBRC G01-001-282), was purchased from the Biological Resource Center (NBRC) at the National Institute of Technology and Eval- uation (NITE) (Kisarazu, Japan). Luria–Bertani (LB) med- ium was used for the cultures of E. coli strains. Substrates for the enzymatic reaction were obtained from Sigma-Aldrich Inc. (St Louis, MO, USA), Wako Pure Chemical Industries, Ltd. (Osaka, Japan) and Peptide Insti- tute Inc. (Minoh, Japan). All of the other chemicals were of analytical grade. Cloning and overexpression of the PhoACY gene The gene of PhoACY was amplified using the following two primers (forward primer, 5¢-GCGGAATTCCA TATGTTGGTGGAAGTCCA-3¢; reverse primer, 5¢-GAA GATCTAACCTTTGAAGTTGAAAGC-3¢; NdeI and BglII sites in italic type) from template DNA. Amplification by PCR was carried out using KOD Plus DNA polymerase (Toyobo Co. Ltd, Osaka, Japan) according to the manufac- turer’s instructions with minor modifications. The amplified DNA fragment was digested by the restriction enzymes and inserted into a pET-11a vector (Merck KGaA, Darmstadt, Germany) digested by NdeI and BamHI. The DNA sequence was confirmed using a DNA autosequencer (ABI PRISM 310 genetic analyzer; Applied Biosystems, Foster City, CA, USA). The transformants were grown in LB medium containing 1mm ZnCl 2 and 50 lgÆmL )1 ampicillin at 37 °C. After incubation with shaking at 37 °C until the absorbance at 600 nm reached around 0.6, an inducer (isopropyl thio-b- d-galactopyranoside) was added to the culture at a final concentration of 1 mm, and the culture lasted for 6 h. The cells were harvested by centrifugation (8000 g for 20 min) and frozen at ) 30 °C. Purification of the recombinant enzyme The cells were resuspended in 50 mm Tris ⁄ HCl buffer (pH 8.0) and disrupted by ultrasonication. The crude extract was heated at 85 °C for 30 min and the supernatant was recovered by centrifugation (20 000 g for 10 min). After dialysis against 50 mm Tris ⁄ HCl buffer (pH 8.0), the crude enzyme was loaded on to a HiTrap Q column (GE Healthcare UK Ltd., Little Chalfont, UK). The column was washed with the same buffer and eluted with a linear NaCl gradient. The fractions containing protein with a molecular mass expected from the amino acid sequence of the enzyme were pooled and loaded again on to a HiTrap Phenyl HP column (GE Healthcare UK Ltd.), followed by elution with 50 mm Tris ⁄ HCl buffer (pH 8.0) with a linear Na 2 SO 4 gradient. The fractions exhibiting a single band, with a molecular mass of 42 kDa determined by SDS- PAGE, were collected. The concentration of protein was determined using a protein assay kit (BioRad Laboratories, Inc., Hercules, CA, USA) with bovine serum albumin as a standard. The molecular mass of the enzyme was determined using SDS-PAGE. The molecular mass of the enzyme in a native form was also determined by HPLC, employing a TSK Gel G3000SWXL column (Tosoh Corp., Tokyo, Japan), with elution using 50 mm Tris ⁄ HCl buffer containing 0.4 m NaCl. Enzyme assay Aminoacylase activity was measured by detecting l-amino acid cleaved from N-acyl-l-amino acid using the colorimet- ric ninhydrin method [28]. In a routine assay, unless other- wise noted, the reaction mixture (990 lL) containing 10 mm AcMet in 50 mm sodium phosphate buffer (pH 7.5) K. Tanimoto et al. Characterization of thermostable aminoacylase FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 1147 was heated in advance at 90 °C for 5 min, and then the enzyme solution (10 lL) was added to the mixture just before starting the reaction at 90 °C for 5 min. The reac- tion mixture was cooled on ice, and 200 l L ninhydrin reagent was added to 400 lL of the reaction mixture, fol- lowed by heating at 100 °C for 15 min. After cooling the solution on ice, 1 mL of 70% ethanol was added to the solution and the absorbance at 570 nm was measured. The amount of methionine produced was determined from a calibration line drawn with the authentic amino acid. A blank test was also performed in the absence of the enzyme. In the enzyme reactions with other substrates, the l-amino acids released were likewise quantified from their respective calibration lines using the corresponding amino acids. One unit of aminoacylase activity (U) was defined as the amount of enzyme which hydrolyzed 1 lmol of l-amino acid per minute. The thermostability of the enzyme was examined by keeping the enzyme in 50 mm Tris ⁄ HCl buffer (pH 8.0) at the indicated temperature. An aliquot of the enzyme solu- tion was withdrawn at the prescribed time and cooled immediately on ice. The residual activity of the enzyme was measured according to the routine assay, as mentioned above. Effect of metal ions on PhoACY activity The enzyme solution was dialyzed against an approximate 300-fold volume of 50 mm Tris ⁄ HCl buffer (pH 8.0) con- taining 20 mm EDTA (disodium salt) for 6 h at 4 °Cto prepare the metal-free enzyme, followed by further dialysis against 50 mm Tris ⁄ HCl buffer (pH 8.0) for 6 h at 4 °Cto remove EDTA from the solution. The enzyme activity was measured by the routine assay in the presence of the indi- cated metal ions. The bound metal ions in the purified enzymes were analyzed by ICP-OES (ULTIMA2, Horiba Jobin Yvon Inc., Edison, NJ, USA) at Tsukuba Technical Center at the National Institute of Advanced Industrial Sci- ence and Technology (AIST). Preparation of mutants Site-directed mutation was introduced to the enzyme by an overlap extension PCR method [29] using the primers listed in Table 4. The overexpression and purification of the mutant enzymes were performed using the same methods as employed for the wild-type enzyme. Homology modeling of PhoACY The optimized homology model of PhoACY was constructed using the automated homology modeling program 3d-jig- saw (http://www.bmm.icnet.uk/servers/3djigsaw/) [30]. This program searched parent sequences from the Protein Data Bank using the program psi-blast, and then AthIAAH (46% identity, PDB ID: 1XMB) was selected as a template for PhoACY. Next, the position-specific scoring matrix (PSSM) of PhoACY and AthIAAH sequences was calculated from the psi-blast analysis, and the PSSM obtained was used for the prediction of secondary structure by the program psipred. After picturing the alignments of target (PhoACY) to parent (AthIAAH) structures, the selection of loops, complete backbone and side-chain rotamers and the energy refinement of the modeling structure were performed in series by the program modules installed in 3d-jigsaw. Acknowledgements The authors wish to thank Dr Noritake Yasuoka (Emeritus Professor of Himeji Institute of Technology) for helpful advice in building the homology model, and Dr Kim Han-Woo (AIST) for technical support in the HPLC analysis. References 1 Schulze B & Wubbolts MG (1999) Biocatalysis for industrial production of fine chemicals. Curr Opin Biotechnol 10, 609–615. 2 Huber H & Stetter KO (1998) Hyperthermophiles and their possible potential in biotechnology. J Biotechnol 64, 39–52. 3 Birnbaum SM, Levintow L, Kingsley RB & Greenstein JP (1952) Specificity of amino acid acylases. J Biol Chem 194, 455–470. 4 Koreishi M, Asayama F, Imanaka H, Imamura K, Kadota M, Tsuno T & Nakanishi K (2005) Purification and characterization of a novel aminoacylase from Streptomyces mobaraensis. Biosci Biotechnol Biochem 69, 1914–1922. 5 Toogood HS, Hollingsworth EJ, Brown RC, Taylor IN, Taylor SJ, McCague R & Littlechild JA (2002) A thermo- stable l-aminoacylase from Thermococcus litoralis : clon- ing, overexpression, characterization, and applications in biotransformations. Extremophiles 6, 111–122. Table 4. Forward primers for the mutagenesis of the PhoACY gene. The italic type shows the nucleotides that were altered to introduce the desired mutations. Mutant Oligonucleotide sequence (5¢-to3¢) H106A CATGCTTGTGGTGCCGACGCCCACACA D107N GCTTGTGGTCACAACGCCCACACAGCA E139Q TTCCAGCCTGCTCAAGAGGGAGGAAAT E140Q CAGCCTGCTGAACAGGGAGGAAATGGA H164A ATCTTTGGGATCGCCGTGTGGGCCGAG H198A GGGAAGGGAGGTGCCGGAGCAGCACCT R260A GAAGGGACATTCGCATTCTTTACCCAG Characterization of thermostable aminoacylase K. Tanimoto et al. 1148 FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 6 Tewari YB (1990) Thermodynamics of industrially- important, enzyme-catalyzed reactions. Appl Biochem Biotechnol 23, 187–203. 7 Sakanyan V, Desmarez L, Legrain C, Charlier D, Mett I, Kochikyan A, Savchenko A, Boyen A, Falmagne P, Pierard A, et al. (1993) Gene cloning, sequence analysis, purification, and characterization of a thermostable aminoacylase from Bacillus stearothermophilus. Appl Environ Microbiol 59, 3878–3888. 8 Curley P, Does C, Driessen AJM, Kok J & Sinderen D (2003) Purification and characterization of a lactococcal aminoacylase. Arch Microbiol 179, 402–408. 9 Yang YB, Hu HL, Chang MC, Li H & Tsai YC (1994) Purification and characterization of l-aminoacylase from Alcaligenes denitrificans DA181. Biosci Biotechnol Biochem 58, 204–205. 10 Lin LL, Chen MH, Chen HCR, Kan SC, Chen CC, Hu HY & Hsu WH (2007) Characterization of a bifunctional aminoacylase ⁄ carboxypeptidase from radioresistant bacterium Deinococcus radiodurans R1. J Biotechnol 128, 322–334. 11 Javid-Majd F & Blanchard JS (2000) Mechanistic anal- ysis of the argE-encoded N-acetylornithine deacetylase. Biochemistry 39, 1285–1293. 12 Story SV, Grunden AM & Adams MWW (2001) Char- acterization of an aminoacylase from the hyperthermo- philic archaeon Pyrococcus furiosus. J Bacteriol 183, 4259–4268. 13 Ishikawa K, Ishida H, Matsui I, Kawarabayasi Y & Kikuchi H (2001) Novel bifunctional hyperthermostable carboxypeptidase ⁄ aminoacylase from Pyrococcus hori- koshii OT3. Appl Environ Microbiol 67, 673–679. 14 Gonzalez JM, Masuchi Y, Robb FT, Ammerman JW, Maeder DL, Yanagibayashi M, Tamaoka J & Kato C (1998) Pyrococcus horikoshii sp. nov., a hyperthermo- philic archaeon isolated from a hydrothermal vent at the Okinawa Trough. Extremophiles 2, 123–130. 15 Colombo S, Toietta G, Zecca L, Vanoni M & Tortora P (1995) Molecular cloning, nucleotide sequence, and expression of a carboxypeptidase-encoding gene from the archaebacterium Sulfolobus solfataricus. J Bacteriol 177, 5561–5566. 16 Lindner HA, Lunin VV, Alary A, Hecker R, Cygler M &Me ´ nard R (2003) Essential roles of zinc ligation and enzyme dimerization for catalysis in the aminoacylase- 1 ⁄ M20 family. J Biol Chem 278, 44496–44504. 17 Liu Z, Zhen Z, Zuo Z, Wu Y, Liu A, Yi Q & Li W (2006) Probing the catalytic center of porcine aminoacylase 1 by site-directed mutagenesis, homology modeling and substrate docking. J Biochem 139, 421– 430. 18 Occhipinti E, Martelli PL, Spinozzi F, Corsi F, Form- antici C, Molteni L, Amenitsch H, Mariani P, Tortora P & Casadio R (2003) 3D structure of Sulfolobus solfa- taricus carboxypeptidase developed by molecular model- ing is confirmed by site-directed mutagenesis and small angle X-ray scattering. Biophys J 85, 1165–1175. 19 Gentzen I, Loffler HG & Schneider F (1980) Aminoacy- lase from Aspergillus oryzae. Comparison with the pig kidney enzyme. Z Naturforsch [C] 35, 544–550. 20 Palm GJ & Rohm KH (1995) Aminoacylase 1 from por- cine kidney: identification and characterization of two major protein domains. J Protein Chem 14, 233–240. 21 Wakayama M, Shiiba E, Sakai K & Moriguchi M (1998) Purification and characterization of l-aminoacy- lase from Pseudomonas maltophila B1. J Ferment Bioeng 85, 278–282. 22 Weiss HM, Palm GJ & Rohm KH (1995) Thermostable aminoacylase from Bacillus stearothermophilus: signifi- cance of the metal center for catalysis and protein sta- bility. Biol Chem Hoppe Seyler 376, 643–649. 23 Lindner HA, Alary A, Boju LI, Sulea T & Me ´ nard R (2005) Roles of dimerization domain residues in binding and catalysis by aminoacylase-1. Biochemistry 44, 15645–15651. 24 Anders MW & Dekant W (1994) Aminoacylases. Adv Pharmacol 27, 431–448. 25 Rowsell S, Pauptit RA, Tucker AD, Melton RG, Blow DM & Brick P (1997) Crystal structure of carboxypeptidase G2, a bacterial enzyme with applications in cancer therapy. Structure 5, 337–347. 26 Liaw SH, Chen SJ, Ko TP, Hsu CS, Chen CJ, Wang AH & Tsai YC (2003) Crystal structure of d-aminoacy- lase from Alcaligenes faecalis DA1: a novel subset of amidohydrolases and insights into the enzyme mecha- nism. J Biol Chem 278, 4957–4962. 27 Lai WL, Chou LY, Ting CY, Kirby R, Tsai YC, Wang AH & Liaw SH (2004) The functional role of the binuclear metal center in d-aminoacylase: one-metal activation and second-metal attenuation. J Biol Chem 279, 13962–13967. 28 Yokoyama S & Hiramatsu J (2003) A modified ninhy- drin reagent using ascorbic acid instead of potassium cyanide. J Biosci Bioeng 95, 204–205. 29 Ho SN, Hunt HD, Horton RM, Pullen JK & Pease LR (1989) Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77, 51–59. 30 Bates PA, Kelley LA, MacCallum RM & Sternberg MJE (2001) Enhancement of protein modeling by human intervention in applying the automatic programs 3D-JIGSAW and 3D-PSSM. Proteins 45, Suppl. 5, 39–46. K. Tanimoto et al. Characterization of thermostable aminoacylase FEBS Journal 275 (2008) 1140–1149 ª 2008 The Authors Journal compilation ª 2008 FEBS 1149 . Characterization of thermostable aminoacylase from hyperthermophilic archaeon Pyrococcus horikoshii Koichi Tanimoto 1 ,. kinetics (Fig. 4). The value of the Michaelis constant for Fig. 1. Alignment of amino acid sequences of P. horikoshii aminoacylase with aminoacylases from various sources