Báo cáo khoa học: Molecular design of a nylon-6 byproduct-degrading enzyme from a carboxylesterase with a b-lactamase fold ppt

Molecular design of a nylon-6 byproduct-degrading enzyme from a carboxylesterase with a b-lactamase fold Yasuyuki Kawashima 1, *, Taku Ohki 1, *, Naoki Shibata 2,3, *, Yoshiki Higuchi 2,3 , Yoshiaki Wakitani 1 , Yusuke Matsuura 1 , Yusuke Nakata 1 , Masahiro Takeo 1 , Dai-ichiro Kato 1 and Seiji Negoro 1 1 Department of Materials Science and Chemistry, Graduate School of Engineering, University of Hyogo, Japan 2 Department of Life Science, Graduate School of Life Science, University of Hyogo, Japan 3 RIKEN Harima Institute, SPring-8 Center, Sayo-gun, Hyogo, Japan Nylon-6 is produced by ring-cleavage polymerization of e-caprolactam, and consists of more than 100 units of 6-aminohexanoate (Ahx). During the polymeriza- tion reaction, some molecules fail to polymerize and remain oligomers, whereas others undergo head-to-tail condensation to form cyclic oligomers. These Ahx olig- omers (designated as nylon oligomers) are byproducts from nylon-6 factories, and contribute to the increase in the amounts of industrial waste material discharged into the environment. Therefore, an efficient system for degradation of these byproducts is required. How- ever, the efficiency of degradation is highly dependent on specific enzymes that can catalyze the desired reac- tions. We have been studying the degradation of the Ahx oligomer by Arthrobacter sp. KI72 as a model system [1–15]. Previous biochemical studies have revealed that the Ahx linear dimer (Ald) hydrolase (NylB) responsible for degradation of the nylon oligo- mers and a carboxylesterase (NylB¢), which has approxi- mately 0.5% of the NylB level of Ald-hydrolytic Keywords 6-aminohexanoate-dimer hydrolase; carboxylesterase; nylon oligomer; X-ray crystallography; b-lactamase Correspondence S. Negoro, Department of Materials Science and Chemistry, Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji, Hyogo 671 2280, Japan Fax / Tel: +81 792 67 4891 E-mail: negoro@eng.u-hyogo.ac.jp Y. Higuchi, Department of Life Science, Graduate School of Life Science, University of Hyogo, 3-2-1 Koto, Kamigori-cho, Ako-gun, Hyogo 678 1297, Japan Fax / Tel: +81 791 58 0179 E-mail: hig@sci.u-hyogo.ac.jp *These authors contributed equally to this work (Received 11 December 2008, revised 13 February 2009, accepted 20 February 2009) doi:10.1111/j.1742-4658.2009.06978.x A carboxylesterase with a b-lactamase fold from Arthrobacter possesses a low level of hydrolytic activity (0.023 lmolÆmin )1 Æmg )1 ) when acting on a 6-aminohexanoate linear dimer byproduct of the nylon-6 industry (Ald). G181D ⁄ H266N ⁄ D370Y triple mutations in the parental esterase increased the Ald-hydrolytic activity 160-fold. Kinetic studies showed that the triple mutant possesses higher affinity for the substrate Ald (K m = 2.0 mm) than the wild-type Ald hydrolase from Arthrobacter (K m =21mm). In addition, the k cat ⁄ K m of the mutant (1.58 s )1 Æmm )1 ) was superior to that of the wild- type enzyme (0.43 s )1 Æmm )1 ), demonstrating that the mutant efficiently con- verts the unnatural amide compounds even at low substrate concentrations, and potentially possesses an advantage for biotechnological applications. X-ray crystallographic analyses of the G181D ⁄ H266N ⁄ D370Y enzyme and the inactive S112A-mutant–Ald complex revealed that Ald binding induces rotation of Tyr370 ⁄ His375, movement of the loop region (N167–V177), and flip-flop of Tyr170, resulting in the transition from open to closed forms. From the comparison of the three-dimensional structures of various mutant enzymes and site-directed mutagenesis at positions 266 and 370, we now conclude that Asn266 makes suitable contacts with Ald and improves the electrostatic environment at the N-terminal region of Ald cooperatively with Asp181, and that Tyr370 stabilizes Ald binding by hydrogen-bonding ⁄ hydrophobic interactions at the C-terminal region of Ald. Abbreviations Ahx, 6-aminohexanoate; Ald, 6-aminohexanoate linear dimer; DD, D-alanyl-D-alanine. FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS 2547 activity, are encoded on a plasmid in Arthrobacter [4,7–9] (Fig. 1). NylB and NylB¢ are composed of 392 amino acid residues, but differ at 46 residues. How- ever, a single G181D [Gly181 (NylB¢ type) to Asp (NylB type)] substitution in the NylB¢ sequence results in a 20-fold increase in hydrolytic activity [10,13]. Also, an additional alteration, H266N [His266 (NylB¢ type) to Asn (NylB type)], increases the Ald-hydrolytic activity to that of the parental NylB enzyme [10,12]. Three-dimensional structures of the enzymes provide not only basic information, such as catalytic mecha- nism and enzyme evolution, but also the information required for improvement of enzyme function. A NylB ⁄ NylB¢ hybrid (designated Hyb-24) gives good crystals suitable for X-ray crystallographic analysis. Hyb-24 includes five amino acid replacements (T3A, P4R, T5S, S8Q, and D15G) in the NylB¢ protein (Table 1), but possesses the NylB¢ level of Ald-hydro- lytic activity [11,14]. X-ray crystallographic analysis of Hyb-24 and Hyb-24DN (a Hyb-24 mutant with G181D ⁄ H266N substitutions) has revealed that these enzymes generate a two-domain structure (a and a ⁄ b) that is similar to the folds of the penicillin-recognizing family of serine-reactive hydrolases, especially to those of the d-alanyl-d-alanine (DD) carboxypeptidase from Streptomyces and the carboxylesterase (EstB) from Burkholderia [11,12]. We have proposed that Ser112, Lys115 and Tyr215 are catalytic residues in Hyb-24 and Hyb-24DN [12]. Tyr215-O g functions as a general base to increase the nucleophilicity of Ser112, which performs nucleophilic attacks on amide compounds. The positively charged Lys115-N f stabilizes the Ser112-O c) anion. However, in Hyb-24DN, additional amino acid residues, Tyr170 and Asp181, which are unnecessary for the esterolytic activi- ties, are required to confer a NylB level of hydrolytic activity towards Ald. In addition, nylon oligomer hydro- lase exhibits unique structural alterations induced by Ald, i.e. movement of the loop region (N167–V177) and flip-flop of Tyr170. On the basis of these findings, we have proposed that amino acid substitutions in the cata- lytic cleft of a pre-existing esterase with a b-lactamase fold result in the evolution of a nylon oligomer hydro- lase, and that catalysis proceeds according to the follow- ing steps: (a) Ald-induced transition from open (substrate-unbound) to closed (substrate-bound) forms; (b) nucleophilic attack on Ald by Ser112 and formation Fig. 1. Reaction catalyzed by NylB ⁄ NylB¢. Hydrolysis of -6-amino- hexanoate linear dimer (Ald) (nylon oligomer hydrolytic activity) (A) and p-nitrophenylacetate (esterolytic activity) (B). Table 1. Enzymes and plasmids. Abbreviations Characteristics Reference NylB Wild-type Ald hydrolase from Arthrobacter sp. KI72 [4,7] NylB¢ Wild-type carboxylesterase with b-lactamase fold from Arthrobacter sp. KI72 (88% homology to NylB) [7–9] Hyb-24 NylB ⁄ NylB¢ hybrid protein constructed from conserved PvuII sites located 24 amino acid residues downstream of the initiation codons (NylB¢ containing T3A ⁄ P4R ⁄ T5S ⁄ S8Q ⁄ D15G substitutions) [11,14] Hyb-24D Hyb-24 containing G181D substitution [13] Hyb-24D-A 112 Hyb-24D containing S112A substitution This study Hyb-24N Hyb-24 containing H266N substitution This study Hyb-24Y Hyb-24 containing D370Y substitution [13] Hyb-24DN Hyb-24 containing G181D ⁄ H266N substitutions [12] Hyb-24DN-A 112 Hyb-24DN containing S112A substitution [12] Hyb-24DY Hyb-24 containing G181D ⁄ D370Y substitutions [13] Hyb-24DNY Hyb-24 containing G181D ⁄ H266N ⁄ D370Y substitutions This study Hyb-24DNY-A 112 Hyb-24DNY containing S112A substitution This study Hyb-24DD Hyb-24 containing G181D ⁄ H266D substitutions This study Hyb-24DG Hyb-24 containing G181D ⁄ H266G substitutions This study pHY3 Hybrid plasmid composed of vector pKP1500 region and a 1.4 kb EcoRI ⁄ HindIII fragment containing the Hyb-24gene [11,14] Molecular design of nylon oligomer hydrolase Y. Kawashima et al. 2548 FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS of a tetrahedral intermediate; (c) formation of an acyl enzyme and transition to an open form; and (d) deacyla- tion [12]. We have concluded that Asp181-COO ) stabilizes substrate binding by electrostatic interactions with Ald-NH þ 3 [12]. However, the role of Asn266 was still unclear. In addition, random mutagenesis experiments with the parental carboxylesterase (Hyb-24) gene have revealed that a D370Y substitution occurring opposite to Gly181 in the catalytic cleft (17.1 A ˚ from Tyr370 at the C a position) enhances Ald-hydrolytic activity eight-fold in comparison to the parental Hyb-24 (0.023 lmolÆmin )1 Æmg )1 protein) [13]. In the current study, we constructed a mutant enzyme by integrating the three amino acid alterations (G181D ⁄ H266N ⁄ D370Y) individually into the Hyb-24 sequence, and examined the cumulative effects on Ald-hydrolytic activity. Moreover, we determined the three-dimen- sional structures of Hyb-24D (Hyb-24 with having the G181D substitution), Hyb-24DNY (Hyb-24 mutant with G181D ⁄ H266N ⁄ D370Y substitutions), and their enzyme–substrate complexes, and analyzed the roles of Asn266 and Tyr370 in comparison with the structures of typical enzymes in the penicillin-recognizing family of serine-reactive hydrolases. Results and Discussion Cumulative effects of amino acid substitution on Ald-hydrolytic activity The catalytic function of mutant enzymes was com- pared on the basis of specific activity for Ald. Assays conducted under standard assay conditions using 10 mm Ald have shown that G181D ⁄ H266 ⁄ D370Y substitutions in Hyb-24 increase the activity 160-fold (3.6 lmolÆmin )1 Æmg )1 ) (Table 2). To determine the individual effects of G181D, H266N and D370Y sub- stitutions on the catalytic function, we constructed Hyb-24 mutants with combinations of the substitu- tions, and determined the kinetic parameters (Table 3). We could not determine K m and V max values for parental Hyb-24 and Hyb-24N (Gly181 enzymes), owing to their low activities, whereas the K m for Hyb- 24D (31 mm for Ald) was found to be close to the value for wild-type NylB (21 mm), suggesting that a single G181D substitution improves Ald binding. Simi- larly, D370Y substitution in Hyb-24 improved Ald binding (K m =39mm in Hyb-24Y). In contrast, a sin- gle H266N substitution in Hyb-24 decreased Ald hydrolytic activity (Table 2; Hyb-24N). In Hyb-24D (Asp181 enzymes), H266N substitution increased the k cat 4.6-fold, but barely affected the K m value (see Hyb-24DN). This result suggests that H266N substitution is effective at increasing the turn- over of the catalytic reaction in combination with Asp181. In contrast, D370Y substitution decreased the K m to 7 mm (23% of the level of Hyb-24D), but barely affected the k cat value (see Hyb-24DY). Thus, D370Y substitutions mainly improved Ald binding. In Hyb-24DN, D370Y substitution further improved substrate binding (K m =2mm). Although the muta- tion had negative effects on k cat (30% of that of Hyb- 24), the k cat ⁄ K m value of Hyb-24DNY (1.58 s )1 Æmm )1 ) was four-fold to five-fold that of Hyb-24DN (0.34 s )1 Æmm )1 ) and wild-type NylB (0.43 s )1 Æmm )1 ). The enzyme–substrate interactions at positions 181, 266 and 370 are discussed below on the basis of the three-dimensional structures. Enzyme–substrate interaction in the catalytic cleft To analyze the structural alterations induced by G181D, H266N and D370Y substitutions in Hyb-24, we per- formed X-ray crystallographic analysis of Hyb-24D and Hyb24DNY, and compared the structures with those of Hyb-24 and Hyb-24DN, which had been identified pre- viously [11–13]. Superimposition of the four molecules revealed that the overall structures share almost identi- cal folding patterns within rmsd values smaller than 0.2 A ˚ . In Hyb-24 (Gly181 enzyme), the region encom- passing D169–A174 had poor electron density, and therefore Tyr170, which is responsible for substrate binding, could not be identified in the three-dimensional models [11]. In contrast, the electron density maps of Hyb-24D and Hyb-24DNY (Asp181 enzyme) at the flex- ible loop region (N167–V177) were clear enough to assign all the side chain atoms. Hydrogen bonding between Tyr170-O g and Asp181-O d fixes the flexible loop in the open form, which has an energy barrier for transition to the closed form (Figs 2 and 3 and Fig. S1). Upon substrate binding, the loop is shifted by approximately 5 A ˚ at Tyr170-C a , and the side chain of Tyr170 is rotated. Through the combined effect, Tyr170-O g moves a total of approximately 11 A ˚ , resulting in the formation of hydrogen bonds with the nitrogen of the amide linkage in Ald. In addition, in the Hyb-24DN-A 112 –Ald complex, the electron density map was poor for the C-terminal half of Ald [12]. In contrast, in Hyb-24D-A 112 –Ald and Hyb-24DNY- A 112 –Ald, the catalytic and binding residues and Ald in the catalytic cleft had clear electron density distribu- tions for which structural models could be determined (Fig. 2A). Thus, the movement of the loop and rotation of Tyr170 cover the active site to generate a Y. Kawashima et al. Molecular design of nylon oligomer hydrolase FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS 2549 closed form, and the modes of these motions were conserved for the three Hyb-24-related proteins (Fig. 3, and Figs S1 and S2). On the bases of these findings, we concluded that dynamic motions induced by Ald play essential roles in Ald hydrolysis. Superimposition of the bound and unbound Ald structures revealed that catalytic residues (Ser112 and Lys115) are conserved at the original positions in Hyb-24D, Hyb-24DN, and Hyb-24DNY. In contrast, upon substrate binding, another catalytic residue (Tyr215) rotates its side chain by approximately 40° around the C b –C c bond, and the phenolic oxygen moves by 0.81–1.1 A ˚ (Fig. 3 and Fig. S1). These results suggest that stable binding of Ald by electro- static interaction with Asp181-COO - causes movement of Tyr215 for suitable positioning in the enzyme– substrate complex. In substrate-unbound Hyb-24D and Hyb-24DN, Asp370-O d forms hydrogen bonds with the His375 imid- azole, and even after Ald binding, no significant move- ments were identified for Asp370 and His375 (Fig. 3A). In contrast, in Hyb-24DNY, Ald binding induced the following structural alterations (Fig. 3B): the His375 side chain rotates by approximately 100° around C a –C b and by approximately 170° around C b –C c , and flips the imidazole ring (6.1 A ˚ at the imidazole nitrogen), to gen- erate a hydrogen-bonding network including Tyr370-O g and the Ald carboxyl (distance approximately 3.2 A ˚ ). Through this effect, Tyr370 moves the aromatic ring (6.4 A ˚ at Tyr-O g ), allowing it to contact Ald (Fig. 3B). These results suggest that binding at the C-terminal region of Ald is improved by the D370Y substitution. The roles of Asn266 and Tyr370 were further examined by site-directed mutagenesis. Roles of Asn266 Superimposition of Hyb-24DNY with the class A b-lactamase revealed that Asn266 of Hyb-24DNY has a similar spatial position to that of Glu166 in the class A b-lactamase (Fig. S3). In the class A b-lactam- ase (Protein Data Bank ID code, 1M40), the distance between Ser70-O c and Glu166-O e2 is 4.5 A ˚ , and the so-called ‘hydrolytic water’ (Wat1004) forms a bridge between the two residues by hydrogen bonding (Fig. S3). Moreover, this network is believed to be responsible for the b-lactam hydrolysis [16–20]. In Hyb-24DNY, Asn266-O d is 5.0 A ˚ away from Ser112- O c , and this distance is slightly larger than the distance between Ser70-O c and Glu166-O e2 of b-lactamase. However, upon substrate binding, water molecules (Wat115, Wat357, Wat375, etc.) in the catalytic cleft of Hyb-24DNY are excluded. Wat397, nearest to Table 3. Kinetic parameters of His-tagged Hyb-24 and its mutant enzymes for Ald. Ald-hydrolytic activity was assayed using the His- tagged purified enzymes under standard assay conditions, except that various concentrations of Ald were used. Kinetic parameters (k cat and K m values) were evaluated by directly fitting data to the Michaelis–Menten equation using GRAPHPAD prism, version 5.01 (GraphPad, San Diego, CA, USA). The k cat values are expressed as turnover numbers per subunit (M r of the subunit: 42 000). Enzyme Kinetic parameters k cat (s )1 ) K m (mM) k cat ⁄ K m (s )1 ÆmM )1 ) Hyb-24D 2.01 ± 0.24 30.9 ± 8.01 0.065 Hyb-24Y 0.61 ± 0.056 39.1 ± 6.73 0.016 Hyb-24DN 9.2 ± 0.24 27.2 ± 1.25 0.34 Hyb-24DY 2.5 ± 0.15 7.1 ± 1.28 0.35 Hyb-24DNY 3.2 ± 0.19 2.0 ± 0.33 1.58 NylB 9.0 ± 1.46 21.0 ± 6.65 0.43 Table 2. Effect of amino acid alternations in His-tagged Hyb-24 on enzyme activity. Enzyme activities of His-tagged proteins were assayed using 10 m M Ald, 0.2 mM p-nitrophenylacetate (C2 ester), 0.2 mM p-nitrophenylbutyrate (C4 ester) and p-nitrophenyloctanoate (C8 ester) as substrates. Details are given in Experimental procedures. The numbers in parentheses indicate the relative activities expressed as a ratio of the specific activity of the Hyb-24 protein. Enzyme Ald-hydrolytic activity (lmolÆmin )1 Æmg )1 ) Esterase activity (lmolÆmin )1 Æmg )1 ) C2 ester C4 ester C8 ester Hyb-24 0.023 (1) 4.81 (1) 2.61 (1) 0.28 (1) Hyb-24D (G181D) 0.47 (20) 3.16 (0.66) 1.28 (0.49) < 0.005 (< 0.02) Hyb-24N (H266N) 0.008 (0.35) 7.63 (1.59) 0.70 (0.27) 0.28 (1.0) Hyb-24Y (D370Y) 0.19 (8.2) 6.31 (1.3) 13.4 (5.1) 1.28 (4.7) Hyb-24DN (G181D ⁄ H266N) 3.53 (153) 5.93 (1.23) 2.73 (1.05) 0.19 (0.68) Hyb-24DY (G181D ⁄ D370Y) 1.79 (78) 3.15 (0.65) 5.44 (2.08) 0.027 (0.096) Hyb-24DD (G181D ⁄ H266D) 1.37 (60) 1.50 (0.31) 1.78 (0.68) 0.051 (0.18) Hyb-24DG (G181D ⁄ H266G) 0.0010 (0.04) 0.48 (0.10) 0.25 (0.096) 0.0038 (0.014) Hyb-24DNY (G181D ⁄ H266N ⁄ D370Y) 3.60 (157) 10.2 (2.12) 4.40 (1.69) 0.088 (0.31) Molecular design of nylon oligomer hydrolase Y. Kawashima et al. 2550 FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS Asn266 (4.7 A ˚ away from Asn266), was identified in the substrate-bound structure, but Wat397 was not connected to Ser112-O c by the hydrogen-bonding network (Fig. S3). In addition, substitution to Asp266 in Hyb-24DN is rather inhibitory for activity, as described below (Table 2; Hyb-24DD). Thus, absence of ‘hydrolytic water’ in the catalytic cleft in Hyb- 24DNY suggests that the role of Asn266 (Asp266) is different from that of Glu166 in b-lactamase, although the possibility remains that the dynamic motion of water molecules, accompanied by open ⁄ closed inter- conversions, is responsible for the catalysis. The roles of Asn266 can be inferred from comparisons between the structure of Asn266 enzymes (Hyb-24DN- A B C Fig. 2. Stereoview of the catalytic cleft of nylon oligomer hydro- lase. (A) 2F o ) F c electron density maps of Hyb-24DNY-A 112 –Ald contoured at 1.0r. The side chains (stick diagram) of some residues (Ala112, Lys115, Tyr170, Asp181, Arg187, Tyr215, Phe264, Asn266, Phe317, Trp331, Ile343, Ile345, Tyr370, and His375), water molecules (Wat18, Wat335, Wat377, Wat397, and Wat430) and the substrate Ald are shown. (B, C) Superimposition of Hyb- 24DNY-A 112 –Ald (blue) on Hyb-24D-A 112 –Ald (green). Structures around the N-terminal region of Ald with the side chains of some residues [Asp181, Phe264, His266 (Asn266)] are shown (B). Struc- tures around the C-terminal region of Ald with the side chains of some residues [Ile343, Tyr370 (Asp370), His375] are shown (C). Hydrogen bonds between two atoms in the enzyme–substrate complex are indicated as red dotted lines, with distance in ang- stroms. Substrate Ald was refined as alternative conformations (Ald A and Ald B ) on the basis of electron density maps. A B C Fig. 3. Stereoviews of Ald-bound and unbound structures of nylon oligomer hydrolases. (A) Superimposition of Hyb-24D-A 112 –Ald (orange) on Hyb-24D (green). (B, C) Superimposition of Hyb-24DNY- A 112 –Ald (orange) on Hyb-24DNY (green). The main chain folding (ribbon diagram) and side chains (stick diagram) of some residues [Ser112 (Ala112), Lys115, Tyr170, Asp181, Tyr215, His266 (Asn266), Asp370 (Tyr370), His375] located in the catalytic cleft are shown (A, B). Hydrogen bonds are indicated as red dotted lines (Hyb-24D-A 112 –Ald and Hyb-24DNY-A 112 –Ald) and magenta dotted lines (Hyb-24D and Hyb-24DNY), with distance in angstroms. Sur- face structures of the entrance of the catalytic cleft are shown (C). Carbon, nitrogen and oxygen atoms in the substrate Ald (space-fill- ing diagram) are shown in yellow, blue, and red, respectively. Ald- bound and unbound structures without superimposition are shown in Fig. S2. Y. Kawashima et al. Molecular design of nylon oligomer hydrolase FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS 2551 A 112 –Ald and Hyb-24DNY-A 112 –Ald) and that of the His266 enzyme (Hyb-24D-A 112 –Ald). In the His266 enzyme, it is likely that the bulky His266 imidazole (2.9 A ˚ away from Ald-NH þ 3 ) creates a steric hindrance effect against substrate binding, and that the positively charged imidazole-NH + creates electrostatic repulsion against Ald-NH þ 3 in the pH range lower than the pK a of the His imidazole (Fig. 2B). Therefore, these effects should destabilize substrate binding, and H266N substi- tution is effective at diminishing the negative effects, resulting in enhancement of Ald-hydrolytic activity. To examine the effects of position 266 on enzyme activity, we constructed mutant enzymes from Hyb- 24D. Hyb-24DG (Gly266 enzyme) possessed only 0.03% of the Hyb-24DN activity (Asn266 enzyme) (Table 2). As Asn266-C a is approximately 6 A ˚ from the substrate Ald at the nearest position (C2), alteration to Gly266 should reduce the effective contact with the sub- strate. This suggests that a suitable contact at position 266 is required to hold the substrate in the catalytic cleft. In addition, the Ald-hydrolytic activity of the Asp266 mutant (Hyb-24DD) was found to be seven-fold higher that of the His266 enzyme (Hyb-24D), but the activity was still only approximately 40% of that of the Asn266 enzyme. This demonstrates that the presence of two acidic residues (Asp181 and Asp266) around Ald-NH þ 3 is rather inhibitory for the activity. We have found that various substitutions at position 181 affect the Ald-hydrolytic activity > 10 4 -fold, but barely affect the esterolytic activity [11]. In contrast, substitutions at position 266 affect both activities, although the extent of the esterolytic activity (for C2 esters, 0.48–5.93 lmolÆmin )1 Æmg )1 ; for C4 esters, 0.25–5.44 lmolÆmin )1 Æmg )1 ; and for C8 esters, 0.038– 0.19 lmolÆmin )1 Æmg )1 ) is smaller than that of the Ald-hydrolytic activity (0.001–3.53 lmolÆmin )1 Æmg )1 ) (Table 2). This may imply that alterations at posi- tion 266, which is close to the catalytic triad (Ser ⁄ Lys ⁄ Tyr), affect both activities more significantly than alterations at position 181. From these analyses, we concluded that Asn266 con- tributes to close contacts with the substrate, and that the electrostatic environment around Ald-NH þ 3 , responsible for efficient Ald binding, is generated mainly by Asp181, and additively by Asn266. Roles of Tyr370 Whereas Ald-hydrolytic activity was enhanced 160-fold through accumulation of three amino acid substitu- tions in Hyb-24, activity against the C2 ester was not as severely affected, and was only 0.65-fold to 2.1-fold higher (Table 2). However, it should be noted that a single D370Y substitution increased the esterase activ- ity against the C4 ester approximately five-fold. More- over, we have found that the activity against tributyrin (glyceryltributyrate) of Hyb-24Y was 30-fold to 50-fold of the activity of Hyb-24 [13]. In contrast, G181D sub- stitution in Hyb-24 decreased the activity against longer acyl chains. In addition, as Hyb-24DY pos- sessed lower esterase activity than Hyb-24Y, the presence of Asp181 is considered to be inhibitory also for esterase activity in Tyr370 mutants (Table 2). The carboxyl-half in the substrate Ald is surrounded by hydrophobic residues, such as Trp331, Phe317, and Ile343 (Fig. 2A,C). This suggests that the hydrophobic interactions stabilize substrate binding. In addition, D370Y substitution should make the environment of the catalytic cleft more hydrophobic, as the water molecules (Wat22, Wat44, Wat242, Wat367, Wat368, Wat400, and so on) found in Hyb-24D are excluded in Hyb-24DNY. To examine the effect of amino acid alterations at position 370 more extensively, we con- structed various mutant enzymes in which Asp370 in Hyb-24 was altered to one of 10 other amino acid residues (Table S1). To simplify the estimation of the specific activity of each enzyme, we quantified the amount of Hyb-24-related protein and Ald-hydrolytic activity in cell extracts, and normalized the data on the basis of the amount of Hyb-24-related protein (see Experimental procedures). Alterations to hydrophobic residues, especially to Trp and Phe, increased the Ald- hydrolytic activity, although the activity was slightly lower than that of Hyb-24Y (Tyr370). In addition, significant enhancements of the activities were found after substitution to Met and Ile. On the basis of these findings, we concluded that substrate binding at the C-terminal region is improved by hydrophobic inter- actions in some mutant enzymes (Phe370, Trp370, Met370 and Ile370 enzymes) rather than by specific hydrogen bonding involving Tyr370-O g . Mutation of Asp370 to hydrophobic residues also increased the substrate specificity for carboxyl esters with longer acyl chains (Table S1). Especially in the Met, Phe and Trp mutants and Hyb-24Y (Tyr370 enzyme), activity against C4 esters and C8 esters was increased 4- to 7-fold over the activity of the Asp370 enzyme. Thus, increased hydrophobic interactions around position 370 can explain the increased binding of esters with longer acyl chains, which results in the alteration of substrate specificity for carboxyl esters. Concluding remarks From the comparisons of the three-dimensional structures of Hyb-24, Hyb-24D, Hyb-24DN, and Molecular design of nylon oligomer hydrolase Y. Kawashima et al. 2552 FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS Hyb-24DNY, we suggest the following enzyme– substrate interactions, where these resulted in stepwise increases in activity: (a) effective substrate binding was achieved by electrostatic interaction between Asp181- COO ) and Ald-NH þ 3 (G181D substitution); (b) Asn266 improves the electrostatic environment cooper- atively with Asp181, and gives suitable contacts with Ald (H266N substitution); and (c) Ald binding induces cooperative movement of Tyr370 ⁄ His375, generating hydrogen-bonding ⁄ hydrophobic interactions at the C-terminal region in Ald (D370Y substitution). Thus, Ald hydrolase activity requires strict substrate binding, achieved by induced fit motion, whereas the enzyme performs the catalytic function as a more relaxed open form for carboxylesterase activity. This model coin- cides with our finding that Ald-hydrolytic activity is significantly affected by amino acid substitutions at positions 170, 181, 266, and 370, which are responsible for Ald binding, whereas amino acid substitutions at these positions do not affect esterase activity as severely. In addition, Ald hydrolase, which is superior to the wild-type enzyme in affinity for Ald and k cat ⁄ K m value, was successfully constructed by integrating G181D ⁄ H266N ⁄ D370Y substitutions into the paren- tal carboxylesterase. This result demonstrates that the mutant efficiently converts the unnatural amide compounds even at low substrate concentrations, and potentially possesses advantages for biotechnological applications. Experimental procedures Site-directed mutagenesis and construction of plasmids expressing mutant enzymes The mutant enzymes and plasmids used in this study are listed in Table 1. To obtain the other mutant enzymes from Hyb-24, site-directed mutagenesis was carried out by PCR [21], using the following primers (mutated sites are underlines): RHmutN1 (5¢-GCCGCCGT TCGCGAAGCC GAA-3¢) (for H266N substitution); RHmutD1 (5¢-GACGC CGCCGT CCGCGAAGCCGAAACCCGT-3¢) (for H266D substitution); RHmutG1 (5¢-GACGCCGCCG CCCGCG AAGCCGAAACCCGT-3¢) (for H266G substitution); and RDmutY1 (5¢-GTGTAGGGATCGGGCCACG-3¢) (for D370Y substitution). To replace Asp370 in Hyb-24 with other amino acids, the mutant primer with NNN at posi- tion 370 (RDmutX1) (5¢-CCGGTGCCAGTGCTCGGT NNNGGGATCGGGCCACGACGACAGC-3¢) was used. After nucleotide sequencing of the mutants we confirmed that isolated mutants possess a single D370N, D370E, D370K, D370T, D370C, D370G, D370I, or D370F muta- tion in the Hyb-24 sequence. For the D370W mutant, primer RDmutW1 (5¢-CCGGTGCCAGTGCTCGGT CCA GGGATCGGGCCACGACGACAGC-3¢) was used. For the D370M mutant, primer RDmutM1 (5¢-CCGGTGCC AGTGCTCGGT CATGGGATCGGGCCACGACGACA GC-3¢) was used. For isolation of S112A mutants, site- directed mutagenesis was performed with the synthetic oligonucleotide 5¢-TGCTGATG GCCGTCTCGAAGT-3¢. The mutant enzymes were expressed in Escherichi coli KP3998, using pKP1500 as the vector [11,12]. Enzyme purification, enzyme assay, and protein concentration For crystallization and X-ray diffraction experiments, native enzymes were purified to homogeneity from cell extracts of E. coli clones by successive chromatography on anion exchange (Hi-Trap Q-Sepharose; GE Healthcare Bio- Science AB, Uppsala, Sweden), gel filtration (Seph- acryl S-200 High Resolution; GE Healthcare Bio-Science AB) and anion exchange (Hi-Trap Q-Sepharose) columns [11]. In order to analyze the specific activities of various mutant enzymes, a His-tagged region was fused to the N-terminus of each mutant enzyme, using the expression vector pQE-80L (Qiagen GmbH, Hilden, Germany). The His-tagged enzymes were expressed in E. coli JM109, and purified to homogeneity [11]. Ald-hydrolytic activities were assayed at 30 °C using 10 mm Ald (chemically synthesized in our laboratory) as sub- strate in 20 mm potassium phosphate buffer (pH 7.3), con- taining 10% glycerol (standard assay condition) [9–13]. Reaction mixtures were fractionated on a C 18 RP-HPLC col- umn (TSK-GEL ODS-80Ts; TOSOH Co., Tokyo, Japan), and the decrease in Ald and increase in Ahx were monitored by absorbance at 210 nm (A 210 nm ). For kinetic studies, the activities were assayed under standard assay conditions, except that different Ald concentrations were used. Esterase activities against 0.2 mm p-nitrophenylacetate (Nakarai tes- que, Kyoto, Japan) (C2 ester), 0.2 mm p-nitrophenylbutyrate (Sigma-Aldrich, MO, USA) (C4 ester) and 0.2 mm p-nitrophenyloctanoate (Wako Pure Chemical Industries, Ltd, Osaka, Japan) (C8 ester) were assayed [11,12]. Protein concentrations were assayed by the Lowry–Folin method. To compare the specific activities of Hyb-24 mutant enzymes with substitutions at position 370, the Ald-hydro- lytic activity in the crude enzyme solution was assayed by HPLC [13]. The amount of the Hyb-24 mutant protein included in the cell extract was quantified by densito- metric analysis of protein bands separated by SDS ⁄ PAGE using nih image analysis software (http://rsb.info. nih.gov/nih-image/) [13]. The specific activity was expressed as the Ald-hydrolytic activity ⁄ amount (mg) of Hyb-24- mutant protein. From the results obtained using cell extracts, the specific Ald-hydrolytic activity of wild-type NylB was estimated to be approximately 180-fold that exhibited by Hyb-24, and this value was almost the Y. Kawashima et al. Molecular design of nylon oligomer hydrolase FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS 2553 same with the ratio obtained from the purified enzymes (specific activity of His-tagged purified NylB/specific activity of His-tagged purified Hyb-24). In addition, no Ald-hydrolytic activity was detected in E. coli harboring the vector (without the NylB ⁄ NylB¢ region) (< 1% of the NylB¢ level of activity), and background esterolytic acti- vity was also quite low as compared to the activity in E. coli clones producing the NylB ⁄ NylB¢-related enzymes. Therefore, we have determined that the estimation based on data from the cell extracts roughly agrees with the results obtained using the purified enzyme. Crystallographic analysis The crystals of Hyb-24D and Hyb-24DNY were grown by the sitting-drop vapor-diffusion method from the protein buffer solution (10–20 mg protein mL )1 , 0.1 m Mes, pH 6.5) (Nakarai tesque, Kyoto, Japan) containing ammo- nium sulfate (2.0–2.2 m) (Nakarai tesque, Kyoto, Japan), lithium sulfate (0.1–0.2 m) (Wako Pure Chemical Industries, Ltd, Osaka, Japan) and glycerol [15–25% (v ⁄ v)] at 10 °C, to a final size of about 0.3 · 0.3 · 0.3 mm, according to the protocol used for Hyb-24 and Hyb-24DN [11,14]. The enzyme–substrate complex for the S112A mutant enzymes (Hyb-24D-A 112 and Hyb-24DNY-A 112 ) was prepared by soaking the crystals in the cryoprotectant [2.0 m ammonium sulfate, 30% (v ⁄ v) glycerol, and 0.1 m Mes buffer, pH 6.5] containing 100 m m substrate (Ald) for 3 h [12]. Diffraction data for the crystals were collected to 1.45–1.70 A ˚ resolu- tion as follows. Diffraction datasets of Hyb-24D were collected at 100 K using the beamline BL41XU (SPring-8, Hyogo, Japan) equipped with the MarCCD detector system. Diffraction datasets of Hyb-24DNY and Hyb-24DNY-A 112 –Ald were collected at 100 K using the beamline BL-5A (Photon Factory, Tsukuba, Japan) equipped with the ADSC Quantum 315r detector system. Diffraction datasets of Hyb-24D-A 112 –Ald were collected at 100 K using the beam- line BL38B1 (SPring-8, Hyogo, Japan) equipped with the Rikagaku Jupiter CCD detector system. Integration of the reflections was performed using the hkl2000 software pack- age [22]. Rigid-body refinement was performed using the coordinates of Hyb-24DN to fit the unit cell of the Hyb-24DNY and Hyb-24D protein crystals, followed by positional and B-factor refinement with cns [23]. The initial model was similarly obtained using the coordinates of Hyb-24DN-A 112 –Ald for protein crystals of Hyb-24D-A 112 – Ald and of Hyb-24DNY-A 112 –Ald. Several cycles of manual model rebuilding were performed by xfit [24]. Results of the crystal structure analysis are summarized in Table 4. The atomic coordinates and structure factors for Hyb- 24D (Protein Data Bank ID code: 2E8I), Hyb-24DNY (Protein Data Bank ID code: 2ZM0), Hyb-24D-A 112 –Ald (Protein Data Bank ID code: 2ZM7) and Hyb-24DNY- A 112 –Ald (Protein Data Bank ID code: 2ZMA) have Table 4. Data collection and refinement statistics. A Hyb-24D and Hyb-24D-A 112 –Ald Hyb-24D Hyb-24D-A 112 –Ald Data collection Space group P3 2 21 P3 2 21 Unit cell parameters a = b (A ˚ ) 96.69 96.68 c (A ˚ ) 112.93 113.16 Wavelength (A ˚ ) 0.8000 0.9000 Resolution (outer shell) (A ˚ ) 50–1.45 (1.50–1.45) 50–1.60 (1.66–1.60) Total reflections 1 152 434 882 657 Unique reflections (outer shell) 107 985 (10 619) 81 044 (7973) Completeness (outer shell) (%) 99.8 (99.0) 100.0 (99.9) R merge (outer shell) (%) a 8.2 (49.6) 6.3 (43.5) <I> ⁄ <r(I)> 25.5 (3.0) 38.5 (4.0) Refinement Resolution (outer shell) (A ˚ ) 41.9–1.45 (1.54–1.45) 31.7–1.60 (1.70–1.60) R work (outer shell) (%) 18.7 (26.0) 17.2 (21.5) R free (outer shell) (%) 19.9 (27.5) 18.7 (22.2) B Hyb-24DNY and Hyb-24DNY-A 112 –Ald Hyb-24DNY Hyb-24DNY-A 112 –Ald Data collection Space group P3 2 21 P3 2 21 Unit cell parameters a = b (A ˚ ) 96.66 96.69 c (A ˚ ) 112.94 112.91 Wavelength (A ˚ ) 1.0000 1.0000 Resolution (outer shell) (A ˚ ) 50–1.51 (1.56–1.51) 50–1.51 (1.56–1.51) Total reflections 1 037 212 549 789 Unique reflections (outer shell) 96 062 (9515) 96 045 (9326) Completeness (outer shell) (%) 99.9 (100) 99.5 (97.5) R merge (outer shell) (%) 5.0 (26.5) 6.3 (47.1) <I> ⁄ <r(I)> 68.3 (9.20) 31.7 (2.2) Refinement Resolution (outer shell) (A ˚ ) 46.8–1.51 (1.60–1.51) 33.6–1.51 (1.60–1.51) R work (outer shell) (%) a 18.6 (23.0) 17.5 (23.6) R free (outer shell) (%) b 19.9 (24.8) 19.0 (24.4) a R ¼ P hkl F obs kÀkF calc P hkl F obs jj ÀÁ       À1 , k: scaling factor. b R ¼ P hkl F obs k ÀkF calc P hkl F obs jj ÀÁ       À1 , k: scaling factor. Molecular design of nylon oligomer hydrolase Y. Kawashima et al. 2554 FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS been deposited in the Protein Data Bank (http://www. rcsb.org/). The structures of Hyb-24 (Protein Data Bank ID code: 1WYB) [11], Hyb-24DN (Protein Data Bank ID code: 1WYC) [12] and Hyb-24DN-A 112 –Ald (Protein Data Bank ID code: 2DCF) [12] have been previously reported. Figures of three-dimensional models of proteins were gener- ated with molfeat (v. 3.6; FiatLux Co., Tokyo, Japan). Acknowledgements This work was supported in part by a Grant-in-Aid for Scientific Research (Japan Society for Promotion of Science), and grants from the GCOE Program, the National Project on Protein Structural and Functional Analyses, the basic research programs CREST type, ‘Development of the Foundation for Nano-Interface Technology’ from JST and the JAXA project. References 1 Negoro S (2000) Biodegradation of nylon oligomers. Appl Microbiol Biotechnol 54, 461–466. 2 Negoro S (2002) Biodegradation of nylon and other synthetic polyamides. Biopolymers 9, 395–415. 3 Kinoshita S, Negoro S, Muramatsu M, Bisaria VS, Sawada S & Okada H (1977) 6-Aminohexanoic acid cyclic dimer hydrolase: a new cyclic amide hydrolase produced by Achromobacter guttatus KI72. Eur J Biochem 80, 489–495. 4 Kinoshita S, Terada T, Taniguchi T, Takene Y, Masu- da S, Matsunaga N & Okada H (1981) Purification and characterization of 6-aminohexanoic acid oligomer hydrolase of Flavobacterium sp. KI72. Eur J Biochem 116, 547–551. 5 Negoro S, Kakudo S, Urabe I & Okada H (1992) A new nylon oligomer degradation gene (nylC) on plasmid pOAD2 from Flavobacterium sp. J Bacteriol 174, 7948– 7953. 6 Kakudo S, Negoro S, Urabe I & Okada H (1993) Nylon oligomer degradation gene, nylC on plasmid pOAD2 from a Flavobacterium strain encodes endo-type 6-aminohexanoate oligomer hydrolase: purification and characterization of the nylC gene product. Appl Environ Microbiol 59, 3978–3980. 7 Kato K, Ohtsuki K, Koda Y, Maekawa T, Yomo T, Negoro S & Urabe I (1995) A plasmid encoding enzymes for nylon oligomer degradation: nucleotide sequence and analysis of pOAD2. Microbiology 141, 2585–2590. 8 Negoro S, Taniguchi T, Kanaoka M, Kimura H & Okada H (1983) Plasmid-determined enzymatic degradation of nylon oligomers. J Bacteriol 155, 22–31. 9 Okada H, Negoro S, Kimura H & Nakamura S (1983) Evolutionary adaptation of plasmid-encoded enzymes for degrading nylon oligomers. Nature 306, 203–206. 10 Kato K, Fujiyama K, Hatanaka HS, Prijambada ID, Negoro S, Urabe I & Okada H (1991) Amino acid alterations essential for increasing the catalytic activity of the nylon-oligomer degradation enzyme of Flavobac- terium sp. Eur J Biochem 200, 165–169. 11 Negoro S, Ohki T, Shibata N, Mizuno N, Wakitani Y, Tsurukame J, Matsumoto K, Kawamoto I, Takeo M & Higuchi Y (2005) X-ray crystallographic analysis of 6-aminohexanoate-dimer hydrolase: molecular basis for the birth of a nylon oligomer degrading enzyme. J Biol Chem 280, 39644–39652. 12 Negoro S, Ohki T, Shibata N, Sasa K, Hayashi H, Nakano H, Yasuhira K, Kato D, Takeo M & Higuchi Y (2007) Nylon-oligomer degrading enzyme ⁄ substrate complex: catalytic mechanism of 6-aminohexanoate- dimer hydrolase. J Mol Biol 370, 142–156. 13 Ohki T, Wakitani Y, Takeo M, Yasuhira K, Shibata N, Higuchi Y & Negoro S (2006) Mutational analysis of 6-aminohexanoate-dimer hydrolase: relationship between nylon oligomer hydrolytic and esterolytic activities. FEBS Lett 580, 5054–5058. 14 Ohki T, Mizuno N, Shibata N, Takeo M, Negoro S & Higuchi Y (2005) Crystallization and x-ray diffrac- tion analysis of 6-aminohexanoate-dimer hydrolase from Arthrobacter sp. KI72. Acta Crystallogr F61, 928–930. 15 Hatanaka HS, Fujiyama K, Negoro S, Urabe I & Okada H (1991) Alteration of catalytic function of 6-aminohexanoate-dimer hydrolase by site-directed mutagenesis. J Ferment Bioeng 71, 191–193. 16 Shimamura T, Ibuka A, Fushinobu S, Wakagi T, Ishiguro M, Ishii Y & Matsuzawa H (2002) Acyl-inter- mediate structures of the extended-spectrum class A b-lactamase, Toho-1, in complex with cefotaxime, cephalothin, and benzylpenicillin. J Biol Chem 277, 46601–46608. 17 Minasov G, Wang X & Shoichet BK (2002) An ultra- high resolution structure of TEM-1b-lactamase suggests a role for Glu166 as the general base in acylation. JAm Chem Soc 124, 5333–5340. 18 Golemi-Kotra D, Meroueh SO, Kim C, Vakulenko SB, Bulychev A, Stemmler AJ, Stemmler TL & Mobashery S (2004) The importance of a critical protonation state and the fate of the catalytic steps in class A b-lactamas- es and penicillin-binding proteins. J Biol Chem 279, 34665–34673. 19 Banerjee S, Pieper U, Kapadia G, Pannell LK & Herz- berg O (1998) Role of the omega-loop in the activity, substrate-specificity, and structure of class-A b-lactam- ase. Biochemistry 37, 3286–3296. 20 Hermann JC, Ridder L, Mulholland AJ & Holtje HD (2003) Identification of Glu166 as the general base in Y. Kawashima et al. Molecular design of nylon oligomer hydrolase FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS 2555 the acylation reaction of class A b-lactamases through QM ⁄ MM modeling. J Am Chem Soc 125, 9590–9591. 21 Ito W, Ishiguro H & Kurosawa Y (1991) A general method for introducing a series of mutations into cloned DNA using the polymerase chain reaction. Gene 102, 67–70. 22 Otwinowski Z & Minor W (1997) Processing of x-ray diffraction data collected in oscillation mode. Meth Enzymol 276, 307–326. 23 Brunger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges N, Pannu NS et al. (1998) Crystallography and NMR system (CNS): a new software system for macro- molecular structure determination. Acta Crystallogr D54, 905–921. 24 McRee DE (1993) Practical Protein Crystallography. Academic Press, San Diego, CA. Supporting information The following supplementary material is available: Fig. S1. Stereoview of the catalytic cleft of Hyb-24DN. Fig. S2. Stereoviews of surface structure of Ald-bound and unbound Hyb-24DNY. Fig. S3. Stereoview of the catalytic cleft of Hyb- 24DNY and class A b-lactamase (TEM-1). Table S1. Effect of amino acid alterations at position 370 in Hyb-24 on enzyme activity. This supplementary material can be found in the online version of this article. Please note: Wiley-Blackwell is not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corre- sponding author for the article. Molecular design of nylon oligomer hydrolase Y. Kawashima et al. 2556 FEBS Journal 276 (2009) 2547–2556 ª 2009 The Authors Journal compilation ª 2009 FEBS . Molecular design of a nylon-6 byproduct-degrading enzyme from a carboxylesterase with a b-lactamase fold Yasuyuki Kawashima 1, *, Taku Ohki 1, *, Naoki. interactions at the C-terminal region of Ald. Abbreviations Ahx, 6-aminohexanoate; Ald, 6-aminohexanoate linear dimer; DD, D-alanyl-D-alanine. FEBS Journal