Ligand Recognition in an

Subscriber access provided by Heriot-Watt | University Library

Article

Origin of Stereoselectivity and Substrate/ligand Recognition in an FAD-Dependent R-Selective Amine Oxidase Shogo Nakano, Kazuyuki Yasukawa, Takaki Tokiwa, Takeshi Ishikawa, Erika Ishitsubo, Naoya Matsuo, Sohei Ito, Hiroaki Tokiwa, and Yasuhisa Asano J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.6b09328 • Publication Date (Web): 20 Sep 2016 Downloaded from http://pubs.acs.org on September 26, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Origin of Stereoselectivity and Substrate/ligand Recognition in an FAD-Dependent R-Selective Amine Oxidase

Shogo Nakano1,2,3, Kazuyuki Yasukawa1,3, Takaki Tokiwa4†, Takeshi Ishikawa5, Erika Ishitsubo6, Naoya Matsuo6, Sohei Ito2, Hiroaki Tokiwa6,7*, and Yasuhisa Asano1,3*

1

Biotechnology Research Center and Department of Biotechnology, Toyama Prefectural University, 5180

Kurokawa, Imizu, Toyama 939-0398, Japan 2

School of Food and Nutritional Sciences, University of Shizuoka, 52-1 Yada, Suruga-ku, Shizuoka 422-8526,

Japan 3

Asano Active Enzyme Molecule Project, ERATO, JST, 5180 Kurokawa, Imizu, Toyama 939-0398, Japan

4

Research Center for Advanced Information Science and Technology, University of Aizu, iikimachi, Tsuruga,

Aizuwakamatsu, 965-8580, Japan 5

Department of Molecular Microbiology and Immunology, Graduate School of Biomedical Sciences, Nagasaki

University, 1-12-4 Sakamoto, Nagasaki, 852-8523, Japan 6

Department of Chemistry, Rikkyo University, Nishi-ikebukuro, Toshimaku, Tokyo 171-8501, Japan

7

Research Center of Smart Molecules, Rikkyo University, Nishi-ikebukuro, Toshimaku, Tokyo 171-8501, Japan

*

Correspondence to Yasuhisa Asano ([email protected]) and Hiroaki Tokiwa ([email protected]).

†

Present address: Department of Chemistry, Graduate School of Science, Tohoku University, Aramaki, Aoba-ku,

Sendai, Miyagi 980-8578, Japan

-1-

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT: Elucidation of the molecular mechanism of amine oxidases (AOx) will help to extend their reactivity by rational design and their application to deracemization of various amine compounds. To date, several studies have been performed on S-selective AOx, but relatively few have focused on R-selective AOx. In this study, we sought to elucidate the mechanism of pkAOx, an R-selective AOx that we designed by introducing the Y228L and R283G mutations into D-amino acid oxidase from pig kidney. Four crystal structures of the substrate-bound protein and first-principles calculations based on the correlated fragment molecular orbital (FMO) indicated that two aromatic residues, Tyr224 and Phe242, form stable π-π stacking interaction with substrates. Enzyme kinetics also supported the importance of Tyr224 in catalysis: the kcat/Km value of the Y224L mutant was reduced by 300-fold than that of wild-type (WT) when utilizing either (R)-methylbenzylamine [(R)-MBA] or (R)-1-(2-naphthyl)ethylamine [(R)-NEA] as the substrate. On the other hand, several Phe242 mutants exhibited higher reactivity toward (R)-NEA than the WT enzyme. In addition, FMO analysis indicated that pkAOx forms a ~ 13 kcal/mol more stable interaction with (R)-MBA than with (S)-MBA; this energy difference contributes to specific recognition of (R)-MBA in the racemate. Through the present study, we clarified three features of pkAOx: the roles of Tyr224 and Phe242 in catalysis, the origin of high stereoselectivity; and the potential to extend its reactivity toward amine compounds with bulky groups.

-2-


Page 2 of 23

Page 3 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

INTRODUCTION Synthesis of chiral compounds is important in industry, because many such compounds are precursors of industrially important chemicals 1. Although chiral compounds are currently obtained primarily by chemical synthesis, methods for enzymatic synthesis are also being developed 2. Enzymatic methods have several advantages over conventional approaches, including the fact that they are more eco-friendly 3, 4 and can yield enantiomerically pure compounds under normal conditions 5. On the other hand, several hurdles must be cleared prior to application of enzymes, including improvements in thermal stability, protein solubility, and reactivity

4, 6

. In recent years,

protein design methods have been developed to achieve such improvements. Consequently, enzymes have been increasingly applied to industrial synthesis 7. In the future, enzymatic synthesis may become the standard method of the synthesis. Various compounds containing chiral amines are industrially important, and several such molecules are used as precursors of pharmaceutical drugs 8. Several enzymes, including transaminases and amine oxidases, are currently being used to synthesize chiral amines

1, 9, 10

. Turner’s group reported several examples of applications of amine

oxidases, e.g., they designed an S-selective amine oxidase (S-AOx) by mutating FAD-dependent monoamine oxidase (MAO)

11

. (S)-methylbenzylamine [(S)-MBA] in the racemic state (rac-MBA) can be successfully

transformed into (R)-methylbenzylamine [(R)-MBA] by a combination of S-AOx catalysis and reaction with a chemical reductant such as NaBH4 11. The same group succeeded in extending the substrate specificity of MAO to amines containing bulky group, such as aminodiphenylmethane, as well as tertiary amines, through an approach that combined directed evolution and structure-based rational design 8, 12-14. Functional analysis and rational design have been extensively performed for S-AOx, but not for R-selective amine oxidases (R-AOx), in part because R-AOx have only been available for a short time. In recent years, two types of R-AOx were reported by distinct groups: the E350L/E352D variant of 6-hydroxy-D-nicotine oxidase by Turner and co-workers 15, and the Y228L/R283G variant of D-amino acid oxidase from pig kidney (pkAOx) by our group 16. These R-AOx have distinctive properties: the former has broad substrate specificity 15, whereas the latter has high specific activity toward a few amine compounds containing benzene rings, such as (R)-MBA and -3-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2-phenylpyrrolidine (PhPyr)

16

Page 4 of 23

. In order to perform rational design to extend their applications, the functions of

R-AOxs should be analyzed. In this study, we analyzed the enzymatic function of a designed variant of pkAOx. Like other DAAOs, pkAOx is an FAD-dependent enzyme. The reaction mechanism of pkAOx was predicted from the crystal structure of the (R)-MBA–bound protein 16. Specifically, this structure revealed that the mechanism of pkAOx is similar to that of other DAAOs: a reactive amine substrate binds to the active site in the anionic form 17, 18

, and the hydrogen on the Cα atom (the α-proton) is transferred as hydride ion to oxidized FAD 19, 20. In fact, the

α-proton of (R)-MBA is transferred to N5 atom of FAD

16

. Although the reaction mechanism was predicted

successfully, the substrate recognition mechanism remained unclear. As with previously studied DAAOs

21-23

,

elucidation of the latter mechanism should facilitate mutation of pkAOx to extend its substrate specificity. Such extended specificity would enable the application of pkAOx to deracemization of various amine compounds. In this study, we performed biochemical, structural, and computational analysis of pkAOx in order to elucidate its mechanism. Overall, we sought to clarify three features of pkAOx: the functional roles of two aromatic residues, Tyr224 and Phe242; the origin of the high stereoselectivity (>99%) of pkAOx when using rac-MBA as a substrate; and the extensibility, by rational design, of its substrate specificity to other amines containing bulky groups.

-4-


Page 5 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


MATERIALS & METHODS Overexpression and purification of pkAOx and their variants. pkAOx plasmids were transformed into E. coli strain JM109, and overexpression of pkAOx was achieved as previously described

16

. Cells were collected and

sonicated after suspension in buffer A (10 mM potassium phosphate [pH 8.0] and 0.2% 2-mercaptoethanol) containing 2% ammonium sulfate (AS). Supernatant was collected after centrifugation (22000 × g, 4°C, 35 min), and AS was added to 30%. The resultant precipitate was collected by centrifugation (22000 × g, 4°C, 30 min), and then solubilized in buffer A. The resultant solution was first applied to a DEAE-Toyopearl 650M column, and the flowthrough was collected. AS was added to the flowthrough at 5%, and the solution was applied to a Butyl-Toyopearl 650M column. The column was washed with buffer A containing 0.8% AS, and pkAOx was eluted with buffer A. Purity was confirmed by SDS-PAGE. Other variants were purified by an identical procedure. Crystallization and X-ray data collection. pkAOx was concentrated to > 30 mg/mL, and 17 µL of 3.5 M AS was mixed with 90 µL of the concentrated sample. The resultant precipitant was discarded by centrifugation, and the supernatant was utilized for crystallization. A 4 µL aliquot of the sample was mixed with 2 µL of reservoir solution for CrystalScreen condition 1-17 (30% [w/v] polyethylene glycol (PEG) 4000, 0.1M Tris-HCl pH 8.5, and 0.2 M Lithium sulfate) (Hampton Research, Aliso Viejo, CA, USA). The resultant crystals were utilized for X-ray data collection. pkAOx crystals were soaked in reservoir solution containing 20% [w/v] PEG400 and 10 mM each of ligands/substrates. For data collection from the apo form, no ligands/substrates were added to the reservoir solution. The soaking times were as follows: 3 min [(S)-MBA], 30 min (3-amino-1-phenylbutane, AmPB), and < 1 min (2-phenylpyrrolidine [PhPyr] and apo form). The crystals were mounted and flash-cooled under a nitrogen stream (-173°C). Diffraction data were collected using a Quantum 315 CCD detector installed in BL17A of the Photon Factory (PF, Tsukuba, Japan) for the apo and PhPyr-bound forms, and using a Rigaku Micro-Max007 CuKα rotating-anode X-ray generator and a Rigaku R-AXISVII image-plate detector for the (S)-MBA– and AmPB-bound forms. X-ray data analysis of all data was performed using identical procedures. Indexing, integration, and scaling were performed using HKL2000 and Scalepack 24. Initial phase was determined by Molrep

25

implemented in the

CCP4 program suite 26, using the (R)-MBA–bound structure of pkAOx (PDB ID: 3WGT) as the template. Model -5-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 23

building and structure refinement were executed by Coot 27. All structural figures in this study were prepared using PyMOL 28. Crystallographic and refinement parameters are provided in Table 1. Fragment molecular orbital calculation. The first- principles (total electronic) calculation based on the fragment molecular orbital (FMO) method was performed for two different pkAOx whose structures were experimentally determined and to which either of two enantiomers [(R)-MBA or (S)-MBA] was bound. The FMO method could reliably evaluate not only electrostatic (ionic) but also weak π-π stacking (dispersion) interaction energies between the protein and the ligand/substrate, using the electron-correlated (post-Hartree-Fock) procedure. pkAOx structures were protonated by Protonate3D implemented in MOE 29. Utilizing the Charmm27 force field, energy minimization was performed by imposing restraint on all atoms in the structures except for hydrogen atoms. The pkAOx protein was divided into one-residue fragments with cut-off points at the Cα carbon of each residue, as described previously 30, and the FAD molecule was divided into four fragments. (Details are provided in the Supplementary Information.) The FMO calculation was performed using the software package PAICS

31

at the resolution of the

identity approximation of the second-order Møller−Plesset perturbation theory (RI-MP2) with the correlation-consistent polarized valence double zeta basis set (cc-pVDZ). HF, vDW, and MP2 indicate electrostatic interactions, van der Waals dispersion, and the sum of these two interactions, respectively. In this study, interaction between ligand/substrate and protein molecule was analyzed by calculating inter-fragment interaction energies (IFIEs) under the FMO scheme. To avoid the basis set superposition error 32 in IFIEs, we adopted the counterpoise (CP) correction method 33. IFIEs with and without CP correction are shown in Supplementary Table 3. All IFIE values given in the main text are CP-corrected, and the IFIEs between ligand/substrate and protein molecule were colored manually in PyMOL. Assay of enzyme activity toward (R)-MBA and (R)-NEA. The activities of pkAOx and their variants were measured by quantifying the amount of peroxide produced by the oxidase reaction. The concentration of pkAOx was estimated using the extinction coefficient of FAD (ε450 = 11300 M-1 cm-1, 34): FAD contents for several variants were altered relative to pkAOx. Enzyme activity towards (R)-MBA and (R)-NEA was measured in reaction buffer (100 mM potassium phosphate [pH 8.0], 1.5 mM 4-aminoantipyrine, 2 mM phenol, 2 U/mg peroxidase from horseradish, and 0.1–15 mM (R)-MBA or (R)-NEA); the given concentrations reflect the final concentrations after -6-


Page 7 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


mixing 100 µL of pkAOx sample with 900 µL of the buffer. The buffer and sample were incubated at 30°C for 30 min. The reaction was started by mixing these solutions in a cuvette, which was placed immediately into a UV-Vis spectrometer (UV-1700 PharmaSpec, SHIMADZU), and the time-dependent change in the concentration of N-ethyl-N-(2-hydroxy-3-sulfopropyl)aniline (extinction coefficient at 505nm: 12700 M-1 cm-1 referred to the reference,

35

) was monitored by measuring absorbance at 505 nm over 1 min. Initial velocities at different

concentrations of substrates were calculated utilizing the ORIGIN software. The kinetic parameters were determined using the Hill equation and the non-linear least-squares method: all plots yielded sigmoidal curves.

-7-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 23

RESULTS Crystal structures of apo and substrate-bound forms of pkAOx. The substrate specificity of pkAOx for various amines was analyzed in a previous study

16

. These compounds could be classified as reactive or

non-reactive with respect to pkAOx: (R)-MBA and PhPyr were reactive, whereas (S)-MBA and AmPB were non-reactive

16

. In this study, reactive and non-reactive compounds are referred to substrates and ligands,

respectively. The ability to categorize amine compounds as substrates or ligands based on their chemical structure would enable us to engineer pkAOx to extend its substrate specificity. However, we could not find any general rules that accurately classify these compounds. Because knowledge of the reaction and ligand/substrate recognition mechanism of pkAOx would enable classification of these compounds, we sought to elucidate these mechanisms. As with other enzymes, a hydrogen-bond network, which is required for pkAOx reactivity, must be formed between pkAOx and its substrates

36, 37

. Determination of this network represents the first step toward elucidating

the mechanism. The network could be predicted based on comparisons among several structures of substrate-bound pkAOx. Ligand/substrate-bound structures of pkAOx were determined by X-ray crystallography. Active-site structures are shown in Figure 1 A–C: (S)-MBA (2.2 Å, Fig. 1A), AmPB (2.5 Å, Fig. 1B), and PhPyr (2.7 Å, Fig. 1C). The electron-density maps clearly indicated that these substrates bind to pkAOx. The (S)-MBA–bound structure indicates that our proposed model, which explains the lack of reactivity toward (S)-MBA 16, is correct: the aromatic ring of (S)-MBA is in the same position as that of (R)-MBA (Fig. 1A). The hydride on the Cα atom of (S)-MBA oriented towards the side of Y224; therefore, (S)-MBA cannot be oxidized by pkAOx. Next, we superimposed the four ligand/substrate–bound structures, including the (R)-MBA–bound structure (PDB ID: 3WGT) (Fig. 1D). All the ligands/substrates have similar binding modes: the benzene rings of the substrates are oriented toward the active-site pocket formed by two aromatic residues (Tyr224 and Phe242) and the xylene ring of FAD (Fig. 1D). The overall and active-site structures of pkAOx were barely altered upon binding of the ligands/substrates: the root-mean-square deviation value for Cα atoms is < 0.20 Å among the structures (Fig. 1D), and the active-site residues are in identical positions (Fig. 1D). These results suggested that pkAOx preferentially binds amine compounds with benzene rings without inducing a dynamic conformational change. -8-


Page 9 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Active-site structures of (S)-MBA– (cyan, A), AmPB- (magenta, B), and PhPyr- (orange, C) bound forms of pkAOx. The 2Fo-Fc electron-density map (blue mesh) is contoured at 0.8 σ. Superimposed structures of apo and substrate binding forms of pkAOx (D). The apo and (R)-MBA binding form are colored in gray and green, respectively. All figures are prepared in PyMOL.

Interaction energy analysis of (R)- and (S)-MBA–bound forms of pkAOx based on FMO method. Next, we quantitatively estimated interactions formed at the active site of pkAOx from IFIEs subjected to counterpoise (CP) -9-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 23

correction under the FMO scheme. IFIEs were analyzed for pkAOx bound to two enantiomers, (R)-MBA and (S)-MBA. Comparison of the sums of the IFIEs calculated by application of Hartree-Fock theory (HF-IFIEs) with IFIEs calculated by application of Møller−Plesset perturbation theory (MP2-IFIEs) suggested that pkAOx recognizes substrate primarily via dispersion interactions, with electrostatic interactions making a smaller contribution to the stabilization. In fact, the sum of HF-IFIEs of (R)-MBA binding structures was 13.8 kcal/mol, whereas the sum of MP2-IFIEs was -38.6 kcal/mol. A similar tendency was also observed in the (S)-MBA structure. Based on these results, we used the MP2-IFIEs to evaluate the dispersion interactions between pkAOx and each substrate. Next, we compared the MP2-IFIEs of the (R)-MBA–bound structure with those of the (S)-MBA–bound structure. Seven residues located within 3 Å of the substrates are shown in Fig. 2A. (R)-MBA binds more stable than (S)-MBA to pkAOx: the sum of MP2-IFIEs of the (R)-MBA–bound structure (-24.8 kcal/mol) is about 13 kcal/mol lower than that of (S)-MBA (-11.6 kcal/mol). Plots of the IFIEs for active-site residues are shown in Fig. 2B [(R)-MBA] and Fig. 2D [(S)-MBA], respectively. In all the molecules, FAD formed the most stable interaction with substrates: the MP2-IFIEs of (R)- and (S)-MBA were -10.16 and -3.73 kcal/mol, respectively (Supplementary Table 2). The residues around the substrates are colored to represent the magnitudes of their IFIE values: (R)- and (S)-MBA–bound structures are shown in Fig. 2C and 2E, respectively. Tyr224 and Phe242 formed strong interactions with substrates (Fig. 2C and 2E): the MP2- IFIE values of the (R)- and (S)-MBA binding forms were respectively -4.18 and -2.77 kcal/mol for Tyr224 and -2.62 and -3.03 kcal/mol for Phe242 (Supplementary Table 2). In both structures, dispersion interactions, mainly π-π stacking interactions involving Phe242, Tyr224 and FAD, make a major contribution in forming stable interactions with substrates. In fact, about 70% (-17.0 kcal/mol, (R)-MBA) to 80% (-9.53 kcal/mol, (S)-MBA) of the IFIEs are resulted from such interactions. This result implied that pkAOx preferentially recognizes substrates with benzene rings. Consistent with this, the reactive substrates of pkAOx are mostly amines containing aromatic rings 16.

- 10 -


Page 11 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Active-site residues of pkAOx positioned within 3 Å of the substrate (A). Bar graph of IFIEs for HF (blue) and MP2 (red) of active-site residues and FAD (B and D). IFIEs of (R)-MBA and (S)-MBA are shown in 2B and 2D, respectively. Representation of IFIEs for MP2 of (R)-MBA (C) and (S)-MBA (E), depending upon the magnitudes of the energy values. Negative and positive IFIEs are colored in red and blue, respectively. Residues forming strong interactions with substrate are colored in deep red. - 11 -



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 23

Enzyme kinetics analysis of pkAOx variants with (R)-MBA and (R)-NEA. Next, by utilizing two aromatic amines [(R)-MBA and (R)-NEA] as substrates, we estimated enzymatic properties of pkAOx and its variants by kinetics analysis. The variants were designed by mutating two aromatic residues, Tyr224 and Phe242. Four of the variants, F242I, F242L, F242V, and Y224L, could be obtained in soluble form at sufficient concentrations to estimate enzymatic parameters; in other cases, low protein solubility and depletion of FAD made it difficult to perform the assay. All assays were performed more than three times. First, the enzyme kinetics were estimated utilizing (R)-MBA as the substrate (Table 1). This mutation drastically reduced the turnover (kcat) value: the values for F242I, F242L, F242V, and Y224L variants were respectively 10-, 20-, 100- and 250-fold lower than that for pkAOx (Table 1). Relative activity (Fig. 3A) and enzyme efficiency (kcat/Km) were also reduced, dependent upon the decrease in kcat value (Table 1). These results, especially the decrease in kcat value, suggested that Figure 3. Relative activity of Tyr224 and Phe242 variants of pkAOx toward (R)-MBA (A) and (R)-NEA

the mutation increased the energy difference between the substrate–bound and transition states of

(B). The specific activity at 15 mM (R)-MBA and 10 mM (R)-NEA was utilized for calculating the relative activity. Specific activities of pkAOx(WT) toward (R)-MBA and (R)-NEA were 9.2 and 1.0 U/mg, respectively. All experiments were performed more than

the enzyme. A previous study showed that the kcat value is dependent on this difference; specifically, a large difference decreases the kcat value 38.

three times. Next, kinetics parameters were estimated utilizing (R)-NEA as substrate (Table 1). (R)-NEA has a bulkier group (a naphthalene ring) than (R)-MBA has. Interestingly, the kcat value was increased by mutation of Phe242 to non-aromatic residues: the value for the Phe242 variants was more than 3-fold higher than that for pkAOx (Table 1). Again, the relative activity of Phe242 variants was more - 12 -


Page 13 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


than 2-fold higher than that for pkAOx (Fig. 3B). This improvement for (R)-NEA was in contrast to the case of (R)-MBA. Consistent with this, the ratio of kcat/Km values (kcat/Km[(R)-NEA])/(kcat/Km[(R)-MBA]) (Table 1), indicated that, in comparison with the enzyme’s efficiency toward (R)-MBA, its efficiency toward (R)-NEA was improved by a decrease in the van der Waals volume of the residue 242 (Table 1). The formation of a larger cavity by the mutation enabled pkAOx to accommodate the naphthalene ring of (R)-NEA appropriately in the active site. On the other hand, for the Y224L variant, both relative activity (Fig. 3A and B) and kcat value (Table 1) were drastically decreased. The Km of Y224L could not be estimated because its initial velocity did not reach a maximum; thus, the Km value of Y224L is higher than those of other variants. We also confirmed that neither pkAOx nor any of its variants had activity toward (S)-MBA and (S)-NEA. This result implies that the reaction mechanisms of the variants are identical to that of pkAOx.

Table 1. Enzymatic properties of pkAOx and its variants using (R)-MBA and (R)-NEA as substrates (R)-methylbenzylamine (R-MBA) Variantsa

kcatb

Kmc

kcat/Km

mM

-1

(R)-1-(2-naphthyl)ethylamine (R-NEA) kcat

Km

kcat/Km

mM

-1

kcat/Km(R-NEA)/ kcat/Km(R-MBA)

s pkAOx

-1

7.4 ± 0.4

s

mM

-1

s

-1

s

mM

-1

7.0 ± 0.4

1.06

0.78 ± 0.06

1.5 ± 0.3

0.52

0.49

Mutation at Phe242 site pkAOx(F242I)

0.52 ± 0.03

7.2 ± 0.3

0.072

3.0 ± 0.2

3.6 ± 0.4

0.83

11.5

pkAOx(F242L)

0.24 ± 0.02

7.3 ± 0.6

0.033

2.6 ± 0.2

4.2 ± 0.5

0.62

18.8

pkAOx(F242V)

0.071 ± 0.006

12.1 ± 0.8

0.006

3.1 ± 0.8

9.5 ± 2.0

0.33

55.0

8.7 ± 0.6

0.003

n.d.

n.d.

< 0.001

< 0.33

Mutation at Tyr224 site pkAOx(Y224L) a

0.026 ± 0.02

Locations of mutated sites in the three-dimensional structure are shown in Figure 3A. Recombinant enzymes were

purified as described in “Experimental Procedures”.

- 13 -



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 23

DISCUSSION In this study, the catalytic mechanism of pkAOx is analyzed by using a combined X-ray crystallography, correlated FMO calculation, and enzyme kinetics approach. Based on the outcome, attempt will be made to address all the issues mentioned in the Introduction. First, we consider the functional roles of Tyr224 and Phe242 in pkAOx. The kinetics analysis revealed that these two aromatic residues play distinct roles in the function of pkAOx. Tyr224 is essential for both enzymatic reactivity and specificity, and therefore could not be mutated in the course of rational design. The IFIE analysis of all the amino-acid residues of pkAOx clearly showed that Tyr224 is the main contributor to stable interactions via a π-π stacking interaction. Furthermore, kinetic parameters indicated that Tyr224 is indispensable for amine oxidase activity. The importance of Tyr224 to the activity of pkAOx is in contrast to the case of DAAO, which retained its activity after mutation of Tyr224 39. Therefore, contrary to DAAO, the hydroxyl group of Tyr224 of pkAOx may contribute to the hydride translation mechanism by forming a hydrogen-bond interaction with the amino group of the substrate. On the other hand, Phe242 primarily regulates the substrate specificity of pkAOx. Thus, this residue could be mutated in order to change the enzyme’s specificity. Although reactivity towards (R)-MBA persisted following the mutation (Table 1), it was reduced when Phe242 was mutated to a residue with a low van der Waals volume, e.g., Val. Based on the IFIE analysis and kinetics parameters, it appears that Phe242 places (R)-MBA appropriately on the xylene ring of FAD via formation of dispersion interactions. In light of these results, two hypotheses could be made to explain this decrease in activity. First, the benzene ring of (R)-MBA could move into the newly generated cavity formed by the Phe242 mutation, e.g., in the F242V variant (arrowed direction in Fig. 4A); the mutation abolishes interactions necessary for the formation of the appropriate geometries with (R)-MBA. Second, (R)-MBA may bind less stably to the active site because the mutation abolishes the π-π stacking interaction. Next, we consider the origin of the high stereoselectivity of pkAOx. The reactivity of pkAOx on rac-MBA was not inhibited when (S)-MBA was in excess relative to (R)-MBA 16. pkAOx could achieve high stereoselectivity in two ways. First, the hydrogen atom on Cα (the α-proton) may have the optimal geometry in (R)-MBA but not in (S)-MBA. The orientation of the α-proton differs between the two substrates: it is directed towards Tyr224 side in - 14 -


Page 15 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the

(S)-MBA–bound

form,

but

towards

and

FAD

in

the

(R)-MBA–bound

form.

Second,

the

enzyme-ligand/substrate complex with (R)-MBA may be more stable than the complex containing (S)-MBA. Comparison of total IFIE values obtained for the structures of pkAOx bound to each enantiomer indicated that (R)-MBA could form a more stable enzyme–substrate complex than (S)-MBA; in other words, pkAOx preferentially recognizes (R)-MBA. From the analysis of IFIEs difference (Supporting Information Fig.1A-C and Supporting Table 2), in comparison to the (S)-MBA binding structure, the relative increase in stabilization for the (R)-MBA–bound structure is provided by FAD (-6.44 kcal/mol) and three amino-acid residues, Ala49 (-2.19 kcal/mol) and Pro54 (-1.50 kcal/mol) (where is the 3rd amino acid?). Thus, these residues may contribute to the high stereoselectivity of pkAOx. Here the large contributions of Ala49 and Pro54 were not considered because these residues did not form any apparent interactions in these structures (Fig. 2C and 2E). In order to achieve high stereoselectivity, recognition of the opposite enantiomer (e.g., (S)-MBA by pkAOx) must be eliminated because such a recognition sometimes inhibits the enzymatic reaction. The mutation of the sites may have little effect to the stereoselectivity because the change in stabilization energy may be compensated by structural changes around the mutation sites. Finally, whether or not the substrate specificity of pkAOx could be changed by mutation is considered. Based on the findings here, it appears that by mutating residues at the aromatic ring, the substrate specificity of pkAOx can be extended to amine compounds containing bulky groups (Fig. 4B, and Fig. 2 of SI, colored in orange) and Cβ sites (Fig. 4B and Supplementary Fig. 2B, colored in green). The aromatic-ring site is formed by six amino acid residues (Tyr55, Phe242, Gly283, and Gly312-Tyr314; Fig. 2A in the SI). Mutations of these sites, such as F242V, could effectively extend the enzyme’s specificity to substrates with bulky groups on the benzene ring, such as (R)-NEA. In addition, the Cβ atom of a substrate molecule is directed towards the Cβ site formed by four hydrophobic residues ([Leu51, Ile215, Ile228, and Ile230], Fig. 2B in SI). Mutation of these residues is expected to broaden the specificity of substrates with bulky group at the Cβ atom; however, this speculative conclusion made in the present study needs additional mutational studies for the confirmation. At present we are working in this direction to expand the reactivity of R-AOx by mutating the assigned two sites by amine compounds containing bulkier groups. At the same time, however, the large contribution of Tyr224, Phe242, and the xylene ring of FAD in - 15 -



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 23

stabilization restricts the range of potential reactive amine substrates, as all substrates must have an aromatic ring and an amine group at the Cα atom. Indeed, reactive substrates consist primarily of these types of amines

16

.

Mutation of other DAAOs may represent a promising approach for obtaining novel R-AOx with broad substrate specificity. Based on the enzyme kinetics of Phe242 variants, reactivity toward (R)-NEA is more tolerant to mutation than (R)-MBA. Therefore, for detection of R-AOx activity by screening, (R)-NEA may be a better substrate than (R)-MBA.

- 16 -


Page 17 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Phe242 mutation changes the interaction between pkAOx and (R)-MBA (A). The interaction of pkAOx (WT) (upper) and pkAOx (F242V) (A, right) was indicated. The optimal geometry is present in pkAOx (WT) (bottom), but broken in pkAOx (F242V) (bottom). Sites of pkAOx that could be mutated to extend substrate specificity (B). Positions of potential mutations in the aromatic ring and Cβ sites are shown as orange and green surfaces, respectively.

- 17 -



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Conclusion

2

In this study, we predicted origin of stereoselectivity of R-AOx by combinational approaches of X-ray

3

crystallography, FMO analysis and biochemical assay. Small energy difference of hydrophobic interaction

4

(including π-π stacking interaction) between R- and S-MBA binding forms in R-AOx appeared to be origin of

5

high stereoselectivity of R-MBA. The prediction could be led for the first time by the combinational

6

approach. In addition, we expect that extension of R-AOx reactivity toward other amine compounds could be

7

achieved based on our results.

8

- 18 -


Page 18 of 23

Page 19 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

Supporting Information Description

2

The Supporting Information is available free of charge on the ACS Publications website.

3

Crystallographic parameters, internal fragment interaction energies (IFIEs) for MP2 between

4

pkAOx and ligands, IFIEs for (R)- and (S)-MBA calculated with or without application of

5

counterpose correction, and the region interacting with the ligands (A, B) are included.

6 7

ACKNOWLEDGEMENTS

8

X-ray data were collected at the synchrotron facilities of the Photon Factory (PF) using beamline BL-17A

9

(proposal No. 2012U005). The authors are grateful to the beamline staff, and to Dr. Yusuke Yamada, for their

10

assistance with the experiments at PF. Computational parts are supported by Rikkyo SFR project, 2014−2016, and

11

MEXT Supported Program for the Strategic Research Foundation at Private Universities, 2013−2018. We also thank Dr.

12

Sundaram Arulmozhiraja, Rikkyo university for fruitful discussions and valuable suggestions.

13 14

- 19 -



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

3

References [1] Turner, N. J. (2010) Deracemisation methods, Curr. Opin. Chem. Biol. 14, 115-121.

4

[2] Reetz, M. T. (2011) Laboratory evolution of stereoselective enzymes: a prolific source of

5

catalysts for asymmetric reactions, Angew. Chem. Int. Ed. Engl. 50, 138-174. [3] Asano., H. G. a. Y. (2012) Introduction - Principle and historial landmarks of enzyme catalysis in organic synthesis., In Enzyme Catalysis in Organic Synthesis, pp 3-42, Springer Verlag. [4] Asano, Y. (2010) Tools for enzyme discovery, In Manual of Industrial Microbiology and

2

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

Biotechnology (Baltz, R. H. D., Julian E.; Demain, Arnold L., Ed.) 3rd ed., American Society for Microbiology, USA. [5] Dadashipour, M., and Asano, Y. (2011) Hydroxynitrile lyases: insights into biochemistry, discovery, and engineering, ACS Catal. 1, 1121-1149. [6] Asano, Y., Dadashipour, M., Yamazaki, M., Doi, N., and Komeda, H. (2011) Functional expression of a plant hydroxynitrile lyase in Escherichia coli by directed evolution: creation and characterization of highly in vivo soluble mutants, PEDS 24, 607-616. [7] Bornscheuer, U. T., Huisman, G. W., Kazlauskas, R. J., Lutz, S., Moore, J. C., and Robins, K. (2012) Engineering the third wave of biocatalysis, Nature 485, 185-194. [8] Ghislieri, D., Green, A. P., Pontini, M., Willies, S. C., Rowles, I., Frank, A., Grogan, G., and Turner, N. J. (2013) Engineering an enantioselective amine oxidase for the synthesis of pharmaceutical building blocks and alkaloid natural products, J. Am. Chem. Soc. 135, 10863-10869. [9] Koszelewski, D., Pressnitz, D., Clay, D., and Kroutil, W. (2009) Deracemization of mexiletine biocatalyzed by omega-transaminases, Org. Lett. 11, 4810-4812. [10] Kohls, H., Steffen-Munsberg, F., and Hohne, M. (2014) Recent achievements in developing the biocatalytic toolbox for chiral amine synthesis, Curr. Opin. Chem. Biol. 19, 180-192. [11] Alexeeva, M., Enright, A., Dawson, M. J., Mahmoudian, M., and Turner, N. J. (2002) Deracemization of alpha-methylbenzylamine using an enzyme obtained by in vitro evolution, Angew. Chem. Int. Ed. Engl. 41, 3177-3180. [12] Dunsmore, C. J., Carr, R., Fleming, T., and Turner, N. J. (2006) A chemo-enzymatic route to enantiomerically pure cyclic tertiary amines, J. Am. Chem. Soc. 128, 2224-2225. [13] Atkin, K. E., Reiss, R., Koehler, V., Bailey, K. R., Hart, S., Turkenburg, J. P., Turner, N. J., Brzozowski, A. M., and Grogan, G. (2008) The structure of monoamine oxidase from Aspergillus niger provides a molecular context for improvements in activity obtained by directed evolution, J. Mol. Biol. 384, 1218-1231. [14] Kohler, V., Bailey, K. R., Znabet, A., Raftery, J., Helliwell, M., and Turner, N. J. (2010) Enantioselective biocatalytic oxidative desymmetrization of substituted pyrrolidines, Angew. Chem. Int. Ed. Engl. 49, 2182-2184. [15] Heath, R. S., Pontini, M., Bechi, B., and Turner, N. J. (2014) Development of an R-selective - 20 -


Page 20 of 23

Page 21 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39


amine oxidase with broad substrate specificity and high enantioselectivity, ChemCatChem 6, 996-1002. [16] Yasukawa, K., Nakano, S., and Asano, Y. (2014) Tailoring D-amino acid oxidase from the pig kidney to R-stereoselective amine oxidase and its use in the deracemization of alpha-methylbenzylamine, Angew. Chem. Int. Ed. Engl. 53, 4428-4431. [17] Tilocca, A., Gamba, A., Vanoni, M. A., and Fois, E. (2002) First-principles molecular dynamics investigation of the D-amino acid oxidative half-reaction catalyzed by the flavoenzyme D-amino acid oxidase, Biochemistry 41, 14111-14121. [18] Fitzpatrick, P. F. (2010) Oxidation of amines by flavoproteins, Arch. Biochem. Biophys. 493, 13-25. [19] Mattevi, A., Vanoni, M. A., Todone, F., Rizzi, M., Teplyakov, A., Coda, A., Bolognesi, M., and Curti, B. (1996) Crystal structure of D-amino acid oxidase: a case of active site mirror-image convergent evolution with flavocytochrome b2, Proc. Natl. Acad. Sci. U S A 93, 7496-7501. [20] Umhau, S., Pollegioni, L., Molla, G., Diederichs, K., Welte, W., Pilone, M. S., and Ghisla, S. (2000) The x-ray structure of D-amino acid oxidase at very high resolution identifies the chemical mechanism of flavin-dependent substrate dehydrogenation, Proc. Natl. Acad. Sci. U S A 97, 12463-12468. [21] Caligiuri, A., D'Arrigo, P., Rosini, E., Tessaro, D., Molla, G., Servi, S., and Pollegioni, L. (2006) Enzymatic conversion of unnatural amino acids by yeast D-amino acid oxidase, Adv. Synth. Catal. 348, 2183-2190. [22] Wong, K. S., Fong, W. P., and Tsang, P. W. (2010) A single Phe54Tyr substitution improves the catalytic activity and thermostability of Trigonopsis variabilis D-amino acid oxidase, N Biotechnol. 27, 78-84. [23] Pollegioni, L., and Molla, G. (2011) New biotech applications from evolved D-amino acid oxidases, Trends Biotechnol. 29, 276-283. [24] Otwinowski, Z., and Minor, W. (1997) Processing of X-ray diffraction data collected in oscillation mode, Macromolecular Crystallography, Pt A 276, 307-326. [25] Vagin, A., and Teplyakov, A. (1997) MOLREP: an automated program for molecular replacement, J. of Appl. Crystallogr. 30, 1022-1025. [26] Winn, M. D. (2003) An overview of the CCP4 project in protein crystallography: an example of a collaborative project, J. Synchrotron Radiat. 10, 23-25. [27] Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010) Features and development of Coot, Acta Cryst. D 66, 486-501. [28] Delano, W., L. (2002) The PyMOL molecular graphics system, Ed.).

(Delano Scientific, S. C., CA.,

[29] Molecular Operating Environment (MOE), 2012.10, (Inc., C. C. G., Ed.), Montreal, QC, Canada. [30] Ohtake, K., Yamaguchi, A., Mukai, T., Kashimura, H., Hirano, N., Haruki, M., Kohashi, S., Yamagishi, K., Murayama, K., Tomabechi, Y., et al. (2015) Protein stabilization utilizing a - 21 -



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

redefined codon, Sci. Rep. 5, 9762. [31] Ishikawa, T., and Kuwata, K. (2009) Fragment molecular orbital calculation using the RI-MP2 method, Chem. Phys. Lett. 474, 195-198. [32] Kamiya, M., Hirata, S., and Valiev, M. (2008) Fast electron correlation methods for molecular clusters without basis set superposition errors, J. Chem. Phys. 128, 074103. [33] Valiron, P., and Mayer, I. (1997) Hierarchy of counterpoise corrections for N-body clusters: generalization of the Boys-Bernardi scheme, Chem. Phys. Lett. 275, 46-55. [34] Harris, C. M., Molla, G., Pilone, M. S., and Pollegioni, L. (1999) Studies on the reaction mechanism of Rhodotorula gracilis D-amino-acid oxidase. Role of the highly conserved Tyr-223 on substrate binding and catalysis, J. Biol. Chem. 274, 36233-36240. [35] Tamaoku, K., Ueno, K., Akiura, K., and Ohkura, Y. (1982) New water-soluble hydrogen donors for the enzymatic photometric determination of hydrogen peroxide. II. N-ethyl-N-(2-hydroxy-3-sulfopropyl)aniline derivatives, Chem. Pharm. Bull. 30, 2492-2497. [36] Wang, L., Fried, S. D., Boxer, S. G., and Markland, T. E. (2014) Quantum delocalization of protons in the hydrogen-bond network of an enzyme active site, Proc. Natl. Acad. Sci. U S A 111, 18454-18459. [37] Katona, G., Carpentier, P., Niviere, V., Amara, P., Adam, V., Ohana, J., Tsanov, N., and Bourgeois, D. (2007) Raman-assisted crystallography reveals end-on peroxide intermediates in a nonheme iron enzyme, Science 316, 449-453. [38] Babtie, A., Tokuriki, N., and Hollfelder, F. (2010) What makes an enzyme promiscuous?, Curr. Opin. Chem. Biol. 14, 200-207. [39] Setoyama, C., Nishina, Y., Mizutani, H., Miyahara, I., Hirotsu, K., Kamiya, N., Shiga, K., and Miura, R. (2006) Engineering the substrate specificity of porcine kidney D-amino acid oxidase by mutagenesis of the "active-site lid", J. Biochem. 139, 873-879.

25 26

- 22 -


Page 22 of 23

Page 23 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents Graphic 1

- 23 -


Ligand Recognition in an

Recommend Documents