Structural Determinants of the Stereoinverting Activity of

Subscriber access provided by EKU Libraries

Article

Structural determinants of the stereo-inverting activity of Pseudomonas stutzeri D-phenylglycine aminotransferase Curtis J Walton, Frédéric Thiebaut, Joseph S Brunzelle, Jean-François Couture, and Roberto A. Chica Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.8b00767 • Publication Date (Web): 28 Aug 2018 Downloaded from http://pubs.acs.org on August 30, 2018

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 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 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.

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

Biochemistry

Structural determinants of the stereo-inverting activity of Pseudomonas stutzeri D-phenylglycine

aminotransferase

Curtis J.W. Walton,1,2 Frédéric Thiebaut,1,2 Joseph S. Brunzelle,3 Jean-François Couture*,2,4 & Roberto A. Chica*,1,2 1

Department of Chemistry and Biomolecular Sciences, University of Ottawa, Ottawa, Ontario, Canada,

K1N 6N5 2

Centre for Catalysis Research and Innovation, University of Ottawa, Ottawa, Ontario, Canada, K1N 6N5

3

Northwestern Synchrotron Research Centers, Life Science Collaborative Access Team, Northwestern

University, Evanston, IL, USA 4

Department of Biochemistry, Microbiology and Immunology, University of Ottawa, Ottawa, Ontario,

Canada

* To whom correspondence should be addressed: Roberto A. Chica, Email: [email protected] Jean-François Couture, Email: [email protected]

Roberto A. Chica, Ph.D. Associate Professor Department of Chemistry & Biomolecular Sciences University of Ottawa 10 Marie-Curie Ottawa, Ontario K1N 6N5 Canada Phone: (613) 562-5800 x1988 Fax: (613) 562-5170 Email: [email protected]

Keywords: structure-function, enzyme kinetics, enzyme mutation, enzyme structure, crystal structure, substrate specificity, X-ray crystallography, biocatalysis

1

ACS Paragon Plus Environment

Biochemistry 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

Aromatic D-amino acids are key precursors for the production of many small molecule therapeutics. Therefore, the development of biocatalytic methods for their synthesis is of great interest. An enzyme that has high potential as a biocatalyst for the synthesis of D-amino acids is the stereo-inverting D-phenylglycine aminotransferase (DPAT) from Pseudomonas stutzeri ST-201. This enzyme catalyzes a unique L to D transamination reaction that produces D-phenylglycine and α-ketoglutarate from benzoylformate and Lglutamate, via a mechanism that is poorly understood. Here, we present the crystal structure of DPAT, which shows that the enzyme folds into a two-domain structure representative of class III aminotransferases. Guided by the crystal structure, we performed saturation mutagenesis to probe the substrate binding pockets of the enzyme. These experiments helped us identify two arginine residues (R34 and R407), one in each binding pocket, that are essential to catalysis. Together with kinetic analyses using a library of amino acid substrates, our mutagenesis and structural studies allow us to propose a binding model that explains the dual L/D specificity of DPAT. Our kinetic analyses also demonstrate that DPAT can catalyze the transamination of β- and γ-amino acids, reclassifying this enzyme as an ω-aminotransferase. Collectively, our studies highlight that the DPAT active site is amenable to protein engineering for expansion of its substrate scope, which offers the opportunity to generate new biocatalysts for the synthesis of a variety of valuable optically-pure D-amino acids from inexpensive and abundant L-amino acids.

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

Biochemistry

Introduction D-amino

acids are important building blocks in the synthesis of many pharmaceuticals1 and

bioactive peptides2 used for the treatment of microbial infections,3 hypertension,4 and cancer.5 The production of optically-pure D-amino acids by biocatalytic approaches is a valuable alternative to traditional chemical synthesis as these enzyme-based methods can increase process efficiency by catalyzing multiple steps in one-pot reactions without requiring intermediate work-ups, cofactor recycling, or toxic metals.6,7 Recently, we developed a biocatalytic cascade to produce unnatural aromatic D-amino acids that couples L-amino acid deaminase (LAAD) with an engineered D-amino acid aminotransferase (DAAT) variant that displays high activity towards D-phenylalanine.8 Although this method enables the synthesis of various D-phenylalanine derivatives with high conversion rates and enantiomeric excesses, it requires a sacrificial D-amino acid donor substrate to yield the desired D-amino acid product because of the strict stereospecificity of DAAT.9,10 To increase the efficiency of our LAAD/DAAT biocatalytic cascade, a biocatalyst capable of performing a stereo-inverting transamination reaction to produce a desired D-amino acid product from an inexpensive and abundant L-amino acid sacrificial amine donor is highly desirable. D-phenylglycine

aminotransferase (DPAT) is, to our knowledge, the only presently known L to D

stereo-inverting aminotransferase.11,12 DPAT is a pyridoxal phosphate (PLP) dependent enzyme that catalyzes the reversible transamination of benzoylformate or p-hydroxybenzoylformate with L-glutamate, yielding α-ketoglutarate and D-phenylglycine or D-4-hydroxyphenylglycine (Fig. 1). Although the DPAT from the soil bacterium Pseudomonas stutzeri has been used in several biocatalytic applications,13–15 its strict specificity towards the D-phenylglycine/D-4-hydroxyphenylglycine and α-ketoglutarate substrates11 limits its application for the synthesis of a broad range of valuable D-amino acids. In addition, the mechanism by which DPAT can catalyze its stereo-inverting transamination reaction is unknown, and no detailed structural analysis has been published on the enzyme, although an incomplete crystal structure has been deposited in the Protein Data Bank (PDB ID: 2CY8).16 To facilitate the engineering of DPAT variants with expanded substrate scope, which would open the door to a broad range of biocatalytic applications, a better understanding of the structural determinants of its unique stereo-inverting activity is necessary.

Figure 1. Stereo-inverting transamination reaction catalyzed by DPAT.

3


Biochemistry 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

Herein, we report the crystal structure of Pseudomonas stutzeri ST-201 DPAT, which we used to identify active-site residues involved in substrate specificity. Saturation mutagenesis of these residues coupled to kinetic characterization of mutants with a library of amine donor substrates enabled us to propose a mechanism by which DPAT can achieve its dual L/D substrate specificity. The results presented here shed light on the unique stereo-inverting activity of this enzyme and open the door to its engineering in order to produce variants capable of synthesizing a variety of valuable D-amino acids from inexpensive and abundant L-amino acids.

Materials and Methods

Materials. All reagents used were of the highest available purity. Restriction enzymes and DNA modifying enzymes were obtained from New England Biolabs. Synthetic oligonucleotides were obtained from Integrated DNA Technologies, and Ni-NTA resin was obtained from Bio-Rad. All aqueous solutions were prepared using water purified with a Barnstead Nanopure Diamond system. The enzyme substrates and cofactors were purchased from Sigma-Aldrich with the exception of 2-methyl-D-phenylalanine (Alfa Aesar, >94%), 3-nitro-D-phenylalanine (Alfa Aesar, >94%), 4-methyl-D-phenylalanine (Alfa Aesar, >97%), 4methoxy-D-phenylalanine (Santa Cruz Biotechnology, >94%), (S)-3-phenyl-β-alanine (Toronto Research Chemicals, >95%), and (S)-4-phenyl-4-aminobutyrate (J & W PharmLab LLC, 97%).

DPAT gene cloning. The codon-optimized (E. coli) gene of Pseudomonas stutzeri ST-201 DPAT (GenBank ID: AY319935.1) was purchased from Genscript and subcloned into the pET11-a (N-terminal his-tag) or pCDFDuet-1 (C-terminal TEV protease cleavage site followed by his-tag) expression vectors (Novagen) via the NdeI/BamHI or NdeI/XhoI restrictions sites, respectively. The resulting plasmids were then transformed into chemically-competent E. coli XL-1 Blue cells (Stratagene). The entire reading frame of each plasmid was verified by DNA sequencing, after which the plasmids were transformed into chemicallycompetent E. coli BL21-Gold (DE3) cells (Stratagene) for protein expression. Mutagenesis. Mutations were introduced into the DPAT gene by overlap extension mutagenesis17 using VentR DNA Polymerase. Briefly, external primers were used in combination with sets of complementary pairs of oligonucleotides containing the mutated codon in individual polymerase chain reactions (PCRs). The resulting overlapping fragments were gel-purified (Omega Biotek) and recombined by overlap extension PCR. The resulting amplicons were digested with NdeI/BamHI, gel-purified, and ligated into pET11-a expression vector with T4 DNA ligase. Constructs were verified by sequencing the entire open

4


Page 4 of 23

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

Biochemistry

reading frame, and were transformed into chemically-competent E. coli BL21-Gold (DE3) cells (Stratagene) for protein expression.

Protein expression and purification. Proteins were expressed in 1 L cultures of E. coli BL21-Gold (DE3) cells transformed with a pET11-a or pCDFDuet-1 vector containing the DPAT gene. When the cultures reached an optical density at 600 nm of 0.6, the temperature was reduced from 37 to 15 °C. Upon reaching the induction temperature, 0.1 mM of isopropyl β-D-1-thiogalactopyranoside was added to the flasks to induce protein expression and the cells were incubated with shaking for 18 hours. Following expression, cells were harvested by centrifugation and lysed with an EmulsiFlex-B15 cell disruptor (Avestin). The proteins were then extracted and purified by immobilized metal affinity chromatography (IMAC), according to manufacturer's protocol. Elution fractions containing the aminotransferases were desalted by gel filtration using EconoPAC 10DG columns (Bio-Rad) into a final buffer solution of 100 mM potassium phosphate (pH 8) or 100 mM N-cyclohexyl-2-aminoethanesulfonic acid (CHES) buffer (pH 9) for samples used in crystallography or enzyme assays, respectively. Protein concentrations were quantified using a modified version of the Bradford assay, where the calibration curve is constructed as a plot of the ratio of the absorbance measurements at 590 nm and 450 nm versus concentration.18 This procedure yielded 12 ± 5 mg L‒1 of wild-type DPAT, and approximately 8 mg L–1 of each of the Q301 mutants.

Protein crystallization. The initial enzyme purification was performed as described above. Additional steps were taken to increase the purity of the wild-type DPAT sample for crystallography. The IMAC purified sample was dialyzed in 4 L of 100 mM potassium phosphate buffer (pH 8), 0.5 mM EDTA, and 1 mM dithiothreitol, overnight at 4 °C with stirring in the presence of TEV protease to cleave the C-terminal histag. Following cleavage of the his-tag, the sample was centrifuged (3000 × g, 30 minutes, 4 °C) to remove any precipitate. Finally, the protein was purified on a Bio-Rad Biologic Duoflow system using a Bio-Rad Enrich SEC 650 size exclusion column with a flow rate of 0.5 mL min–1 of 100 mM Tris-Cl Buffer (pH 7.4) supplemented with 100 mM sodium chloride. Fractions containing the enzyme were pooled and concentrated to 10 mg mL–1 using Amicon Ultra centrifugal filters. Trigonal crystals were grown in 0.2 M sodium phosphate buffer, pH 6.2 or 6.6, with 0.2–1.3 M ammonium sulfate using vapour diffusion. The crystals grew in 2 to 3 weeks and were transferred to a 35% cryoprotection solution of ethylene glycol with mother liquor. Crystals were subsequently flash-frozen with liquid nitrogen.

Crystal data collection and processing. A full data set was collected at the 17-ID-D beamline of LS-CAT (Advanced Photon Source, Argonne National Laboratories). The data set was indexed with XDS19 and scaled using Aimless.20 A molecular replacement solution was found with Phaser21 using the available

5


Biochemistry 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

DPAT structure as a search model (PDB ID: 2CY8). The model was completed using iterative rounds of refinement and model building using PHENIX22 and Coot,23 respectively. Quality of the models was assessed using MolProbity.24

Preparation of clarified lysate. DNA libraries of saturation mutants were transformed into chemically competent E. coli BL21-Gold (DE3) cells (Stratagene). Colonies were picked into individual wells of V96 MicroWell polypropylene plates (Nunc) containing 200 µL of medium (Luria-Bertani with 100 µg mL–1 ampicillin supplemented with 10% glycerol). The plates were covered with a sterile, breathable rayon membrane (VWR) and incubated overnight at 37 °C with shaking. After incubation, these mother plates were used to inoculate sterile Nunc V96 MicroWell polypropylene plates (“daughter” plates) containing 300 µL of Overnight Express Instant TB media (Novagen) supplemented with ampicillin. Daughter plates were sealed with breathable membranes and incubated overnight (37 °C, 250 rpm). Cells were harvested by centrifugation (3000 × g, 30 minutes, 4 °C) and the resulting pellets were washed twice with phosphate buffered saline pH 7.4. Washed cell pellets were resuspended in lysis buffer (100 mM CHES buffer, pH 9, containing 1× Bug Buster Protein Extraction Reagent (Novagen), 25 U mL–1 Benzonase Nuclease (EMD), and 1 mg mL–1 lysozyme). The clarified lysate was collected following centrifugation and stored at 4 °C until used in the screening assay.

Mutant library screening assay. The screening assay is fully described in reference

25

. Briefly, 200-µL

reactions at 37 °C in 100 mM CHES buffer (pH 9) were prepared. The reaction mixture contained 16 µM pyridoxal phosphate, 5 mM D-phenylglycine, 1 mM α-ketoglutarate, 1 U of glutamate dehydrogenase from bovine liver (Sigma), and 0.5 mM NAD+. Reactions were initiated by addition of 10 µL clarified cell lysate prepared as described earlier, and monitored by measuring the absorbance of NADH at 340 nm (extinction coefficient = 6220 M–1 cm–1) for 30 minutes in individual wells of flat-bottom 96-well plates (Greiner BioOne) using an Infinite M1000 plate reader (Tecan).

Substrate specificity profiles. The substrate specificity profile for each DPAT variant was determined as described in the previous paragraph with the following modifications. The reaction mixture contained 10 mM amine donor. Reactions were initiated using 2 mU of purified enzyme (activity determined using 5 mM of D-phenylglycine, 1 mM α-ketoglutarate, 1 U of glutamate dehydrogenase from bovine liver, and 0.5 mM NAD+ in 100 mM CHES buffer, pH 9, 37 °C). Negative controls containing all reaction components except DPAT were used. Solution mixtures were pre-incubated at 37 °C for 30 minutes prior to initiating the reaction by the addition of DPAT. The enzymatic reaction was monitored as described above and

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

Biochemistry

performed in triplicate from a single enzyme preparation. The path lengths for each well were determined ratiometrically using the difference in absorbance between 900 and 998 nm.

Steady-state kinetics. Kinetic assays were performed as the substrate specificity profile assays but with the following modifications. The concentration of the substrate being analyzed was varied and approximately 2 mU of purified enzyme was added to each well to initiate the reaction. For all blanks, the aminotransferase was substituted for 100 mM CHES buffer, pH 9. The reactions were monitored by measuring the absorbance of NADH at 340 nm for 30 minutes in individual wells of 96-well plates (Greiner Bio-One) using an Infinite M1000 plate reader (Tecan). Nonlinear regression analysis of the initial rates as a function of substrate concentrations fit to the Michaelis-Menten equation was performed with Graphpad Prism software. For fitting of the kinetic data obtained using the (S)-4-phenyl-4-aminobutyrate substrate, a modified rate equation that takes into account substrate inhibition was used: vo = vmax [S]/(KM + [S] + [S]2/ Ki).

Results

Crystal structure. To investigate the structural causes of the stereo-inverting activity of DPAT, we solved the crystal structure of the enzyme from P. stutzeri ST-20111 (UniProtKB Q6VY99) to a resolution of 1.82 Å, which yielded a single polypeptide chain in the asymmetric unit (data collection and refinement statistics are summarized in Table 1). In solution, this enzyme forms a homodimer comprising two 49-kDa subunits of 453 amino acids.11 Analysis of the DPAT dimer structure demonstrates that it belongs to the fold-type I group of PLP-dependent enzymes,26 and that it is a class III aminotransferase (Fig. 2A).27

7


Biochemistry 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

Table 1. Data Collection and Refinement Statistics for Pseudomonas stutzeri ST-201 DPAT Data Collection Space group Cell dimensions Unit cell (Å) Resolution (Å) Rmeas I / σI Completeness (%) Redundancy Refinement Resolution (Å) No. unique reflections Rwork / Rfree No. Atoms Protein Sulfate/Ethylene Glycol/Tris/Phosphate/Acetate Water B-factors (Å2) Protein Sulfate/Ethylene Glycol/Tris/Phosphate/Acetate Water Rmsds Bond lengths (Å) Bond angles (°) Molprobity scores Ramachandran favored (%) Ramachandran allowed (%) PDB ID Values in () are for the last resolution shell.

P 3 12 1 a = 74.64, b = 74.64, c = 149.95 49.9 – 1.82 2.4 (71.1) 78.4 (3.6) 99.8 (100) 8.5 (9.3) 49.9 – 1.82 44,043 17.9/22.2 3238 25/44/8/10/4 263 31.6 47.4/40.4/62.9/39.8/42.6 40.4 0.006 0.754 1.47 96.6 3.4 6DVS

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

Biochemistry

Figure 2. DPAT structure. (a) DPAT dimer with one active site per monomer (indicated by circle). Two phosphate ions (shown as sticks) are bound at the dimer interface. Spheres indicate chain breaks caused by disordered regions. (b) The DPAT subunit comprises a large PLP binding domain (blue) and a smaller NC domain, which is divided into N- (light green) and C-terminal (dark green) lobes. Chain breaks caused by the disordered regions are indicated by spheres, with the residue number of adjacent amino acids indicated. The N and C termini are also indicated. (c) Topology diagram of the DPAT structure. Helices, strands, and loops are represented by cylinders, arrows, and black lines, respectively. Numbers indicate N- and Cterminal residues of each structural element. The disordered regions are indicated with thick red lines. Dashed lines are used when two lines intersect. The N- and C-terminal lobes of the NC domain, and the PLP-binding domain are colored light green, dark green, or blue, respectively. The DPAT subunit consists of a large and a small domain (Fig. 2B).27 The large domain (residues 75–324) is the PLP-binding domain, which forms a three-layer α/β/α sandwich comprised of nine α-helices surrounding a central seven-stranded mixed β-sheet (strands β4–β10–β9–β8–β7–β5–β6), with β10 being the sole anti-parallel strand (Fig. 2C). The small domain, also called the NC domain because it is divided into N- and C-terminal lobes (residues 1–74 and 325–453), consists of two three-stranded anti-parallel β-sheets surrounded by two or four α-helices, respectively. The N-terminal lobe is formed by β1–3, which is capped by helices α1 and α2, while the C-terminal lobe is formed by β11–13, which are surrounded by helices α12–15. In contrast to the deposited 2CY8 structure, our DPAT structure also contains a phosphate ion bound at the dimer interface (Fig. 2A), close to where the phosphate group of PLP binds in other class III aminotransferases.28,29 This phosphate ion is coordinated by three water molecules, the side chains of S121 and T123, and the backbones of G122 and T123. The phosphate is also located close to catalytic residue K269, with a distance of 3.9 Å between the lysine Nε and phosphate oxygen atoms, which is consistent with a weak salt bridge interaction.30 9


Biochemistry 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

In its apo state, the DPAT structure contains four disordered regions for which no electronic density is observed (Fig. 2): residues 28–36, 150–176, 292–302, and the C-terminus (445–453). Together, these disordered regions represent approximately 12% of the total polypeptide chain. Interestingly, the three disordered regions closest to the N-terminus are located in or near the enzyme cleft and are expected to contribute residues to the active site (Fig. 2A). To gain a better understanding of the structural features of the active site, we aimed to solve the structure of the holo enzyme with bound substrate. Although the apo enzyme crystallized in a matter of weeks to form trigonal crystals, we were unable to find conditions that yielded crystals of the holo enzyme bound to the benzoylformate or α-ketoglutarate substrates, and soaking experiments did not yield density for the cofactor or substrates in the density maps. In the absence of the holo enzyme structure with bound substrates, we compared DPAT to homologous enzymes to gain insights on active-site organization and possible substrate binding modes.

Comparison with homologous enzymes. We performed sequence and structural alignments of DPAT using PDBeFold31 to identify similar structures in the Protein Data Bank (Table S1). The highest scoring, and thereby most structurally similar structure, was that of glutamate-1-semialdehyde-2,1-aminomutase (GSAM, UniProtKB P24630) from Synechococcus elongatus (PDB ID: 2HOY; Cα RMSD of 1.56 Å; 28% sequence identity; Z-score = 18.3).32 This enzyme catalyzes the isomerization of glutamate-1-semialdehyde to the non-proteinogenic amino acid 5-aminolevulinate. The highest scoring aminotransferase is β-phenylalanine aminotransferase (MesAT, UniProtKB A3EYF7) from Mesorhizobium sp. LUK (PDB ID: 2YKX; Cα RMSD of 1.82 Å; 30% sequence identity; Z-score = 17.1),28 which catalyzes the reversible transamination of α-ketoglutarate with (S)-3-phenyl-β-alanine. Like DPAT, these homologous enzymes are classified as PLP-dependent fold-type I and class III aminotransferases, and share many common structural (Fig. S1 and S2) and sequence features (Fig. 3).33–36

10


Page 10 of 23

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

Biochemistry

Figure 3. Sequence alignment of DPAT with homologous enzymes. Fully and partially conserved residues are indicated by white letters on a red background or red letters, respectively. Red lines indicate disordered regions for which no electronic density was observed in the apo state. Colored lines below alignment indicate the NC domain (green) and the PLP-binding domain (blue). The catalytic lysine is indicated by a star. O- and P-pocket residues that participate in substrate binding are indicated by circles or squares, respectively. The sequence alignment was generated using ESPript 3.0 (http://espript.ibcp.fr) 36.

11


Biochemistry 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

Substrate binding pockets. Aminotransferases have evolved ways to accommodate two substrates with side chains of different shapes and properties at the same active site while discriminating against other molecules. Molecular recognition by aminotransferases is typically controlled by a carboxylate binding pocket, which binds the α-carboxylate group of amino acid substrates, and a side-chain binding pocket, which can bind both the charged γ-carboxylate side chain of L-glutamate and a neutral or charged side chain of another amino-acid substrate. Dual substrate recognition by the side-chain binding pocket is commonly achieved through one of two strategies: an arginine switch or hydrogen-bond network rearrangement.37 In the case of DPAT, the enzyme also needs to achieve the L to D stereo-inversion to produce Dphenylglycine. Previously, the C-4′ proton transfer required for transamination of D-phenylglycine by DPAT was shown to occur exclusively on the si-face of the PLP cofactor.38 Therefore, to achieve the stereo-inversion, D-phenylglycine must bind in an inverted orientation relative to L-glutamate. In this orientation, the phenyl side-chain of D-phenylglycine would be bound in the carboxylate pocket that binds the α-carboxylate group of L-glutamate, while the α-carboxylate of D-phenylglycine would be bound in the side-chain pocket, where the γ-carboxylate group of L-glutamate also binds. Thus, binding of Dphenylglycine and L-glutamate by DPAT may require both the carboxylate and side-chain binding pockets to undergo some structural rearrangement, in contrast to most other aminotransferases, which only require rearrangement of the side-chain binding pocket.39 This hypothesis is supported by the missing electron density observed in each pocket in our structure of the DPAT apo state, which suggests a high degree of conformational flexibility in these regions. As each pocket in DPAT may bind an α-carboxylate group (of either D-phenylglycine or L-glutamate),

we will use the O- and P-pocket nomenclature28 to analyze the binding pockets of DPAT

instead of the traditional carboxylate/side-chain binding pocket terminology. In this nomenclature, the Oand P-pockets are defined by their proximity to the 3′-oxygen or phosphate group of PLP. We start our analysis by superimposing substrate-bound structures of GSAM (PDB ID: 2HP2)32 or MesAT (PDB ID: 2YKX)28 with that of DPAT (Fig. 4). Inspection of the superimposed structures allowed us to identify DPAT residues H66, H213, and R407 as potential substrate binding residues in the O-pocket. For the Ppocket however, it was not possible to identify substrate binding residues as the crystal structure of the DPAT apo state is missing electron density in this region of the active site (residues 28–36 and 292–302). Therefore, it was necessary to rely on sequence alignments to identify potential substrate-binding residues in DPAT’s P-pocket. Both GSAM and MesAT utilize an arginine residue in the P-pocket to coordinate the carboxylate of their respective substrates (Fig. 4). This arginine residue in GSAM aligns with R34 of DPAT, whereas the arginine residue in MesAT is located three residues upstream in the sequence alignment (Fig. 3), suggesting that the P-pocket of DPAT may be organized in an arrangement similar to

12


Page 12 of 23

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

Biochemistry

that of GSAM. Inspection of the P-pockets of GSAM and MesAT show that they both contain an Ala residue whose side chain is in close proximity to the substrate (Fig. 4). Interestingly, the residue at the corresponding position in DPAT is a glutamine (Q301*, where * indicates a residue from the other subunit), which is rare among class III aminotransferases.40 Because this polar residue can act as a hydrogen bond donor, it may contribute to the binding of the D-phenylglycine or α-ketoglutarate substrates and was thus analyzed further.

Figure 4. Comparison of the DPAT active site with that of homologous class III aminotransferases. The active site of DPAT (blue) is overlaid with that of the (a) MesAT internal aldimine bound to αketoglutarate (PDB ID: 2YKX, yellow), (b) MesAT (S)-3-phenyl-β-alanine external aldimine (PDB ID: 2YKY, green), and (c) GSAM (4S)-4,5-diaminopentanoate external aldimine (PDB ID: 2HP2, salmon). DPAT residues, including the phosphate ion bound at the active site (PO4), are labeled in italics. An asterisk indicates an active site residue located on the other subunit. Hydrogen bonds are indicated by dashed lines. The catalytic lysine residue is shown below the PLP cofactor.

Probing the active site by mutagenesis. Based on our comparison of DPAT with the homologous GSAM and MesAT enzymes, we performed saturation mutagenesis to produce five mutant libraries of residues in the O- (H66, H213, R407) and P-pockets (R34, Q301). Library screening was performed in 96-well plates using a coupled enzyme assay that we previously developed (Fig. S3).25 However, due to the low solubility of wild-type DPAT, expression in 96-well plate format did not yield sufficient quantities of the enzyme to distinguish its activity from the negative control containing no DPAT. To solve this issue, we introduced two surface mutations (N439D/Q444E) into DPAT, which were previously shown to improve solubility without causing significant changes to the specific activity of the enzyme.41 Introduction of these mutations into our saturation libraries resulted in an increased amount of soluble protein produced in 96-well format, which was sufficient to enable screening of clarified lysates.

13


Biochemistry 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

Screening of the O-pocket saturation libraries for transamination activity with D-phenylglycine as the amine donor and α-ketoglutarate as acceptor yielded several active mutants at positions 66 and 213 (Fig. S4), most of which displayed reduced activity compared to the wild type. However, all mutations introduced at position 407 yielded variants with indistinguishable activity to that of the negative controls, in agreement with the observation that an arginine residue in this pocket is required for aminotransferases to bind substrates [Mehta, Perdeep K., Terence I. Hale, and Philipp Christen. "Aminotransferases: demonstration of homology and division into evolutionary subgroups." European Journal of Biochemistry 214.2 (1993): 549-561]. A similar result was observed for MesAT when the corresponding R412 residue was mutated to an alanine, which caused a large decrease in specific activity (

Structural Determinants of the Stereoinverting Activity of

Recommend Documents