Structural Basis of the Substrate Range and Enantioselectivity of

Subscriber access provided by UNIV OF TECH SYDNEY

Article

Structural basis of substrate range and enantioselectivity of two (S)-selective #-transaminases Niels van Oosterwijk, Simon C Willies, Johan Hekelaar, Anke C Terwisscha van Scheltinga, Nicholas J Turner, and Bauke W. Dijkstra Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.6b00370 • Publication Date (Web): 18 Jul 2016 Downloaded from http://pubs.acs.org on July 20, 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.

Biochemistry 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 39

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 basis of substrate range and enantioselectivity of two (S)-selective ωtransaminases Niels van Oosterwijka, Simon Williesb, Johan Hekelaara, Anke C. Terwisscha van Scheltingaa, Nicholas J. Turnerb and Bauke W. Dijkstraa* a

Laboratory of Biophysical Chemistry, Groningen Biomolecular Sciences and Biotechnology

Institute (GBB), University of Groningen, Nijenborgh 7, 9747 AG Groningen, The Netherlands b

School of Chemistry, Manchester Institute of Biotechnology, University of Manchester, 131 Princess Street, Manchester, M1 7DN, UK

* To whom correspondence may be addressed: Bauke W. Dijkstra Laboratory of Biophysical Chemistry, University of Groningen, Nijenborgh 7, 9747 AG Groningen, The Netherlands E-mail: [email protected] Phone +31 50 363 4381 / 4378, Fax +31 50 363 4800

ACS Paragon Plus Environment

1

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 2 of 39

Abstract

ω-Transaminases are enzymes that can introduce an amino group in industrially interesting compounds. We determined crystal structures of two (S)-selective ω-transaminases, one from Arthrobacter sp. (Ars-ωTA), and one from Bacillus megaterium (BM-ωTA), which have 95% sequence identity, but somewhat different activity profiles. Substrate-profiling measurements using a range of (R)- and (S)-substrates showed that both enzymes have a preference for substrates with large, flat cyclic side groups, for which the activity of BM-ωTA is generally somewhat higher. BM-ωTA has a significantly higher preference for (S)-3,3-dimethyl-2butylamine than Ars-ωTA, as well as a less enantiopreference towards 1-cyclopropylethylamine. The crystal structures showed that, as expected for (S)-selective transaminases, both enzymes have the typical transaminase type I fold, and have spacious active sites to accommodate largish substrates. A structure of BM-ωTA with bound (R)-α-methylbenzylamine explains the enzymes’ preference for (S)-substrates. Site-directed mutagenesis experiments revealed that the presence of a tyrosine, instead of a cysteine, at position 60 increases the relative activities on several small substrates. A structure of Ars-ωTA with bound L-Ala revealed that the Arg442 side chain has repositioned to bind the L-Ala carboxylate. Compared to the arginine switch residue in other transaminases, Arg442 is shifted by six residues in the amino acid sequence, which appears to be a consequence of extra loops near the active site that narrow the entrance to the active site.

Keywords: S-selective, enantioselectivity, ω-transaminase, crystal structure


2

Page 3 of 39

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

Transaminases (EC class 2.6.1.-) catalyse the transfer of an amino group from a donor substrate to an acceptor molecule, replacing a carbonyl group on the acceptor molecule. Transaminases use pyridoxal-5’-phosphate (PLP) as a cofactor, which, in the resting state of the enzyme, is covalently attached to the ε-amino group of a lysine residue in the active site. The catalysed reaction consists of two half reactions. In the first half reaction the enzyme transfers the amine group from the donor substrate to the PLP cofactor, after which, in the second half reaction, the PLP-bound amino group is transferred to the acceptor substrate. The amine donors can vary, but an often-used amino donor is L-alanine, which is converted into pyruvate (Figure 1). Transaminases can be distinguished according to the position of the amino group that is transferred. α-Transaminases (αTAs) transfer the amino group at the α-position of an amino acid, whereas ω-transaminases (ωTAs) act on an amino group at any other position than the αposition1. In contrast to α-transaminases, ωTAs do not necessarily require a carboxylate group in their substrates, which makes them much more versatile in biocatalytic transformations than αTAs. Moreover, they often have well-defined enantioselectivity, which puts them among the most useful enzymes for the production of amine-containing, chiral intermediates for the pharmaceutical and fine chemicals industry1,2. An impressive example of the application of an ωTA is in the production of the anti-diabetic compound sitagliptin, where an extensively engineered (R)-selective ω-transaminase from an Arthrobacter species replaced a rhodium-catalysed asymmetric enamine hydrogenation, which suffered from inadequate stereoselectivity and a product stream contaminated with rhodium 3. A second example, using (S)-selective ωTAs, is the total synthesis of S-rivastigmine, one of the most potent drugs for the treatment of early stages of Alzheimer’s disease4. ωTAs are also


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

Page 4 of 39

applied in the production of halogenated aromatic chiral amines, which are building blocks for potent potassium channel openers with application in the treatment of e.g. hyperparathyroidism5. Nevertheless, the enantioselectivity of ω-transaminases is often substrate-dependent and not always high enough for unnatural substrates. Furthermore, the enzymes must be able to cope with the, sometimes harsh, process conditions like high temperatures, non-physiological pH, and the presence of organic solvents. Adaptation of enzymes to new substrates and process conditions is an important goal to make them suitable for application. Moreover, the transaminase reaction is often affected by unfavourable steady state kinetics in the direction of the ketone. Changing the kinetics is essential for increasing the amine product yield. This can be done by, for example, removing the co-product pyruvate by lactate dehydrogenase or pyruvate decarboxylase, or by employing an amine donor (alanine) regeneration system6-8. As an alternative, the product yield can be increased by replacing the amine donor alanine by another amine donor like α-methylbenzylamine (α-MBA) or 2-propylamine2, which usually also requires adaptation of the enzyme. To aid rational modification of transaminases for biocatalytic applications, various crystal structures have been elucidated over the years. In 1989 the structure of the (S)-selective ωamino-acid:pyruvate transaminase from Pseudomonas sp. F-126 was solved9. Examples of more recently determined structures are the structure of the S-selective Chromobacterium violaceum transaminase (CV-ωTA)10, the gabaculine-bound structures of CV-ωTA and the βalanine:pyruvate aminotransferase from P. aeruginosa (PA-ωTA)11, and the recent structure of an (R)-selective ωTA from an Arthrobacter sp.12. These structures showed that (S)- and (R)selective ωTAs have very different three-dimensional structures. The (S)-selective enzymes all belong to fold type I of PLP-dependent enzymes, and (R)-selective ωTAs have fold type IV.


4

Page 5 of 39

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

Despite these crystal structures, the structural determinants that govern the stereoselectivity and substrate range of ωTAs are not fully understood. Here, we report on the ~54 kDa S-selective ωtransaminases from Arthrobacter sp. (Ars-ωTA)13 and Bacillus megaterium (BM-ωTA)14, which are 95% identical in amino acid sequence, but have somewhat different activity profiles. Extended substrate-profiling experiments, using a range of (R)- and (S)- substrates, and crystal structures revealed several amino acid residues involved in the different substrate preferences.


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 39

Materials and Methods

Expression of BM-ωTA(His6) and Ars-ωTA(His6) For expression the BM-ωTA and Ars-ωTA genes were cloned into pET21a(+) vectors containing a C-terminal His6-tag to facilitate purification. Expression was tested in E. coli strains C4315, BL21* (Invitrogen), Rosetta II (DE3) (Novagen), and Rosetta II pLysS (Novagen), and the best systems, E. coli Rosetta II (DE3) for BM-ωTA(His6), and E. coli C43 for Ars-ωTA(His6), were used for protein production. For BM-ωTA(His6) production, cells were grown at 37 °C in 1 litre of Luria Bertani broth containing 100 µM vitamin B6, 100 µg/ml ampicillin and 34 µg/ml chloramphenicol. At an OD600 of 0.6-0.8 protein expression was induced with 0.5 mM isopropyl-β-D-thiogalactopyranoside. Three hours after induction, the cells were harvested, and resuspended in 40 ml lysis buffer containing 50 mM Tris-HCl, pH 8.0, 300 mM NaCl, 50 µM PLP, a catalytic amount of DNase I, and one EDTA-free protease inhibitor tablet (Roche Applied Science, according to protocol). The resuspended cells were passed twice through a French press (Thermo Scientific), and, after centrifugation, the supernatant was used for protein isolation and purification. Expression of Ars-ωTA(His6) was carried out in a similar fashion, without the addition of chloramphenicol, but, to obtain soluble protein, a cold shock (10 min, 0 °C), followed by 6 hours of expression at 25 °C was required.


6

Page 7 of 39

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

Protein purification and characterization The purification protocols were similar for both transaminases. Cell-free extracts were loaded on a Hi-Trap IMAC (immobilized metal affinity chromatography) column (GE Healthcare, Buckinghamshire, UK), pre-washed with 6 column volumes (CV) of IMAC buffer (50 mM TrisHCl, pH 8.0, 300 mM NaCl and 50 µM PLP). After washing with 6 CV of IMAC buffer containing 30 mM imidazole, the target protein was eluted with 3 CV of IMAC buffer containing 300 mM imidazole. The eluted fractions were analyzed for protein content and purity, using Coomassie blue stained SDS-PAGE gels. Fractions containing ω-transaminase were pooled, and diluted to a NaCl concentration of less than 50 mM by adding AEX buffer (50 mM Tris-HCl, pH 8.0, 50 µM PLP). The diluted sample was loaded on an anion-exchange column (ResourceQ, GE Healthcare) equilibrated with AEX buffer. After washing with 5 CV of AEX buffer, the protein was eluted using a 0 to 500 mM NaCl gradient (20 CV) in AEX buffer. Fractions containing ωtransaminase were identified by SDS-PAGE, and pooled and concentrated using an Amicon Ultra 10K concentrator (Milipore). Purified enzymes were characterized by recording absorption spectra (NanoDrop1000, Thermo Scientific) in the 250 to 600 nm range. The absorption at 280 nm was used to estimate the protein concentration (ε = 86290 for Ars-ωTA, 89270 for BM-ωTA, based on their amino acid contents). Ars-ωTA was concentrated to 32 mg ml-1, and BM-ωTA to 20 mg ml-1. Both transaminases were characterized by analytical size exclusion chromatography (SEC) using a Superdex200 column (GE Healthcare), with a 50 mM Tris-HCl buffer, pH 8.0, containing 200 mM NaCl and 50 µM PLP. The purified protein samples were divided into aliquots, frozen in liquid nitrogen, and stored at -20 °C until use.


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 39

Crystallization Crystallization conditions of both proteins were screened for, using the JCSG+ (Qiagen Corporation) and PACT (Qiagen Corporation) screens, set up with a Mosquito (TTP Labtech) pipetting robot. For each condition two drops were pipetted, one consisting of 125 nl protein solution and 75 nl reservoir solution, and the other of 75 nl protein and 125 nl reservoir solution. The resulting drops were monitored using a RockImager imaging system (Formulatrix). For Ars-ωTA, PACT screen conditions, containing a range of PEGs, 0.1 M Bis-Tris propane buffer, pH 6.5, and 0.2 M of salt (NaNO3, Na-malonate, NaF or NaCl), showed the most promising crystals. After optimization, diffraction quality crystals were obtained from sitting drops of 1 µl protein solution and 1 µl reservoir solution (13-18 % PEG 3350, 0.2 M Namalonate and 100 mM MMT buffer (DL-malic acid, MES and Tris base), pH 5.0, and 10 mM PLP). The crystals appeared after approximately 2-3 days, and were cryoprotected using 20% glycerol. For alanine binding studies, Ars-ωTA crystals were soaked for ~30 sec in cryo-mother liquor containing 200 mM alanine. Co-crystallization and longer soaks did not result in diffracting crystals. BM-ωTA crystals were obtained from a JCSG+ screen condition, consisting of 30 % PEG300 and 50 mM Na-phosphate/citrate, pH 4.2, and 10 mM PLP. Crystals suitable for data collection grew in approximately 1 day, using the hanging drop vapor diffusion method with drops consisting of 1 µl protein solution and 1 µl reservoir solution. Special care was taken to use fresh crystals for data collection, as diffraction quality deteriorated after about three weeks of storage at room temperature. Crystals of BM-ωTA complexed with substrate were obtained by soaking crystals for ~1 minute in mother liquor saturated with (R)-α-methylbenzylamine ((R)-α-MBA,


8

Page 9 of 39

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

~60 mM). Since the mother liquor contained 30 % PEG300, further cryoprotection was not needed.

Small-angle X-ray scattering Small-angle X-ray scattering experiments SAXS were performed using BM-ωTA at the X33 beamline at the EMBL outstation at DESY (Hamburg). Data was collected on samples using a concentration range from 1.8 to 5.5 mg ml-1 in 50 mM Tris-HCl, pH 8.0, 150 mM NaCl and 50 µM PLP. The BM-ωTA SAXS data was processed using ATSAS16 (scattering curves and Guinier plots shown in Figure S1).

Data collection and processing All crystals were cryo-cooled in liquid nitrogen, and data sets were collected at 100 K. Diffraction images were processed and scaled with XDS17 and SCALA18 from the CCP4 package19. Native and (R)-α-MBA soaked BM-ωTA crystals have the same space group (P212121) with highly similar cell parameters. The native Ars-ωTA crystals had space group P21 and the crystal used for the alanine soak had space group P1. An overview of the data collection statistics can be found in Table 1.


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

Page 10 of 39

Transaminase activity assay Assays were carried out in 96 well micro-titre plates, with 10 µl substrate stock solution (200 mM in 100% DMSO (dimethyl sulfoxide)), 95 µl assay stock solution and 95 µl enzyme solution. The assay stock solution consisted of 10 U D-amino acid oxidase, 1 mg horseradish peroxidase, 2 mg pyruvate, 3 mg 4-aminoantipyrine, 2 mg 2,4,6-tribromo-3-hydroxybenzoic acid, all dissolved in 10 ml 100 mM HEPES buffer, pH 8.0. Pyruvate is used as amino acceptor in the assay, and is converted into D-alanine. Subsequently, the D-alanine is oxidized by a Damino acid oxidase (Sigma) and molecular oxygen, which regenerates the pyruvate with concomitant formation of ammonia and H2O2. The formed H2O2 is a substrate for horseradish peroxidase, which converts equimolar amounts of hydrogen peroxide, 4-aminoantipyrine, and 2,4,6-tribromo-3-hydroxybenzoic acid into a quinone imine dye20. The increase in absorption at 510 nm due to the formation of the dye was monitored with a spectrophotometer at 30 °C for 30 minutes.

Phasing & Refinement The structure of BM-ωTA was solved by molecular replacement with PHASER21 using a putative aminotransferase from Mesorhizobium loti as search model (25% sequence identity, PDB entry 3gju (DOI:10.2210/pdb3gju/pdb)). The refined structure of BM-ωTA was subsequently used to solve the structure of Ars-ωTA. Both molecular replacement solutions were good enough to allow automatic model building by ARP/wARP22. Models obtained from ARP/wARP were improved by manual rebuilding using COOT23 alternated with refinement using Refmac524. Initially, non-crystallographic symmetry (NCS)


10

Page 11 of 39

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

restraints were applied to the four molecules in the asymmetric unit during model building and refinement. TLS (Translation/Libration/Screw) refinement25 was added in the final steps for all structures presented here. At this stage, the NCS restraints were also released, except for the 12 molecules in the P1 Ars-ωTA alanine structure. Restraints for the alanine PLP adduct were generated using JLigand26, and for the (R)-α-MBA PLP adduct with PRODRG27. The occupancy of the alanine was estimated by matching the temperature factors of the alanine with those of the PLP cofactor. An overview of the refinement statistics, final R-factors and the various nonprotein molecules found in the structures can be found in Table 2. The quality of the final models and the geometrical restraints were checked using Molprobity28. The coordinates and structure factor amplitudes have been deposited in the Protein Data Bank with accession codes 5g0a (BMωTA - native), 5g09 (BM-ωTA - (R)-α-MBA) and 5g2p (Ars-ωTA - native) and 5g2q (Ars-ωTA - alanine). Figures of the structures were created using PYMOL29.


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

Page 12 of 39

Results and Discussion

Protein expression, purification, crystallization and data collection Expression tests for BM-ωTA and Ars-ωTA, using several different E. coli strains, showed that for BM-ωTA highest expression was achieved in E. coli RII (DE3). In contrast, similar tests with Ars-ωTA produced protein in inclusion bodies. To obtain soluble Ars-ωTA the growth temperature was reduced to 25 °C, which gave best results using E. coli strain C43. For the crystallization experiments, Ars-ωTA and BM-ωTA were both purified by immobilized metal affinity and anion exchange chromatography. Whereas Ars-ωTA showed the typical yellow colour of a PLP-containing protein, the BM-ωTA cell pellet and purified protein were pink. A UV-VIS spectrum showed absorption at 550 nm, indicative of a quinonoid form of the PLP30. BM-ωTA crystals obtained using these protein samples obtained the yellow colour usually observed for PLP-enzyme complexes after about one week, and were subsequently used for data collection (Table 1). The crystals diffracted to a resolution better than 2.0 Å. Crystallization of Ars-ωTA yielded crystals that diffracted to 1.9 Å resolution.

Substrate ranges of Ars-ωTA and BM-ωTA The activity of Ars-ωTA and BM-ωTA was tested on a range of (S)- and (R)-amines with pyruvate as the amino acceptor (see Methods for details). Both enzymes show comparable activities for most of the substrates tested (Figure 2). In particular, both proteins prefer substrates with large R-groups like (S)-naphthylethylamines and halogenated (S)-methylbenzylamines,


12

Page 13 of 39

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

although the activity of BM-ωTA on these compounds is generally somewhat higher than that of Ars-ωTA. BM-ωTA is also significantly more active on (S)-3,3-dimethyl-2-butylamine. Both enzymes are not only active on (S)-α-methylbenzylamine, but also on (S)-α-ethylbenzylamine, a substrate with a slightly larger secondary side group. The enzymes show high (S)enantioselectivity. The only exception is 1-cyclopropylethylamine, of which both enantiomers are converted with about equal efficiency by BM-ωTA (Figure 2). To analyse the molecular basis of these intriguing activity differences we determined crystal structures of Ars-ωTA and BM-ωTA.

3D-structures of BM-ωTA and Ars-ωTA As expected from their 95% sequence identity, the crystal structures of BM-ωTA and Ars-ωTA (see Table 2 for refinement statistics) are highly similar. The monomers exhibit the typical transaminase type I fold, consisting of two main domains (Figure 3). The large domain (residues 69-351) binds the PLP cofactor, and contains a mostly parallel seven-stranded β-sheet flanked by α-helices on both sides. The small domain comprises residues 1-68 and 352-474. The latter part contains a three-stranded antiparallel β-sheet flanked by α-helices, and residues 430 to 434 form a mixed four-stranded β-sheet with residues 37-55. The N-terminal tail (residues 1-36) forms a long loop and an α-helix, and is involved in dimer contacts. Native BM-ωTA and Ars-ωTA crystallized with a homotetramer in the asymmetric unit. The protomers of this tetramer have virtually identical structures, as shown by the pairwise RMSDs of Cα positions, which are all less than 0.6 Å. Analytical gel filtration yielded single peaks for both proteins, corresponding to an apparent molecular weight of approximately 150 kDa, which


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

Page 14 of 39

would indicate a trimer. However, SAXS data indicate a molecular weight of 210 kDa, in agreement with a tetramer, with the Rg (~ 4.0 nm) and Dmax (~ 12 nm) values matching those of the tetramer found in the X-ray structures (Rg = 3.6 nm , Dmax = 11.7 nm), see Table S1. Since also a PISA analysis31 indicates that the proteins likely oligomerize as tetramers, organized as a dimer of dimers (D2 symmetry; buried surfaces of 5600-5800 Å2 within the dimer, and of 2850 A2 between the dimers) we conclude that the tetramers in the crystals are in agreement with the oligomerization states of the proteins in solution, as observed for other fold type I transaminases32.

The active sites of BM-ωTA and Ars-ωTA In both BM-ωTA and Ars-ωTA the PLP co-factor is covalently attached to the Lys298 side chain. PLP binds in a pocket lined with residues conserved in both proteins (Figure S2), and all its non-carbon atoms are involved in hydrogen bonding interactions. In particular, the phosphate group is hydrogen bonded to the backbone amides of Gly121 and Ser122, and to the Ser122 side chain. Thr330-B, a residue coming from the B monomer in the dimer, hydrogen bonds to the phosphate group with its backbone amide and side chain. Four water molecules that are hydrogen bonded to the protein complete the hydrogen bonding interactions of the phosphate group. The phenolate oxygen has a hydrogen bond to a water molecule held in place by Glu237 and Ala242. The N1 nitrogen of the PLP is at hydrogen bonding distance to the Asp269 side chain. This interaction promotes protonation of N1, which is beneficial for PLP enzymes involved in transaminase reactions33. Thr330-B and the carbonyl of Leu59 are close to the Schiff base, and


14

Page 15 of 39

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

bind the Lys298 Nζ group, when it is not involved in Schiff base formation with the PLP (see Figure 5). The transaminase substrate-binding site contains two pockets, in which the substrate’s R1 and R2 side chains (Figure 4) bind34. The original naming based on the size of the pockets is no longer consistent with the size observed in various transaminase structures. Therefore, we prefer to name the pockets according to their proximity to either the 3’-oxygen (O pocket, previously L) or the phosphate group (P pocket, previously S) of the PLP35. To analyse substrate binding, a 1.9 Å resolution crystal structure was elucidated of (R)-α-MBA bound to BM-ωTA. In all four molecules in the asymmetric unit the compound is covalently attached to PLP without any large conformational changes of the protein or PLP. The phenyl ring of (R)-α-MBA is bound in the hydrophobic O pocket, between the side chains of Leu59 and Ala242 (Figure 5). The bound (R)-α-MBA completely occludes the PLP from the solvent, limiting solvent access to the Schiff base. The methyl group of (R)-α-MBA is bound in the small P pocket, located at the dimer interface, and lined by Val328-B, Thr330-B and Leu59. In both BM-ωTA and Ars-ωTA the O pocket is the last part of a funnel leading up to the PLP. The funnel provides space for substrates with large side chains, but close to the PLP co-factor it becomes narrower, defining a slit-like space. Leu59 and Ala242 (a Val in Ars-ωTA) function as “gate residues” that restrict access to the PLP co-factor to flat molecules. This explains the preference of BM-ωTA and Ars-ωTA for substrates with large, planar aromatic substituents like (S)-naphthylethylamines and (S)-methylbenzylamines. At the edge of the (R)-α-MBA phenyl ring, near Leu272 and Val328-B, space is available to provide some freedom of orientation of the bound substrate, and to accommodate substrates with even larger substituents.


15

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 16 of 39

Covalent binding of (R)-α-MBA frees Lys298 of which the Nζ group now forms hydrogen bonds with the backbone carbonyl oxygen of Leu59 and the hydroxyl group of Thr330-B. In the PLP(R)-α-MBA complex, the C4’ carbon of the PLP moiety clearly shows sp2 hybridization and the C7 atom is sp3 hybridized. This confirms that the adduct is the external aldimine intermediate (see Figure 6).

(R)-α-MBA binding explains activity towards S-substrates For most substrates, BM-ωTA shows higher activity towards S-compounds than to Rcompounds. The structure of the (R)-α-MBA external aldimine intermediate allows rationalizing the enzyme’s (S)-enantioselectivity. To proceed from the external aldimine to the ketimine intermediate (Figure 6), a proton needs to be transferred from the C7 atom of the external aldimine adduct to the C4’ carbon atom of the PLP. In the (R)-α-MBA-bound structure, no proton acceptor is available, and the C7 proton points away from the Lys298 Nζ group, precluding involvement of Lys298 in the reaction. This is consistent with (R)-α-MBA being trapped in the external aldimine state in the crystal structure. In contrast, BM-ωTA readily processes (S)-compounds. Assuming that the preferred substrate (S)-α-MBA binds similarly to (R)-α-MBA (Figure S3), with its phenyl and methyl groups in the O and P pockets, respectively, the C7 proton would be directed towards the Lys298 Nζ group (at ~3.5 Å), and the scissile C7-H bond would be perpendicular to the π-system of the PLP ring system, which facilitates easy hydrogen abstraction33. Thus, with an (S)-enantiomeric substrate, Lys298 would be in a perfect position for hydrogen abstraction and promoting the formation of


16

Page 17 of 39

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

the ketimine intermediate. In conclusion, the position of the lysine with respect to the substrate’s scissile C7-H bond is crucial for the enantioselectivity of the enzymes. BM-ωTA is not enantioselective on 1-cyclopropylethylamine; both enantiomers are converted with about equal efficiency (Figure 2). Possibly, the similar size and apolar nature of the R1 and R2 substituents could allow each of them to bind in the O and P pockets, such that the C7 proton of both enantiomers can be directed towards the active site Lys298 Nζ group, resulting in the loss of enantioselectivity towards this substrate. However, Ars-ωTA does not show this lack of enantioselectivity on 1-cyclopropylethylamine. Because its P-pocket is indistinguishable from that of BM-ωTA, and the peptide backbone conformations around the active site are virtually identical, the difference in enantioselectivity is likely due to variations in the amino acid side chains in the O pocket. Indeed, as shown by site-directed mutagenesis experiments (see below) the O pocket residue Tyr60 is an important determinant of the lack of enantioselectivity of BMωTA towards 1-cyclopropylethylamine.

Determinants of activity and enantioselectivity BM-ωTA and Ars-ωTA differ by 20 residues, of which four are in the active site. Three of these active site differences (Tyr/Cys60, Tyr/Phe164 and Val/Ala436) result in a larger O pocket in Ars-ωTA. Yet, both enzymes are equally well able to process naphthylethylamines (Figure 2), which were the largest substrates we tested. Thus, the different size of the O pocket may only be of consequence for even larger substrates than naphthylethylamines. The fourth difference, the Ala/Val gate residue at position 242, which is also situated in the vicinity of the O-pocket, narrows the slit-like binding site for flat groups in Ars-ωTA. This latter amino acid difference


17

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 18 of 39

may affect the enzymes’ preferences for aromatic vs. aliphatic ring systems, but, unfortunately, such compounds were not included in our tests. To probe the importance of the amino acid differences in the active sites of Ars-ωTA and BMωTA we constructed variants at positions 60 and 242 (BM-ωTA: Y60C, A242V; Ars-ωTA: C60Y, V242A), and investigated their substrate specificities (Figure 7 and 8). Whilst neither of these mutations has completely switched the substrate specificities, interesting differences were found. For instance, the introduction of a tyrosine at position 60 increases the relative activities on cyclohexylamine, S-3,3-dimethyl-2-butylamine, and R-1-cyclopropylamine, whereas the presence of a cysteine results in decreased activities (Figure 7). Tyr60 is situated in the O pocket, and the phenyl ring of the bound (R)-α-MBA lies in the same plane as the aromatic ring of Tyr60 at about 3.8 Å, contacting it via edge-to-edge interactions (Figure 5). Assuming that cyclohexylamine binds in a similar way as (R)-α-MBA an interaction of the aliphatic ring with Tyr60 can be envisaged which may contribute to the affinity for this substrate. Mutation of Tyr60 to a cysteine will remove those interactions, and could explain the lower activity. S-3,3dimethyl-2-butylamine is smaller than (R)-α-MBA, and modelling (using COOT23) shows that it is unlikely to interact with the side chain of Tyr60. Yet, also with this compound a Tyr to Cys mutation reduces the activity. Possibly, the additional space in the active site created by the Tyr to Cys mutation causes less specific binding resulting in lower activity. The introduction of an Ala at position 242 in Ars-ωTA (Figure 8) results in a significantly increased activity of Ars-ωTA towards (S)-3,3-dimethyl-2-butylamine. In contrast, the reverse mutation in BM-ωTA (A242V) has no effect on the activity. Modelling shows that the tri-methyl group of (S)-3,3-dimethyl-2-butylamine barely fits between the gate residues. The alanine present in Ars-ωTA may provide extra space to accommodate this substrate, and thus may


18

Page 19 of 39

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

explain the higher activity. The lack of any effect of the reverse mutation in BM-ωTA may be due to the presence of Tyr60, which is beneficial for activity on (S)-3,3-dimethyl-2-butylamine (see above). BM-ωTA and Ars-ωTA differ only to a limited extent in enantioselectivity. This is reflected in the mutational data. Only for 1-cyclopropylethylamine, a substrate for which BM-ωTA lacks enantioselectivity, the Y60C mutation conferred the (S)-enantioselectivity found in Ars-ωTA. Apparently a tyrosine at position 60 in BM-ωTA allows binding and conversion of both enantiomers of 1-cyclopropylethylamine. It is conceivable that the small apolar side chains of this compound can be accommodated in both the O and P pockets, thus explaining the lack of enantioselectivity of BM- ωTA. Alternatively, a water molecule present in the active site, at 4.1 Å from the C7 atom of (R)-alpha-MBA (Figure 5), may function as a nucleophile to hydrolyse the ketimine intermediate of the non-preferred substrate. This water molecule could be activated by Tyr148 and/or the PLP phosphate group (Figure 5). With preferred enantiomers, hydrolysis has been proposed to occur by a water molecule directly activated by the catalytic lysine side chain (Figure 6). However, where this water is positioned, is currently unclear and further studies are required.

Structure of Ars-ωTA with a bound alanine: the role of the arginine switch To study the interaction of Ars-ωTA with an amino donor, we soaked a crystal with alanine. The resulting structure revealed electron density close to the PLP cofactor in several of the 12 molecules in the asymmetric unit. Although weak, an alanine could be modelled in the strongest density (molecule C, estimated occupancy 60 %). The alanine is covalently bound to the C4’


19

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 20 of 39

atom of the PLP, and its carboxylate group forms a salt bridge with the guanidinium group of Arg442 in the hydrophobic O pocket (Figure 9). To accommodate the hydrophobic side chain of a substrate as well as the carboxylate group of the amino donor alanine in the O pocket, ωtransaminases often use an arginine residue that switches between “in” and “out” positions in the active site32. In the Ala-bound Ars-ωTA structure, Arg442 has the “in” position, but in the native structure the residue has moved aside into the “out” position. There it is held in place by hydrogen bonds to the carbonyl groups of Gly241, Gln240, Trp383 and the side chain of Ser239. The equivalent arginine in BM-ωTA, Arg442, could take the same “in” position, although a minor movement of the Tyr60 side chain would be needed. The structures of Ars-ωTA clearly demonstrate the role of Arg442 as the arginine switch in these ω-transaminases. Arg442 as the arginine switch residue is in contrast to the arginine switch identified in other ω-transaminases. A NCBI BLAST search of BM-ωTA against the PDB-database identified entries 3HMU, 3I5T, 3FCR, 3GJU (all unpublished) and CV-ωTA (4AH3)10,11 as the entries with the highest sequence identity. Structural alignment of these fold type I transaminases using the PDBeFold server36 showed that Ars-ωTA and BM-ωTA have two extended loops near the active site (residues 163178 and 402-416, respectively), which are absent in the other proteins. These loops result in the more closed active site funnels found in Ars-ωTA and BM-ωTA. If the arginine switch residue would have the same position in the sequence as in the canonical ω-transaminases, it would restrict access to the active site tunnel of Ars-ωTA and BM-ωTA, when the arginine side chain is in the out-position, prompting the observed six-residue sequence shift of the arginine switch residue in Ars-ωTA and BM-ωTA.


20

Page 21 of 39

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

Acknowledgements We thank Prof. Wolfgang Kroutil (Graz) for providing the Ars-wTA and BM-wTA genes. We thank Kor H. Kalk and the beam-line scientists of the ID14-1, ID23-2 (ESRF, Grenoble) and PX3 (Swiss Light Source, Villingen) beamlines for their help during data collection. We thank Harshwardan Poddar and Paul Tucker for helpful discussions.

Supporting information Available Figure S1: SAXS scattering curves and Guinier plots Figure S2. Interactions of the PLP co-factor Figure S3. Configurations around the C7 atom of the external aldimine intermediate


21

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 22 of 39

REFERENCES (1) Malik, M. S., Park, E.-S., and Shin, J.-S. (2012) Features and technical applications of ωtransaminases. Appl. Microbiol. Biotechnol. 94, 1163–1171. (2) Tufvesson, P., Lima-Ramos, J., Jensen, J. S., Al-Haque, N., Neto, W., and Woodley, J. M. (2011) Process considerations for the asymmetric synthesis of chiral amines using transaminases. Biotechnol. Bioeng. 108, 1479–1493. (3) Savile, C. K., Janey, J. M., Mundorff, E. C., Moore, J. C., Tam, S., Jarvis, W. R., Colbeck, J. C., Krebber, A., Fleitz, F. J., Brands, J., Devine, P. N., Huisman, G. W., and Hughes, G. J. (2010) Biocatalytic Asymmetric Synthesis of Chiral Amines from Ketones Applied to Sitagliptin Manufacture. Science 329, 305–309. (4) Fuchs, M., Koszelewski, D., Tauber, K., Kroutil, W., and Faber, K. (2010) Chemoenzymatic asymmetric total synthesis of (S)-Rivastigmine using omega-transaminases. Chem. Commun. (Camb.) 46, 5500–5502. (5) Malik, M. S., Park, E.-S., and Shin, J.-S. (2012) Features and technical applications of ωtransaminases. Appl. Microbiol. Biotechnol. 94, 1163–1171. (6) Höhne, M., Kühl, S., Robins, K., and Bornscheuer, U. T. (2008) Efficient asymmetric synthesis of chiral amines by combining transaminase and pyruvate decarboxylase. Chembiochem 9, 363–365. (7) Truppo, M. D., Rozzell, J. D., and Turner, N. J. (2009) Efficient production of enantiomerically pure chiral amines at concentrations of 50 g/L using transaminases. Org. Process Res. Dev. 14, 234–237. (8) Galkin, A., Kulakova, L., Yoshimura, T., Soda, K., and Esaki, N. (1997) Synthesis of optically active amino acids from alpha-keto acids with Escherichia coli cells expressing


22

Page 23 of 39

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

heterologous genes. Appl.Environ.Microbiol. 63, 4651–4656. (9) Watanabe, N., Sakabe, K., Sakabe, N., Higashi, T., Sasaki, K., Aibara, S., Morita, Y., Yonaha, K., Toyama, S., and Fukutani, H. (1989) Crystal structure analysis of omega-amino acid:pyruvate aminotransferase with a newly developed Weissenberg camera and an imaging plate using synchrotron radiation. J.Biochem.(Tokyo). 105, 1–3. (10) Humble, M. S., Cassimjee, K. E., Håkansson, M., Kimbung, Y. R., Walse, B., Abedi, V., Federsel, H.-J., Berglund, P., and Logan, D. T. (2012) Crystal structures of the Chromobacterium violaceum ω-transaminase reveal major structural rearrangements upon binding of coenzyme PLP. FEBS J. 279, 779–792. (11) Sayer, C., Isupov, M. N., Westlake, A., and Littlechild, J. A. (2013) Structural studies of Pseudomonas and Chromobacterium ω-aminotransferases provide insights into their differing substrate specificity. Acta Crystallogr. D Biol. Crystallogr. 69, 564–576. (12) Guan, L.-J., Ohtsuka, J., Okai, M., Miyakawa, T., Mase, T., Zhi, Y., Hou, F., Ito, N., Iwasaki, A., Yasohara, Y., and Tanokura, M. (2015) A new target region for changing the substrate specificity of amine transaminases. Sci Rep 5, 10753. (13) Kariskoga, C. (2007, January 29) Thermostable omega-transaminases. US Patent Office. (14) Koszelewski, D., Göritzer, M., Clay, D., Seisser, B., and Kroutil, W. (2010) Synthesis of Optically Active Amines Employing Recombinant ω-Transaminases in E. coli Cells. ChemCatChem 2, 73–77. (15) Miroux, B., and Walker, J. E. (1996) Over-production of proteins in Escherichia coli: mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J. Mol. Biol. 260, 289–298. (16) Petoukhov, M. V., Franke, D., Shkumatov, A. V., Tria, G., Kikhney, A. G., Gajda, M.,


23

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 24 of 39

Gorba, C., Mertens, H. D. T., Konarev, P. V., Svergun, D. I., IUCr. (2012) New developments in the ATSAS program package for small-angle scattering data analysis. J Appl Crystallogr 45, 342–350. (17) Kabsch, W. (2010) XDS. Acta Crystallogr. D Biol. Crystallogr. 66, 125–132. (18) Evans, P. (2006) Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 62, 72–82. (19) Winn, M. D., Ballard, C. C., Cowtan, K. D., Dodson, E. J., Emsley, P., Evans, P. R., Keegan, R. M., Krissinel, E. B., Leslie, A. G. W., McCoy, A., McNicholas, S. J., Murshudov, G. N., Pannu, N. S., Potterton, E. A., Powell, H. R., Read, R. J., Vagin, A., and Wilson, K. S. (2011) Overview of the CCP4 suite and current developments. Acta Crystallogr. D Biol. Crystallogr. 67, 235–242. (20) Moshides, J. S. (1988) Enzymic determination of the free cholesterol fraction of highdensity lipoprotein in plasma with use of 2,4,6-tribromo-3-hydroxybenzoic acid. Clin. Chem. 34, 1799–1804. (21) McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C., and Read, R. J. (2007) Phaser crystallographic software. J Appl Crystallogr 40, 658–674. (22) Langer, G. G., Cohen, S. X., Lamzin, V. S., and Perrakis, A. (2008) Automated macromolecular model building for X-ray crystallography using ARP/wARP version 7. Nature Protocols 3, 1171–1179. (23) Emsley, P., Lohkamp, B., Scott, W. G., and Cowtan, K. (2010) Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501. (24) Murshudov, G. N., Skubák, P., Lebedev, A. A., Pannu, N. S., Steiner, R. A., Nicholls, R. A., Winn, M. D., Long, F., and Vagin, A. A. (2011) REFMAC5 for the refinement of


24

Page 25 of 39

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

macromolecular crystal structures. Acta Crystallogr. D Biol. Crystallogr. 67, 355–367. (25) Winn, M. D., Isupov, M. N., and Murshudov, G. N. (2001) Use of TLS parameters to model anisotropic displacements in macromolecular refinement. Acta Crystallogr. D Biol. Crystallogr. 57, 122–133. (26) Lebedev, A. A., Young, P., Isupov, M. N., Moroz, O. V., Vagin, A. A., and Murshudov, G. N. (2012) JLigand: a graphical tool for the CCP4 template-restraint library. Acta Crystallogr. D Biol. Crystallogr. 68, 431–440. (27) Schüttelkopf, A. W., and van Aalten, D. (2004) PRODRG: a tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallogr. D Biol. Crystallogr. 60, 1355– 1363. (28) Chen, V. B., Arendall, W. B., Headd, J. J., Keedy, D. A., Immormino, R. M., Kapral, G. J., Murray, L. W., Richardson, J. S., and Richardson, D. C. (2010) MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. D Biol. Crystallogr. 66, 12–21. (29) Schrödinger. The PyMOL Molecular Graphics System. LLC. (30) Metzler, C. M., Harris, A. G., and Metzler, D. E. (1988) Spectroscopic studies of quinonoid species from pyridoxal 5'-phosphate. Biochemistry 27, 4923–4933. (31) Krissinel, E. (2010) Crystal contacts as nature's docking solutions. J Comput Chem 31, 133– 143. (32) Eliot, A. C., and Kirsch, J. F. (2004) Pyridoxal phosphate enzymes: mechanistic, structural, and evolutionary considerations. Annu. Rev. Biochem. 73, 383–415. (33) Toney, M. D. (2011) Controlling reaction specificity in pyridoxal phosphate enzymes. Biochim. Biophys. Acta 1814, 1407–1418. (34) Höhne, M., Schätzle, S., Jochens, H., Robins, K., and Bornscheuer, U. T. (2010) Rational


25

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 26 of 39

assignment of key motifs for function guides in silico enzyme identification. Nat. Chem. Biol. 6, 807–813. (35) Wybenga, G. G., Crismaru, C. G., Janssen, D. B., and Dijkstra, B. W. (2012) Structural Determinants of the β-Selectivity of a Bacterial Aminotransferase. J.Biol.Chem. 287, 28495– 28502. (36) Krissinel, E., and Henrick, K. (2004) Secondary-structure matching (SSM), a new tool for fast protein structure alignment in three dimensions. Acta Crystallogr. D Biol. Crystallogr. 60, 2256–2268. (37) Weiss, M. S. (2001) Global indicators of X-ray data quality. J Appl Crystallogr 34, 130– 135.


26

Page 27 of 39

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

O

NH2 -transaminase

R1

(R/S)

R2

R1

NH2

R2

O -O

-O

O alanine

O pyruvate

Figure 1. The overall ω-transaminase reaction. The transaminase reaction consists of two half reactions, one using an amide acceptor and the other using an amide donor. Here the alaninepyruvate amide donor substrate-product pair is shown, but several other donor acceptor pairs can be used. Like α-transaminases, ω-transaminases can be (R)- or (S)-selective. However, in contrast to the α-enzymes, they can perform the transamination on substrates without a carboxylic acid group and can thus accept a wide range of substrates with various substituents on the R1 and R2 positions.


27

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 28 of 39

Figure 2. Relative activities of Ars-ωTA and BM-ωTA for a range of amine donors. The activity on 4-bromo-α-methylbenzylamine was taken as 100%. For most of the substrates both the (R)and the (S)- enantiomers were tested.


28

Page 29 of 39

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

a)

b)

Figure 3 - a) The BM-ωTA tetramer (dimer of dimers) found in the asymmetric unit. The four molecules are shown in red, blue, purple and green. In each BM-ωTA monomer a PLP molecule is bound covalently to Lys298 (both shown in CPK). b) The BM-ωTA dimer. One of the two molecules is coloured according to the secondary structure and the other monomer coloured blue. As in the dimeric ω-transaminases the active site


29

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 30 of 39

is formed by residues from two monomers. The PLP, shown in yellow, is covalently attached to Lys298 and bound at the dimer interface.

Figure 4. Tunnel leading up to the active site. The PLP cofactor is accessible by a ~ 15 Å deep tunnel in the protein. A surface representation of the native enzyme with bound PLP is shown in gray, and the other monomer that contributes to the active site is shown in blue. The internal aldimine is shown in gray. For clarity, the aligned (R)-alpha-MBA substrate (pink) has been added. The red circle marks the P pocket where the methyl group of (R)-α-MBA binds and the blue circle the O pocket where its phenyl group is located. Leu59 and Ala242 (not shown) are located above and below the plane of the phenyl group, respectively.


30

Page 31 of 39

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 5. The 1.9 Å structure of BM-ωTA dimer bound with (R)-α-MBA. One of the monomers is coloured blue, the other according to secondary structure as in Figure 3B. For clarity, the side chain of Val328-B is not shown. The external aldimine between the PLP cofactor and (R)-αMBA (yellow) is formed through a Schiff-base. The phenyl ring of (R)-α-MBA is bound between the side chains of gate residues Leu59 and Ala242. Lys298 is positioned below the adduct at 3.9 Å from the C7 carbon (yellow dashes), held in place by hydrogen bonds (red dashes) with the side chain Thr330-B and to the carboxyl group Leu59. It is positioned opposite to the C7 proton of (R)-α-MBA, explaining the minimal activity towards R-substrates. The two water molecules closest to the C7 carbon of the (R)-α-MBA are at 5.3 Å and 4.1 Å are shown (yellow dashes). Of these two, the one at 4.1 Å has a hydrogen-bonding pathway via the PLP phosphate and Thr330-B to lysine 298.


31

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 32 of 39

Lys298

1

Lys298

3

2 NH2

Lys298

NH3+ H

NH2

NH+

*

C4’

O

N+ H

N+ H

Internal aldimine

O-

O

O-

PO32-

CH3

NH+

NH+

-

PO32-

H2O

C7

CH3

External aldimine

OH

PO32N H

CH3

Quinonoid intermediate

Asp269

4

5

Lys298

H

O

H

Lys298

NH2

NH3+

H2N

NH+ H2O

O-

PO32N+ H

Ketimine

CH3

O

O-

PO32N+ H

CH3

PMP

Figure 6. Proposed reaction mechanism for Ars-ωTA and BM-ωTA based on our X-ray structures. Here the half-reaction for the reversible de-amination of the ωTA model substrate (S)α-MBA to acetophenone is shown. This reaction consists of several steps. 1) The amide group of (S)-α-MBA performs a nucleophilic attack on the C4’ atom of the PLP cofactor. Formation of the external Schiff base between the substrate and PLP frees Lys298. Its resulting neutral form allows Lys298 to act as a base in the second step. 2) Lys298 abstracts the proton from the C7 carbon of the substrate. Access of Lys298 to this proton is proposed to be a key step in the stereospecificity of the protein. The resulting negative charge on the substrate can be delocalized over the π system of the external aldimine. 3) A minor movement of the Nζ group of the lysine would then allow the donation of the proton to the C4’ atom of PLP. 4) A water molecule, possibly activated by Lys298, attacks the C7 atom of the substrate. 5) An internal rearrangement results in the formation of the keto product acetophenone, leaving the amine group on the PMP cofactor.


32

Page 33 of 39

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 7. Comparison of relative activities of variants at position 60 for Ars-ωTA and BM-ωTA with a range of amine substrates. The activity on 4-bromo-α-methylbenzylamine was taken as 100%.


33

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 34 of 39

Figure 8. - Comparison of relative activities of variants at position 242 for Ars-ωTA and BMωTA with a range of amine substrates. The activity on 4-bromo-α-methylbenzylamine was taken as 100%.


34

Page 35 of 39

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 9 - The arginine switch in Ars-ωTA. Structure of Ars-ωTA with alanine covalently bound to the PLP cofactor. The 2Fo-Fc electron density for alanine-PLP adduct (yellow) is contoured at 1.2 σ. Arg442 is shown both in the “in” position (alanine-bound structure - green) and in the “out” position (native structure - teal).


35

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 36 of 39

Table 1. Data collection statistics BM-ωTA

BM-ωTA

Ars-ωTA

Ars-ωTA

native

(R)-α-MBA

native

Alanine

Beamline

PX3 (SLS)

ID23-2 (ESRF)

ID14-1 (ESRF)

ID23-2 (ESRF)

Wavelength (Å)

1.0

0.87260

0.93340

1.710

Space group

P212121

P212121

P21

P1

a (Å)

119.1

119.3

64.4

64.3

b (Å)

124.6

124.6

142.7

99.8

c (Å)

126.8

127.0

99.7

217.3

α (º)

90

90

90

81.6

β (º)

90

90

105.3

89.1

γ (º)

90

90

90

75.5

Resolution (Å)

43.4-1.7 (1.80-1.70)

44.5-1.9 (2.00-1.90)

48.1-1.9 (2.00-1.9)

43.0-2.3 (2.42-2.30)

Rmerge†

3.1 (38.6)

9.3 (62.1)

9.1 (64.1)

6.8 (41.4)

Rp.i.m††

1.7 (21.4)

5.0 (33.4)

5.0 (35.5)

6.2 (39.2)

Wilson B-factor

22.04

22.93

18.17

29.81

Total No. of observations

847462 (110578)

649835 (92531)

580893 (82384)

450169 (57344)

Total No. of unique reflections

200029 (26554)

149076 (21375)

137172 (19735)

214456 (29180)

Mean I/σ(I)

27.4 (3.6)

11.0 (2.5)

14.1 (2.2)

8.5 (1.8)

Completeness (%)

96.9 (89.0)

99.8 (98.8)

99.8 (92.4)

93.8 (87.5)

Multiplicity

4.2 (4.0)

4.4 (4.3)

4.2 (4.2)

2.1 (2.0)

Unit-cell parameters

Values within parentheses are for the last resolution shell and the formulas for Rmerge and Rp.i.m. were taken from Weiss et al. 37.


36

Page 37 of 39

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

Table 2. Refinement statistics BM-ωTA native Refinement Resolution (Å) R† Rfree† Number of atoms Protein Ligands Water B-factors Protein Solvent RMSD deviations Bond lengths (Å) Bond angles (°)

†

R = ∑ Fobs − Fcalc hkl

BM-ωTA

Ars-ωTA

(R)-α-MBA

native

Ars-ωTA Alanine

43.4-1.7 17.3 20.0

44.5-1.9 16.6 20.3

48.1-1.9 15.6 20.1

48.0-2.3 19.2 24.9

14911 187 1096

15964 225 918

14571 4 1165

43519 252 790

32.9 39.2

24.9 28.4

23.2 29.3

74.7 65.4

0.01 1.32

0.01 1.42

0.01 1.40

0.01 1.52

∑ Fobs , Rfree is the R-factor calculated with 5% of the reflections hkl

excluded from the refinement.


37

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 38 of 39

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Phone: +31503634381 Present Addresses †If an author’s address is different than the one given in the affiliation line, this information may be included here. Funding Sources This research was funded from the European Union Seventh Framework Programme (FP7/20072013) under grant agreement n° KBBE-245144.


38

Page 39 of 39

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

For Table of Contents Use Only

Structural basis of substrate range and enantioselectivity of two (S)-selective ωtransaminases Niels van Oosterwijk, Simon Willies, Johan Hekelaar, Anke C. Terwisscha van Scheltinga, Nicholas J. Turner and Bauke W. Dijkstra


39

Structural Basis of the Substrate Range and Enantioselectivity of

Recommend Documents