Engineering Erg10 Thiolase from Saccharomyces cerevisiae as a

Subscriber access provided by RYERSON UNIVERSITY

Article

Engineering Erg10 thiolase from Saccharomyces cerevisiae as a synthetic toolkit for the production of branched-chain alcohols Pamela Torres-Salas, Vicente Bernal, Fernando López-Gallego, Javier Martínez-Crespo, Pedro Alejandro Sánchez-Murcia, Víctor Barrera, Rocío Morales-Jiménez, Ana GarcíaSánchez, Aurora Mañas-Fernández, José Miguel Seoane, Marta Sagrera Polo, Juande D. Miranda, Javier Calvo, Sonia Huertas, José Luis Torres, Ana Alcalde-Bascones, Sergio González-Barrera, Federico Gago, Antonio Morreale, and María del Mar González-Barroso Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.7b01186 • Publication Date (Web): 23 Jan 2018 Downloaded from http://pubs.acs.org on January 28, 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 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 38 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

Engineering Erg10 thiolase from Saccharomyces cerevisiae as a synthetic toolkit for the production of branched-chain alcohols Pamela Torres-Salas1,*,‡, Vicente Bernal1,*,‡, Fernando López-Gallego1,2,3, Javier MartínezCrespo1, Pedro A. Sánchez-Murcia4, Víctor Barrera1, Rocío Morales-Jiménez1, Ana GarcíaSánchez1, Aurora Mañas-Fernández1, José M. Seoane1, Marta Sagrera Polo5, Juande D. Miranda1, Javier Calvo2, Sonia Huertas1, José L. Torres1, Ana Alcalde-Bascones1, Sergio González-Barrera1, Federico Gago4, Antonio Morreale1, María del Mar González-Barroso1 1

Centro de Tecnología de Repsol. REPSOL S.A. Calle Agustín de Betancourt, s/n. 28935.

Móstoles, Madrid (Spain). 2 CIC biomaGUNE. Paseo de Miramón n 182. 20014. San Sebastián, (Spain). 3 ARAID Foundation, Zaragoza, (Spain). 4 Departamento de Ciencias Biomédicas and “Unidad Asociada IQM-CSIC”, Universidad de Alcalá, E-28805 Alcalá de Henares, Madrid (Spain). 5Centro de Investigaciones Biológicas (CSIC), Calle Ramiro de Maeztu 9, 28040 Madrid. * Corresponding authors: P. Torres-Salas, V. Bernal. KEYWORDS Synthetic biology, metabolic engineering, protein engineering, thiolase, branched compounds, 2-ethyl-1-butanol.

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

ABSTRACT:

Page 2 of 38

Thiolases catalyze the condensation of acyl-CoA thioesters through the Claisen

condensation reaction. The best described enzymes usually yield linear condensation products. Using a combined computational/experimental approach, and guided by structural information, we have studied the potential of thiolases to synthesize branched compounds. We have identified a bulky residue located at the active site that blocks proper accommodation of substrates longer than acetyl-CoA. Amino acid replacements at such position exert effects on the activity and product selectivity of the enzymes that are highly dependent on protein scaffold. Among the set of five thiolases studied, Erg10 thiolase from Saccharomyces cerevisiae showed no acetylCoA/butyryl-CoA branched condensation activity, but variants in position F293 resulted the most active and selective biocatalysts for this reaction. This is the first time that a thiolase has been engineered to synthesize branched compounds. These novel enzymes enrich the toolbox of combinatorial (bio)chemistry, paving the way for manufacturing a variety of α-substituted synthons. As a proof of concept, we have engineered Clostridium’s 1-butanol pathway to obtain 2-ethyl-1-butanol, an alcohol which is interesting as branched model compound.


2

Page 3 of 38 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 Despite the growing interest in biomanufacturing chemicals, fuels and materials with novel properties, the natural-occurring metabolism of microorganisms is not always sufficient to obtain any desired structure. Synthetic biology has emerged as a novel paradigm in biotechnology, allowing the production of non-natural products by creating new capabilities1–5. Many of these new synthetic approaches rely on engineered enzymes. Protein engineering allows expanding the space of chemical reactions catalyzed by enzymes6,7. In Nature, enzymes from the thiolase superfamily catalyze the Claisen condensation of thioesters. Among these, strictly coenzyme A-dependent enzymes catalyze the reaction in both directions, with thiolysis being thermodynamically favored8,9. Two subfamilies are distinguished based on substrate specificity and physiological role. Thiolase I subfamily (E.C. 2.3.1.16) encompasses degradative thiolases, which catalyze the removal of an acetyl group from βketoacyl-CoA in fatty acid β-oxidation10 with broad substrate range (C4-C16). Thiolase II subfamily (E.C.2.3.1.9) covers synthetic enzymes which have strict specificity for the condensation of acetyl-CoA for the synthesis of the C4 product acetoacetyl-CoA9, substrate for the synthesis of terpenoids. The mechanism of Zoogloea ramigera thiolase comprises two consecutive steps. In the first step, an acyl group is transferred from the substrate to Cys89. Then, this acyl group is shifted to a second substrate by means of a Claisen condensation that represents the rate limiting step11,12. The Claisen condensation starts by the abstraction of a proton from the incoming acyl-CoA by Cys378, giving rise to an enolate intermediate that is stabilized by oxyanion hole I. Next, the nucleophilic enolate attacks the carbonyl carbon from the acylated Cys89 forming a C-C bond with concomitant release of Cys89. During this reaction step, a negatively charged intermediate


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 38

is created and stabilized by the so-called oxyanion hole II. Finally, His348 donates a proton to Cys89, so that the catalytic site becomes neutral, releasing the acylated substrate and rearranging for the next catalytic cycle. In agreement with this mechanism, synthesis of branched carbon skeletons requires at least three-carbons long acyl-CoAs as second substrate. Thiolases have different substrate specificities. Some thiolases can accept longer acyl-CoAs, giving rise to linear condensation products. This is the case of BktB from Cupriavidus necator H16 (formerly Ralstonia eutropha13), which is able to condense acetyl-CoA with propionyl-CoA and butyryl-CoA14,15 to yield linear compounds. The ability of thiolases to condense various structurally diverse acyl-CoAs has been recently exploited for the synthesis of a plethora of functionalized molecules16. Few works have dealt with the use of thiolases to synthesize branched compounds: the cleavage of 3-keto-2-branched acyl-CoAs have been described17–19 and the branched condensation activity has been demonstrated using native enzymes, although showing low selectivity16,20. Therefore, the design of efficient and selective (bio)catalysts for the synthesis of branched carbon skeletons may fill an unmet need in biotechnology that would open new avenues for the synthesis of branched compounds. In this work, we engineered thiolases to synthesize branched compounds from acyl-CoAs. Using a combined computational/experimental strategy, we identified the key residues to achieve this novel activity. As proof of concept, we designed and implemented a new pathway for the production of 2-ethyl-1-butanol in E. coli based on Clostridium’s 1-butanol route (Scheme 1), to demonstrate the usefulness of these new enzymes in synthetic biology.


4

Page 5 of 38 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

Scheme 1 Engineered pathway for 2-ethyl-1-butanol synthesis in E. coli and competing pathways.

Thl, thiolase; Hbd, 3-hydroxybutyryl-CoA dehydrogenase; Crt, crotonase; Bcd, butyryl-CoA dehydrogenase; Ter, trans-enoyl-CoA reductase; Aldh, aldehyde dehydrogenase; Adh, alcohol dehydrogenase; 3KB-CoA, 3-ketobutyryl-CoA; 3HB-CoA, 3-hydroxybutyryl-CoA; crot-CoA, crotonyl-CoA; But-CoA, butyryl-CoA; 3K2EB-CoA, 3-keto-2-ethylbutyryl-CoA; 3H2EB-CoA, 3-hydroxy-2-ethylbutyryl-CoA; 2E-crot-CoA, 2-ethylcrotonyl-CoA; 2EB-CoA, 2-ethylbutyrylCoA; 2EBd; 2-ethylbutyraldehyde; 3KH-CoA, 3-ketohexanoyl-CoA; 3HH-CoA, 3hydroxyhexanoyl-CoA; 2-enH-CoA, 2-hexenoyl-CoA; Hex-CoA, hexanoyl-CoA.


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 38

MATERIALS AND METHODS Detailed methods (and associated references) are available in SI Sections 1 to 8. Bacterial strains and plasmids. E. coli BL21(DE3)T1 strain was used as protein expression host. A derivative of E. coli K-12 BW25113 strain was used as parental strain for metabolic engineering studies of the 2-ethyl-1butanol production pathway. All protein genes used throughout this work were cloned into pET family vectors (pET28b, pET32, pCDF-Duet-1) and expressed either using IPTG or autoinduction media for protein induction. For a complete description of bacterial strains and plasmids, refer to SI Section 1. Expression and purification of proteins Expression and purification were carried out either in standard (full scale) assays or in high throughput assays as described in the following sections. Standard expression and purification of thiolases Cultures in 50 mL of NZY Auto-Induction LB medium (NZYTech, Portugal) were supplemented with 100 µg/mL Kan or Amp (according to plasmid used) and inoculated to an OD600nm of 0.1. The cultures were grown aerobically at 37°C for 7 h and at 25°C until 24 h. Hereafter the cells were collected by centrifugation (4,500g, 10 minutes) and kept at -20°C until further purification. Pellets (50 mL) were homogenized with 50 mM phosphate pH 7.5 (10 mL) and placed in an ice bath for sonication (10 minutes, pulse mode, 5 seconds ON/OFF, 20% amplitude). Lysates were centrifuged at 4°C, 10,000g for 30 minutes. Supernatants containing the fusion protein


6

Page 7 of 38 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

were purified by IMAC using Talon® resin. Proteins were eluted in 1 mL elution buffer (300 mM imidazole, 50 mM sodium phosphate pH 7.5), quantified and used for enzyme reaction. High throughput expression and purification of thiolases Cultures were carried out as in the standard expression and purification procedure, but at 5 mL scale. After growth and induction, cells were aliquoted in 96-deep-well plates and centrifuged at 4,500g for 10 minutes. Pellets were kept at -20°C until further purification. For lysis, pellets were resuspended with 250 µL lysis buffer (50 mM phosphate buffer pH 7.5, Bugbuster®, 10 µg/mL DNase I, 20 mM MgSO4) and incubated for 20 minutes. Lysates were centrifuged at 10,000g for 30 minutes. Supernatants containing the fusion protein were incubated with 100 µl Talon® resin for 15 minutes. Proteins were eluted with 50 µL elution buffer (300 mM imidazole, 50 mM sodium phosphate buffer pH 7.5), quantified and used for enzyme reaction. Expression

and

purification

of

3-hydroxy-2-methylbutyryl-CoA-dehydrogenase

from

Pseudomonas putida (PpFadB2) Protein expression was carried out as described in the standard procedure. IMAC-purified protein was eluted with 3 mL elution buffer (300 mM imidazole, 50 mM sodium phosphate pH 7.5), dialysed and concentrated by 10 kDa MWCO ultrafiltration membrane (Amicon®) to a volume of ca. 0.5 mL. Purified protein was quantified and kept at 4°C until used for enzyme reaction. Expression and purification of the 2-ethyl-1-butanol pathway enzymes Protein expression was carried out as described in the standard procedure, but using 50 mL of LB medium (Sigma), supplemented with 50 µg/mL Kan or 100 µg/mL Amp (according to the corresponding plasmid). The cultures were grown aerobically at 37°C until OD600nm reached 0.4-


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 38

0.6 and induced with 0.1 mM IPTG. Temperature was lowered to 25°C and cultures were incubated overnight for protein expression. Hereafter, the cells were harvested by centrifugation at 4,500g for 10 minutes and stored at -20°C until further purification. Proteins were extracted by sonication and IMAC-purified. For enzyme assays, buffer was exchanged with phosphate buffer (pH 7.5) by three rounds of filtration through Vivaspin Turbo 15 (Sartorius) 10 KDa cut-off filters. Proteins were quantified and used for enzyme reactions. In vitro thiolase enzymatic assay: acetyl-CoA/butyryl-CoA condensation assay The acetyl-CoA/butyryl-CoA condensation activity of thiolase variants was evaluated in the synthetic direction by a coupled enzyme assay with product profiling by HPLC-MS. The condensation reaction was coupled with 3-hydroxy-2-methylbutyryl-CoA dehydrogenase from Pseudomonas putida (PpFadB2) to drive the reaction in the forward direction. Reactions were carried out in 2-(N-morpholino)ethanesulfonate (MES) buffer (pH 6.0), 100 µM acetyl-CoA, 100 µM butyryl-CoA, 1 mM NADH, 0.15 mg/mL PpFadB2 and 0.19 mg/mL of thiolase variant. Reaction mixtures were prepared in 96 well microplates in a 200 µL volume, and incubated for 1 h at 37°C with agitation (300 rpm in Thermomixer). The enzyme was removed by filtration through 10 KDa cut-off filtration plates (Millipore). Samples were stored at -20°C until further analysis. The activity of enzyme variants was reported as the titer of product (3HB-CoA, 3H2EB-CoA or 3HH-CoA) after 1 hour of reaction at 37°C. UPLC and mass spectrometry detection of coenzyme A thioesters Chromatography was performed in an Acquity UPLC system using an Acquity BEH C18 column (100x 2.1 mm, 1.7 µm) from Waters (Milford, MA, USA) and equipped with photodiode array detector (PDA). The gradient elution buffers were A (100 mM of ammonium formate in water) and B (acetonitrile). The gradient method was: 0-1 minutes, isocratic at 95% A; 1-14


8

Page 9 of 38 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

minutes, gradient to 80% A; 14-15 minutes, gradient to 10% A; 15-17 minutes, isocratic at 10% A; 17-17.5 minutes, gradient back to 95% A; 17.5-20 minutes, stabilization at 95% A. The UV detector wavelength was set at 254 nm and the injection volume was 10 µL. Total run time was 20 minutes and the flow rate was set at 300 µL·min-1. The mass spectrometry detection was carried out using a time-of-flight mass spectrometer (ESI-TOF) LCT Premier XE from Waters (Milford, MA, USA) with an electrospray ionization source, working in positive /V mode. The MS range acquired was between m/z 100-1,000. The capillary and cone voltages were set at 3,000 and 100 V, respectively. Desolvation gas temperature was 220°C and source temperature was 120°C. The desolvation gas flow was set at 600 L·h-1 and cone gas flow was set at 50 L·h-1. For quantification and data analysis, Masslynx v4.1 software was used (Waters, Milford, MA, USA). All the analytes were identified by mass spectrometry. The quantification of 3HH-CoA and 3H2EB-CoA was performed by ESI-TOF MS. To avoid signal saturation, the quantification of acetyl-CoA, butyryl-CoA and 3HB-CoA was performed by UV absorbance at 254 nm. Fermentation culture conditions for 2-ethyl-1-butanol production experiments The screening of pathway constructions for production of 2-ethyl-1-butanol was carried out in Terrific Broth (TB), supplemented with 90 mM glucose following a modification of previously reported procedures21,22. Production cultures in TB medium supplemented with the appropriate antibiotics (Amp 100 µg/mL, Kan 100 µg/mL, Strep 50 µg/mL) were inoculated at an initial OD600nm of 0.1. The cultures were grown at 37°C with agitation (250 rpm) to an OD600nm of 0.40.6. Cultures were then induced with 0.1 mM IPTG and incubated for additional 3-5 h at 25°C. At that point, glucose was added to a final concentration of 90 mM and 60 mL of cultures were transferred to 50 mL Schott bottles and incubated at 25°C without agitation for up to 72 h. The


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 38

bottle stoppers were loosened in order to avoid pressure build up due to gas production. To analyze the fermentation profile in time course experiments, parallel cultures were set in 50 mL Schott bottles and one bottle was sacrificed at each sampling time. GC-MS analysis of aldehydes and alcohols Alcohols (1-butanol, 1-hexanol and 2-ethyl-1-butanol) and aldehydes (butyraldehyde and 2ethylbutyraldehyde) were analyzed by headspace GC-MS. Samples and standards were prepared in 10 mL headspace vials. NaCl (1 g/vial) was added to “salt-out” the solutes. An automated dynamic headspace system (DHS) installed on an MPS2/TDU unit (Gerstel GmbH, Mülheim an der Ruhr, Germany) in combination with a 7890GC – 5975MSD (Agilent Technologies, Wilmington, USA) system was used. During DHS, the sample (5 mL) is incubated at 90°C during 10 min with a flow of 50 mL/min helium while the vial is agitated at 250 rpm. Flush time: 60 s. The purged solutes are trapped at -30 °C on a TENAX adsorbent (Gerstel GmbH), placed in a TDU (thermal desorption unit) liner. After dynamic headspace extraction, the trap is desorbed at 250°C during 5 min and the released solutes are cryo-focussed in a programmable temperature vaporizer (PTV – CIS-4, Gerstel) interface operated at –30°C in the splitless mode using a Tenax packed liner. Finally, the CIS-4 is programmed to 250°C (3 min hold) for injection in splitless mode. Separation was done on a 30 m x 250 µm x 0.25 µm HP Innowax 3352.69405 column (Agilent Technologies) using helium as carrier gas at 1.0 mL/min (4.7564 psi at 40°C). The GC oven was programmed from 40°C (10 min) at 5°C/min to 150°C and at 50°C/min to 200°C (10 min). MS-detection was done in synchronous acquisition mode: Scan/SIM. Scan mode: Low Mass: 20.0; High Mass: 300.0. SIM mode using ions 57 and 72 for butyraldehyde with dwell times of 40 ms for each ion, 57 and 72 for 2-ethylbutanal with dwell times of 40 ms for each ion, 41 and


10

Page 11 of 38 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

56 for butanol with dwell times of 100 ms for each ion, 43, 56 and 70 for 2-ethylbutanol with dwell times of 100 ms for each ion and. Electron ionization (70 eV) at 230°C source temperature was used. Computational methods Model Systems. All thiolase enzymes were simulated as dimers, with the nucleophilic Cys residue acetylated and in complex with acetyl-CoA or butyryl-CoA as ligands. Biosynthetic thiolases (Erg10 and FbThl) were modeled through Phyre 2.023 Web server using as template the X-ray structure of the biosynthetic thiolase from Z. ramiguera (PDB entry 1DM3). For degradative thiolases AtoB, BktB and TtThl, their corresponding crystal structures were used (PDB entries 5F0V, 4NZS and 1ULQ, respectively). Protonation state of His, Asp and Glu residues was calculated with H++ Web server24 using pH=7.0. Butyryl-CoA was modeled and placed into the active sites of the former enzymes in the same conformation as acetyl-CoA at the active site of the biosynthetic thiolase of Z. ramiguera. PpFadB2, CaCrt, PfAldh and LlAdh were modelled with Phyre 2.0 Web server using PDB entries 4O5O, 1MJ3, 4C3S and 3KRT, respectively, as templates, and the X-ray structure (PDB entry 4GGO) for TdTer. To generate the structures for the combinatorial library of variants, an automatic pipeline for obtaining the best rotamer of a given variant at a specific position in a given structure was built. We used a modified version of the mutate.py code (Thomas Holder, MPI for Developmental Biology, 2010; modified by V. Barrera, Repsol SA, 2016) that allows saving the coordinates for this rotamer. Specific positions on BktB enzyme (Leu89, Met290, Ile352 and Met379) were substituted by Ala, Met, Val, Leu, Ile, Phe, Ser and Thr residues. The rest of the variants were built using PyMOL25. For the sake of clarity, we will refer to residue numbers from one of the


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 38

two catalytic sites. However, all the analysis was performed taking into account residues from both sites.


12

Page 13 of 38 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

RESULTS Exploring the acetyl-CoA/butyryl-CoA condensation ability of wild type thiolases We selected five thiolases with low sequence identity as potential biocatalysts for the acetylCoA/butyryl-CoA branched condensation (Figure 1, SI Section 9), including representatives of the synthetic (Erg10 and FbThl) and degradative (BktB, AtoB and TtThl) subfamilies. Their ability to condense different acyl-CoAs was evaluated in a coupled assay with 3-hydroxy-2methylbutyryl-CoA dehydrogenase from Pseudomonas putida (PpFadB2)17. Previously, the ability of PpFadB2 to use both 3-ketobutyryl-CoA or 3-keto-2-ethylbutyryl-CoA as substrate was experimentally confirmed (SI Section 15). Depending on the order of substrate binding, thiolases may yield three different products from the acetyl-CoA and butyryl-CoA condensation reaction: (i) 3-hydroxybutyryl-CoA (3HB-CoA), through the condensation of two acetyl-CoA molecules; (ii) 3-hydroxyhexanoyl-CoA (3HHCoA), through linear butyryl-CoA/acetyl-CoA condensation,; and (iii) 3-hydroxy-2-ethylbutyrylCoA (3H2EB-CoA) through an unprecedented branched acetyl-CoA/butyryl-CoA condensation (Scheme 1). Product selectivity of enzymes was defined across this study as the ratio between the molar concentration of a desired product and the sum of molar concentrations of all three products quantified (3HB-CoA, 3HH-CoA and 3H2EB-CoA), expressed in percentage. From the five tested thiolases, FbThl and Erg10 were selective for the synthesis of 3HB-CoA. On the contrary, BktB, AtoB and TtThl also produced 3HH-CoA (Figure 2b). Remarkably, trace amounts of 3H2EB-CoA were detected in the reactions catalyzed by BktB and TtThl, the results being slightly more consistent in the former (Figure S4). In the light of these results, BktB turned into the most promising candidate to be rationally engineered for the production of 3H2EB-CoA.


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 38

Figure 1. Thiolases used in this study. (a) Identification, classification and properties (physiological function and annotated EC numbers) of the 5 thiolases selected, based on information annotated in public databases KEGG26, Biocyc/Metacyc27 and BRENDA28. (b) Heatmap of the sequence homology of the 5 thiolases. The percentage of sequence identity is given by the colour scale shown in the figure. The heatmap was built using pheatmap R package. (c) Phylogenetic relationships between the enzymes. Phylogenetic tree was constructed using the online tool www.phylogeny.fr 29.

Thiolase engineering: computer-guided residue identification, variants modelling, ranking and experimental assessment Site-directed mutagenesis experiments were planned following a rational-design approach aided by molecular modelling of BktB (SI Section 10). After visual inspection and energy analysis of the interactions between butyryl-CoA and the amino acids forming the active site


14

Page 15 of 38 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 S10), we identified residues surrounding the butyryl moiety but not participating in substrate binding nor affecting the interaction network of catalytic residues. We selected Leu89, Met290, Ile352 and Met379 as potential hot positions for site-directed mutagenesis with a view to enlarging the active site. Since products generated in the assay cannot be easily discriminated using a high throughput adaptable assay and given the high sensitivity required, we resorted to HPLC-MS analysis to detect product titers in the range of nM. Given the throughput limitations imposed by HPLC-MS, we prioritized in silico variants using MD to reduce the experimental efforts. The selected residues of C. necator thiolase were replaced in silico by Ala, Met, Val, Leu, Ile, Phe, Ser and Thr in a combinatorial fashion (generating all 4096 possible variants). All variants were built with the aid of an in-house developed computational pipeline. Virtual variants were simulated (2 ns), analyzed and ranked using an ad hoc assembled scoring function (f) based on key distances and an angle crucial for the mechanism (SI Section 7). Accordingly, 10 variants were selected as representative of improved (6) or equivalent/worse (4) performance compared to the wild type enzyme. To improve the predictive power of the computational approach, we carried out extended MD simulations for the 10 selected variants (3 x 30 ns) and calculated new f scores. The new values placed BktB_Var10, with great difference, as the best variant (SI Section 10, Table S10). The

10

chosen

variants

were

experimentally

assayed

and

only

BktB_Var10

(BktB_L89I_M290A_M379L) catalyzed the acetyl-CoA/butyryl-CoA branched condensation, showing a 6-fold increase of activity regarding the wild type enzyme (Figure 2a). Visual inspection of a homology model of the 3D structure of the enzyme variant evidenced a better


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 38

accommodation of butyryl-CoA in the pre-reactive configuration of BktB_Var10 (data not shown). Characterization of acyl-CoA condensing activity of thiolases and the corresponding Var10 analogues We investigated the effect of BktB_Var10 mutations on the other thiolases studied herein (SI Section 11, Table S11). FbThl_Var10 displayed no acetyl-CoA/butyryl-CoA condensation activity, while TtThl_Var10 showed a 3H2EB-CoA activity close to the quantification limits. Consequently, these enzymes were no longer considered. Both AtoB_Var10 and Erg10_Var10 exhibited the highest acetyl-CoA/butyryl-CoA branched condensation activity, with product selectivity around 10% (Figure 2).

Figure 2. Engineering thiolases for acetyl-CoA/butyryl-CoA branched condensation. (a) 3H2EB-CoA product titers using wild type enzymes and their corresponding triple variants (Var10). (b) Product selectivities observed in the acetyl-CoA/butyryl-CoA condensation by different thiolase variants. Bars are mean values ± SE. All experiments were carried out at least in triplicates. Different letters show significant differences at p

Engineering Erg10 Thiolase from Saccharomyces cerevisiae as a

Recommend Documents