The role of streptavidin and its variants in catalysis by biotinylated secondary amines

Here, we combine the use of host screening, protein crystallography and QM/MM molecular dynamics simulations to investigate how the protein structure affects iminium catalysis by biotinylated secondary amines in a model 1,4 conjugate addition reaction. Monomeric streptavidin (M-Sav) lacks a quaternary structure and the solvent-exposed reaction site resulted in poor product conversion in the model reaction with low enantio- and regioselectivities. These parameters were much improved when the tetrameric host T-Sav was used; indeed, residues at the symmetrical subunit interface were proven to be critical for catalysis through a mutagenesis study. The use of QM/MM simulations and the asymmetric dimeric variant D-Sav revealed that both Lys121 residues which are located in the hosting and neighboring subunits play a critical role in controlling the stereoselectivity and reactivity. Lastly, the D-Sav template, though providing a lower conversion than that of the symmetric tetrameric counterpart, is likely a better starting point for future protein engineering because each surrounding residue within the asymmetric scaffold can be refined for secondary amine catalysis.

Plasmid miniprep-kit and gel extraction-kit were purchased from Qiagen. DNA oligos were purchased from Sigma-Aldrich.
T-Sav (Streptavidin Streptomyces avidinii recombinant, tetramer, M w ≈ 52 kDa, "core" Streptavidin with amino acids 13-139) was obtained commercially from ProSpec (PRO-791) as lyophilized powder in 10 mM KP i pH 6.5 and stored at −23 °C upon receipt until further use. According to the supplier T-Sav has the following amino acid sequence:
Racemic samples of S1 were obtained following a known procedure, using piperidine as catalyst. S1 was transformed into S2 by reduction with NaBH 4 in MeOH as previously reported (vide infra). Protein concentrations were determined using a Thermo Scientific NanoDrop One spectrophotometer measuring the absorption at 280 nm.

Expression and Purification
Tetrameric reduced streptavidin (T-rSav) and relative mutants were expressed using an E. coli expression system with the following protocol. Plasmid pTSA-13 containing the desired T-rSav gene in a pET-3a vector was transformed into calcium competent BL21(DE3) pLysS cells and grown for 16 h on LB agar plates containing 100 μg/mL ampicillin and 34 μg/mL chloramphenicol. A single colony from the plate was picked to inoculate a 15 mL MTP (per 1 L: 10 g tryptone, 10 g NaCl, 5 g yeast extract, 2.2 g Na 2 HPO 4 , 1 g KH 2 PO 4 , pH = 6.9) starter culture containing 100 μg/mL ampicillin and 34 μg/mL chloramphenicol, which was incubated at 37 °C and 180 rpm overnight. The culture was diluted to 40 mL with 20% glucose and then added to 1 L MTP medium containing 100 μg/mL ampicillin and 34 μg/mL chloramphenicol, yielding a final glucose concentration of 0.05%. The cultures were grown at 37 °C and 220 rpm to an OD 600 of 1.0−1.2 and induced with IPTG at a final concentration of 1 mM. The culture was grown at 25 °C for 16 h and the cell pellet was harvested after centrifugation at 4000 rpm at 4 °C for 25 min and stored at −20 °C.
The pellet was subjected to a freeze-thaw cycle, resuspended in 25 mL of lysis buffer 1 (50 mM Tris, 100 mM NaCl, 1 mM PMSF, pH 8.0) and lysed by sonication (7 min, 5 s on, 10 s off). The insoluble fraction was isolated by centrifugation at 15000 rpm for 25 min at 4 °C. The supernatant was discarded and the insoluble fraction was washed with wash buffer 1 (4× resuspension in 50 mM Tris, 110 mM EDTA, 1.5 M NaCl, 1 mM PMSF, 0.1% Triton X-100, pH 8.0 and pellet isolation by centrifugation at 11000 rpm for 15 min and 4 °C) and wash buffer 2 (4× resuspension in 50 mM Tris, 110 mM EDTA, 1.5 M NaCl, 1 mM PMSF, pH 8.0 and pellet isolation by centrifugation at 11000 rpm for 15 min and 4 °C). The insoluble fraction was resuspended in denaturing buffer 1 (5 mL / g pellet, 6 M GdnHCl, 50 mM Tris-HCl, pH 1.5) and incubated at 37 °C and 180 rpm for 16 h. The insoluble fraction was removed by centrifugation at 15000 rpm for 25 min and 4 °C. The supernatant was diluted to 200 mL with denaturing buffer 2 (6 M GdnHCl, 50 mM Tris-HCl, pH 6.5) and dialysed against 3 L of 6 M GdnHCl, 50 mM Tris-HCl, pH 6.5 for 3 h at room temperature. The dialysis bag (3.5 kDa cut-off) was then placed into fresh 3 M GdnHCl, 50 mM Tris-HCl, pH 6.5 (denaturing buffer 2 was reused up to 5 times). T-rSav was refolded by gradient dialysis, pumping in refolding buffer (0.5 mg/L catalyst 1, 10 mM KP i , pH 7.0) at 4 mL/min, constant stirring and removal of the mixture at 4 mL/min for 48 h at room temperature. Towards the end of this process a varying amount of precipitation was observed. The precipitate was removed via centrifugation at 15000 rpm for 25 min at 4 °C and the supernatant was concentrated to 20 mL by Amicon ultra centrifugation using a 3.5 kDa cutoff. The concentrated solution was transferred into a centrifugal concentrator with a 10 kDa cutoff and the buffer was exchanged five times by concentration to 2.5 mL and refilling to 20 mL (10 mM KP i , pH 7.0). The protein solution was finally concentrated to obtain a protein concentration of 2 mg/mL as determined by nanodrop measurement at 280 nm. This was used for catalysis of the Michael addition without further purification. A sample of the solution was loaded on SDS-PAGE to check the purity of the protein (15% w/v).
In case of purified T-rSav (WT) the concentrated protein solution obtained after Amicon ultra centrifugation was applied to size exclusion chromatography (Generon ProSEC 26/60 3-70 HR column, 10 mM KP i , pH 7.0). Fractions containing protein (analysis by following 215, 255 and 280 nm UV traces) were collected and the samples loaded on SDS-PAGE to check the purity of the protein (15% w/v). Fractions containing T-rSav were pooled, transferred to a centrifugal concentrator with a 10 kDa cutoff and the buffer was exchanged to 10 mM KP i , pH 7.0.). The protein solution was finally concentrated to obtain a protein concentration of 2 mg/mL as determined by nanodrop measurement at 280 nm.

Site-directed mutagenesis
The Leu124, Lys121, or Ser112 mutations were introduced by site-directed mutagenesis PCR using PrimeStar HS DNA polymerase (Takara) and the accompanying buffers, dNTPs and primers mentioned in Table S1 below. Due to the high GC content of the region of interest a variety of methods and temperatures had to be screened, as primer insertions were observed, especially for mutations at Lys121. Hence a 50 µL PCR was prepared according to the instructions and the reaction mixture distributed equally (12.5 µL) over 4 PCR tubes. These were then subjected to the following conditions, using a gradient to achieve a different annealing temperature for each tube: In case of the Lys121 mutations, Method 1 and 3 were also applied using 3% DMSO, if no positive results were obtained without DMSO. The mutant constructs were confirmed by DNA sequencing (Eurofins Genomics) using the T7 promoter primer. Table S1. List of primers used for the introduction of mutations in T-rSav at positions Ser112, Lys121, and Leu124.

S112E
Forward Experimental Details for the Preparation and Purification of D-Sav and mutants.

Expression and Purification
Dimeric Streptavidin (D-Sav) and relative mutants were expressed using an E. coli expression system with the following protocol. Plasmid pRSF-scdSav(SARK)mv2 1 containing the desired D-Sav gene was transformed into calcium competent BL21(DE3) cells and grown for 16 h on LB agar plates containing 50 μg/mL kanamycin. A single colony from the plate was picked to inoculate a 100 mL LB starter culture containing 50 μg/mL kanamycin, which was incubated at 37 °C and 180 rpm overnight. 10 mL of the culture were added into 1 L LB medium containing 50 μg/mL kanamycin. The cultures were grown at 37 °C and 220 rpm to an OD 600 of 0.6−0.8 and induced with IPTG at a final concentration of 1 mM. The culture was grown at 25 °C for 16 h and the cell pellet was harvested after centrifugation at 4000 rpm at 4 °C for 25 min and stored at −20 °C.
The pellet of a 1 L culture was subjected to a freeze-thaw cycle, resuspended in 50 mL of lysis buffer 2 (20 mM Tris, pH 7.4, 1 mg/mL lysozyme, 4 μg/mL DNase-I) and stirred vigorously at room temperature (20 °C) for 3 h. The soluble lysate was isolated by centrifugation at 15000 rpm for 25 min at 4 °C. The supernatant was diluted to 200 mL with denaturing buffer 2 (6 M GdnHCl, 50 mM Tris-HCl, pH 6.5) and dialysed against 3 L of 6 M GdnHCl, 50 mM Tris-HCl, pH 6.5 for 12 h at 4 °C. The dialysis bag (3.5 kDa cut-off) was then placed into 3 L of neutralising buffer (20 mM Tris, pH 7.4) for 12 h at 4 °C and finally into 3 L of imino-biotin binding buffer (50 mM NaHCO 3 , 500 mM NaCl, pH 9.8) for 12 h at 4 °C. The lysate was subjected to centrifugation at 15000 rpm for 25 min at 4 °C. The supernatant was filtered using 0.22 μm syringe filters and the filtrate was incubated with 5 mL imino-biotin resin (Thermo Scientific Pierce™ Iminobiotin Agarose) for 30 min at 4 °C with gentle mixing. The resin was washed with imino-biotin binding buffer (2× 1.5 resin volume) and the protein was eluted with 1% acetic acid (5× resin volume). The elution fractions were collected in ice-cold 50 mL centrifuge tubes containing 5 mL Tris buffer (1M, pH 8.0). Samples of the wash and elution fractions were collected and run on SDS-PAGE gel (15% w/v). Elution fractions containing D-Sav were pooled and dialyzed 5 times for 12 h against 20× the volume against deionised water at 4 °C. The protein was isolated after freezing with liquid nitrogen and lyophilisation. The lyophilized D-Sav was stored at 4 °C.

Site-directed mutagenesis
The Lys121 A and Lys121 B mutations were introduced by site-directed mutagenesis PCR using PrimeStar HS DNA polymerase (Takara) and the accompanying buffers, dNTPs and primers mentioned in Table S2 below. A 50 µL PCR was prepared according to the instructions and the reaction mixture distributed equally (12.5 µL) over 4 PCR tubes. These were then subjected Method 2 or Method 3, see 2.3.
The mutant constructs were confirmed by DNA sequencing (Eurofins Genomics) using the T7 promoter and terminator primers. Table S2. List of primers used for the introduction of mutations in D-Sav at positions Lys121 A and Lys121 B .

Mutation
Primer (5' to 3') Experimental Details for the Preparation and Purification of M-Sav and mutants.

Expression and Purification
Monomeric streptavidin (M-Sav) and relative mutants were expressed using an E. coli expression system with the following protocol. Plasmid pRSET-mSA containing the desired M-Sav gene in a pRSET-A vector was transformed into calcium competent BL21 AI cells and grown for 16 h on LB agar plates containing 100 μg/mL ampicillin. A single colony from the plate was picked to inoculate a 5 mL LB culture overnight. The starter culture was diluted into 500 mL of LB medium containing 100 μg/mL ampicillin. The culture was grown at 37 °C and 225 rpm to an OD 600 of 0.8−1.0 and induced with a final concentration of 0.5 % w/v Larabinose. The culture was grown overnight at 20 °C. The pellet was harvested by centrifugation at 4000 rpm at 4 °C for 25 min, resuspended in 10 mL of wash buffer 3 (50 mM Tris-HCl, 100 mM NaCl, pH 8.0) and lysed via sonication (7 min, 5 s on, 10 s off). 10 mL of lysate buffer 1 (50 mM Tris-HCl, 100 mM NaCl, and 6 M GdnHCl, pH 8) were added to the suspension and left to dissolve incubating 16 hours at 4 °C. The insoluble fraction was removed by centrifugation at 4000 rpm at 4 °C for 25 min, and the supernatant was mixed with 3 mL of Ni-NTA affinity resin for six-His affinity purification. After incubation at 25 °C for 1.5 h with occasional stirring, the resin was washed twice with 1.5 volumes of resin wash buffer 1 (50 mM Tris-HCl, 100 mM NaCl, 6 M GdnHCl, and 10 mM imidazole, pH 7.5). M-Sav was eluted with elution buffer (5× resin volume, 50 mM Tris-HCl, 150 mM NaCl, 6 M GdnHCl, and 300 mM imidazole, pH 8.0). Samples of the wash and elution fractions were collected and run on SDS-PAGE gel (15% w/v). The elution fractions were added drop by drop to 5 times their total volume of ice-cold refolding buffer 2 (50 mM Tris-HCl, 150 mM NaCl, 0.3 mg/mL catalyst 1, 2, or biotin, 0.2 mg/mL oxidized glutathione, and 1 mg/mL reduced glutathione) under rapid stirring to refold the protein by stirring overnight. The precipitates were removed by centrifugation at 15000 rpm at 4 °C for 25 min. The refolded protein solution was concentrated to 5 mL using Amicon ultra centrifugation with a 10 kDa cut-off. The concentrated protein solution was applied to size exclusion chromatography (Hi-Load TM 16/600 Superdex 200 pg, 50 mM Tris-HCl, 150 mM NaCl, pH 8.0). Fractions containing protein (analysis by following 215, 255 and 280 nm UV traces) were collected and the samples loaded on SDS-PAGE to check the purity of the protein (15% w/v). Fractions containing M-Sav were pooled, transferred to a centrifugal concentrator with a 10 kDa cut-off and the buffer was exchanged to the respective buffer (see 5.2 and see 14.1, Tables S8-S10). The protein solution was finally concentrated to obtain a protein concentration of 2 mg/mL as determined by nanodrop measurement at 280 nm.

Site-directed mutagenesis
The mutations were introduced by site-directed mutagenesis PCR using PrimeStar HS DNA polymerase (Takara) and the accompanying buffers, dNTPs and primers mentioned in Table  S3 below. For single mutants at Y111 the mentioned forward primers are used in conjunction with reverse primer GGTCAGGTTCCACTGGTGTTG. For all single mutants, the following protocol was used:

Circular Dichroism experiments
The circular dichroism spectra of M-Sav were recorded on a Chirascan TM CD spectrophotometer (Applied Photophysics) with a temperature controller using a cuvette with a path length of 1 mm. A spectral bandwidth of 1 nm was used for data collection. M-Sav was dissolved in PBS buffer (pH 7.4) at a concentration of 50 μM using 500 μM of the appropriate ligand. The blank measurement was prepared using 500 μM of the appropriate ligand or just using the buffer. To induce heat denaturation, the temperature was increased from 4 to 96 °C, and the CD spectra were recorded between 200 and 400 nm for every 2 °C increment with 1 nm steps. Melting temperatures were checked at 234 nm. The tube was taken out of the incubator and the aqueous mixture was extracted once with CDCl 3 (700 µL). The organic phase was transferred into an NMR tube and directly subjected to 1 H NMR analysis.

Runs in KP i Buffer / MeOH
A stock solution of the respective biotinylated organocatalyst 1 or 2 (2.59 mM in KP i buffer, pH 7.0, 10 mM) was prepared. 0.58 mg commercial T-Sav (38 nmol active sites, 1.2 mol%, contains 75 µg KP i mixture, pH 6.5) were weighed into a 1.5 mL Microcentrifuge tube. 200 µL of buffer (KP i buffer, pH 7.0, 10 mM) as well as 12 µL of the organocatalyst stock solution (31 nmol, 1.0 mol%) were added to T-Sav. The suspension was mixed by inversion, spun down in a microcentrifuge tube centrifuge (13,000 g), and the pH was adjusted carefully to 7.0 by using 0.1 M NaOH and 0.1 M HCl stock solutions. The mixture was filled up to 250 µL with buffer (KP i buffer, pH 7.0, 10 mM). Stock solutions of cinnamaldehyde (1.33 M in MeOH, HPLC grade) and nitromethane (6.61 M in MeOH, HPLC grade) were prepared. Subsequently, 2.5 µL of the nitromethane stock solution (16.5 µmol, 5.0 eq) and 2.5 µL of the aldehyde stock solution (3.3 µmol, 1.0 eq) were added to the catalyst solution. 250 µL of MeOH were added. The microcentrifuge tube was placed in a Falcon tube, which was placed inside an incubator shaker (50 rpm, 25 °C) for 18 h. The phases were separated, the aqueous phase was extracted (3 × 500 µL EtOAc, 3 × 500 µL CHCl 3 ), and the organic phases were combined. The volatiles were removed under reduced pressure, the residue was taken up in 700 µL CDCl 3 , and transferred to an NMR tube. This was directly subjected to 1 H NMR analysis.

Catalyst Background Runs
To determine the background reactivity of 1 and 2 reactions were performed exactly as stated above without the use of T-Sav. The phases were separated, the aqueous phase was extracted (3 × 500 µL EtOAc, 3 × 500 µL CHCl 3 ), and the organic phases were combined. The volatiles were removed under reduced pressure, the residue was taken up in 700 µL CDCl 3 , and transferred to an NMR tube. This was directly subjected to 1 H NMR analysis.

General Procedure for all Runs
A stock solution of the respective biotinylated organocatalyst 1 or 2 (2.59 mM in KP i buffer, pH 7.0, 10 mM) was prepared. At least 1.33 mg of lyophilized D-Sav from 3.1 (38 nmol active sites, 1.2 mol%) were weighed into a 1.5 mL Microcentrifuge tube. 238 µL of buffer (KP i buffer, pH 7.0, 10 mM) as well as 12 µL of the organocatalyst stock solution (31 nmol, 1.0 mol%) were added to D-Sav. The suspension was mixed by inversion and spun down in a microcentrifuge tube centrifuge (13,000 g). Stock solutions of cinnamaldehyde (1.33 M in MeOH, HPLC grade) and nitromethane (6.61 M in MeOH, HPLC grade) were prepared. Subsequently, 2.5 µL of the nitromethane stock solution (16.5 µmol, 5.0 eq) and 2.5 µL of the aldehyde stock solution (3.3 µmol, 1.0 eq) were added to the catalyst solution. 250 µL of MeOH were added. The microcentrifuge tube was placed in a Falcon tube, which was placed inside an incubator shaker (50 rpm, 25 °C) for 18 h. The phases were separated, the aqueous phase was extracted (3 × 500 µL EtOAc, 3 × 500 µL CHCl 3 ), and the organic phases were combined. The volatiles were removed under reduced pressure, the residue was taken up in 700 µL CDCl 3 , and transferred to an NMR tube. This was directly subjected to 1 H NMR analysis. 1, Tables S8-S10). The mixture was homogenized by shaking at 500 rpm for 5 min at 25 °C and the mixture was shaken at 300 rpm at 25 °C for 18 h. CH 2 Cl 2 (500 µL) was added, the biphasic mixture was shaken vigorously for 1 min, and the organic phase was isolated. This operation was repeated three times, the organic phases were pooled, and evaporated under reduced vacuum. The crude of reaction was dissolved in CDCl 3 and subjected to 1 H NMR analysis. Figure S1. 1 H NMR Based Screening for Yield -Exemplary Analysis. Yields were calculated as followed using the above indicated integrals of the starting material cinnamaldehyde, the 1,4-addition product, the 1,2-addition product, 4 and an unidentified impurity.

Enantioselectivity Determination for all Sav Variants
After determining the conversion by 1 H NMR, the samples of each triplicate were combined, and the solvent was removed. The crude material was purified by preparative TLC (nhexane:EtOAc 75:25) using a complete sheet. The part containing S1 (checked via racemic reference sample by UV fluorescence deletion and permanganate stain) was cut out, the silica scratched from the aluminium plate, and stirred in CH 2 Cl 2 for several minutes. The silica was filtered off, washed with CH 2 Cl 2 , the filtrate was concentrated under reduced pressure, and the purified product was stored at -23 °C under inert atmosphere until further usage The purified compound S1 was converted to alcohol S2 according to 3 via reduction with NaBH 4 , using MeOH as solvent instead of ethanol. The crude material obtained according to Ref 3 was purified by preparative TLC (n-hexane:EtOAc 66:33) using a complete sheet. The part containing S2 (checked via racemic reference sample by UV fluorescence deletion and permanganate stain) was cut out, the silica scratched from the aluminium plate, and stirred in CH 2 Cl 2 for several minutes. The silica was filtered off, washed with CH 2 Cl 2 , the filtrate was concentrated under reduced pressure, and the purified product was used for analytical chiral HPLC to determine the enantioselectivity of the respective catalyst. Crystals of T-Sav:2 were grown by adapting protocols for the preparation of T-Sav:1 crystals. A solution of commercial T-Sav (Prospec 791, 10 mg/mL) was incubated with 2 (5 equivalents per active biotin binding site) for 10 minutes. Residual KP i from the commercial lyophilised protein was removed by buffer exchange via spin column (10k MWCO) to produce a solution of T-Sav:2 in deionised water (40 mg/mL). This was used as protein stock solution to prepare samples for crystallisation.
The crystallisation experiments were performed in a 24-well plate employing 4 different conditions in each quadrant of 6 wells, with slightly varied conditions within the quadrants (see Table S4). All crystals were grown using the hanging-drop vapour-diffusion method. Each well was filled with 1 mL of the precipitant solution (Table S4). On the respective cover slip, three drops were produced by mixing the protein solution (1, 2, and 3 µL) with the same volume of precipitant solution. The assay was stored in an incubator at 20 °C and the crystals appeared within 10 days.

X-ray data collection, processing and structure solution
Crystals grown from various drops were fished and flash-frozen with liquid nitrogen. Diffraction data were collected at 100 K at beamlines I04 of the Diamond Light Source, Didcot, UK, with all images integrated with XDS and scaled with Aimless in xia2. The crystals grown in 52% MPD gave the best resolution. The structures were solved by molecular replacement using MOLREP in CCP4i2, using PDB 6GH7 as a search model. Model was refined numerous cycles with maximum likelihood refinement using REFMAC and manually corrected using COOT.

7
Experimental Details for the Structure Determination of M-Sav:2

Procedure for Crystallization of M-Sav:2
Stock solutions of M-Sav:2 (20 mM Tris, 50 mM NaCl, pH 8.0) were obtained by the protocol of Chapter 3.1 but applying a second size exclusion chromatography run to obtain highly pure material. Crystals of M-Sav:2 were grown in 96 well crystallization plates by employing a crystallization robot and a commercial screen (Molecular Dimensions, Structure Screen 1 + 2 HT-96 Single Reagent) using the sitting drop method. The assay was stored in an incubator at 20 °C.

X-ray data collection, processing and structure solution
Crystals grown from various drops were fished and flash-frozen with liquid nitrogen. Diffraction data were collected at 100 K at beamlines I03 of the Diamond Light Source, Didcot, UK, with all images integrated with XDS and scaled with Aimless in xia2. The crystals grown in 0.2 M (NH 4 ) 2 SO 4 , 0.1 M Na(CH 3 ) 2 AsO 2 , 30% w/v PEG 8000, pH 6.5 gave the best resolution. The structures were solved by molecular replacement using MOLREP in CCP4i2, using PDB 4JNJ as a search model. Model was refined numerous cycles with maximum likelihood refinement using REFMAC and manually corrected using COOT.

8
Computational Methods

Computational Details of the Molecular Models Set Up
The origin of atom coordinates of streptavidin with bound biotin was adapted from X-Ray structure as available in Protein Data Bank (PDB 1STP). 5 Biological assembly, missing atoms, protonation state of titratable amino acids, optimization and molecular dynamics (MD) simulations performed for the T-Sav with biotin-catalyst were described in our previous paper. 3 The last structure of these MD simulations was used in order to build up the model to study the molecular mechanism of the iminium catalysis based on Quantum Mechanics/Molecular Mechanics (QM/MM) methods. After optimization of the system, those residues located 20 Å far away from any of the substrate atoms were kept frozen in the remaining calculations. Potential energy surfaces, free energy surfaces and spline corrections have been performed using fDynamo library 6 together with the OPLS force field. 7 A cut-off for non-bonding interactions was applied using a smooth switching function between 14.5 Å to 16 Å. Parameters for parts of the catalyst included in the MM part were generated using the SwissParam web server. 8 Additionally, two new models have been prepared to study the mutations of the residues Lys121 (Lys121Ala) and Ser112 (Ser112Ala) by Ala. MD simulations were performed at MM level using NAMD software. 9 The behaviour of the system was controlled by using AMBER ff-03 parameters. 10 NPT MD of 5 ns for Lys121Ala and Ser112Ala models with time step of 1 fs at 300 K were carried out after previous optimization, heating (from 0 to 300 K with 0.001 K temperature increment) and equilibration processes of 100ps. The constant temperature and pressure were controlled using the Langevin piston method. 11 Periodic boundary conditions (PBC) using the particle mesh Ewald method were applied. A Cut-off for nonbonding interactions was applied using a smooth switching function with between 14.5 to 16 Å.

Computational details of the QM/MM simulations
In this work, an additive hybrid QM/MM scheme was employed for the construction of the total Hamiltonian, , where the total energy is obtained from the sum of each contribution to the ê nergy. (S1)

Potential Energy Surfaces
Exploration of the Potential Energy Surfaces (PES) was carried out by choosing the appropriate combination of internal coordinates ( i ) in every single step of the reaction. A harmonic constraint was used to maintain the proper interatomic distances along the reaction coordinate, and a series of conjugate gradient optimizations and L-BFGS-B optimization algorithms were applied to obtain the final potential energy of the minimized constrained geometry. The QM sub-set of atoms were described first by the Austin Model 1 (AM1) 12 semiempirical Hamiltonian.
After the exploration of stationary points on the PES, the structures corresponding to reactants and products states (RS and PS, respectively), intermediates (Is) and transition state (TSs) were localized applying Baker's algorithm. 13 Minimum energy path was traced down to reactants and products following the Intrinsic Reaction Coordinate (IRC) method from every localized TS structure. The QM sub-set of atoms in these energy minimization were treated by the AM1 Hamiltonian and later by the M06-2X functional 14 with the standard 6-31+G(d,p) basis set using the Gaussian09 program. 15

Free Energy Surfaces
FESs were obtained, in terms of Potentials of Mean Force (1D-and 2D-PMF), for every step of the reaction using the Umbrella Sampling approach 16 combined with the Weighted Histogram Analysis Method (WHAM). 17 Series of MD simulations were performed adding a constraint along the selected reaction coordinates with an umbrella force constant of 2500 kJ·mol -1 ·Å -2 . In every window QM/MM MD simulations were performed with a total of 5 ps of equilibration and 20 ps of production at 303 K using the Langevin-Verlet algorithm with a time step of 1 fs. Structures obtained in previously computed PESs were used as starting points for the MD simulations in every window. The resulting Free Energy Surfaces are shown in Figures S5A-S5C while average of key inter-atomic distances of every state involved in the reaction are listed in Table S7.

Spline Corrections
A correction term is interpolated to any value along the reaction coordinates in the FES. A continuous energy function is used to obtain the corrected PMFs: where S is the two-dimensional spline function and is the difference between the energies Δ obtained at low-level (LL) and high-level (HL) of theory of the QM part. The AM1 semiempirical Hamiltonian was used as LL method, while a density functional theory (DFT)-based method was selected for the HL energy calculation. In particular, HL energy calculations were performed by means of the hybrid M06-2X 11 functional using the standard 6-31+G(d,p) basis set. These calculations were carried out using the Gaussian09 program.         Figure S13. SDS-PAGE gels of M-Sav at different stages of purification. Figure S14. SDS-PAGE gel of D-Sav during Ni-affinity purification.

ESI-MS of the Expressed M-Sav
Expected MS: 15730 Da Figure S16. Deconvoluted (left) and raw (right) mass spectrum of M-Sav.

Chiral HPLC Data of Activity and Selectivity Screening
Signals in the range from 5 to 10 minutes are impurities from the solvents used for preparative TLC purification. Signals appear higher than usual for S112V and S112Y due to the low amount of product S2 obtained from catalytic runs. Small shifts in retention time were observed due to the use of two different machines, column regeneration procedures, and replacement of the guard column. The exact times were confirmed by running the standard S2 sample before and after each measuring sequence.