An alginate-confined peroxygenase-CLEA for styrene epoxidation

Oxyfunctionalisation reactions in neat substrate still pose a challenge for biocatalysis. Here, we report an alginate-confined peroxygenase-CLEA to catalyse the enantioselective epoxidation of cis-β-methylstyrene in a solvent-free reaction system achieving turnover numbers of 96 000 for the biocatalyst and epoxide concentrations of 48 mM.


Enzyme expression and purification
The laboratory-evolved expression variant PaDa-1 of the peroxygenase form Agrocybe aegerita (rAaeUPO) was heterologously expressed in Pichia pastoris and isolated following a previously described procedure. [1] S1.2.2 SDS-PAGE analysis rAaeUPO (5 μg) in Tris/HCl buffer (20 mM, pH 7.0) was mixed with 6x SDS loading buffer. Samples were boiled for 10 min at 99 °C and centrifuged for 5 min at 13000 g. A SDS-PAGE was loaded with 5 μL rAaeUPO and 5 μL protein ladder. The page was run for 20 min at 80 V and for 1 h at 150 V. Proteins were detected via Coomassie staining.

S1.2.3 Bradford assay
General protein concentrations were determined by Bradford assay in 1 mL cuvettes with 20 µl protein solution in Tris/HCl buffer (20 mM, pH 7.0) and 980 µl Bradford dye reagent. Absorbance was detected at 595 nm in a UV-Vis spectrophotometer. Concentrations were quantified based on a standard curve of horseradish peroxidase.

S1.2.4 Determination of rAaeUPO concentration
Absorbance at 420 nm As previously described, [2] the millimolar extinction coefficient 115 mM -1 cm -1 at 420 nm was used to determine the concentration of purified rAaeUPO. Herein, the UV−Vis spectrum between 350 nm and 500 nm of different dilutions of purified rAaeUPO in 1 mL Tris/HCl buffer (20 mM, pH 7.0) was recorded, normalised and evaluated accordingly.

CO difference spectra
To determine the amount of active rAaeUPO, the extinction coefficient of rAaeUPO from the ferrous carbon monoxide binding difference spectra (CO difference spectra) was utilised. Purified rAaeUPO was prepared in 1 mL Tris/HCl buffer (20 mM, pH 7.0) at different dilutions. Na 2 S 2 O 4 was added in a final concentration of 10 mM to reduce the heme−thiolate enzyme. Subsequently, the UV−Vis spectrum between 400 nm and 500 nm was recorded in a UV-Vis spectrophotometer. After saturating the enzyme solution with CO, the spectrum between 400 nm and 500 nm was measured again to record the absorbance shift which occurs due to heme-CO adduct formation in active rAaeUPO. The millimolar extinction coefficient 91 mM -1 cm -1 at 445 nm [3] [4] was utilized to calculate the concentration and mass of active imm-rAaeUPO after normalisation.

Activity of rAaeUPO and rAaeUPO-CLEAs
Activity of rAaeUPO and rAaeUPO-CLEAs was determined by ABTS assay as describes earlier [5]. Measurements were performed in 1 mL of 150 mM sodium phosphate citrate buffer at pH 4.4 with 0.3 mM ABTS and 1 mM H 2 O 2 . Absorbance was determined every 6 sec over 60 sec at 420 nm in an UV-Vis spectrophotometer Genesys 150 from Thermo Scientific. The activities were calculated in U ABTS (1 U ABTS = 1 µmol ABTS min -1 ) with the millimolar extinction coefficient of ABTS ( = 36.8 mM -1 cm -1 ).

Activity of imm-rAaeUPO
Activity of imm-rAaeUPO was assayed in 10 mL of 100 mM MES buffer at pH 5 with 0.3 mM ABTS and 1 mM H 2 O 2 . Absorbance was measured every 30 sec over 90 sec at 420 nm in UV-Vis spectrophotometer Genesys 150 from Thermo Scientific. The activities were calculated in U ABTS (1 U ABTS = 1 µmol ABTS min -1 ) with the millimolar extinction coefficient of ABTS ( = 36.8 mM -1 cm -1 ).

Recovered activities
Recovered activities of agg-rAaeUPO and imm-rAaeUPO were calculated with use of the starting rAaeUPO mass in mg and expressed in percentage: Recovered activity of agg -r Recovered activity of imm -r × 100 S1.3 Optimisation of the enzyme immobilisation process S1.3.1 Optimised protocol for the preparation of imm-rAaeUPO 5 mg rAaeUPO (112.6 nmol) were diluted in a volume of 2.5 mL Tris/HCl buffer (20 mM, pH 7) containing 10 mg low-viscosity chitosan. The solution was precipitated with 9 times the volume iso-propanol (22.5 mL) and cross-linked with 5000 molar protein equivalents of glutaraldehyde (563 µmol).
After 2 h of incubation, the samples of rAaeUPO-CLEAs were centrifuged for 5 min at 4000 rpm, 4 °C and the supernatant was discarded. The pellet was washed with 5 mL Tris/HCl buffer (20 mM, pH 7) and resuspended in 500 µl Tris/HCl buffer (20 mM, pH 7). The resuspended rAaeUPO-CLEAs were clarified with a 40 µm sieve. A solution of 2.5% (w/v) low viscosity alginate in MilliQ water was prepared and purified by filtration through a 5 µm membrane filter. The 500 µl rAaeUPO-CLEAs were mixed with 9.5 mL of alginate solution.
For preparation of the imm-rAaeUPO alginate beads, a semi-automated encapsulator (Encapsulation Unit VAR-J30 from Nisco Engineering Inc.) was used with the following parameters: frequency 3.35 kHz, amplitude 0%, 21 hPa, 4.34 mA. The prepared alginate-enzyme suspension was transferred into a 20 mL syringe and extruded through a 150 µm diameter nozzle with 3.3 mL min -1 speed. The droplets were captured in 100 mL of 100 mM CaCl 2 hardening solution under constant agitation. After 30 min of hardening, the imm-rAaeUPO alginate beads were collected with a 40 µm sieve and washed with 10 mL of 5 mM CaCl 2 washing solution. imm-rAaeUPO was stored in Tris/HCl buffer (20 mM, pH 7) containing 5 mM CaCl 2 at 4°C.

S1.3.2 Optimisation of the precipitation reagent
To determine the optimal precipitation reagent, 1 mg rAaeUPO (22.5 nmol) was diluted in a volume of 500 µL Tris/HCl buffer (20 mM, pH 7). The solution was precipitated with 9 volumes of iso-propanol, acetonitrile, ethanol, or saturated ammonium sulphate solution. Afterwards, 5000 times molar excess (112.6 µmol) of glutaraldehyde was added. In addition, the same experiment was conducted without any precipitation reagent.
After 2 h of incubation, the samples of rAaeUPO-CLEAs were treated as described in the optimised protocol. After clarification, activity of the rAaeUPO-CLEAs was determined by ABTS assay and expressed as recovered activity relative to free rAaeUPO.

S1.3.3 Optimisation of the aggregation reagent
In order to determine the optimal cross-linking protocol, five different aggregation reagents were tested. 1 mg rAaeUPO (22.5 nmol) was diluted in a volume of 500 µL Tris/HCl buffer (20 mM, pH 7). The solution was precipitated with 9 volumes of iso-propanol (4.5 mL) and aggregated with glutaraldehyde, terephthalaldehyde, starch oxide/ polyaldehyde starch, formaldehyde, or iminodiacetic acid. All aggregation reagents were applied in 5000 times molar excess (112.6 µmol) compared to rAaeUPO. In addition, the same experiment was conducted without adding any aggregation reagent. All subsequent steps follow the optimised protocol. Activity of the resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads .
The starch oxide/ polyaldehyde starch was freshly prepared as the following [6], [7] : 5 mL of a 0.7 M solution of NaIO 4 was brought to pH 4.0 by addition of sulfuric acid. Starch was added to reach a 1:1 molar ratio with NaIO4 (568 mg of starch). The solution was constantly stirred at 37°C -40°C during the addition. After stirring for 3 -4 h at 37°C -40°C, acetone was added to the solution.
The precipitate was washed with solvent and water, and subsequently dried to recover a white powder of oxidized starch.

S1.3.4 Optimisation of the co-aggregator
1 mg rAaeUPO (22.5 nmol) was diluted in a volume of 500 µL Tris/HCl buffer (20 mM, pH 7). Bovine serum albumin (BSA) and chitosan were tested separately and in combination as coaggregators. The compounds were used in double the quantity (2 mg) of rAaeUPO and added to the prepared solution. In addition, the same experiment was performed without addition of a coaggregator. The mixture was precipitated with 9 times the volume iso-propanol (4.5 mL) and aggregated with 5000 molar protein equivalents of glutaraldehyde (112,6 µmol/ 22.5 µl of a 50% stock). All subsequent steps follow the optimised protocol. Activity of the resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads . S1.3.5 Optimisation of the glutaraldehyde concentration 1 mg rAaeUPO (22.5 nmol) and 2 mg chitosan were diluted in a volume of 500 µL Tris/HCl buffer (20 mM, pH 7). The mixture was precipitated with 9 times the volume iso-propanol (4.5 mL). To test different concentrations of the aggregation reagent glutaraldehyde, 500 molar rAaeUPO equivalents, 5000 molar rAaeUPO equivalents, and 50000 molar rAaeUPO equivalents were used for aggregation of the protein-chitosan mix. All subsequent steps follow the optimised protocol. Activity of the resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads .

S1.3.6 Optimisation of the catalyst loading
In order to determine the optimal catalyst loading, the immobilisation of 1 mg, 2 mg, 5 mg and 10 mg rAaeUPO was analysed. 1 mg / 2 mg / 5 mg / 10 mg rAaeUPO and 2 mg / 4 mg / 10 mg / 20 mg chitosan were diluted in a volume of 500 µL / 1 mL / 2.5 mL / 5 mL Tris-HCl buffer (20 mM, pH 7). The mixture was precipitated with 9 times the volume iso-propanol and aggregated with 5000 molar protein equivalents of glutaraldehyde. All subsequent steps follow the optimised protocol. Activity of the resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads . Moreover, recovered activity relative to free rAaeUPO was determined.

S1.3.7 Optimisation of the alginate concentration
Resuspended rAaeUPO-CLEAs were prepared as given in the optimised protocol und clarified with a 40 µm sieve. Solutions of 2.5%, 3%, 3.5%, and 4% low viscosity alginate in MilliQ water were prepared and purified by filtration through a 5 µm membrane filter. The 500 µl rAaeUPO-CLEAs were mixed with 9.5 mL of the alginate solutions. All subsequent steps follow the optimised protocol. Activity of the resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads .

S1.3.8 Optimisation of the counterions
An alginate-enzyme suspension was prepared as described in the optimised protocol. It was transferred into a 20 mL syringe and extruded through a 150 µm diameter nozzle with 3.3 mL min -1 speed using a semi-automated encapsulator set-up as outlined in the optimised protocol. In order to test the effect of counterions used for hardening of the alginate beads, 50 mM CaCl 2 , 100 mM CaCl 2 , 200mM CaCl 2 , 100mM BaCl 2 , and 100 mM SrCl 2 were examined as hardening solutions. The alginate-enzyme droplets were captured in 100 mL of these solutions under constant agitation. After 30 min of hardening, the imm-rAaeUPO alginate beads were collected with a 40 µm sieve and washed with 10 mL of 5 mM CaCl 2 , 5mM BaCl 2 , or 5 mM SrCl 2 washing solution. imm-rAaeUPO was stored in 20 mM Tris/HCl buffer (pH 7) containing 5 mM CaCl 2 , 5mM BaCl 2 , or 5 mM SrCl 2 at 4°C. Activity of the resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads .

S1.3.9 Optimisation of the alginate bead diameter
The immobilisation was performed as described in the optimised protocol. Next to using a 150 µm diameter nozzle in the semi-automated encapsulation unit, nozzles of 50 µm, 100 µm, 200 µm, and 400 µm diameter were utilized to examine the effect of beads size on the immobilisation efficiency.
For manual preparation of imm-rAaeUPO, an alginate-enzyme suspension was prepared as described in the optimised protocol and was transferred into a 20 mL syringe connected to a 0.5 mm x 16 mm cannula. With use of a syringe pump, the alginate-enzyme suspension was extruded at a constant speed of 3.3 mL min -1 . The droplets were captured in 100 mL of 100 mM CaCl 2 hardening solution under constant agitation. Hardening, washing, and storage were performed as in the optimised protocol. Activity of all resulting immobilisates was tested by ABTS assay and expressed in U ABTS g -1 wet beads .

S1.4.1 Determination of the immobilisation yield
The immobilisation yield was obtained by using two different approaches. First, the general rAaeUPO concentration within the immobilisates was quantified with use of the millimolar extinction coefficient at 420 nm. Moreover, the concentration of active heme-sites inside imm-rAaeUPO was determined by CO difference spectra.
Herein, 200 mg of dry imm-rAaeUPO beads were dissolved in 5 mL of 150 mM sodium phosphate citrate buffer at pH 4.4 and incubated for 30 min. The solution was neutralised with NaOH and the concentration of rAaeUPO inside the solution was determined with use of the millimolar extinction coefficient at 420 nm. Afterwards, the solution was concentrated to a volume of 1 mL using a centrifugal filter unit Ulra-4 with 10 kDa molecular weight cut-off from Amicon. The rAaeUPO concentration inside the resulting solution was quantified by CO difference spectra. The immobilization yield was determined relative to the starting mass of rAaeUPO of 5 mg:

S1.4.3 Light and fluorescence microscopy
Samples analysed by light microscopy were prepared following the optimized immobilisation protocol. Samples for fluorescence microscopy were prepared respectively with the following deviations from the protocol: 1 mg sfGFP [8] was diluted in a volume of 300 µl Tris/HCl buffer (20 mM, pH 7). The suspension was clarified with a 40 µm sieve. A solution of 2.5% low viscosity alginate in MilliQ water was prepared and purified by filtration through a 5 µm membrane filter. The 300 µl GFP was mixed with 4.7 mL of alginate solution.
Light microscopy pictures were taken with use of the white light channel of Microscope Axio Observer A1 from ZEISS. Fluorescence microscopy pictures were taken as overlay of the while light channel and GFP channel. Size measurement was performed with the inbuilt ZEISS software. S1.5 Gas chromatography (GC) S1.

Chemical synthesis of GC standard 2-methyl-3-phenyloxirane
In order to serve as standard chemical for analytics, racemic 2-methyl-3-phenyloxirane was synthesised as previously described [9]. After synthesis, qualitative analysis of the product purity was performed with Thin-Layer-Chromatography (TLC). Herein, silica paper was used as solid phase. A 98:2 mixture of n-heptane and ethyl acetate including 1% triethylamine for basification of the silica paper was used as mobile phase.
Purification of the product mixture was undertaken with an automated Reveleris X2 flash purification system. A 4 g Reveleris silica column was used as solid phase and a 98:2 mixture of n-heptane and ethyl acetate including 1% triethylamine was used as single mobile phase at 15 mL min -1 flow rate. Product fractions were collected within the first 3 minutes and analysed by TLC and gas chromatography. Fractions with high purity were pooled and remaining solvent was evaporated using a rotary vacuum evaporator.

S1.5.2 GC for cis-β-methylstyrene epoxidation
Product mixtures of cis-β-methylstyrene epoxidation were analysed using a Shimadzu GC-2014 gas chromatograph with a flame ionization detector and a CP Sil 5 CB column (dimensions: 50.0 m x 0.53 mm x 1.00 µm nominal). Carrier gas: Nitrogen; Column flow: 20 mL/min. The utilised temperature profile and retention times of all analysed compounds are given in the following (Table S1).
Concentrations of 2-methyl-3-phenyloxirane, benzaldehyde and phenylacetone were quantified based on calibration lines and with use of an internal standard tetradecane. Calibration lines were set up in duplicates by preparing respective concentrations including 100 mM tetradecane in styrene as solvent. 5 µL of this mixture was diluted in 1 mL of methyl tert-butyl ether and dried with MgSO 4 before GC analysis. Commercially available standard chemicals were used for the calibration lines of benzaldehyde and phenylacetone. For 2-methyl-3-phenyloxirane, a selfsynthesised standard was used. In all cases, standard purities based on peak area ratios were taken into account by only using the peak area at expected retention times for the calibration lines.

S1.5.3 Chiral GC for cis-β-methylstyrene epoxidation
To determine enantiomeric excess (ee), product mixtures of cis-β-methylstyrene epoxidation were analysed as previously described [9] by chiral GC (Shimadzu GC-2010) with a flame ionization detector and a Lipodex E 1b (Macherey-Nagel) column (dimensions: 50.0 m × 0.25 mm × 0.25 μm). Carrier gas: Helium; Column flow: 2.16 mL/min; Split ratio: 100; Linear velocity: 38 cm/s. The utilised temperature profile and retention times of all analysed compounds are given in the following (Table S2). Enantiomers were quantified by peak area integration and the ee was calculated as: Table S2. Overview on the method for chiral GC analysis of the cis-β-methylstyrene epoxidation reactions. . Helium served as carrier gas and the injection volume was 5 μL. Product formation was determined as GC conversion, obtained with peak area integrations of the product mixture. The utilised temperature profile and retention times of all analysed compounds are given in the following (Table S3). Table S3. Overview on the standard method for GC analysis of the styrene epoxidation reactions.

.1 Epoxidation of cis-β-methylstyrene
The reaction was performed in small GC vials at room temperature, shaking at 99 rpm with 60° angle in an overhead rotator ( Figure S1). tert-Butyl hydroperoxide ( t BuOOH) was fed continuously with use of a syringe pump and 1 mL syringes which were connected with tubes to the reaction vials.
Before start of the reaction, cis-β-methylstyrene was supplemented with 100 mM tetradecane as internal standard and saturated with TRIS/HCl buffer (20 mM, pH 7). Moreover, a 5 M stock of t BuOOH in decane was supplemented with 100 mM internal standard tetradecane. If further dilution of t BuOOH was required, decane was used as solvent.
A GC vial was prepared with 0.5 µM rAaeUPO. For the case of imm-rAaeUPO, 180 mg of carefully dried immobilisates were utilised. For the case of free rAaeUPO in a two liquid phase system, 0.5 µM enzyme was prepared in an aqueous phase of 180 µL TRIS/HCl buffer (20 mM, pH 7). For the case of free rAaeUPO in a micro-aqueous system, 0.5 µM enzyme was prepared in an aqueous phase of 6.6 µL TRIS/HCl buffer (20 mM, pH 7). Figure S1. Reaction set-up for the enzymatic epoxidation of cis-β-methylstyrene. On the left, syringe pumps for continuous t BuOOH supply. Syringes are connected to tubes which lead to the reaction mixture inside GC vials on the right. An overhead rotator is used for shaking of the reaction at 99 rpm, 60 ° at room temperature.
To start the reaction, the prepared cis-β-methylstyrene was added to the GC vial to reach a final volume of 750 µL. Subsequently, the vials were connected to the continuous t BuOOH feed. The utilised feed rate is indicated in the caption of the respective figures. 5 µL samples were taken at indicated time points, diluted in 1 mL methyl tert-butyl ether, and dried with MgSO 4 . Product mixtures were analysed by gas chromatography. Turnover frequencies (TOF) and turnover numbers (TN) were calculated as the following:

Continuous t BuOOH feed
The same set-up as for the epoxidation of cis-β-methylstyrene was used. Styrene was prepared with 2% internal standard tetradecane and saturated with TRIS/HCl buffer (20 mM, pH 7). If required, dilution of t BuOOH was performed with decane as solvent. A GC vial was prepared with 0.8 µM carefully dried imm-rAaeUPO beads. 1 mL of styrene was added into the vial. The reaction was started with a pulse of t BuOOH of 1 mM final concentration. Subsequently, the vials were connected to the continuous t BuOOH feed. The feed rate during the biotransformation is indicated in the caption of the respective figures. Samples were taken at indicated time points, diluted in 1 mL methyl tert-butyl ether, dried with MgSO 4 , and analysed by gas chromatography.

t BuOOH pulse feeding
Styrene and t BuOOH were prepared as for the reactions with continuous feed. 0.2 µM rAaeUPO were provided in a GC vial as carefully dried imm-rAaeUPO or in a micro-aqueous system with 2 µL phase of TRIS/HCl buffer (20 mM, pH 7). 500 µl of styrene were added and the reaction was started with the addition of t BuOOH in pulses of 1.25 mM final concentration every 15 min. Samples were taken at indicated time points, diluted in 1 mL methyl tert-butyl ether, dried with MgSO 4 , and analysed by gas chromatography.

S1.7 E-factor analysis
The E-factors were calculated according to literature [10]. All waste that was produced during a reaction was taken into account, including water:

Precipitation
Iso-propanol was selected as optimal precipitation reagent due to a high recovered activity, its easy handling, and its low hazardousness.  Figure S3. Selection of the optimal precipitation reagent, used for precipitation prior to CLEA formation. Recovered activity of rAaeUPO-CLEAs after precipitation with different precipitation reagents was determined relative to free rAaeUPO by ABTS-activity assay in aqueous environment. w/o = no precipitation reagent; IPA = iso-propanol; ACN = acetonitrile; EtOH = ethanol; Amm.sulf. = saturated ammonium sulphate solution. At conditions, indicated with asterisk (*), no enzyme precipitation was observed. Data represents an average of duplicates.

Cross-linking
5000 molar equivalents of glutaraldehyde were chosen for cross-linking and chitosan was selected as co-aggregator since respective immobilisates yielded highest ABTS oxidation activities.

Catalyst loading
An enzyme loading of 0.5 mg Protein g -1 beads was used in the optimised immobilisation protocol. Figure S6. Optimisation of imm-rAaeUPO enzyme loading. Activities of imm-rAaeUPO at different loadings of rAaeUPO were determined by ABTS-activity assay in aqueous environment. Data represents an average of duplicates. Catalyst loading of 1 mg Protein g -1 beads did not yield functional catalyst beads. (a) Absolute activity of imm-rAaeUPO. (b) Recovered activity of imm-rAaeUPO, determined relative to free rAaeUPO.

Alginate concentration and counter ions
In the final immobilisation protocol, 2.5% alginate was utilised and 100 mM CaCl 2 was used for hardening of the immobilisates.

Diameter of the immobilisate
In the final immobilisation protocol, imm-rAaeUPO beads of 440 µm diameter were prepared as they were most homogeneous in size and shape. Moreover, an even distribution of protein inside the alginate beads was observed after encapsulating non-cross linked GFP.   Table S4. Key values of the immobilisation efficiency. Data represented an average of duplicates.
Here, expressed relative to free rAaeUPO.

Initial TOF [c] [s -1 ]
Overall TN [b] Free rAaeUPO in two-liquid phase system 10  [a] Reaction robustness is defined as timepoint when no product formation was observed anymore. [b] Final epoxide concentration and turnover number (TN = mol Product  mol rAaeUPO -1 ) were quantified at reaction stop. [c] Turnover frequency (TOF = mol Product  mol rAaeUPO -1  s -1 ).

S2.4 Epoxidation of styrene
Scheme S1. Reaction equation for the epoxidation of styrene by imm-rAaeUPO with tert-butyl hydroperoxide as oxidant.