Enzyme‐Catalyzed Synthesis of Esters in Water

MsAcT catalyzes the esterification of primary alcohols in water. When utilizing acid and alcohol as starting materials low yields dictated by thermodynamics were observed. However, with activated esters such as ethyl acetate and vinyl acetate very high yields of the desired ester can be achieved in combination with the appropriate alcohol. This study investigated both the intrinsic kinetic properties of MsAcT for the hydrolysis and transesterification of esters in water as well as the thermodynamics of the reaction. In comparison to the chemical or enzymatic ester synthesis using either toxic reagent, and harsh organic solvents, the MsAcT‐catalyzed synthesis of esters of primary alcohols can be achieved efficiently in water without neutralization steps.


Introduction
The synthesis of esters is textbook-knowledge and is performed according to standard protocols that are taught unaltered already for many decades. [1] An activated acid, i. e. an acid chloride or anhydride, reacts with an alcohol in the presence of a base, typically pyridine, often catalyzed by DMAP, in an organic solvent. Alternatively, the acid is activated in situ with reagents such as DCC. In all cases considerable amounts of waste are generated: the solvent and stoichiometric amounts of salt from neutralization steps. Furthermore, the activation, in situ or not, also generates considerable amounts of waste. The direct esterification catalyzed by an acid requires azeotropic removal of water to shift the equilibrium towards ester formation and the reaction conditions are so drastic that it is limited to stable starting materials and products. Moreover, it still leaves the problem of the organic solvent and the neutralization steps generating salt waste. [1a-c] Similarly the enzyme catalyzed synthesis of esters is performed in dry organic solvents with activated acids (acyl donors), often generating much waste which needs to be removed in additional work up steps (Scheme 1). [2] Since the activated acids in the enzyme catalyzed ester synthesis are esters themselves, these reactions are transesterifications.
To address the problem the acyltransferase from Mycobacterium smegmatis (MsAcT) was employed as catalyst. [3] This enzyme was described to enable the synthesis the esters in water. With MsAcT it should thus be possible to avoid the use of organic solvents and to eradicate the need for neutralization steps commonly employed to remove the base utilized in ester syntheses. In all cases described to date, the ester synthesis catalyzed by MsAcT is strictly speaking a transesterification since the reactions all utilize ethyl acetate or more reactive acid derivatives. [3] While this is similar to the classical ester synthesis which utilizes acid chlorides or anhydrides it raises the question whether MsAcT catalyzes the direct synthesis of esters, too. This has been described for lipases and other catalysts; [4] but in all those examples the yield was determined by the thermodynamic stability of the product under reaction conditions. Therefore, the direct esterification and transesterification of alcohols catalyzed by MsAcT in water was studied in parallel. As a catalyst MsAcT should accelerate the reaction but not alter its overall equilibrium. [5] High transesterification yields with MsAcT can only be obtained under kinetic control using activated acids.
MsAcT is known to be an excellent catalyst for the transesterification of primary alcohols and amines with a variety of acyl donors. [3d,6] Benzyl alcohol and isobutyl alcohol were chosen for this study in combination with acetic acid in the form of potassium acetate buffer or ethyl-, vinyl-and phenylacetate as activated acids. The kinetic properties of MsAcT were investigated for both the hydrolysis and transesterification rate of vinyl acetate. This acyl donor irreversibly tautomerizes upon

Results and Discussion
To establish the thermodynamic equilibrium of the reaction and to probe whether MsAcT had the potential to overcome the macroscopic thermodynamic forces of the reaction [3b] the direct esterification of benzyl alcohol with acetic acid at pH 7.5 in potassium phosphate buffer (KPi) was studied. Investigation of both the synthesis (10 mM benzyl alcohol) and hydrolysis (10 mM benzyl acetate) reaction catalyzed by MsAcT revealed a final concentration of benzyl acetate of 0.12 mM. A K eq of 0.00012 could thus be established. The uncatalyzed control reactions yielded essentially no product. MsAcT is thus a classical catalyst, catalyzing a reaction but not altering it.
To prepare the desired ester it is therefore necessary to employ activated acids. As was demonstrated earlier this can be simple ethyl acetate, a benign solvent that can be used monophasic, i. e. dissolved in water, or as a separate layer. [3] Here, all studies were performed under monophasic conditions with the acyl donor dissolved in the pH 7.5 KPi buffer. In order to characterize the affinity and activity of activated acyl donor ethyl, vinyl and phenyl acetate to MsAcT, the optimal enzyme concentration was investigated. At less than 50 ng mL À 1 MsAcT the conversion of benzyl alcohol to benzyl acetate was not complete. In line with earlier observations, [3d] a significant amount of benzyl acetate was hydrolyzed again at high enzyme concentrations ( Figure 1). The trend in conversion of vinyl acetate > phenyl > ethyl acetate after one hour of reaction time was in line with the reactivity of the acyl donor. [2a,b] Vinyl acetate clearly gave the highest conversion for ester synthesis. Moreover, it is synthesized via an atom efficient catalytic process from acetic acid and ethylene, making its production environ-mentally benign. [7] Additionally it is very readily available because it is a monomer for polymer synthesis. Also, the side product acetaldehyde does not require special work up steps in laboratory scale. On industrial scale measures to handle it are well established. [8] Therefore, this activated acyl donor is relatively environmentally benign. It does not require neutralization steps in down-stream processing nor are toxic compounds produced at a large scale. The characterization of this acyl donor was pursued in more detail in water to also avoid organic solvents necessary when used in combination with lipases.
To firmly establish all possible products and side products, such as hemiacetal esters, in-situ 1 H-NMR analysis was performed. In a previous study, we have shown that MsAcT catalyzes the transesterification of isobutanol in water. [3d] Based on those results and since transesterification of benzyl alcohol and vinyl acetate could not be followed via in-situ 1 H-NMR due to overlapping NMR signals, the acylation of 11 mM isobutanol with 212 mM vinyl acetate with less than a μg mL À 1 of MsAcT was studied. Full conversion of isobutanol toward isobutyl acetate was observed [3d] and no ester of the gem-diol was observed ( Figure S6-S9). Next, the substrate concentration was increased to 97 mM isobutanol and 20-fold more MsAcT was added. Under these conditions the starting materials are all soluble, but the product will be insoluble in water. During the course of the reaction a second organic phase appeared due to the insolubility of isobutyl acetate ( Figure 2, Figure S10, and the video will be uploaded together with the proofs). No hemiacetal ester was observed. After prolonged reaction time the reaction mixture became clear again demonstrating the hydrolysis of isobutyl acetate. This occurred after all acyl donor had been consumed, indeed no isobutyl acetate was detectable after 18 hours of reaction time at this stage of the reaction. MsAcT catalyzed only the hydrolysis of the product ester, leading to the final reaction equilibrium; that of acid, alcohol, ester and water.
The transesterification of alcohols with activated acids catalyzed by MsAcT follows overall a synthesis/hydrolysis pattern as is well established for the amide bond synthesis. [9] Initially the alcohol is the preferred substrate of MsAcT and the desired ester, the kinetic product, is formed. In parallel both the activated acid and the ester are also subject to hydrolysis. Once all acyl donor is consumed the hydrolysis reaction leads to the final reaction equilibrium and thermodynamics dictate the ester yield (Scheme 2 and Figure 3a).
After the synthesis of esters in water with different acyl donors or enzyme concentrations the kinetic parameters of MsAcT catalyzed transesterification in water was investigated. To firmly establish the transesterification of benzyl alcohol with vinyl acetate the ester formation was monitored using GC analysis for the determination of the Michaelis-Menten constants of two-substrate reactions. [10] The MsAcT-catalyzed transesterification of benzyl alcohol with vinyl acetate gave for benzyl alcohol a K m of 9.1 mM � 2.8 mM and k cat of 3.16 × 10 3 s À 1 , demonstrating high transesterification efficiencies. More interestingly, the transesterification was revealing a K M and k cat of 30.1 mM � 6.3 mM and 2.49 × 10 3 s À 1 respectively.
More interestingly, using a coupled spectrophotometric assay the hydrolysis of benzyl acetate was measured demonstrating high affinity and uncompetitive inhibition for MsAcT. When benzyl acetate was added to the hydrolysis of vinyl acetate the apparent affinity towards vinyl acetate was significantly lower representing competitive inhibition (Figure 4). Recently, a computational study suggested that benzyl acetate would have high affinity towards the aromatic active site. [6b] In this study, we demonstrated that the kinetics of the enzymatic reaction are severely inhibited by the addition of benzyl acetate indicating that ligand exchange severely lowers the reaction rates.

Conclusions
The synthesis of esters in high yields can be achieved in water with activated acids using MsAcT. The transesterification in water proceeds under kinetic control and careful monitoring of the reaction is necessary to avoid reverse hydrolysis. The kinetic parameters for MsAcT demonstrated a high synthesis to hydrolysis ratio. If the reaction is performed at high concentrations of benzyl alcohol, the product will form a separate layer leading to in-situ product removal essential for reducing enzyme inhibition under synthesis conditions. From a reaction engineering perspective, since the side product acetaldehyde is volatile, the laboratory scale reaction is straightforward to perform. On a larger scale standard measures to prevent acetaldehyde from escaping into the environment need to be taken. [8] These are well-established for industrial scale. Overall, organic solvents and wasteful neutralization steps can be avoided.

SDS-PAGE Analysis
The purity of MsAcT (24 kDa) was analyzed with SDS-PAGE analysis. The protein samples were denatured using a XT Sample Buffer (Bio-Rad) and XT Reducing Agent (Bio-Rad) at 95°C for 5 to 10 minutes. A Criterion XT Bis-Tris MOPS 4-12 % precast gel (Bio-Rad) equipped with a Precision Plus Protein Unstained Standard (Bio-Rad) and denatured protein samples was run at 180 V in MES buffer (Bio-Rad) for 40 minutes. The gel was stained with SimplyBlue SafeStain (ThermoFisher).

BCA Assay
Protein content was determined with the bicinchoninic acid (BCA) protein quantitation kit (Thermo Scientific, Carlsbad, USA). Standard curves were prepared with bovine serum albumin (BSA) in the range of 0.003 to 1.38 mg mL À 1 in a (poly)styrene 96-well plate. Samples were measured in triplicate and monitored at 562 nm utilizing a microtiter plate spectrophotometer (Synergy 2, BioTek).

MsAcT Activity Assay with Neopentyl Glycol (NPG)
The activity assay was started with the addition of 25 μL from a solution of MsAcT (0.02 mg mL À 1 ) dissolved in KPi buffer (200 mM, pH 8) to 975 μL ethyl acetate containing the external standard dodecane (10 mM) and NPG (100 mM). The biphasic reaction mixtures were shaken for 30 seconds at 2500 rpm at room temperature. Reactions were quenched after 0, 2, 4, 6 and 8 minutes by addition of excess sodium sulfate. 50 μL of each reaction mixture was pipetted into a GC vial and 950 μL ethyl acetate was added prior to GC analysis. The amount of NPG monoacetate was quantified using external calibration curve, as is shown in Figure S5. Activity is reported as micromoles of NPG monoacetate produced per minute per milligram of purified MsAcT.

MsAcT Activity assay with p-Nitrophenylbutyrate (pNPB)
The activity assay was started by adding 20 μL of MsAcT to 180 μL buffer containing KPi buffer (50 mM, pH 7.5) in transparent 96-well polystyrene plates. The rate of hydrolysis was monitored at 37°C at 405 nm using a multimode spectrophotometer plate reader (Synergy 2, BioTek). Buffer and 96-well plate were pre-heated to 37°C before the start of the reaction. An external calibration curve of p-nitrophenol was constructed between 0.00-1.00 mM.

Esterification of Benzyl Alcohol with Potassium Acetate under Monophasic Reaction Conditions
For the synthesis of benzyl acetate from benzyl alcohol and potassium acetate the following reaction conditions were applied. MsAcT (12000 ng) was added to KPi buffer (200 mM, pH 7.5) containing 10 mM benzyl alcohol and 100 mM potassium acetate to yield a monophasic reaction mixture of 1 mL. For the hydrolysis of benzyl acetate to benzyl alcohol and potassium acetate the following reaction conditions were applied. MsAcT (12000 ng) was added to KPi buffer (200 mM, pH 7.5) containing 10 mM benzyl acetate and 100 mM potassium acetate to yield a monophasic reaction mixture of 1 mL. The reaction mixture was shaken at 1000 rpm at 21°C. Samples were taken after 1, 2, 3, 4, 5 and 24 hours. Samples were quenched by addition of 500 μL diethyl ether containing dodecane (10 mM) as external standard. The mixture was rapidly vortexed for 30 seconds and centrifuged at 13 000 rpm to remove enzyme precipitates. The organic layer was collected, dried with an excess of dry MgSO 4 , and analyzed by GC. The amount of benzyl alcohol and benzyl acetate was quantified using an external calibration curve, as is shown in Figure S4.

Reversible Hydrolysis and Synthesis of Isobutyl Acetate by MsAcT
The reaction was initiated by the addition of MsAcT to a reaction solution resulting in the final concentration of 11 mM isobutanol, 212 mM vinyl acetate, KPi (200 mM,pH = 7.5), and 20 μg mL À 1 MsAcT. The solution was resuspended vigorously during addition for 1-3 seconds and measured immediately with 1 H-NMR (Agilent, 400 MHz). Also, the reaction was repeated and recorded with a dual pixel 12MP OIS (F1.7) camera to show the reversible synthesis and hydrolysis. For the synthesis of isobutyl acetate from isobutanol and vinyl acetate the following reaction conditions were applied. The reaction was initiated by the addition of MsAcT to a reaction solution resulting in the final concentration of 11 mM isobutanol,212 mM vinyl acetate,200 mM KPi (200 mM,pH = 7.5), and 0.5 μg mL À 1 MsAcT. The solution was resuspended vigorously during addition for 1-3 seconds and measured immediately with 1 H-NMR (Agilent, 400 MHz).

Optimizing MsAcT Concentration for the Acylation of Benzyl Alcohol with Vinyl-, Phenyl-, and Ethyl Acetate
MsAcT (50-12000 ng) was added to KPi buffer (200 mM, pH 7.5) containing 10 mM benzyl alcohol, and 100 mM acyl donor being either vinyl-, phenyl-, and ethyl acetate to yield a monophasic reaction mixture of 1 mL. The mixture was shaken at 1000 rpm at 21°C for 60 minutes. Product and substrate were extracted twice by addition of 500 μL diethyl ether containing dodecane (10 mM) as external standard. The mixture was rapidly vortexed for 30 seconds and centrifuged at 13 000 rpm to remove enzyme precipitates. The organic layer was collected, dried with an excess of dry MgSO 4 , and analyzed by GC. The amount of benzyl alcohol and benzyl acetate was quantified using external calibration curve ( Figure S4). Initial rates were calculated from the linear slope of benzyl acetate concentration over time.

Determination Kinetic Parameters for Vinyl Acetate and Benzyl Alcohol using GC Analysis
For the kinetic parameters for benzyl alcohol the following reaction conditions were applied. MsAcT (50 ng) was added to KPi buffer (200 mM, pH 7.5) containing either 0, 1, 2. 5,5,10,15, and 20 mM benzyl alcohol, and 100 mM vinyl acetate to yield a monophasic reaction mixture of 1 mL. For the kinetic parameters for vinyl acetate the following reaction conditions were applied. MsAcT (50 ng) was added to KPi buffer (200 mM, pH 7.5), 10 mM benzyl alcohol, and 0, 10, 25, 50, 100, and 200 mM vinyl acetate to yield a monophasic reaction mixture of 1 mL. The mixture was shaken at 1000 rpm at 21°C for 0, 2, 5, 10, 20, and 30 minutes. Product and substrate were extracted twice by addition of 500 μL diethyl ether containing dodecane (10 mM) as external standard. The mixture was rapidly vortexed for 30 seconds and centrifuged at 13 000 rpm to remove enzyme precipitates. The organic layer was collected, dried with an excess of dry MgSO 4 , and analyzed by GC. The amount of benzyl alcohol and benzyl acetate was quantified using external calibration curves ( Figure S4) and the data is shown in Figure S14. Initial rates were calculated from the linear slope of benzyl acetate concentration over time. The curves were fitted to the Michaelis-Menten equation using the fit function of Gnuplot 5.2, [13] as is shown in Table S1.

Coupled Spectrophotometric Activity Assay with ScADH of Vinyl Acetate Hydrolysis with MsAcT
The assay was performed in polyacrylate 1 cm cuvettes by monitoring the conversion of 0.25 mM NADH at 340 nm and 20°C with an extinction coefficient of 6.221 mM À 1 cm À 1 . Before the addition of MsAcT, the reaction was monitored until stable. Enzymatic reactions were started by the addition of MsAcT with a final concentration of vinyl acetate (0-10 mM) in KPi (200 mM, pH 7.5), ScADH (50 U mL À 1 ), MsAcT (24 ng μL À 1 ). All measurements were performed in triplicates. The addition of additional ScADH did not result in a higher response for indirect acetaldehyde detection. The Michaelis Menten curve is shown in Figure S11. The curves were fitted to the Michaelis-Menten equation using the fit function of Gnuplot 5.2, [13] as is shown in Table S1.

Coupled Spectrophotometric Activity Assay with ScADH of Benzyl Acetate Synthesis with MsAcT
The assay was performed in polyacrylate 1 cm cuvettes by monitoring the conversion of 0.25 mM NADH at 340 nm and 20°C by using an extinction coefficient of 6.221 mM À 1 cm À 1 . Before the addition of MsAcT, the reaction was monitored until stable. Enzymatic reactions were started by the addition of MsAcT with a final concentration of benzyl alcohol (1-100 mM), vinyl acetate (10 mM) in KPi (200 mM, pH 7.5), ScADH (50 U mL À 1 ), MsAcT (0.24 ng μL À 1 ). All measurements were performed in triplicates. The addition of additional ScADH did not result in a higher response for indirect acetaldehyde detection. The Michaelis Menten curve is shown in Figure S12. The curves were fitted to the Michaelis-Menten equation using the fit function of Gnuplot 5.2, [13] as is shown in Table S1.

Coupled Spectrophotometric Activity Assay with HLADH-E of Benzyl Acetate Hydrolysis with MsAcT
The assay was performed in polyacrylate 1 cm cuvettes by monitoring the conversion of 0.25 mM NAD + at 340 nm and 20°C by using an extinction coefficient of 6.221 mM À 1 cm À 1 . Before the addition of MsAcT, the reaction was monitored until stable. Enzymatic reactions were started by the addition of enzyme with a final concentration of benzyl acetate (1-10 mM), in KPi (200 mM, pH 7.5), HLADHÀ E (5 U mL À 1 ), MsAcT (0.24 ng μL À 1 ). All measurements were performed in triplicates. The addition of additional HLADHÀ E did not result in a higher response for indirect benzyl alcohol detection. The Michaelis Menten curve with non-competitive substrate inhibition is shown in Figure S13. The curves were fitted to the non-competitive Michaelis-Menten equation using the fit function of Gnuplot 5.2. [13] Coupled Spectrophotometric Activity Assay with ScADH of Competitive Vinyl-and Benzyl Acetate Hydrolysis with MsAcT The assay was performed in polyacrylate 1 cm cuvettes by monitoring the conversion of 0.25 mM NADH at 340 nm and 20°C by using an extinction coefficient of 6.221 mM À 1 cm À 1 . Before the addition of MsAcT, the reaction was monitored until no chemical background hydrolysis occurred. Enzymatic reactions were started by the addition of MsAcT with a final concentration of benzyl acetate (0.00-5.6 mM), vinyl acetate (5.4 mM), NADH (0.25 mM), KPi (200 mM, pH 7.5), MsAcT (24 ng mL À 1 ), ScADH (50 U mL À 1 ). The error bars show the standard deviation of triplicates. All measurements were performed in triplicates. The addition of additional ScADH did not result in a higher response for indirect acetaldehyde detection.

Coupled Spectrophotometric Activity Assay with ScADH of Benzyl Acetate Synthesis with MsAcT in the Presence of Varying Amounts of Product
The assay was performed in polyacrylate 1 cm cuvettes by monitoring the conversion of 0.25 mM NADH at 340 nm and 20°C by using an extinction coefficient of 6.221 mM À 1 cm À 1 . Before the addition of MsAcT, the reaction was monitored until no chemical background hydrolysis occurred. Enzymatic reactions were started by the addition of MsAcT with a final concentration of benzyl acetate (0.0-12.0 mM), vinyl acetate (5.4 mM), benzyl alcohol (80 mM), NADH (0.25 mM), KPi (200 mM, pH 7.5), MsAcT (24 ng mL À 1 ), ScADH (50 U mL À 1 ). All measurements were performed in triplicates. The addition of additional ScADH did not result in a higher response for indirect acetaldehyde detection.