Enantioselective Transesteriﬁcation of Allyl Alcohols with ( E )-4-Arylbut-3-en-2-ol Motif by Immobilized Lecitase ™ Ultra

: Lecitase ™ Ultra was immobilized on four di ﬀ erent supports and tested for the ﬁrst time as the biocatalyst in the kinetic resolution of racemic allyl alcohols with the ( E )-4-arylbut-3-en-2-ol system in the process of transesteriﬁcation. The most e ﬀ ective biocatalyst turned out to be the enzyme immobilized on agarose activated with cyanogen bromide (LU-CNBr). The best results (E > 200, ee s and ee p = 95–99%) were obtained for ( E )-4-phenylbut-3-en-2-ol and its analog with a 2,5-dimethylphenyl ring whereas the lowest ee of kinetic resolution products (90%) was achieved for the substrate with a 4-methoxyphenyl substituent. For all substrates, ( R )-enantiomers were esteriﬁed faster than their ( S )-antipodes. The results showed that LU-CNBr is a versatile biocatalyst, showing high activity and enantioselectivity in a wide range of organic solvents in the presence of commonly used acyl donors. High operational stability of LU-CNBr allows it to be reused in three subsequent reaction cycles without negative e ﬀ ects on the e ﬃ ciency and enantioselectivity of transesteriﬁcation. This biocatalyst can become attractive to the commercial lipases in the process of the kinetic resolution of allyl alcohols. of the reaction catalyzed by LU-CNBr has been shown [34]. Similar to the hydrolysis reaction, the best resolution in the process of transesterification was observed for substrates with unsubstituted benzene ring and with a 2,5-dimethylphenyl substituent; very good results were also obtained for compounds with a p methylphenyl substituent. Noteworthy is the significantly shorter time (6–8 h) needed for effective separation of alcohols 1b and 1c by transesterification compared to the hydrolysis of their butyrates (24–48 h) [34]. In contrast to the non-enantioselective hydrolysis of butyrate containing a methoxy substituent on a benzene ring [34], transesterification of alcohol bearing this substituent ( 1d ) was enantioselective and it was possible to obtain both products with relatively high enantiomeric excess.


Introduction
Using enzymes as the biocatalysts of chemical reactions is a subject of interest for many researchers worldwide due to their wide spectrum of activity and high stereo-and regioselectivity [1][2][3]. New environmental requirements concerning "green chemistry" have also made the enzymes attractive industrial catalysts, as they can be used in mild reaction conditions (aqueous phase, normal pressure, room temperature) [4][5][6]. However, different enzyme features, including activity, selectivity, and operational stability, can be improved using different metagenomics tools [7][8][9]. One of them is a production of chimeric enzymes by mixing the genes of two different proteins to join their desired properties [10][11][12].
An example of such an approach is a preparation named Lecitase™ Ultra (LU), produced initially by Novo Nordisk A/S by the fusion of genes of the lipase from Thermomyces lanuginosus and phospholipase A 1 from Fusarium oxysporum [13]. This enzyme was designed for the degumming of plant oils, which is the first step in the rafination of vegetable oils. In this process, both hydratable

Comparison of Immobilized Preparations of Lecitase™ Ultra as the Catalysts of Transesterification
Since enzyme-catalyzed transesterification requires an anhydrous environment, it is not possible to use directly the commercially available aqueous Lecitase™ Ultra solution for this process. Thus, the first step of the study involved the immobilization of the enzyme on four different carriers, using methods of immobilization on the surface or inside the carrier. In the first case, LU was physically adsorbed on Supelite™ DAX-8 polyacrylic resin (LU-DAX), covalently bound to cyanogen bromide-activated agarose (LU-CNBr), covalently bound to bacterial cellulose that was produced by Komagataeibacter xylinus and modified with polyethyleneimine, saturated by superparamagnetic Fe 3 O 4 particles, and activated by glutaraldehyde (LU-MBC). In order to entrap the enzyme inside the carrier, LU was immobilized in calcium alginate to produce LU-ALG. The activity and enantioselectivity of immobilized preparations in the process of transesterification were compared using (E)-4-phenylbut-3-en-2-ol (1a) as a model substrate, obtained previously in a two-step synthesis consisting of the Claisen-Schmidt condensation of the benzaldehyde with acetone and reduction of the obtained α,β-unsaturated ketone [62]. The enzymatic reactions were carried out in diisopropyl ether (DIPE) at room temperature with vinyl propionate 2 as an acyl group donor.
All the preparations catalyzed an effective kinetic resolution of the studied substrate 1a (Table 1), and for three preparations, high enantioselectivity of the reaction was observed (E > 200) (Entries 1-3). Shorter reaction times (4-8 h), necessary to obtain both unreacted (S)-alcohol 1a and (R)-propionate 3a with high enantiomeric purity, were observed in the reactions catalyzed by the LU-DAX and LU-CNBr, compared to those catalyzed by LU-ALG and LU-MBC (48-72 h). The best kinetic resolution was achieved using LU-CNBr, and after 8 h, it resulted in the production of (S)-alcohol 1a and (R)-ester 3a with ee 94% and >99%, respectively. Extending the reaction time to 24 h allowed the increase of ee of unreacted (S)-alcohol 1a to 97%. Lower optical purity of (S)-alcohol 1a (ee 90-92%) and still high ee of (R)-propionate 3a (98%) were determined in the reactions catalyzed by LU-ALG and LU-MBC after 48 h and 72 h, respectively. The lowest enantioselectivity of the process was observed when LU-DAX was used. In the case of this preparation, the reaction should be stopped before exceeding 50% conversion of the substrate (after 2 h) to obtain (R)-propionate 3a with highest optical purity. When the reaction proceeds further, the optical purity of (S)-alcohol 1a increases but the enantiomeric purity of (R)-propionate 3a is reduced (Entry 4).
Immobilization strategy can influence enzyme catalytic properties, including their enantiospecifity. A different enantiopreference was observed for immobilized LU during hydrolysis of racemic esters of mandelic acid. For example, LU immobilized on Immobead-350, activated by ethylenediamineglutaraldehyde groups, preferred hydrolysis of the R isomer of methyl mandelate. On the contrary, the immobilization of LU on Immobead-350, activated with epoxy-or ethylenediamine-divinylsulfone groups, changed the preference for the S isomer [55]. In another study, LU adsorbed on octyl-agarose hydrolyzed preferentially the (R)-isomer of 2-O-butanoyl-2-phenylacetic acid, whereas LU-CNBr exhibited an inversion of the enantiopreference, yielding (S)-mandelic acid [29]. In the studies described herein, regardless of immobilization method, no changes of enantiopreference were noticed. Comparison of specific rotation signs of obtained isomers with literature data confirmed the preference of LU, predicted on the Kazlauskas rule, towards esterification of (R,E)-4-phenylbut-3-en-2-ol (1a), which results in the formation of (R)-propionate 3a and unreacted (S)-alcohol 1a.

Effect of Enzyme Dosage
Considering the highest optical purity of both isomers obtained after 24 h of the kinetic resolution of (E)-4-phenylbut-3-en-2-ol (1a) using LU-CNBr, this preparation was selected for further research to determine the impact of different factors such as enzyme dose, a type of organic solvent, and a type of an acyl donor on the process efficiency.
The first of these experiments was aimed at determining the minimum dose of the enzyme that is necessary to achieve the effective resolution of alcohol 1a (Figure 1).

Effect of Enzyme Dosage
Considering the highest optical purity of both isomers obtained after 24 h of the kinetic resolution of (E)-4-phenylbut-3-en-2-ol (1a) using LU-CNBr, this preparation was selected for further research to determine the impact of different factors such as enzyme dose, a type of organic solvent, and a type of an acyl donor on the process efficiency.
The first of these experiments was aimed at determining the minimum dose of the enzyme that is necessary to achieve the effective resolution of alcohol 1a (Figure 1).   Regardless of the number of enzyme units used for the reaction, the enantioselectivity of the reaction was very high (E > 200). The lowest effective dose of the LU-CNBr, which allowed the achievement of about 50% of substrate conversion and high enantiomeric excesses of both the obtained ester 3a (99%) and unreacted alcohol 1a (97%), was 0.03 U of preparation. Lower amounts of the biocatalyst resulted in a significantly lower degree of substrate conversion, hence a clearly reduced enantiomeric purity of unreacted alcohol 1a.

Effect of Solvent
The environment has a great impact on the catalytic properties of lipases [63,64]. These enzymes, which in natural conditions act at the water-organic phase interface, also show catalytic activity in organic solvents [65]. Polar organic solvents, e.g., methanol or DMSO, are far more enzyme-deactivating than more hydrophobic solvents. They disrupt the hydrophobic interactions responsible for the natively folded structure of proteins and strip off the water from protein molecules, which is crucial for maintaining the structure, stability, and activity of the enzymes. These factors lead to a decrease in enzyme stability that can result in denaturation [66]. Therefore, in the lipase-catalyzed reactions employed in organic synthesis, apolar solvents such as ethers, alkanes, cycloalkanes, and aromatic hydrocarbons are generally used [67].
To check the influence of the solvent on the transesterification of 4-phenylbut-3-en-2-ol (1a) with vinyl propionate and with LU-CNBr as the biocatalyst, the process was carried out for 24 h in six organic solvents commonly used in the lipase-catalyzed kinetic resolution of racemates ( Figure 2).
Catalysts 2020, 10, x FOR PEER REVIEW 5 of 17 Regardless of the number of enzyme units used for the reaction, the enantioselectivity of the reaction was very high (E > 200). The lowest effective dose of the LU-CNBr, which allowed the achievement of about 50% of substrate conversion and high enantiomeric excesses of both the obtained ester 3a (99%) and unreacted alcohol 1a (97%), was 0.03 U of preparation. Lower amounts of the biocatalyst resulted in a significantly lower degree of substrate conversion, hence a clearly reduced enantiomeric purity of unreacted alcohol 1a.

Effect of Solvent
The environment has a great impact on the catalytic properties of lipases [63,64]. These enzymes, which in natural conditions act at the water-organic phase interface, also show catalytic activity in organic solvents [65]. Polar organic solvents, e.g., methanol or DMSO, are far more enzymedeactivating than more hydrophobic solvents. They disrupt the hydrophobic interactions responsible for the natively folded structure of proteins and strip off the water from protein molecules, which is crucial for maintaining the structure, stability, and activity of the enzymes. These factors lead to a decrease in enzyme stability that can result in denaturation [66]. Therefore, in the lipase-catalyzed reactions employed in organic synthesis, apolar solvents such as ethers, alkanes, cycloalkanes, and aromatic hydrocarbons are generally used [67].
To check the influence of the solvent on the transesterification of 4-phenylbut-3-en-2-ol (1a) with vinyl propionate and with LU-CNBr as the biocatalyst, the process was carried out for 24 h in six organic solvents commonly used in the lipase-catalyzed kinetic resolution of racemates ( Figure 2). No significant differences were observed in the activity between the analyzed solvents. For all of them, about 50% conversion of the substrate and high enantioselectivity (E > 200) were obtained, resulting in very high ee of propionate 3a and unreacted alcohol 1a (95-100%). The lower enantiomeric excesses of both isomers were determined in toluene (ees = eep = 96%), compared with those obtained in cyclohexane (ees = eep = 98%). Optically pure alcohol 1a and propionate 3a with ee 95-96% were obtained in the reactions carried out in hexane and methyl t-butyl ether (MTBE). In turn, the use of tetrahydrofuran (THF) and DIPE resulted in slightly higher optical purity of the resulting propionate 3a (ee 98-99%) compared to the unreacted alcohol 1a (ee 96-97%). No significant differences were observed in the activity between the analyzed solvents. For all of them, about 50% conversion of the substrate and high enantioselectivity (E > 200) were obtained, resulting in very high ee of propionate 3a and unreacted alcohol 1a (95-100%). The lower enantiomeric excesses of both isomers were determined in toluene (ee s = ee p = 96%), compared with those obtained in cyclohexane (ee s = ee p = 98%). Optically pure alcohol 1a and propionate 3a with ee 95-96% were obtained in the reactions carried out in hexane and methyl t-butyl ether (MTBE). In turn, the use of tetrahydrofuran (THF) and DIPE resulted in slightly higher optical purity of the resulting propionate 3a (ee 98-99%) compared to the unreacted alcohol 1a (ee 96-97%).
Catalysts 2020, 10, 798 6 of 17 One of the main criteria for solvent selection in the enzymatic reactions is hydrophobicity, which is measured by the logarithm of the partition coefficient between octanol and water (log P). It is generally regarded that the activity and enantioselectivity of enzymes is low in solvents with a log P < 2, average in solvents with log P between 2 and 4, and high when log P exceeds 4 [68]. In the performed experiment, a panel of solvents with a wide range of polarity was tested, including hexane (log P = 3.5), cyclohexane (log P = 3.2), toluene (log P = 2.5), DIPE (log P = 1.52), MTBE (log P = 1.3), and THF (log P = 0. 49). No clear correlation between the hydrophobicity of the solvent and the catalytic activity and enantioselectivity of the reaction was found. Very high enantiomeric purity of the resolution products was achieved both in reactions carried out in aliphatic or aromatic hydrocarbon solvents (hexane, cyclohexane, toluene) and more polar ether solvents (MTBE, DIPE, THF). These results confirm that log P cannot be the only criterion for a solvent selection in enzymatic reactions and other important aspects should also be taken into account, including the solubility of the substrate and the possible negative effect of the solvent on the carrier.

Effect of Acyl Donor
The type of acyl group donor has a great influence on the course of transesterification processes [69]. To achieve a satisfactory alcohol conversion and high optical purity of the kinetic resolution products in a relatively short time, the reaction should be irreversible. For this reason, the most widely used acyl donors are enol esters, because the immediate tautomerization of the resulting enol to the carbonyl compound shifts the equilibrium of the esterification reaction towards product formation, making the process irreversible [70]. The common reagents from this group are vinyl esters [71][72][73][74][75], but the activity of many lipases is inhibited as a result of imines formation in the reaction between amino groups of the lysine residues and acetaldehyde, a product of tautomerization of vinyl alcohol formed during transesterification. The solution of this inconvenience is the use of isopropenyl esters because released propen-2-ol isomerizes to acetone, which does not affect the lipase activity [76,77].
To check the effect of an acyl donor on the enantioselectivity of the reaction catalyzed by the LU-CNBr, transesterification of (E)-4-phenyl-but-3-en-2-ol was carried out using three commonly used enol esters: two vinyl esters with different acyl chain length (vinyl acetate 4 and vinyl propionate 2) and isopropenyl acetate 5. They have been successfully used as donors of acyl groups in the kinetic resolution of racemic alcohols with the (E)-4-arylbut-3-en-2-ol system catalyzed by commercially available lipases [58,[78][79][80][81][82].
All tested reagents proved to be effective acylating agents ( Figure 3). The use of vinyl propionate (2) resulted in 50% substrate conversion, and unreacted alcohol 1a and propionate 3a were obtained with ee 97% and 99%, respectively. In the reactions with vinyl acetate (4) and isopropenyl acetate (5), substrate conversion exceeded 50%, allowing the obtainment of optically pure alcohol 1a. Higher enantiomeric excess of obtained ester 6 (96%) was observed using isopropenyl acetate, compared to that achieved with vinyl acetate (92%).
In the area of research on the kinetic resolution of alcohols with the (E)-4-arylbut-3-en-2-ol system, a significant effect of the type of enol ester on the activity and enantioselectivity of the process was previously reported for the transesterification of (E)-4-(benzo[d][1',3']dioxol-5'-yl)-but-3-en-2-ol, catalyzed by commercially available lipases [81]. The use of isopropenyl acetate reduced the reaction rate and the degree of substrate conversion, which, in the case of the reaction catalyzed by lipase B from Candida antarctica, resulted in an increase of ee for unreacted (S)-alcohol and the resulting (R)-ester, from 78% and 83% to >99% and 91%, respectively, compared with the reaction in which the acyl donor was vinyl propionate. In the studies described herein, no similar effect was found, and the kinetic resolution was not significantly influenced by acyl donors. These results proved a high versatility of LU-CNBr towards common enol esters. Irrespective of the acylating agent, both isomers obtained by kinetic resolution possessed high enantiomeric purity.

Enzyme Reuse
The possibility of recovery and multiple uses of the biocatalyst is one of the most important advantages of immobilized enzymes. Due to the very good results of the kinetic resolution of (E)-4phenylbut-3-en-2-ol (1a) by LU-CNBr, the operational stability and the possibility of reuse of this biocatalyst in this process were checked. After each reaction, the enzyme was separated by filtration, residual solvent was removed, and the biocatalyst was applied to the resolution of a new substrate portion. For each reaction cycle, the conversion of alcohol 1a, enantiomeric excesses of both kinetic resolution products, and the residual activity of LU-CNBr were determined. The results are shown in Figure 4.

Enzyme Reuse
The possibility of recovery and multiple uses of the biocatalyst is one of the most important advantages of immobilized enzymes. Due to the very good results of the kinetic resolution of (E)-4-phenylbut-3-en-2-ol (1a) by LU-CNBr, the operational stability and the possibility of reuse of this biocatalyst in this process were checked. After each reaction, the enzyme was separated by filtration, residual solvent was removed, and the biocatalyst was applied to the resolution of a new substrate portion. For each reaction cycle, the conversion of alcohol 1a, enantiomeric excesses of both kinetic resolution products, and the residual activity of LU-CNBr were determined. The results are shown in Figure 4.
After two subsequent reaction cycles, LU-CNBr retained 97% of its initial activity, which correlated with a very high enantiomeric purity of both isomers produced by kinetic resolution of alcohol 1a (96-99%). A rapid drop in residual activity (77% of its initial value), was observed after the third use of the enzyme, which in the next cycle, correlated with a slightly lower degree of substrate conversion (47%) and a lower enantiomeric excess of unreacted alcohol 1a (84%). After the fourth reaction cycle, the activity of the preparation was reduced to almost half of its initial value, and the very low degree of substrate conversion made the kinetic resolution of alcohol 1a ineffective.  After two subsequent reaction cycles, LU-CNBr retained 97% of its initial activity, which correlated with a very high enantiomeric purity of both isomers produced by kinetic resolution of alcohol 1a (96-99%). A rapid drop in residual activity (77% of its initial value), was observed after the third use of the enzyme, which in the next cycle, correlated with a slightly lower degree of substrate conversion (47%) and a lower enantiomeric excess of unreacted alcohol 1a (84%). After the fourth reaction cycle, the activity of the preparation was reduced to almost half of its initial value, and the very low degree of substrate conversion made the kinetic resolution of alcohol 1a ineffective.
The activity of the LU-CNBr, expressed in units of enzymatic activity (U), is a reflection of the hydrolytic activity of enzymes towards p-nitrophenyl palmitate, while the degree of conversion is a measure of the activity in the process of transesterification of (E)-4-phenylbut-3-en-2-ol. The results indicate that the enzyme retains high esterification activity longer than hydrolytic activity. The decrease of the activity may be the result of partial denaturation of the enzyme [38], but changes in the protein conformation have a greater influence on the substrate fitting to the enzyme active site in the case of p-nitrophenyl palmitate than of (E)-4-phenylbut-3-en-2-ol. To date, there is no data on the operational stability of the LU in the kinetic resolution of racemic mixtures. Such studies have been carried out so far in the case of LU immobilized in a copolymer of styrene and divinylbenzene, catalyzing the esterification of myristic and caprylic acid with ethanol [39]. The activity of the preparation in this reaction decreased in subsequent reaction cycles, resulting in a complete lack of activity after the sixth cycle. In turn, for LU immobilized on Amberlite XAD2 and XAD4 resins, and used in the esterification reaction of 1,2-O-isopropylidene glycerol, a rapid decrease in activity was observed after the first reaction cycle [49]. Our results indicate that LU-CNBr can be used three times as an effective biocatalyst in the process of kinetic resolution of (E)-4-phenylbut-3-en-2-ol (1a).

Resolution of (E)-4-Phenylbut-3-en-2-ol Analogues (1b-d)
In the last experiment, the immobilized LU-CNBr was used for the kinetic resolution of (E)-4phenylbut-3-en-2-ol analogues with various substituents on the aromatic ring, such as the methyl group in the para position (alcohol 1b), two methyl groups in the 2 and 5 positions (alcohol 1c), and a methoxy group in the para position (alcohol 1d) ( Table 2). Similarly to alcohol 1a, they were The activity of the LU-CNBr, expressed in units of enzymatic activity (U), is a reflection of the hydrolytic activity of enzymes towards p-nitrophenyl palmitate, while the degree of conversion is a measure of the activity in the process of transesterification of (E)-4-phenylbut-3-en-2-ol. The results indicate that the enzyme retains high esterification activity longer than hydrolytic activity. The decrease of the activity may be the result of partial denaturation of the enzyme [38], but changes in the protein conformation have a greater influence on the substrate fitting to the enzyme active site in the case of p-nitrophenyl palmitate than of (E)-4-phenylbut-3-en-2-ol. To date, there is no data on the operational stability of the LU in the kinetic resolution of racemic mixtures. Such studies have been carried out so far in the case of LU immobilized in a copolymer of styrene and divinylbenzene, catalyzing the esterification of myristic and caprylic acid with ethanol [39]. The activity of the preparation in this reaction decreased in subsequent reaction cycles, resulting in a complete lack of activity after the sixth cycle. In turn, for LU immobilized on Amberlite XAD2 and XAD4 resins, and used in the esterification reaction of 1,2-O-isopropylidene glycerol, a rapid decrease in activity was observed after the first reaction cycle [49]. Our results indicate that LU-CNBr can be used three times as an effective biocatalyst in the process of kinetic resolution of (E)-4-phenylbut-3-en-2-ol (1a).

Resolution of (E)-4-Phenylbut-3-en-2-ol Analogues (1b-d)
In the last experiment, the immobilized LU-CNBr was used for the kinetic resolution of (E)-4-phenylbut-3-en-2-ol analogues with various substituents on the aromatic ring, such as the methyl group in the para position (alcohol 1b), two methyl groups in the 2 and 5 positions (alcohol 1c), and a methoxy group in the para position (alcohol 1d) ( Table 2). Similarly to alcohol 1a, they were synthesized previously from corresponding aromatic aldehydes via Claisen-Schmidt condensation with acetone, followed by the reduction of a carbonyl group [81][82][83].
The presence of substituents on the aromatic ring affected the rate of transesterification of the tested alcohols and their kinetic resolution differently. The time necessary for reaching about 50% conversion was shorter for the derivative with the p-methylphenyl substituent 1b (6 h), while the effective transesterification of alcohol containing the methoxy group 1d required 24 h. Enantioselectivity of transesterification of alcohols with methyl substituents on the benzene ring (1b,1c) was as high as the enantioselectivity of transesterification observed for (E)-4-phenylbut-3-en-2-ol (1a) (E > 200, Table 1), and enantiomerically pure (R)-propionates 3b,c were obtained. Higher optical purity of unreacted alcohol (95% vs. 91%) was observed for the substrate with the 2,5-dimethylphenyl substituent (1c) than its p-methylsubstituted analog (1b). The presence of the methoxy group on the aromatic ring resulted in a lower enantioselectivity and lower optical purity of both isomers (ee s = ee p = 90%) (Entry 3). In all cases, the signs of specific rotations of obtained isomers (Section 3.8.4) confirmed the R configuration of propionates 3b-d and S configuration of unreacted alcohols 1b-d. Table 2. Transesterification of (E)-4-arylbut-3-en-2-ols (1b-d) with different substituents on a benzene ring catalyzed by LU-CNBr.
Catalysts 2020, 10, x FOR PEER REVIEW 9 of 17 synthesized previously from corresponding aromatic aldehydes via Claisen-Schmidt condensation with acetone, followed by the reduction of a carbonyl group [81][82][83]. The presence of substituents on the aromatic ring affected the rate of transesterification of the tested alcohols and their kinetic resolution differently. The time necessary for reaching about 50% conversion was shorter for the derivative with the p-methylphenyl substituent 1b (6 h), while the effective transesterification of alcohol containing the methoxy group 1d required 24 h. Enantioselectivity of transesterification of alcohols with methyl substituents on the benzene ring (1b,1c) was as high as the enantioselectivity of transesterification observed for (E)-4-phenylbut-3-en-2-ol (1a) (E > 200, Table 1), and enantiomerically pure (R)-propionates 3b,c were obtained. Higher optical purity of unreacted alcohol (95% vs. 91%) was observed for the substrate with the 2,5dimethylphenyl substituent (1c) than its p-methylsubstituted analog (1b). The presence of the methoxy group on the aromatic ring resulted in a lower enantioselectivity and lower optical purity of both isomers (ees = eep = 90%) (Entry 3). In all cases, the signs of specific rotations of obtained isomers (Section 3.8.4) confirmed the R configuration of propionates 3b-d and S configuration of unreacted alcohols 1b-d.
In our previous studies on the hydrolysis of (E)-4-arylbut-3-en-2-yl esters, a similar effect of the substituents on the benzene ring on the activity and enantioselectivity of the reaction catalyzed by LU-CNBr has been shown [34]. Similar to the hydrolysis reaction, the best resolution in the process of transesterification was observed for substrates with unsubstituted benzene ring and with a 2,5dimethylphenyl substituent; very good results were also obtained for compounds with a pmethylphenyl substituent. Noteworthy is the significantly shorter time (6-8 h) needed for effective separation of alcohols 1b and 1c by transesterification compared to the hydrolysis of their butyrates (24-48 h) [34]. In contrast to the non-enantioselective hydrolysis of butyrate containing a methoxy substituent on a benzene ring [34], transesterification of alcohol bearing this substituent (1d) was enantioselective and it was possible to obtain both products with relatively high enantiomeric excess.

1b
Catalysts 2020, 10, x FOR PEER REVIEW 9 of 17 synthesized previously from corresponding aromatic aldehydes via Claisen-Schmidt condensation with acetone, followed by the reduction of a carbonyl group [81][82][83]. The presence of substituents on the aromatic ring affected the rate of transesterification of the tested alcohols and their kinetic resolution differently. The time necessary for reaching about 50% conversion was shorter for the derivative with the p-methylphenyl substituent 1b (6 h), while the effective transesterification of alcohol containing the methoxy group 1d required 24 h. Enantioselectivity of transesterification of alcohols with methyl substituents on the benzene ring (1b,1c) was as high as the enantioselectivity of transesterification observed for (E)-4-phenylbut-3-en-2-ol (1a) (E > 200, Table 1), and enantiomerically pure (R)-propionates 3b,c were obtained. Higher optical purity of unreacted alcohol (95% vs. 91%) was observed for the substrate with the 2,5dimethylphenyl substituent (1c) than its p-methylsubstituted analog (1b). The presence of the methoxy group on the aromatic ring resulted in a lower enantioselectivity and lower optical purity of both isomers (ees = eep = 90%) (Entry 3). In all cases, the signs of specific rotations of obtained isomers (Section 3.8.4) confirmed the R configuration of propionates 3b-d and S configuration of unreacted alcohols 1b-d.
In our previous studies on the hydrolysis of (E)-4-arylbut-3-en-2-yl esters, a similar effect of the substituents on the benzene ring on the activity and enantioselectivity of the reaction catalyzed by LU-CNBr has been shown [34]. Similar to the hydrolysis reaction, the best resolution in the process of transesterification was observed for substrates with unsubstituted benzene ring and with a 2,5dimethylphenyl substituent; very good results were also obtained for compounds with a pmethylphenyl substituent. Noteworthy is the significantly shorter time (6-8 h) needed for effective separation of alcohols 1b and 1c by transesterification compared to the hydrolysis of their butyrates (24-48 h) [34]. In contrast to the non-enantioselective hydrolysis of butyrate containing a methoxy substituent on a benzene ring [34], transesterification of alcohol bearing this substituent (1d) was enantioselective and it was possible to obtain both products with relatively high enantiomeric excess.

1c
Catalysts 2020, 10, x FOR PEER REVIEW 9 of 17 synthesized previously from corresponding aromatic aldehydes via Claisen-Schmidt condensation with acetone, followed by the reduction of a carbonyl group [81][82][83]. The presence of substituents on the aromatic ring affected the rate of transesterification of the tested alcohols and their kinetic resolution differently. The time necessary for reaching about 50% conversion was shorter for the derivative with the p-methylphenyl substituent 1b (6 h), while the effective transesterification of alcohol containing the methoxy group 1d required 24 h. Enantioselectivity of transesterification of alcohols with methyl substituents on the benzene ring (1b,1c) was as high as the enantioselectivity of transesterification observed for (E)-4-phenylbut-3-en-2-ol (1a) (E > 200, Table 1), and enantiomerically pure (R)-propionates 3b,c were obtained. Higher optical purity of unreacted alcohol (95% vs. 91%) was observed for the substrate with the 2,5dimethylphenyl substituent (1c) than its p-methylsubstituted analog (1b). The presence of the methoxy group on the aromatic ring resulted in a lower enantioselectivity and lower optical purity of both isomers (ees = eep = 90%) (Entry 3). In all cases, the signs of specific rotations of obtained isomers (Section 3.8.4) confirmed the R configuration of propionates 3b-d and S configuration of unreacted alcohols 1b-d.
In our previous studies on the hydrolysis of (E)-4-arylbut-3-en-2-yl esters, a similar effect of the substituents on the benzene ring on the activity and enantioselectivity of the reaction catalyzed by LU-CNBr has been shown [34]. Similar to the hydrolysis reaction, the best resolution in the process of transesterification was observed for substrates with unsubstituted benzene ring and with a 2,5dimethylphenyl substituent; very good results were also obtained for compounds with a pmethylphenyl substituent. Noteworthy is the significantly shorter time (6-8 h) needed for effective separation of alcohols 1b and 1c by transesterification compared to the hydrolysis of their butyrates (24-48 h) [34]. In contrast to the non-enantioselective hydrolysis of butyrate containing a methoxy substituent on a benzene ring [34], transesterification of alcohol bearing this substituent (1d) was enantioselective and it was possible to obtain both products with relatively high enantiomeric excess.

1d
Catalysts 2020, 10, x FOR PEER REVIEW 9 of 17 synthesized previously from corresponding aromatic aldehydes via Claisen-Schmidt condensation with acetone, followed by the reduction of a carbonyl group [81][82][83]. The presence of substituents on the aromatic ring affected the rate of transesterification of the tested alcohols and their kinetic resolution differently. The time necessary for reaching about 50% conversion was shorter for the derivative with the p-methylphenyl substituent 1b (6 h), while the effective transesterification of alcohol containing the methoxy group 1d required 24 h. Enantioselectivity of transesterification of alcohols with methyl substituents on the benzene ring (1b,1c) was as high as the enantioselectivity of transesterification observed for (E)-4-phenylbut-3-en-2-ol (1a) (E > 200, Table 1), and enantiomerically pure (R)-propionates 3b,c were obtained. Higher optical purity of unreacted alcohol (95% vs. 91%) was observed for the substrate with the 2,5dimethylphenyl substituent (1c) than its p-methylsubstituted analog (1b). The presence of the methoxy group on the aromatic ring resulted in a lower enantioselectivity and lower optical purity of both isomers (ees = eep = 90%) (Entry 3). In all cases, the signs of specific rotations of obtained isomers (Section 3.8.4) confirmed the R configuration of propionates 3b-d and S configuration of unreacted alcohols 1b-d.
In our previous studies on the hydrolysis of (E)-4-arylbut-3-en-2-yl esters, a similar effect of the substituents on the benzene ring on the activity and enantioselectivity of the reaction catalyzed by LU-CNBr has been shown [34]. Similar to the hydrolysis reaction, the best resolution in the process of transesterification was observed for substrates with unsubstituted benzene ring and with a 2,5dimethylphenyl substituent; very good results were also obtained for compounds with a pmethylphenyl substituent. Noteworthy is the significantly shorter time (6-8 h) needed for effective separation of alcohols 1b and 1c by transesterification compared to the hydrolysis of their butyrates (24-48 h) [34]. In contrast to the non-enantioselective hydrolysis of butyrate containing a methoxy substituent on a benzene ring [34], transesterification of alcohol bearing this substituent (1d) was enantioselective and it was possible to obtain both products with relatively high enantiomeric excess.

Chemicals and Enzyme
Racemic alcohols 1a-d were synthesized from corresponding aromatic aldehydes as described 24 50 90 90 58 In our previous studies on the hydrolysis of (E)-4-arylbut-3-en-2-yl esters, a similar effect of the substituents on the benzene ring on the activity and enantioselectivity of the reaction catalyzed by LU-CNBr has been shown [34]. Similar to the hydrolysis reaction, the best resolution in the process of transesterification was observed for substrates with unsubstituted benzene ring and with a 2,5-dimethylphenyl substituent; very good results were also obtained for compounds with a p-methylphenyl substituent. Noteworthy is the significantly shorter time (6-8 h) needed for effective separation of alcohols 1b and 1c by transesterification compared to the hydrolysis of their butyrates (24-48 h) [34]. In contrast to the non-enantioselective hydrolysis of butyrate containing a methoxy substituent on a benzene ring [34], transesterification of alcohol bearing this substituent (1d) was enantioselective and it was possible to obtain both products with relatively high enantiomeric excess.

Enzymatic Activity Assays
Activity of the immobilized LU preparations was determined using p-NP assay [84]. The reaction mixture contained 75 µL of p-NPP isopropyl alcohol solution (1 mM), 5.75 mL of 10 mM Tris/HCl buffer (pH 8.0), and 30 mg of immobilized enzyme. After 30 min of incubation at 37 • C, 1 mL of cold ethanol was added to quench the reaction and the enzyme was allowed to settle. The clear supernatant was transferred to cuvettes of the spectrophotometer and absorbance of the samples was measured spectrophotometrically against a blank sample (without the enzyme). The amount of p-NP was calculated using a standard curve prepared for p-NP in the concentration range of 0-0.3 mM. Activity of immobilized enzymes was expressed in units per gram of immobilized enzyme (U/g), where one unit corresponds to an amount of enzyme that liberated 1 µmol of p-NP per minute under the conditions described.

Immobilization of LU in Calcium Alginate Beads
The mixture of Lecitase™ Ultra solution (2.2 mL) and 6% (w/v) solution of sodium alginate (10 mL) was stirred for 1 h and dropped through a syringe into 500 mL of 0.3 M CaCl 2 solution. After 1 h of hardening, the beads were filtered and washed with distilled water. After 24 h of freeze-drying using an LYO GT2 Freeze Dryer (SEIB Industrie, Goeteborg, Sweden), the immobilized beads of Lecitase™ Ultra (0.04 U/g) were collected and stored at 4 • C until use.

Adsorption of LU on Polyacrylic Resin (Supelite™ DAX-8)
Supelite™ DAX-8 was washed with the Tris-HCl buffer followed by distilled water (4 × 20 mL) and dried for 24 h at room temperature. Lecitase™ Ultra (4 mL) in 10 mL of the Tris-HCl buffer was shaken for 24 h with a 1 g of pretreated Supelite™ DAX-8 in a twisted 20 mL vial using an end-over-end shaker. The immobilized enzyme (0.06 U/g) was filtered, washed with the Tris-HCl buffer (2 × 10 mL), freeze-dried for 24 h, and stored at 4 • C until use.

Immobilization of LU on Cyanogen Bromide-Activated Crosslinked 4% Agarose
Commercial Lecitase™ Ultra (2 mL), 12 mL of the coupling buffer (0.1 M NaHCO 3 , pH 8.3) containing 0.5 M NaCl, and 3 g of wet cyanogen bromide were shaken on a rotary shaker for 1 h at 20 • C. The immobilized enzyme (0.06 U/g) was filtered on a Schott G3 funnel, washed with 50 mL of the coupling buffer, resuspended at 1 M ethanolamine (pH 8.0), and incubated for 2 h at 20 • C. Then, it was filtered and washed with an acetate buffer (pH 4.0, 40 mL) followed by the coupling buffer (50 mL), freeze-dried for 24 h, and stored at 4 • C until use.

Immobilization of LU on Modified Bacterial Cellulose
The details of the carrier preparation and immobilization procedure were described previously [56]. Briefly, magnetic bacterial cellulose spheres modified by polyethyleneimine were activated by 1% glutaraldehyde solution in 100 mM of a phosphate buffer (pH 7.0). In the next step, 5 mL of activated carrier were mixed with 10 mL of commercial preparation Lecitase™ Ultra and incubated for 24 h at 4 • C at a roller shaker. After incubation time, the immobilized enzyme was separated by using a magnetic separator, thoroughly washed with the phosphate buffer, and incubated with buffered NaBH 4 solution at 4 • C for 1 h. Next, the preparation was collected by using the magnetic separator and flushed with the phosphate buffer containing 100 mM NaCl and 0.25% Triton-X100. Finally, the preparation (0.045 U/g) was washed again by the phosphate buffer, freeze-dried for 24 h and stored at 4 • C until use.

General Procedure
Racemic alcohol 1a-d (0.2 g) and 2 mL of acyl donor (2,4 or 5) were dissolved in 6 mL of organic solvent followed by the addition of immobilized Lecitase™ Ultra (0.06 U). Reactions were carried out in 10 mL screw-cap glass vials on a magnetic stirrer (750 rpm) at room temperature. Samples (0.6 mL) were taken from the reaction mixture at the specified intervals and filtered through Celite 560 using diethyl ether to wash the adsorbent bed. After removal of excess solvent using a rotary vacuum evaporator, the concentrated samples were analyzed on a GC chiral column. Prior to GC analysis, the samples were treated with acetyl or propionyl chloride to derivatize inseparable enantiomers of alcohols 1a-d into corresponding esters according to the procedure described elsewhere [82]. Following this general protocol, the effect of the solvent and acyl donor on the transesterification of rac-(E)-4-phenylbut-3-en-2-ol (1a) using LU-CNBr was investigated.

Effect of Enzyme Dosage
Transesterification of racemic alcohol 1a (0.1 g) in DIPE (3 mL) was carried out for 24 h as described in Section 3.8.1 using different doses of LU-CNBr (0.01 U, 0.02 U and 0.03 U) and 2 mL of vinyl propionate.

Enzyme Reusability
Racemic (E)-4-phenylbut-3-en-2-ol (1a) was subjected to the transesterification using LU-CNBr according to the procedure described in Section 3.8.1. After the reaction, the enzyme preparation was recovered by filtration and dried at room temperature to remove the residual solvents. The residual activity of the recovered enzyme was measured as described in Section 3.3 and the enzyme was used as a biocatalyst in the next reaction cycle, with a fresh portion of substrate 1a. Totally, the enzyme was used in five reaction cycles.

Isolation of Products Obtained by Transesterification of Racemic Alcohols (1a-d)-General Procedure
Transesterification of racemic alcohols 1a-d (0.2 g) in DIPE (6 mL) was carried out for 24 h as described in Section 3.8.1, using 0.06 U of LU-CNBr and 2 mL of vinyl propionate. Reaction mixtures were filtered through Celite 560 and the organic solvent was evaporated under vacuo. Products were separated by preparative TLC (hexane:acetone, 10:1). They were known from previous literature reports and their spectroscopic data ( 1 H NMR, 13 C NMR, IR) were consistent with those reported earlier [81,82,85]. For confirmation of the structures, 1 H NMR spectra of the products were included in Supplementary Materials (Figures S1-S8).

Conclusions
The work describes for the first time the use of immobilized Lecitase™ Ultra preparations as biocatalysts for the kinetic resolution of racemic (E)-4-phenylbut-3-en-2-ol and its structural analogues by transesterification. The best results were obtained in the reaction catalyzed by the preparation immobilized on cyanogen bromide-activated agarose (LU-CNBr), which resulted in both resolution products with high yields and very high enantiomeric excesses (≥96%). This preparation turned out to be an effective and versatile biocatalyst, showing high activity and enantioselectivity in a wide range of organic solvents in the presence of commonly used acyl donors. An additional advantage of this biocatalyst is an easy recovery and high operational stability, allowing for its reuse in three subsequent reaction cycles without negative effects on the efficiency and enantioselectivity of transesterification. Due to the low cost of the preparation and simple immobilization, it becomes a promising alternative to the commercially available lipases commonly used for the resolution of racemic alcohols.
Future research will focus on the determination of the substrate specificity of LU towards a variety of alcohols, as well as the use of this enzyme in unconventional reaction media (ionic liquids, eutectic solvents, supercritical liquids, etc.).