Chemoenzymatic Synthesis of Synthons as Precursors for Enantiopure Clenbuterol and Other β 2 -Agonists

: Clenbuterol is a β 2 -agonist used in the veterinary treatment of asthma in several countries. The drug is listed on the World Antidoping Agency’s prohibited list due to its effect on increased protein synthesis in the body. However, racemic clenbuterol has recently been shown to reduce the risk of Parkinson’s disease. In order to reveal which one (or both) of the enantiomers that cause this effect, pure enantiomers need to be separately studied. ( R )-1-(4-Amino-3,5-dichlorophenyl)-2-bromoethan-1-ol has been synthesised in 93% enantiomeric excess ( ee ) by asymmetric reduction of the corresponding ketone catalysed by a ketoreductase and nicotinamide adenine dinucleotide phosphate (NADPH) as the cofactor in dimethyl sulfoxide (DMSO). ( S )- N -(2,6-Dichloro-4-(1-hydroxyethyl)phenyl)acetamide has been synthesised in >98% ee by the same system. Both synthons are potential precursors for clenbuterol enantiomers.


Introduction
The need for enantiopure compounds for the treatment of human diseases and combatting microbial attacks on, for example, crops can be explained by the fact that all living organisms in nature are chiral. Protein biosynthesis and most metabolic processes are mediated by enzymes that are specific for a particular isomeric form of a certain substrate, and this is essential for ensuring the high degree of three-dimensional organization that is found in structures within cells. It is presumably evolutionary chance that has determined that life is based on L-rather than D-amino acids [1].
When drug-receptor interactions are considered, it is postulated that the lower the effective dose of a drug, the greater the difference in the pharmacological effect of the optical isomers [2]. The ratio of the more active enantiomer (eutomer) compared to the less active enantiomer (distomer) is defined as the eudismic ratio; the higher the eudismic ratio is, the higher the effectiveness of the drug [3].

Results
In order to obtain a precursor for enantiopure (R)-clenbuterol, bromo alcohol 1-(4-amino-3,5dichlorophenyl)-2-bromoethan-1-ol (3) was synthesised in a 45% yield (Scheme 1) for the purpose of performing lipase-catalyzed kinetic resolution with lipases suitable for industrial-scale production. Bromination of substrate ketone 1 produced bromoketone 2 in an 86% yield. Various methods for the α-bromination of ketone 1 were performed: Addition of N-bromosuccinimide (NBS) (1.0 eq.) to 1 with para-toluenesulfonic acid (0.1 eq.) in methanol at 30 °C produced brominated ketone 2 in an 11% Clenbuterol (Ventipulmin ® , Boehringer Ingelheim Vetmedica, Inc., Duluth, GA, USA) is a long-acting, selective β 2 -agonist with anabolic properties used in the treatment of shock and airway-obstructing diseases in veterinary medicine [7,8]. It has not been approved by the U.S. Food and Drug Administration (FDA) for human or food-producing animals use, although it is illegally added to livestock feed to promote the growth of lean meat [9]. Clenbuterol can reduce body fat, and it is extensively used by athletes and bodybuilders even though it is listed on the World Antidoping Agency's prohibited list [10]. In addition to the desired effects of smooth-muscle relaxation, clenbuterol can cause several side effects, such as heart palpitations, muscle tremors, and nervousness [8].
Previously, it was reported that (S)-clenbuterol had neuroprotective properties, reduced blood pressure, and enhanced blood-glucose levels in rats. While (R)-clenbuterol did not promote these properties, it was found that it caused decreased motor activity, head twitches, and tremors [11,12]. Upon examination of human urine of participants who ingested racemic clenbuterol, it was found that the S-enantiomer was retained for longer in the body [13]. Recently, Mittal et al. found that the brains of patients with Parkinson´s disease contained accumulations of α-synuclein protein, the so-called Lewy bodies. β 2 -agonists metaproterenol, clenbuterol, and salbutamol lowered the levels of α-synuclein gene mRNA and α-synuclein protein levels in human SK-N-MC neuroblastoma cells [14]. Only racemic compounds have been studied, but pure enantiomers of the drugs should be independently studied in order to reveal if enantiomers would show different effects.
Our attempts to synthesize enantiopure precursors for (R)-clenbuterol include both lipase catalysed kinetic resolutions of racemates and asymmetrizations of substituted acetophenones catalysed by ketoreductases. Previously, we utilised chemoenzymatic methods in syntheses of enantiopure building blocks for compounds that are related to clenbuterol [18,19].

Kinetic Resolutions
The kinetic resolutions of 3 were performed in dichloromethane with vinyl butanoate as the acyl donor catalysed by commercial immobilised lipases A and B from Candida antarctica (CALA and CALB, Sigma-Aldrich, Oslo, Norway). The reactions showed no conversion after five days. We also attempted to resolve racemic clenbuterol with CALA and CALB from Sigma-Aldrich and vinyl butanoate as the acyl donor in tetrahydrofuran, Scheme 2. However, these reactions resulted in several acylation products after five days and the enzyme showed low enantioselectivity. Scheme 2. Kinetic resolution of clenbuterol catalysed by commercial immobilised lipases A and B from Candida antarctica (CALA and CALB) in dry tetrahydrofuran (THF) with vinyl butanoate (VB) as acyl donor, showed several acylation products. The E-value for the desired acylation was E = 1.3 after five days.
The electropherogram from the CALA-catalysed kinetic resolution of clenbuterol after 78 hours ( Figure 2) shows the peaks of (R)-clenbuterol at retention times (tR) tR 9.70 min and (S)-clenbuterol at tR 11.50 min, and the butanoic esters of (S)-clenbuterol at tR 12.52 min and ester of (R)-clenbuterol at tR 12.98 min. Elution orders are consistent with previously reported retention times [20]. The remaining peaks have not been identified, but are most likely the amide products from the acylation of the aromatic or aliphatic amines, which would produce two enantiomers each.

Kinetic Resolutions
The kinetic resolutions of 3 were performed in dichloromethane with vinyl butanoate as the acyl donor catalysed by commercial immobilised lipases A and B from Candida antarctica (CALA and CALB, Sigma-Aldrich, Oslo, Norway). The reactions showed no conversion after five days. We also attempted to resolve racemic clenbuterol with CALA and CALB from Sigma-Aldrich and vinyl butanoate as the acyl donor in tetrahydrofuran, Scheme 2. However, these reactions resulted in several acylation products after five days and the enzyme showed low enantioselectivity. yield. Bromination of 1 with bromine (1.1 eq.) in acetic acid at 60 °C gave ketone 2 in a 27% yield, while reacting substrate 1 with NBS (1.3 eq.) catalysed by silica gel (10% w/w) in refluxing methanol produced 2 in a 30% yield. Finally, we were able to produce α-bromo ketone 2 in an 86% yield by the slow addition of Br2 (1.1 eq.) to 1 in dichloromethane and methanol (2:1 v/v) at room temperature. Subsequent reduction of 2 with sodium borohydride gave 3 in a 45% yield.

Kinetic Resolutions
The kinetic resolutions of 3 were performed in dichloromethane with vinyl butanoate as the acyl donor catalysed by commercial immobilised lipases A and B from Candida antarctica (CALA and CALB, Sigma-Aldrich, Oslo, Norway). The reactions showed no conversion after five days. We also attempted to resolve racemic clenbuterol with CALA and CALB from Sigma-Aldrich and vinyl butanoate as the acyl donor in tetrahydrofuran, Scheme 2. However, these reactions resulted in several acylation products after five days and the enzyme showed low enantioselectivity. Scheme 2. Kinetic resolution of clenbuterol catalysed by commercial immobilised lipases A and B from Candida antarctica (CALA and CALB) in dry tetrahydrofuran (THF) with vinyl butanoate (VB) as acyl donor, showed several acylation products. The E-value for the desired acylation was E = 1.3 after five days.
The electropherogram from the CALA-catalysed kinetic resolution of clenbuterol after 78 hours ( Figure 2) shows the peaks of (R)-clenbuterol at retention times (tR) tR 9.70 min and (S)-clenbuterol at tR 11.50 min, and the butanoic esters of (S)-clenbuterol at tR 12.52 min and ester of (R)-clenbuterol at tR 12.98 min. Elution orders are consistent with previously reported retention times [20]. The remaining peaks have not been identified, but are most likely the amide products from the acylation of the aromatic or aliphatic amines, which would produce two enantiomers each.
+enantiomer +enantiomer +enantiomer Scheme 2. Kinetic resolution of clenbuterol catalysed by commercial immobilised lipases A and B from Candida antarctica (CALA and CALB) in dry tetrahydrofuran (THF) with vinyl butanoate (VB) as acyl donor, showed several acylation products. The E-value for the desired acylation was E = 1.3 after five days.
The electropherogram from the CALA-catalysed kinetic resolution of clenbuterol after 78 h ( Figure 2) shows the peaks of (R)-clenbuterol at retention times (t R ) 9.70 min and (S)-clenbuterol at t R 11.50 min, and the butanoic esters of (S)-clenbuterol at t R 12.52 min and ester of (R)-clenbuterol at t R 12.98 min. Elution orders are consistent with previously reported retention times [20]. The remaining peaks have not been identified, but are most likely the amide products from the acylation of the aromatic or aliphatic amines, which would produce two enantiomers each.
We anticipate that the α-bromo substituent in 3 could be one reason for the low selectivity and low reaction rate in the transesterification. In order to reveal the influence of an electron-rich substituent next to the stereocenter compared to bromine, transesterification of the propargyl alcohol 8 was performed in dry hexane with vinyl butanoate as the acyl donor and CALB as the catalyst (Scheme 3). The catalyst showed excellent enantioselectivity towards 8 after 2.5 h with an E-value >200 resulting in the conversion of (R)-8 to (R)-9 with no sign of (R)-8 on the chromatogram, giving (S)-8 in 99% ee. The reaction progress is shown in Figure 3. From this test reaction, it seems that either the chlorosubstituents in meta-position or the amino substituent in para-position on the benzene ring on 3 must be the reason for the low selectivity and slow esterifications of 3 and clenbuterol with CALB. We anticipate that the α-bromo substituent in 3 could be one reason for the low selectivity and low reaction rate in the transesterification. In order to reveal the influence of an electron-rich substituent next to the stereocenter compared to bromine, transesterification of the propargyl alcohol 8 was performed in dry hexane with vinyl butanoate as the acyl donor and CALB as the catalyst (Scheme 3). The catalyst showed excellent enantioselectivity towards 8 after 2.5 hours with an E-value >200 resulting in the conversion of (R)-8 to (R)-9 with no sign of (R)-8 on the chromatogram, giving (S)-8 in 99% ee. The reaction progress is shown in Figure 3. From this test reaction, it seems that either the chlorosubstituents in meta-position or the amino substituent in para-position on the benzene ring on 3 must be the reason for the low selectivity and slow esterifications of 3 and clenbuterol with CALB.     We anticipate that the α-bromo substituent in 3 could be one reason for the low selectivity and low reaction rate in the transesterification. In order to reveal the influence of an electron-rich substituent next to the stereocenter compared to bromine, transesterification of the propargyl alcohol 8 was performed in dry hexane with vinyl butanoate as the acyl donor and CALB as the catalyst (Scheme 3). The catalyst showed excellent enantioselectivity towards 8 after 2.5 hours with an E-value >200 resulting in the conversion of (R)-8 to (R)-9 with no sign of (R)-8 on the chromatogram, giving (S)-8 in 99% ee. The reaction progress is shown in Figure 3. From this test reaction, it seems that either the chlorosubstituents in meta-position or the amino substituent in para-position on the benzene ring on 3 must be the reason for the low selectivity and slow esterifications of 3 and clenbuterol with CALB.   We anticipate that the α-bromo substituent in 3 could be one reason for the low selectivity and low reaction rate in the transesterification. In order to reveal the influence of an electron-rich substituent next to the stereocenter compared to bromine, transesterification of the propargyl alcohol 8 was performed in dry hexane with vinyl butanoate as the acyl donor and CALB as the catalyst (Scheme 3). The catalyst showed excellent enantioselectivity towards 8 after 2.5 hours with an E-value >200 resulting in the conversion of (R)-8 to (R)-9 with no sign of (R)-8 on the chromatogram, giving (S)-8 in 99% ee. The reaction progress is shown in Figure 3. From this test reaction, it seems that either the chlorosubstituents in meta-position or the amino substituent in para-position on the benzene ring on 3 must be the reason for the low selectivity and slow esterifications of 3 and clenbuterol with CALB.

Asymmetrisations
Asymmetrisations of ketones with suitable enzymes is another way of obtaining enantiopure compounds. Depending on the catalysts stereoselectivity towards a chosen substrate, a theoretical yield of 100% might be obtained. Scheme 4 shows the asymmetrisation of ketones 1, 2, and 4-7 with ketoreductase KRED 228 (Syncozymes Co., Ltd., Shanghai, China). Several regeneration systems were tested in order to find a suitable system to regenerate nicotinamide adenine dinucleotide phosphate (NADP + ), and we were successful with glucose-6-phosphate dehydrogenase with glucose-6-phosphate as the cosubstrate [21,22]. The cosolvent was DMSO in all asymmetrisation reactions.
The enzyme-catalysed kinetic resolutions were performed in a New Brunswick

Achiral Chromatographic Analyses
Achiral GLC analyses were performed on an Agilent 7890A gas chromatograph, with an Agilent

Chiral GLC-Analyses
The chiral GLC analyses of 3 and 6a were performed on an Agilent 7890B gas chromatograph

Determination of Enantiomeric Excess (ee), Conversion (c), and Enantiomeric Ratio (E)
Chiral GLC and HPLC analyses gave ee s -and ee p -values from which the degree of conversion was calculated according to c = ee s /(ee s + ee p ). Enantiomeric ratios, E, were calculated based on ping-pong bi-bi kinetics using the computer program E and K Calculator 2.1b0 PPC (Department of chemistry, NTNU, Trondheim, Norway) [23]. Two or more replicates of the transesterification and asymmetrisation reactions were performed.

Synthesis of N-(4-Acetyl-phenyl)acetamide (5)
To a stirred solution of 1-(4-aminophenyl)ethan-1-one (0.53 g, 3.92 mmol) in CH 2 Cl 2 (15 mL), a solution of AcCl (0.56 mL, 7.84 mmol) in CH 2 Cl 2 (2 mL) was added dropwise at RT. After 5 min, a white precipitate started to form, and a solution of Et 3 N (1.09 mL, 7.84 mmol) in CH 2 Cl 2 (2 mL) was added dropwise, after which the solution cleared and turned a dark-yellow color. Full conversion was observed after 24 h by TLC: R f (5) = 0.05 (1:4 EtOAc/n-pentane), and the solution was extracted with brine (3 × 15 mL). The aqueous layers were combined and extracted with EtOAc (2 × 15 mL), before the combined organic layers were dried over MgSO 4 , filtered, and the solvents were removed under reduced pressure. The crude product was recrystallised from EtOAc to afford 5 in 53% yield (0.37 g, 2.09 mmol) and 98% purity (GLC). 1  To a stirred solution of 1-(4-amino-3,5-dichlorophenyl)ethan-1-one (1) (6.00 g, 29.40 mmol) in CH 2 Cl 2 (150 mL) AcCl (10.49 mL, 147.01 mmol) in CH 2 Cl 2 (15 mL) was added. To the stirred solution Et 3 N (4.29 mL, 30.78 mmol) in CH 2 Cl 2 (15 mL) was added dropwise. The reaction was monitored by TLC. R f (6) = 0.64 (1:4 EtOAc/n-pentane). After 48 h, full conversion was observed. The reaction mixture was washed with satd. K 2 CO 3 (2 × 100 mL) and brine (2 × 100 mL), before the organic layers were collected and dried over anhydrous MgSO 4 , filtered, and the solvent removed under reduced pressure. The crude product was recrystallized from EtOAc, and the purified compound dried in vacuo overnight to afford 6 as white crystals in 58% yield (4.20 g, 17.07 mmol) in 99% purity. 1  19 mmol) in THF (2.5 mL) was added, and the flask washed out with THF (0.5 mL), which was then added to the reaction mixture. The cooling bath was removed, and the reaction mixture was stirred at room temperature (RT) for 5 min. The reaction was quenched by addition of satd. aq. NH 4 Cl (10 mL), and the mixture was extracted with Et 2 O (2 × 20 mL). The combined organic extracts were dried with Na 2 SO 4 , filtered, and concentrated in  Aliquots of 100 L were collected every 30 min for the first 5 h, then every hour for 3 h, and every 24 h for 5 days. The solvent in the aliquots was removed with N 2 before 0.100 M pH 2.5 phosphate buffer (0.5 mL) was added, and the samples were analysed by chiral CE, from which the enantiomeric excess of (R)-clenbuterol was calculated to 11% ee.

General Procedure
The ketones 1, 2 and 4-7 (2.5-60.8 mg, 0.01-0.16 mmol) were dissolved in DMSO (0.10 mL) and transferred to a solution of ketoreductase 228 (KRED 228), NADPH, glucose-6-phosphate (G6P) and glucose-6-phosphate dehydrogenase (G6P-DH) in pH 7.0 phosphate buffer (0.9 mL). The reactions were incubated at 30 • C and 300 rpm for 1-28 days, and monitored by TLC (1:2 n-pentane/EtOAc) every 24 h. After full conversion of the starting material, the mixtures were extracted with EtOAc (4 × 10 mL) and the organic phases were combined, washed with water (2 × 20 mL), dried over MgSO 4 , and filtered before the organic solvent was removed under reduced pressure. The crude products were either purified by flash chromatography (silica, EtOAc/n-pentane) to afford the pure alcohols, or the crude product was analysed by chiral GLC or chiral HPLC.

Conclusions
Asymmetrisation of ketones 2 and 6 with KRED 228 were the only asymmetrisations in the series of ketones 1, 2, and 4-7 that produced the respective alcohol with high enantiomeric excess, giving (R)-2a in a 93% ee (low yield) and (S)-6a in a 98% ee and 31% yield. (R)-2a may be reacted with t-butylamine, yielding (R)-clenbuterol in high enantiomeric excess. (S)-6a may be a precursor for important enantiopure compounds.
The absolute configuration of the alcohols was determined by comparing elution orders on GLC with similar compounds. KRED 228 from Syncozymes Co., Ltd. was found to be a suitable enzyme for the reduction of substitued arylketones with several substituents on the benzene ring, however when an α-halogen is present the enzyme is not fully efficient. In addition, protection of the amino group on the benzene ring seems to be cruicial.
Kinetic resolutions of the racemic bromo alcohol 3 with CALA and CALB were not successful. The transesterification reactions showed low reactivity and low stereoselectivity with both lipases. Due to the excellent selectivity of the CALB catalysed kinetic resolution of propargyl alcohol 8 with vinyl butanoate in hexane it was anticipated that both the amino and the chloro substituents in meta-positions of 3 cause the problem. Both steric hindrance and electronic effects of these substituents leading to an irreversible substrate-enzyme complex might be the reason for the low conversion of the two lipases.